You are on page 1of 448


rni +ons
Professor of Classics, The University of Vermont
caxniioci uxiviisir\ iiiss
Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, São Paulo
Cambridge University Press
The Edinburgh Building, Cambridge cn: :iu, UK
First published in print format
isnx-:, ,;ï-c-,::-ï,:,:-,
isnx-:, ,;ï-c-,::-:,,,o-,
© Cambridge University Press 2004
Information on this title:
This publication is in copyright. Subject to statutory exception and to the provision of
relevant collective licensing agreements, no reproduction of any part may take place
without the written permission of Cambridge University Press.
isnx-:c c-,::-:,,,o-o
isnx-:c c-,::-ï,:,:-,
Cambridge University Press has no responsibility for the persistence or accuracy of uiis
for external or third-party internet websites referred to in this publication, and does not
guarantee that any content on such websites is, or will remain, accurate or appropriate.
Published in the United States of America by Cambridge University Press, New York
eBook (NetLibrary)
eBook (NetLibrary)
List of tables page ix
List of maps x
Preface xi
i x+nontc+i ox :
I Sex. Julius Frontinus :
II The De Aquaeductu ¸
Its date ¸
Its content and form 8
Its audience and purpose :.
The curator aquarum and the emperor :¡
The sources .o
III Language and style .:
Lexicon of water quality .¸
Formulaic presentation .¡
Rhetorical style .¸
IV The textual tradition ¸o
The Middle Ages ¸o
Poggio’s quest ¸:
The Codex Hersfeldensis ¸¡
The Codex Casinensis and Peter the Deacon
of Monte Cassino ¸¸
The manuscript tradition prior to C ¡¡
The recentiores ¡6
V Editions and commentaries ¸.
VI Editorial conventions and the apparatus criticus ¸8
+rx+ \xn cni +i c\r \rr\n\+ts 6¸
s i or\ 6¡
nr \¸\\rn\c+\ \nni s nox\r 6¸
coxxrx+\nv ::q
\rrrxni crs ¸¸¸
A Poggio’s use of the De Aquaeductu ¸¸¸
B Inscriptions pertinent to Frontinus’ text ¸¸q
C The impossibility of reaching an exact value for
the Roman quinaria measure, by Christer Bruun,
University of Toronto ¸¡.
References ¸6o
: Selected editions of De Aquaeductu ¸6o
. Translations ¸6:
¸ Abbreviations ¸6.
¡ Other works ¸6¡
Literary and epigraphical citations ¡o¡
Index ¡:¸
: Lengths of the aqueducts (Chapters ¸–:¸) page ¸¸o
. Fractions ¸¸:
¸ Small adjutages relative to the quinaria
(Chapter .6.¸–¸) ¸¸:
¡ Pipe-sizes (Chapters ¸q–6¸) ¸¸.
¸ Quinariae assigned to the various aqueducts
(Chapters 6¸–¸¸) ¸¸¡
6 Categories of distribution (Chapter ¸8) ¸¸¸
¸ Castella and distributions (Chapters ¸8–86) ¸¸¸
8 Distribution by aqueduct (extra urbem) (Chapters
¸8–86) ¸¸6
q Distribution by aqueduct (intra urbem) (Chapters
¸8–86) ¸¸¸
:o Distribution by regiones (Chapters ¸q–86) ¸¸8
:: Curatores aquarum (Chapter :o.) ¸¸q
: Extra-urban routes of the ancient aqueducts
based on Peter Aicher’s Guide to the Aqueducts of Ancient
Rome (:qq¸), with permission of Bolchazy-Carducci
Publishers, Inc. page ¸¡¸
. Routes of the aqueducts within Rome
based on Harry Evans’ Water Distribution in Ancient
Rome: The Evidence of Frontinus (:qq¡), with permission
of the University of Michigan Press ¸¡8
¸ Settling-tanks near the seventh milestone ¸¡q
Par tibi, Roma, nihil, cum sis prope tota ruina;
quam magni fueris integra, fracta doces. . . .
non tamen annorum series, non flamma, nec ensis
ad plenum potuit hoc abolere decus.
Hildebert of Lavardin, c. ::oo cr
Metropolitan Rome, the domina orbis, can to this day point with
especial pride to one of the gems in her imperial crown: a copi-
ous, ever-flowing supply of public water. And beginning at least
with Strabo, visitors to the Eternal City have not failed to admire
the architectural grandeur of the aqueducts. ‘Der sch¨ one große
Zweck, ein Volk zu tr¨ anken durch eine so ungeheure Anstalt!’
wrote Goethe in November :¸86. ‘Diese Menschen arbeiteten
f ¨ ur die Ewigkeit, es war auf alles kalkuliert, nur auf den Unsinn
der Verw¨ uster nicht, dem alles weichen mußte.’
In the year q¸ cr Julius Frontinus was appointed by the em-
peror Nerva to the post of curator aquarum for the City of Rome.
Frontinus exemplifies the ideal of a high-ranking senator who
works closely with his prince in service to the commonwealth.
He sees the aqueducts under his charge as monuments of Roman
greatness, for their practical value more wonderful even than
the fabled pyramids. In the present booklet, De Aquaeductu Urbis
Romae, Frontinus sets forth his duties, responsibilities and accom-
plishments during approximately one year in office as curator.
By the time he is writing, Nerva has died and Rome awaits the
arrival of the newemperor Trajan, in whose accession Frontinus
himself seems to have played no small role.
Our author sketches the history of Rome’s aqueducts, fur-
nishes a wealth of technical data on supply and delivery, quotes
verbatimfromlegal documents and touches on a variety of other
topics incidental to his administrator’s viewpoint. Yet he is not
composing a treatise on the engineering of aqueducts, he barely
concerns himself with fiscal aspects of management, nor does
he compile what might comprise a comprehensive administra-
tive manual of use to a successor. Scholars who are grateful for
such information as he provides are nonetheless prone to consult
this text rather than to read it. Frontinus, in consequence, has
been alternatively under-rated and over-rated both as a tech-
nical writer and as an administrator. In plain truth we do not
surely understand what purpose he might have intended for the
De Aquaeductu and the work remains something of an enigma.
Nothing quite like it is known, let alone survives, from the an-
cient world.
This edition of the De Aquaeductu is the first in eighty years
to be based on the single authoritative witness, that sadly blem-
ished exemplar which Poggio discovered at Monte Cassino in
:¡.q. ‘Authors surviving in a solitary MS. are by far the easiest
to edit,’ wrote Housman. ‘They are the easiest, and for a fool
they are the safest.’ But since Fritz Krohn in :q.., no editor has
chosen the easy pathway of reliance on this unique manuscript,
for all have been misled in vain attempts to retrieve an indepen-
dent tradition amongst its fifteenth-century progeny. From the
starting-point of the Codex Casinensis there is progress still to
be made, I believe, especially by taking into account the idiosyn-
crasies of its twelfth-century scribe, Peter the Deacon of Monte
Cassino, a man notorious for literary affectation but nonethe-
less an intriguing figure in the long process by which classical
antiquity was rediscovered and appreciated.
No full commentary on the De Aquaeductu has been written
since Giovanni Poleni’s masterpiece of :¸.., and the task is a
daunting one – not least because his credentials were those of a
hydraulic engineer and professor of mathematics. In the words
of the late Pierre Grimal, ‘Plus que nul autre texte, le trait´ e de
Frontin impose ` a l’´ editeur une compr´ ehension minutieuse de
chaque mot, chaque phrase, et oblige de d´ epasser la critique
verbale pure et simple pour saisir les realia.’ Indeed, the realia
of which he speaks are themselves richly varied. They encom-
pass not only the stuff of history, archaeology and technology
but extend to such matters as the exacting details of Roman law
and the intricacies of fractions in Roman arithmetical compu-
tation. Under such circumstances, a commentator may perhaps
be forgiven superficiality of a sort on the one hand and a certain
speculative latitude on the other. I have done what I could to
give appropriate attention to content and interpretation as well
as to text and language.
My engagement with the text of Frontinus began a quarter
century ago in conjunction with a seminar in Latin epigraphy
at the University of California, Berkeley, in :q¸8, and the initial
stages of my work were supported by grants from that univer-
sity’s Committee on Research and from the American Philo-
sophical Society. I profited enormously from the resources of
the Harvard College Library during a term as Visiting Lecturer
in :q8o, and in :q86 I enjoyed the congenial hospitality of the
Department of Classical Studies at the University of Michigan.
The edition and commentary took on a preliminary form dur-
ing the year :q86–¸, the period for which I was honoured to be
a John Simon Guggenheim Memorial Fellow. The Foundation’s
generosity made possible a trip to Italy in May :q8¸, with the op-
portunity for study in the Vatican Library and to re-examine the
codex unicus in the abbey library at Monte Cassino. As a Fellowin
Byzantine Studies at Dumbarton Oaks in :qq6 I discovered that
the architecture and topography of Rome constitute an unad-
vertised strength of its library. In the final throes of preparation
I received welcome subsidy for cartographic assistance from the
dean’s fund for professional development in the College of Arts
and Sciences of the University of Vermont. To all these institu-
tions, and to the supportive band of colleagues and friends who
comprise them, I express sincere and lasting gratitude.
Bothtext andcommentary are by their very nature tralaticious
endeavours, and far beyond reckoning are the debts I owe to my
predecessors. That I specially admire the accomplishments of
Poleni and B¨ ucheler should be apparent on every page of this
edition, and to the loving labours of Thomas Ashby I have felt
the keenest attraction. The bibliography will reveal some at least
of the crucial help I have garnered from women and men who
represent an extensive range of scholarly expertise over a period
of more than five centuries. Of closer friends those to whom I
can no longer render thanks in person include Arthur and Joyce
Gordon, Peter Marshall, George Goold and John D’Arms.
For help and support of various sorts over many years I re-
spectfully acknowledge Crawford Greenewalt Jr, W. Kendrick
Pritchett, Richard Thomas, John Humphrey, Bruce Frier,
Ruth Scodel, Christina Kraus, John Peter Oleson, James
Clauss, Robert Arns, Andrea Salgado, Francis Newton, Z.
Philip Ambrose, Peter Aicher, William Mierse, Jane Chaplin,
Jacques Bailly, Cyrus Rodgers, Audrey Hunt, Eleanor Rodgers,
Jonathan Huener and Lutz Kaelber. Among those who pa-
tiently criticised discrete parts of this work I owe special thanks
to Charles Murgia, Harry Evans, Trevor Hodge, Michael
Crawford, Christer Bruun, Rabun Taylor, Roger Cooke and
Michael Peachin. It goes without saying that none of these
persons is responsible for any follies in which I have persisted.
Helena Fracchia and Maurizio Gualtieri accompanied me on
pleasant outings among the remains of the aqueducts and will at-
test that I was totally unprepared for their breathtaking majesty.
Don Faustino Avagliano, librarian and archivist, graciously
received me on two separate visits to Monte Cassino. Long and
pleasant hours were spent in great libraries at Harvard, Ann
Arbor and Berkeley; in many cases I found rewarding resources
in their numerous branches, notably the Houghton Library at
Harvard and the Bancroft Library at Berkeley. Amongst indi-
vidual librarians, I am specially indebted to Irene Vaslef and
Mark Zapatka (Dumbarton Oaks), Jean Hannon (Harvard Law
School) and Luminita Florea (Robbins Collection, Boalt Hall,
Berkeley). My own library at the University of Vermont has
proudly maintained a strong collection in classical studies; for
books not available here I am grateful to Connell Gallagher
in Special Collections for an occasional purchase, and to Nancy
Rosedale, Lisa King, Barbara Lamonda and Daryl Purvee in the
interlibrary loan department for constant labours on my behalf.
The map showing the extra-urban courses of Rome’s aque-
ducts is based upon a similar map in Peter Aicher’s Guide to the
Aqueducts of Ancient Rome (:qq¸), with permission of Bolchazy-
Carducci Publishers, Inc. That for the network within the City
is based upon one appearing in Harry Evans’ Water Distribution
in Ancient Rome: The Evidence of Frontinus (:qq¡), with permission of
the University of Michigan Press. For expert cartographic modi-
fications to these and for the map showing piscinae I acknowledge
the cheerful collaboration of my colleague Lesley-Ann Dupigny-
Giroux. It was a welcome relief when Christer Bruun agreed to
let me include his discussion on the value of the quinaria (appear-
ing here as Appendix C), for it spared me the frustrating task of
covering the same dreary ground.
Professor Michael Reeve has awaited the final version of
this book with far more patience than I deserve. For his care-
ful scrutiny, gentle corrections and wise suggestions I am more
grateful than I can say. Staff of the Cambridge University Press
have been consistently helpful: among those who merit special
thanks are commissioning editors Pauline Hire and Michael
Sharp and production editors Neil de Cort and Alison Powell.
Copy-editor Linda Woodward bent to her task with a singular
diligentia by which she has deserved well of Frontinus.
My wife Barbara Saylor Rodgers has had to hear all of my
thoughts from their first tentative expressions, for I rely con-
stantly upon her ability as an historian and a Latinist. Her steady
encouragement has been, I hope, to good effect, and for my
faults she can bear no blame. Warmest of all is my heartfelt ap-
preciation for the long, unselfish and never lessening interest of
Professor Herbert Bloch: that I am still his disciple is a special
Obscurity veils the origins and early career of Julius Frontinus.
As praetor urbanus he convened the Senate on : January in the
year ¸o cr, but he soon yielded the post to Domitian (Tac. Hist.
i\.¸q.:–.). A suffect consulship followed soon thereafter, prob-
ably in ¸¸.
His birth can with reasonable certainty be set in
the later years of Tiberius’ reign. In all likelihood he came from
Narbonese Gaul.
He may have spent his early years as aneques-
trian officer, perhaps with military service in the Parthian cam-
paigns of the late ¸os, perhaps as procurator in Spain and/or
Africa in the 6os. His behaviour in the political events of the
year 6q is entirely unclear. Syme suggests that Galba adlected
him into the Senate for swift adherence to his cause, but rapid
promotionunder Vespasianmight point ina different direction.
Between praetorship and consulship he may have held a mil-
itary command (presumably as legatus legionis), if he was on the
scene of the Rhineland revolt and received the surrender of
the Lingones (Str. i\.¸.:¡).
After the consulship he was almost
For his praenomen see commentary. Biographical data and testimonia are
conveniently collected in RE Julius, no. .¡¸ (:q:¡, Kappelmacher) ¸q:–¸,
RE Suppl. :¡: .o8 (:q¸¡, Eck); PIR
(:q66) i ¸... The best accounts of his
career are those of Birley (:q8: ) 6q–¸. and Eck (:q8.a) ¡¸–¸.; succinctly,
Bruun (:qq:) :o–::.
Degrassi in II :¸.:: :¸8. He can hardly have been consul before ¸¸: Eck (:q¸o)
8: and n..¸. A consulship in ¸¡ is unlikely, since he succeeded Cerialis in
Britain in the spring of that year.
CIL :..:8¸q (Vienne) names a senator Q. Valerius Lupercus Iulius Frontinus;
cf. Syme (:q¸¸) ¸:8, (:q¸8) ¸qo.
Syme (:q¸8) ¸qo and ¸q.; Eck (:q8.a) ¸o–:. Conjectures to account for the
short interval between praetorship and consulship have included the unlikely
possibility that F. was a patrician: so Birley (:q6¸) 6¸, but abandoned (Birley
(:q8: ) ¸o).
The authenticity of Str. i\ has been questioned; see RE loc.cit., with bib-
liography, also below n. ¸¸. That his legion was II Adiutrix, later taken to
Britain, is an interesting hypothesis: Ward-Perkins (:q¸¸) :o.–¡.
immediately appointed legatus Augusti pro praetore for Britain,
whither he went as successor to Petillius Cerialis in ¸¡ and
where he remained until the arrival of Agricola in ¸¸ or ¸8.
Tacitus’ biography of the successor (Agr. :¸..) describes Fronti-
nus’ command only briefly: subiit sustinuitque molem Iulius Frontinus,
vir magnus quantum licebat, validamque et pugnacem Silurum gentem armis
subegit, super virtutem hostium locorum quoque difficultates eluctatus. His
achievements in Britain may have earnt for him the triumphalia
References in the Strategemata (i.:.8, ¸.:o, ii.::.¸) suggest that
Frontinus was personally acquaintedwithDomitian’s campaigns
in Germany in 8¸, and it is probably to this period that one
shouldassigna dedicatory inscriptionfromVetera (CIL:¸.86.¡).
Whether Frontinus held the governorship of Germania inferior
is not clear; he may have been legatus legionis or comes of the em-
At the appropriate stage he received the proconsulship
of Asia, long regarded (with that of Africa) as the pinnacle of a
senatorial career. Coins from Smyrna with the legend óvûutó-
:cu 1pcv:ivcu have long been known,
and a recently studied
inscription from Hierapolis in Phrygia (modern Pamukkale) at-
tests him in that office.
His tenure can now with reasonable
certainty be fixed to 8¡/8¸.
The grand gates that bear his
name had largely been completed under his predecessor.
During the later years of Domitian’s reign Frontinus seems to
have played no major role in public life. There is not a shred of
evidence, however, to suggest that he was in disgrace, or that he
On the date of Agricola’s succession see Ogilvie–Richmond (:q6¸) ¸:8;
Birley (:q8: ) ¸¸–8. The most recent work on Frontinus’ governorship in
Britian is Boni (:qq8).
Eck (:q8.a) ¸¸.
Eck (:q¸o) ¸¸–8:, with bibliography; cf. Eck (:q8¸) :¡:.
BMC Ionia, .¸o, nos. :¸¸–¸; cf. Kowalewski (:qq¸).
AE :q6q/¸o, ¸q¸. First published by Monaco (:q6¸–¡); improved by Eck
(:q¸o) ¸¸–8:, and C. P. Jones (:q¸¸) 688.
Eck (:q8¸) .o8; cf. Thomasson (:q8¡) i: .:¸ no. ¸¸.
One can compare the inscription set up in the forum at Verulamium in ¸q
(AE :q¸¸, :6q): the name is that of the current governor (Agricola), although
Frontinus must have had a large part in the building project.
had deliberately chosen to maintain a distance from Domitian.
Quite the contrary: it is to the later years of Domitian that Pliny
refers in speaking of Corellius Rufus (cos. ¸8) and Frontinus as
prominent statesmen: Ep. \.:.¸ quos tunc civitas nostra spectatissimos

In this period Frontinus may first have turned to literary
activities, an amusement somewhat traditional for the senatorial
class. The poet Martial writes of having spent time in his com-
pany at Anxur (Tarracina), where the two friends enjoyed the
leisure of letters (Mart. x.¸8doctas tecumcelebrare vacabat/Pieridas).

Aelian writes of having consulted him on military matters, and
Pliny discussed legal topics with him in the early qos.

The lit-
erary treatises (which must surely be dated to these years) look
to be products of a contented retirement. A theoretical work on
military tactics, highly praised in Antiquity, has not survived.
His Strategemata reveal their author’s antiquarian bent; like his
gromatical writings,

they were ‘safely apolitical’.
With the reign of Nerva comes the final and most impres-
sive stage of Frontinus’ career. In q¸ he accepted the post of
curator aquarum (Aq. : and :o..:¸), and in the same year he
served on a senatorial commission looking for economies (Pliny,

See Eck (:q8.a) ¸8. Southern(:qq¸) ¡o goes sofar as tosuggest that Frontinus
might have been a member of the ‘privy council’ under Domitian.

White (:q¸¸) .q¸–6 n.¡: is unconvinced that Martial’s Frontinus is our man.

Plin. Ep. \.:.¸; Aelian, Tact. pr. ¸: tcpc 1pcv:ivc ttionucv otc:iscv
nutpc, :ivc, oit:pi¸c, oc¸cv óttvt,scutvc ttpi :nv tv :cï, tcìtuci,
tuttipicv . . . topcv cos tìó::cvc otcuonv tycv:c ti, :nv tcpc :cï,
Lììnoi :tûtcpnutvnv uóûnoiv.
Frontinus refers to this work in Str. i pr., and it was used by Aelian, Tact. pr.
¸ and Vegetius, i.8 and ii.¸.

These survive only in part, confused to some extent with a commentary by
Agennius Urbicus: see Dilke (:q¸:) :.6–¸., Campbell (.ooo) xxvii–xxxiii,
Chouquet-Favory (.oo:) .:–¡. For the possibility that the Frontinus of the
Corpus agrimensorumis not the same as our man, see Keppie (:q8¸) :.. Personal
experiences in Spain and Africa may underlie parts of this work, and Eck
has suggested an official assignment under Domitian: Eck (:q8.a) ¸¸, (:q8¸)
Syme (:q¸8a) 68: he means, I suppose, that they reveal nothing of their
author’s personality or public status. For a recent reviewof Frontinus’ literary
works, see Del Chicca (.oo.).
Pan. 6...).
He held a second (suffect) consulship in February
q8, with Trajan as colleague.
His son-in-law Sosius Senecio
was consul ordinarius in qq. And, a year later, Frontinus himself
was marked with the signal honour of a third consulship,
time ordinarius and again with Trajan as his colleague. That this
honour was not in fact unique only underscores the remarkable
status which Frontinus enjoyed. In q8 and :oo Julius Ursus also
received second and third consulships, both times in immediate
succession to Frontinus.
Pliny in his Panegyric (6:.¸, ¸) mentions
these iterated honours (the first by Nerva, the second by Trajan):
duos pariter tertio consulatu, duos collegii tui sanctitate decorasti. Theirs
was a praemium – Pliny is unambiguous – for eximia in toga merita,
by which he means that they had stood behind Trajan: utriusque
cura, utriusque vigilantia obstrictus es (Pan. 6o.6). Ursus was Frontinus’
junior by roughly a decade, but his marriage to Hadrian’s sister
reveals a special relationship to the princeps. We know of no
similar relationship linking Frontinus to Trajan, but one might
well have existed. Syme is probably safe in his speculation that
Frontinus might have been acquainted with Trajan’s father;

and there is little doubt that the third consulship was a reward
for his part in approving – we should perhaps say arranging –
the elevation of Trajan as Nerva’s heir and successor.

Frontinus’ death can be fixed in :o¸/:o¡ by Pliny’s succession
to his place in the College of Augurs (Ep. i\.8.¸).

His daughter
was married to Q. Sosius Senecio (cos. ord. qq, suff. :o¸); further
descendants appear in later inscriptions.
Cf. Syme (:q¸o).
CIL :6.¡.; II :¸.:, :q¡; cf. Mart. x.¡8..o.
II :¸.:, :q¸; CIL 6....., 8.¸o66 ( =ILAlg. ..6¸.), AE :q¸¸, .6.
Zevi (:q¸q).

Syme (:q¸8) ¸¸.

Recent discussions onthe succession, withgoodbibliography, are Berriman–
Todd (.oo:), Eck (.oo.).

For the date see Sherwin-White (:q66) .¸..
McDermott (:q¸6). Unclear is Frontinus’ exact relationship to his younger
contemporary, P. Calvisius Ruso Iulius Frontinus; Eck (:q8.a) 6o not un-
reasonably proposes that the connection was one of testamentary adoption.
Calvisius Ruso (PIR
c ¸¸o) was consul in ¸q, proconsul of Asia in q./q¸:
see E. Birley (:q8¸), R´ emy (:q8¸), Syme (:q8¸) ¡6–¸.
The spectacular sunset of Frontinus’ life was the product of a
combination of political circumstances. For others this era had a
radiance of dawn, caught for posterity in the artistic penstrokes
of panegyric and propaganda: these are the fellow senators who
call Frontinus vir magnus, princeps vir. But it is Frontinus himself
who invites us to see his career as one of long and sincere de-
votion to public duty. Personal satisfaction prompted his seem-
ingly un-Roman request

that admirers dispense entirely with
a tombstone:
impensa monumenti supervacua est; memoria nostri dura-
bit, si vita meruimus.
This is not modesty. It is rather the proud
statement of a man confident of the place awaiting him in the
fields of Elysium; there he will join those [qui] sui memores alios
fecere merendo.
Its date
This booklet – in its present form– cannot have been completed
until sometime early in the year q8. Frontinus’ appointment as

Presumably fromFrontinus’ will: Sherwin-White (:q66) ¸o¸; Eck (:q8.a) ¡¸,
Champlin (:qq:) :¸o. Baldwin (:qq¡) ¡86 notes the strikingly similar finale
of Tacitus, Agr. ¡6.¸ simulacra vultus imbecilla ac mortalia sunt (with reference to
Ogilvie–Richmond ad loc.).
Plin. Ep. ix.:q.6–8, where context (dum mavult videri contempsisse) makes plain
that Frontinus had not renounced claims to gloria; cf. DeLaine (:qq¸) :¸o.
Gloria, in the end, was what really mattered. A monument by itself, how-
ever traditional and valued (cf. Plin. Ep. \i.:o.¸, distressed that the ashes
of Verginius Rufus had lain for near ten years sine titulo, sine nomine), was no
guarantee that memory would be permanent. With roots of a sterile fig tree
the satirist can shatter the record of a lifetime. Juvenal (:o.:¡o–6) might con-
ceivably have had Frontinus in mind; his antithesis of fama and virtus closely
resembles Tac. Ann. i\.¸8. See further H¨ ausle (:q8o) esp. .q–¡o, Champlin
(:qq:) :6q–8..
Emphasis, as one would expect, is on the final verb. Merita are accomplish-
ments for the public good; because it implies recognition the word is a
stronger (and less-objective) alternative to res gestae.
Virgil, Aen. \i.66¡: Rodgers (:qq.). Note also Hor. C. ii..o..¸–¡ sepulchri /
mitte supervacuos honores.
curator aquarum took effect in q¸ (:o..:¸).
His first task was to
discover what the office entailed: primum ac potissimum existimo . . .
nosse quod suscepi (:). Study of his curatorial responsibilities and a
personal review of the water-system will have taken some time.
He speaks, for example, of monitoring conditions during the
summer months (q.8, ¸¡.¸). The text we have was completed
only after Nerva’s death in January q8: that prince is twice styled
divus (:o..¡ and ::8.¸).
Short of Frontinus’ death (:o¸/:o¡), a
terminus ante quem cannot be fixed. But the content and form of
the work itself are wholly consistent with the view that Frontinus
prepared it for circulation at no long interval after Nerva’s death.
He rehearses detailed instances of unhappy practices he has
detected, and he mentions reforms introduced as well as plans
undertaken but not yet complete.
The work thus seems to
reflect what the curator has learnt from (and accomplished in)
something like a year’s experience.
The date of compositioncannot be entirely separatedfromthe
question of Frontinus’ term as curator. The post had originally
been given to eminent men of some seniority and was an ap-
pointment for life; in later practice curators had been younger
consulars serving shorter terms as part of the senatorial cur-
sus (:o..:n.). Exactly what had been the pattern just prior to
Frontinus’ appointment is unfortunately not at all clear
(:o..:¸n.). Some have supposed that Frontinus kept the office
until his death – a view that might be supported by the sense
of traditionalism represented in his approach to the office and
the projected ideal of intimate and continuing cooperation be-
tween princeps and curator. Others have argued that the of-
fice could not be held simultaneously with a consulship and
Nomination late in q6 cannot be excluded. Fromchapter :o. we can observe
a pattern of successors assuming the curatorship with the consuls of the
following year – at least for curators who died in office.
Trajan’s name occurs but once in the transmitted text – an instance which
I judge to be an interpolation (q¸.¡n.).
Note especially 88.¡: revised figures are not yet available. The changes
Frontinus has outlined will have required a not inconsiderable length of
that Frontinus’ term as curator must therefore have come to an
end when he assumed the fasces for a second time in February
of q8.
If Frontinus relinquished his curatorship at the time of his sec-
ond consulship, he might naturally have taken this opportunity
to pass on to a successor his collection of data, freshly updated,
along with a catalogue of reforms already set in motion and
initiatives projected for the future. We could then interpret the
words in his prologue (..¸) to mean: ‘I made some notes for my
own benefit, starting at the beginning of my term, and these
will now perhaps be useful to my successor.’ It is scarcely credi-
ble, however, that Frontinus’ term as curator ended early in q8.
Nerva had more important things on his mind than to replace,
after so short a term, an official who was dutifully and effec-
tively addressing important problems of urban administration
(even if the major aims of reform were successfully under way) –
merely to avoid the overlap with a suffect consulship of very
limited duration. Add the fact of Nerva’s death and Trajan’s ab-
sence (until sometime in qq), and it becomes altogether easier
to suppose that Frontinus continued as curator at least until his
third consulship (:oo) or even until his death.
Nor can I see
any drawback to supposing that he gave this booklet its present
form while continuing in office. On the one hand, his statement
‘This will be useful to me’ (..¸) is better taken as rhetorical lib-
erty (with its author still in office) than as rhetorical fiction (from
the pen of one who has already retired). On the other hand, his
closing words (:¸o) give no hint at all that it will be someone
other than himself who will uphold the trust of the curatorial
Cantarelli (:qo:) .o:, Syme (:q¸o) ¸¸, Ashby (:q¸¸) .o, Grimal ix, xvi.
Eck (:q8.a) 6o, speaks of the years after the third consulship, ‘in denen er
wohl weiterhin als curator aquarum t¨ atig war’.
The tenses in chapter :¸o are present perfect (laboravimus, fuisse) and the last
main verb is opto; see commentary for difficulties with the final word (C
has praestitit). Note also 88.¡, where the future tense (adiunxerimus) seems to
indicate that Frontinus is still in office.
Let us be satisfied with dating the literary form of the De
Aquaeductu as it stands to sometime in q8, after Nerva’s death
and with Trajan not yet come to Rome. Frontinus in this period,
during his second consulship and in the months that followed,
was among the small circle of senatorial leaders in whose hands
lay control of the state’s constitutional helm. Not only did he
retain the office of curator aquarum, but he was simultaneously
one of the emperor’s vice-gerents at Rome.
Its content and form
In his prologue to the De Aquaeductu
Frontinus refers to this
work as a commentarius, and explains its genesis as a collection of
material made primarily for self-instruction and personal refer-
ence (...–¸). Let us look first at what the booklet contains and
then at how it fits the definition of a commentarius.
The contents fall readily into two categories. The first em-
braces the matters that Frontinus outlines later in his prologue
(¸.:–.), while the second category could be called ‘editorial
remarks’, Frontinus’ commentary, as it were, on the data he
has collected. These comments are indeed so extensive that
the rhetorical modesty of the prologue (..¸) comes very near
to embarrassing untruth. ‘For his own benefit’ Frontinus hardly
needed to record the delinquencies he had observed and the re-
forms he had made in the course of his initial months in office.
These, plainly, were included for the edification of some other
In the ‘table of contents’ of chapter ¸ Frontinus promises the
following material (in parentheses are the chapter references to
the work itself ):
: Data onindividual aqueducts: persons who built them; dates
when they were built; location of the sources; length of the
conduits (broken down into types of construction); heights
of the terminal delivery points. (Chapters ¸–..)
The title is that found in C, the unique manuscript; it is not without difficulty
(see commentary).
. Data on distribution: pipes and their sizes; quantities deliv-
ered according to the supply of water available;
of delivery (imperial properties, public uses of various sorts,
private persons); distribution among the wards of the City.
(Chapters .¸–86)
¸ Legal matters pertinent to the right of drawing public water;
precautions for upkeep of the channels; penalties for abuse.
(Chapters q¡–:¸o)
For much of this material, most indeed of the first two categories,
a modern writer would have chosen a tabular format. The infor-
mation thus collected all served an administrative aim. Frontinus
recognised its potential usefulness and the importance of having
it readily to hand.
It is primarily to this material that he refers
when he states that he has collected information in commentarium
quem pro formula administrationis respicere possem (...).
Not explicitly announced in the prologue are those portions
of the work which represent Frontinus’ critical review of the
data he has collected and his administrative analysis of the sys-
tem he has undertaken to superintend. This is nowhere more
noticeable than in his exhaustive scrutiny of the official figures
for the quantity of available water (6¡–¸6) and in his optimistic
account of projected improvements (8¸–q¸).
But comments of
an explanatory or editorial nature are not limited to such obvi-
ous addenda. They occur throughout the work, combined for
See commentary to ¸.. pro suo modo.
Of only marginal usefulness perhaps were the ages of the aqueducts and
the names of the builders. But the auctores were an essential element of
identification, and the dates relate to types of construction used for different
aqueducts (and consequently to peculiarities of their upkeep).
Evans (:qq¡) ¸¸–¸.
Grimal ix–x, ¡¸ n.:, 6¸ n.. thinks that these two passages are unannounced
in the prologue. That the second is absent there is true enough, but the
same could be said of many other passages (e.g. ::6–:8 on the familiae). As
was noted by Rubio (:q6¸) ¸¸–¡, Frontinus probably in fact does announce
chapters 6¡–¸6, althougha textual difficulty unhelpfully occurs at the crucial
point (¸..n.).
the most part so harmoniously
that, were it not for Frontinus’
explicit statements in the prologue, one would suspect that the
booklet was indeed composed post experimenta et usum (..¸).
We find, then, in our text of the De Aquaeductu both data and in-
terpretive matter, either or both of which are accepted as normal
for the contents of a commentarius. ‘A commentarius could be a pub-
lished composition in the plain style, or lack polish altogether:
there was no firmtradition, as for the major genres of prose. The
subject matter and the author’s personality, rather than rules of
genre, determined the character and quality of the writing.’
Goodyear’s definition, which he applies to Frontinus, is a good
one, but more can be said. In a strict sense, the term commentarius
describes a text that accompanies and explains something. So
in the present work chapters ¸–:¸ are the text to which Fron-
tinus could refer for details as he and others interpreted the
maps or diagrams that he tells us he had prepared (:¸).
chapters ¸q–6¸ are an official listing of the calibres authorised
for delivery-pipes. Speaking more generally, commentarius was the
term applied to notes and records of many sorts, some of which
might remain in the form of data such as lists or compendia,
while others might be incorporated into a quasi-archival series
or be polished for wider circulation. Thus, on the one hand,
we have Caesar’s commentarii, while on the other we know of
commentarii of individual magistrates and priests which formed
the libri ‘records’ of magistracies and priesthoods.
An exam-
ple from the present work: Frontinus speaks of the commentarii
‘records’ of Agrippa (qo.¸), which, beginning with Augustus, had
Chapters ¸¸–6 are a glaring exception, and I take their awkwardness as
an indication that they were an afterthought, included at the last minute:
Rodgers (:qq:).
Goodyear (:q8¸) 6¸..
Evans (:qq¡) ¸q, DeLaine (:qq¸) ::8.
On commentarii in general see Von Premerstein (:qo:) and B¨ omer (:q¸¸);
cf. R¨ upke (:qq.) esp. .o.–¡. For records and archives of land distributions
Nicolet (:qq:) esp. :¡q–6¸, Moatti (:qq¸); grain distributions, Tarpin (:qq8);
provincial governorships, Haensch (:qq.); magistrates and senate, Coudry
(:qq¡) esp. 6¸–8¸; priesthoods, Sini (:q8¸), Scheid (:qq¡), North (:qq8);
cf. also Panciera–Virlouvet (:qq8) and below, n. ¸:.
apparently been maintained and supplemented up to his own
day and to which he refers as the commentarii principum (¸:.., 6¡.:,
Agrippa kept many records besides those of his water-
management cares, and upon such materials as these no doubt
he drew in composing a commemoratio of his aedileship, whether
or not that was part of a larger autobiography.
In short, with no
clear distinction required, the term commentarius could embrace
both data and interpretive matter. This is precisely what Fronti-
nus’ booklet contains. But such a combination is not the rule for
a commentarius and, given the rather abstruse subject-matter of
water-conduits and water-rights, oversight and upkeep, the De
Aquaeductu is in fact unique as a specimen of Roman literature,
and even perhaps of the ancient world as a whole.
Unsurprising, then, is the lack of consensus over how to cat-
egorise this work. Traditional views have been that it is a piece
of ‘technical writing’ or an ‘administrative manual’. Such cate-
gorisations reflect nothing so much as the perspectives of their
proponents. Archaeologists and historians of technology have
come to Frontinus for what he can offer them by way of techni-
cal information; students of Roman law and government have
looked to him for evidence of legal and administrative proce-
dures and practices. Recent scholars have taken a wider view,
and it can now be agreed that Frontinus should not be accorded
the status of a technical writer, much less an authority, just be-
cause he administered a vast hydraulic systemand therefore had
occasion to write about things that were indeed technical. Nor
because he is an administrator of high standing must we assume
that his text was anything like an administrative guide:
it is not,
I think, entirely a mark of ironic modesty that even he is unsure
that a successor will find his booklet to be of any real use (..¸).
Certainly we must be aware that Frontinus is selective in what
he includes and that he can be prescriptive as well as descriptive;
Reinhold (:q¸¸).
Pliny, HN xxx\i.:.: adicit ipse aedilitatis commemoratione. Agrippa’s autobiog-
raphy is mentioned in Serv. Dan. ad Georg. ii.::.: Roddaz (:q8¡) ¸68–¸:.
Bruun (:qq:) :¸–:q, ¸6q–¸o, Hodge (:qq.) :6–:8, Evans (:qq¡) ¸¸–6¡.
those who use him must do so critically.
Similarities between
the De Aquaeductu and administrative texts such as the Gnomon of
the Idios Logos, libri mandatorum and libri de officiis known from later
centuries all seem more apparent than real.
However much a
pioneer Frontinus might have been,
the plain fact is that noth-
ing like his De Aquaeductu is known, let alone survives, from the
ancient world.
Its audience and purpose
To account for apparent inconsistencies between the prologue
and the text as a whole, or for the presence of both promised data
and unannounced commentary, it is easy to assume that Fronti-
nus’ plans somehowchanged between the outset of the work and
its eventual completion. Grimal puts it this way: ‘Frontin avoue
avoir commenc´ e son trait´ e pour sa propre instruction. Mais, peu
` a peu, son projet a chang´ e: il r´ efl´ echit sur les constatations qu’il
fait, critique les renseignements qu’il trouve, et cela modifie la
r´ edaction mˆ eme de son ouvrage, si bien que celui-ci finit par
apparaˆıtre comme une sorte de ‘journal’ de sa gestion.’
administrative report, then, instead of a manual?
To a report of findings about the water commissioner’s office
and of initiatives undertaken the label commentarius remains en-
tirely appropriate. But ‘report’ implies an audience for whom
that report has interest or import. If the commentarius before us
goes beyond the stated aim to be merely a work for personal
reference, no more could it have been written for an audience
Bruun (:qq:) :¡ is pained that he must cite no less a scholar than Hirschfeld
(:qo¸) .¸¸: ‘die Abhandlung des Sextus Iulius Frontinus . . . kann durch in-
schriftliche Dokumente kaum eine Erweiterung erfahren; vielmehr k¨ onnen
diese nur als Zeugnis f ¨ ur die treue Darstellung herangezogen werden’.
Bruun (:qq:) :¸. On the Gnomon see Riccobono (:q¸o); for mandata, Finley
(:q¸¡), Dell’Oro (:q6ob), Marotta (:qq:); for libri de officio, Dell’Oro (:q6oa).
Eck (:q8.a) 6: notes that ‘Amtshandb¨ ucher f ¨ ur verschiedene Aufgabenbere-
iche hat es zumindest f ¨ ur diese Zeit und f ¨ ur die stadtr¨ omischen
Amter nicht
gegeben.’ For the possibility of Frontinus as a pioneer in gromatical writings,
see Campbell (.ooo) xxvi–xxxi.
Grimal x.
of one, a curator who might be Frontinus’ successor. The audi-
ence, then, must have been wider, and for a man of Frontinus’
standing that audience can quite accurately be determined: the
senatorial class as a whole and – by no means incidentally – the
new princeps.
‘Administrative report’ or not, we still must ask what was Fron-
tinus’ purpose in putting this text into somewhat careful literary
form for this audience at this time. In this context we cannot
neglect that second component of Goodyear’s commentarius – the
author’s personality.
The De Aquaeductu, as McElwain puts it,
gives us glimpses beyond the concerns of an officious adminis-
trator: ‘It depicts a man; it depicts motives and ideals, the springs
of conduct.’
Ashby sees Frontinus’ senatorial sympathies as in-
forming all of his writings, and most especially the De Aquaeductu,
in which this author ‘epitomised for all time the ideal of an ef-
ficient civil servant’, a representative of ‘the best answer that
citizen-government could give to the alternative of professional
administration by free or slave delegates of the princeps’.
views this work as ‘un ´ ecrit politique’, ‘un manifeste officieux’,
and its author as a propagandist for the regime: ‘Frontin n’est
que le porte-paroles du Prince.’
Others find a sense of self-
enhancement pervading this work;
Bruun and DeLaine go so
far as to suggest that memorialising himself might have been
a distinct motive for Frontinus to put this booklet into literary
DeLaine also has perceptively noted that literary form
and rhetorical features are likely to be more than icing on what
had initially been a tedious if data-rich cake. She thinks that
Christ (:q8q) looks more broadly than have others at Frontinus’ personality.
McElwain (:q.¸) xv.
Ashby (:q¸¸) .6–¸¸.
Grimal xv–xvi; Hodge (:qq.) :6–:¸ is in essential agreement. Grimal does
not say which prince: Nerva, whose initiatives Frontinus reports, now dead?
Trajan, absent from Rome at the time of publication? Hodge speaks of
Nerva and Trajan, with the tendency of modern scholars to oversimplify the
continuum. DeLaine (:qq¸) :.o n.:., :¸¸ n.¸¸ remarks on the ambiguity.
Bruun (:qq:) :¸8–q, :86–¸, ¸¸o; Evans (:qq¡) ¸¸–8; cf. Baldwin (:qq¡)
Bruun (:qq:) ¸¸o, DeLaine (:qq¸) :.q–¸o, :¸6–¸.

between a commentarius proper (for personal use) and what we
have (intended for some – but hardly extensive – circulation)
came a speech to the senate of which the present text would
be the published version.
Professor Michael Peachin contem-
plates a speech as well, but perhaps to a broader audience, along
the lines of the contio of Caelius Rufus (mentioned by Frontinus
in chapter ¸6). He proposes that our present work might most
fittingly be described as a pamphlet, addressed to fellow senators
and persons with significant commercial interests involving wa-
ter, an apology for – but an unambiguous announcement of –
the watchful restoration of policies and penalties that had been,
but no longer will be, overlooked by responsible officials. On
this view, the final paragraph of the De Aquaeductu, often seen as
puzzlingly abrupt, assumes a meaningful significance, and the
author ends his work on a note of respectful firmness.
The curator aquarum and the emperor
Whether or not it was part of his purpose in producing our text
of the De Aquaeductu – I very much doubt that it was the sole pur-
pose – Frontinus’ work describes a close relationship between
the emperor and himself as curator aquarum, that is a senior con-
sular named to important office as the emperor’s administrative
The importance of this partnership is always near the
surface – fromthe opening words (res ab imperatore delegata, mihi ab
Nerva Augusto) to the very end (where he is the immediate inter-
cessor for those who seek indulgentia principis). This booklet reveals
the persona at least of a dutiful and diligent public servant.
One essential element in successful administration, as
Frontinus presents it, is energetic and personal involvement on
the part of the curator. Only intimate familiarity with the system
DeLaine (:qq¸) :¸.–8.
Prof. Peachin has generously shared his views with me, both in personal
conversation and in a preliminary draft.
The fact that this post was filledby the emperor’s ownnominee is not without
significance: see commentary on :oo.:.

under his management can free himfromdependence upon un-
derlings. Thus far Frontinus himself in the prologue, explicit and
unequivocal. Within the text, wide-ranging reforms to eliminate
fraus in lower levels of the administration testify to his diligentia,
in pointed contrast to the inertia and segnitia of his immediate
predecessors. The prologue heralds a second, no less important,
theme. Only by shouldering the responsibilities of his office could
the curator fulfil the expectations of the emperor who had ap-
pointed him – and the emperor it was who bore final responsi-
bility for the reliability and adequacy of the urban water system.
Since the burden was borne by two men, success could result
only from a close cooperation between princeps and curator.
Inequalities of the partnership were not to be overlooked, but
these were trifling and insignificant in the face of the common
goal: by any standards, Rome’s water supply was one of the most
magnificent gems in the City’s imperial crown.
A collaborative personal relationship between the emperor
and his curator was no idealist’s dream; for Frontinus it was
an inherent part of the curatorial office as established by
Augustus in :: ncr. Indeed, the origin and nature lay further
back, for this particular Augustan cura was deliberately fash-
ioned as a means of perpetuating the public services of Marcus
Agrippa, lifelong friend and apparently selfless ally of the first
princeps. Beginning at least with his munificent aedileship in
¸¸ ncr, Agrippa had single-handedly assumed an overall re-
sponsibility for Rome’s water supply. From his own purse he
had paid for new construction, and he kept a gang of slaves
as a standing maintenance crew. On Agrippa’s death Augustus
inherited this gang, and with it he accepted the full range of
Agrippa’s responsibilities. Parts of Agrippa’s ‘cura’ the prin-
ceps kept for himself: the willingness to cover costs of ma-
jor building and repairs, as well as the privilege of granting
public water to certain private parties. To a senatorial agent
he entrusted the routines of administration with concomitant
powers (some of which had fallen to Republican censors).
The office was not unduly onerous, and the high prestige

it carried was no doubt intended as a deliberate tribute to the
lasting achievements of Marcus Agrippa. More than political
tact lay behind Augustus’ choice of Messala Corvinus as the first
curator aquarum.
Over a century had elapsed by the time Frontinus came to the
post, and important changes had taken place. Most dramatic of
these was that celebrated by Claudius in ¸. cr, on completion
of two new aqueducts which roughly doubled the water supply
and added more than a hundred Roman miles of channel. Fol-
lowing the Augustan rule, Claudius himself paid the vast sums
for construction. Following the example of Agrippa, he insti-
tuted an administrative system for their upkeep, which seems in
some ways to parallel that under the management of the sen-
atorial curator. Composed of imperial slaves and managed by
the emperor’s own freedmen, the new branch of administra-
tion may gradually have eclipsed the older system. Whatever
circumstances might have been in the years just prior to his ap-
pointment, Frontinus unmistakably depicts – with or without
exaggeration – an administration in shambles.
The diligentia that Frontinus brought to his post had positive
results which went well beyond restoring discipline among the
work crews and clarifying the distinct responsibilities of lesser
members of the administrative service (including, of course, the
procurator). Reform of the curatorial post itself forms no small
part of his accomplishment. Nosse quod suscepi: to avoid the faults
he found in his predecessors, Frontinus sought the best possible
definition of his job, that contained in the series of legislative
acts establishing the curator’s office.
With these documents
(and an appreciation for the historical circumstances that lay
behind them) he had, in effect, a ready-made model for his own
performance, not to fashion a new office but to restore to it the
senatorial dignity, perhaps even the prestige that it once had
The commentarii of the Secular Games, Pighi (:q6¸), begin with senatus consulta
of :¸ ncr relating to their organisation; Scheid (:qq¡) :8o.
The grand example of Marcus Agrippa must never have been
far from Frontinus’ mind. Scholars duly note that Frontinus’
maps of the aqueduct network (:¸.¸–¡) remind us of the more
extensive maps set up by Agrippa.

I would go further: when
Frontinus tells us that he has cast his eyes over even the stretches
in outlying gorges and tortuous mountain tracts, it is hard not
to recall that Marcus Agrippa once took a boat ride through the
sewers of Rome (Dio, xrix.¡¸.:).

His deference to the princeps
is no less telling, for nothing can mask Frontinus’ honest pride in
the fact that his own administrative diligence in increasing the
water supply was practically equivalent to the engineering feat
required to develop entirely new sources.

Again, dispensing
with one of the trappings of his office, Frontinus explains fides
nostra et auctoritas a principe data pro lictoribus erit (:o:.¡). With such
apparent modesty he can underscore the security of his own per-
sonal relationship with the sovereign, and we should not fail to
remember that for more than two decades Agrippa had served
his prince with comparable modesty. Finally, inasmuch as it is a
rehearsal of his accomplishments as curator, Frontinus’ commen-
tarius recalls the published memoir of Agrippa’s aedileship.
In recounting his own administrative activities, Frontinus
never fails toemphasise that he is acting at the behest or onbehalf
of the emperor who appointed him. When amending fraudulent
mistreatment of the Tusculans’ rights to the aqua Crabra (which,
incidentally, Agrippa had respected), Frontinus proceeds iussu
imperatoris (q.¡–¸). He undertakes an exhaustive scrutiny of the
water available praeeunte providentia . . . principis Nervae (6¡.:). His
own discovery that far more water was available than had been
known he modestly concedes to the providentia diligentissimi prin-
cipis (8¸..). In a matter of financial irregularity we read of the

Evans (:q8.), Dilke (:q8¸) ¸q–¸¡, :o:, Nicolet (:qq:) 8¸–:.., Evans (:qq¡)

Evans (:qq¡) 6o–., Baldwin (:qq¡) ¡q6–¸, DeLaine (:qq¸) :¸¸.

Note especially ¸¸.: nova quadam adquisitione aquarum and 8¸.. quasi nova
inventione fontium.
Pliny, HN xxx\i.:.: (see above, n. ¡¸).

iustitia divi Nervae immediately pursued by nostra sedulitas (::8.¸).
Official calibrations of pipe-sizes must be recognised and made
standard, not least because they are consistent with data in com-
mentariis principis (¸:..). Because it will involve construction (an
imperial prerogative carried out at imperial expense), he credits
to Nerva the plan to improve the quality of Anio Novus and he
promises that a prominent titulus will in due course recognise the
emperor as auctor of the completed project (q¸.:–¡).
Such repeated stress on his own close relationship with the
emperor lends to this work a self-laudatory tone which might,
although I think need not, imply that Frontinus is smug, arro-
gant, or deliberately self-serving. The atmosphere in the first
half of q8 was not one in which responsible political leaders
were likely to be comfortable, relaxed and confident. Human
beings in their seventh decade are not immune to ambition, and
Frontinus had a son-in-law whose career was moving forward.
Nor was that nexus of which he formed a part an insignificant
one: men looked to him as a patron – and he had played a role
in Trajan’s succession. It might be fairer to Frontinus to call him
an idealist in the cause of the Roman elite, preaching that happy
possibilities for the commonwealth could emerge from cooper-
ative good will between princeps and senate. Frontinus was one
whose seniority and status permitted him to speak for an entire
order whose pride and purpose in statesmanship had recovered
dramatically in the months since Domitian’s death.

regime had kindled new senatorial hopes; however confidently
one may have looked to Trajan as a successor, this flickering
flame needed special care.
Praise of Rome’s aqueducts constitutes praise of the Roman
state itself. Frontinus sketches their history along annalistic lines
and recites their auctores, great figures in Roman history,

Note the enthusiasm voiced by Tacitus Agr. ¸, a work almost exactly con-
temporary with the De Aquaeductu. It is unfortunate that Frontinus finds no
mention in Wirszubski (:q¸o) :.¡–¸:.
Bruun (:qq:) :8, Evans (:qq¡) ¸¸, Baldwin (:qq¡) ¡qo–., DeLaine (:qq¸)
indeed he treats as if they were his direct predecessors in a long
series of curatores aquarum. There is a touch of the antiquarian
in our curator, not surprising perhaps in a novus homo, which
finds expression in his mention of the tergiversations of Appius
Claudius (¸.¸), his interest in the triumphal success of Marcius
Rex in bringing water to the Capitol (¸.¸), the statutory gems he
culls from the archives (q¡.¸, q¸.¸), the snippet of legal trivia he
cites from Ateius Capito (q¸..–¸). But he saves his finest rhetor-
ical flourishes for the masonry structures that brought water to
Rome. Because they are no less practical than they are grand,
for him they outrank the most spectacular wonders of Rome’s
predecessors (:6). They are a plain and visible symbol of all that
Rome is and all that she stands for. Their maintenance and
upkeep in and of itself is a matter worthy of especial care, cum
magnitudinis Romani imperii vel praecipuum sit indicium (::q.:). The
water they furnish supplies the public demands of a luxuriant
metropolis, serves as a resource crucial to the security of its citi-
zenry, and assures a wholesome atmosphere worthy of the regina
et domina orbis (88.:).
Frontinus, senior consular and partner of the emperor, im-
presses in the end as a man firmly righteous in the vein of an
old-time senator. Self-serving or not, he shows himself both will-
ing and able to stand before his fellow citizens to exemplify the
energies of a lifetime devoted to the best interests of the Roman
state. The memoir of anadministrator canhardly be more thana
modest platform; the speaker’s tones may fall short of resonance;
but there is nonetheless in the De Aquaeductu a quiet eloquence in
the affirmation of a senatorial idealism. Frontinus was not alone
among his contemporaries in comfortably and capably uphold-
ing long traditions of senatorial dignity.
His voice is that of a
man proud of himself, proud of his City, proud of her monu-
ments, proud of her standing as queen of the world. If he flatters
himself by asserting that his present responsibility is an office
administratum semper per principes civitatis nostrae viros (:), it is equally
Cf. Talbert (:q8¡) ¡qo.
true that per quos (¸.:) has a Livian ring beyond the whims of an
The sources
There is no great complexity to the question of Frontinus’
Answers to some questions posed by a new cura-
tor could presumably have been found in the archives of his
own and closely related bureaux.
(Convenient accessibility of
the information is another matter, although Frontinus’ decision
to include certain types of data does not in itself suggest that
these were especially hard to retrieve.) For the data that relate to
supplies and deliveries Frontinus explicitly draws upon the com-
mentarii principum (6¡.:), records that were maintained by the im-
perial staff but the origin of which lay in the personal commentarii
kept by Marcus Agrippa (qq.¸). Official figures for pipe-sizes
were also safely recorded in imperial commentarii (¸:..); these
sizes had been standard since the time of Agrippa (.¸.:, qq.¡).
At his own disposition the curator had a clerical staff whose
records must surely have contained a rich miscellany of highly
specific data.
Frontinus gives no hint of having taken personal
measurements of the channel lengths, for instance, a fact which
strongly suggests that these figures were already available.
same can be said for location of the sources, and the manner
in which he gives directions to Marcia’s spring implies that he
used written records (¸.6n.). The nucleus of these records, like
that of the imperial registers, was presumably formed during
Agrippa’s lifetime. Copies of relevant legal texts might well have
been available in the curator’s office (along with mandata issued
Baldwin (:qq¡) ¡q.–¸.
For archives and record-keeping, see Posner (:q¸.), Talbert (:q8¡) ¸o.–
¸¸, Culham (:q8q), Haensch (:qq.), Coudry (:qq¡), Crawford (:qq6) .¸,
DeKleijn (.oo:) ::o–:¡. See also above n. ¡¸.
The practice of detailed accountability dated back to Republican times (q6,
Probably even broken down into categories as Frontinus presents them
by the emperors), although Frontinus’ remarks on the adminis-
trative practices under the Republic (q¡.. dum altius repeto leges)
reveal that he had undertaken some archival research.
For historical details or information to clarify uncertainties
he encountered Frontinus could turn to the ordinary range of
literary sources. Of these one imagines that the works of Marcus
were specially useful: from such a primary source
Frontinus must have learnt the particular circumstances that at-
tended the discovery of Virgo’s springs and the date on which its
water was introduced into the City (:o..–¡).
The works of his-
torians and annalists sometimes yielded details of similar interest
for the earlier aqueducts. Frontinus seems not always to have re-
lied on a single source (cf. ¸.6 constantius traditur),
and in one
instance (the origin of the quinaria as a standard: .¸.:–¸) he ex-
presses an equal dissatisfaction with two available explanations,
both of which may derive from oral inquiries. Citing a source by
name occurs but twice (¸.¡ Fenestella, q¸.. Ateius Capito), and
in both cases the point is parenthetical: neither author was his
sole or even his main source. Likewise, his reference to the contio
of Caelius Rufus (¸6.:) is an aside, for rhetorical point.
Studies of Frontinus’ language and style have until recently been
focused on the question of the authenticity of Book i\ of his
Above, n. ¡¸.
Unless one supposes Frontinus to have recreated the story from oral tradi-
tions surrounding a picture set up near the springs (:o.¡). The date might
equally well have been preserved in an annalistic context: he can be similarly
precise about the dedication of Claudia and Anio Novus (:¸..), and it can
hardly have been novelty that the Fasti Ostienses record a specific date for the
introduction of Aqua Traiana in the year :oq.
Grimal xiii: ‘Pour chaque aqueduc, Frontin a recours aux dossiers contem-
porains de l’adduction.’ The contention bears no close scrutiny and provides
an extremely precarious basis for arguments set forth by Roncaioli Lamberti
A comprehensive index verborum appeared in :q8¸,
and more sophisticated resources are now available electroni-
A few scholars have looked closely at Frontinus’ vocab-
ulary and lexicon,
others at his prologue or the structure of
the De Aquaeductu as a whole.
Judgements of his style have
embraced the briefly dismissive and the unfairly contemptu-
Most critics, however, have appreciated that his material
prescribed for Frontinus an intentionally mixed style, which he
accomplishes well enough but without distinction.
In der Tat ist Frontin ein Fachschriftsteller. Mehr als die Form inter-
essiert ihn der Stoff. Seine Sprache ist freilich zumeist klar und gew¨ ahlt,
bewegt sich gern in ziemlich sorgf¨ altig gebauten Perioden und ist sogar
mit rhetorischenStilmittelnwie AlliterationundProsametrik versehen.
Aber auf stilistische K¨ unstlerschaft kann er irgendwelche Anspr¨ uche
nicht erheben, und sein Stil ist . . . manchmal schwerf¨ allig, unsch¨ on
und holperig.
Il commentario sul De aquis Urbis Romae ci presenta una prosa degna
di nota: Frontino in questa sua opera, in vero poco conosciuta dalla
critica, ci si revela con uno stile forbito e alquanto singulare per il
suo tempo; ci dimonstra anche un fondo culturale considerevole. Il
Chief among the challengers of Frontinus’ authorship are Wachsmuth
(:86o), W¨ ollflin (:8¸¸), Gundermann (:88:). Those whose responses have
dealt primarily with philological detail are Fritze (:888), Esternaux (:88q),
Kortz (:8q¸), and especially Bendz (:q¸8, :q¡¸).
Costas Rodr´ıguez (:q8¸), Frontini index. Of electronic materials, those
with which I am most familiar are the Packard Humanities Institute’s
PHI computer file: cn nox #¸.¸ (:qq:) and the IntraText website
<¸88/>, the latter permits one to
search an author’s word-usage by frequency, inverse alphabetical order, and
word-length, as well as providing a range of statistical information.
Hern´ andez-Gonz´ alez (:q8¸, :q8¸), L´ opez Moreda (:qq:), Espinilla Buis´ an
(:qq¸), Del Chicca (:qq6, :qq¸).
Santini (:qq.), Del Chicca (:qq¸–¸, :qq¸), DeLaine (:qq¸); Baldwin (:qq¡)
treats more comprehensively of both language and structure.
McElwain (:q.¸) xv ‘absolute lack of of stylistic charm’. Goodyear (:q8¸)
6¸. ‘in general unaffected, though one finds occasional embellishments’.
Hodge (:qq.) ‘one of the driest [works] ever written . . . wholly devoid of
literary pretensions or elegance whatever’.
Bendz (:q¸8) :¸., speaking of both the Strategemata and the De Aquaeductu.
suo periodo a volte ` e breve, conciso, anzi scarno; a volte ha un tono
solenne e l’ampiezza della prosa ciceroniana. . . . Inoltre, ci si accorge
immediatamente d’avere a che fare con un uomo colto e tecnicamente
preparato. Frontino infatti in questa sua opera non fa che misurare,
controllare e descrivere; poche sono le volte in cui cerca di staccarsi
dall’ arida materia trattata. Ma il suo periodo ` e sempre sorvegliato e
corretto, curato, chiaro nel suo significato e a volte anche solenne.

Detailed remarks on Frontinus’ language and style will be found
throughout the commentary. Here one may examine a few se-
lective features which more extensively inform his writing in the
De Aquaeductu.
Lexicon of water quality
Attention to Frontinus’ lexicon has been directed for the most
part to his use of technical vocabulary, but in this work the author
has occasion to speak of water quality and we can observe that
his usage overlaps with both specialist writers and with poets.
On water quality in general he notes how it is an important
criterion of distribution for different categories of use (q. secun-
dum suam quaeque qualitatem, in contrast to Marcia, reserved potui
tota). (One can compare Pliny, HN xxxi.¡., who speaks of the
preference given to Virgo for its tactus and to Marcia for its haus-
tus.) When appropriate he makes mention of poor quality, as of
Alsietina (::.: nullius gratiae, parum salubrem) or of the Anio River
(:¸.: limosus et turbulentus, q¸.¡ deformis ac turbidae). Special boni-
tas characterises the water of Marcia’s Fons Augustae (:..:) and
Claudia’s Fons Albudinus (:¡..). Both abundance and reliability
are singled out for notice. Claudia, drawing fromfontes amplissimi
(:¡.:) is abundantior aliis; the river water of Anio Novus is in primis
abundans (q:..); Agrippa’s engineers found Virgo to have ingentem
aquae modum(:o.¸); even without illicit supplement, Tepula’s flow
was maintained quamvis notabili siccitate (q.8).
Urban salubritas is one of the benefits of a good water sup-
ply (:, 88.:, :::; cf. ¡.. fontium salubritas). (This characteristic is

Panimolle (:q68) :¡–:¸.

stressed by both Vitruvius, \iii.¡.., 6.:¸ and Pliny, HN xxxi.¸¸,
¸q.) Alsietina is parum salubris while others are salubriores (::.:),
and Anio Vetus ranks low in quality because it is minus salubris
(q.). Good water has gratia (:¸.¸) while poor is described as nul-
lius gratiae (::.:). Gratior aquarum sinceritas (qo..) is appreciated, and
the supply from Marcia is called gratissima (q:.¸).
Purity is of course a desirable quality. The superlative puris-
simus is used to describe the Rivus Herculaneus (:¸.¸) as well as
the lake at Subiaco (qo.:). (Compare Horace, Epist. i.:o.:q pu-
rior in vicis aqua.) Purity of Claudia is indicated by sincerus (:¸.6),
and the tam felix proprietas projected by means of improvements
to Anio Novus will make it sinceriorem et iucundiorem (8q.:). Water
taken from the Anio River below the dam at Subiaco is mi-
nus limpida (qo.:), while that purified in the lake is limpidissima
Appearance and coldness are two recurrent features that de-
note excellence. Marcia pleases et rigore et splendore (q:.¸), and
water in the high reaches of the Anio is frigidissimus simul ac splen-
didissimus (q¸.¸n). We find splendor applied to the spring water of
the Rivus Herculaneus (:¸.¸) as well as to Marcia and the nearby
spring of Claudia (8q.¡, q:.¸). (Compare Lucretius, i\..:: splen-
dor aquai, Horace, C. iii.:¸.: splendidior vitro, Silius, Pun. \ii.:¡¸
aquae splendor.) Claudia’s springs are speciosi and one of them is
named Caerulus for its colour (:¡.:); water at Marcia’s source is
green (¸.¸).
Formulaic presentation
In a certain sense, Frontinus assists his reader by adopting a
strictly formulaic pattern when he deals with repetitive mate-
rial, the kind for which a modern writer would abandon con-
nected prose altogether in favour of a more visually accessible
tabular format. (Our author realises that data presented in such
a way, while useful for reference, can strike a reader as inordi-
nately dull: ¸¸.¸–¸ cuius comprehensionem scio non ieiunam tantum sed
etiam perplexam videri posse . . . iis quibus sufficiet cognovisse summam

licebit transire leviora.) The most immediately noticeable examples
of Frontinus’ ‘tabular’ style are his list of pipe-sizes (chapters
¸q–6¸), data on distribution from individual aqueducts (¸8–86),
and the roster of curatores aquarum (:o.), all of which are printed
distinctively in editions since Krohn’s.

Almost equally formal
are chapters 6¸–¸¸, in which he discusses the available sup-
ply of each aqueduct and accounts for discrepancies between
the imperial records and his own measurements. Even chapters
¸–:¸, in which Frontinus proceeds chronologically, are as much
statistical as historical.

For each aqueduct, from the earliest
to the most recent, he systematically gives the location of the
source and then follows with data on the length and type of
Chapter ¸ can serve to illustrate Frontinus’ virtually formulaic
procedure. In §¡ he locates the starting-point of this aqueduct: (a)
concipitur Appia (b) in agro Lucullano (c) via Praenestina (d) inter miliari-
um septimum et octavum (e) deverticulo sinistrosus passuum septingentorum
(a) The sentence usually begins with concipitur + name of the
aqueduct (¸.¡, 6.¸, ¸.6, :o.¸, and cf. ::.¸ concipitur ex lacu
Alsietino); for two others we find the name before the verb
(8.., :¡.:); and for the last the verb is delayed until after
part (c) and then given a different prefix (:¸.: Anio novus . . .
excipitur ex flumine).
(b) This part does not invariably appear (only here, 6.¸, and
::.¸; but cf. :o.¸ palustribus locis, :¸.: excipitur ex flumine, :¸.¡
trans flumen viamque).
(c) In the case of Aqua Marcia (¸.6) there are directions along
two different roads: the Via Valeria, present when the water
had been brought (:¡¡ ncr), and the Via Sublacensis, built
under Nero.

Frontinus interjects an occasional comment in addition to essential data
(e.g. :o..¡), and sometimes he affects a pointed use of variatio (8¸.:, 8¡.:,
contrasted with 8o.:, 8:.:, etc.).

Evans (:qq¡) ¸6.
Chapters ¸.¡–¸, 6.¸–6, ¸.6–8, 8.., q.¸, :o.¸–¸, ::.¸–¡, :¡.:–¡, :¸.:–6.

(d) The form here (inter miliarium + ordinals) is unique. Fronti-
nus’ commonest construction is ad miliarium + ordinal (¸.6,
8.., :o.¸, :¡.:, :¸.:), but there are two instances of the abla-
tive miliario (6.¸, ::.¸).
(e) A side-road leading from highway to source is specified ei-
ther to the right (¸.¡, ¸.6, 8.., ::.¸) or to the left (¸.6, :¡.:):
ablative deverticulo +number of passus in the genitive. In two
instances the transmitted text notes that the directional ref-
erences are for persons travelling from Rome (¸.6, 8..). For
Anio Vetus, Virgo and Anio Novus there is no deverticulum
(6.¸, :o.¸, :¸.:).
In ¸.¸ we follow Frontinus’ pattern of indicating first the length
of the conduit as a whole and then the lengths of its various parts
broken down by type of construction. (a) ductus eius habet longi-
tudinem (b) a capite usque ad Salinas, qui locus est ad portam Trigeminam,
(c) passuum undecim milium centum nonaginta. (d) <ex eo (e) rivus est>
sub<t>er<raneus pas>suumundecimmiliumcentumtriginta, (f) supra ter-
ram (g) substructio et arcuatura (h) proximum portam Capenam passuum
(a) ductus eius habet longitudinem (¸.¸, 6.6, ¸.8, :o.¸, and cf. :¡.¡
Claudiae ductus habet longitudinem), ductus eius efficit longitudinem
(::.¡, and cf. q.¸ with Iuliae instead of pronoun), venit per
longitudinem (:o.¸). Slightly different is :¸.6 ductus Anionis novi
efficit (cf. :o.8 rivi subterranei efficiunt, :..¸ ductus . . . efficit).
(b) a capite ad Salinas . . . (¸.¸), a capite ad urbem(¸.8), anexplanatory
phrase (ita exigente libramento 6.6), otherwise omitted(q.¸, :o.¸,
::.¡, :¡.¡, :¸.6).
(c) The construction habet longitudinem is followed with passuum
(genitive) + number, that of effecit with passus (accusative).
(d) At this point the phrase ex eo is transmitted at 6.6, :o.¸,
:¡.¡, :¸.6 (cf. ¸.¸ <ex eo>). In two cases it is followed by
nominatives (¸.¸, 6.6), in three by ablatives (:o.¸, :¡.¡, :¸.6).
The ablative is used without ex eo in two instances (¸.8, q.¸),
possibly also a third (::.¡).
(e) rivus subterraneus (6.6, restored at ¸.¸), otherwise ablative (¸.8,
:o.¸,:¡.¡, :¸.6, supplied q.¸, ::.¡).
(f) supra terram substructio (¸.¸, cf. substructio supra terram 6.¸); opere
supra terram (¸.8, q.¸, :¡.¡, :¸.6), all followed by passuum +
number; supra terram per passus . . . :o.¸; omitted ::.¡.
(g) Another ex eo appears at this point inq.¸, :o.¸, :¡.¡, :¸.6 (also
¸.8 [ex] eo), to introduce subcategories of construction above
ground. ¸.¸ and 6.6 are anomalous, both because of nomi-
natives (substructio, arcuatura) and because there is less detail.
Elsewhere the ablative is consistent: substructione (¸.8, q.¸),
substructione rivorum (:o.¸, :¡.¡, :¸.6), substructionibus (:¸.6);
opere arcuato (¸.8, q.¸, :o.¸, ::.¡, :¡.¡, :¸.6). All are followed
by passuum except :¡.¡, :¸.6 per passus.
(h) In three chapters there is a further breakdown between rural
and suburban sections: longius ab urbe pluribus locis per vallis
(¸.8), superiori parte (:¡.¡, :¸.6); propius urbem a septimo miliario
(¸.8, :¡.¡, :¸.6); cf. also q.¸ proximis urbi locis a septimo miliario,
:o.¸ locis compluribus.
Rhetorical style
No reader of the De Aqueductu has ever failed to be struck by
the enthusiastic outburst in chapter :6: Tot aquarum tam multis
necessariis molibus pyramidas videlicet otiosas compares aut cetera inertia
sed fama celebrata opera Graecorum.

There is no lack of rhetoric
in Frontinus’ prologue (:–.), where such might in any case be
But there are throughout the work sections written,
if not carefully, then at least in a manner that reveal an author
whose education was worthy of a Roman senator. By way of
example, we can look to a passage where Frontinus is intending
to write persuasively, but not necessarily with literary elegance.

De Laine (:qq¸) :.¡, for instance, draws attention to ‘the rhetorical nature
of the passage and its supercilious tone’.
Grimal xvi speaks of ‘le ton solennel de l’introduction’; see further Santini
(:qq.), Del Chicca (:qq¸–¸).

–. Non dubito aliquos admiraturos quod longe maior
copia actis mensuris inventa sit quamerat in commentariis prin-
cipum. cuius rei causa est error eorum qui ab initio parum
diligenter uniuscuiusque fecerunt aestimationem. ac ne metu
aestatis aut siccitatumin tantuma veritate eos recessisse credam,
obstat id quod ips[e actis] mensuris Iulio mense hanc unius-
cuiusque copiamquae supra scripta est tota deinceps aestate du-
rantemexploravi. quaecumque tamen est causa quae praecedit,
illudutique detegitur decemmilia quinariarumintercidisse, dum
beneficia sua principes secundum modum <in> commentariis
adscriptum temperant. sequens diversitas est quod alius modus
concipitur ad capita, alius nec exiguo minor in piscinis, minimus
deinde distributione continetur. cuius rei causa est fraus aquario-
rum, quos aquas ex ductibus publicis inprivatorumusus derivare
Frontinus begins with rhetorical understatement,
his introduction to this portion of the work (6¡.¡ ante omnia itaque
capita ductuum metiri adgressus sum, sed longe, id est circiter quinariis
decem milibus, ampliorem quam in commentariis modum inveni, ut per
singulas demonstrabo). A reason for the discrepancy he states as a
matter of fact: cuius rei causa est, predicated by abstract noun with
subjective genitive (error eorum), the pronoun defined in a relative
clause. Word-order in the relative clause builds a crescendo of
indignation: prepositional phrase functioning as temporal ad-
verb (ab initio), modified modal adverb (parum diligenter), pronoun
in genitive (uniuscuiusque with ellipse of aquae), verb and emphatic
direct object (fecerunt aestimationem).
The next sentence starts with his purposeful dismissal of what
might have been an excuse for his forerunners, and ends with
confident self-satisfaction. Both the initial clause (ne . . . credam
+ indirect statement) and the sentence as a whole end with
first-person verbs, the latter’s present perfect emphasised by its
With the litotes non dubito, cf. also in this passage parum diligenter, nec exiguo
collocation with present participle (durantem exploravi). One al-
most takes the main sentence to be ipse . . . exploravi, nearly every
feature of which recalls by contrast the preceding sentence. First
are the repetitions (actis mensuris, uniuscuiusque). Next, the two
temporal ablatives (Iulio mense, tota . . . aestate) are more specific
than ab initio, and they intertwine with copiam . . . durantem; note
also the hyperbaton and alliteration in tota deinceps aestate duran-
Finally, hanc . . . quae supra scripta est points to demonstrable
and thereby reliable data in the preceding nine chapters. Yet the
whole sentence is built around an entirely prosaic obstat id quod.
An unexciting prosaic style informs the whole of the following
sentence (quaecumque . . . temperant), the only striking feature of
which is the emphatic decem milia quinariarum and a tiresome rec-
ollection of the discrepancy with imperial records that Frontinus
himself has detected and demonstrated.
But a second set of numerical differences allows Frontinus
now to sound one of his favourite notes, the fraus aquariorum.
The facts are stated with an elaborate tricolon, replete with
anaphora and ellipsis (alius modus . . . alius . . .), alliteration (con-
cipitur ad capita . . . continetur, minor . . . minimus, deinde distributione)
and variation (ad capita, in piscinis, distributione). Again, the declar-
ative statement of fact as we saw above: cuius rei causa est, now
followed by a noun with far stronger negative force and with a
more immediate subjective genitive. The relative clause nowhas
as its subject the writer himself, and the tense authoritatively is
the present.
Note the chiastic word-order of the prepositional
phrases ex ductibus publicis in privatorum usus, and the alliterative
finale derivare deprehendimus.
Recall the hendiadys aestatis ac siccitatum near the beginning of this sentence.
It may not be coincidental that Frontinus speaks in q.8 of notabilis siccitas in
a context of his personal monitoring of water supplies (the verb there also
in the present perfect).
Cf. ab eo quod .¸.¸, ex eo quod 6¸.¸, 6¸.q, ::¸.¡.
It matters little whether deprehendimus is present or (more likely) present
The Middle Ages
Virtually nothing is known of the fate of Frontinus’ commenta-
rius from the time of its publication until it was discovered in
the fifteenth century. From its mere survival we can surmise
that it attracted some attention in late Antiquity, and one can
guess at reasons: a lasting prestige which attached to the au-
thor; a tone that might have appealed to imperial idealists (or
to bureaucrats); a subject-matter which never completely lost its
relevance, given the practical necessity of maintaining an essen-
tial service for Rome. The title De Aquaeductu possibly dates to
this same period, and there are tantalising hints that it might
even have been familiar to an administrative audience.

Yet the
booklet can never have achieved widespread circulation, and its
very existence must always have been precarious.

Codex Casinensis ¸6: (C), our oldest manuscript of this text,
was written at Monte Cassino about the year ::¸o. The copyist,
as it happens, was no ordinary scribe: he was Petrus Diaconus,
an enigmatic but remarkable monk who left his erratic footsteps
indelibly impressed on the history of that venerable abbey.

Peter the Deacon’s interest in Frontinus did not end with simple
transcription. Into his Chronica consulum, dictatorum et imperatorum,
a curious but very important compilation made a fewyears later,
Peter inserted references to four of Rome’s earliest aqueducts.

For the title, see commentary. DeLaine (:qq¸) :¸:, :¡¸ points to possible
verbal echoes of this booklet and something of its flavour in the Formula
comitivae formarum Vrbis of Cassiodorus (Var. \ii.6), but she acknowledges that
they may likely be no more than coincidental. For another possible link to
late Antiquity, but still more tenuous, see ¸.¸n. arcuatura.

By contrast, Frontinus’ Strategemata was far better known in the Middle Ages:
see Reynolds (:q8¸) :¸:–.; cf. J. Martin (:q¸¸).

Both manuscript and scribe are discussed at greater length below (pp. ¸¸–
Codex Casinensis .¸¸, p.q (ed. Florilegium Casinense ¸.: (:8q¡), p.¸8). The
special interest which attaches to these entries was first revealed by Bloch
(:q8¡) esp. ¸o–: and Plate :.
He can have drawn these references only from a text of the De
Aquaeductu, presumably from the copy he had himself exscribed.
The Monte Cassino manuscript, as it happens, is nowthe surviv-
ing archetype, and Peter’s references constitute the only known
use of this text by any writer in the Middle Ages. The next person
to reveal an acquaintance with this work of Frontinus was the
enthusiastic book-hunter Poggio Bracciolini (:¸8o–:¡¸q),

Poggio himself deserves full credit for its rediscovery – at Monte
Cassino and in the very copy made there three centuries earlier
by Peter the Deacon.
Poggio’s quest
By spring of :¡.¸ Poggio had heard – we do not know how –
of the possibility that a Frontinus manuscript was to be found
at Monte Cassino.
He wrote to his friend Niccol ` o Niccoli
on :¡ June :¡.¸ (Epist. ii..6), ‘Nudius tertius locutus sum cum
administratore monasterii Cassinensis satis diligenter de Iulio
Pollicitus est se missurummihi librum, cumprimum
redierit, dummodo reperiatur; nam multos deperditos paucis
ante annis dicit. Petiit a me titulum libri, tradam ei ante re-
cessum suum et confido nos habituros librum.’ In subsequent
letters to Niccol ` o during the summer of :¡.¸ Poggio mentions

DBI (:q¸:) xiii: 6¡o–6.
We rely on Poggio’s letters to Niccol ` o for tracing this story. For convenience
numbered references are to Tonelli’s edition (:8¸.); the letters have been
newly edited by Helene Harth (:q8¡).
‘Fronto’ for ‘Frontinus’ is to be explained in part by ignorance (the work had
as yet not been seen), but this name occurs in the tradition of the Strategemata
(see Gundermann’s preface, xii) – andthere is evenanepigraphic attestation:
CIL q.6o8¸ Sex. Iuli Frontoni. Poggio uses ‘Frontinus’ for the first time in
Epist. ii.¸¡ (Nov. :¡.¸), not perhaps coincidentally just after receiving a
report of the Hersfeld manuscript (see below). Panormita’s ‘Iulius Fronto’,
occurring some months later, seems to reflect adherence to the form that
had gained initial currency. As late as :¡¸. Traversari writes (Epist. \iii.¡¸)
‘Frontonem de aquaeductibus’, but corrects himself on the spot: ‘Est tamen
id opus non Frontonis, ut putavimus, sed Frontini.’
a delay in acquiring the text,
and finally on ¸ November he
reports that the search has been unsuccessful (Epist. ii.¸¡): ‘Iulius
Frontinus non reperitur in monasterio Cassinati, nam rescripsit
nobis ille, cui curam demandaramus, se diu quesisse librum, sed
minime inveniri.’ But, he continues in high spirits, ‘Hec autem
minima est iactura, nam aliunde expiscabimur.’
The newhope arose because Poggio had been approached by
‘quidam monachus amicus meus ex quodam monasterio Ger-
maniae’. The monk was Heinrich von Grebenstein, present in
Rome on one of several official visits on behalf of the abbey of
but with a sideline interest in trading books. Poggio
lost no time in sharing with Niccol ` o two highlights on the list
he had received from the monk: ‘inter ea volumina est Iulius
Frontinus et aliqua opera Cornelii Taciti nobis ignota’.
Contacts with the Hersfeld monk were protracted for more
than three years. In the fall of :¡.6 Poggio asked for an ‘inven-
tarium cuiusdam vetustissimi monasterii in Germania, ubi est
ingens copia librorum’ (Epist. iii.:), and he seems to have re-
ceived it in the spring of :¡.¸ when Heinrich von Grebenstein
was once again in Rome. On :¸ May Poggio wrote that the new
inventory was a disappointment (Epist. iii.:.): he could do no
better for Niccol ` o than to send him ‘partem inventarii sui, in
quo describitur volumen illud Cornelii Taciti, et aliorum quibus
Hopes of acquiring the treasures from Hersfeld remained un-
realised: in February :¡.q the monk again arrived in Rome
Epist. ii..¸ (.¸ June): ‘Si Iulius Fronto veniet, qui procul dubio, nisi perditus
est, veniet, hec per proprios tabellarios deferentur ad Poggium.’ Epist. ii..q
(:8 August): ‘De Monte Cassino, hoc est Iulio Frontone sollicitus sum, sed
mirum est; tam pauci eo accedunt aut inde ad nos veniunt.’ Epist. ii.¸. (.q
September): ‘Iulium Frontonem aliquando eruemus ex agro illo Cassinati,
sed durumest impellere istos nostros barbaros, ut aliquid sit eis dulce preter
That the ‘German monastery’ was Hersfeld emerges from other letters:
Epist. iii.:. (:¸ May :¡.¸) and iii..q (.6 Feb. :¡.q). For the monk’s identity
see Pralle (:q¸.) .¡.
The existence of these same works is reported by Panormita to Guarino in
a letter which probably dates from April :¡.6 (see below).
absque libro (Epist. iii..q),
and we hear fromPoggio of no further
contacts. Not long afterwards, however, he had an opportunity
to visit Monte Cassino. This time he could search for himself,
and on q July he wrote from Anagni that he had found the text
of Frontinus (Epist. iii.¸¸): ‘Vidi autem bibliothecam monasterii,
repperique librum, in quo erat Iulius Frontinus De aqueductu
urbis et item Firmici Matheseos libri \iii . . . Portavi volumen
hoc mecum, ut transcribamlibellumFrontini, cumsit mendosus
et pessimis litteris adeo ut vix queam legere.’
Poggio was less than punctual about returning the borrowed
codex to Monte Cassino. In December :¡.q he writes to Niccol ` o
that he is about to return the book: he has copied Frontinus and
in its other contents he has no interest.
But Niccol ` o seems to
have pleaded for a delay,
and a letter of Ambrogio Traversari
indicates that the book might still have been in Rome as late as
April of :¡¸..
The manuscript of Frontinus Poggio discovered was the copy
made by Peter the Deacon, and the volume he borrowed was
none other than Codex Casinensis ¸6: (although not exactly in
its present state). Poggio mentions that the manuscript contained
Although there is still a promise that the book will come later. Note
the singular liber: perhaps this means ‘with no book at all’, although
Poggio and Niccol ` o might have been mainly interested in the Hersfeld
Cf. Epist. iii.¸8 (:¸ July): ‘Scripsi item noviter ex Anania de itione mea ad
Sanctum Germanum et de libello Frontini.’
Epist. i\.. ‘Liber Montis Cassini repetitur a me; itaque remittam eum.
Transcripsi enim, ut nosti, De aqueductibus, quod mihi cure erat. Reliqua
non magnopere me delectant. Illis ergo equo animo carebo.’ For the other
contents at this time, see Bloch (:q8¡) :o¸–:6.
Cf. Epist. i\.¡ (.¸ Dec. :¡.q). Harth (above, n.q8), no.¸8, p.:o¸, dates this
letter to May :¡¸o.
Epist. \iii.¡., quoted by Sabbadini (:8qq) :¸¸: ‘Poggius Frontonem de
aquaeductibus secum habet; eum pollicitus est mittere ad me, sed non-
dum promissioni suae satisfecit.’ Epist. \iii.¡¸: ‘Poggius . . . Frontonem de
aquaeductibus secum fortassis adtulit. Est tamen . . . non Frontonis . . .
sed Frontini, ut in exemplari antiquissimo, quod secum adtulisse debuit,
notavi.’ Observe that Traversari had seen the venerable antique only on a
prior occasion; secum habet may not have reflected present reality.
both Frontinus and Firmicus, and we know that part at least of
Firmicus was once to be found in Casinensis ¸6:.
casual reference confirms beyond doubt: ‘de Frontino et frag-
mento Arati’ in Poggio’s letter of December :¡.q (Epist. i\.¡)
is no accident, for the opening words of the codex were once
verses beginning ‘Ab Iove . . .’ – from Germanicus’ translation
of Aratus, the initial line of which read in full ‘Ab Iove principium
magno deduxit Aratus.’
Niccol ` o and Traversari can hardly have been the only friends
who expressed an interest in the newly discovered Frontinus.
Copies could now circulate amongst humanist scholars, and the
work could be read by those who studied and loved the mon-
uments of ancient Rome. Appropriately enough, it was Poggio
himself who first used Frontinus as an historical source, and
he credits himself with rediscovering the text: ‘quem libellum
ipse paulo ante reperi absconsum abditumque in Monasterio
The Codex Hersfeldensis
Poggio, as we have seen, met with no success in his attempt to
acquire manuscripts from Hersfeld. Two authors in particular
had aroused his interest: Frontinus and Tacitus. The Hersfeld
codex which contained the minor works of Tacitus and Sueto-
nius’ De grammaticis et rhetoribus was eventually brought to Italy,
and a part of that volume survives to this day.
On the other
It appears in an inventory made in the fifteenth century (see below, n. :..):
see Bloch (:q8¡) :o8, ::.–:¸; cf. Rinaldi (.oo.) ¸8–¸¸.
For the date, see note :o6 above.
Bloch (:q8¡) ::..
Poggio writes of Rome’s aqueducts in the first book of his De varietate fortunae,
a work begun as early as :¡¸:. An annotated text of the relevant portion is
found below, in Appendix A.
Brought to Italy by Enoch of Ascoli, and seen at Rome in :¡¸¸ by Pier
Candido Decembrio. Only part of this ninth-century manuscript survives
(that containing the Agricola) in Rome, Bibl. Naz. Vitt. Em. :6¸: (until
recently Codex Aesinas lat. 8). For rehearsal of the evidence see Robinson
hand, there is no evidence that the Hersfeld Frontinus ever
reached humanist circles, nor has it come to light in modern
What we know of the Hersfeldensis is thus limited entirely to
information Poggio received from the Hersfeld monk Heinrich
von Grebenstein. The first of his reports was that which Poggio
mentioned to Niccol ` o in November :¡.¸.
To this same first
report we no doubt owe the more detailed description related by
Panormita (Antonio Beccadelli): ‘QuinetiamSex. Iulii Frontonis
liber de aquaeductibus qui in urbem Romam inducuntur; et est
litteris aureis transcriptus. Item eiusdem Frontonis liber alter,
qui in hunc modum iniciatur: “Cum omnis res ab imperatore
delegata mentionem exigat” et cætera.’
Here we have a work
clearly identical to that which survives in the Casinensis, divided
into two books (as it is also in C), as well as an apparent title (‘De
aquaeductibus qui in urbem Romam inducuntur’). The ‘golden
letters’ are a mystery.
The Hersfeld monk supplied a second inventariumin the spring
of :¡.¸, a portion of which Poggio sent on to Niccol ` o. In :¡¸:
Niccol ` o drew up a list of desiderata to be sought in five German
and Danish monasteries. The Hersfeld data in this commentarium
almost certainly were derived from the select inventory Niccol ` o
had received from Poggio in May of :¡.¸.
(:q¸¸) :–.o, Murgia (:q8¡); for a close look at the chronology, Stok (:q8¸).
A comprehensive review of the tradition of Tacitus’ minor works is that of
R¨ omer (:qq:) .¸..–¸o.
See below ¸¸. Of a Hersfeld Ammianus there survives only a fragment,
now in the Landesbibliothek at Kassel.
Epist. ii.¸¡ ‘et nomina librorum mittit interclusa’.
Sabbadini (:8qq) :.¸, without date, but conjectured to have been written
in April :¡.6.
This bit of the description may possibly reflect a touch of salesmanship on
the part of the monk. A similar motive might conceivably explain the fact
that two libri are listed separately and in reverse order.
Epist. iii.:.. Niccol ` o’s commentarium is edited by Robinson (:q.:). Variant
readings (cited as J) are taken from an independent witness to the monk’s
inventory, a letter of Poggio’s son written c. :¡¸¸: Rubinstein (:q¸8). At
In Monasterio hispildensi* haud procul ab alpibus continentur haec
opuscula. videlicet.
Repertus.* Julii Frontini* De aquae ductis* quae* in urbem inducunt*
liber .j. Incipit sic. PERSECVTVS ea quae de modulis dici fuit nec-
essarium. Nunc ponam quemadmodum queque aqua ut principium*
commentariis comprehensumest usque ad nostramcuramhabere visa
sit &c. Continet hic liber XIIj.*
Repertus.* Item eiusdem frontini* liber incipit sic. Cum omnis res ab
imperatore delegata interiorem* exigat & curam, & me seu naturalis
solicitudoseufides sedula, nonaddiligentiammodo, verumadmorem*
commisse rei instigent, sitque mihi nunc ab nerva augusto, nescio dili-
gentiore an amantiore rei .p. imperatore aquarum iniunctum officium
& ad usum &c. Continet. XI. folia.
hisfildensi J om. J Frontoni J aqueductibus J qui J Indu-
cuntur J principum J folia add. J om. J om. J Intentionem
J amorem J
If this is any indication, the second inventory was indeed
‘plenum verbis’. For each liber there is an extensive incipit and an
indication of the number of folia it occupied in the manuscript.
The ‘first’ book can now be identified with what is called Book
. in C (beginning with chapter 6¡); and although there is no ex-
plicit, it is unlikely that the Hersfeld codex contained more than
C (which ends with chapter :¸o). The text, one guesses, closely
resembled that found in C (even more closely, no doubt, that in
the exemplar from which C was copied): note 6¡.: quem modum]
quemadmodum HC.
Aclose relationship between the Hersfeld manuscript and the
text of Frontinus preserved at Monte Cassino should come as
no surprise: copies of the work, after all, were extremely rare.
We know of connections that existed between Monte Cassino
and Germany, especially from the days of Abbot Richer of
Niederaltaich (:o¸8–¸¸). Although we can never recover the
the end of his letter Jacopo mentions that the Frontinus has been found,
although he is wrong (perhaps only through carelessness) in putting the
discovery ‘post patris mortem’ (i.e. after :¡¸q).
particular details of the transmission of Frontinus, analogy with
the Tacitus traditions
makes it very tempting to imagine that
the text came toMonte CassinofromnorthernEurope, probably
in the eleventh century.
The Hersfeldensis of Frontinus does not survive, nor was its
existence ever reported by anyone other than the Hersfeld monk
who was in contact with Poggio. Silence does not of course pre-
clude the possibility that the codex was copied or that readings
from it were used to correct the blemished text which Poggio
had discovered at Monte Cassino. Careful study of the recentiores,
however, reveals that these manuscripts are all descended from
the Casinensis. Unless the Hersfeld manuscript was itself a copy
of C, the lost codex cannot have left any progeny.
The Codex Casinensis and Peter the Deacon
of Monte Cassino
Monte Cassino, Archivio della Badia, cod. ¸6:, ff. ..
¡¸–6¸), s. xii
./¡ :.o
Our sole authority for the text of De Aquaeductu is the copy made
by Petrus Diaconus of Monte Cassino about the year ::¸o.
This was the Frontinus discovered and borrowed by Poggio in
Lowe (:q.q); cf. Bloch (:q¸:) ¸8:–.. A text of the Agricola (recovered from
Hersfeld: above, n. ::.) had also been available at Monte Cassino, where it
was used by Peter the Deacon in the twelfth century: Bloch (:q¡:); see also
below, n. :¸:.
It is reasonable to suppose that the exemplar from which Cwas copied was
a Carolingian manuscript. The most economical surmise would be that
this single manuscript was parent to both H(a copy made before the book
was taken to Italy) and C (copy made at Monte Cassino).
For a detailed catalogue description, see Inguanez (:q¸¡) .o8–:.. Al-
though Inguanez provides foliation, it is worth noting that manuscripts
at Monte Cassino are usually cited by page. The Latin spelling of the ad-
jective is Casinensis (one s), consistently used at the abbey throughout the
Middle Ages: editors of Frontinus unfortunately perpetuate Poggio’s form
Some portions may have been penned by other hands, and my own sense
is that the text was copied before ::¸o: see below ¡:–..
:¡.q. The manuscript appears ina catalogue of the abbey library
made between :¡6¸ and :¡¸:.
Parts of the book were copied
not long afterwards, and it might perhaps have been dismantled
at this time.
In any case, several works which it then contained
are wanting in the present volume.
Mabillon saw the codex
and copied the Frontinus on a visit to Monte Cassino in Novem-
ber :68¸.
Don Erasmo Gattola, the learned and reverend
scholar who assisted Mabillon, was later to furnish a second
copy to Poleni, whose milestone edition appeared in :¸...
excellent facsimile of the Frontinus was published by Clemens
Herschel in :8qq; a second, prepared by Don Mauro Inguanez,
was issued at Monte Cassino on the occasion of the abbey’s
fourteenth centenary (which coincided with the five hundredth
anniversary of Poggio’s discovery).
Vat. lat. ¸q6:, f. :¡
; ed. Inguanez (:q¡:) ¸8.
Naples, Bibl. Naz. cod. i\ n .. bis contains the work of three scribes, two of
whom (ff. :–6., 6¸–8¸) were copying from Casinensis ¸6:. The third (Ar-
naldus de Steccatis de Bruxella), whose work is presumably contemporary,
entered dates: :¡¸¡ (f.:oq
) and :¡8¡ (f.::¡).
Bloch (:q8¡) :o¸–:6, has recovered much of the original contents by care-
ful comparison of the library inventory (Vat. lat. ¸q6:), the present Cod.
Cas. ¸6:, the Naples manuscript, and Peter the Deacon’s autobiographical
Mabillon (:68¸) :.¸: ‘quem [sc. codicem Frontini], quia deficiente edito
conferre nobis non licuit, integrum descripsimus. Manus ad scribendum,
& se ipsum totum in nostros usus impendit pius & cordatissimus Domnus
Erasmus a Ca¨ıeta’. The fate of Mabillon’s transcript is not known.
OnGattola, animportant figure inhis ownright, see DBI (:qqq) rii: 6¸8–6o.
Poleni writes warmly of his assistance (pref. p. :q), ‘ab . . . Abbate Gattola
exemplum habui Codicis illius Cassinensis descriptum diligentissime. Pro
ea vero, qua praeditus est Abbas ille humanitate, mihi deinde praestitit
summo studio quaecumque sum ab eo sciscitatus’.
Herschel (:8qq, .nd ed. :q:¸); Sexti Julii Frontini De aquaeductu urbis Romae,
editio phototypica ex cod. Cas. , saec. xii (Montecassino :q¸o). Portions of the
Frontinus have appeared elsewhere: e.g., f.¸o
( = p. ¸q) in Steffens (:qoq)
pl. ¸qb; f..6
(p. ¸:) in Krohn’s edition (:q..); f..¸
(p. ¡6) in Valentini–
Zucchetti, vol. i (:q¡o) pl. i; f.¸:
(p. 6.) in Meyvaert (:q¸¸) pl. iia; ff.¸:

(pp. 6:–¸). Chapter-length facsimiles of the text are available online
at rome/Images/
Roman/Texts/Frontinus/De Aquis/ms*/
Since scribes and scholars play so crucial a role in the trans-
mission of an ancient text, it is a matter of special interest that we
happen to be well acquainted with the man responsible for our
archetype of the De Aquaeductu.
Peter the Deacon was born in
Rome (probably in ::o¸) and was presented by his father Egidius
as a puer oblatus to the monastery of Monte Cassino when he
was five years old. This Egidius Tusculanensis was somehow re-
lated to the counts of Tusculum, one of the most illustrious noble
houses of Italy in the tenth and eleventh centuries. Peter empha-
sises his connection with Rome: Egidius was natione Romanus, the
son (he says) of Gregorius ‘Romanorum patricius et consul’.
Peter had the advantage of an outstanding education, and at
his disposal were the rich treasures of the abbey library, among
them the impressive collection of texts copied half a century
earlier under the great Abbot Desiderius (:o¸8–8¸). Radical
changes taking place in the twelfth-century church were sharply
and painfully manifested at Monte Cassino in the ::.os. Sud-
denly the abbey was plunged from its pinnacle of cultural and
political preeminence,
and the young Peter shared in the dis-
grace which befell his monastic home. For a period of some three
and a half years (::.8–¸:) he was banished fromthe abbey. Upon
his return he was entrusted with responsibilities for the library
and archives and he was encouraged in his literary and histori-
cal studies. But bitterness over the exile lingered unmistakably; it
may have riveted his focus forever on bygone glories, the aureum
I would have shortened the discussion that follows had there existed any
general treatment on Peter the Deacon to which users of this edition could
conveniently turn.
Peter left three autobiographies. The two earliest are in his autograph: Cod.
Cas. ¸6:, p. :¡¸, and Cod. Cas. .¸¸, pp. ¸o–:. The third, as it were official,
version is incorporated into the abbey chronicle: Chron. Cas. i\.66, ed. H.
Hoffmann (:q8o), Monumenta Germaniae Historica, Scriptores ¸¡: ¸.q–¸:. For
Peter’s family relationships, see Hoffmann (:q¸:) esp. 6o–¸¸.
For a general introduction to the intellectual and cultural standing of
Monte Cassino, see Bloch (:q¸., :q8.). For more detailed discussion, see
Bloch (:q86) i: :–:¸6. On the events of the ::.os and their affects on
Monte Cassino, see Bloch (:q¸.), revised in his (:q86) ii: q¡¸–:o¡q; see also
Hoffmann (:q¸:) ¸¡–:o¡.
patris Desiderii seculum
and all that it represented – not least of
which was a new kind of fascination with ancient Rome.
The prodigious volume and range of Peter’s writings in the
subsequent decades (his death came sometime after ::¸q) dis-
close a man of wide learning and indefatigable energy, not with-
out a generous share of originality and imagination. That such
signal talents were readily applied to forgery and fabrication
may in large part be excused by the standards of his age. That
he suffered from what might now be diagnosed as psycholog-
ical disorders may explain the exaggerated sense of his own
importance, along with some of the more bizarre contents of his
To personality and temperament one must ascribe as
well the haste and carelessness which are such conspicuous and
consistent features of every product in his literary career. But,
for all his defects, Peter the Deacon must not go unrecognised as
a significant figure in the renaissance of the twelfth century.
Holste seems to have been first to identify the scribe of Cod.
Cas. ¸6:.
The hand which wrote most of this manuscript is
The phrase is from the prologue to the Vita S. Severi, ed. Rodgers (:q¸.)
:¸–:8 (cf. Patr.Lat. :¸¸: :o¸.c), composed c. ::¸¸. Memory of humiliation
unambiguously underlies Peter’s allusion to Tacitus’ Agricola in the same
prologue: Bloch (:q¡:). The opening words of this vita are no less telling:
‘Casinensis igitur arcis sublimitas tanto olimculmine viguit, ut Romani cel-
situdo imperii philosoficis studiis illum in evum dicaret,’ whereupon Peter
devotes an inordinate space to rehearsing the literary accomplishments of
Publius [sic] Marcus Terentius Varro, ‘omnium Romanorum doctissimus,
considerantissimus, acutissimus, et diligentissimus’ (copied verbatim from
Augustine, De civitate Dei \i..–¸). Peter’s fascination with this particular
ancient author is plainly remarkable. He knew at first hand only the De
lingua Latina (of which the unique manuscript Mediceus ¸:.:o is of Cassi-
nese origin), and in fact copied a fragment of this text into Cod. Cas. ¸6:
alongside Frontinus and Vegetius. Peter had found and pitched upon the
perfect model for his own scholarly ambitions in Varro the Roman, Varro
the polymath, Varro ‘Casinensis’ (whose villa at Casinum was known from
Cic. Phil. ii.¡o–:, available to monastic readers at Monte Cassino in what
is now Cod. Vat. Lat. ¸..¸).
Bloch (:q8¡) ::¸–.6.
Bloch (:q8¡) esp. 66–8¸, :¡:–¸q; more generally, Bloch (:q8.).
Barb. lat. :.:, f. ¡:
: ‘ut ex opusculis adiunctis coniicio Petri Diaconi manu
scriptus est’ (cf. Barb. lat. :¸:, f.:).
also found in Cod. Cas. .¸¸, and the contents of both books are
primarily works compiled by Peter the Deacon. That these texts
were in fact Peter’s autographs was suggested by Wattenbach,
and definitively established by Meyvaert.
The script is best
described by those most familiar with Peter the Deacon. Caspar
speaks of ‘eine kleine, breite, in der Linienf ¨ uhrung nicht ganz
sichere, etwas hastige und unsorgf¨ altige Minuskel’.
stresses that Peter’s is an untrained hand, more that of a scholar
than a calligrapher.
By comparison with other specimens of Peter’s handwriting
one sees clearly that this transcription of Frontinus belongs to his
early years. The hand is unpractised and the overall impression
is decidedly juvenile. There are experimental touches, some of
them playful and whimsical. In the general unevenness, more-
over, there is something besides the variety to be expected from
normal interruptions in the copying process: one has the feeling
that this scribe alternates between care and boredom, patience
and haste.
Here we should perhaps call attention to those places on fo-
lio ¸: (pages 6:–.) where there is a marked difference in the
writing. Meyvaert saw here three additional hands: two he de-
scribed as those of scribes who had been trained in the Ben-
eventan script, while the third employed Beneventan features
not normally found in Peter’s writing.
The use of amanuenses
is attested elsewhere in Peter’s work: on rare occasions when he
was working from a preexisting draft, he would call upon others
Wattenbach (:8¡6), ¸q¡ n.8; cf. Bethmann (:8¸¡) ¸o., ¸o6.
Meyvaert (:q¸¸). Of equal importance is Meyvaert’s demonstration that
Peter wrote only Caroline minuscule. He never mastered the more elegant
and prestigious Beneventan, although the latter might fairly be called the
‘official’ script of Monte Cassino; cf. Lowe (:q:¡) 8¡–q..
Caspar (:qoq) :q. Caspar’s monograph is fundamental to all subsequent
studies concerned with Peter the Deacon.
Perceptively remarked, solely on palaeographical grounds, by Steffens
(:qoq) pl. ¸qb.
Meyvaert (:q¸¸) :.¸–6: f.¸:
(p.6:), lines :–8 quinquaginta . . . curatoribus;
(p.6.), lines :–¸ fistule . . . ducit; lines ¸o–¸ opera . . . reditu.
to copy a few lines here and there.
This might have happened
in the Frontinus, although it is my belief that Peter himself wrote
these lines.
Apart from the handwriting, there is another indication that
this transcription is a product of Peter’s youth. Bloch’s recon-
struction of Codex ¸6: in its earlier state allows us to see the
copy of Frontinus in a new context – that of a series of excerpts
and compilations, many of which were concerned with Roman
This entire portion of the original codex conveys
the impression of a compendium gathered and preserved for
largely personal reasons. Texts which Peter had copied or ex-
cerpted were often those to which he was later to turn in creating
his own literary works. Taking these circumstances into consid-
eration (along with the palaeographical evidence), I strongly
incline to date Peter’s copy of Frontinus to the years when he
had graduated from schoolboy to young scholar, in other words
to the period prior to his exile – perhaps as early as the mid
What attracted Peter to Frontinus can only be surmised. His
fascination with ancient Rome would no doubt have given this
Examples from Cod. Cas. ¸6:: p.:¸¸ (Vita S. Apollinaris); p.:q6 (Vita S. Gebi-
zonis); p..o¸ (Vita S. Aldemarii). The first two vitae, which Peter had originally
written as separate works, were being incorporated into his Ortus et vita ius-
torum cenobii Casinensis. The Vita S. Aldemarii (not by Peter) was subsequently
appended to the same work.
In overall quality they resemble nothing so much as Peter’s own uneven
and careless script. The view that this is Peter’s own script I advance with
some hesitation and only because my initial feeling has gained considerable
strength over several years of close work with Peter’s handwriting. The issue
requires minute reexamination of Cod. Cas. ¸6: (including portions not
available in facsimile). Although it has no bearing on the present edition,
the matter is of some interest for an understanding of Peter’s workmanship
and it might serve to throw welcome light on the reasons why Peter seems
not to have used the more prestigious Beneventan script. (I am indebted
to Professor Francis Newton for several pleasant and helpful conversations
on this subject.)
Bloch (:q8¡) :o¸–.¸.
Meyvaert (:q¸¸) pl. iia, dates the script ‘c.::¸.–¸¸’ – in other words, to the
period shortly after Peter’s exile.
text great interest. References to Tusculum (8.:, q.¸–¸) might
have struck a personal note. The mention of Subiaco (q¸..)
might have seemed important as this was the locality where
St Benedict had tarried prior to making his settlement at Monte
Peter’s copy of Frontinus follows immediately upon
a partial copy of Vegetius, and the conjunction is not perhaps
insignificant: that text contains words of praise for Frontinus
which must have piqued the interest of a budding writer.
tone and attitude of Frontinus himself might have left a deep
‘Mendosus et pessimis litteris, adeo ut vix queam legere’:
Poggio’s words were intended no doubt primarily to describe
the physical condition of the Monte Cassino manuscript, but
they might be applied with equal justice to the quality and intel-
ligibility of the text in this codex. Because we know the scribe so
well, it is tempting to believe that many of the blemishes are the
fault of Peter the Deacon. A good number of slips could readily
be understood – some of them even excused – in the product of
a youthful, careless and impatient copyist whose transcription
was made for largely personal reasons. Unfortunately, there is no
objective means of isolating Peter’s role from that of his prede-
cessors, nor yet of reaching a general assessment of his reliability
as a copyist. In contrast to the ample material by which one
Paulus Diac., Hist.Lang. i..6; Chron.Cas. i.:. Note especially Gregory the
Great’s life of St Benedict (Dial. ii.:, p.¸¸.6 Moricca): deserti loci secessum
petiit, cui Sublacus vocabulum est, qui a Romana urbe quadraginta fere milibus distans,
frigidas adque perspicuas emanat aquas. quae illic videlicit aquarum abundantia in
extenso prius lacu collegitur, ad postremum viro in amne dirivatur.
Veg. ii.¸ nam unius aetatis sunt, quae fortiter fiunt; quae vero pro utilitate rei publicae
scribuntur, aeterna sunt. idem fecerunt alii complures, sed praecipue Frontinus, divo
Traiano ab eius modi comprobatus industria.
The imperial familiarity of Frontinus’ prologue, of course, and his pride
in the aqueducts as monuments of Roman greatness (e.g. :6, 88). Note
especially ::q.: rem [sc. tutelam ductuum] enixiore cura dignam, cum magnitudinis
Romani imperii vel praecipuum sit indicium – a sentence unmistakably echoed in
the opening lines of Peter’s own Liber dignitatum Romani imperii: Bloch (:q8¡)
:¸o. For Peter’s use of the De Aquaeductu as an historical source, see above,
judges him as a literary figure (be he forger, plagiarist, cavalier
compiler), we have from his pen but two texts which purport
to be straightforward transcriptions. For neither Vegetius nor
Frontinus is there an extant exemplar against which one can
measure his ability as a scribe. But that he expanded the text
of Vegetius – and how he did so – is of some importance for
an editor of Frontinus. Note in Vegetius i..8.. ipsos <invictissimos
atque excellentissimos>progenuere Romanos, 8 <tirrano et scelesto>Han-
nibali, iii.:o.. provinciae <status et celsitudo terribilis Romani imperii>
conservatur {imperium}, iii..:.¸ <gloriosissimi> Scipionis.
interpolations in Frontinus are ¸.. {Iulii} Caesaris and 88.: quae
terrarum dea consistit cui par nihil et nihil secundum, both of which have
long been recognised.
In one sense, an editor’s treatment of the archetype remains
the same, regardless of the scribe’s identity. To the process of
examinatio it matters little whether the faults of this manuscript
are due to Peter the Deacon or were already in his exemplar.
It serves no useful purpose to vilipend the scribe to whom we
owe this text, and one should not discredit a copy Peter made
in his youth merely in reaction against the complex fantasies
of his mature years. An editor of Frontinus, on the other hand,
ought not to ignore the dangers of placing undue confidence in
the authority of a manuscript written by a man whose attitude
and purposes are always questionable and whose concern for
exactitude is never conspicuous.
The manuscript tradition prior to C
Stages of the tradition prior to C can be perceived by analysis
of certain types of copyists’ errors.
It is abundantly clear that
some of C’s faulty readings are due to misreadings in minus-
cule script (note, for example, ... servato] struato, .¸.¸ ne rivus]
Reeve (.ooo) .¸.–¸; for further examples (alsofromReeve) see commentary
to ¸:...
Kunderewicz ix–xprovides afewexamples; more extensive lists inGonz´ alez
Rol´ an li–lv.
neminis, ¸¸.¡ desit iis] desitus, :o..:¸ Vibius] albius). Errors that in-
volve abbreviations, suspension strokes, or marks of punctuation
also point to a copy in minuscules (although their frequency is
perhaps due to carelessness on the part of Peter the Deacon).
These are especially frequent in the case of final m (e.g. 6.:
aquam] aqua, q.6 moderatione] moderatione , :8.¸ libram] libram, q6
curam] cura), sometimes with confusion between -us and -um (e.g.
¸.6 ramus] ramum, ¸.: datum] datus). Plural and singular are some-
times reversed in this way, as are active and passive (e.g. :8.¡
aequat] equata, 8q.¡ debeant] debeat, :.:.: sustinentur] sustinetur, ¸..¸
dat] datur, :o¡.¡ videbatur] videbant). Faulty word-division reveals
vestiges of scriptura continua (e.g. ¸.¸ senatu M.] senatu, ¸¸ breve
spatium] breves. At cum, 68.¸ erogatione], ¸¸.¸ comprehen-
sionem scio] compreensio.nescio).
Some errors may reflect earlier stages in the tradition: ¸.¡
spatium] statium, :¡.. adiectione sui] adlectiones sex, :8.¸ quarum]
duarum, .¸.¸ saepius] septus, 8¸.¸ duplicata] publicata, :.¸.. gelatio]
celatio, :.8.. per suffossa] per suetossa. Blank spaces left in C in-
dicate a scribe’s inability to decipher the text he was copying.
Although the scribe might have been Peter the Deacon, it is no
less plausible that the blanks were already present in the exem-
plar from which Peter was working. Those that occur near the
end of the text (ch. :.q) probably result from difficulties with the
unfamiliar legal language. Blanks in the opening chapters, on
the other hand, are rather evenly spaced (¡.¸–¸.:, ¸.6, 6..–¸,
¸.¸, ¸.¸), and this fact suggests that they represent physical dam-
age in an ancestral codex. One indication that the damaged and
indecipherable ancestor was not C’s immediate exemplar is the
unexpected appearance of majuscule letters in conjunction with
blank spaces (at 6.¸, :.q.q). It is far more reasonable to suppose
that these meaningless letters have been copied as faithfully as
possible through several stages of the transmission than it is to
see them as very early examples of litterae tonsae (such as were
employed in the papal chancery in the thirteenth century).
Gundermann (:qo¸) :¡¸¡.
The recentiores
Scholars’ searches have so far brought to light a total of eleven
fifteenth-century manuscripts containing the De Aquaeductu.
U Vatican, Urb. lat. :¸¡¸, ff.:–¸.
. Florentine calligraphy of
the :¡¸os, attributed to Francesco de’ Contugi (whose notar-
ial sign appears on f.¸.
). See Pellegrin (:q8.) ii..: 6¸.. Cited
by Poleni and subsequent editors. I use my own collation.
V Vaticanlat. ¡¡q8, ff.:–.o
, late s.x\. The closeness of V’s text
to the editio princeps has long been recognised (B¨ ucheler vi),
and one can note that its text of Tacitus’ Agricola (ff.6¸
derives from Vat. lat. ¸¡.q.
Cited by Poleni and subse-
quent editors. I use my own collation.
M Paris N.A.L. 6.6 (formerly Middlehillensis ¸¸o6), s.x\
(the date ‘xvi Iu. :¡¸¸’ occurs ina marginal note addedlater).
The text is divided into three books (chapters :–6¸, 6¡–86,
8¸–:¸o). This manuscript also contains Frontinus’ Stratege-
mata. Noted by Aly (:q:¸); cited by Grimal and Gonz´ alez
Rol´ an.
E Escorial S.iii..¸, ff.:–¸.
, subscribed (f.¸.
) ‘Rome anno
a nat. dni. M CCCC L quinto per me Joannem Vynck
clericum Colonien. dioc. transcriptum feliciter’. See An-
tol´ın (:q:o–:6) i\: ¸q. The text of Frontinus was noted by
Rampolla (:qo.) and more thoroughly investigated by Ru-
bio (:q6¸). Cited by Kunderewicz and Gonz´ alez Rol´ an (the
latter using the siglum S).
A Milan, Ambros. i..q sup., ff.¸–:¸
, with subscription (f.:¸
‘Romae VII Kl. Iunii . . . anno MCCCCLIIII’. Briefly noted
by Sabbadini (:qo¸) ¸o6 and Bloch (:q¡8) ¸q. Studied by
Rubio (alongside E) and used by subsequent editors. I have
collated a microfilm copy.
Listed here in the order of their use by scholars and editors of Frontinus.
I do not include Paris lat. 6:.¸\ or Vatican Barb. lat. :¸:, both of which
are manuscript editions: see below n. :6:, :6¸.
See Murgia (:q¸¸).
Est. Modena, Est. lat. :¸. (formerly c.T.6.:¡), s.x\. Mentioned
by Grimal and Kunderewicz, but its readings are reported
for the first time by Pace and Gonz´ alez Rol´ an (both of
whom give it the siglum E). I have collated a microfilm
B Berlin, Hamilton .¸¡, ff..¸–¡¸
, s.x\
. Corrections and
conjectures (B

) are in the hand of Pietro Donato (died
:¡¡¸). The anonymous scribe had worked for Donato at
Basel: De la Mare (:q¸8) ..:. Reeve has shown that Bwas
‘the one exemplar’ used by Giocondo (see below). I have
collated a microfilm copy.
O Vatican, Ottob. lat. .o8q, ff.qo–:::
, s.x\
. See Pellegrin
(:q¸¸) i: ¸8.–¸. I use my own collation.
F Vatican lat. ¸:.., ff.:–.¡
. The hand that wrote Fronti-
nus continues to the end of Festus (f.6¸
), where it gives
the date, .o Aug. :¡68. See Ruysschaert (:q¸8) ¸¸¸. The
Maffei arms appear on f.: (Pace, fig. :¸¡).
H British Library, Harley ¸.:6, s.x\
. A slovenly copy, far
outshone by U (its nearest relative).
S Siena L.v..6, ff..6, s.x\
. Described by Terzaghi (:qo¸)
¡:o, its readings for Frontinus first reportedby Pace (whose
siglum is S).
The relationships of these manuscripts (to each other and to
C) and their consequent editorial value have been variously ex-
plained. Hopes of retrieving among them a tradition indepen-
dent of the Casinensis have at last been firmly dispelled,
Reeve has established a comprehensive stemma.
I am perhaps too optimistic: coot, tóv :i, ts vtspcv óvco: n, ttioûnocv-
:ci. For the better part of a century the Hersfeldensis proved an almost
irresistible mirage appearing to those seeking a draught of independent
truth available in the fifteenth century. Cf. Sabbadini (:8qq) :¸¸: ‘Si ritiene
comunemente che i codici del De aquaeductibus del sec. x\ discendano
tutti dal cassinese [he cites B¨ ucheler]; ma se ci ` o vale per quelli che furono
esattamente collazionati, rimane da vedere se fra quelli finora non esam-
inati ce ne sia qualcuno di lezione diversa, il quale potrebbe risalire all’
esemplare hersfeldese.’
Reeve (:q8¸) :66–¸o.
E Est.
Reeve is at pains to point out that the case for independence
among the recentiores must rest upon readings of their common
ancestor (which, for convenience, I have called c) and not on
readings peculiar to one or another manuscript. Renaissance
emendation (right or wrong) must be recognised as such, and it
cannot be brought as evidence for an independent transmission.
Among the readings that were to be found in c, in fact, are some
which can only be explained as misinterpretations in copying
from C.
Note that a is the source of the recentiores – except for B

, a
corrector whose evidence now deserves a closer look. As might
be expected, B

often agrees with CU, setting right an error of
BOA. For example: ¸.. in urbe: urbem, sanciant: faciat, :¸.: excip-
itur: concipitur, :q.¸ piscina: pisana, q initur (anitur U): oritur, .¸..
vires om. BOA, ¸o.. sint: fuit, ¸¸ aut oneranda esse erogatione om.
BOA, ¸¸.. que om. BOA, 8q.: supervenerant: pervenerant, :o..¡
etiam om. BOA, :o¸.: instruendum: instituendum, :.8.: vindicaren-
tur: iudicarentur, :.q.:: permiserint om. BOA. More interesting
is the case where CU are wrong: :o¸.: hominum (hominem B

dominum BOA (rightly); B

has part way returned to the para-
dosis but has nonetheless written an accusative required by the
Inadequacies in Rodgers (:q¸8) were rightly noted by Reeve (:q8¸) :6q–¸o.
More often, B

offers the same reading as C, or with only a
slight change (usually for the better), adjusting for errors that B
inherited from a:
¸.. erogationes habiles: habilis erogationes, is (his B

): quis, ¡.. et om.
a, ¸.8 gemellos: gemellas, 6.: Iterum: item, ¸.¡ millies] mine: nunc, q.:
alteriusque: alterius, :q.: contectis: contentis, limum om. a, :q.6 ibi:
ubi (recte), ...¡ post: est, .6.. ud (ut B

): et, .6.¸ modulus: modus,
.q.: area: arta, 6¡.: quantumque: quantumcumque, ¸..¸ quinarias om.
a, ¸6.¸ etiam si inter: etiam inter (om. si), ¸¸.¸ ieiunum (ieiunam B

om.sp.rel. a, q¸.¡ tam (tum B

): tamen, :o..¡ auus: cuius, :o¸.:
pertinentia: pertineulia a, :o¸.. ducende: dicendio (dicende O), :o¡..
interdiu: interdici (interdiei U), ::o.: manationibus: minationibus, :::..
nam: non, ::¸.¸ solute: solite, ::¸.¡ et ut . . . coartantur om. a, ::¸.. ad:
adit, ::8.: Quomoda (commoda B

) om. a, ::q.. quia non: qui (quia U)
ante, :.o quid om. a, :.:.: in: ideo, :.:.: arcuationibus: circuitionibus,
:...: enim om. a, :.... pile: parve, :.¸ tollerentur . . . exportarentur
om. a, res om. a, :.8.. suetos: suetas, :.q.¡ que edificia (edificiis B

urbi continentia sunt erunt in his hortis prediis om. a.
Three readings in the preceding list (::8.: Quomoda, :.:.: in,
:.... pile) were amongst the compelling instances adduced by
Reeve to show that the a manuscripts derive from C and that

had access to a non-a source. We have a similar instance
at ::.¸, where C first wrote quarto decimo, but at once added in
over the ar which provides the variant quinto: while a read quarto,

in the margin has xv. Nothing so far can establish that B

is independent of C. To examine this question I have found a
total of twenty-six readings where B

differs from both C and a
(those which I print are marked with an asterisk).
*¸.: {Romam} del. B

; *¸.¸ idem B

A: ibidem CU; *¸.¡ spatium

M: statium CU; ¸.¸ substructionibus B

: sub[. . .]bus Ca; *:q.¸
uiminalis B

: [. .]minalis Ca; :q.q is B

, Dederich: se Ca; *....
I do not include :o8 oporteret B

: oporte(re) C: oportet a, which may be my own
misinterpretation of B’s oportet altered by a superscript -re intended to give
oportere (the reading of C). B

also agrees with C in numerical readings in
the following places: 6.:, ::.¸, :¡.¸, :¸.¡, :¸.6, ¸..¸.
luculianis B

(-ulli- Jocundus): lucilianis CU (-illi- BOA); 66.¸ per-
veniret B

: veniret Ca; *¸¸.¸ <sed et> B

(sed only Jocundus); *88.¸
sepe B

: se C: om. a; q¸.¸ saxeos B

: saxosos CUA: saxos BO; q¡.¸
{et} legis B

: et leges CUBO: eius legis A; *q¡.¡ publicum penderetur

: publico impenderetur Ca; :o..¸ satrius B

: tarius CBA: tario O;
:o..¡ satrio B

: taurio C: tario a; *:o..¡ {et} B

; :o..¡ sergio B

: se-
rio C: segrio U: segerio BOA; ::8.. ex ortis aedificiisve B

: exolie.difficisve
C: ex oli edicisve a; *::q.: enixiore B

A: enixiorem CUBO; *:.¸ viri
<boni> B

: viri . . . [rasura] C, viri a; *:.8.: utilitate B

: utilitatis
C: utilitas a; :.8.: eripuerint B

, Dederich: eripuerunt Ca; :.q.¸ ea
quae B

: eaea que C: ea aqua U: ea eaqua BOA; :.q.6 dare damnas
dedito in marg. B

: det Ca; *:.q.¸ obponito etc. B

, Schulz: obponit
etc. Ca; *:.q.q liceto B

, Bruns: licet Ca
These include trivial, straightforward corrections that might be
made by any intelligent reader facing anobvious difficulty inBas
well as what might be more thoughtful emendations introduced
by a fifteenth-century scholar.
Examples of the former are
q¸.¸, *::q.:, :.q.¸. Of the latter we have *¸.¸, *¸.¡, ¸.¸, *:q.¸,
:q.q, *...., *¸¸.¸, q¡.¸, *q¡.¡, *:o..¡ {et}, ::8.., *:.8.: (bis),
:.q.¸, :.q.6, *:.q.¸, *:.q.q (that ¸.¸ idem and ¸.¡ spatium at least
are not beyond the capacity of such readers is clear from the
coincidental appearance of these same readings in A and M
respectively, notorious amongst recentiores for their enthusiastic
editorial activity). But against these reasonable improvements
we must set Satrius and Sergius (:o..¸–¡), of which both look
like guesswork and neither is impressive: Sergius is a Roman
gentilicium (a praenomen is wanted), and one Satrius Rufus was
known to Renaissance scholars from Pliny, Ep. i.¸.::, ix.:¸.:¸.
These contributions on the part of B

number twenty-two in
all, of which twelve appear in my text along with parts of three
others (q¡.¸ legis, ::8.: aedificiisve, :.q.6 dare damnas).
This is
I acknowledge that the distinction is entirely subjective, and that some will
disagree with what I judge to be ‘improvements’.
In fairness I should perhaps point also to :q.q is (where I print hic), for this
was independently conjectured by Dederich and has long been accepted.
not a bad showing, since c already shows the fruits of major, if
largely superficial, editorial attention.
I have thus far withheld four readings which obviously do not
derive fromCand which may be too good for conjecture. These
*¸.: {Romam} The word might have been deleted in B

cause it was missing in the manuscript being consulted (which
was not C, where it is plainly to be read). Equally well it could
have been suppressed for stylistic reasons (on comparison with
¡.¸ in urbem?). I remove the word because I judge it to be an
interpolation of Peter the Deacon.
66.¸ perveniret In 6¸.¸, an exactly parallel passage, C has per-
veniret (the source of B

’s reading?), and perhaps one should print
it here. The compound verb present in C’s exemplar could have
lost its prefix through carelessness (after piscinam?) on the part of
C. In either case, it may be noted that B

does not catch the dif-
ficulty with eroga<ba>ntur earlier in the sentence, although Est.
does so (presumably alerted by the tense of the subjunctive).
*88.¸ s(a)epe Conjecture can, I think, quite readily account for

’s saepe (C has se), good as it is, especially since it offers no
other variants in this muddled sentence.
*:.¸ viri <boni> The formulaic adjective could probably have
been supplied no less promptly by a fifteenth-century student
than by his counterpart in ancient Rome.
A fistful of good readings alone cannot establish that B

resents a tradition independent of C. I have sought, but not
found, instances where B

reveals evidence of something more
telling, a reading (good or bad) sufficiently different from what
C offers and where C’s reading would otherwise have raised no
suspicions. (The closest are ¸.: and 66.¸, just discussed.) Most
important, perhaps, is the fact that not a single one of C’s numer-
ous problematic passages can be brightly and clearly illuminated
by any testimony which B

offers. While B

may indeed have
overlooked some things in his exemplar, not being himself a
copyist, it is precisely in C’s troublesome passages that a correc-
tor or collator would have been alert to discover useful readings.
Conjecture can account, I believe, for everything unique that
we find in B

: whatever manuscript was its source contained
nothing that deserves recognition as independently transmitted
Eliminandi haud omnino spernendi. Although they have no worth
as textual witnesses (and C has suffered no physical damage for
whichthey might compensate), the Renaissance manuscripts still
require an editor’s attention for the conjectures they contain.
The text in C is remarkably poor, and readers (beginning with
Poggio) have resorted willy-nilly to emendation. Along with the
now clearer understanding of the nature of the recentiores comes
the observation that the quality of their conjectures is not very
high. The bulk of this activity deserves no place in a critical
apparatus, but one must in fairness report those instances of
successful improvements introduced by anonymous scribes and
scholars of the fifteenth century.
There are few highlights in the editorial history of this work.
The editio princeps, as it happened, represents an already inferior
state of the text which had in effect been in circulation for only
half a century. No improvement at all can be discerned until
the Juntine edition of :¸:¸, prepared by the famous Veronese
architect Giovanni Giocondo.
Although he was fortunate that
the one manuscript he used took him closer to the archetype,
Giocondo’s text is marred by sometimes radical changes of his
own occasioned by his relatively greater interest in the subject-
matter than the language of his author.
But it had at least the
Note the wisdom of Murgia (:q¸¸) ¸.6: ‘In a text of sufficient length,
uncomplicated by a plethora of contaminated MSS, a truly independent
witness always announces itself in no uncertain terms.’
On Fr` a Giocondo, recent studies are Ciapponi (:q6: ), Fontana (:q88);
cf. DBI (.oo:) r\i: ¸.6–¸8.
‘Quem[sc. Frontinum] cumuno dumtaxat exemplare contuli’, in his prefa-
tory letter. He used the distinctive text of B as corrected by B

, and Reeve
(:q8¸) :¸o n.:8, has tentatively identified Giocondo’s own hand at work
merit of being a moderately readable version, and for over a
century new editions were little more than reissues.
Lucas Holste (:¸q6–:66:), native of Hamburg and distin-
guished at Rome both as scholar and librarian,
was the next
serious student of Frontinus. His death prevented publication of
an annotated edition
that would have been remarkable for
two reasons. Holste had collated the Casinensis and recognised
its value.
He also ventured conjectural emendations of a high
order, some of which anticipate those of later scholars and are
now accepted as standard. Editors of Frontinus have never been
aware of the extent or the importance of Holste’s work.
survives only in manuscript, virtually ready for a printer, but
perhaps delayed deliberately to enable him to correlate it with
extensive but still unfinished topographical studies.
in this manuscript. In fairness to Giocondo, it should be noted that his
sentence continues, ‘non quod me existimem illum seu etiam Vitruvium
ad integrum redegisse, sed utrosque tam minus dilacæratos pleribusque
scaturiginibus repurgatos protulisse’.
Lucas Holste (Barb. lat. :¸:, f.:) had collated a manuscript ‘in bibliotheca
PP. Theatinorum Neapoli ad SS. Apostolos’ (‘scriptus est manu recentiori
et quidem hominis Gallicani ante annos circiter c, ut credam Guilielmi
Philandri fuisse, aut Jacobi Metelli’). This codex is now Paris lat. 6:.¸\,
which contains on f.¸ a similar note signed by Holste, but naming Demon-
tesius instead of Jacob Metellus. Despite a cancelled note on f.: (‘Anno
domini a Nativitate :66¡ fecit’), the book itself cannot date, as editors have
said, from :66¡. The text is clearly a manuscript edition, as observed by
B¨ ucheler (pref., p.v), and only a few of its readings are worthy of note.
I am indebted to my colleague Z. Philip Ambrose for careful inspection of
relevant parts of this manuscript.
Rietbergen (:q8¸); Serrai (.ooo) ::–:oo, with bibliography.
Vatican, Barb. lat. :.: (interleaved with Basel edition of :¸¸o), and (a neater
copy) Barb. lat. :¸:.
Barb. lat. :¸:, f.:
: ‘exemplar omnium quotquot extant puto optimum,
unde plurima passim restitui’. He also collated a codex at Naples (above,
n. :6:), there are occasional references to U (e.g., Barb. lat. :.: f.¡q
) and
he gives extensive reports of conjectures made by Fulvio Ursino and Paulus
Manutius. I have not attempted to trace the present whereabouts of the
annotated editions fromwhich he drewthese reports (cf. Barb. lat. :¸: f.:
Except (very briefly) Pace (:q8¸) :¸ and .:o.
See Ashby (:q¸¸) :–.. The latest reference that caught my eye in
:.: (f. ¸:
) is to May :6¸q.
The elegant edition of Giovanni Poleni (Padua :¸..) is in all
respects a landmark. Aware that this text had been discovered
at Monte Cassino and that Mabillon had seen the manuscript
there only a fewdecades earlier, Poleni obtained a copy of C, and
was thus the first editor to publish readings from the archetype
Improvement was perhaps inevitable: it was particu-
larly noticeable because Poleni’s own patience and sobriety of
judgement restored a textual balance that was badly needed.
(His, incidentally, was the first edition to divide the text into
chapters.) An alert and wide-ranging commentary incorporated
the linguistic, historical and archaeological data available at the
time; this material has been largely superseded, yet for this por-
tion of his work Poleni has had no worthy successor. The overall
achievement is all the more remarkable because Poleni’s inter-
ests were technical rather than historical or philological (he was
professor of mathematics), and it is ironic that the most serious
shortcoming of his edition lies in a misdirected attempt to re-
compute the pipe-sizes by using a more exact figure for the ratio
of the circumference of a circle to its diameter (t).
No forward steps were taken for a hundred years. Unfin-
ishedstudies of ChristianSchultz (:¸8o–:8¸¡) andKarl Heinrich
(:¸¸¡–:8¸8) formed the basis of an edition with textual commen-
tary produced by Andreas Dederich (Wesel :8¡: ).
plined and uncritical as it sometimes was, the work of these three
scholars reveals a serious interest in advancing the understand-
ing of Frontinus’ booklet, and Schultz in particular approached
His remarks and his practice both suggest that Poleni valued the Casinensis
primarily for its antiquity and its use by Poggio, not as sole authority. His
other manuscripts (U and V) are cited throughout, as are earlier editions;
but we are, after all, a century in advance of the ‘Lachmannian revolution’.
B¨ ucheler x–xiii. Poleni used the fraction ¸¸¸/::¸ adopted by Adrien Metius
(:¸¸:–:6¸¸), although it is clear from Frontinus’ own text (see commentary
to .¡.¸) that his figure was ../¸. (Incidentally, the name pi (t) standing
for ‘periphery’ or ‘perimeter’ was first used by William Jones in his Synopsis
palmariorum matheseos, London :¸o6. The great mathematician Leonhard
Euler fixed the usage by his consistent use of the letter t from :¸¸¸ on.)
For Poleni’s place in the history of hydraulic science see Franke (:q8¸).
Second edition (Leipzig :8¸¸) included also the Strategemata.
the text with freshness and independence of thought.
the glaring defects of Dederich’s edition were turned to profit,
for these attracted the attention of the only good Latinist ever
to edit Frontinus. Franz B¨ ucheler (:8¸¸–:qo8) was twenty-one
years of age when his slender but monumental edition appeared
at Leipzig in :8¸8.
‘Vir de Frontino omnium optime meritus’: B¨ ucheler’s judge-
ment on Poleni can with even greater justice be applied to
B¨ ucheler himself. Here we have the first and best critical edition
of Frontinus’ work. With magisterial firmness B¨ ucheler declared
that C was the archetype,
and he radically but judiciously
thinned the luxuriant growth of conjectures that had pullulated
over four hundred years. The product was a sound and sensi-
ble text supported by a clear and coherent apparatus, the only
weakness of which is that B¨ ucheler was unable to inspect the
Casinensis for himself.
It can fairly be said that the text of Frontinus has seen very lit-
tle improvement since B¨ ucheler’s edition. Fritz Krohn’s Teubner
text (:q..) is a piece of honest workmanship, but it adheres too
Schulz was acquainted with Goethe, who speaks of the projected edition
in a letter to him written in May :8.q. In July :8¸: Goethe’s diary tells us
‘Wurde gestern mit Herrn Schultz seine neue Ausgabe des Frontin und die
Einrichtungder r¨ omischenundorientalischenWasserleitungenbesprochen
. . . Ichlas inFrontins Werke vondenWasserleitungen.’ See Grumach(:q¡q)
Preface, p.v: ‘Poggius in complura exemplaria transcribendum[sc. codicem
Casinensem] curavit ita ut a Cassinensi repetenda sit origo omnium quae
supersunt octo’ (the total number known to him). This falls short of a de-
tailed demonstration such as modern scholarship would require (especially
in light of more sophisticated studies on the worth of fifteenth-century
manuscripts), but it has left the burden of proof squarely on the shoulders
of those who would claim authority for manuscripts other than C. Behind
B¨ ucheler’s decision lay the value traditionally accorded to C by Poleni,
Rondelet (:8.o) v, as well as the view expressed by Haupt (:8¡¸) ¸::: ‘ich
sage, die Handschrift, denn wo von Ueberlieferung die Rede ist kann nur
die Handschrift von Monte Cassino in Betracht kommen, wenn die Kritik
eine regelrechte und sichere sein soll’.
He relied primarily on an apograph of C procured by Schultz (and made
by the eminent epigraphist Olaf Kellermann).
closely to C,
and its editor was gifted with neither imagina-
tion nor critical skill. Sober and sane were the labours of Roberto
Valentini and Giuseppe Zucchetti, who included an edition of
chapters :–.. and ¸q–q¸ intheir collectionof texts onthe topog-
raphy of Rome.
In the Bud´ e edition of :q¡¡
Pierre Grimal
introduced a convenient division of sections within chapters, but
his text shows neither care nor critical ability. Chief among its
demerits is the editor’s mistaken belief that there are fifteenth-
century witnesses to a tradition independent of C. In this belief,
alas, Grimal has not been alone: the same error informs all sub-
sequent editions and most of the recent textual studies.
result has been distraction from the archetype and scholarly at-
tention diverted from the need for conjectural improvements.
The time has come for editors to recognise the realities of the
tradition: a new and closer scrutiny of the Casinensis is long
overdue. Frontinus will best be served by those who squarely
face the dangers of dealing with a badly corrupt tradition and
who do not avoid conjecture out of fear that they may err.
* * *
Of the commentary contained in Poleni’s edition something has
already been said. That provided by Dederich was largely con-
fined to textual points, an uncooked stew with most of its ingre-
dients fetched from the random notes of Schultz and Heinrich.
It would not be too serious an oversimplification to say that it amounts
to little more than a diplomatic transcript, the need for which had been
removed with the publication of Herschel’s facsimile (above, n. :.¸).
Codice topografico della citt` a di Roma, vol. i (Fonti per la Storia d’italia 8:, Rome
:q¡o) q–¸6.
Reissued in :q6: (complete with typographical errors) as a ‘second edition’.
Bruun (:qq:) ¸8:–¡ judiciously surveys the editorial history. For Grimal V
and M are the representatives of this supposedly independent tradition.
Rubio (:q6¸) elegantly demonstrated that VM derive from A, but argued
that Aand Ewere independent witnesses. Kunderewicz uncritically follows
Rubio. Gonz´ alez Rol´ an, whose collations are marked by noticeably greater
accuracy, draws a stemma in which the source of all the recentiores is given
equal status with C. Pace (:q8¸) too enthusiastically embraces the Siena
manuscript (his ‘discovery’: :o, .o6, .:o): ‘E di certo indipendente dal
Although not strictly speaking a commentary, the great work
of Rudolfo Lanciani, I comentarii di Frontino intorno le acque e gli
aquedotti [sic], which appeared in :88:, incorporated the fruits
of nineteenth-century historical, epigraphical and archaeologi-
cal scholarship and was a fitting complement to B¨ ucheler’s edi-
tion. With Lanciani began the modern topographical study of
Rome’s aqueducts, magnificently continued – one is tempted to
say completed – by Thomas Ashby’s Aqueducts of Ancient Rome,
posthumously published in :q¸¸.
Ashby took up a suggestion
of Lanciani to trace the aqueducts’ courses by searching for piles
of deposit cleared from their channels. He supervised the metic-
ulous operations whereby levels were taken of most of the extant
thus aiding both identification and technical under-
standing. He worked in close conjunction with Esther Boise Van
Deman, a pioneering student of construction techniques, whose
own important monograph, The Building of the Roman Aqueducts,
appeared in :q¸¡.
In the purposefully restricted notes which accompany his
Bud´ e text, Grimal drew with profit upon these recent archae-
ological advances. Nor did he neglect questions of a techni-
cal nature, responding to what was beginning to develop as a
more sophisticated interest in the history of technology. This
interest has grown considerably in recent decades, with the
focus of research understandably moving away from the text
of Frontinus.
Scholars in other areas have also begun to
sidestep the shadow of Frontinus, and his authority in matters
The thoroughness of Ashby’s work, carried out over more than three
decades, inevitably leaves less for his successors, and of what he saw much
has since disappeared. But his cannot of course be the last word: for a case
in point, see commentary to 6.¸.
Reina et al. (:q:¸).
For the Ashby–Van Deman partnership, see Bull-Simonsen (:q86) and
Claridge–Cozza (:q86).
The study of Roman aqueducts takes one, of course, far afield from Rome,
and some of the most interesting features (and problems) are to be encoun-
tered elsewhere. Hodge (:qq.) has become the standard comprehensive
work in English on the subject.
of Roman administration is receiving a critical assessment long
It appears unlikely that the labours of philologists, archae-
ologists, historians, and students of Roman law, administration
and technology will be closely correlated in future, but – thanks
not a little to the enthusiasm that Rome’s aqueducts have al-
ways inspired in amateurs – the booklet Frontinus wrote will
perhaps continue to provide a common meeting ground for
representatives from increasingly specialised branches of learn-
ing. Wasserversorgung im antiken Rom, published by the Frontinus-
Gesellschaft in :q8., admirably exemplifies the positive benefits
of such contact: the text of Frontinus (in Latin and in translation)
is juxtaposed with historical and technical studies contributed by
scholars whose backgrounds and specialties are richly varied.
Frontinus himself would perhaps have smiled on the happy col-
laboration represented by this splendid work and by others that
have followed. When there is much to be known, it is good that
willing learners are many.
Codex Casinensis ¸6: (C) is the sole basis on which a critical edi-
tion of this text may be constituted. The nature of this archetype
invites the use of somewhat modified editorial conventions.
Bruun (:qq:), building on the work of Werner Eck and others, and ac-
knowledged by Evans (:qq¡), DeLaine (:qq¸) and Rodgers (:qq6).
Pietrantonio Pace’s Gli acquedotti di Roma (:q8¸) represents the opposite
extreme. This book is also elegant and appeals to the interest of an interdis-
ciplinary audience, but its author (‘ingegnere nucleare’ and by confession
an amateur) ought never to have tried his hand at editing: in the Latin
text and apparatus lie the book’s greatest weakness; for another see Bruun
(:qq:) ¸¸–¡.
The conventions I have adopted will not, I trust, prove to be unduly trou-
blesome: they are, after all, familiar from papyrological and epigraphical
texts and they are those recommended by West (:q¸¸) 8o.
It is particularly important to indicate blank spaces left in C
and to distinguish these from other, purely conjectural, lacunae.
C’s blank spaces are of two kinds: those left by the scribe when
his exemplar was indecipherable or might itself have had a blank
(e.g. ..:, ¡.¸) and those required by the condition of the parch-
ment (e.g. .o.¸). The latter, although noticeable in facsimile, are
of no consequence. I have consistently noted instances of the
former, where appropriate in the text (with square brackets) and
always in the apparatus. After much hesitation, I have expressed
the extent of such blanks as an approximate number of letter-
spaces. I amfully aware that this practice has serious drawbacks:
the script itself is erratic and irregular, and one would never as-
sume that this scribe devoted special care to blank spaces. We
are not dealing with physical damage to an extant copy, and I
cannot overemphasise that the extent of C’s gaps can serve only
as a very rough guideline for conjectural restorations.
For conjectural lacunae (those for which no traces are to
be seen in C) I use pointed brackets. At ¸.., for example, I
print <ducendarum>, noting in the apparatus that this word was
first supplied by Schultz. For conjectural deletions, where nor-
mal practice in most traditions allows the use of square brack-
ets, I have instead used braces: thus at ¸.. {Iulii} Caesaris (the
apparatus will reveal that the intrusive word was passed over
by c).
For simple legibility and to avoid pedantic distractions, I have
within the text permitted myself certain inconsistencies in the
use of pointed brackets and braces. One example may suffice.
Because C so often omits or wrongly adds a stroke for final m, I
have readily forgone the tedium of printing such exactitudes as
urbe<m> or urbe{m} (where urbem and urbe are what the reader
really wants to see). In all such cases, of course, the apparatus
should make clear what is C’s reading and what is editorial
conjecture. Where the error is of a type found less frequently
(or I am perhaps less confident in the solution) I may leave the
typographical distractions to alert a reader to the conjectural
nature of what appears on the printed page. Finally, for places
where the text is highly conjectural I resort to the use of italics,
alerting the reader to grave uncertainty.
Some will have expected the apparatus to report each and ev-
ery divergence of my text from that found in the archetype. The
fact that C is readily accessible in facsimile, however, has made
such detail largely otiose.
I regularly pass over peculiarities
of C which are purely orthographical: these include unambigu-
ous occurrences of e for ae or oe, the interchange of i and y, the
omission or insertion of h.
Since other manuscripts derive fromC, they are cited only for
conjectures of critical interest. The stemma of recentiores proves
helpful in assigning credit to otherwise anonymous scribes and
readers; but I have chosen not to use the apparatus to set forth
evidence justifying the stemma, nor yet to reveal the develop-
ment and character of the vulgate text in the fifteenth century.
I do not avoid using the siglum a where the readings of that
hypothetical codex can be restored with confidence. For practi-
cal purposes a may be associated with Poggio’s emended copy
of C, thus revealing the range of improvements introduced by
the first of Renaissance readers. Except perhaps for B

, subse-
quent fifteenth-century editorial activity, on the other hand, is
less dramatic and less impressive.
A multiplication of Greek
sigla serves no useful purpose and readings are reported directly
from the manuscript(s) in which they occur.
Selecting which conjectures to report I have found to be
amongst the most difficult of editorial tasks. Many of the al-
terations introduced by early editors can be left in obscurity, for
they were working with a very poorly transmitted text and with-
out good understanding of the textual tradition. On occasion,
nonetheless, such scholars hit upon something which advanced
Nor did it seemworth the effort to relegate archetypal trivia to an appendix.
Those whose disappointment is most severe will appreciate that editors have
to make some practical decisions: life is hard.
Any reader who works through Gonz´ alez Rol´ an’s (or even Grimal’s) appa-
ratus will chide me for not qualifying this statement. MVare nothing if not
impressive – for their poor quality and ungrounded pretentiousness.
understanding or which anticipated a reading later confirmed
in C, and for this they deserve credit. I have tried never to sup-
press conjectures that have won a measure of respect, even if
their merit is little more than exempli gratia. We sometimes learn
most from those whose views are different, and at the risk of
being scorned for lack of judgement I have piously recorded
suggestions, implausible in themselves, which have helped me
understand the author I study. It was a deliberate decision to
print my own conjectures more often than is seemly – more of-
ten, indeed, than I find comfortable.
I stand ready to face a
variety of charges for such irresponsible behaviour; it seems a
fair price to pay for the opportunity to draw upon Frontinus’
text the attention it deserves.
I have not forgotten that Bloch (:q¡8) ¸q n.:6 wrote, ‘Philologists who suffer
from horror vacui had better keep their hands from this treatise,’ and that ‘It
would have been better if Grimal had refrained from introducing highly
conjectural emendations of his own into the text which deserve hardly a
place in the apparatus or in the notes.’ To my beloved master I can only
reply that my text does not appear as part of a series in which such restraint
is required.
C Casinensis ¸6:, in eodem coenobio anno c. ::¸o a Petro
Diacono exaratus
a Fons codicumrecentiorum, anno c. :¡¸o ex Cdescriptus
U Vaticanus Urbinas lat. :¸¡¸
H Londinensis Harleianus ¸.:6
B Berolinensis Hamiltonianus .¸¡
O Vaticanus Ottobonianus lat. .o8q
A Ambrosianus i..q sup.
S Senensis L.v..6
E Escorialensis S.iii..¸
Est. Estensis lat. :¸.
M Parisinus N.A.L. 6.6
V Vaticanus lat. ¡¡q8
F Vaticanus lat. ¸:..

Coniecturae et variae lectiones manu Petri Donati ad-
P Parisinus lat. 6:.¸A, saec. XVI(?)
edd. editio princeps (curantibus Pomponio Laeto et
Johanne Sulpitio, Romae :¡86–qo) aliaeque ante Jo-
cundi Florentinam (:¸:¸)
* * * * *
<> per coniecturam addenda
{} per coniecturam delenda
[ ] lacuna vel spatium a librario vacuum relictum
| finis vel versus vel paginae

CVM OMNIS RES ab imperatore delegata intentiorem exi- r [f.zz
gat curam, et me seu naturalis sollicitudo seu fides sedula non
ad diligentiam modo verum ad amorem quoque commissae
rei instigent, sitque nunc mihi ab Nerva Augusto, nescio dili-
gentiore an amantiore rei publicae imperatore, aquarum in-
iunctum officium, ad usum tum ad salubritatem atque etiam
securitatem urbis pertinens, administratum per principes sem-
per civitatis nostrae viros, primumac potissimumexistimo, sicut
in ceteris negotiis institueram, nosse quod suscepi. neque enim z
ullum †omnis† actum certius fundatum crediderim, aut aliter
quae facienda quaeque vitanda sint posse decerni, aliudve tam
indecorum tolerabili viro quam delegatum officium ex adiuto-
rum agere praeceptis, quod fieri necesse est quotiens imperitia
praepositi ad i[nferi]orumdecurrit usum, quorumetsi necessariae
partes sunt, ad ministeriumtamen ut manus quaedamet instru-
mentum agentis <adhibentur>. quapropter ea quae ad univer- .
samrempertinentia contrahere potui, more iamper multa mihi
officia servato, in ordinem et velut {in hunc} corpus deducta in
hunc commentarium contuli, quem pro formula administratio-
nis respicere possem. in aliis autemlibris, quos post experimenta ¸
et usum composui, succedentium res acta est; huius commen-
tarii pertinebit fortassis et ad successorem utilitas, sed cum inter
initia administrationis meae scriptus sit, in primis ad meam
Incipit prologus iulii frontini in libro de aqueductu urbis romae C
r <cum> (tum a) post officium add. Krohn civitatis a: celuitatis C z.:
omnis] omis C: homini Haupt: an hominis? actumego: actus C: satius Haupt
fundatumego: fundatus C: fundamentuma: fundatius Haupt quod a: quo C
praepositi Schultz: precositei C ad inferiorumscripsi: adi[c. : litt.]orua C: ad
illorumSchultz <adhibentur>Reeve: post agentis spat. c. ¡ litt. C: esse debent
add. Jocundus: sunt Grimal . servato a: struato C {in hunc} B¨ ucheler
(hunc C: hoc a: unum Polenus)

institutionemregulamque proficiet. ac ne quid ad totius rei per- j
tinens notitiampraetermisisse videar, nomina primumaquarum
quae in urbem {Romam} influunt ponam; tum per quos
quaeque earum et quibus consulibus, quoto post urbem condi-
tamannoperducta sit, deinquibus ex locis et †a†
<quot passus ductus cuiusque efficiat, et ex eo>quantumsubter-
raneorivo, quantumsubstructione, quantumopere arcuato; post .
altitudinemcuiusque, modulorumque <rationes, quemmodum
quaeque aqua habere visa sit quantumque erogaverit, quaeque>
erogationes habiles factae sint; quantum extra urbem, quan-
tum <in> urbe unicuique regioni pro suo modo unaquaeque
aquarum serviat; quot castella publica privataque sint, et ex
is quantum publicis operibus, quantum muneribus – ita enim
†cultiores† appellantur –, quantum lacibus, quantum nomine
{Iulii} Caesaris, quantum privatorum usi<bus> beneficio prin-
cipis detur; quod ius <ducendarum> tuendarumque sit earum;
quae id sanciant poenae lege, senatus consulto et mandatis prin-
cipum inrogatae.
Ab urbe condita per annos quadringentos quadraginta unum q
contenti fuerunt Romani usu aquarum quas aut ex Tiberi aut
ex puteis aut ex fontibus hauriebant. fontium memoria cum .
sanctitate adhuc exstat et colitur (salubritatemaegris corporibus
afferre creduntur), sicut Camenarum et †Apollinaris† {in} et
j.: {Romam} seclusi sit Polenus: sint C a quoto <miliario> Jocundus
cepisse. | at C: cepissent a: coepisset Polenus: <initium> capessat Krohn: capta
sit Bennett: concipiatur Grimal ‘addendum quot passus ductus cuiusque effi-
ciat’ B¨ ucheler et ex eo addidi substructione a: -tione C . rationes
quaeque Grimal (rationes denique quae Schultz): rationema quemmodum
. . . erogaverit exempli causa scripsi (cf. ó¸.. ) erogationes habiles C: ab
illis erogationes a: erogationes ab illis edd. in urbe UB
: urbe C: intra
urbem Polenus privataque del. Schultz lacus post cultiores ins. Lanciani,
salientes Krohn appellantur C: appellant ed. Argent. iulii recte praetermisit
a usibus Heinrich: usi C: usui a <ducendarum> Schultz poenae
lege Holstenius, B¨ ucheler: pena.elige C: p(o)ena e lege a: poenae ex legibus Hein-
rich consulto C: consultis a Explicit prologus C q.. Camoenarum
Dederich: caminaras (ut vid.) C apollinaris C: Apollinis Dederich et Iutur-
nae Dederich: inetiuturne (ut vid.) C
Iuturnae. nunc autem in urbem confluunt aqua Appia, Anio ¸
vetus, Marcia, Tepula, Iulia, Virgo, [Alsietina] (quae eadem vo-
catur Augusta), Claudia, Anio novus.
M. Valerio Maximo P. Decio Mure consulibus, anno post j
initium Samnitici belli tricesimo, aqua Appia in urbem in-
ducta est <ab> Appio Claudio {Crasso} censore, cui postea
[Caeco] fuit cognomen. [idem eo anno] et viam Appiam [a
porta] Capena usque a[d urbem] Capuammuniendamcuravit.
collegam habuit C. Plautium, cui ob inquisitas eius aquae ve- .
nas Venocis cognomen datum est. sed quia is intra annum et ¸
sex menses, deceptus a collega tamquam {ib}idem facturo, ab-
dicavit se censura, nomen aquae ad Appii tantum honorem
pertinuit, qui multis tergiversationibus extraxisse censuram tra-
ditur, donec et viam et huius aquae ductum consummaret.
concipitur Appia in agro Lucullano via Praenestina inter mi- ¡
liarium septimum et octavum, deverticulo sinistrosus passuum
septingentorum octoginta. ductus eius habet longitudinem a ¸
capite usque ad Salinas, qui locus est ad portam Trigeminam,
passuum | undecim milium centum nonaginta. <ex eo rivus [f.zz
est> sub<t>er<raneus pas>suum undecim milium centum tri-
ginta, supra terram substructio et arcuatura proximum portam
Capenam passuum sexaginta. iungitur ei ad S[p]em veterem 6
in confinio hortorum Torquatianorum et [. . . . .]norum ramus
Augustae ab A[ugusto] in supplementumeius additus, [cui lo]co
¸ cfluunt C: influunt a anio (ut vid.) a: quam C alsietina a: [spat. c. ç
litt.] C j.: valerius maximus C: em. a tricesimoSigonio praeeunte B¨ ucheler:
vicesimo C <ab> edd. crasso seclusi C(a)eco A, Jocundus: [spat. c. ó
litt.] C idem eo anno scripsi: [spat. c. , litt.] C: qui a a porta a: [spat. c. 8
litt.] C ad urbem a: [spat. c. , litt.] C muniendam a: -da C . Plau-
tium Opsopoeus, Holstenius: plautum C ¸ is a: his C idem B
A: ibidem
C ¡ lucullano a: luculano C prenestina a: prenestrina C: Collatina
Lanciani ¸ ex eo . . . passuum restituit B¨ ucheler: supersuum C: subterraneo
rivo passuumJocundus miliumU: milia C arcuatura suspectum: <opus>
arcuatum B¨ ucheler (substructione et arcuatione Polenus) 6 ante iungitur spat.
c. ¡ litt. C, sed ex c. .¡.¸ colligere licet nihil fere deesse Spem Holstenius, Polenus:
s[c. : litt.]em C [c. ¸ litt.]norum C: Epaphroditianorum Corradinus de Al-
lio: Taurianorum Carcopino ramus ed. Basil.: -um C Augusto Holstenius,
Polenus: a[c. ¸ litt.] C: Agrippa Schultz additus P: -um C cui loco Grimal
(loco iam Holstenius): [c. ¡ litt.]to C: imposito Polenus

co<g>nomen [aquarii de]derunt Gemellorum. hic via Praenestina ¸
ad miliarium sextum deverticulo sinistro<sus> passuum non-
gentorum octoginta proxime viam Collatinam accipit fontem.
cuius ductus usque ad Gemellos efficit rivo subterraneo passus 8
sex milia trecentos octoginta. incipit distribui [in] imo Publicii q
clivo ad portam Trigeminam, qui locus Salinae appellantur.
Post annos quadraginta quamAppia perducta est, anno ab urbe 6
condita quadringentesimo octogesimo primo, M’. Curius Den-
tatus, qui censuram cum L. Papirio Cursore gessit, Anionis qui
nunc vetus dicitur aquam perducendam in urbem ex manu-
biis de Pyrrho captis locavit, Sp. Carvilio L. Papirio consulibus
iterum. post biennium deinde actum est in senatu de consum- .
mandoeius aquae opere, †irefent [. . . . . .]nocumi[. . . . .] praetor.†
tum ex [sena]tus [cons]ulto duumviri <a>quae perducendae ¸
<cre>ati sunt Curi[us, qui eam] locaverat, <et Q.>Fulvius Flac-
cus. Curius intra quintum diem quam erat duumvirum creatus ¡
decessit; gloria perductae pertinuit ad Fulvium. concipitur Anio ¸
vetus supra Tibur <via Valeria> vicesimo miliario extra por-
tam [. .]RRL[. .]nam, ubi partem [dat] in Tiburtium usum.
ductus eius habet longitudinem, ita exigente libramento, pas- 6
suum quadraginta trium milium: ex eo rivus est subterraneus
cognomen BOA: co | nom C (unde loco nomen B¨ ucheler): cognomine U (pro-
bat Polenus) aquarii dederunt ego: [c. ¸ litt.]denti C: inde habenti Holstenius:
respondenti Polenus: ideo datur Grimal ¸ sinistrosus a: sinistro C Col-
latinam Schultz: collatiam C 8 passus ego: passuum C q in scripsi: [spat.
c. : litt.] C: Appia Polenus publicii C: PublicioSchultz, Becker clivoPolenus:
ciluo C qui . . . appellantur del. Dederich (ad portam . . . appellantur del.
Schultz) 6.: primo scripsi: uno C M’.] Man. Polenus: M. C lucio C
aquam a: aqua C urbem a: urbe C spurio C lucio C .
irefent[c. ¸ litt.]nocumi[c. ¸ litt.]ptor C: referente . . . praetore B¨ ucheler ¸
senatus consulto A: [c. ¡ litt.]tus [c. ¡ litt.]ulto C aqu(a)e a: que C creati
a: ati C curius a: curi. C qui eam Polenus: [c. ¡ litt.] C et a: om.
C Q. addidi ¡ pertinuit a: pertimuit C ¸ <via Valeria>addidi praee-
untibus Blackman et Mari: lacunam post Tibur ind. Blackman (fort. <. . .-tur>) vi-
cesimo corruptum esse suspicatur Grimal [spat. c. : litt.]RRL[spat. c. : litt.]nam
C: Baranam Cassio: Varianam Fea: Tiburtinam Grimal dat Polenus: [spat. c.
¡ litt.] C usum BOA: usu C 6 quadraginta trium] LIII Ashby: LXIII
Roncaioli Lamberti
passuum quadraginta duum milium septingentorum septua-
ginta novem, substructio supra terrampassuumducentorumvi-
ginti unius.
Post annos centum viginti septem, id est anno ab urbe con- ¡
dita sexcentesimo octavo, Ser. Sulpicio Galba {cum} L. Aurelio
Cotta consulibus, cum Appiae Anionisque ductus vetustate
quassati privatorum etiam fraudibus interciperentur, datum
est a senatu negotium <Q.> Marcio, qui tum praetor inter
cives ius dicebat, eorum ductuum reficiendorum ac vindi-
candorum. et quoniam incrementum urbis exigere videbatur .
ampliorem modum aquae, eidem mandatum a senatu est ut
curaret alias aquas quatinus {quas} posset in urbem perdu-
cere. [itaque pri]ores ductus ref[ecit et] tertiam illa<m pro>prio ¸
ri[vo in urbem per]duxit, cui ab auctore Marciae nomen est.
legimus apud Fenestellam in haec opera Marcio decretum ¡
sestertium millies octingenties, et quoniam ad consumman-
dum negotium non sufficiebat spatium praeturae, in annum
alterum est prorogatum. eo tempore decemviri, dum aliis ex ¸
causis libros Sibyllinos inspiciunt, invenisse dicuntur non esse
[fas] aquam Marciam seu potius Anionem – de hoc enim
constantius traditur – in Capitolium perduci; deque ea re in
senatu M. Lepido pro collegio verba faciente actum, Appio
Claudio Q. Caecilio consulibus; eandemque post annumtertium
¡.: Ser. a: S. C sulpicio a: -us C {cum} B¨ ucheler lucio C ani-
onisque a: amion- C datum a: datus C <Q.> Holstenius, B¨ ucheler
Marcio Polenus: marco C . quatinus alias aquas C: transp. Sauppe
{quas} Schultz perducere Schultz: perduceret C ¸ itaque (Jordan: ille
Holstenius: qui Grimal: Marcius Kunderewicz) priores (Holstenius, B¨ ucheler):
[c. 8 litt.]ores C refecit et Holstenius: rei[c. ¡ litt.] C: restituit et
B¨ ucheler illam (Schultz) proprio rivo (ego): illiobrior(um) C: illis salu-
briorem Holstenius, Bennett: illis uberiorem Sauppe, Jordan in urbem
Schultz: eamaquamPolenus perduxit Schultz: deduxit Holstenius auctor¯ e
artiae C ¡ millies octingenties (Schultz) et B¨ ucheler: mine octingente. Set
C spatium B
M: statium C ¸ decemviri a: decemvira C invenisse
MV: inventi C fas Sch¨ one: [c. ¡ litt.] C perduci C: perducendam Jocun-
dus (-dum B¨ ucheler) senatu M. B¨ ucheler (senat ¯ u iam a): senatu C col-
legio Pighius: collega C Q. a: que C eandemque MV: eademque C
a L. Lentulo retractatam, C. Laelio Q. Servilio consulibus;
sed utroque tempore vicisse gratiam Marcii Regis, atque ita in
Capitolium esse aquam perductam. concipitur Marcia via Va- 6
leria ad miliarium tricesimum sextum deverticulo euntibus ab
urbe Roma dextrosus miliumpassuumtrium, Sublacensi autem,
quae sub Nerone principe primum strata est, ad miliarium
tricesimum octavum sinistrosus intra passus ducentos. †fontin ¸
[. . . . .]sub[. . . .]bus petrei[. . . .]† stat im[mobilis] stagni
mo[do] colore perviridi. ductus eius habet longitudinema capite 8
ad urbem passuum sexaginta milium et mille septingentorum
decem et semis: rivo subterraneo passuum quinquaginta quat-
tuor milium ducentorum | quadraginta septem semis, opere [f.zj
supra terram passuum septem milium quadringentorum sex-
aginta trium: [ex] eo longius ab urbe pluribus locis p<e>r
vallis opere arcuato passuumquadringentorumsexaginta trium,
propius urbem a septimo miliario substructione passuum quin-
gentorum viginti octo, reliquo opere arcuato passuum sex mi-
lium quadringentorum septuaginta duum.
Cn. Servilius Caepio et L. Cassius Longinus, qui Ravilla appella- 8
tus est, censores anno post urbem conditam sexcentesimo vice-
simo septimo, M. Plautio Hypsaeo M. Fulvio Flacco consulibus,
aquam quae vocatur Tepula ex agro Lucullano, quem quidam
Tusculanumcredunt, Romamet inCapitoliumadducendamcu-
raverunt. Tepula concipitur via Latina ad decimum miliarium .
Q. a: que C 6 miliumed. Argent.: milia C passus ducentos B¨ ucheler: pas-
suumducentorumC: spatium(pro passuum) A, unde spatiumpassuumducento-
rumJocundus ¸ locus vix sanabilis fontin[c. ó litt.] C: fontium[spat.] BOA:
<aqua> Grimal sub [c. ¸ litt.]bus C: substructionibus B
: sub fornicibus
Schultz: s. specibus B¨ ucheler: s. rupibus Grimal: s. arcubus Kunderewicz pe-
trei[c. ¡ litt.] C: petraeis Schultz statim [c. ¸ litt.]stagnimo[c. ¡ litt.] C (suppl.
Schultz) perviridi Polenus: pviridi C 8 cum et semis tum semis suspectum,
an ex eo scribendum? ex Polenus: [c. : litt.] C per vallis Holstenius, B¨ ucheler:
p.R. vallis C: per p(opuli) R(omani) vallis a substructione Est: -tione C
reliquo MV: reliqua C sex] sex.s. C (sexs B¨ ucheler) 8.: Caepio edd.: scy-
pio C plautio a: plautius C Hypsaeo (Scaliger) M. B¨ ucheler: hypsaponi
ut vid. C lucullano a: luculano C . fortasse miliarium decimum (c. ¡.¸,
¡.,, etc.)
deverticulo {euntibus ab Roma} dextrosus [. . .] milium pas-
suum duu[m]. inde [rivo] suo in urbem perducebatur.
Post[ea M.] Agrippa aedilis post primum consulatum, impe- q
ratore Caesare Augusto. II. L. Volcatio consulibus, anno post
urbem conditam septingentesimo nono decimo, ad miliarium
ab urbe duodecimum via Latina <deverticulo> {euntibus ab
Roma} dextrosus miliumpassuumduum, alterius <a>quae pro-
prias vires collegit et Tepulae rivumintercepit; adquisita<e>que .
ab inventore nomen Iuliae datum est, ita tamen divisa eroga-
tione ut maneret Tepulae appellatio. ductus Iuliae efficit longi- ¸
tudinem passuum quindecim milium quadringentorum viginti
sex: <rivo subterraneo passuum octo milium quadringentorum
viginti sex>, opere supra terram passuum septem milium: ex
eo in proximis urbi locis a septimo miliario substructione pas-
suumquingentorumviginti octo, reliquo opere arcuato passuum
sex milium quadringentorum septuaginta duum. praeter caput ¡
Iuliae transfluit aqua quae vocatur Crabra. hanc Agrippa omisit, ¸
seu quia improbaverat sive quia Tusculanis possessoribus relin-
quendamcredebat. haec namque est quamomnes villae tractus
eius per vicem in dies modulosque certos dispensatam accipi-
unt. sed non eadem moderatione aquarii nostri par[tem] eius 6
semper in supplementum Iuliae vindicaverunt, nec ut Iuliam
augerent, quam hauriebant largiendo compendi sui gratia.
exclusa ergo Crabra et tota iussu imperatoris reddita <est> ¸
{euntibus ab Roma} ego ab a: ad C ante milium c. ¡ litt. spatium
in C, sed cum voces apte cohaereant nihil omnino deesse crediderim duum a: duu [c.
: litt.] C rivo suo Jocundus (suo rivo B¨ ucheler): [c. . litt.] suo C: suo iure Grimal
q.: Postea M. Polenus (praenomen iam suppl. a): Post [c. ¡ litt.] C: Post annos
LXXXXII M. B¨ ucheler: post hos M. Kunderewicz (cf. .¡.. ) consulibus a:
consule C <deverticulo>Schultz {euntibus ab Roma} ego alterius
aquae Schultz: alteriusque C . adquisitaeque] aquisit(a)eque A: adquisita-
que C: adquisitae aquae Schultz ¸ sex] sex.s. C (sex S. B¨ ucheler) <rivo . . .
viginti sex> ego urbi Drechsler: urbis C: urbem B¨ ucheler ¸ haec] Ec
C (primus recte vidit Krohn): Ea a 6 moderatione a: -tione C par [c.
¡ litt.] C: partem insequente spatio a, unde p. maximam (Polenus) vel potiorem
(Schultz) addere voluerunt vindicaverunt (a praeeunte) Dederich: udindicaverunt
C compendi Jocundus: complendi C ¸ exclusa C: exclusi Dederich:
exclusa est A, Jocundus crabra C: crabram UBO tota Jocundus: tot¯ a C
reddita (Jocundus) <est> scripsi: reddidit C: reddidi Schultz
Tusculanis; qui nunc forsitan non sine admiratione eamsumunt,
ignari cui causae insolitam abundantiam debeant. Iulia autem 8
revocatis derivationibus per quas subripiebatur modum suum
quamvis notabili siccitate servavit. eodem anno Agrippa duc- q
tus Appiae Anionis Marciae paene dilapsos restituit et singulari
cura compluribus salientibus {aquis} instruxit urbem.
Idem cum iam tertium consul fuisset, C. Sentio <Q.> Lucretio ro
consulibus, post annum tertium decimum quam Iuliam dedux-
erat, Virginem quoque in agro Lucullano collectam Romam
perduxit. die quo primum in urbem responderit, quinto Idus .
Iunias invenitur. Virgo appellata est, quod quaerentibus aquam ¸
militibus puella virguncula venas quasdam monstravit, quas se-
cuti qui foderent ingentem aquae modum vocaverunt. aedicula ¡
fonti adposita hanc originem pictura ostendit. concipitur Virgo ¸
via Collatina ad miliarium octavum palustribus locis signino
circumiecto continendarum scaturriginum causa. adiuvatur et 6
compluribus aliis adquisitionibus. venit per longitudinem pas- ¸
suum decem quattuor milium centum quinque: ex eo rivo
subterraneo passuum decem duum milium octingentorum
sexaginta quinque, supra terram per passus mille ducentos
quadraginta: ex eo substructione rivorumlocis compluribus pas-
suum quin | gentorum quadraginta, opere arcuato passuum [f.zj
septingentorum. adquisitionum ductus rivi<s> subterranei<s> 8
efficiunt passus mille quadringentos quinque.
Quae ratio moverit Augustum, providentissimum principem, rr
perducendi Alsietinam aquam, quae vocatur Augusta, non satis
perspicio: nullius gratiae, immo etiam parum salubrem ideoque
nusquam in usus populi fluentem; nisi forte dum opus nau-
machiae adgreditur, ne quid salubrioribus aquis detraheret,
q {aquis} Keuchenius ro.: tertium a: tertio C <Q.> Polenus . die
C: dies a, fort. recte in urbe C: in urbe Polenus responderit ‘corruptum
videtur’ B¨ ucheler quinto C: quintus Dederich ¸ foderent Reeve: foderant
C vocaverunt C: invenerunt a ¸ VirgoSchultz: ergoC collatia C(cf.
¡.,) circumiecto Jocundus: -lecto ut vid. C 8 rivis subterraneis B¨ ucheler:
Rivi subterranei C rr.: dumKrohn: cumC adgreditur C: aggrederetur
hanc proprio opere perduxit et quod naumachiae coeperat su-
peresse hortis adiacentibus et privatorum usibus ad inrigandum
concessit. solet tamen ex ea in Transtiberina regione, quotiens .
pontes reficiuntur et a citeriore ripa aquae cessant, ex neces-
sitate in subsidium publicorum salientium dari. concipitur ex ¸
lacu Alsietino via Claudia miliario quinto decimo deverticulo
dextrosus passuumsex miliumquingentorum. ductus eius efficit ¡
longitudinem passuum viginti duum milium centum septua-
ginta duorum, <rivo subterraneo passuum viginti milium et
mille octingentorum decem et quattuor>, opere arcuato pas-
suum trecentorum quinquaginta octo.
Idem Augustus in supplementum Marciae, quotiens siccitates rz
egerent auxilio, aliam eiusdem bonitatis opere subterraneo per-
duxit usque ad Marciae rivum, quae ab inventore appellatur
Augusta. nascitur ultra fontem Marciae. cuius ductus, donec .,¸
Marciae accedat, efficit passus octingentos.
Post hos C. Caesar, qui Tiberio successit, cum parum et pub- rj
licis usibus et privatis voluptatibus septem ductus aquarum
sufficere viderentur, altero imperii sui anno, M. Aquila
Iuliano P. Nonio Asprenate consulibus, anno urbis conditae
septingentesimo <unde>nonagesimo duos ductus incohavit.
quod opus Claudius magnificentissime consummavit dedi- .
cavitque <Fausto> Sulla <Salvio> Othone consulibus, anno
post urbem conditam octingentesimo tertio Kalendis Augustis.
alteri nomen, quae ex fontibus Caerulo et Curtio perduce- ¸
batur, Claudiae datum. {haec bonitatis proxima est Marciae.} ¡
perduxit a: perducit C ¸ quarto statim in supra additum ut quinto le-
geretur C: quarto a (xv in marg. B
) ¡ duorum suspectum (duum exspec-
tares) <rivo . . . quattuor> ego rz.: egerent Jocundus: agerent C
aliam Jocundus (qui et aquam addidit): alia C rj.: undenonagesimo
B¨ ucheler (DCCLXXXVIIII Pagio praeeunte Polenus): nonagesimo C . Fausto
Sulla Salvio Othone scripsi: sulla et tian C: sulla et tutiano a (titiano A)
tertio (DCCCIII) Pagio praeeunte Polenus: sexto (i.q. DCCVI) C: quarto Valentini-
Zucchetti ¸ quae . . . perducebatur secl. Schultz: ‘addendum aquae’ B¨ ucheler
¡ {haec . . . Marciae} Heinrich bonitatis C: bonitate A proxima C:
proximae B¨ ucheler
altera, quoniamduae Anionis in urbemaquae fluere coeperant, ¸
ut facilius appellationibus dinoscerentur, Anio novus vocitari
coepit {alia omnes praecedit}; priori Anioni cognomen veteris
Claudia concipitur via Sublacensi ad miliarium tricesimum rq
octavum deverticulo sinistrosus intra passus trecentos ex fon-
tibus duobus amplissimis et speciosis, Caerulo (qui a simili-
tudine appellatus est) et Curtio. accipit et eum fontem qui .
vocatur Albudinus, tantae bonitatis ut Marciae quoque adiu-
torio quotiens opus est ita sufficiat, ut adiectione sui nihil
ex qualitate eius mutet. Augustae fons, quia Marciam sibi ¸
sufficere adparebat, in Claudiam derivatus est, manente ni-
hilominus praesidiario in Marciam, ut ita demum Claudiam
aquam adiuvaret Augusta, si eam ductus Marciae non caperet.
Claudiae ductus habet longitudinem passuum quadraginta sex ¡
milium <quadringentorum sex>: ex eo rivo subterraneo pas-
suum triginta sex milium ducentorum triginta, opere supra ter-
ram passuum decem milium <centum> septuaginta sex: ex eo
opere arcuato in superiori parte pluribus locis passuum trium
milium septuaginta sex et propius urbem a septimo miliario
substructione rivorum per passus sexcentos novem, opere arcu-
ato passuum sex milium quadringentorum nonaginta et unius.
Anio novus via Sublacensi ad miliariumquadragesimumsecun- rj
dum in Simbruino excipitur ex flumine, quod cum terras cultas
circa se habeat soli pinguis et inde ripas solutiores, etiam sine
pluviarum| iniuria limosumet turbulentumfluit. ideoque a fau- [f.zq
cibus ductus interposita est piscina limaria, ubi inter amnem et
specum consisteret et liquaretur aqua. sic quoque, quotiens im- ¸
bres superveniunt, turbida pervenit in urbem. iungitur ei rivus ¡
Herculaneus oriens eadem via ad miliarium tricesimum
¸ novus a: nous C {alia . . . praecedit} Heinrich alia C: alias MV: alti-
tudine alias Krohn rq.: {qui . . . est} Schultz . adiectione sui Jocundus:
adlectiones sex C: adiectiones vi V (adiect- iam a) ¡ quadr. sex] CCCCVI
add. Polenus centum] C add. Polenus propius ego (cf. ,.8, .¡.ó): prope C
unius B¨ ucheler: unum C rj.: Simbruino B¨ ucheler: s¯ ubriuno C ¡ ei
Jocundus: et C
octavum e regione fontium Claudiae trans flumen viamque.
natura e<st> purissimus, sed mixtus gratiam splendoris sui ¸
amittit. ductus Anionis novi efficit passus quinquaginta octo 6
milia septingentos: ex eo rivo subterraneo passuum quadra-
ginta novem milium trecentorum, opere supra terram passuum
novem milium quadringentorum: ex eo substructionibus aut
opere arcuato superiori parte pluribus locis passuum duum
{decim} milium trecentorum, et propius urbem a septimo mi-
liario substructione rivorum <per> passus sexcentos novem,
opere arcuato passuum sex milium quadringentorum nona-
ginta unius. hi sunt arcus altissimi, sublevati in quibusdam locis ¸
pedes centum novem.
Tot aquarumtammultis necessariis molibus pyramidas videlicet r6
otiosas compares aut cetera inertia sed fama celebrata opera
Non alienum mihi visum est longitudines quoque rivorum r¡
cuiusque ductus etiam per species operum complecti. nam .
cum maxima huius officii pars in tutela eorum sit, scire prae-
positum oportet quae maiora impendia exigant. nostrae qui- ¸
dem sollicitudini non sufficit singula oculis subiecisse; formas
quoque ductuum facere curavimus ex quibus adpar<er>et ubi
valles quantaeque, ubi flumina traicerentur, ubi montium late-
ribus specus adpliciti maiorem adsiduamque †petendi ac mu-
¡ niendi vi† exig<er>ent curam. hinc illa contingit utilitas, ut rem
¸ natura est B¨ ucheler: nature C: natura A purissimus ut vid. a: pessimus C
6 Ductu C: em. a passus scripsi (cf. ¡.8, .o.8, .:.¡): passuum C milium
trecentorum ego: milia trecentos C milia quadringentos C: em. ego pas-
suumC: passus a duumscripsi: duo C {decim} Polenus miliumtre-
centorumego: milia trecentos C per addidi (cf. .¸.¸) milia quadringentos
C: em. ego unius ego: unumC r6.: tammultis del. censuit Bergk opera
edd.: opere C r¡.: quoque del. Heinrich: ‘suppl. fere <quantas> quomque
deinde per<lustrare et oculis singulas> species’ Krohn . sit a: scid C ¸
lacunam ante nostrae ind. Krohn nostrae . . . sollicitudini Jocundus: nostra . . .
sollicitudine C sufficit C: suffecit Jocundus adpareret scripsi: adparet a:
abparet C traicerentur C (‘suspectum’ B¨ ucheler) adpliciti Jocundus: ad-
plicite C petendi C(obelumadhibuit Krohn): protegendi P: perterendi Jocundus:
tuendi B¨ ucheler: tegendi Ehlers vi C: ii Jocundus: rivi Heinrich: ‘fortasse ductus’
B¨ ucheler exigerent scripsi: exigant C
statim velut in conspectu habere possimus et deliberare
tamquam adsistentes.
Omnes aquae diversa in urbem libra perveniunt. inde r8.:,.
s<erv>iunt quaedam altioribus locis et quaedam erigi in emi-
nentiora non possunt; nam et colles sensim propter frequen-
tiamincendiorumexcreverunt rudere. quinque sunt quarumal- ¸
titudo in omnempartemurbis adtollitur, sed ex his aliae maiori,
aliae leviori pressura coguntur. altissimus est Anio novus, prox- ¡
ima Claudia, tertium locum tenet Iulia, quartum Tepula, de-
hinc Marcia, quae capite etiam Claudiae libram aequat. sed
veteres humiliore derectura perduxerunt, sive nondum ad sub-
tile explorata arte librandi, seu quia ex industria infra terram
aquas mergebant, ne facile ab hostibus interciperentur, cum
frequentia adhuc contra Italicos bella gererentur. iam tamen ¸
quibusdam locis, sicubi ductus vetustate dilapsus est, omisso cir-
cuitusubterraneovalliumbrevitatis causa substructionibus arcu-
ationibusque traiciuntur. sextum tenet librae locum Anio vetus, 6
similiter suffecturus etiam altioribus locis urbis, si, ubi vallium
summissarumque regionumcondicio exigit, substructionibus ar-
cuationibus<que> {veteris} erigeretur. sequitur huius libram ¸
Virgo, deinde Appia: quae cum ex urbano agro perduceren-
tur, non in tantumaltitudinis erigi potuerunt. omnibus humilior 8
Alsietina est, quae Transtiberinae regioni et maxime iacentibus
locis servit.
Ex his sex via Latina intra septimum miliarium contectis pisci- rq
nis excipiuntur, ubi quasi respirante rivorum cursu limum de-
ponunt. modus quoque earum mensuris ibidem positis initur. .
una autem emergunt Iulia Marcia Tepula (quae intercepta, sicut ¸
supra demonstravimus, rivo Iuliae accesserat, nunc a piscina
¡ velut ego: veluti C r8.. serviunt scripsi (fit ut quaedam a. l. serviant Jo-
cundus): siunt C: fiunt a, unde fluunt ed. pr. sensim B¨ ucheler: si sint C ¸
quarum Schultz: duarum C ¡ (a)equat a: equata C 6 si ubi Polenus:
sicubi C valliuma: uallumC arcuationibusque a: arcuationibus C: ar-
cuationibusve B¨ ucheler veteris del. Polenus: arcuationibusve in is Krohn ¸
libramBOA: libramC rq.: his a: is C ¸ una C: tres B¨ ucheler emer-
gunt scripsi: earum C: lacunam ante earum ind. Krohn, vocem seclusit Grimal
eiusdem Iuliae modum accepit ac proprio canali et nomine
venit). hae tres a piscinis in eosdem arcus recipiuntur: summus ¡
rivus est Iuliae, inferior Tepulae, dein Marciae. quae ad libram ¸
[collis Vi]minalis †con[. . . .]ntea[. . .]entes† ad Viminalem
usque portamdeveniunt, ubi rursus emergunt. prius tamen pars 6,¸
Iuliae ad Spem veterem excepta castellis Caelii montis diffundi-
tur. Marcia autempartemsui post hortos Pallantianos | in rivum 8 [f.zq
qui vocatur Herculaneus deicit. hic per Caelium ductus, ipsius q
montis usibus nihil ut inferior subministrans, <f>initur supra
portam Capenam. Anio novus et Claudia a piscinis in altiores zo
arcus recipiuntur ita ut superior sit Anio. finiuntur arcus earum .
post hortos Pallantianos et inde in usumurbis fistulis diducuntur.
partem tamen sui Claudia prius in arcus qui vocantur Neroni- ¸
ani ad Spem veterem transfert. hi derecti per Caelium montem ¡
iuxta templum divi Claudii terminantur. modum quem accepe- ¸
runt aut circa ipsum montem aut in Palatium Aventinumque et
regionem Transtiberinam dimittunt. Anio vetus citra quartum zr
miliariumin tramite qui a <via>Latina in Labicanaminter arcus
traicit et ipse piscinam habet. inde intra secundum miliarium .
partem dat in specum qui vocatur Octavianus et pervenit in re-
gionemviae Novae ad hortos Asinianos, unde per illumtractum
distribuitur. rectus vero ductus secundum Spem veniens intra ¸
portamEsquilinamin altos rivos per urbemdiducitur. nec Virgo zz
nec Appia nec Alsietina conceptacula, id est piscinas, habent.
arcus Virginis initium habent sub hortis Lucullianis; finiuntur .
in campo Martio secundum frontem Saeptorum. rivus Appiae ¸
sub Caelio monte et Aventino actus emergit, ut diximus, infra
accepit C: accipit Jocundus ¡ rivus Reeve: his C: ex is Schultz: iniis P Mar-
tiae Jocundus: marcia C ¸ collis Viminalis Jocundus (viminalis tantumB
): [c. ó
litt.]minalis C c¯ o[c. ¸ litt.]nte|a[c. ¡ litt.]entes C: coniunctim infra terram euntes
Polenus: continenter una fluentes Grimal (fluentes iam B¨ ucheler, fort. accipiendum)
6 ubi a: ibi CB
q hic scripsi: se C: is B
, Dederich finitur Rubenius: initur
C zr.: in tramite scripsi: intra novie C: infra (Dederich) novum B¨ ucheler: in
transitu viae Holstenius via add. Dederich lavicanamC . octavianus
A (cf. .ç.8): octavianum C vienove C suspectum ¸ Spem <veterem> Jo-
cundus exquilinam C zz.: id est piscinas secluserim . Lucullianis
Jocundus (luculianis B
): lucilianis C ¸ rivus A: riuos C
clivumPublicii. Alsietinae ductus post naumachiam, cuius causa ¡
videtur esse factus, finitur.
Quoniam auctores cuiusque aquae et aetates, praeterea ori- zj
gines et longitudines rivorum et ordinem librae persecutus
sum, non alienum {autem modi} mihi videtur etiam singula
subicere et ostendere quanta sit copia quae publicis privatisque
non solum usibus et auxiliis verum etiam voluptatibus sufficit,
et per quot castella quibusque regionibus diducatur, quantum
extra urbem, quantum in urbe, et ex eo quantum lacibus, quan-
tum muneribus, quantum operibus publicis, quantum nomine
Caesaris, quantum privatis usibus erogetur. sed rationis exis- .
timo, priusquam nomina quinariarum centenari<ar>umque et
ceter<or>um modulorum per quos mensura constituta est pro-
feramus, {et} indicare quae sit eorum origo, quae vires et quid
quae<que> appellatio significet, propositaque regula, ad quam
ratio eorum initur et computatur, ostendere qua ratione dis-
crepantia invenerim et quam emendandi viam sim secutus.
Aquarum moduli aut ad digitorum aut ad unciarum mensu- zq
ram instituti sunt: digiti in Campania et in plerisque Italiae
locis, unciae in Apulia adhuc observa<n>tur. est autem di- .
gitus, ut convenit, sextadecima pars pedis, uncia duodecima.
quemadmodum autem inter unciam et digitum diversitas, ita ¸
et ipsius digiti duplex observatio est: alius vocatur quadratus, ¡
alius rotundus. quadratus tribus quartisdecumis suis rotundo ¸
maior, rotundus tribus undecumis suis quadrato minor est, sci-
zj licet quia anguli detrahuntur. postea modulus nec ab uncia nec
ab alterutro digitorum originem accipiens, inductus, ut quidam
zj.: auctores edd.: auctoris C origines Ursinus, Schultz (iam in marg. A): or-
dines C alienum a: alienia¯ u modi C: alieni modi Polenus diducatur
Dederich: deducatur C in urbe Schultz: in urbe C: intra urbem BOA .
centenariarumque Jocundus: centenariumque C ceterorum A: ceterum C
{et} seclusi qu(a)eque a: que C initur et dubitanter Krohn: et initium C
zq.: Apulia Ursinus, Scaliger: papula (pr. a ex p) C adhuc Heinrich: citahuc C:
ita haec ed. pr.: ita hoc Dilke observantur MV: observatur C ¸ duplex
ego: simplex C: non ante est add. Jocundus (iam supra lin. A), ante simplex B¨ ucheler
¸ detrahuntur Lanciani: deteruntur C
putant, ab Agrippa, ut alii, a plumbariis per Vitruviumarchitec-
tum, in usumurbis exclusis prioribus venit, appellatus quinariae
nomine. qui autem Agrippam auctorem faciunt dicunt quod .
quinque antiqui moduli exiles et velut puncta, quibus olimaqua
cum exigua esset dividebatur, in unam fistulam coacti sint; qui
Vitruvium et plumbarios, ab eo quod plumbea lammina plana
quinque digitorumlatitudinemhabens circumacta in rotundum
hunc fistulae modulum efficiat. sed hoc incertum est, quoniam ¸
cum circumagitur sicut interiore parte adtrahitur, ita per illam
quae foras spectat extenditur. maxime probabile est quinariam ¡
dictam a diametro | quinque quadrantum; quae ratio in se- [f.zj
quentibus quoque modulis usque ad vicenariamdurat, diametro
per singulos adiectione singulorumquadrantumcrescente: ut in
senaria quae sex quadrantes in diametro habet, et septenaria
quae septem, et deinceps simili incrementousque advicenariam.
Omnis autem modulus colligitur aut diametro aut perimetro z6
aut areae mensura, ex quibus et capacitas adparet. differentiam .
unciae, digiti quadrati et digiti rotundi, et ipsi<us> quinariae ut
facilius dinoscamus, utendum est substantia quinariae, qui mo-
dulus et certissimus et maxime receptus est. unciae ergo modu- ¸
lus habet diametri digitum unum et trientem digiti; capit plus
quam <quinaria>quinariae octava, hoc est sescuncia quinariae
et scripulis tribus et bese scripuli. digitus quadratus in rotun- ¡
dumredactus habet diametri digitumunumet digiti sescunciam
sextu<la>m; capit quinariae dextantem. digitus rotundus habet ¸
diametri digitum unum; capit quinariae septuncem semun-
ciamsextulam. ceterummoduli qui a quinaria oriuntur duobus z¡
generibus incrementum accipiunt. est unum cum ipsa<e> .
zj.: plumbariis BOA: plumbaris C quinariae B¨ ucheler: quinario C .
sint Heinrich: sunt C ¸ vicenariama: vicinariamC adiectione a: -tione
C sex] sex.s. C (sexs B¨ ucheler) z6.: areae mensura Holstenius, Polenus:
eree mensure C . ipsius a: ipsi C substantia Holstenius, Polenus: sub-
stantie C ¸–¸ repetuntur infra cap. ¡8 ¸ <quinaria> B¨ ucheler: <quinaria
plus>Grimal sescuncia Jocundus: seseuncia C ¡ sextulamJocundus: sex-
tum C z¡.: quinaria A: quinarie C . est Heinrich: et C unum
Polenus: una C: una <
> Krohn ipsae ego: ipsa C
multiplicantur, id est eodem lumine plures quinariae includun-
tur, in quibus secundum adiectionem quinariarum amplitudo
luminis crescit. est autem fere tunc in usu cum plures quina- ¸
riae impetratae, ne rivus saepius convulneretur, una fistula ex-
cipiuntur in castellum, ex quo singuli suum modum recipiunt.
alterum genus est quotiens non ad quinariarum necessitatem z8
fistula incrementum capit sed ad diametri sui mensuram, se-
cundum quod et nomen accipit et capacitatem ampliat: ut puta
quinaria, cum adiectus est ei ad diametrum quadrans, sena-
riam facit. nec iam in solidum capacitatem ampliat; capit enim .
quinariam unam et quincuncem sicilicum. et deinceps eadem ¸
ratione quadrantibus diametroadiectis, ut supradictumest, cres-
cunt septenaria, octonaria, usque ad vicenariam. subsequitur zq
illa ratio quae constat ex numero digitorum quadratorum, qui
area, id est lumine, cuiusque moduli contine<n>tur, a quibus
et nomen fistulae accipiunt. nam quae habet areae {id est lu- .
minis in rotundum coacti} digitos quadratos viginti quinque,
vicenum quinum appellatur; similiter tricenaria et deinceps per
incrementum digitorum quadratorum usque ad centenum vi-
cenum. in vicenaria fistula, quae in confinio utriusque rationis jo
posita est, utraque ratio paene congruit. nam habet, secundum .
eam computationem quae in <an>tecedentibus modulis ser-
vanda est, in diametro quadrantes viginti, cumdiametri eiusdem
digiti quinque sint; et secundum eorum modulorum rationem
qui sequuntur, aream habet digitorum quadratorum exiguo
minus viginti.
Ratio fistularum quinariarum usque ad centenum vicenum per jr
omnes modulos ita se habet, ut ostendimus, et omni genere inita
multiplicantur C: multiplicatur MV ¸ usu Schultz: us¯ u C ne rivus Jor-
dansius: neminis C saepius Jocundus: septus C z8.: necessitatem a:
necessitate C capacitatem Jocundus: capacitatis C: capacitatis <modum>
Krohn senariam Schultz: senarium C . sicilicum <scripulum> Krohn
zq.: id est lumine B¨ uchelero suspectum continentur Ursinus, Polenus: con-
tinetur C . id est . . . coacti seclusi per incrementum (ex-torum)
C: pari incremento <quinorum> B¨ ucheler (<V> Krohn) jo.. habet edd.:
habes C in antecedentibus Polenus: intecedentibus C aream Holstenius,
B¨ ucheler: adeam C
constat sibi. convenit et cum is modulis qui in commentariis .
{invictissimi et piissimi} principis positi et confirmati sunt. sive ¸
itaque ratio sive auctoritas sequenda est, utroque commenta-
riorum moduli praevalent. sed aquarii cum manifestae rationi ¡
<in> pluribus consentiant, in quattuor modulis novaverunt:
duodenaria et vicenaria et centenaria et cent<en>um vicenum.
in duodenaria quidem nec magnus error nec usus frequens est. jz
cuius diametro adiecerunt digiti semunciam sicilicum, capaci- .
tati quinariae †ebesem.† <in> reliquis autem tribus | modulis ¸ [f.zj
plus deprenditur. vicenariam exiguiorem faciunt diametro ¡
digiti semisse, capacitate quinariis tribus et semuncia. quo mo- ¸
dulo plerumque erogant. centenariam autem et centenum vi- 6
cenum, quibus adsidue accipiunt, non minuunt sed augent.
diametro enim centenariae adiciunt digiti besem et semun- ¸
ciam, capacitati quinarias decembesemsemunciam{sicilicum}.
centenumvicenumdiametro adiciunt digitos tres septuncemse- 8
munciam<sicilicum>, capacitati quinarias sexaginta sexsextan-
tem. ita dumaut vicenariae, qua subinde erogant, detrahunt aut jj
cent<en>ariae et centenum vicenum adiciunt, quibus semper
accipiunt, intercipiuntur in centenaria quinariae viginti septem,
in centenum vicenum quinariae octoginta sex {uncia}. quod .
cumratione adprobetur, re quoque ipsa manifestumest. namex ¸
vicenaria, quamCaesar pro quinariis sedecimadsignat, nonplus
erogant quam tredecim, et ex centenaria quam ampliaverunt
jr.. seclusi ¡ cum edd.: cur C <in> temptavit Polenus centenum
edd.: centumC jz.: in ego: et C duodenaria C: -ae Jocundus . ebe-
sem C: quadrantem Polenus ¸ <in> Polenus deprenditur C: deprehen-
ditur a, fort. recte ¡ semisse <et semuncia>. . . <et quadrante>et semuncia
B¨ ucheler ¸ erogant ego: erogatur C 6 centenariam scripsi: centenaria C
minuunt C: minuuntur MV augent ego: augentur C ¸ digiti bessem
Polenus: digitibus bese C: digiti plus besem Krohn semunciam <et scripu-
los duos> Ehlers besem (vel potius S =) Krohn: semissem C sicilicum hic
delevit, post semunciam infra transposuit B¨ ucheler: traditam lectionem def. Ehlers 8
centenumvicenumPolenus: centeno viceno C semunciamedd.: semuncia C
sicilicum huc transp. B¨ ucheler quinarias Jocundus: quinaria C sextantem
C: semunciam duellam Ehlers jj.: qua Polenus: quas C centenarie a:
centarie C {uncia} B¨ ucheler ¸ ex
ego: pro C: et B¨ ucheler ampli-
averunt a: amplaverunt C
aeque certum est illos non erogare nisi ad artiorem numerum,
quia Caesar secundum suos commentarios, cum ex quaque
centenaria explevit quinarias octoginta unam se[missem] (item
ex centenum vicenum quinarias nonaginta octo) tamquam ex-
hausto modulo desinit distribuere. in summa moduli sunt XX jq
quinque. omnes consentiunt et rationi et commentariis, exceptis .
his quattuor quos aquarii novaverunt. omnia autem quae men- ¸
sura continentur certa et immobilia congruere sibi debent; ita
enimuniversitati ratio constabit. et quemadmodumverbi gratia ¡
sextarii ratio ad cyathos, modii vero et ad sextarios et ad cy-
athos respondet, ita et quinariarummultiplicatio in amplioribus
modulis servare consequentiae suae regulamdebet. alioqui cum ¸
in erogatorio modulo minus invenitur, in accepto <ve>ro plus,
adparet non errorem esse sed fraudem.
Memineramus omnem aquam, quotiens ex [altiore loc]o venit jj
et intra breve spatiumin castellumcadit, non tantumrespondere
modulo suo sed etiam exuberare; quotiens vero ex humiliore, id
est minore pressura, longius ducitur, segnitia ductus modum
quoque deperdere; <e>t ideo secundum hanc rationem aut
oneranda<m> esse erogatione aut relevanda<m>. sed et posi- j6
tio habet momentum. in rectumet ad libramconlocatus modum .
servat; ad cursumaquae obpositus et devexus, id est ad haustum
pronior, amplius rapit; ad latus praetereuntis aquae conversus
et supinus, segniter et exiguum sumit. est autem calix modulus ¸
aeneus, qui rivo vel castello inditur; huic fistulae adplicantur.
eque C: eque <cent.vicenum> B¨ ucheler (ex cent. et cent. vic. quas ampl.
Schultz) semissem Polenus: se[c. ¸ litt.] C jq.¸ continentur a: conti-
netur C ¡ respondet B¨ ucheler: respondent C ¸ accepto <ve>ro scripsi:
acceptore C: acceptorio UA jj Memineramus C: -erimus edd. al-
tiore loco Ursinus, Pithoeus: [c. ¸ litt.]o C breve spatium Jocundus: breves.
At cum C ducitur B¨ ucheler: ducatur C deperdere (A) et Schultz:
deperderet C onerandam . . . relevandam ed. Basil.: -da . . . -da C
erogatione C: erogationem ed. Basil. j6.: positio C (moduli scilicet, quod sub-
audiri potest): <calicis> positio Jocundus . servat Jocundus: servavit C id
est . . . pronior huc transp. Krohn (post supinus legitur in C): interpretandi causa
(ad obp. et dev.) adscripta iudicavit B¨ ucheler segniter et B¨ ucheler: segniter
C ¸ inditur Krohn: induitur C: inducitur Ursinus, Fea
longitudinis habere debet digitos non minus duodecim, lumen, ¡
id est capacitate<m>, quanta impe<t>rata fuerit. excogitatus ¸
videtur quoniam rigor aeris difficilior ad flexum non temere
potest laxari vel coartari.
Formulas modulorumqui sunt omnes viginti et quinque subieci, j¡
quamvis in usu quindecim tantum frequentes sint, derectas ad
rationem de qua locuti sumus, emendatis quattuor quos aquarii
novaverant. secundum quod et fistulae omnes, quae opus fa- .
cient, derigi debent aut, si haec fistulae manebunt, ad quinarias
quot capient computari. qui non sint in usu moduli in ipsis est ¸
{Ed diametri trientemdigitumdici quamqui quinarie sescuncia j8
et scripulis tribus et bes scripuli. Digitus quadratus inlatitudine et .
longitudine equalis est. Digitus quadratus in rotundumredactus ¸
habet diametri digitum unum et digiti sescuncia; capit quinarie.
Digitus rotundus | habet diametri digitum unum; capit quinarie ¡ [f.z6
septuncen et semiunciam sextam.}
Fistula quinaria: diametri digitum unum <= –, perimetri> di- jq
gitos tres S = = – ∋ III; capit quinariam unam.
Fistula <senaria>: diametri digitum unum semis, perimetri di- qo
gitos IIII S = £ ∋ II; capit quinariam unam = = – ∋ <VI>.
Fistula septenaria: diametri digitum I S = –, perimetri digitos qr
V S; capit quinariam I S = = – £; in usu non est.
Fistula octonaria: diametri digitos duos, perimetri digitos sex qz
<= – ∋ X>; capit quinarias II S £ ∋ quinque.
Fistula denaria: diametri digitos duos et semis, perimetri digitos qj
septem S = = ∋ VII; capit quinarias IIII.
¡ longitudinis ego: longitudo eius C lumen C: lumine Grimal id est secl.
B¨ ucheler: eius Schultz capacitatem A: capacitate C impetrata ego: impe-
rata C ¸ temere edd.: timeri C coartari a: coortari C j¡.: subieci
Schultz: subiecti C derectas (di- Schultz) B¨ ucheler: derectam C Cap. ¡8 e
cap. :ó male repetitum ideoque hoc loco delendeum indicavit Schultz jq ‘desideramus
verbum habet’ Dederich <= –, perimetri> Polenus S = = – ∋ III Polenus
quinariamunamJocundus: quinaria una C qo <senaria>Jocundus ∋ II
B¨ ucheler: ∋ III ut vid. C quinariam unam Polenus: quinarias nouem C ∋
VI Polenus: ∋ VII B¨ ucheler qr V S Polenus: sex C I S = = – £ Polenus
qz = – (Polenus) ∋ X add. B¨ ucheler (∋ VIII Polenus) ∋ quinque C: sicilicum
(i.e. ∋ VI) Polenus qj ∋ VII B¨ ucheler: ∋ VIII C: sicilicum Polenus

Fistula duodenaria: diametri digitos <III, perimetri digitos> qq
VI<III> = = – ∋ <III>; capit quinarias quinque S = – ∋
<II>; in usu non est. {alia} apud aquarios habebat diametri .
digitos III £ ∋ VI, capacitatis quinarias sex.
Fistula quinum denum: diametri digitos III S = –, perimetri qj
digitos XI S = – ∋ X; {alia} capit quinarias novem.
Fistula vicenaria: diametri digitos quinque £∋ , perimetri digitos q6
XV S = = – ∋ VI; capit quinarias sedecim = – £. apud .
aquarios habebat diametri digitos IIII S, capacitatis quina<ria>s
Fistula vicenum quinum: diametri digitos quinque S – £ ∋ V, q¡
perimetri digitos decem et septem S = £ ∋ VI; capit quinarias
XX = = ∋ VIIII; in usu non est.
Fistula tricenaria: diametri <digitos> sex = ∋ III, perimetri q8
digitos decemet novem = = –; capit quinarias viginti quattuor
= = – ∋ quinque.
Fistula tricenum quinum: diametri digitos sex S = ∋ II, qq
perimetri digitos <XX> S = = – £ ∋ V; capit quinarias
XX<VIII S∋ III>; in usu non est.
Fistula quadragenaria: diametri digitos septem– £∋ III, perime- jo
tri digitos XXII = = – < ∋ II>; capit quinarias XXXII S –.
Fistula quadragenum quinum: diametri digitos septem S £ ∋ jr
octo, perimetri digitos XXIII S = – ∋ <X>; capit quinarias
XXXVI S <–> £ ∋ octo; in usu non est.
qq.: <III, perimetri digitos> Polenus VIIII Polenus: sex C <III>
B¨ ucheler (II Polenus) <II> B¨ ucheler . {alia} Dederich £ ∋ VI Polenus:
S ∋ V C qj III S = – Polenus XI S = – (Polenus) ∋ X B¨ ucheler (∋
VIII Polenus) {alia} edd. q6.: quinque £ ∋ B¨ ucheler: quinque Polenus
S = = – ∋ VI B¨ ucheler: bessem semuniciam Polenus: S = £ ∋ II Krohn
sedecim = – £ B¨ ucheler: XVI Polenus . IIII S post Polenum B¨ ucheler: octo z C
quinarias edd.: quinas C XIII B¨ ucheler: IIII C: XII deuncem semunciam
Polenus q¡ S = £ B¨ ucheler: ∋ VI Krohn (∋ VII ut vid. C, B¨ ucheler) XX =
= ∋ VIIII B¨ ucheler q8 digitos om. C ∋ III C: ∋ IIII Krohn qq S =
∋ II B¨ ucheler XX (viginti) add. Polenus S = = – £ ∋ V (deuncem iam
Polenus) Lindow–Krohn: ∋ IIII C XXVIII S ∋ III post Polenum B¨ ucheler jo
‘exactius < ∋ II>’ Krohn jr S £ ∋ octo Polenus XXIII Polenus: XXIIII C
∋ <X> Krohn: £ ut vid. C: duellam Polenus S <–> £ ∋ octo B¨ ucheler

Fistula quinquagenaria: diametri digitos septem S = = – £ ∋ jz
quinque, perimetri digitos XXV £ ∋ <V>IIII; capit quinarias
XL S = £ ∋ V.
Fistula quinquagenum quinum: diametri digitos octo = = ∋ jj
decem, perimetri digitos XXV<I> = – £<∋ I>; capit quinarias
XLIIII S = – £ ∋ II; in usu non est.
Fistula sexagenaria: diametri digitos octo S = £ ∋ novem, jq
perimetri digitos XXVII = = <–> £ <∋ I>; capit quinarias
XL octo S = = ∋ XI.
Fistula sexagenum quinum: diametri digitos novem – ∋ III, jj
perimetri <digitos> XX octo S –; capit quinarias quinquaginta
duas S = = – ∋ octo; in usu non est.
Fistula septuagenaria: diametri digitos novem = = <–> ∋ sex, j6
perimetri digitos XXIX S =; capit quinarias L<V>II ∋ V.
Fistula septuagenum quinum: diametri digitos novem S = – ∋ j¡
sex, perimetri digitos XXX S = £; capit quinarias LXI – ∋ II;
in usu non est.
Fistula octogenaria: diametri digitos decem – ∋ II, perimetri j8
digitos XXXI S = £ <∋ I>; capit quinarias LXV =.
Fistulaoctogenum quinum: diametri digitos decem = = £ ∋ jq
septem, perimetri digitos XXXII S = ∋ VI; <capit quinarias
LXVIIII = £ ∋ VIII>; in usu non est.
Fistula nonage<na>ria: diametri digitos decem S = ∋ X, 6o
perimetri digitos triginta tres S – £ ∋ III; capit quinarias sep-
tuaginta tres = – £ ∋ V.
jz ∋ <V>IIII Lindow–Krohn: ∋ IIII C: ∋ VII B¨ ucheler ∋ V B¨ ucheler: ∋ IIII
C jj = = ∋ decemcum Poleno B¨ ucheler: = = £ ∋ decemC XXV<I>
Polenus: XXV C < ∋ I> Krohn ∋ II B¨ ucheler: ∋ nouem C jq octo S
= £ ∋ novemPolenus: novemS Z£ ∋ octo C = = <–>£ Polenus ‘exac-
tius . . . <∋ I>’ Krohn ∋ XI B¨ ucheler: ∋ octo C jj digitos om. C duas
B¨ ucheler: duo C S = = – ∋ octo B¨ ucheler j6 = = <–> ∋ sex B¨ ucheler
L<V>II Polenus: LII C ∋ V B¨ ucheler: ∋ sex C j¡ LXI Polenus: quadrag-
inta unum C ∋ II B¨ ucheler: ∋ IIII C j8 XXXI S post Polenum B¨ ucheler:
XXXII C ‘exactius . . . <∋ I>’ Krohn jq ∋ VI Krohn: ∋ IIII C capit
quinarias LXVIIII add. Polenus = £ ∋ VIII Lindow–Krohn: = - (‘exactius =
£ ∋ XI’) B¨ ucheler 6o nonagenaria a: nonageria C S – £ ∋ III Krohn (S –
£ ∋ II iam Polenus): S = = – ∋ IX C = – £ ∋ V B¨ ucheler: = = Z ∋ IIII C

Fistula nonagenum quinum: diametri digitos X S = = – £ ∋ 6r
XI, perimetri digitos XXX<IIII> S £ <∋ V>; capit quinarias
LXXVII = = £ ∋ II; in usu non est.
Fistula centenaria: diametri digitos XI = – ∋ VIIII, 6z
perimetri digitos XXXV = = – £; capit quinarias oc-
toginta unam = = – ∋ X. apud aquarios habebat diametri .
digitos XII, capacitatis quinarias nonaginta II – £ ∋ <X>.
Fistula centenum vicenum: diametri digitos duodecim = = 6j
∋ VI, perimetri digitos XXXVIII S = =; capit quinarias
LXXXXVII S = –. | apud aquarios habebat diametri digi- [f.z6
tos XVI, capacitatis quinarias centum sexaginta tres S <=>
= – (qui modus duarum centenariarum est).
Persecutus ea quae de modulis dici fuit necessarium, nunc 6q
ponam quem {ad} modum quaeque aqua, ut principum com-
mentariis comprehensum est, usque ad nostram curam habere
visa sit quantumque erogaverit; deinde quem ipsi scrupulosa in-
quisitione praeeunte providentia {optimi} diligentissimi{que}
Nervae principis invenerimus.
Fuerunt ergo in commentariis in universo quinariarum decem .
duo milia septingentae quinquaginta quinque, in erogatione de-
cem quattuor milia decem et octo: plus in distributione quam
<in> accepto computabatur quinariis mille ducentis sexaginta
tribus. huius rei admiratio, cumpraecipuumofficii opus inexplo- ¸
randa fide aquarum atque copia crederem, non mediocriter me
convertit ad scrutandum quemadmodum amplius erogaretur
quaminpatrimonio, ut ita dicam, esset. ante omnia itaque capita ¡
ductuummetiri adgressus sum, sed longe, id est circiter quinariis
decem milibus, ampliorem quam in commentariis modum in-
veni, ut per singulas demonstrabo.
6r ‘exactius est XI’ B¨ ucheler: ∋ VIIII C XXXIIII Polenus: XXX C ∋ V
add. Krohn (sextulam add. Polenus) ∋ II B¨ ucheler: ∋ IIII C 6z.. <X>
Krohn 6j.: S = = Polenus LXXXXVII Polenus: octoginta septem C
. S <=> = – B¨ ucheler liber primus explicit. liber secundus incipit. C
6q.: quem modum Polenus: quemadmodum C optimi et -que seclusi (cf.
8,.:) . <in> edd. ¸ ut a: a¯ u C
Appiae in commentariis adscriptus est modus quinariarum 6j
octingentarum <quadraginta> unius. cuius <a>quae ad caput .
inveniri mensura nonpotuit, quoniamexduobus rivis constat. ad ¸
Gemellos tamen, qui locus est intra Spem veterem ubi iungitur
cum ramo Augustae, inveni altitudinem aquae pedum quinque,
latitudinem pedis unius dodrantis: fiunt areae pedes octo do-
drans; centenariae viginti duae et quadragenaria, quae effici-
unt quinarias mille octingentas viginti quinque: amplius quam
<in> commentariis habet quinariis nongentis octoginta quat-
tuor. erogabat quinarias septingentas quattuor: minus quam in ¡
commentariis adscribitur quinariis centum triginta septem, et
adhuc minus quam ad Gemellos mensura respondet quinariis
mille centum viginti una. intercidit tamen aliquantum e ductus ¸
vitio, qui cum sit depressio<r> non facile manationes ostendit;
quas esse ex eo adparet quod in plerisque urbis partibus probata
aqua observatur {id quod ex ea manat}. sed et quasdam fis- 6
tulas intra urbem inlicitas deprehendimus. extra urbem autem ¸
propter pressuram librae, cum sit infra terram ad caput pedibus
quinquaginta, nullam accepit iniuriam.
Anioni veteri adscriptus est incommentariis modus quinariarum 66
mille quingentarum quadraginta unius. ad caput inveni quat- .
tuor milia trecentas nonaginta octo praeter eum modum qui in
proprium ductum Tiburtium derivatur: amplius quam in com-
mentariis est quinariis duobus milibus octingentis quinquaginta
septem. eroga<ba>ntur antequam ad piscinam veniret quina- ¸
riaeducentaesexagintaduae. modus inpiscina, qui per mensuras ¡
positas initur, efficit quinariarum duo milia trecentas sexaginta
6j.: <quadraginta> Polenus . aquae B¨ ucheler: -que C ¸ intra C: infra
B¨ ucheler pedum Schultz: pedes C in (MV) commentariis habet Schultz:
commentariis abent C: commentarii habent Dederich ¸ cum edd. (quom
B¨ ucheler): quam C depressior edd.: depressio C esse var. lect. A, Polenus:
eise C: ei inesse ed. Florent. probata C dubius retinui: prolata Polenus: praebita
Jocundus {id . . . manat} seclusi: id est quae Dederich: {id} quae B¨ ucheler
¸ cum edd. (quom B¨ ucheler): quam C (cf. §¡ supra) infra terram Polenus:¯ aterr¯ a C 66.: quingentarumSchultz: quadringentarumC ¸ ero-
gabantur Est, Jocundus: erogantur C veniret C: perveniret B
(cf. ó,.¡)
quinariae a: quinariis C

duas. intercidebant ergo inter caput et piscinam quinariae ¸
mille septingentae septuaginta quattuor. erogabat post piscinam 6
quinarias mille trecentas quadraginta octo: amplius quam in
commentariis conceptionis modumsignificari diximus quinariis
sexaginta novem, minus quam recipi in ductum post piscinam
posuimus quinariis mille decemquattuor. summa quae inter ca- ¸
put et piscinam et post piscinam intercidebat quinariarum duo
milia septingentae octoginta octo. quod errore mensurae fieri
suspicarer, nisi invenissem ubi averterentur.
Marciae in commentariis adscriptus est modus quinariarum 6¡
duum milium centum sexaginta duarum. ad caput mensus in- .
veni quinarias quattuor milia sexcentas nonaginta: amplius |
quam in commentariis est quinariis duobus milibus quingen- [f.z¡
tis viginti octo. erogabantur antequam ad piscinam perveniret ¸
quinariae nonaginta quinque, et dabantur in adiutorium Tepu-
lae quinariae nonaginta duae, item <in> Anionem quinariae
centum sexaginta quattuor. summa quae erogabatur ante pisci- ¡
namquinariae trecentae quinquaginta una. modus qui inpiscina ¸
mensuris positis initur cum eo qui circa piscinam ductus eodem
canali in arcus excipitur efficit quinarias duo milia nongentas
quadraginta quattuor. summa quae aut erogatur ante piscinam 6
aut {quinarias} <in> arcus recipitur quinariarum tria milia
ducentae nonaginta quinque: amplius quam in conceptis com-
mentariorum positum est quinariis mille centum triginta tribus,
minus quam mensurae ad caput actae efficiunt quinariis mille
trecentis nonaginta quinque. erogabat post piscinam quinarias ¸
mille octingentas quadraginta: minus quam in commentariis
6 piscinam
Schultz: iisei.nam C quinariis Polenus: quinarias C ¸
quinariarum Jocundus, qui et millium scripsit: quinarias C: quinariae B¨ ucheler
6¡.¸ <in> B¨ ucheler ¡ summa a: s¯ um¯ a C ¸ cuma: quomC qui A:
quod C: quid UBO ‘ex Poleni interpretatione scribendum fuit circa piscinamduc-
tus’ Dederich: circa piscine ductum C: ‘fort. citra piscinam duct ¯ us’ Krohn ar-
cus Jocundus: arcu C nongentas BOA: nonaginta C 6 {quinarias} <in>
(Jocundus) arcus (MV) Polenus: quinarias arcuo C: ‘fort. <inde a pis>cina ri<vo>
Marciae’ Krohn ¸ octingentas a: octingenta C
conceptionis <modum> significari diximus quinariis ducentis
viginti septem, minus quam ex piscina in arcus recipiuntur
{sunt} quinariis mille centum quattuor. summa utraque quae 8
intercidebat aut inter caput et piscinamaut post piscinamquina-
riarum duo milia quingentae; quas sicut in ceteris pluribus locis
intercipi deprehendimus. non enim eas cessare manifestum est q
et ex {hoc} eo quod <ad> caput praeter eam mensuram quam
comprehendisse nos capacitate ductus posuimus effunduntur
amplius trecentae quinariae.
Tepulae in commentariis adscriptus est modus quinariarum 68
quadringentarum. huius aquae fontes nulli sunt; venis quibus- .
dam constabat, quae interceptae sunt in Iulia. caput ergo eius ¸
observandum est a piscina Iuliae. ex ea enim primum accipit ¡
quinarias centumnonaginta, deinde statimex Marcia quinarias
nonaginta duas, praeterea ex Anione novo ad hortos Epaphrodi-
tianos quinarias centum sexaginta tres. fiunt omnes quinariae ¸
quadringentae quadraginta quinque: amplius quam in com-
mentariis quinariis quadraginta quinque, quae in erogatione
Iuliae in commentariis adscriptus est modus quinariarum sex- 6q
centarum quadraginta novem. ad caput mensura in<ven>iri .
non potuit quoniamex pluribus adquisitionibus constat. sed [ad]
sextum ab urbe miliarium universa in piscinam recipitur, ubi
modus eius manifestis mensuris efficit quinarias mille ducentas
sex: amplius quam in commentariis quinariis quingentis quin-
quaginta septem. praeterea accepit prope urbem post hortos ¸
Pallantianos ex Claudia quinarias centum sexaginta duas. est ¡
omne Iuliae inacceptis: quinariae mille trecentae sexaginta octo.
ex eo dat in Tepulam quinarias centum nonaginta; erogat suo ¸
<modum>Schultz (cf. óó.ó) {sunt} Schultz quinariis Schultz: quinarie C
8 quingente C: ID B¨ ucheler praeeunte Poleno q eas edd.: eos C hoc seclusi:
eo secl. Polenus: hoc <et ex>eo Schultz <ad>Polenus 68.: quadringen-
tarum a: quadragintarum C ¸ erogatione Polenus: erogaturne C 6q..
inveniri var. lect. in ed. Basil.: iniri C ad Jocundus (spat. c. ¡ litt. C) ‘videtur
scribendum ad septimum’ B¨ ucheler ¸ accepit C: accipit B¨ ucheler
nomine octingentas tres. fiunt quas erogat quinariae nongentae 6
nonaginta tres: amplius quam in commentariis habet quinariis
trecentis quadraginta quattuor, minus quam in piscina habere
posuimus ducentis decem tribus, quas ipsas apud eos qui sine
beneficiis principis usurpabant deprehendimus.
Virgini in commentariis adscriptus est modus quinariarum sex- ¡o
centarum quinquaginta duarum. huius mensura ad caput in- .
veniri non potuit quoniam ex pluribus adquisitionibus constat
et lenior rivum intrat. prope urbem tamen ad miliarium septi- ¸
muminagro qui nunc est Ceionii Commodi, ubi velocioremiam
cursumhabet, mensuramegi quae efficit quinariarumduo milia
quingentas quattuor: amplius quam in commentariis quinariis
mille octingentis quinquaginta [du]abus. adpro | batio nostra ¡ [f.z¡
expeditissima est; erogat enim omnes quas mensura compre-
hendimus, id est duo milia quingentas quattuor.
Alsietinae conceptionis modus nec in commentariis adscriptus ¡r
est nec in re praesenti certus inveniri potuit, cum ex lacu Alsi-
etino et deinde circa Careias ex Sabatino <tantum accipiat>
quantum aquarii temperaverunt. Alsietina erogat quinarias tre- .
centas nonaginta duas.
Claudia abundantior aliis maxime iniuriae exposita est. ¡z
in commentariis habet non plus quinariis duobus milibus octin- .
gentis quinquaginta quinque, cum ad caput invenerim quina-
riarum quattuor milia sexcentas septem: amplius quam <in>
commentariis mille septingentis quinquaginta duabus. adeo ¸
autem nostr[a cer]tior est mensura ut ad septimum ab urbe
6 quinariis trecentis Schultz: quinarias trecentas C deprehendimus a: de-
preensimus C ¡o.: Virgini ed. Argent.: Virginis C . huius Ursinus,
B¨ ucheler: minus C mensura BOA: mensur¯ a C inveniri a: invenire C
potuit C: potui B¨ ucheler rivum] rivom B¨ ucheler: rivo C ¸ septimum] III
(vel II) prop. Polenus Ceionii Pithoeus: celony C iam B¨ ucheler: sam C: sane
Polenus duabus post Schultzium B¨ ucheler (. . . II iam e calculis edd.): [c. ¸ litt.]inbus
(vel nibus) C: omnibus ed.Florent. ¡ comprehendimus scripsi: deprendimus C:
deprehendimus a ¡r.: conceptionis a: conceptioni C Sabatino edd.:
abatino C <tantum>B¨ ucheler <accipiat>Schultz: <habeat>Holstenius,
Polenus ¡z.: Claudia ed. Argent.: Claudiae C . sexcenta C <in> a
¸ ante adeo spat. c. ¡ litt. C (nihil deesse videtur) nostra certior a: nostr[c. ¡
litt.]tior C
miliarium in piscina, ubi indubitatae mensurae sunt, invenia-
mus quinarias tria milia trecentas decem duas: plus quam in
commentariis quadringentis quinquaginta septem, quamvis et
ex beneficiis ante piscinam eroget et plurimum subtrahi depre-
henderimus, ideoque minus inveniatur quam revera esse debeat
quinariis mille ducentis nonaginta quinque. et circa erogationem ¡
{autem} fraus adparet, quae neque ad commentariorum fi-
dem neque ad eas quas ad caput egimus mensuras neque ad
illas saltem <quae> ad piscinam {post tot iniurias} <positae>
sunt convenit. sola<e> enim quinariae mille septingentae quin- ¸
quaginta erogantur: minus quam commentariorum ratio dat
quinariis mille centum quinque, minus autem quam mensu-
rae ad caput actae demonstraverunt quinariis duobus milibus
octingentis quinquaginta septem, minus etiam quam in piscina
invenitur quinariis mille quingentis sexaginta duabus. ideoque 6
cum sincera in urbem proprio rivo perveniret, in urbe misce-
batur cumAnione novo, ut confusione facta et conceptio earum
et erogatio esset obscurior. quod si qui forte me adquisitionum ¸
mensuris blandiri putant, admonendi sunt adeo Curtium et
Caerulumfontes aquae Claudiae sufficere adpraestandas ductui
suo quinarias quas significavi quattuor milia sexcentas septem,
ut praeterea mille sexcentae effunda<n>tur. nec eo infitias quin 8
ea quae superfluunt non sint proprie horum fontium; capiuntur
enim ex Augusta, quem inventum in Marciae supplementum
dum illa non indiget adiecimus fontibus Claudiae, quamvis ne
huius quidem ductus omnem aquam recipiat.
Anio novus in commentariis habere ponebatur quinarias tria ¡j
milia ducentas sexaginta tres. mensus ad caput repperi quinarias .
trecenta C quinaris C ¡ et del. Grimal: lacunam post et signavit Krohn
{autem} Jocundus ‘scribendum videtur saltem quae [quae iam Holstenius] ad
piscinam post tot iniurias positae sunt, nisi quis malit quae – positae post t. i.
sunt’ B¨ ucheler: quae factae ante sunt add. Krohn piscinam Polenus: piscinas
C {post tot iniurias} seclusi sunt om. MV, del. Jocundus ¸ solae a:
sola C dat a: datur C actae Ursinus, Schultz: facte C quinaris (bis) C
¸ cerolum C sexcentas] -ta C effundantur a: effundatur C 8 ea
C: eae Polenus proprie C: propriae Heinrich quem inventum C: quam
inventam Holstenius, Dederich
quattuor milia septingentas triginta octo: amplius quam in
conceptis commentariorum est quinariis mille quadringentis
septuaginta quinque. quarum adquisitionem non avide me ¸
amplecti quo alio modo manifestius probem quam quod in
erogatione ipsorum commentariorum maior pars earum con-
tinetur? erogantur enim quinariarum quattuor milia ducentae ¡
{undecim}, <cum> alioquin in eisdem commentariis inve-
niatur conceptio non amplius quam trium milium ducentarum
sexaginta trium. praeterea intercipi non tantum quingentas ¸
XXXVIII, quae inter mensuras nostras et erogationem inter-
sunt, <sed et> longe amplioremmodumdeprehendi<mus>. ex 6
quo adparet etiam exuperare comprehensam a nobis mensu-
ram: cui<us> rei ratio est quod vis aquae rapacior, ut ex largo
et celeri flumine excepta, velocitate ipsa ampliat modum.
Non dubito aliquos admiraturos quod longe maior copia ac- ¡q
tis mensuris inventa sit quam erat in commentariis principum.
cuius rei causa est error eorum qui ab initio parum diligen- .
ter uniuscuiusque fecerunt aestimationem. ac ne metu aestatis ¸
aut siccitatum in tantum a veritate eos recessisse credam, obstat
id quod ips[e actis] mensuris Iulio mense hanc uniuscuiusque
copiam quae supra scripta est tota deinceps aestate durantem
exploravi. quaecumque tamen est causa quae praecedit | illud ¡ [f.z8
utique detegitur decem milia quinariarum intercidisse, dum
beneficia sua principes secundum modum <in> commentariis
adscriptum temperant. sequens diversitas est quod alius modus ¡j
concipitur ad capita, alius nec exiguo minor in piscinis, mi-
nimus deinde distributione continetur. cuius rei causa est fraus .
¡j.¸ erogatione edd.: -tione C ¡ erogantur Holstenius, Polenus: negantur
C ducentae cum (quom iam var. lect. not. Holstenius) Poleno et Schultzio auc-
toribus Dederich: ducen undec´ C non MV: nam C ¸ XXXVIII praee-
unte Poleno Schultz: viginti septem C <sed et> B
(sed tantum Jocundus)
deprehendimus ego: deprendi C: deprehendi a 6 exuperare ego: exube-
rare C cuius MV: cui C ¡q.: admiraturos scripsi: adnotaturos C .
uniuscuiusque <aquae> Schultz ¸ obstat (Heinrich) id ego: obstantib; C:
obstat illud Grimal ips[c. ¡ litt.] C: suppl. B¨ ucheler Polenum secutus ¡ <in>
Schultz ¡j.: deinde C: denique Schultz, fort. recte in ante distrib. add. Est,
ed. Basil.
aquariorum, quos aquas ex ductibus publicis in privatorumusus
derivare deprehendimus. sed et plerique possessorum, †e quo- ¸
rum agris aqua circumducitur† {unde}, formas rivorum per-
forant. unde fit ut ductus publici hominibus privatis †vel ad
oritorum† itinera suspendant. ac de vitiis eiusmodi nec plura ¡6
nec melius dici possunt quama Caelio Rufo dicta sunt in ea con-
tione cui titulus est ‘de aquis’. quae nunc nos omnia simili licentia .
usurpata utinam non per offensas probaremus: inriguos agros,
tabernas, cenacula etiam, corruptelas denique omnes perpetuis
salientibus instructas invenimus. namquod falsis titulis aliae pro ¸
aliis aquae erogabantur etiam si inter leviora ceteris vitia, inter
ea tamen quae emendationemvidebantur exigere numerandum
est. quod fere circa montem Caelium et Aventinum accidit. qui ¡,¸¸
colles, priusquam Claudia perduceretur, utebantur Marcia et
Iulia. sed postquam Nero imperator Claudiam opere arcuato 6
ad Spem exceptam usque ad templum divi Claudii perduxit
ut inde distribueretur, priores non ampliatae sed omissae sunt.
nulla enim castella adiecit, sed isdem usus est, quorum quamvis ¸
mutata aqua vetus appellatio mansit.
Satis iam de modo cuiusque et velut nova quadam adquisitione ¡¡
aquarum et fraudibus et vitiis quae circa ea erant dictum est.
superest ut erogationem quam confertam {et}, ut sic dicam, in .
massam invenimus, immo etiam falsis nominibus positam, per
nomina aquarum, uti quaeque se habet, et per regiones urbis
digeramus. cuius comprehensionem scio non ieiunam tantum ¸
sed etiam perplexam videri posse. ponemus tamen quam bre- ¡
vissime, ne quid velut formulae officii desit. iis quibus sufficiet ¸
cognovisse summam licebit transire leviora.
¸ e secl. Mommsen circumducitur ‘vix sanum’ B¨ ucheler {unde} B¨ ucheler:
inde a: subinde Dederich oritor(um) meis ipsius oculis legi, alii ortorumviderunt,
unde ad hortorum <usum> Polenus (vel <usus> B¨ ucheler) ¡6.6 ad Spem
B¨ ucheler: ascus C: altius Polenus exceptam a: excepta C ¸ nulla a: a-
nulla C ¡¡.: aquarum a: quarum C ea C: eas ut vid. a . et om. a
mass¯ a C: massa B¨ ucheler positama: posita C ¸ comprehensionemscio
a: compreensio. | nescio C ieiunamB
: ieiunumC ¡ officii edd.: officis
C desit iis A: desitus C ¸ summ¯ a C: summa edd. verba non praeterit
. . . fuerint (quae infra c. 88.¸ leguntur) huc transferenda censuit Krohn

Fit ergo distributio quinariarum quattuordecim milium decem ¡8
et octo ita ut quinariae DCCLXXI, quae ex quibusdamaquis in
adiutoriumaliarumdantur et bis in speciemerogationis cadunt,
semel in computationemveniant. ex his dividuntur extra urbem .
quinariae quattuor milia sexaginta tres: ex quibus nomine Cae-
saris quinariae mille septingentae decem et octo, privatis quina-
riae ∞∞ CCCXXXXV. reliquae <quinariae novem> milia ¸
nongentae quinquaginta quinque intra urbem distribuebantur
in castella ducenta quadraginta septem: <ex>quibus erogaban-
tur {sub} nomine Caesaris quinariae mille septingentae septem
semis, privatis quinariae tria milia octingentae quadraginta
septem, usibus publicis quinariae quattuor milia quadringentae
una: ex eo castris †ducenti† <qui>nariae ducentae septuaginta
novem, operibus publicis nonaginta quinque quinariae ∞∞
CCCI, muneribus triginta novem quinariae CCCLXXXVI,
lacibus quingentis nonaginta uni quinariae ∞trecenta<e> tri-
ginta quinque. sed et haec ipsa dispensatio per nomina aquarum ¡
et regiones urbis partienda est.
Ex quinariis ergo quattuordecim milibus decem et octo, ¡q
quam summam erogationibus omnium aquarum <es>se
posuimus, dantur nomine Appiae extra urbem quinariae tan-
tummodo quinque, quoniam †humiliortur etia.metitoribus.†
reliquae quinariae sescentae nonaginta novem intra urbem di- .
videbantur per regiones secundamIIXVIIII XI XII XIII XIIII
in castella viginti: ex quibus nomine Caesaris quinariae centum
¡8.: Fit Schultz: Ud ( =ut) C milia C ut Schultz: et C quinariae
DCCLXXI praeeunte Poleno B¨ ucheler: quadraogen (ao in in corr.) triginta
sex.quia unus C adiutorium a: adiutori uis C veniant Schultz: veni-
unt C ¸ relique intra urb¯ e milia C: intra urbem transposuit Schultz ut ante
distribuebantur legeretur <quinariae> cur noster vocem hoc loco praetermissu-
rus fuerit non video (cf. ,ç.:, 8o,:, etc.) <novem> milium numerum (sc. VIIII)
add. Polenus <ex> Schultz sub seclusi ducentinarie C (quinariae ves-
tigia agnovit Polenus): numerum castrorum varie restituerunt (X & VIIII, i.e. decem
et novem, probabilius Polenus, X Schultz, duo<devig>inti Krohn) ducenta C
nonagintaquinque (sc. LXXXXV) Polenus: septuaginta quinque C: XCVI
Schultz quingente nonaginta unum C trecenta C ¡q.: ‘lego esse
posuimus uti Neap[olitanus codex]’ Holstenius (esse exposuimus P): seposuimus C
locus nondum sanatus

quinquaginta una, privatis quinariae centum nonaginta quat-
tuor, <usibus> publicis quinariae trecenta<e> quinquaginta
quattuor: ex eo castris I quinariae quattuor, operibus publi-
cis quattuordecim quinariae | centum viginti tres, muneri uni [f.z8
quinariae duae, lacibus nonaginta duo<bus> quinariae ducen-
tae viginti sex.
Anionis veteris erogabantur extra urbem nomine Caesaris 8o
quinariae centum sexaginta novem, privatis quinariae CCCCI-
III. reliquae quinariae mille quingentae octo semis intra urbem .
dividebantur per regiones primam III IIII V VI VII VIII VIIII
XII XIIII in castella triginta quinque: ex quibus nomine Cae-
saris quinariae sexaginta IV S, privatis quinariae CCCCXC,
<usibus> publicis quinariae quingentae LII: ex eo castris unis
quinariae quinquaginta, operibus publicis XIX quinariae cen-
tum nonaginta sex, muneribus novem quinariae octoginta octo,
lacibus nonaginta quattuor quinariae ducentae decem et octo.
Marciae erogabantur extra urbem nomine Caesaris quinariae 8r
CCLXI S <privatis quinariae
>. reliquae quinariae mille .
quadringentae septuaginta duae intra urbem dividebantur per
regiones primamtertiamquartamVVI VII VIII VIIII XXIIII
in castella quinquaginta unum: ex quibus nomine Caesaris
quinariae CXVI, privatis quinariae quingentae quadraginta
tres, <usibus publicis quinariae
: ex eo>castris IIII quinariae
XLII S, operibus publicis quindecim quinariae XLI, muneribus
XII quinariae CIIII, lacibus CXIII quinariae CCLVI.
Tepulae erogabantur extra urbem nomine Caesaris quinariae 8z
LVIII, privatis quinariae quinquaginta sex. reliquae quina- .
riae CCCXXXI intra urbem dividebantur per regiones quar-
tam V VI VII in castella XIIII: ex quibus nomine Caesaris
. una] unum C <usibus> Dederich trecenta C duae] duo C
duobus] duo C ducentas C 8o.. IV S Krohn: Ius C: VI S B¨ ucheler
<usibus> B¨ ucheler quingentae] quingentas C LII Polenus: tres C
lacibus B¨ ucheler (lacubus iam a): lacessus C 8r.: <privatis quinariae
addidi Polenum secutus . usibus publicis quinariae CCCCXXXVIIII add.
Polenus, sed numerus vix certus ex eo add. B¨ ucheler quinariae a: quinariis C
XLII S B¨ ucheler: C xii C, nisi forte xu dispici crederes 8z.. regiones
Dederich: regionem C

quinariae XXXIIII, privatis quinariae CCXXXVII, usibus
publicis quinariae quinquaginta: ex eo castris I quinariae
duodecim, operibus publicis III quinariae septem, lacibus XIII
quinariae XXXII.
Iulia fluebat extra urbem nomine Caesaris quinariis LXXX 8j
quinque, privatis quinariis CXXI. reliquae quinariae quingen- .
tae quadraginta octo intra urbem dividebantur per regiones
secundam III V VI VIII X XII in castella decem et septem:
ex quibus nomine Caesaris quinariae decem et octo, <privatis
>, usibus publicis quinariae CCCLXXXIII: ex eo
castris <
> quinariae sexaginta novem, operibus publicis <
quinariae CXXCI, muneribus <
>quinariae sexaginta septem,
lacibus viginti octo quinariae sexaginta quinque.
Virginis nomine exibant extra urbem quinariae ducentae. 8q
reliquae quinariae duo milia trecentae quattuor intra urbem .
dividebantur per regiones septimam nonam quartamdecimam
in castella decem et octo: ex quibus nomine Caesaris quina-
riae quingentae novem, privatis quinariae CCCXXXVIII, usi-
bus publicis <quinariae> ∞ centum sexaginta septem: ex eo
muneribus II quinariae XXVI, lacibus viginti quinque quina-
riae quinquaginta una, operibus publicis sedecim quinariae ∞
CCCLXXX: in quibus per se Euripo cui ipsa nomen dedit ¸
quinariae CCCCLX.
Alsietinae quinariae trecentae nonaginta duae. haec tota extra 8j
urbem consumitur: nomine Caesaris quinariae trecentae quin-
quaginta IIII, privatis quinariae centum triginta octo.
Claudia et Anio novus extra urbem proprio quaeque rivo 86
erogabantur, intra urbem confundebantur. et Claudia quidem .
8j.: Iulia fluebat C: Iuliae fluebant edd. quinariis
Krohn: quinarie C
C: quinarie a . regiones a: regionem C privatis Polenus,
quinariae Schultz addiderunt numerum CLXXXXVI add. Polenus castro-
rum numerum II prop. Krohn, III Polenus, IV Schultz operum publ. numerum XI prop.
Schultz, X Polenus muneribus <III> Polenus 8q.. quartadecimam C
<quinariae> Kunderewicz ∞ centum sexaginta septem C: MCDLVII
B¨ ucheler unum C 8j trecenta C duo C trecenta C: CC Polenus
extra urbem dabat nomine Caesaris quinarias CCXLVI,
privatis quinarias CCCCXXX novem; Anio novus nomine
Caesaris quinarias septingentas viginti octo <privatis quina-
>. reliquae utriusque quinariae tria milia quadringen- ¸
tae nonaginta octo intra urbem dividebantur per regiones
urbis XIIII in castella nonaginta duo: ex quibus nomine
Caesaris quinariae octingentae quindecim †V†, privatis quina-
riae ∞ <D> sexaginta septem, usibus publicis quinariae
∞ XV: ex eo castris novem | quinariae centum quadra- [f.zq
ginta novem, operibus publicis decem et octo quinariae
CCCLXXIIII, muneribus XII quinariae centum septem,
lacibus <C>C viginti sex quinariae CCCCXXCV.
Haec copia aquarumadNervamimperatoremusque computata 8¡
ad hunc modum discribebatur. nunc providentia diligentissimi .
principis quicquid aut fraudibus aquariorumintercipiebatur aut
inertia pervertebat<ur>quasi nova inventione fontiumadcrevit.
ac prope duplicata ubertas est, et tam sedula deinde partitione ¸
distributa, ut regionibus quibus singulae serviebant aquae plures
darentur, tamquam Caelio et Aventino, in quos sola Claudia
per arcus Neronianos ducebatur, quo fiebat ut quotiens refec-
tio aliqua intervenisset celeberrimi colles sitirent. quibus nunc ¡
plures aquae et in primis Marcia reddita amplo opere a S<p>e
in Aventinum usque perducitur. atque etiam omni parte urbis ¸
lacus tam novi quam veteres plerique binos salientes diver-
sarum aquarum acceperunt, ut si casus alterutram impedis-
set, altera sufficiente non destitueretur usus. sentit hanc curam 88
86.. quinarias a: quinarie C CCXLVI B¨ ucheler: ccxiui C quinarias a:
quinarie C quinarias septingentas edd.: quinarie.Septingente C <pri-
vatis quinarias
> post Polenum addidi ¸ quindecim.u. C: XVIII B¨ ucheler:
XVI Krohn D addendum esse vidit B¨ ucheler ∞xu C: ‘scribendum MCXII’
B¨ ucheler CC MV, Polenus: centum C CCCCXXCV Krohn: ccccxx.etcu
C: CCCCXXCII B¨ ucheler 8¡.: discribebatur B¨ ucheler: de- C intercipi-
ebatur edd.: intercapiebatur C . pervertebatur P: pervertebat C ¸ du-
plicata Holstenius, Fea: publicata C et tam B¨ ucheler: tam et C ut var. lect.
A: aut C refectio C: defectio a ¡ ‘post reddita videtur intercidisse quae’
B¨ ucheler opere a: operi C a Spe Schultz: ase C 88.: curam a: cura

{imperatoris piissimi} Nervae principis sui regina et domina
orbis in dies, {quae terrarum dea consistit, cui par nihil et nihil
secundum}, et magis sentiet<ur> salubritas eiusdem aeternae
urbis aucto castellorum, operum, munerum et lacuum numero.
nec minus ad privatos commodum ex incremento beneficiorum .
eius diffunditur; illi quoque qui timidi inlicitam aquam duce-
bant, securi nunc ex beneficiis fruuntur. ne pereuntes quidem ¸
aquae otiosae sunt: alia munditiarumfacies, purior spiritus est, et
remotae sunt causae gravioris caeli quibus apud veteres saepe urbis
infamis aer fuit. non praeterit me deberi operi novae erogationis ¡
ordinationem; sed haec cum incremento adiunxerimus, intelligi
oportet non esse ea ponenda nisi consummata fuerint.
Quid quod nec hoc diligentiae principis, quam exactissimam 8q
civibus suis praestat, sufficit parumpraesidii <usibus>ac volup-
tatibus nostris contulisse sese credentis, quod tantam copiam
adiecit, nisi eam ipsam sinceriorem iocundioremque faciat?
operae pretiumest ire per singula, per quae ille occurrendo vitiis .
quarundam universis adiecit utilitatem. etenim quando civitas ¸
nostra, cum vel exigui imbres supervenerant, non turbulentas
limosasque aquas habuit? nec quia haec universis ab origine ¡
natura est aut quia istud incommodum sentire debeant quae
capiuntur ex fontibus (inprimis Marcia et Claudia {ac reliquae})
quarum splendor a capite integer nihil aut minimum pluvia in-
quinatur, si putea exstructa et obtecta sint. duae Anienses minus qo
permanant limpidae, nam sumuntur ex flumine ac saepe etiam
sereno turbantur, quoniam Anio, quamvis purissimo defluens
{imperatoris piissimi} ego {quae . . . secundum} deleri iussit Lipsius
sentietur ego: sentiet C aeternae urbis del. B¨ ucheler ¸ alia a (alia iam
Polenus): alla C est et remotae sunt causae scripsi: et cause C, qui post fuit
habet est remotus: sunt remot(a)e OA sepe B
: se C: om.a: semper Krohn
urbis a: urbi C ‘fortasse veteres infamis urbis’ B¨ ucheler ¡ incremento
C: incrementa (et mox {ea}) Sauppe 8q.: nec ed. Argent.: ne C prae-
sidii C <usibus> B¨ ucheler: praesidiis Polenus eam ipsam a: ea ipsa C .
quarundam Schultz: quorundam C: ‘fortassse scribendum aquarum quarundam’
B¨ ucheler ¡ origine a: origine C debeant edd.: debeat C {ac reli-
quae} seclusi inquinatur Jocundus: inquinatus C obtecta Polenus (putei
obtecti iam Holstenius): obiecta (nisi ab- legas) C qo.: Anienses Schultz: an-
tensis C: aniensis ut vid. a permanant: Bergk: permanent C
lacu, mob<i>lit[ate] tamen cedentibus ripis aufert aliquid quo
turbetur priusquam deveniat in rivos. quod incommodum non .
solum hibernis ac vernis sed etiam aestivis imbribus sentit, quo
tempore exi<g>it<ur> gratior aquarum sinceritas {exigitur}.
et alter quidem ex his, id est Anio vetus, cum plerisque libra qr
sit inferior, incommodum intra se tenet. novus autem Anio vi- .
tiabat ceteras; nam cum editissimus veniat et in primis abun-
dans defectioni aliarum succurrit. imperitia vero aquariorum ¸
deducentium in alienos eum specus frequentius quam exple-
mento opus erat etiam sufficientes aquas inquinabat, maxime
Claudiam, quae per multa milia passuum proprio ducta rivo
Romae demum cum Anione permixta in hoc tempus perdebat
proprietatem. adeoque obvenientibus non succurrebat{ur}, ut ¡
pleraeque accerserentur per imprudentiam non ut {in}dignum
erat aquas par | tientium. Marciam ipsam et rigore et splendore [f.zq
gratissimam balneis ac fullonibus et relatu quoque foedis mini-
steriis deprehendimus servientem. omnes ergo discerni placuit, qz
tum singulas ita ordinari ut in primis Marcia potui tota serviret
et deinceps reliquae secundum suam quaeque qualitatem aptis
usibus adsignarentur, sic ut Anio vetus pluribus ex causis, quo
inferior excipitur minus salubris, in hortorum rigationem atque
in ipsius urbis sordidiora exiret ministeria.
Nec satis fuit principi nostro ceterarumrestituisse copiamet gra- qj
tiam; Anionis quoque novi vitia excludi posse vidit. omisso enim .
flumine <aquam> repeti ex lacu qui est super villam Nero-
nianam Sublaquensem, ubi limpidissima est, iussit. nam cum ¸
oriatur Anio supra Trebam Augustam, seu quia per saxosos
mobilitate ego: moblib|[c. ¸ litt.] C: mobilibus (Jocundus) aquis Grimal: mollibus
Ursinus, Dederich . exigitur scripsi: exiit C: vocem del. Dederich: ex<imie, licet
nunquam non> sit Krohn: scilicet Walther {exigitur} ego qr.¡ obve-
nientibus . . .] ‘fere op<tima quaeque usibus con>venientibus’ Krohn, qui et
obelum adposuit succurrebat a: succurrebatur C: secernebatur Krohn ut
{in}dignum ego: ud ´dign¯ u C: uti dignum edd. ¸ et rigore et splendore
scripsi: frigore (P. Manutius, Schultz) et splendore Krohn: splendore (ex -ere) ´ C foedis edd.: fideis C qz inferior Holstenius, Dederich:
interior C qj.. <aquam> Heinrich limpidissima C: -mus Schultz
¸ Trebam edd.: tribam C
montes decurrit paucis circa ipsumoppidumobiacentibus cultis,
seu quia lacuum altitudine in quos excipitur velut defaecatur,
imminentium quoque nemorum opacitate inumbratus, frigidis-
simus simul ac splendidissimus eo pervenit. haec tam felix pro- ¡
prietas aquae omnibus dotibus aequatura Marciam, copia vero
superatura, veniet in locum deformis illius ac turbidae, novum
auctoremimperatoremCaesaremNervam{Traianum} Augus-
tum praescribente titulo.
Sequitur ut indicemus quod ius ducendae tuendaeque sit aquae, qq
quorum alterum ad cohibendos intra modum impetrati bene-
ficii privatos, alterum ad ipsorum ductuum pertinet tutelam. in .
quibus dum altius repeto leges de singulis datas, quaedam apud
veteres aliter observata inveni<o>. apud antiquos omnis aqua ¸
in usus publicos eroga<ba>tur et cautum ita fuit: ‘ne quis pri-
vatus aliam ducit<o> quam quae ex lacu humum accidit’ (haec
enim sunt verba eius legis), id est quae ex lacu abundavit; eam
nos caducam vocamus. et haec ipsa non in alium usum quam in ¡
balnearum aut fullonicarum dabatur, eratque vectigalis statuta
mercede quae in publicumpenderetur. aliquid et in domos prin- ¸
cipum civitatis dabatur concedentibus reliquis. ad quem autem qj
magistratum ius dandae vendendaeve aquae pertinuerit in iis
ipsis legibus variatur. interdum enim ab aedilibus, interdum a .
censoribus permissum invenio; sed apparet quotiens in re pub-
lica censores erant, ab illis potissimum petitum, cum autem non
erant, aedilium eam potestatem fuisse. ex quo manifestum est ¸
quanto potior cura maioribus communium utilitatium quam
paucis a: paucihis C: paucissimis Schultz oppidum Rubenius: oppidis (pr. p
ex b) C altitudine edd.: -ne C quos Holstenius, Polenus: quo C opa-
citate a: -te C ¡ (a)equatura A: aquatura C Traianum seclusi qq.:
quod a: qua C . datas Crook: quilata C: perlatas Polenus: aquis latas B¨ ucheler
invenio ego: inveni C ¸ antiquos C: quos Dederich omnis] omnes C
erogabatur Jocundus: erogatur C et Polenus: ea C aliam <aquam> add.
B¨ ucheler praeeunte Graevio ducito Crook: ducat C accidit Graevius: accedit
C eius legis A: et leges C: {et} legis B
, Polenus ¡ publicum pendere-
tur B
, Jocundus: publico impenderetur C qj.. autem scripsi: duodecim C:
(h)ii a ¸ ‘ex quo . . . pertineret in caput ç¸ extr. post penderetur transponenda’
B¨ ucheler
privatarumvoluptatiumfuerit, cumetiamea aqua quamprivati
ducebant ad usum publicum pertineret.
Tutelamautemsingularumaquarumlocari solitaminvenio posi- q6
tamque redemptoribus necessitatem certum numerum circa
ductus extra urbem, certuminurbe servorumopificumhabendi,
et quidem ita ut nomina quoque eorum quos habituri essent in
ministerio per quasque regiones in tabulas publicas deferrent;
eorumque operum probandorum curam fuisse penes censores,
aliquando et aediles, interdum etiam quaestoribus eam provin-
ciamobvenisse, ut apparet ex S.C. quod factumest C. Licinio et
<Q.>Fabio consulibus. quanto opere autemcurae fuerit ne quis q¡
violare ductus aquamve non concessam derivare auderet, cum
ex multis apparere potest, tum et ex hoc quod Circus Maximus
ne diebus quidem ludorum circensium nisi aedilium aut censo-
rum permissu inrigabatur. quod durasse etiam postquam res ad .
curatores transiit sub Augusto apud AteiumCapitonemlegimus.
agri vero qui aqua publica contra legem essent inrigati publica- ¸
bantur. mancipi etiam, si clam eo quem <constaret> adversus ¡
legem fecisse, multa dicebatur. in isdem legibus adiectum est ¸
ita: ‘ne quis aquam oletato dolo malo ubi publice saliet. si quis 6
oletarit, sestertiorum decem milia multa{tum} esto’. {oletato ¸
videtur esse olidam facito.} cuius rei causa aediles curules 8
iubebantur per vicos singulos ex iis qui in unoquoque vico
habitarent praediave haberent binos praeficere, quorum arbi-
tratu aqua in publico | saliret.
Primus M. Agrippa post aedilitatemquamgessit consularis ope- q8
rum suorum et munerum velut perpetuus curator fuit. qui iam .
copia permittente discripsit quid aquarum publicis operibus,
quid lacibus, quid privatis daretur. habuit et familiam propriam ¸
voluptatium] voluptatumA: voluntatiumC q6 certumin Jocundus: centum
ab C curam a: cura C penes Pithoeus: per C S. C. Pithoeus: eo C
C. Licinio et Q. Fabio coss. Polenus: clycynio fabio censoribus C
q¡.: potest a: posset C et om.a permissu a: -s¯ u C ¡ clamB¨ ucheler:
cum C constaret addendum prop. Mommsen fecisse C: fecisset Holstenius,
Dederich dicebatur ut vid. a: dicebantur C 6 multa a: multatumC ¸
{oletato . . . facito} B¨ ucheler oletito C 8 causa a: caus¯ a C q8..
discripsit B¨ ucheler: de- C quid (ante lac.) a: quo C
aquarum quae tueretur ductus atque castella et lacus. hanc Au- qq
gustus hereditate ab eo sibi relictam publicavit. post eum Q. .
Aelio Tuberone Paulo Fabio Maximo consulibus in re, quae
usque in id tempus quasi potestate acta certo iure eguit, senatus
consulta facta sunt ac lex promulgata. Augustus quoque edicto ¸
complexus est quo iure uterentur qui ex commentariis Agrip-
pae aquas haberent, tota re in sua beneficia translata. modulos ¡
etiam, de quibus dictum est, constituit et rei continendae ex-
ercendaeque curatoremfecit MessalamCorvinum, cui adiutores
dati Postumius Sulpicius praetorius et Lucius Cominius pedari-
us. insignia eis quasi magistratibus concessa, deque eorumofficio ¸
senatus consultum factum quod infra scriptum est. {S.C.}
Quod Q. Aelius Tubero Paulus Fabius Maximus cos. V. F. de iis roo
qui curatores aquarum publicarum ex consensu senatus a Cae-
sare Augusto nominati essent ornandis, D. E. R. Q. F. P. D. E. R.
I. C. placere huic ordini eos qui aquis publicis praeessent cum
eius rei causa extra urbem essent lictores binos et servos publi-
cos ternos, architectos singulos et scribas et librarios, accensos
praeconesque totidem habere quot habent ii per quos frumen-
tum plebei datur; cum autem in urbe eiusdem rei causa ali- .
quid agerent ceteris apparitoribus isdempraeterquamlictoribus
<uti>. utique quibus apparitoribus ex hoc senatus consultocura- ¸
toribus aquarum uti liceret eos diebus decem [pr]oximis quibus
senatus consultum factum esset ad aerarium deferrent; quique
ita delati essent iis praetores aerarii mercedem cibaria quanta
qq.: hereditate U: ereditati C . consulibus edd.: consuli C in re que C:
cumres B¨ ucheler eguit Dederich: eguisse C: eguisset edd. consulta Polenus:
consulto C facta Polenus (iamP. Manutius, Ursinus): acta C ¸ tota a: tuta C
¸ deque a: de quo C consultum a (c. tantum UBO): consulto C S. C.
suspectum Poleno del. Dederich roo.: Quod Q. a: Quodq(ue) C(et sic saepius: hic
semel moneo) Paulus Pighius (P. iama): pullus C cos.] consul C ornan-
dis Bergk: ordinandis C I. Jocundus: E. C eius a: ius C scribas {et}
librarios Mommsen . agerent a: aget (aut ageret aut agent explicari potest) C
<uti> Brissonius ¸ proximis a: [c. : litt.]oximis C consulto C de-
ferrent Pithoeus: deferenti C delati a: dilati C pretores a: pretoris C
mercedem Casaubon: mercede C: mercedes Brissonius
praefecti frumento dando dare deferreque solent annua darent
et adtribuerent; isque eas pecunias sine fraude sua capere liceret.
utique tabulas chartas ceteraque quae eius curationis causa opus ¡
essent iis curatoribus {praebenda} Q. Aelius Paulus Fabius cos.
ambo[alte]rve si is videbitur [adhi]bitis praetor[ibus] qui aerario
praesint, ea praebenda locent. itemque cum viarum curatores ror
†que frumentique† parte quarta anni publico fungerentur mi-
nisterio, ut curatores <a>quarum iudiciis vac<ar>ent privatis
apparitores et ministeria, quamvis perseveret adhuc aerarium .
in eos erogare, tamen esse curatorum videntur desisse inertia ac
segnitia non agentiumofficium. egressis autemurbemdumtaxat ¸
agendae rei causa senatus praesto esse lictores iusserat. nobis ¡
circumeuntibus rivos fides nostra et auctoritas a principe data
pro lictoribus erit.
Cumproduxerimus remad initiumcuratorum, non est alienum roz
subiungere qui post Messalam huic officio ad nos usque
Messalae successit Planco et Silio consulibus Ateius Capito. .
Capitoni [C. Asinio Pollione] C. Antistio Vetere consulibus ¸
Tarius Rufus.
Tario {et} Ser<v>io Cornelio Cethego L. Visellio Varrone ¡
consulibus M. Cocceius Nerva, divi Nervae avus, scientia etiam
iuris inlustris.
prefecti frumentoC: prefectis frumenti Bergk capere Pithoeus: facere C ¡
{praebenda} B¨ ucheler (‘praeeunte Casaubono qui alterum praebenda delevit’ ): opus
esset . . . praeberi . . . ea praebenda Brissonius: op. esset . . . praeberi, ea . . . {ea}
praebenda Dederich Aelius] elius a: eius C consul C alterve a: [c.
¸ litt.]rve C adhibitis pretoribus a: [c. ¡ litt.]bitis pretor[c. ó litt.] C ea
Brissonius: et C: voc. deleri iussit Casaubon ror.: viarum curatoresq(ue) fru-
mentiq(ue) C: u. cur. {que} frumentique Polenus: u. curatores <curatores>que
frumenti qua parte {quarta} anni Hirschfeld fungerentur scripsi: fungeban-
tur C: fungantur Dederich ut C: ea et Hirschfeld aquarum a: quarum C
vacarent Amatucci, Grimal: vacent C . erogare edd.: eroget C roz..
Planco et Silio edd.: plancus et silius C consulibus UOA: consules CB ¸
C. Asinio Pollione Polenus: [spat. c. ¡ litt.] C consule C ¡ tario a: taurio
C et del. B
, Jocundus Servio edd.: serio C
huic successit Fabio Persico L. Vitellio consulibus C. Octavius ¸
Laenati Aquila Iuliano et Nonio Asprenate consulibus M. Por- 6
cius Cato.
huic successit †post quem† Ser. Asinio Celere [Sex.] Nonio ¸
Quintiliano consulibus <

consulibus> A. Didius Gallus.
Gallo Q. Veranio | et Pompeio Longo consulibus Cn. Domitius 8 [
Afro Nerone Claudio Caesare. IIII. et Cosso Cossi f. consulibus q
L. Piso.
Pisoni Verginio Rufo et Memmio Regulo consulibus Petronius :o
Turpiliano Crasso Frugi et Laecanio Basso consulibus P. Marius. ::
Mario Luccio Telesino et Suetonio Paulino consulibus Fonteius :.
Agrippae Silio et Galerio Trachalo consulibus Vibius Crispus. :¸
Crispo Vespasiano .III. et Cocceio Nerva consulibus Pompeius :¡
Silvano <Domitiano .II.> Valerio Messalino consulibus :¸
Tampius Flavianus.
Flaviano Vespasiano .V. Tito .III. consulibus Acilius Aviola. :6
post quem imperatore Nerva .III. et Verginio Rufo .III. con- :¸
sulibus ad nos cura translata est.
¸ favio C consule C 6 Iuliano Polenus: iunianus C nonio a: nonius
C ¸ cruces adposui (post † quemKrohn): postea Polenus: post quattuor menses
H¨ ulsen (coni. ex fastis Ost.) Ser. Asinio Celere Polenus: serasinius celera C
Sex. Nonio suppl.H¨ ulsen: [c. ¸ litt.]ionio C lacunam statui gallus a: gallius
C 8 gallo q. a: galloq(ue) C veranio et pompeio a: -ius bis C Longo
edd.: longus C: fort. Long<in>o consules C :. Luccio B¨ ucheler: lucius
(sed e additum supra) C telesino V (celesino a): celesinus C suetonius
paulinus C: em. a consules C :¸ Vibius Pighius: albius C :¡ ves-
pasiano a: -us C cocceio a: cocceia C consule C :¸ Domitiano II
add. Ursinus, B¨ ucheler :6 mihi persuasum habeo aliquot curatorum nomina excidisse
:¸ imperatore Est, edd. (-em a): imperator C
Nunc quae observare curator aquarum debeat et legem sena- roj
tusque consulta ad instruendum actum pertinentia subiungam.
circa ius ducendae aquae inprivatis observanda sunt, ne quis sine .
litteris Caesaris, id est ne quis aquampublicamnon impetratam,
et ne quis amplius quam impetravit ducat. ita enim efficiemus ¸
ut modus quem adquiri diximus possit ad novos salientes et
ad nova beneficia principis pertinere. in utroque autem magna ¡
cura multiplici opponenda fraudi est: sollicite subinde ductus
extra urbem circumeundi ad recognoscenda beneficia; idem in
castellis et salientibus publicis faciendum ut sine intermissione
diebus <et noctibus> aqua fluat. quod senatus quoque consulto ¸
facere curator iubetur, cuius haec {quoque} verba sunt.
<Quod Q.> Aelius Tubero Paulus Fabius Maximus cos. V. F. roq
de numero publicorum salientium qui in urbe essent intraque
aedificia urbi coniuncta, quos M. Agrippa fecisset, Q. F. P.
D. E. R. I. C. neque augeri placere nec minui <numerum>
publicorum salientium quos nunc esse retulerunt ii quibus ne-
gotium a senatu est imperatum ut inspicerent aquas publicas
inirentque numerum salientium publicorum. itemque placere .
curatores aquarum, quos {S. C.} Caesar Augustus ex senatus
auctoritate nominavit, dare operam uti salientes publici quam
adsiduissime interdiu et noctu aquamin usumpopuli funderent.
in hoc senatus consulto crediderim adnotandum quod sena- ¸
tus tam augeri quam minui salientium publicorum numerum
vetuerit. id factum existimo quia modus aquarum quae is tem- ¡
poribus in urbem veniebant, antequam Claudia et Anio novus
perducerentur, maiorem erogationem capere non videbatur.
roj.: consulta a: consulto C . aqu(a)e a: eaque C impetravit C: im-
petraverit Schultz ¡ sol(l)icite a: sollicites C intermissione a: -ne C
<et noctibus> Sauppe: <noctibusque> Jocundus ¸ –roq.: haec – Aelius
Dederich: hec quoque verba sunt.Elius C: Quod Q. iam addiderant Brissonius et Op-
sopoeus, quod in quoque latere primus vidit Dederich roq.: consul C <nu-
merum> Jocundus nunc <CV> esse prop. B¨ ucheler . S. C. suspectum
Poleno del. Schultz publici a: -cis C ¡ is] his C: iis a videbatur a:
videbant C
Qui aquam in usus privatos deducere volet, impetrare eam de- roj
bebit et a principe epistulam ad curatorem adferre; curator
deinde beneficio Caesaris praestare maturitatemet <ad>procu-
ratorem eiusdem officii libertum Caesaris protinus scribere.
procuratorem autem primus Ti. Claudius videtur admovisse, .
postquam Anionem novum et Claudiam induxit. quid con- ¸
tineat epistula vilicis quoque notum fieri debet, ne quando ne-
glegentiam aut fraudem suam ignorantiae colore defendant.
procurator calicemeius moduli qui fuerit impetratus adhibitis li- ¡
bratoribus signari cogitet, diligenter intendat mensurarum quas
supra diximus modum, et positionis notitiam habeat, ne sit in
arbitrio libratorum interdum maioris luminis interdum minoris
pro gratia personarum calicem probare. sed nec statim ab hoc ¸
liberum subiciendi qualemcumque plumbeam fistulam permit-
tatur arbitrium, | verum eiusdem luminis quo calix signatus est [f.jr
per pedes quinquaginta, sicut senatus consulto quod subiectum
est cavetur.
Quod Q. Aelius Tubero Paulus Fabius Maximus cos. V. F. ro6
quosdam privatos ex rivis publicis aquam ducere, Q. D. E.
R. F. P. D. E. R. I. C. ne cui privato aquam ducere ex
rivis publicis liceret, utique omnes ii quibus aquae ducen-
dae ius esset datum ex castellis ducerent, animadverterent-
que curatores aquarum quibus locis intra extra urbem apte
castella privati facere possent ex quibus aquam ducerent quam
ex castello communem accepissent a curatoribus aquarum.
ne cui eorum quibus aqua daretur publica ius esset intra quin- .
quaginta pedes eius castelli ex quo aquam ducerent laxiorem
fistulam subicere quam quinariam.
in hoc S.C. dignum adnotatione est quod aquam non nisi ex ¸
roj.: <ad>scripsi libertum(ut vid.) a: liberiumC . Ti. B: titus C ¸
villicis Polenus: vilici C quoque notum fieri transposui: fieri quoque notum
C: q. f. n. Heinrich ¡ positionis Schultz: posuimus C ro6.: consul C
extra secl. B¨ ucheler: extra<ve> Sauppe . ne C: neu Dederich: neve B¨ ucheler
cui Brissonius: qui C quinquaginta a: quinonaginta C quinariam] ‘vereor
ne ab interpolatore additum sit cum intercidisset ea sententia quam cap. .o¡ extr. Frontinus
significavit’ B¨ ucheler
castello duci permittit, ne aut rivi aut fistulae publicae frequenter
Ius impetratae aquae neque heredem neque emptorem neque ro¡
ullum novum dominum praediorum sequitur. balneis quae .
publice lavarent privilegium antiquitus concedebatur ut semel
data aqua perpetuo maneret; sic ex veteribus senatus consultis ¸
cognoscimus, ex quibus unum subieci. nunc omnis aquae cum
possessore instauratur beneficium.
Quod Q. Aelius Tubero Paulus Fabius Maximus cos. V. F. ro8
constitui oportere quo iure intra [extra]que urbem duce-
rent aquas quibus adtributae essent, Q. D. E. R. F. P. D.
<E. R.> I. <C.> uti usque eo maneret adtributio aquarum,
exceptis quae in usum balinearum essent datae aut haustus
nomine, quoad idem domini possiderent id solum in quod ac-
cepissent aquam.
Cum vacare aliquae coeperunt aquae, adnuntiatur et in com- roq
mentarios redigitur, qui respiciuntur ut petitoribus ex vacuis dari
possint. has aquarii statim intercipere solebant, ut medio tem- .
pore venderent aut possessoribus praediorum aut aliis etiam.
humanius tamen visum est principi nostro, ne praedia subito ¸
destituerentur, triginta dierum spatium indulgeri, intra quod ii
ad quos res pertineret <
* * *
>. de aqua in praedia sociorum ¡
data nihil constitutuminvenio. perinde tamen observatur ac iure ¸
cautum ut, dum quis ex iis qui communiter impetraverant su-
peresset, totus modus praediis adsignatus flueret et tunc demum
renovaretur beneficiumcumdesisset quisque ex iis quibus datum
erat possidere. impetratam aquam alio quam in ea praedia in 6
quae data erit aut ex alio castello quamex quo epistula principis
ro¡.: dominum BOA: hominum C ¸ sic C: sicut a ro8 consul C
extra a: [c. ¡ litt.]qui C F. Jocundus: I. C E. add. Jocundus: R. add.
ed. Panviniana C. add. Jocundus uti usque eo B¨ ucheler: ut ius eoque C
que A: quam C quoad a: quod ad C roq.: alique BOA: alieque C
. aquarii scripsi: aquas C intercipere Polenus: intercidere C ¸ tamen
Sauppe: etiamC: autemPolenus: voc. del. Opsopoeus lacunamstatuit Polenus ¸
impetraverant ego: impetraverunt C quisque edd.: quisquam C 6 im-
petratam aquam Polenus: Impetrata aqua C erit B¨ ucheler: erat C
continebit duci palamst non oportere, sed et mandatis prohi-
betur. impetrantur autem et eae aquae quae caducae vocantur, rro
id est quae aut ex castellis aut ex manationibus fistularum. quod .
beneficiuma principibus parcissime tribui solitum. sed fraudibus
aquariorum obnoxium est; quibus prohibendis quanta cura de-
beatur ex capite mandatorum manifestum erit quod subieci.
Caducam neminem volo ducere nisi qui meo beneficio aut rrr
priorum principum habent. nam necesse est ex castellis ali- .
quam partem aquae effluere, cum hoc pertineat non solum ad
urbis nostrae salubritatem sed etiam ad utilitatem cloacarum
Explicitis quae ad ordinationem aquarum privati usus per- rrz
tinebant, non ab re est quaedam ex iis quibus circumscribi
saluberrimas constitutiones in ipso actu deprehendimus exem-
pli causa attingere. ampliores quosdam calices quam impetrati .
erant positos in plerisque castellis inveni et ex iis aliquos ne
signatos quidem. quotiens autem signatus calix excedit legiti- ¸
mam mensuram ambitio procuratoris qui eum signavit detegi-
tur. cum vero ne signatus quidem est, manifesta culpa omnium, ¡
maxime accipientis, deprehenditur, deinde vilici. in quibusdam, ¸
cum calices legitimae mensurae signati essent, statim amplioris
moduli | fistulae subiectae fuerunt, unde acciderat ut aqua [f.jr
non per legitimum spatium coercita sed per brevis angustias
expressa facile laxiorem in proximo fistulam impleret. ideoque 6
illud adhuc, quotiens signatur calix, diligentiae adiciendum est
ut fistulae quoque proximae per spatium quod S. C. compre-
hensum diximus signentur. ita demum enim vilicus, cum scierit ¸
non aliter quam signatas conlocari debere, omni carebit excu-
satione. circa conlocandos quoque calices observari oportet ut rrj
ad lineam ordinentur nec alterius inferior calix alterius superior
rro.: e(a)e a: aee C <effluunt>post castellis Jocundus: <abundant>Sauppe:
lacunam post fistularum statuit B¨ ucheler rrr.: ante caducam :o fere litt. spatium
in C: ‘deest titulus capitis’ B¨ ucheler rrz.¡ ‘deprehenditur suspectum, certe quidem
transponi debet ante maxime’ B¨ ucheler ¸ acciderat a: acciderant C 6 illud
a: illuc C ¸ excusatione a: -ne C rrj.: inferior edd.: interior C
ponatur. inferior plus trahit; superior, quia cursus aquae ab infe- .
riore rapitur, minus ducit. in quorundam fistulis ne calices qui- ¸
dem positi fuerunt. hae fistulae solutae vocantur et ut aquario ¡
libuit laxantur vel coartantur. adhuc illa aquariorum intolera- rrq
bilis fraus est: translata in novum possessorem aqua foramen
novum castello imponunt, vetus relinquunt quo venalem ex-
trahunt aquam. in primis ergo hoc quoque emendandum cu- .
ratori crediderim. non enim solum ad ipsarum aquarum cus- ¸
todiam sed etiam ad castelli tutelam pertinet, quod subinde et
sine causa foratum vitiatur. etiam ille aquariorum tollendus est rrj
reditus quem vocant puncta. longa ac diversa sunt spatia per .
quae fistulae tota meant urbe latentes sub silice. has comperi ¸
per eum qui appellabatur a punctis passim convulneratas om-
nibus in transitu negotiationibus praebuisse peculiaribus fistulis
aquam. quo efficiebatur ut exiguus modus ad usus publicos per-
veniret. quantum ex hoc modo aquae servatum sit aestimo ex ¡
eo quod aliquantum plumbi sublatis eiusmodi ramis redactum
Superest tutela ductuum, de qua priusquam dicere incipiam rr6
pauca de familia quae huius rei causa parata est explicanda
sunt. familiae sunt duae: altera publica, altera Caesaris. pub- .,¸
lica est antiquior, quam ab Agrippa relictam Augusto et ab eo
publicatam diximus; habet homines circiter ducentos quadra-
ginta. Caesaris familiae numerus est quadringentorum sexa- ¡
ginta, quamClaudius cumaquas inurbemperduceret constituit.
utraque autem familia in aliquot ministeriorum species diduci- rr¡
tur: vilicos, castellarios, circitores, silicarios, tectores aliosque
opifices. ex his aliquos extra urbem esse oportet ad ea quae .
non sunt magnae molitionis, maturum tamen auxilium viden-
tur exigere. homines in urbe circa castellorum et munerum ¸
stationes opera quaeque urgebunt, in primis ad subitos casus ut
rrj.¸ comperi a: compari C negotiationibus C: negociatoribus Jocun-
dus ¡ servatum C: sublatum Jocundus: surreptum B¨ ucheler: serivatum Krohn
sublatis Ursinus, Polenus: sublati C rr6.. ante familie .¡ fere litt. spatium C
¡ familie a: familia C rr¡.: diducitur Jocundus: de- C ¸ homines
B¨ ucheler: omnes C urbe a: urbe C
ex compluribus regionibus in quam necessitas incubuerit con-
verti possit praesidium aquarum abundantium. tam amplum ¡
numerum utriusque familiae, solitum ambitione aut neglegen-
tia praepositorum in privata opera diduci, revocare ad aliquam
disciplinam <statuimus>, et publica ministeria ita instituimus ut
pridie quid esset actura dictaremus et quid quoque die egisset
actis comprehenderetur. commoda publicae familiae ex aerario rr8
dantur, quod impendium exoneratur vectigalium reditu ad ius
aquarum pertinentium. ea constant ex locis aedificiisve quae .
sunt circa ductus aut castella aut munera aut lacus. quem redi- ¸
tum prope sestertiorum ducentorum quinquaginta milium alie-
natumac vagum, proximis vero temporibus inDomitiani loculos
conversum, iustitia divi Nervae populo restituit, nostra seduli-
tas ad certam regulam redegit, ut constaret quae essent ad hoc
vectigal pertinentia loca. Caesaris familia ex fisco accipit com- ¡
moda, unde et omne plumbum et omnes impensae ad ductus et
castella et lacus pertinentes erogantur.
Quoniam quae videbantur ad familiam pertinere exposuimus, rrq
ad tutelamductuumsicut promiseramdivertemus, remenixiore
cura dignam, cummagnitudinis Romani imperii vel praecipuum
sit indicium. multa atque ampla opera subinde nascuntur, quibus .
ante succurri debet quam magno auxilio egere incipiant,
plerumque tamen prudenti temperamento sustinenda quia
non semper opus aut facere aut ampliare quaerentibus cre-
dendum est. ideoque non solum scientia peritorum sed et ¸
proprio usu curator instructus esse debet, nec suae tan-
tum stationis architectis uti sed plurium advocare non minus
fidem quam subtilitatem, ut aestimet quae repraesentanda,
abundantium Heinrich: abundantius C ¡ ambitione aut neglegentia edd.:
ambitione aut neglegenti¯ a C <statuimus> scripsi rr8.: commoda B
Jocundus: Quom(od)a eadem manu deletum C . constant edd.: constat C ex
locis (B¨ ucheler: hortis Jocundus: solo Holstenius) aedificiisve (B
, Jocundus): exo-
lie.difficisve C ¸ alienatum Polenus: alientem C constaret edd.: cons-
tarent C ad a: ab C ¡ omne a: omne C rrq.: enixiore B
A: enixio-
rem C sit C: sint (‘scil. ductus’) prop. Krohn . nascuntur C: dilabuntur
B¨ ucheler sustinenda C: festinanda Krohn
quae differenda sint, et rursus quae per redemptores ef-
fici debeant, quae per domesticos artifices. nascuntur | opera rzo
ex his causis: aut vetustate corrumpitur quid aut impo-
tentia possessorum aut vi tempestatium aut culpa male
facti operis, quod saepius accidit in recentibus. fere aut rzr
vetustate aut vi <eae>partes ductuumlaborant quae arcuationi-
bus sustinentur aut montium lateribus adplicatae sunt, et ex ar-
cuationibus eae quae per flumen traiciuntur. ideoque haec opera .
sollicita festinatione explicanda sunt. minus iniuriae subiacent ¸
subterranea, nec gelicidiis nec caloribus exposita. vitia autem ¡
eiusmodi sunt ut aut non interpellato cursu <su>bveniatur eis
aut emendari nisi averso non possint, sicut ea quae in ipso alveo
fieri necesse est. haec duplici ex causa nascuntur: aut enimlimo rzz
concrescente, qui interdum in crustam indurescit, iter aquae
coartatur, aut tectoria corrumpuntur, unde fiunt manationes
quibus necesse est latera rivorum et substructiones vitiari. pilae .
quoque ipsae tofo exstructae sub tam magno onere labuntur.
refici quae circa alveos rivorum sunt aestate non debent, ne ¸
intermittatur usus <eo> tempore quo praecipue desideratur,
sed vere vel autumno et maxima cum festinatione, ut scilicet
ante praeparatis omnibus quam paucissimis diebus rivi cessent.
neminem fugit per singulos ductus hoc esse faciendum, ne si
plures pariter avertantur desit aqua civitati. ea quae non inter- rzj
pellato aquae cursu effici debent maxime structura constant,
quamet suis temporibus et fidelemfieri oportet. idoneumstruc- .
turae tempus est a Kalendis Aprilibus in Kalendas Novembres,
ita ut optimum sit intermittere eam partem aestatis quae nimiis
rzo his edd.: in C quid aut Krohn: aut quid C: aut imp. poss. quid cor-
rumpitur aut vet. aut vi B¨ ucheler: aut quid vet. corr. aut imp. poss. Dederich
rzr.: tempestatium add. post vi Jocundus <eae> B¨ ucheler sustinentur
a: sustinetur C eae quae edd.: ea que A: aque C ¡ aquae ante cursu
add. Heinrich cursu subveniatur a: cursubveniatur C averso Polenus:
aversa C: <aqua> aversa Dederich rzz.¸ sunt rivorum cum transp. signis C
<eo> addidi rzj.. calendis et calendas C (K- a) eam A: iam C
caloribus incandescit, quia temperamento caeli opus est ut <et
humorem>ex commodo structura conbibat et in unitatemcon-
rob<or>etur; nonminus autemsol acrior quamgelatio praecipit
materiam. nec ullum opus diligentiorem poscit curam quam ¸
quod aquae obstaturum est; fides itaque eius per singula secun-
dum legem notam omnibus sed a paucis observatam exigenda
est. illud nulli dubium esse crediderim, proximos ductus, id est rzq
qui a sexto miliario lapide quadrato consistunt, maxime custo-
diendos, quoniam et amplissimi operis sunt et plures aquas sin-
guli sustinent. quos si necesse fuerit interrumpere, maiore{m} .
parte{m} aquarumurb{i}s destituet<ur>. remedia tamen sunt ¸
et huius difficultati<s: o>pus inchoatumexcitatur ad libramde-
ficientis, alveus vero plumbatis canalibus per spatium interrupti
ductus rursus continuatur. porro quoniamfere omnes specus per ¡
privatorum agros derecti erant et difficilis videbatur futurae im-
pensae praeparatio nisi aliqua iuris constitutione succurreretur,
simul ne accessuadreficiendos rivos redemptores a possessoribus
prohiberentur, S.C. factum est quod subieci.
Quod Q. Aelius Tubero Paulus Fabius Maximus cos. V. F. de rivis rzj
specibus fornicibus <a>quae Iuliae Marciae Appiae Tepulae
Anienis reficiendis, Q. D. E. R. F. P. D. E. R. I. C. uti cum ii
rivi <specus> fornices quos Caesar Augustus se refecturum im-
pens<a s>ua pollicitus senatui est reficerentur [. . .] ex agris
<et humorem> (Heinrich) ex commodo (Sch¨ one) scripsi: ex commodi C: ex
humore commode ed. pr. unitatem B¨ ucheler: unitate C corroboretur
a: conrobetur C sol acrior quam C: quam sol acrior Dederich gela-
tio A: celatio C ¸ aquae Jocundus: eque C rzq.: sexto C: VII Polenus
singuli sustinent a: singulis sustinet C . maiore . . . destituetur scripsi:
maiore parte aquar¯ u urbis destituet C: destituent Corradinus de Allio: maior
pars aquarum urbem destituet ed. pr.: maiorem partem urbis aqua tum des-
tituet Krohn ¸ huius C: his edd.: huiusmodi Schultz difficultatis (Heinrich)
opus (Polenus): difficultatibus C alveus A, Polenus: alvea C ¡ aliqua iuris
Schultz: alicuius C rzj consul C aquae B¨ ucheler: que C anyenis
C <specus> Schultz Caesar Augustus scripsi: augustus cesar C im-
pensa sua pollicitus a: impensua sollicitus C post reficerentur spat. c. 8 litt.
C: ‘suppl. fere et necesse esset requirere’ Krohn
privatorum terram limum lapidem testam harenam ligna cete-
raque quibus ad eamremopus esset unde quaeque eorumprox-
ime sine iniuria privatorum tolli sumi <ex>portari possint viri
<boni>arbitratu aestimata darentur tollerentur sumerentur ex-
portarentur; et ad eas res omnes exportandas earumque rerum
reficiendarum causa quotiens opus esset per agros privatorum
sine iniuria eorum itinera actus paterent darentur.
plerumque autem vitia oriuntur ex impotentia possessorum, rz6
qui pluribus <modis> rivos violant. primum enim spatia quae .
circa ductus aquarum ex S. C. vacare debent aut aedificiis aut
arboribus occupant. arbores magis nocent, quarum radicibus ¸
et concamerationes et latera | solvuntur. dein vicinales vias [f.jz
agrestesque per ipsas formas derigunt. novissime aditus ad tute- ¸
lam praecludunt. quae omnia S. C. quod subieci provisa sunt. 6
Quod Q. Aelius Tubero Paulus Fabius Maximus cos. V. F. rz¡
aquarum quae in urbem venirent itinera occupari monumentis
et aedificiis et arboribus conseri, Q. F. P. D. E. R. I. C. cum ad
reficiendos rivos specusque <
* * *
>, per quae et opera publica
corrumpantur, placere circa fontes et fornices et muros utraque
ex parte quinos denos pedes patere, et circa rivos qui sub terra
essent et specus intra urbem et {extra} urbi continentia aedifi-
cia utraque ex parte quinos pedes vacuos relinqui, ita ut neque
monumentum in is locis neque aedificium post hoc tempus
ponere neque conserere arbores liceret; <et> si quae nunc es-
sent arbores intra id spatium exciderentur praeterquam si quae
villae continentes et inclusae aedificiis essent. si quis adversus .
terr¯ a lim¯ u C: terra limus Dederich lapide C: lapides B¨ ucheler testam a:
testa C <ex>portari ego: portari C possint C: possent Sauppe boni
, Jocundus: post arbitratu rasuram ¸ fere litterarum praebet C, in qua vestigia istius
boni me dispexisse non dubito rz6.: <modis> a rz¡.: Q. (quintus a):
que C consul C Q. Jocundus: que C cum del. Pithoeus lacunam statuit
B¨ ucheler per que C: iter aquae Gundermann: per<tineat, ut spatiumcirca eos
pateat neve quicquamad eos ponatur, quo impediantur a>quae Mommsen: per
quae . . . corrumpantur post conseri transp. Heinrich et del. (‘nisi ea mavis’)
B¨ ucheler quinos Pithoeus: c.uinos C (c.quinos a) {extra} urbi B¨ ucheler:
extra urbem C <et> si quae (edd.) scripsi: sique C exciderentur Polenus:
exciperentur C
ea commiserit, in singulas res poena HS dena milia essent, ex
quibus pars dimidia praemium accusatori daretur cuius opera
maxime convictus esset qui adversus hoc S. C. commisisset, pars
autem dimidia in aerarium redigeretur; deque ea re iudicarent ¸
cognoscerentque curatores aquarum.
posset hoc S.C. aequissimum videri, etiam <si> ex rei tan- rz8
tum publicae utilitate ea spatia vindicarentur, multo magis cum
maiores nostri admirabili aequitate ne ea quidemeripuerint pri-
vatis quae ad <com>modum publicum pertinebant, sed cum
aquas perducerent, si difficilior possessor in parte vendunda
fuerat, pro toto agro pecuniam intulerint et post determinata
necessaria loca rursus eum agrum vendiderint, ut in suis finibus
propriumius <tam>res publica quamprivati haberent. plerique .
tamen, non contenti occupasse{nt} fines, ipsis ductibus manus
adtulerunt. per suffossa latera passim<in>cursu<
ii qui> ius aquarum impe<t>ratum habent, quam ii qui quan-
tulacumque beneficii occasione ad expugnandos rivos abutuntur.
quid porro fieret si non universa ista diligentissima lege prohibe- ¸
rentur poenaque non mediocris contumacibus intentare<tu>r?
qu<ar>e subscripsi verba legis. ¡
T. Quin<c>tius Crispinus consul [de S. S.] populum iure rzq
rogavit populusque iure scivit in foro pro rostris aedis divi Iulii
p. K. Iulias. tribus Sergia principium fuit. pro tribu Sex. .,¸¸
. poena HS Polenus: pena hes C ¸ deque a: dique C rz8.: <si> Jo-
cundus rei tantum publice utilitate B
, Jocundus: re tantum publice utili-
tatis C cum C: autem B¨ ucheler eripuerint B
, Dederich: eripuerunt C
commodum van der Vliet: modum C <tam> ed. pr. quam privati ego:
quamprivata C: privatique B¨ ucheler . occupasse a: occupassent C suf-
fossa Heinrich: suetossa C passim <in>cursu< . . .>ius (passi in- interpr.
Schultz, ius iam a) ego: passim|cursusus C non minus ego: tam Polenus ei
qui Polenus impetratum a: imperatum C hii C quantulacumque
ed. pr.: quantulumcumque C: quantuli- B¨ ucheler occasione ed. pr.: occasionis
C rivos B¨ ucheler: n¯ u C: nunc a: fort. expugnandum rivum ¸ intentare-
tur edd.: intentarer C ¡ quare Schultz: que C: <ideo>que Grimal rzq.:
quinctius O, edd.: quintius C de S(enatus) S(ententia) post B¨ uchelerum Momm-
sen: [spat. c. ¡ litt.] C p.K.] pr(idie) K(alendas) B¨ ucheler: P.R. C . tribus
Brissonius: tribui C ¸ tribu Brissonius: tribus C
<Vibidius> L. f. Virro <primus scivit>. Quicumque post hanc ¡
legem rogatam rivos specus fornices fistulas tubulos castella la-
cus aquarum publicarum quae ad urbem <Romam> ducuntur
<ducentur>sciens dolo malo foraverit ruperit foranda rumpen-
dave curaverit peiorave fecerit quo minus eae aquae <e>arumve
quae <a>qua in urbem Romam ire cadere fluere pervenire
duci <possit> quove minus in urbe Roma et in iis locis qua
aedificia urbi continentia sunt erunt, in iis hortis praediis locis
quorum hortorum praediorum locorum dominis possessoribus
V. F. aqua data {vel} adtributa est {vel} erit, saliat distribu-
atur dividatur in castella lacus inmittatur, is populo Romano
<HS> centum milia dare damnas esto; et qui <S.> D. {a} M. ¸
quid eorum ita fecerit, id omne <re>sarcire reficere restituere
<red>aedificare <re>ponere excidere demolire damnas esto
{sine dolo malo} <e>aque omnia ita ut<i quod recte factum
esse volet> quicumque curator aquarum est erit <aut> si cura-
tor aquarum nemo erit tum is praetor qui inter cives et peregri-
nos ius dicet multa{m} pignoribus cogito exerceto, eique cura-
tori aut si curator non erit tum ei praetori eo nomine cogendi
exercendi multa<e> dicenda<e> {sunt} pignoris capiendi | ius
nomen deesse vidit Heinrich: Vibidius Dessau: Visellius B¨ ucheler: S. Sextius Gunder-
mann <primus scivit> Brissonius ¡ <Romam> et <ducentur> Schultz
peiorave Schultz: peioremve ut vid. C: puteumve temptavi eae aquae earumve
Polenus: eaeaquaerumve C quae <a>qua Crawford: quequa C: {que}qua
Polenus: quaeque Schultz: quae pars B¨ ucheler: quaqua quae Mommsen: quae
que<at> Gundermann fluere Opsopoeus, Brissonius: fluis C <possit>
B¨ ucheler: <possint> ed. pr. locis qua aedificia Mommsen: locis que edificia C:
aedificiis quae loca aedificia B¨ ucheler: {locis} aedificiis quae Jocundus V(su)
F(ructuariis) expl. Ursinus, Gundermann: possessoribusve Pithoeus vel (bis) delevi
<HS> B¨ ucheler ¸ <S> (i.e. sine) Crook qui d(olo) {a} m(alo) B¨ ucheler:
quid¯ a C <re>sarcire ego <red>aedificare <re>ponere Crawford: edifi-
care ponere C: {aedificia} reponere Heinrich excidere Mommsen: et celere C:
{e}tollere Gundermann: et delere temptavit B¨ ucheler demolire C: demoliri Gun-
dermann {sine dolo malo} seclusi eaque Mommsen: aqu(a)e C: atque a
lacunam ind. Heinrich: ut<i quod> ego, <recte – volet> Mommsen <aut>
Brissonius is a: his C dicet B¨ ucheler: dicit C multa Brissonius: mult¯ a
C exerceto Gundermann: exercito C: coercito edd. pretori A: pretorio C
cogendi exercendi Crawford: coge. Decoercenda C: cogendi coercendi Polenus
multae dicendae Jocundus: multa dicenda C {sunt} Crawford: sive Jocundus
potestasque est<o>. si quid eorum servus fecerit, dominus eius 6
HS centum milia populo <Romano> D. D. E. si qui <locu>s ¸
circa rivos specus fornices fistulas tubulos castella lacus aquarum
publicarumquae adurbemRomamducuntur {et}ducentur ter-
minatus <e>st {et}erit, ne {que}quis ineolocopost hanc legem
rogatamquid obponit<o>molit<o>obsaepit<o>figit<o>sta-
tuit<o> ponit<o> conlocat<o> [in]arat<o> serit<o>, neve in
eum quid inmittit<o> praeterquam earum <rerum> facien-
darum reponendarum causa {praeterquam} quod hac lege
licebit oportebit. qui adversus ea quid fecerit {et} adversus eum 8
siremps lex ius{su} causaque omnium rerum omnibusque esto,
atque uti esset esse{q}ve oportere<t> si is adversus hanc legem
rivum specum rupisset forassetve. quo minus in eo loco <qui q
locus><circa fontes et for{tu}ni[ces] et muros et rivos et specus
terminatus est> <erit pecus> pascere herbam fenum secare
sentes [tollere liceat quove minus in eo loco] curatores aquarum
qui nunc sunt quique erunt {circa fontes et for{tu}ni[ces] et
muros et rivos et specus terminatus est} arbores vites vepres
esto a: est C 6 R. (i.e. Romano) post populo add. B¨ ucheler D(are)
D(amnas) E(sto) praeeunte Ursino B¨ ucheler: det C: dare damnas dedito in marg.
¸ qui locus B¨ ucheler: quis C et seclusi est et erit B¨ ucheler (et del.
Grimal ): steterit C ne Schultz: neque C: neve Gundermann obponito &
cet.] imperativum modum rest. B
, Schultz [in]arato scripsi: arat C immittito
Schulz: immidtit C earum <rerum> ego: earum C (‘scilicet aquarum publi-
carum’ Gundermann): rerum(pro earum) Mommsen {praeterquam} Opsopoeus
quod C: quae <quidem>Opsopoeus 8 {et} Heinrich: et adv. eumdel. Scaliger:
ei (deleto adv. eum) Ribbeck: ei adv. eum (‘scilicet locum terminatum’) Gunder-
mann siremps lex ius Scaliger: si rempublicamex iussu C: fort. siremps <res>
lex ius omnibusque C: omninoque Ribbeck esseve oporteret Ritschl: es-
seque oportere C q qui locus B¨ ucheler praeeunte addidi circa fontes . . .
terminatus est huc transtuli (legitur in C post quique erunt) fornices et muros
Polenus: fortuni[c. ¸ litt.] et muror(um) C terminatus C: <qua>terminatum
Crawford <erit> Crawford, <pecus> ego tollere liceat (Mommsen) quove
minus (Crawford) in eo loco scripsi: [spat. c. .¡ litt. ante Curatores] C circa
fontes quique erunt cum transponendi signis C
sentes ripae maceria<e> salicta harundineta tollantur exci-
dantur effodiantur excodicentur <curent>, uti quod recte fac-
tum esse vole<n>t <E. H. L. N. R.>, eoque nomine iis pig-
noris capio multae dictio <coerciti>o <exe>rciti<o>que esto,
idque iis sine fraude sua facere licet<o> ius potestasque esto.
quo minus vites arbores quae villis aedificiis maceriisve in- :o
clusae sunt maceriae<ve> quas curatores aquarum causa cog-
nita ne demolirentur dominis permiserunt quibus inscripta in-
sculpta{q}ve essent ipsorum qui permisissent curatorum no-
mina maneant <eius> hac lege nihilum rogatur. quo minus ::
ex iis fontibus rivis specibus fornicibus aquam sumere hau-
rire iis quibuscumque curatores aquarum permiserunt permi-
serint praeterquam rota coclea machina licea<t> dum ne qui
puteus ne{q}ve foramen novum fiat eius hac lege nihilum
utilissimae legis contemptores nonnegaverimdignos poena quae rjo
intenditur. sed neglegentia longi temporis deceptos leniter revo-
cari oportuit. itaque sedulo laboravimus ut quantum in nobis .
fuit etiam ignorarentur qui erraverant. is vero qui admoniti ad ¸
indulgentiam imperatoris decucurrerunt possumus videri causa
impetrati beneficii fuisse. in reliquo<s> vero opto ne executio ¡
legis necessaria sit, cum officii fidem etiam per offensas tueri
‘fortasse maceriae’ B¨ ucheler: maceria C <curent> Crawford recte a: rec-
tumC volent Schultz: volet C <E.H.L.N.R.>Mommsen capio Dede-
rich: capto C: captio edd. multae dictio Scaliger: multa edici C <coer-
citi>o <exe>rciti<o>que esto Crawford: po.[c. : litt.]o[c. : litt.]R citi questo
C: coercitioque esto Opsopoeus liceto B
, Bruns: licet C: liceat edd. esto
a: isto C ante quominus (§.o) spat. c. .¡ litt. C: ‘sed duo haec enuntiata apte
cohaerent’ B¨ ucheler :o maceriaeve Mommsen: macerie C curatores a: cu-
ratorum C insculptave Mommsen: -que C <eius> post B¨ ucheler Crawford
rogatur Crawford: rogatio C: rogato Mommsen: rogator Scaliger :: coclea
scripsi: calice C liceat MV: licea C neve ego: neque C rogatur Craw-
ford: rogato C: rogator Scaliger rjo.. quanta C: em. a ¸ qui a: aut C
causa a: caus¯ a C ¡ reliquos ego: reliquo C: reliquum Schultz praestet
Jocundus: prestitit C: praestiterit Schultz
Iulii Frontini F.’s praenomen is known only from epigraphic sources (CIL
6..... = ILS 6o¸¡, 8.¸o66, :¸.¸¸::, :¸.¸¡¸¡, :6.¡.: see PIR
i ¸..); it is
found neither in C nor in the report of the Hersfeld codex nor yet in the text
traditions of the Strategemata or the Agrimensores. Tacitus, Hist. i\.¸q.: and Pliny,
Ep. i\.8.¸ have ‘Iulius Frontinus’; elsewhere he is simply Frontinus.
De aquaeductu urbis Romae There is no reliable evidence as to the title
F. assigned to this work. The only possible rival to C’s rubric is that reported
from the Hersfeldensis (Introd. ¸¡): de aquae ductibus quae in urbem inducuntur. H’s
heading does not inspire much confidence, for it has clearly been composed
fromthe opening chapters of the work (note especially ¸.:, ¡.¸); the ed. princeps, it
might be noted, borrowed more exactly (De aquis quae in urbeminfluunt). B¨ ucheler,
who rightly questioned even C’s authority in this instance, gave (after Heinrich)
the title De aquis urbis Romae. This title does in fact represent not only F.’s text
at ¸.:, but the plural aquae which we find consistently (from : aquarum . . .
officium), referring to the metropolitan water-system as a whole. Baldwin (:qq¡)
¡86–¸, not inappropriately, points to de aquis, the title of a speech by Caelius
Rufus (¸6.:); cf. Serv. ad Aen. i\..¡¡ (Cato speaking de aqua). But editors have
been more cautious since Krohn pointed out (pref. vi) that de aquaeductu is
found in Cod. Theod. x\.. and Cod. Just. xi.¡¸, as a heading to sections which
deal with the upkeep of aqueducts and the rights of individuals who have
received grants of public water. These topics are indeed among those covered
in F.’s work, and the succinct title is not wholly inappropriate. Still, F. himself
nowhere uses the word aquaeductus in this sense (meaning either ‘water supply’
or ‘water-rights’ – akin, it would seem, to the ancient legal term aquae ductus,
a praedial servitude (cf. Cic. Caecin. ¸¡): see ¸..n. ductus and cf. TLL s.v. aquae
ductus. C’s de aquaeductu is probably a title applied to this text in late Antiquity, for
someone in the Middle Ages might be expected to have used the plural de aquae
r–j Prologue explaining circumstances of composition and announcing the
contents. For a general discussion of the literary prologue in Latin, see Janson
(:q6¡); on this prologue, Santini (:qq.), Del Chicca (:qq¸–¸). F. adheres to the
affected modesty such prefaces required, and he lays conventional stress upon
the practical purposes for which he has collected the material. Two conven-
tions, however, are modified: (:) ‘Author’s relationship to someone requesting
the work or to whom the work is dedicated’ is here replaced by the announce-
ment that Nerva has appointed F. to a responsible office. F. addresses the
work to himself, but not necessarily because Nerva is now dead. (.) ‘Author’s
competence in the subject’ is defined as non-existent, and F.’s work becomes
The rhetorical embellishment of these prefatory remarks is a clear indi-
cation that the work was prepared for circulation of a sort (Introd. :¸). The
style is copious, even abundant. Among the particular devices may be noted
the following: (:) Elaborate tricola: Cum . . . exigat / et . . . instigent / sitque . . .
iniunctum in the first sentence, neque ullum . . . certius / aut aliter / aliudve in the
second, in both of which the third member is greatly lengthened by subor-
dinate matter (paratactically in the former, hypotactically in the latter). (.) A
bevy of alternative expressions: seu / seu, non modo / verum quoque, nescio / an,
alongside the correlative tam / quam. (¸) Synonyms used for variety (delegata /
commissae / iniunctum, adiutorum / inferiorum) alongside hendiadys (manus quaedam
et instrumentum, institutionem regulamque). (¡) Alternation between verbs and corre-
sponding abstract nouns (institueram / institutionem, nosse / notitiam); cf. diligentiam
/ diligentiore, amorem / amantiore. (¸) Chiasmus (naturalis sollicitudo / fides sedula,
ad . . . pertinens / administratum per . . . , imperitia praepositi . . . ad inferiorum usum)
and alliteration (salubritatem . . . securitatem, primum ac potissimum, succedentium . . .
successorem). Virtually nothing is put simply. There is perhaps some deliberate
attention to prose rhythm (e.g. exigat curam – ∪ – – x, seu fides sedula – ∪ – – ∪ x,
dili]gentiam modo – ∪ – ∪ x): Del Chicca (:qq¸–¸) ¸:–.. The overall effect (at
least in chapters :–.) is more tedious and ponderous than formal and grave;
Baldwin (:qq¡) ¡8¸ describes it as ‘a touch pompous and fussy’, but would for-
give this in an administrator. In chapter ¸, by contrast, the extended parataxis
is aptly suited to a recitation of the contents.
r Cum omnis res A deceptively simple opening. For the initial cum-
circumstantial clause cf. e.g. F. Str. (cum . . . accesserim), the more elaborate
series of such clauses in Vitr., and the deliberately prosaic cum tot sustineas of
Horace, Epist. ii.:.:. A beginning with ‘all’ is frequent in Aristotle, e.g., Metaph.
tóv:t, cvûpctci, Pol. tcocv tcìiv . . . sci tcocv scivcvivcv, Part. An. ttpi
tcocv ûtcpicv :t sci utûcocv, Eth. Nic. tcoc :tyvn sci tcoc utûcoc,; cf.
Sall. Cat. :, Caes. BGall. i.:.
res . . . delegata Cf. ..: delegatum officium and Pliny, Ep. x.¸.: delegato
mihi officio, 8.¸ curis delegati a vobis officii (also iii.¸.q, and cf. \.:¡..). The word
delegata establishes a tone of respectful deference, but makes explicit the special
nature of the position, a cooperative effort on the part of both the emperor
and his deputy. A high level of responsibility demands superior performance.
For the tone of the prologue as a whole, cf. Pliny, Ep. iii..o.:. sunt quidem
cuncta sub unius arbitrio, qui pro utilitate communi solus omnium curas laboresque suscepit;
quidam tamen salubri temperamento ad nos quoque velut rivi ex illo benignissimo fonte
intentiorem exigat curam For curam exigere (emphasised by hyperba-
ton) cf. :¸.¸ maiorem adsiduamque . . . exig<er>ent curam, :.¸.¸n. diligentiorem poscit
curam. I find the expression elsewhere only in Columella (i\..q.:o, \.¸.:¸,
xii.:o.6). Cura intentior occurs three times in Curtius (iii.:¸.8, i\.:¸.:, ::), twice
as often in Livy with at least as many again related collocutions, e.g. \iii.8.:
custodiae vigiliaeque et ordo stationum intentioris ubique curae erant; xrii.¸..6 dilectum
consules multo intentiore quam alias cura habebant, \.8.. cura omnium in Veiens bel-
lum intenta est (∼ neglectum Anxuri praesidium). Livy, xx\....¡ and Quint. x.:.:::
alone have intentissima cura. Note also Pliny, Ep. ii.:o.¸ tantum curae intentionisque
suscipere; Balb. Expos. et ratio omn. formarum p.q:.:¸ L (p..o¡.:o Campbell), quod
si . . . parum diligentem adhibitam curam esse credideris; as well as a few instances
in Tacitus, which imply by ironic contrast that F.’s use is a commonplace of
courteous diction: Hist. ii.6¸.. numquam ita ad curas intento Vitellio ut voluptatum
oblivisceretur; i\...: (Domitian), Ann. i\.6¸.¸ (Tiberius), xiii.¸¡ (Nero). Behind
such usage, however, lies a long history leading backward at least to Aristotle’s
otcuocïc, óvnp (e.g. Eth. Nic. iii.:::¸a.q); for a useful summary see Kunkel
me The first personsingular is consistent throughout the prologue. For F.’s
practice elsewhere see q.¸n. exclusa, .¸..n. invenerim, 6¡.:n. invenerimus, 6¡.¸n.
naturalis sollicitudo, fides sedula naturalis (gúoti, ‘by nature’) is
paired with sedula (‘attentive’, ‘deliberate’). Throughout the work F. uses these
and similar words to describe attention to public duty. See :¸.¸ nostrae sollici-
tudini, :o¸.¡ sollicite, :.:.. sollicita festinatione; :o:.¡ fides nostra, :¸o.¡ officii fides;
::8.¸ nostra sedulitas, :¸o.. sedulo (cf. 8¸.¸ sedula partitione). Note also 6¡.: scrupu-
losa inquisitione, ::q.. prudenti temperamento. In contrast to such laudable qualities,
F. points to imperitia (..:), inertia ac segnitia (:o:..), and neglegentia (::¸.¡; cf. :¸o.:),
which had informed the administration prior to his appointment. Words of
condemnation are applied with still greater force to the behaviour of lesser of-
ficials: imperitia (q:.¸), imprudentia (q:.¡), neglegentia (:o¸.¸), ambitio (::..¸, ::¸.¡),
and above all fraus (¸¡.¸, ¸..¸, ¸¸.., ¸¸.:, 8¸.., :o¸.¡, :o¸.¸, ::o.., ::¡.:). For
the tone and sentiment, one can note what are perhaps F.’s own words in De
contr. agrorum p.¡6..6–¸: Campbell: in iudicando autem mensor[em] bonum virum et
iustum agere debet neque ulla ambitione aut sordibus moveri, servare opinionem et arti et
moribus. quidam enim per imperitiam, quidam per imprudentiam peccant: totum autem hoc
iudicandi officium et hominem et artificem exigit egregium. erat aequissimum et in advocatione
eandem fidem exhiberi a mensoribus.
ad diligentiam . . . ad amorem Note the variation in diligentiore an
amantiore just below, deferential in tone because res publica outweighs commissa
res and because the comparatives are applied to enhance in the emperor those
same qualities that F. has claimed for himself. F. toys with the semantic proxim-
ity of diligere and amare, the latter being stronger in force (cf. TLL ¸.:: ::¸6.¡q).
Although traditional as a public virtue of the Romans, diligentia, devotion to
the competent performance of duty (see Iglesias, :qq:), seems at the time F. is
writing to have been something of a slogan in senatorial circles. Trajan several
times uses the word to Pliny (Ep. x..o.., 6., ¸8..; note esp. ¸8, where Sherwin-
White rightly discerns the vigour of Trajan’s personal style). F. mentions the
diligentia of the prince at 8q.: and gives him the epithet diligentissimus at 6¡.:
and 8¸... Equally important, F. requires diligentia on the part of the procurator
at :o¸.¡ (see also ::..6); and at ¸¡.. parum diligenter contains a strong reproach.
Amor patriae / rei publicae is a long established topos; for amor in relation to a task
cf. Livy, pr.:: amor negotii suscepti, Ovid, Met. x\.¸ huius amor curae.
mihi ab Nerva Augusto . . . imperatore Neither this juxtaposi-
tion nor the passive construction is unintentional. The appositional phrase,
especially with nescio, underscores a deliberate modesty while pointedly repre-
senting a special closeness between author and prince – F. was appointed in
the year q¸ (:o..:¸). Later he twice refers to Nerva as divus (:o..¡, ::8.¸), so we
take it that the work in its present form was completed after Nerva’s death in
January q8. For his termin office and the date of composition, see Introd. ¸–8.
The specific use of the name Nerva (as opposed to, say, Caesar Augustus, as in
Pliny, Pan. ¡.¸, ¸.., etc.) is perhaps to distinguish between (unnamed) Domi-
tian on the one hand and Trajan on the other. Despite the fact that the latter
was also Nerva (cf. Tac. Agr. ¸.: Nerva Caesar, then Nerva Traianus), in this
document, marked throughout by a certain immediacy, the reigning sovereign
would perhaps have been called princeps or imperator noster (but cf. q¸.:n.). –
Except for ab Roma (8..n., q.:) and ::..: ab re, this is the sole appearance of
ab before a consonant (note that ¸.¸ a Lucio is the sole instance of a before a
liquid); cf. L&S s.v., Leumann :¸¸.
aquarum iniunctum officium Aquarum officium is ‘position with re-
sponsibility for the (public) water supply’ (thus normally F.’s meaning of plural
aquae). This genitive is not frequent with officium, but it is analogous to that
with cura, munus, negotium, and the like. Officium has come to mean ‘office’ (qq.¸,
:o..:; cf. Suet. Aug. ¸¸ nova officia excogitavit), but it retains much of its original
sense (duty, responsibility). The verb iniungere (TLL ¸.:: :66q.¸¡) is synonymous
with delegare and with mandare: cf. Colum. xi.:.¸ (on choosing a vilicus) potentis-
simum est autem in eo magisterio scire et existimare quale officium et qualis labor sit cuique
iniungendus. If it retains any connotation of onus here, it would be to emphasise
that such a burden is willingly borne: Pliny, Ep. iii.:8.: officium consulatus iniunxit
mihi ut rei publicae nomine principi gratias agerem; \.:¡.. mandatum mihi officium (of
the cura alvei Tiberis et riparum et cloacarum).
ad usum. . . tum With ad usum F. conveys a commonplace (Vitr. \iii.:.:.
[aqua] est enimmaxime necessaria et ad vitamet ad delectationes et ad usumcotidianum), but
tum singles out two reasons why a reliable water supply is necessary (salubritas
and securitas). Loss of cum might be readily explained by haplography, but
(despite non . . . modo, sed . . . quoque) it is unnecessary: see OLD s.v. tum q, H–Sz
6.6. An et is reported here from the Hersfeldensis (Introd. ¸6), but the word
would bring nothing but awkwardness.
adsalubritatem. . . securitatemurbis Note the alliterative pairing:
W¨ olfflin (:88:) .8o. Salubritas involved not only the quality of drinking water
(::.:, 8q–q¸), but also the benefits derived from urban cleanliness (cf. Pliny,
Ep. x.qo.. et salubritati et amoenitati valde sitientis coloniae, q: plurimum ea res et
salubritati et voluptati eius [coloniae] collatura); see 88.¸n., :::..n. On sanitation
and health, see Cardini (:qoq), Garrison (:q.q), Squassi (:q¸¡), Scobie (:q86),
Winkelmann (:q88), Shaw (:qq6). Securitas presumably refers to the constant
threat of fire (this is not, however, the sense of subitos casus at ::¸.¸), for which a
reliable supply of water was deemedessential (cf. Tac. Ann. x\.¡¸.¡). Bothwords
may have political overtones: for securitas cf. e.g. Tac., Agr. ¸.:; for salubritas cf.
::..: saluberrimas constitutiones, also Syme (:q86) ¡¡8. And securitas, in particular,
reminds one of crises involving the public grain supply, e.g. that faced by
Claudius in ¸: cr (Tac. Ann. xii.¡¸.:); cf. Rickman (:q8o) .::.
per principes semper civitatis nostrae viros Hyperbaton empha-
sises the grandeur of sentiment. The postponed adverb is a regular feature of
F.’s style, e.g. 6.. post biennium deinde actum est, ¸.: privatorum etiam fraudibus, :8.6
suffecturus etiam altioribus locis, ¸o.¸ velociorem iam cursum, 8¸.¸ sedula deinde parti-
tione, :o..¡ scientia etiam iuris; cf. Kortz (:8q¸) .q–¸o. While the closest parallel
to our passage is Colum. civitatis nostrae principes, the common – never
trite – phrase is frequent in dignified prose, especially oratory and history,
e.g. Cic. i Verr. ¸¸, Cat. ..¸, :¸, Red. Pop. :¸, Dom. ¡., Livy, iii.:... (applied
to T. Quinctius Capitolinus et al.), xx\ii.::.:: (Q. Fabius Maximus), Pliny, HN
xxx\i.::6 (M. Scaurus pater), Sen. Controv. \, Pliny, Ep. iii.¸.¡ (Silius
Italicus), i\.¸.: (Arrius Antoninus). Very typical is its use in Suet. Aug. .q.¡.
Tacitus regularly uses primores civitatis, e.g. Hist. i.¸:.¸, iii.6¡.:, Ann. iii.6¸..,
x\..¸... Cf. also proceres civitatis Colum., Tac. Ann. xi\.¸¸.¸ (Seneca).
For discussion of the term, and the absence of any tension between princeps (the
emperor) and principes civitatis, see B´ eranger (:q¸¸) ¡:–¸, citing especially Pliny,
Ep. i\.¸.: (Arrius Antoninus). F. here makes unmistakable reference to the use-
ful services furnished by officials of the senatorial class. Fromthe establishment
of the cura aquarum under Augustus (qq.¡), the chairmanship had apparently
been reserved to consulars. Mommsen (:88¸) ii: :o¡q suggests that the curatores
aquarum ranked higher than other senatorial commissions in part because they
continued certain censorial responsibilities, but for the nature and prestige of
the post in general see :o..:n. F. lists his predecessors in chapter :o., a few of
whom are known to have been principes civitatis in the sense that phrase is used
(e.g. Suet. Tib. ¸¸, Tit. ¸..) to apply to the ‘privy council’ of amici Caesaris, on
which see, in general, Crook (:q¸¸), Amarelli (:q8¸). Note especially Vibius
Crispus (:o..:¸), whose status is described by Tacitus, Dial. 8.¸ as one who per
multos iam annos potentissimi sunt civitatis ac . . . principes in amicitia Caesaris . . . atque ab
ipso principe cum quadam reverentia diliguntur. The Republican administration was
less clearly defined, but most of the comparable duties had fallen to censors or
aediles (¸.:n., q¸); see also Helm (:q¸¡), Flores (.ooo).
civitatis C’s celuitatis apparently represents an attempt at correction (the
l deriving froma long i meant to replace e): cf. ..: praepositi] precositei C. There is
no justification for introducing archaic spellings, ceivitatis and praepositei: Haupt
(:8¡¸) ¸:..
primumac potissimum Cf. Quint. \.¸.¸ (Domitius Afer) verissime prae-
cepit primum esse in hac parte officium oratoris ut totam causam familiariter norit; quod
sine dubio ad omnia pertinet. The alliterative pair is frequent in Livy (e.g., \.:..:.,
xxi.::.6, xx\i.¡o.:): see W¨ olfflin (:88:) .¸., Santini (:qq.) q.¸.
sicut in ceteris negotiis institueram Cf. ... more iam per multa mihi
officia servato; also Str. hoc opus, sicut cetera. On F.’s career see Introd. :–¸. For
the pluperfect cf. ::q.: sicut promiseram.
nosse quod suscepi A commonplace of sorts, but aptly applied to civil
service; cf. Seneca, Dial. x.:8.¸ (to Pompeius Paulinus, praefectus annonae): satius
est vitae suae rationem quam frumenti publici nosse. Observe the first-person active,
which emphatically responds to delegata / commissa res and iniunctum officium. Cf.
Colum. i.:.. caput est in omni negotio nosse quid agendum sit.
z.r neque enim . . . The sentence clarifies the habit of nosse quod suscepi.
Knowledge of one’s task is the first step in training, the basis for decisions, the
prerequisite for shouldering responsibility. Cf. again Colum. xi.:.¡ (on vilicus)
quisquis autem destinabitur huic negotio, sit oportet idem scientissimus robustissimusque, ut
et doceat subiectos et ipse commode faciat quae praecipit. siquidem nihil recte sine exemplo
docetur aut discitur praestatque vilicum magistrum esse operariorum, non discipulum, cum
etiam de patre familiae prisci moris exemplum Cato dixerit: male agitur cum domino, quem
vilicus docet.
neque . . . ullum . . . fundatum The difficulties here centre on om-
nis (which may not even be right: C has omis) and fundatus. For the latter a’s
funda<men>tum has long seemed satisfactory (Poleni: ‘neque enim crediderim
ullum fundamentum omnis actus certius esse cognitione rei agendae seu su-
ceptae’). Seen for the conjecture that it is, however, we find it to be a less
than convincing solution, for certius (which should be its predicate) stands in
what is normally an attributive position. The transmitted fundatus looks like
the participle of fundare, a verb not inappropriate in sense, while the position of
certius suggests that it is an adverb: something ‘has been more surely grounded’
(or ‘established’). Indirect statement (with crediderim) requires the change -us
to -um (no worrisome step in a text such as this) both in the participle and in
its required noun – apparently actus ‘performance of task’ (this sense emerg-
ing from agentis at the end of the sentence; cf. :o¸.: ad instruendum actum). An
objective genitive is normal with actus (cf. TLL :: ¡¸..¸q), but context hardly
requires one here. A form of omnis (omnium or omnino?) might suit the negative
generalisation; but the manuscript reading can be otherwise interpreted, and a
generalising hominis might counter what is potentially restrictive in per principes
viros. My version can be translated ‘No <person’s> performance of duty is
more surely grounded,’ i.e. there can be no better training for any position of
administration than by knowing what the job entails. Professor Reeve rightly
observes that with my text the comparison seems rather to be ‘than the actus
just mentioned’. I acknowledge this objection, but feel that such awkwardness
and ambiguity result from F.’s own rhetorical contortions: the comparative
certius introduced (instead of alium?) to vary with aliter and aliud.
crediderim For ellipsis of esse, cf. 8.: quem quidam Tusculanum credunt, q.¸
quia . . . possessoribus relinquendam [sc. aquam] credebat, 6¡.¸ cum praecipuum officii
opus . . . crederem, :o¡.¸ crediderim adnotandum quod . . . , ::¡.. hoc . . . emendandum
curatori crediderim; but note :.¡.:.
quae facienda quaeque vitanda sint Cf. Livy, pr.:o omnis te exempli
documentum . . . intueri, inde . . . quod imitere capias, inde . . . quod vites. Del Chicca
(:qq¸–¸) ¸¡n. adduces Sen. Ep. q¸.:¸ antiqua sapientia nihil aliud quam facienda ac
vitanda praecepit, et tunc longe meliores erant viri.
tolerabili viro I can find no good parallel for the unusual use of toler-
abilis. Poleni explains: ‘Puto Frontinum intellexisse pro viro tolerabili, virum
praeditum mediocri virtute ac habilitate ad agendum; qui licet excellens non
esset, tolerari tamen posset.’ I should be more generous, taking tolerabilis as
‘honourable’, ‘respectable’ (or ‘decent’: Loeb transl.); note Tac. Hist. i\.8..
bonos imperatores voto expetere, qualiscumque tolerare. The word tolerabilis looks pri-
marily upward: the appointee will be ‘tolerable’ in the eyes of his superior for
responsible performance of the delegatum officium. It may, however, look down-
ward (or perhaps sideward) at F.’s fellow senators, members of the Roman
elite against whom a dutiful official might find it necessary to take unwelcome
actions (cf. per offensas ¸6..n., :¸o.¡n.). It is unlikely that F. would care that a
firm and judicious superintendent might also be tolerable to underlings for
assuming rather than shifting the burden which is his: the aquarii (q.6n.) are
rather the object of F.’s displeasure – precisely because they have been acting
without supervision.
adiutorum The whole staff, not merely senatorial adiutores, who might
not even have existed in F.’s day (qq.¡n.). Included are the procurator (:o¸.:),
various clerical and technical specialists (:o¸.¡, ::¸.:, ::q.¸), as well as the gangs
of workmen (::6–:¸).
imperitia praepositi . . . i[nferi]orum . . . usum For the senti-
ment cf. ::q.¸ non solum scientia peritorum sed et proprio usu curator instructus esse
debet. Here usus means ‘experience:’ OLD s.v. 6a ‘actual performance, practice’
(opp. knowledge or theory). Renaissance conjecture is markedly unimpressive
(BOA have imperitia praecessit ei a divo Nerva). Transmitted praecositei may have
come from the preceding praeceptis; Schultz’s praepositi is a splendid correction
(cf. :¸..), but the quorum which follows requires something stronger than his
illorum ( =adiutorum). For inferiorum ‘underlings’ see TLL ¸: :¸q¸.¸o. Inciden-
tally, ad inferiorum usumis nicely chiastic with imperitia praepositi. On the antithesis
between praepositus and subordinates, cf. Del Chicca (:qq6) .¸6. Santini (:qq.)
q.. cites Pl. Ti. ¡¡d¸–8: :c0:c c v0v stgcìnv ttcvcuóçcutv, c ûtic:c:cv
:t to:iv sci :cv tv nuïv tóv:cv ototc:c0v· c sci tcv :c ocuc tcpt-
ococv otnptoicv co: c ouvcûpciocvc:t, ûtci, sc:cvcnocv:t, c:i tcocv
coci sivnoti, tocv:ci ut:tyci. F.’s attitude as a public servant is very like that
we read of the Younger Cato’s quaestorship (6¸ ncr) in Plut. Cat. Min. :6.:–¸:
Eck (:q8.a) 6:, Evans (:qq¡) ¸¸ n.... Cato would not stand for office until
he had learnt its nature and authority, from both written and oral authori-
ties. Once in office, when Cato discovered that the clerical staff had grown
accustomed to taking advantage of the inexperience of the magistrates them-
selves to the point that they rather than the elected officials were effectively in
charge, he was able to put them in their place: otnpt:ci,, cttp nocv, nçicu
ypnoûci :cï, ,pcuuc:t0oi, :c utv tçtìt,ycv scscup,c0v:c, co:cú,, :c o
cucp:óvcv:c, óttipic oioóoscv.
ad ministerium tamen Punctuation has varied with editors, but the
phrase replies to the concessive etsi necessariae partes sunt (the word-order of which
specially emphasises the predicate adjective). F. regularly avoids beginning a
clause with tamen: q.., :q.¸, 6¸.¸, ¸6.¸, etc. (the single exception in this text is
:o:..). Note also Str. i\.pr. quia quamvis clara [sc. haec exempla sunt] diversae tamen
erant substantiae, :.:¸ quamvis . . . esset necessarius, ne tamen . . . corrumperetur, . . .
manus quaedamet instrumentum Hendiadys: ‘tool in the hand’ of
the craftsman.
agentis <adhibentur> The substantive use of this present participle
(TLL :: :¸q¸.¸¸) borders on ‘officialese’: Adams (:q¸¸) :.¸, :¸¸. Rather than
being a mere synonym of praepositus, here it seems to have the meaning of
(specialised) ‘performer, operator’: Del Chicca (:qq6). C has a blank space
and no mark of punctuation after agentis. Giocondo supplied esse debent, Grimal
sunt; Poleni made no addition, but punctuated after partes; Dederich inserted
sint earlier in the sentence, after tamen. None of these solutions satisfies F.’s
style: Del Chicca (:qq6) .¸6. While a form of esse can readily be understood
from the preceding sunt, the ellipse is hard to justify in this laboured prologue.
For Professor Reeve’s supplement see the similar constructions cited in TLL ::
6¸8.¸q, and note also :oo.¡ adhibitis praetoribus, :o¸.¡ adhibitis libratoribus.
z.z ad universamrem There apparently existed no convenient, up-to-
date collection of data to which a curator could refer, let alone a systematic
guide (formula) to matters that might arise in the course of his official duties.
On those records which might have been available to F. (in his own and related
bureaux), see Introd. ¡.
more iam per multa mihi officia servato Alliteration is no doubt
deliberate. Unusual is the dative of agent, perhaps for a kind of formality
(to lend a tone of responsibility?); elsewhere I find it only at Str. :.... silvam
intemptatam ante militi nostro.
velut {in hunc} corpus Cf. e.g. Vitr. \¡ quorum ex commentariis,
quae utilia esse his rebus animadverti, collecta in unum collegi [v.l. coegi] corpus; Scrib.
Larg. Comp. (ep):: in hunc librum contuli. This metaphorical use of corpus (TLL ¡:
:o.o.6.) is too well established to require an apologetic velut. Perhaps F. inserts
it here to avoid an awkwardness after the somewhat different metaphor in
ordinem, although he is not sparing of the device anywhere in this text: Baldwin
(:qq¡) ¸o¡–¸. Note velut :¸.¡, .¸.., ¸¸.:, ¡, q¸.¸, q8.:; quasi :q.:, 8¸.., qq..,
¸; tamquam :¸.¡, ¸¸.¸; ut ita dicam 6¡.¸, ut sic dicam ¸¸.., nova quadam adquisitione
¸¸.:. By contrast, it occurs but twice in Str., in close succession in the prologue ipso velut acervo rerum confuderunt legentem. . . . ut ipsum quod exigitur quasi ad
interrogatum exhibeat (elsewhere in Str. the words velut, quasi, tamquam introduce,
not surprisingly, deceptive manoeuvres). The phrase in hunc I take to be a
scribal anticipation of the same words. There is no merit whatsoever in the
reading ultra hoc in (E), pace Rubio (:q6¸) .8.
in hunc commentarium Cf. §¸ huius commentarii. Used not so much to
indicate the modest genre of the present work as to stress the convenience and
practical value of the collected data. The singular commentarius suggests that
the division into two libri (:–6¸, 6¡–:¸o) was introduced at a stage of tradition
subsequent to initial ‘publication’.
pro formula administrationis Cf. ¸¸.¡ velut formulae officii; note also
¸¸.: formulas modulorum. F. seems to be likening this work to the kind of guide-
lines familiar in legal contexts (e.g. formula cognitionis); cf. Cic. QRosc. .¡ sunt
iura, sunt formulae de omnibus rebus constitutae, ne quis aut in genere iniuriae aut <in>
ratione actionis errare possit. expressae sunt enim ex unius cuiusque damno, dolore, incom-
modo, calamitate, iniuria publicae a praetore formulae, ad quas privata lis accommodatur,
Cic.Off. iii.:q ut sine ullo errore diiudicare possimus . . . formula quaedam constituenda
est; quam si sequemur in comparatione rerum, ab officio numquam recedemus. We ought
perhaps not to press too hard to determine a precise concrete sense for F.’s
word formula, for its essential meaning is that of forma ‘sketch’ (or outline: OLD
s.v. :¸c). An apologetic velut formula occurs in Colum. i.¡.: tradita est, iii.:..:
conscriptam; cf. \.:.:¸ quasi formulis. Note also an earlier use in Varro, Rust. i.:8.:
scribens . . . ut duas formulas. Perhaps the point is partly to have something in
respicere possem The hand is that of Peter the Deacon, but these two
words seem to have been written later in a space originally left blank. For
respicere, ‘refer to, consult’, cf. :oq.: (commentarii) respiciuntur, Sic. Flacc. Cond.
agr., p.:.o..q Campbell sanctuarium Caesaris respici solet.
z.j in aliis . . . libris Cf. Str. cum hoc opus sicut cetera usus potius aliorum
quam meae commendationis causa adgressus sim.
experimenta et usum Hendiadys; for usum see §:n. Cf. ::q.¸ non solum
scientia peritorum sed et proprio usu curator instructus esse debet.
succedentium res acta est Poleni: ‘succedentium utilitati consultum
est’. For the idiom res agitur see Reeve (:q8o). Note, however, that acta est recalls
actus and agentis from §:.
fortassis F. might have felt that this form had a grander sound than
fortasse (cf. TLL 6.:: ::¡¸..6), or he might have wished to avoid hiatus (with et).
j Contents of the present work
§: primum nomina . . . influunt chapter ¡
tum, deinde per quos . . . arcuato ¸–:¸
§. post altitudinem :8–..
modulorumque <rationes> .¸–6¸
<quem modum> . . . factae sint 6¡–¸6
quantum . . . detur ¸¸–q¸
quod ius . . . inrogatae q¡–:¸o
(Resumptive remarks occur at :¸.:, .¸.:–., ¸¸.:, 6¡.:, ¸¸.:–., q¡.:, :o..:,
:o¸.:, ::..:, ::6.:, ::q.:.) For portions of the text not apparently announced in
this prologue, see below and Introd. q.
j.r ne quid . . . praetermisisse For the didactic phrase, cf. Cic. Brut.
.¡. ne quem vocalem praeterisse videamur, Quint. iii.6..8, \i.¸.:, Balbus, Expositio
formarum p..o6.¡ Campbell nequid nos praeterisse videamur, Hyg. Grom. De cond.
agr. p.qo.:. Campbell ne quid sit quod praeterisse videamur, S.H.A. Pescennius Niger q.¸
ac ne quid ex his, quae ad Pescennium pertinent, praeterisse videamur; Divus Claudius :¸..
ne[c] ea, quae scienda sunt, praeteriise videamur. Contrast F.’s deliberate selectivity in
Str. multa et transire mihi ipse permisi. Modern readers will miss in this text
any but the most incidental references to financial and economic aspects of
administration (cf. ::8.:–¡). Technical matters are also largely ignored, except
when they relate directly to the curator’s responsibility (cf. ¸¸–6, ::q–.¡). For
the drawbacks of this work as an administrative manual, see Bruun (:qq:) ¸6q–
¸:, and for a good list of technical questions for which F. offers at best tantalising
hints, see Blackman–Hodge (.oo:) :¸.–q. Onthe other hand, whether because
information was less readily available, or because its need could be satisfied
in other ways, it is important to realise that F.’s point on comprehensiveness
is one that had meaning to him and his contemporaries. He seems to include
data and texts which he expected to have, if not at his fingertips, at least in a
single succinct collection for ready reference.
in urbem {Romam} B
deletes Romam, perhaps by conjecture (on
comparison with ¡.¸?). I have deleted it because it strikes me as both awkward
and unnecessary (we have urbis alone in the opening chapter) and because it
is characteristic of the interpolations of Peter the Deacon (who added Roma
and Romani at numerous points in his transcription of Vegetius: see Introd. ¡¡);
note the intrusive {Iulii} in §¸ just a few lines below. On F.’s ‘definition’ of urbs,
see §.n. extra urbem.
ponam ‘I shall put down’ inwriting (OLDs.v.:8): 6¡.: ponam, 66.6 posuimus
(also 6¸.q, 6q.6, ¸q.:), ¸¸.¡ ponemus. Cf. also ¸:.. in commentariis positi, 6¸.6 in
conceptis commentariorum positum, ¸¸.: in commentariis ponebatur, ¸¸.. erogationem
positam, 88.¡ ea ponenda, q6 positam necessitatem. F. only seldom uses compounds:
..¸ composui, ::q.: exposuimus (cf. ¸q.:n.).
per quos . . . quoto . . . quibus . . . quantum Extensive parataxis
with repetitive interrogatives recurs also at .¸.:–.. The device serves where
modern writers would perhaps use a table, and the repetitive style functions
as an aid to the reader.
per quos . . . perducta sit Cf. .¸.: auctores cuiusque aquae et aetates.
(On AUC chronology and possible irregularities, see ¡.:n.) The auctores are
the magistrates or emperors on whose initiative and by whose agency (hence
per + accusative) the aqueducts had been built. Dates of construction were of
some administrative relevance insofar as they might indicate need for upkeep
(cf. :.on. vetustate), and names of historical figures were a typically Roman
way of identification, but F.’s antiquarian bent informs his presentation overall
(especially noticeable at ¸.¸, ¸.¸).
quoto post urbem conditam anno Quoto (interrogative: H–Sz :q¡)
introduces the third colon in a sentence laden with formulae of Roman history:
per quos (cf. Livy, pr.q per quos viros), quibus consulibus, quoto post u.c. anno.
perducta sit Ducere and its compounds are ubiquitous in this text. F.
uses the simplex verb in two meanings: (:) ‘draw water’, i.e. from the public
supply, relates always to privati – either explicitly (88.., q¡.:, ¸, q¸.:, :o¸..,
:o6.:–.) or implicitly (¸.., :o6.¸, :o8, :oq.6, :::.:). This sense represents legal
usage, as is clear from its appearance in formal documents (q¡.¸, :o6, :::),
sometimes alongside the word ius (:o6.:, :o8). (.) ‘carry or convey water’, i.e.
in a conduit, used either in the active voice (::¸.. superior [calix] minus ducit), or
the passive (:q.q [rivus Herculaneus] per Caelium ductus, ¸¸ [aqua] longius ducitur,
6¸.¸n[modus] circa piscinamductus, 8¸.¸, q:.¸ Claudia . . . ducebatur /ducta, :.q.¡, ¸);
cf. Cic. QFr. iii.:.¡ aqua per fundum eius ducenda. Of this second usage deducere is
a variant: :o.: postquam Iuliam deduxerat, q:.¸ aquariorum [aquam] deducentium in
alienos specus, :o¸.: qui aquam in usus privatos deducere volet; cf. Cic. Div. ii.6q, Tac.
Ann. xi\....¡ (also Ann. xi.:¸.. fontis aquarum deductos). Perducere ‘deliver into use,
to carry to its destination’ in virtually every case involves new construction
(6.:–¡, ¸..–¸, ::.:, :..:, :¸.¸, :8.¡, :8.¸, ¸6.¸–6, 8¸.¡, :o¡.¡, ::6.¡, :.8.:), but
see 8..n. perducebatur. Cf. also Livy, Per. q, Pliny, HN iii.¸¡, Ep. x.¸¸–8, qo–:,
Suet. Aug. ¡..:; Vitr. \iii.¸..¸ fontes perducere. Inducere bears the sense ‘introduce,
bring in’ (something new): ¸.: Appia in urbem inducta est, .¸.: modulus inductus,
:o¸.. [Claudius] Anionem novam et Claudiam induxit happens to be in a context
of novelty (F. is speaking of the appointment of a new official, the procurator).
Diducere means ‘divide and distribute’: .o.. fistulis (ablative), .:.¸ in altos rivos,
.¸.: regionibus (dative); the word is applied also to the crew of workers (::¸.:,
¡). Adducere (8.:n.) and circumducere (¸¸.¸n.) are singular occurrences.
quibus ex locis F. indicates the source of each aqueduct in general terms
(e.g. ¸.¡ in agro Lucullano, 6.¸ supra Tibur), then gives more precise directions
on how to reach it (¸.¡n.). The damaged text (beginning with a quoto) has
generally been taken with the words which precede. Giocondo’s supplement
miliario thus corresponds to F.’s references to milestones along the main roads.
The transmitted has been variously emended: the temporal force of
coepisse(n)t is inappropriate (construed absolutely by Poleni), and logic recom-
mends Grimal’s concipiatur (in the corresponding chapters F. regularly uses the
verb concipitur: ¸.¡n.). The difficulties may, however, be more serious. After lo-
cating the source, F. indicates a total length of the conduit before breaking this
down per species operum (:¸.:n.). As B¨ ucheler saw, the present passage should
contain an explicit announcement of that procedure: note .¸.: origines et longi-
tudines rivorum. For his suggested addendum cf. ¸.8, :..¸, :¸.6 (efficit passus), but
note that F. varies the formula: habet longitudinem passuum (¸.¸, 6.6, ¸.8, :¡.¡),
efficit longitudinem passuum (q.¸, ::.¡), and cf. venit per longitudinem passuum (:o.¸).
Before the first quantum F. might also have added et ex eo (‘and of this overall
length’: so 6.6, :o.¸, :¡.¡, :¸.6). (For the whole cf. §. quot castella . . . et ex is
If a quoto miliario were the transmitted reading, I should have no trouble in
defending the phrase as meaning ‘from what distance’: cf. CIL 6.:.¸6 (cited
in Appendix B, no. ¸), ILS q¸¸o flumen Sebaston ab Schedia induxit a milliario XXV
(tti o:coicu, oicscoicu,), Pliny, Ep. x.qo.: a sexto decimo miliario posse perduci
(see also TLL 8: q¡q). This would leave us with F. promising to specify the
general location of the sources and the total length of the conduit, but
would still remain unexplained (for perducta sit would be sufficient to govern
both prepositional phrases). My positioning of the daggers reflects hesitation
on two points: a quoto is highly suspicious (so close upon quoto . . . anno), and
cepisse could be concealing some corruption of passus.
<ductus cuiusque> The noun ductus in F.’s text always means ‘channel’
or ‘conduit’, whether used alone (e.g. ¸.¸, :¸.:, :o¸.¡, ::6.:), with genitive of
aqua (¸.¸, :¸.:, :.6..; cf. ¸.:, q.q proper names of aqueducts), or with pronoun
or relative (¸.¸, ¸.8, 6.6, ¸.:, :..¸). Cf. also Cic. Off. ii.:¡, 6o, Vitr. \iii.6.:,
Pliny, HN xxx\i.:.:, Pliny, Ep. :o.¸¸.:, Tac. Ann. i.¸¸.., Suet. Claud. .o.:. For
F. ductus and rivus (in the sense of ‘artificial watercourse’) are often virtually
synonymous: note ductus publici (¸¸..–¸) alongside rivi publici (:o6.¸; cf. rivi in
the S.C. of :o6.:–.), ductus longitudinem habet (¸.¸, 6.6, etc.) matches longitudines
rivorum cuiusque ductus (:¸.:; cf. .¸.:), and circa ductus (:.6..) parallels circa rivos
(:.¸.:, :.q.¸, both from legal documents). The distinction is that ductus refers
more precisely to the structure by which the water is conveyed (note ¸.¸ huius
aquae ductum, :¸.: ductus aquarum, where we have our ‘aqueduct’), while rivus can
sometimes be used close to its basic meaning ‘stream’ or ‘supply of running
water’. Variety of style results, for F. can say a faucibus ductus (:¸..) and ductus
vitio (6¸.¸), but rivorum cursu (:q.:) and rivi cessent (:...¸).
subterraneo rivo . . . substructione . . . opere arcuato Levels of
the course andthe nature of the terrainrequireddifferent types of construction:
see Vitr. \iii.¸–6. When an underground tunnel was impossible, the channel
of the aqueduct could be raised (up to five or six metres) on substructure,
and even higher on free-standing arch-work – for bridging valleys in outlying
areas (e.g. ¸.8) or, most dramatically, in the final stretch to the City (:¸.¸n.). F.’s
interest arises because upkeep and repair would naturally vary with the type
of construction (:¸..–¸, :.:.:–¸).
subterraneo rivo Even for underground channels F. uses the more gen-
eral word rivus (see note on ductus just above) far more commonly than specus
(:¸..n.). For the variety of subterranean construction possibilities, see Ashby
(:q¸¸) ¡.–¡, Hodge (:qq.) q¸–:o6, :.6–q.
substructione This word (abstract for concrete) is fairly rare except
in building contexts: Vitr. i.¸.¸, \.¸.¡, \i.8.¸–¸, \iii.¸.¸, Livy, xxx\iii..8.¸,
Pliny, HN xxx\i.:o¡, Colum. i.¸.q, Pliny, Ep. x.¸q... F. seems to use the
word in the singular to refer to uninterrupted stretches (¸.¸, 6.6, ¸.8, q.¸),
sometimes with rivorum (:o.¸, :¡.¡, :¸.6) when more than one channel is
superposed (:.¡.:). The plural applies to separate sections (:¸.6, :8.¸–6; cf.
:...:). F.’s use is more specialised than that of Vitruvius, for whomsubstructiones
(\iii.¸.¸) and substructum (\iii.6.6) would cover anything raised above ground
level: Espinilla Buis´ an (:qq¸) 6¡¸ n.¸. For what F.’s ‘substructure’ is see Hodge
(:qq.) :.q.
opere arcuato The phrase seems precisely limited to style of construc-
tion: apart from F. its only appearance is Pliny, Ep. x.¸¸... Besides chapters
¸–:¸, we find it only at ¸6.6 (applied to the urban series of Neronian arches);
elsewhere F. uses arcus (:¸.¸n.; cf. :.¸.:n. fornices), arcuationes (:8.¸n.) and arcu-
atura (questionable: ¸.¸n.): Espinilla Buis´ an (:qq¸) 6¡¸–q.
j.z altitudinem cuiusque See :8.:–. omnes aquae diversa in urbem libra
perveniunt; inde serviunt quaedam altioribus locis et quaedam erigi in eminentiora non
modulorumque <rationes> F. discusses size and capacity of pipes
(.¸..–¸:.¸), ending with comments on irregularities he has detected (¸:.¡–¸¡);
chapters ¸¸–6 add a few hydraulic remarks. In chapters ¸¸–6¸ he lists official
measurements (formulae) for each pipe. Modulus is F.’s standard word for a
‘calibrated pipe’, i.e. one of a specified size: Hern´ andez Gonz´ alez (:q8¸) .6¸.
Note especially q.¸ modulos certos, qq.¡ [Augustus] modulos constituit, :o¸.¡ calicem
eiusmodi qui fuerit impetratus. For the relationship of the diminutive to the noun
modus (cf. note below) see OLDs.vv. a’s rationempresumably derives fromthe use
of that word at ¸¸.: (cf. .¸.., ¸o–.); Schulz’s plural makes the scribal omission
a trifle more explicable.
< . . . quem modum . . . quantumque erogaverit> Cf. 6¡.: nunc
ponam quem modum quaeque aqua . . . habere visa sit quantumque erogaverit; ¸¸.: satis
iam de modo cuiusque . . . superest ut erogationem . . . Delivery (erogatio) plainly relates
to the supply conveyed (modus), about which we need a mention at this point
(anticipating pro suo modo just below). My tentative addition derives directly
from 6¡.:, omitting the detail which F. gives there about the source of his data
in the imperial registers (principum commentariis).
<quaeque> erogationes habiles factae sint ‘and what deliveries
have been made possible’. For erogatio = distributio (cf. 6¡..), see TLL ¸..:
¸q8.6o. (Pace .¸6 prints habiles but translates ‘quali erogazioni dei moduli
siano state confermate’.) Grimal’s reason for thinking that chapters 6¡–¸6
are unannounced seems primarily to be the assumption that modulorum is an-
tecedent of illis and that these erogationes refer to (potential) deliveries (akin
to F.’s use of the verb erogare in ¸.–¸). Kunderewicz punctuates with a semi-
colon after rationem and follows Rubio in suppressing factae sunt, thus taking
this phrase with what follows, a kind of introduction to the detailed break-
down by categories (quantum . . . detur); cf. Gonz´ alez Rol´ an xix–xxii. There
are several difficulties unresolved. (:) The tense of factae sint contrasts with the
(expected) present subjunctives below (serviat, sint, detur, sit) and ought to re-
semble more closely that of perducta sit above. (.) Erogationes (accusative) facere
(or nominative + fieri) is not F.’s normal way of expressing ‘to make distri-
butions, or deliveries’. For this he uses the verb erogare, commonly with aqua
as subject (e.g. 6¡.:, 6¡.¸, 6¸.¡, 66.¸, etc., but used with aquarii as subject in
¸¸). The expression erogationem facere makes only an occasional appearance in
legal texts (Dig. \.¸.¸, xxx\i.:.¸, xxxix...¡; cf. TLL ¸..: ¸q8.¸6.
(¸) Since the verb erogare is constructed with ablative of means (e.g. ¸..¸ quo
modulo, ¸¸.: qua [sc. fistula]) or with ex (¸¸.¸ ex centenaria), we should expect
either the genitive or ex instead of ab + ablative. (¡) So firm a footing has
A’s ab illis gained that too little heed has been paid to the word habiles, which
appears in Cfollowing erogationes. This adjective is primarily used predicatively
(TLL 6.¸: .¡6..8o), as it is here. An appropriate sense can be found for habilis
(OLD s.v. :d) ‘able to be readily accomplished or accommodated’ – roughly
‘possible’ – and this is precisely what F. proclaims has happened (6¡.:, 8¸..–¸,
quantum . . . serviat An unobjectionable anacoluthon: with the first
quantum we anticipate a verb such as detur (below) or erogetur (.¸.:), but then –
because servire is intransitive (+dative) – the second quantum is used adverbially
(‘to what extent’).
extra urbem In chapters ¸8–86 it is further specified how much of the
extra-urban distribution is nomine Caesaris and how much to privati. F.’s extra /
intra urbemdistinction is consistent in chapters ¸8–86, but elsewhere we see extra
urbem / in urbe: .¸.:, q6, ::¸..–¸; cf. 6¸.¸–¸ in plerisque urbis partibus / extra urbem,
:o¸.¡ extra urbem / in castellis et salientibus publicis. Except in certain specific cases
(the senatus consulta of :oo.:–., :o6.:, :o8; cf. also :o:.¸) it is difficult to pin down
what F. means by urbs. He seems to use the term rather loosely, sometimes at
least embracing the continentia tecta (see :o¡.:n. aedificia urbi coniuncta); cf. ¸.8n.
propius urbem; 6q.¸, ¸o.¸ prope urbem. For further discussion see Champlin (:q8.)
q¸–8; Fr´ ezouls (:q8¸) ¸8:–¸, Bruun (:qq:) :¡¸–8, Taylor (.ooo) ¸6–q, De
Kleijn (.oo:) ¸:–¡, .¡6–¸..
<in> urbe unicuique regioni Cf. .¸.: quibus regionibus, ¸¸.¸ per regiones
urbis. Because chapers ¸q–86 refer to fourteen numbered regiones, it is clear that
these are the Augustan wards (Suet. Aug. ¸o.:, Dio, r\.8.¸); LTUR i\: :q¸–.o¡.
pro suo modo ‘in relation to the available supply’, as rightly noted by
Poleni: ‘Significatur autem pro sua copia, pro mensura suae quantitatis.’ This
sense of modus (OLD s.v. :a) is frequent and regular (:q.., 6¡.¡, ¸¸.¸, ¸¸.:, etc.);
cf. ::¸.¡n. This use for quantities of water derives from the more common use
in gromatical contexts, e.g. Hyg. Grom. De condicionibus agrorum p.86.¸T p.qo.¸
Campbell hunc igitur modum quattuor limitibus mensura s(upra) s(criptum) inclusum
vocamus medimna.
quot castella Tanks fromwhich water was delivered within the City. Ac-
cording to ¸8.¸ these totalled .¡¸. The translation ‘reservoirs’ is misleading,
for the purpose of castella was not storage but regulation of the flow in closed
pipes that issued from them. Vitruvius, \iii.6.:–., the first attestation of this
sense, describes a simple system: from a castellum located on the town wall (the
height creating a greater head) three discharge pipes are appropriately posi-
tioned to serve public basins, public baths and private consumers (see further
q8..n.). At Rome the system was far more complex, for castella numbered in
the hundreds (¸8.¸n.).
Modern scholars speak regularly of a terminal castellum (or castellum divi-
sorium), which ‘is in effect a junction box, marking the end of the aqueduct
proper and the start of the urban distribution process’: Hodge (:qq.) .¸q–q:.
F. never explicitly mentions such a castellum, although he often makes clear at
what point distribution begins (e.g. ¸.q, .o–.); see also .o..n.
Secondary castella, probably not radically dissimilar to the ‘water towers’
at Pompeii, were fed by conduits (channels perhaps sometimes, more often
pipes) leading from the main castellum, and it was from these that deliveries
were made. The entire systemissuing froma secondary castellumwill have been
one of closed pipes (fistulae), hydraulically different from the open or free flow
of rivi but thereby subject to what was probably fairly sophisticated regulation
and adjustment (cf. :8.¸n. pressura, ¸¸–6, :o¸.¡–¸, ::¸.¸). See Wasserversorgung :
(:q8.) :q6, pls. 6q–¸o; Hodge (:qq.) .q:–¸o¸; cf. Bruun (:qq:) ::6–.o, Evans
(:qq¡) 6–8, De Kleijn (.oo:) ¸.–8.
publica privataque The antithesis is a commonplace (see, for exam-
ple, Nisbet and Hubbard on Horace, C. ii.:¸.:¸). Neither in .¸.: nor in the
corresponding chapters (¸8–86) does F. distinguish some castella as ‘private’,
hence the readiness to accept Schultz’s deletion of privataque. But why then
does one need publica? Water granted to privati (¸8.¸, ¸q.., etc.) could only be
drawn from castella: :o6.: quibus locis . . . apte castella privati facere possent, ex quibus
aquam ducerent quam ex castello communem accepissent. ‘Private tanks’ would be those
erected by privati, or serving in the distribution of water to privati. ‘Public tanks’
were, of course, those which formed part of the main service. Both were used
in the distribution of public water, and thus combined without distinction in
F.’s figures for erogatio.
publicis operibus . . . detur At ¸8.¸ F. has three categories for urban
distribution: nomine Caesaris, privati, usus publici, the last of whichis further divided
into castra (¸8.¸n.), opera publica, munera, lacus. His figures indicate that deliveries
under the heading of ‘public uses’, all within the City, amounted to ¸:.¡ per
cent of the total supply (¡,¡o: quinariae out of :¡,o:8: Table 6). For categories
of delivery see Petrucci (:qq6) :¸:–8..
publicis operibus Public works must have embraced (:) buildings be-
longing to the state, and (.) structures, monuments and places open for public
use that were not specifically under the control of Caesar or his fiscus (::8.¡n.):
Lanciani (:88:) ¸8:, Evans (:qq¡) q–:o. The only specific opus publicum that
F. names is the Euripus of Virgo in the Campus Martius (8¡.¸). Perhaps also
by tradition some establishments that served public needs continued under
this category (balneae come to mind, on which see :o¸..n.). The great imperial
thermae would have received their water nomine Caesaris, and those who operated
privately owned bathing facilities accessible to the public would have received
theirs as privati: Evans (:qq¡) :o, Fagan (:qqq) ¸..
muneribus . . . appellantur The parenthesis indicates that F. uses
munera in an unfamiliar sense, but the textual difficulty obscures his intended
definition. Castellorum et munerum stationes at ::¸.¸ suggests that munera played
some role in regulating distribution. The conventional view is that these were
castella of a special sort, fitted out with elaborately decorated monumental
fountains. As best I can tell, this notion hangs precariously on the word cul-
tiores transmitted in the ita-clause. Support is sought in Pliny, HN xxx\i.:.o:
(Agrippa) lacus septingentos fecit, praeterea salientes centum quinque, castella centum
triginta, complura etiam cultu magnifica; operibus iis signa trecenta aerea aut marmorea
imposuit, columnas ex marmore quadringentas (cf. Suet. Claud. .o.: plurimos et ornatis-
simos lacus). On this viewmunera essentially keeps its familiar sense (an elaborate
private undertaking presented for the public good), and this is how F. applies
the word to Agrippa’s building projects at q8.: operum suorum et munerum velut
perpetuus curator fuit (cf. TLL 8: :666.¸¡). For a more comprehensive discussion
of these ‘fontane artistiche’ (or ‘ornamentali’) see Aicher (:qq¸), Del Chicca
(:qq¸) .¡6–¸¸. As transmitted, however, the ita-clause does not yield a com-
patible meaning. Scholars have approached the problem from two directions.
(:) Make the verb active, with cultiores taken as ‘the more polite’ (Loeb; cf. TLL
¸: :6q..¡.). This implies that munera is either a euphemism (for public latrines:
Grimal (:q¸¡); Robinson (:q8o) ¸¸; Blackman–Hodge (.oo:) :.o) or a touch
of fashionable jargon (shorthand for deum munera: Baldwin (:qq¡) ¸o¸, citing
Pliny, HN xxxi.¡:, where Marcia is inter reliqua deum munera urbi tributa). Both
are absurd. (.) Keep the passive but adjust cultiores, either by turning it to a
neuter (sc. opera) or by supplying an appropriate substantive. ‘More elegant’
could perhaps be applied to such things as fountains (cf. TLL ¸: :6q:.¸q),
although no emendation proposed has given the requisite neatness of style:
cf. Del Chicca (:qq¸) .¡6 n.q.. Even if the term munera may have been a part
of the watermen’s argot, ita enim . . . appellantur ought to reveal a synonym in
common use, and the fault thus seems to lie in cultiores.
lacibus Public basins, serving in much the same way as Greek spnvci:
T¨ olle-Kastenbein (:qqo) :¸:. They were very much public places: Cic. Rosc.
Am. 8q multos caesos non ad Trasumennum lacum sed ad Servilium vidimus, Hor. Serm.
i.¡.¸6–8 omnes / gestiet a furno redeuntes scire lacuque / et pueros et anus. Some lacus
were simple stone troughs or tubs: see Wasserversorgung : (:q8.) :q¸ pl. 68; others
more elaborate (Pliny, HN xxx\i.:.o cited in preceding note). For a survey of
attested usage of this term, see Del Chicca (:qq¸) .¸:–¡o. On the apparent
city-wide uniformity of delivery to lacus see ¸8.¸n. Closely associated with lacus
are salientes (q.qn.).
nomine {Iulii} Caesaris The intrusive Iulii exemplifies the kind of
thoughtless pretension characteristic of Peter the Deacon (Introd. ¡¡). It is
worth noting that this category obviously postdates the Agrippan pattern of
distribution (q8..n.). We may surmise, but we cannot be sure, that it was de-
vised as part of the Augustan administrative reform (see qq..–¡). Deliveries
nomine Caesaris presumably were – or could be – made to imperial properties
and interests of all kinds. In addition, they might well have included places and
purposes of more ‘public’ nature; cf. Bruun (.oo:b) ¸¸–8. Pipes bearing stamps
with the emperor’s name are no reliable guideline for determining delivery
patterns, for the epigraphic evidence from fistulae does not fit neatly into F.’s
categories: Bruun (:qq:), .¸¡–6. Nor is it easy in any case to extricate im-
perial interests per se from projects that emperors undertook for the public
good: Taylor (.ooo) 8¸–¸. From ¸8.¸ we reckon that deliveries nomine Caesaris
amounted to ¡..¸ per cent outside the City, :¸.. per cent within, overall .¡.¡
per cent (Tables 6, 8 and q).
privatorum usi<bus> beneficio principis Note the chiasmus, in
which contrasting singulars to plurals is perhaps a subtle hint of power – or
generosity (cf. F.’s closing words :¸o.¸ is vero qui admoniti ad indulgentiam impera-
toris decucurrerunt possumus videri causa impetrati beneficii fuisse: in reliquos vero . . .).
Privati were, technically, all persons not members of the imperial family and
who held no magistracy. Grants to privati were known as beneficia (qq.¸n.), and
they were made only by the process of impetratio (:o¸.:n.). Plainly, as data
from surviving pipes demonstrate, privati who received concessions for wa-
ter were for the most part persons of standing: Eck (:q8.c), Bruun (:qq:)
66–¸:; De Kleijn (.oo:) :q¸–..¸. Some were owners of luxurious suburban
villas and elegant townhouses or lavish urban horti, others persons to whom
the water was an important industrial commodity for agriculture, manufac-
turing or trade. From ¸8.¸ we reckon that deliveries to privati amounted to
¸¸.¸ per cent outside the City, ¸8.6 per cent within, overall ¡¡.. per cent
(Tables 6, 8 and q).
detur This use of dare (for a more general one see ::..n.) is semantically
consistent with other terms such as impetrare and beneficium (:oq.¡–6), but there
is usually in addition at least a semi-legal sense connoting a formal and official
grant of public water: q¡.¡, q8.., :o¸.., :oq.:. Note especially its appearance
in legal texts: :o6.: aquae ducendae ius esset datum, . quibus aqua daretur publica,
particularly in combination with adtributio (:o8, :.q.¡). One can compare a
similar use of dare applied to public grain distributions: :oo.: frumentum plebei
datur, ¸ praefecti frumento dando.
ius <ducendarum> tuendarumque See q¡.: (with its definition):
Sequitur ut indicemus quod ius ducendae tuendaeque sit aquae, quorum alterum ad co-
hibendos intra modum impetrati beneficii privatos, alterum ad ipsorum ductuum pertinet
tutelam. In theory the ius ducendae aquae ‘right to draw (public) water’ granted
to privati (q¡..–q¸, :o¸–:¸) was quite distinct from tutela ductuum ‘respon-
sibility for the conduits’ (q6, ::6–¸o), but see :o6.:n., where privati with
such ius are still prohibited from taking water directly from public chan-
nels. Tutela embraced the overall maintenance – including the reliability of
the water supply and safeguarding the priorities for its use. This is clear
from F.’s discussion of Republican precedent (q¸.:–¡) and from his occa-
sional remarks on illicit taps (¸.:, ¸¸.¸, :o6.¸, :.6.:, :.8..). Note especially
the twofold fraus exposed at ::¡.:–¸: non enim solum ad ipsarum aquarum custodiam
sed etiam ad castelli tutelam pertinet, quod subinde et sine causa foratum vitiatur. Along
with their legal flavour, words like tutela and cura may carry socio-political
poenae . . . inrogatae ‘Penalties called for’ are of direct concern to the
curator, for his judicial powers include imposing such fines (:.q.¸). For poena
(distinguished from multa: q¸.¡n.) cf. Varro, Ling. \.:¸¸ poena a poeniendo aut quod
post peccatum sequitur. Note that the three legal authorities F. names reflect the
legislative roles of both populus and senatus as well as the force of the emperor’s
lege The singular can be taken generally as ‘by law’ (or ‘statute’, for F.
surely means a lex rogata voted by the populus), but it may be used specifically in
reference to the Lex Quinctia of q ncr, quoted verbatim in chapter :.q, which
provided for a poena non mediocris of :oo,ooo sesterces. It may not be accidental
that the words poenae lege are juxtaposed here: the lex is the last document F.
provides (:.q) and executio legis appears in his final sentence (:¸o.¡).
senatus consulto The senatus consultum was in essence a form of instruc-
tions for a magistrate: RE Suppl. 6: 8oo, Daube (:q¸6) ¸8–8o. C’s singular may
be accidental, an expansion of an abbreviation, but I prefer it to a’s plural.
A series of senatus consulta in :: ncr established the office of curator aquarum
and defined its powers. F. cites six of these documents (there were likely more:
qq..n.), but in only one of them is a poena mentioned (:.¸..).
mandatis principum Internal instructions (to which F. refers at :oq.6
and ::o..) addressed by the princeps to an official in his service, but regarded as
having the force of law: RE:¡: :o.¸; Finley (:q¸¡), Marotta (:qq:). Mandata thus
used occurs only in the plural: TLL 8: .6¸.66; cf. H–Sz :8. For the similarity
of the Gnomon of the Idios Logos to such mandata, see Bruun (:qq:) :¸.
q–r6 F. presents the aqueducts of Rome in chronological order and intro-
duces their auctores as familiar names in Roman history. By giving the lengths
of the conduits in this section, he can demonstrate howthe hydraulic systemof
the City has grown from small beginnings to a magnificence worthy of fable.
q.r Ab urbe condita A solemn and formal phrase, emphatically ‘annal-
istic’ in tone (cf. Livy, pr.6, Tac. Hist. i.:.:), but far commoner in the Elder
Pliny than in the historians, as noted by De Laine (:qq¸) :..–¸. For reports
of novelties at Rome cf. Pliny, HN \iii.:6 elephantos Italia primum vidit Pyrrhi regis
bello et boves Lucas appellavit in Lucanis visos anno urbis CCCCLXXII, x\.:o. cerasi
ante victoriam Mithridaticam L. Luculli non fuere in Italia, ad urbis annum DCLXXX,
x\iii.:o¸ pistores Romae non fuere ad Persicum usque bellum annis ab urbe condita super
DLXXX. F.’s use of AUC dates is less a matter of solemnity than of chronology.
Since he is not writing in annalistic form, for him consular names alone are
not adequate and AUCdates serve his purpose better as he leaps fromcentury
to century.
‘Through ¡¡: years’ suggests that Rome’s first aqueduct was introduced
in the following year (AUC ¡¡.). At ¸.: the date of Appia is securely fixed
by the consuls of ¸:. ncr, which is AUC ¡¡. in the traditional (Varronian)
reckoning. But F.’s AUC dates – as transmitted – do not follow the Varronian
AUC (F.) AUC (Varro) consuls
Appia (¸.:) <?¡¡:> ¡¡. ¸:. ncr
Anio Vetus (6.:) ¡8: ¡8. .¸.
Marcia (¸.:) 6o8 6:o :¡¡
Tepula (8.:) 6.¸ 6.q :.¸
Julia (q.:) ¸:q ¸.: ¸¸
Virgo (:o.:) <¸¸¸> ¸¸¸ :q
Claudia-Anio Novus (:¸.:) ¸qo ¸q: ¸8 cr
(:¸..) 8o6 8o¸ ¸.
The two earliest dates seem to differ from the Varronian standard by one
year, the next four by two years. Emendation is required (at least for inter-
nal consistency) at :¸.:, .. To account for the phenomena is not an issue
of gripping interest, since consular dates leave no room for doubts as to the
chronology. F.’s source(s) may not have followed the Varronian system, and/or
numeral forms may have suffered in transmission. On F. and Roman dat-
ing practices in general, see Trieber (:8q.), Leuze (:qoq) .o¸, Samuel (:q¸.)
contenti fuerunt Cf. Juv. ¸.¸:¡ uno contentam carcere Romam. F. prefers the
perfect of esse, perhaps to emphasise completion of a state (¸.:, 6¡.., 88.¸, q¸.:,
q8.:, :¸o..; cf. :.8.: fuerat); it is more frequent than the imperfect (q¡.¡, q¸..).
Verbal force in contentus (<contineo) has evaporated (TLL ¡: 6¸8..:), but for F.’s
use of perfect with passive participles see q¡.¸n. cautum fuit.
quas . . . hauriebant The tense contrasts with nunc . . . confluunt (§¸). For
the hydrogeology of Rome, where draining had long taken precedence over
importing water, see Lanciani (:88:) .:¸–¡¸, Thomas (:q8q), Purcell (:qq6a)
:q¡. F.’s emphasis is, of course, on fresh, or ‘living’ water, but it is perhaps
noteworthy that F. does not mention cisterns or pools (cf. Tac. Hist. \.:..:, of
Jerusalem, fons perennis aquae, cavati sub terra montes et piscinae cisternaeque servandis
imbribus). Although storage might not have been a matter of necessity in early
Rome, there did exist the piscina publica hard by Porta Capena (Livy, xxiii.¸..¡;
NTD .q.). In F.’s day cisterns may well have existed, but they would have
been attached to private properties and not part of the public system under
his care. On cisterns in general, Hodge (:qq.) ¸8–66; on their occurrence in
agricultural writers, White (:q¸o) :¸:–6o.
ex Tiberi Tiber water was notoriously turbid: cf. e.g. Virgil, Aen. \ii.¸:
multa flavus harena (the conventional epithet perhaps fromEnnius). But this defect
was intermittent, largely seasonal, and overall water quality was good – prior
to the advent of modern pollutants. Pliny speaks of continued reliance upon
river water (HN iii.¸¡ nec minus tamen aquis ac tot fontibus in urbem perductis), and
for some uses it might even have been preferable to that from the aqueducts.
See in general LeGall (:q¸¸).
ex puteis Wells, from which water must be manually or artificially lifted
(distinct from fontes, springs): Varro, Ling. \..¸ is si quamvis deorsum in terra, unde
sumi pote, puteus; nisi potius quod Aeolis dicebant ut tú:cucv sic tú:tcv a potu, non
ut nunc gptcp (cf. \i.8¡). Although F. does not say so, these were sometimes
accorded festal attention along with springs; cf. Varro, Ling. \i.6 Fontanalia a
Fonte, quod is dies feriae eius; ab eo tum et in fontes coronas iaciunt et puteos coronant. On
wells in general, Hodge (:qq.) ¸:–8.
ex fontibus Springs, from which water flows naturally: Varr. Ling. \.:.¸
fons unde funditur e terra aqua viva. On springs, Hodge (:qq.) ¸.–q. Cf. Cic.
Rep. ii.:: (Romulus) locum delegit et fontibus abundantem et in regione pestilenti
q.z fontium . . . colitur Natural springs and capita fluminum, especially
those with constant flow in all seasons, attracted the devotion of primitive
peoples (echoes of which are, e.g., Virgil, Ecl. :.¸. fontis sacros, Hor. Epist. i.:6.:.–
:¡); cf. Ninck (:q6o) :–¡6, T¨ olle-Kastenbein(:qqo) .8. Pliny, Ep. \iii.8 describes
the cult at Clitumnus (note his praesens numen and fons ille deusque celebratur). To
the source of Marcia enough sanctity could be attributed to permit Tacitus,
Ann. xi\....¡ to encourage a shudder from his readers: (Nero) fontem aquae
Marciae ad urbem deductae nando incesserat, videbaturque potus sacros et caerimoniam
loci corpore loto polluisse; secutaque anceps valetudo iram deum adfirmavit. Curative
properties associated with these waters reinforced and perpetuated the ancient
veneration. F. has singled out three fontes, ones which no doubt antedated the
introduction of Aqua Appia and were located in or near the oldest settled
portions of the City.
salubritatem . . . creduntur a inserts enim, not unappealing and per-
haps easily lost after -tem. Of course the whole clause may be an interpolation.
Camenarum On this source and the cults associated with it cf. Livy,
i..:.:–¸ (with Ogilvie’s note), Symmachus, Ep. :..o.: Camenarum religio sacro
fontis advertitur. From here Vestal Virgins drew their water (Plut. Numa :¸.:–.;
Serv. ad Aen. i.8). Vitruvius, \iii.¸.: mentions fontalis ab Camenis alongside Marcia
saliens as models of excellence. The spring is traditionally placed in the Vallis
Egeriae to the left of Via Appia just outside Porta Capena (cf. Juv. ¸.:.–.o).
Richardson (:q¸8) .¡¡ argues that the prescription by Numa would place the
spring within the ‘Servian’ wall and suggests that it lay on the slope of the
Caelian just NE of the present-day church of San Gregorio Magno: NTD
6¸–¡, LTUR i: .:6 (Rodr´ıguez Almeida).
†Apollinaris Obelised because the form does not cohere with the other
genitives. A spring with healing properties associated with the cult of Apollo
is otherwise unknown, but to delete the word entirely is surely too drastic:
so Jordan (:8¸: ) ii: ¡8. Lanciani (:88:) ..¸ guessed at a location near the
Piscina Publica, between the Caelian and Aventine: cf. H¨ ulsen (:qo¸) .o6 n.:6.
The -ar- may be totally irrelevant, but it is tempting to think of some connex-
ion with the Apollinare, a cult centre of Apollo between the Circus Flaminius
and the Forum Holitorium. A temple of Apollo was vowed in ¡¸¸ ncr and
dedicated in ¡¸: (Livy, iii.6¸.¸, i\..q.¸), and the introduction of the cult
seems to have been closely related to outbreaks of plague (Livy, i\..¸.¸): see
Gag´ e (:q¸¸) 6o–::¸. Apart from a ttpippcv:npicv of Apollo near the Forum
(Plut. Sulla ¸..¡; LTUR i\: ¸q–8o), which is of questionable relevance, there
is no evidence for a spring or fountain associated with this site. Gag´ e (:q¸¸)
¸¸–¡ refuses to identify F.’s fons with this location (‘l’invocation apollinienne
est banale pour l’´ epoque imp´ eriale’), but his objection can be countered by
remarking that F. refers to a cult that antedates the year AUC ¡¡:. More
telling is the objection that fontes usually had divinities of their own: NTD :¸¸
Iuturnae Cf. Varro, Ling. \.¸: multi aegroti . . . hinc aquam petere solent; Serv.
ad Aen. xii.:¸q fons saluberrimus. The spring-fed Lacus Iuturnae was located in
the Roman Forumat the foot of the Palatine, adjacent to the Temple of Castor:
Nash ii: q–:¸, NTD .¸o–:, LTUR iii: :68–¸o (Steinby). For a temple of Juturna
connected with the statio aquarum see ::q.¸n., Coarelli (:qq¸) .¡¸–¸o.
q.j nunc autem . . . Contrasts with contenti fuerunt and hauriebant (§:).
F. names the nine aqueducts which existed in his day. Subsequent to F. two
new aqueducts were constructed: the Aqua Traiana, dedicated in :oq (Ashby
(:q¸¸) .qq–¸o¸, van Deman (:q¸¡) ¸¸:–¡o, Nash i: ¸.–¡), and the Aqua
Alexandriana, built by Alexander Severus c. ..6 (Ashby (:q¸¸) ¸o8–:¸, Van
Deman (:q¸¡) ¸¡:–6o). For interpretation of later texts with larger numbers
and other names, see Ashby (:q¸¸) :¡–:¸.
confluunt Cf. ¸.: in urbem Romam influunt. F.’s point here is that the City
enjoys continuous delivery from nine separate aquae. The verb fluere and its
compounds all refer to water in constant flow. For the simplex: ::.: in usus
populi fluentem, :¸.¸ in urbem fluere, :¸.: (Anio River) turbulentum fluit, 8¸.: Iulia
fluebat, :o¸.¡ ut sine intermissione diebus <et noctibus> aquam fluat, :oq.¸ totus
modus . . . flueret, :.q.¡ fluere. Compounds are rare: q.¡ transfluit, ¸..8 superfluunt,
qo.: defluens, ::: effluere.
aquaAppia Below¸, ¸.:, q.q, :8.¸, ...:, ¸, 6¸, ¸q, :.¸. RE.: .:¸; P–A.:;
NTD ¸–6; LTUR i: 6:–. (Mucci); Lanciani (:88:) .¡6–¸¸; Van Deman (:q¸¡)
.¸–8, ¸68, ¸8¸; Ashby (:q¸¸) ¡q–¸¡; Panimolle (:q68) ¸q–¡¸; Hainzmann
(:q¸¸) ¸8–8.; Pace (:q8¸) ::8–:q; Evans (:qq¡) 6¸–¸¡; Aicher (:qq¸) ¸¡–¸.
Anio vetus Below6, ¸.:, q.q, :8.6, .:, 66, 6¸.¸, 8o, qo, q:.:, q., :.¸. P–A
:.–:¸; RE :: ..::; NTD ::; LTUR i: ¡¡–¸ (Mari); Lanciani (:88:) .¸¸–¸6; Van
Deman (:q¸¡) .q–66, ¸6q, ¸8¸–qo; Ashby (:q¸¸) ¸¡–8¸; Panimolle (:q68) ¡q–
6:; Hainzmann (:q¸¸) 8¸–qo; Pace (:q8¸) :.:–¡; Evans (:qq¡) ¸¸–8.; Aicher
(:qq¸) ¸¸–6.
Marcia Below ¸, q.q, :., :¡..–¸, :8.¡–¸, :q.¸–q, 6¸, 68.¡, ¸..8, ¸6.¸, 8:,
8¸.¡, 8q.¡, q:.¸, q., :.¸. RE:¡: :¸o¸; KP :: q.6 (Forsch); P–A.¡–¸; Nash i: ¡8–
¸:; NTD :¸–:8; LTUR i: 6¸–q (Cattalini); Lanciani (:88:) .¸6–q¸, .q8–¸:¸;
VanDeman(:q¸¡) 6¸–:¡6, ¸¸o–:, ¸qo–¡o.; Ashby (:q¸¸) 88–:¸8; Astin(:q6: );
Panimolle (:q68) 6¸–q¸; Hainzmann (:q¸¸) q:–:o6; Bieber (:q6¸); Morgan
(:q¸8); Pace (:q8¸) :.¸–¸¸; Tortorici (:qq¸); Evans (:qq¡) 8¸–q¸; Aicher (:qq¸)
¸6–¸; Volpe (:qq6).
Tepula Below8, q.:, :8.¡, :q.¸–¡, 6¸.¸, 68, 6q.¸, 8., :.¸. P–A.¸–8; NTD
:8; LTUR i: ¸o (Cattalini); Lanciani (:88:) .q¸–¸, ¸:¸–:¡; Van Deman (:q¸¡)
:¡¸–¸6, ¸¸., ¡o¸; Ashby (:q¸¸) :¸q–6o; Panimolle (:q68) qq–:o.; Hainzmann
(:q¸¸) :o¸–8; Pace (:q8¸) :¸8–q; Evans (:qq¡) q¸–8; Aicher (:qq¸) ¸8.
Iulia Below q, :8.¡, :q.¸–¡, 68..–¸, 6q, ¸6.¸, 8¸, :.¸. P–A .¸–¡; Nash
i: ¡¸; NTD :¸; LTUR i: 66–¸ (Cattalini); Lanciani (:88:) .q¸–8, ¸:¡–:¸; Van
Deman (:q¸¡) :¸¸–66, ¸¸¸, ¡o¸–¡; Ashby (:q¸¸) :6:–6; Panimolle (:q68) :o¸–
:.; Hainzmann (:q¸¸) :oq–:¸; Pace (:q8¸) :¡o–¸; Evans (:qq¡) qq–:o¸; Aicher
(:qq¸) ¸8–q.
Virgo Below :o, :8.¸, ...., ¸o, 8¡. RE .o: :¸o¡; P–A .8–q; Nash i: ¸¸–6;
NTD :q; LTUR i: ¸.–¸ (Le Pera); Lanciani (:88:) ¸¸.–¡.; Van Deman (:q¸¡)
:6¸–8, ¸¸¡–¸, ¡o¡–¸; Ashby (:q¸¸) :6¸–8.; Panimolle (:q68) ::¸–.:; Quilici
(:q68); Hainzmann(:q¸¸) ::¡–:8; Lloyd(:q¸q); Pace (:q8¸) :¡¡–6; Evans (:qq¡)
:o¸–q; Aicher (:qq¸) ¸q–¡:.
[Alsietina] Below::, :8.8, ...¡, ¸:, 8¸. RE :: :6¸8, Suppl. 8: .; P–A.o–
:; Nash i: ¸¸–6; NTD :¸; LTUR i: 6: (Liberati Silverio); Lanciani (:88:) ¸¡.–¸;
Van Deman (:q¸¡) :¸q–86, ¸¸6–¸, ¡o¸; Ashby (:q¸¸) :8¸–q; Panimolle (:q68)
:.¸–¸o; Hainzmann (:q¸¸) ::q–.o; Pace (:q8¸) :¡¸–q; Evans (:qq¡) :::–:¸;
Aicher (:qq¸) ¡:; Taylor (:qq¸); Taylor (.ooo) :6q–.oo. For a to supply the
proper name from ::.: was easy enough; that Peter the Deacon did not do
likewise is good evidence for his carelessness.
Claudia Below:¸–:¡, :¸.¡, :8.¡, .o, 6q.¸, ¸., ¸6.¸–6, 86.:–., 8¸.¸, 8q.¡,
q:.¸, :o¡.¡, :o¸... RE¸: .8¸o, Suppl. 8: .o; P–A..–¸; Nash i: ¸¸–¡6; NTD:6–
:¸; LTUR i: 6¸–¸ (Mari, Jolivet); Lanciani (:88:) ¸¡¸–q, ¸¸6–6¡; Van Deman
(:q¸¡) :8¸–.¸o; Ashby (:q¸¸) :qo–.¸:; Panimolle (:q68) :¸:–¡6; Hainzmann
(:q¸¸) :.:–8; Pace (:q8¸) :¸o–¸¸; Evans (:qq¡) ::¸–.8; Aicher (:qq¸) ¡.–¸.
Anio novus Below :¸, :¸, :8.¡, .o.:–., ¸..6, ¸¸, 86, q:.:–¡, q¸, :o¡.¡,
:o¸... RE :: ..:.; P–A ::–:.; NTD ::; LTUR i: ¡.–¡ (Mari); Lanciani (:88:)
¸¸o–6¡; Van Deman (:q¸¡) .¸:–¸¸o, ¸8o, ¡:¸–.:; Ashby (:q¸¸) .¸.–q8;
Panimolle (:q68) :¡¸–6o; Hainzmann (:q¸¸) :.q–¸¡; Pace (:q8¸) :¸6–8¸;
Evans (:qq¡) ::¸–.8; Aicher (:qq¸) ¡¸–¡.
j–rj Historical background for the individual aqueducts, location of their
sources, length of their channels broken down according to construction type;
cf. ¸.: per quos quaeque earumet quibus consulibus, quoto post urbemconditamanno perducta
sit, dein quibus ex locis et †a quoto† <quot passus ductus cuiusque efficiat, et ex
eo> quantum subterraneo rivo, quantum substructione, quantum opere arcuato. DeLaine
(:qq¸) :..–¡ observes that in these ‘potted histories’ (for their relevance to
the curator aquarum see ¸.:n. per quos) F. mimics ‘the highest tradition of prose
literature, annalistic history’. He does so in small ways, for these chapters
conformlargely to a ‘tabular’ pattern (Introd. .¸). He is unusual in giving years
ab urbe condita as well as consuls’ names for dating; he identifies the auctores of
the aqueducts as men of the highest standing; he refers in passing to important
events in Roman history; only later does he explicitly state that the hundreds of
miles of physical structures are themselves monuments of Rome’s greatness (:6,
cf. ::q.:).
j.r M. Valerio Maximo P. Decio Mure ¸:. ncr, the AUC date (¡¡.)
being understood from ¡.:; cf. MRR i: :¸q. C’s frequent nominatives (i.e. -us
for -o) in consular dates are presumably due to scribal carelessness, as likely as
not on the part of Peter the Deacon (see Introd. ¡¸).
anno . . . tricesimo The First Samnite War began in ¸¡¸ ncr, AUC¡::
(Livy, \ii..q.:), and tricesimo (‘in the ¸oth year’) brings us precisely to ¡¡:/¡¡..
Besides the promised consular and AUC dates (¸.:), F. generally gives a third
chronological reference (6.:, ¸.:, q.:, :o.:, :¸.:). A meaningful point de rep`ere for
the first aqueduct was the first of Rome’s great wars (cf. Livy, \ii..q..).
Samnitici belli The adjective is extremely rare: Suet. Vit. :.¸, Florus,
i.:8...: Baldwin (:qq¡) ¸o¸.
aqua Appia The appellation (like that of the road) is unusual in that it is
based on the praenomen, not the gentile. But the praenomen Appius was restricted
to the gens Claudia and it seems to have been regarded as a nomen in its own
right (note especially CIL 6.:.8o Appios consol); cf. TLL .: .8q.6¡.
<ab> Appio Claudio {Crasso} censore Appius’ censorship in this
year: MRR i: :6o. The cognomen Crassus is nowhere else attested for Appius
Claudius (RE Claudius no.q:), although it is found among others of the Appii
Claudii: TLL Onom. s.vv. Claudius, Crassus; cf. Kajanto (:q6¸) .¡¡. Its ap-
pearance here betrays every sign of Peter the Deacon’s work, for it is to an
Appius Claudius Crassus that he appends the note hic viam appiam stravit in his
Chronica consulum, dictatorum et imperatorum (Cod. Cas. .¸¸, p.q, ed. Florilegium
Casinense ¸.: (:8q¡), p.¸¸); cf. Bloch (:q8¡) 66–¸q. In C the name Crassus was
written along with the rest of the sentence, but the cen- of censore was added
later by Peter in a space left blank (cf. ...n. respicere possum).
cui postea [Caeco] fuit cognomen Cf. Str. ii.¸.¡ Scipio, cui postea
Africano cognomen fuit; i\.:.¡o Manlius, cui Imperioso postea cognomen fuit. Caecus
is unique as a cognomen: Kajanto (:q6¸) .¸8. On its origin note especially
Livy, ix..q.:: sed censorem etiam Appium memori deum ira post aliquot annos luminibus
[idem eo anno] A temporal phrase is more appropriate than a second
(unparallel) relative clause. For idem cf. ¸.. eidem and :..: idem Augustus; for
eo anno cf. q.q eodem anno. Appius’ two great public projects are closely linked
in Roman tradition (e.g. Livy, ix..q.6, Paul. Fest. p..¸ L, Eutrop. ii.q), al-
though the name Via Appia does not appear before Cicero (Leg. Man. ¸¸, in
66 ncr). Both are firmly associated with Appius’ censorship: note especially
the scornful words to Clodia put into Appius’ mouth by Cic. Cael. ¸¡: ideo
aquam adduxi ut ea tu inceste uterere, ideo viam munivi ut eam tu alienis viris comitata
viam Appiam RE .: .¸8; NTD ¡:¡; LTUR \: :¸o–¸ (Patterson).
a[durbem] Capuam The regina viarumwas the principal route to south-
ern Italy and beyond (Strabo, \i.¸.¸). Diodorus, xx.¸6.. also credits Appius
with building the road as far as Capua during his censorship (ótc Pcun, utypi
Kctún,). There may have been some simplification in the tradition, although
curavit means no more than that Appius undertook the project in ¸:.. Radke
(:q¸¸) :¡q¸–:¸o: has plausibly argued that during his censorship Appius built
only the stretch to Formiae (with ForumAppi as its mid-point); in a later magis-
tracy (cos. ¸o¸, pr.ii .q¸) he extended the route to Capua. Cf. MacBain (:q8o);
Schneider (:q8.) ¡6–¸o.
muniendam curavit The verb munire means loosely ‘to build’, that is
to determine the route, acquire right of way, and make a passable track for
vehicular traffic: see Wiseman (:q¸o) :¸q, Radke (:q¸¸) :¡¡o. The word was
used with viam from early times (Leg. xii ¸.¸, Cato, Agr. ..¡) and probably
conveyed the sense ‘by public authority’ or ‘in the public interest’ (cf. moenia,
munus). Both Diodorus (xx.¸6.. ìiûci, o:tptcï, sc:to:pcotv) and the elogium
for Appius (CIL :
.:q. viam Appiam stravit) are in error about the paving. Even
near Rome this came later: Livy, x..¸.:., xxx\iii..8.¸.
j.z C. Plautium RE no. ¸., MRR i: :6o; cf. MacBain (:q8o) ¸¸:.
venas Subterranean veins of water (OLD s.v. ¸) are distinguished from
fontes (¡.:n.) in that they do not rise naturally or in noticeable quantity to the
surface. At :o.¸ we learn that diggers were used to pursue the venae of Virgo.
Once these veins have been captured F. uses the word adquisitiones (‘feeders’):
:o.¸–8, 6q.., ¸o...
Venocis cognomen Fast. Cap. preserve the names of both censors,
adding for Plautius qui in hoc honore Venox appellatus est. While the word-play
venas Venocis is clear, there is probably no connexion with venator as ‘Hunter’
(Bennett) and ‘Chasseur’ (Grimal) imply. Kajanto (:q6¸) does not include this
cognomen. If Appia’s source was indeed fifty feet below ground (6¸.¸n.), the
appellation is not at all inappropriate: Evans (:qq¡) 6¸ n.:.
j.j intra annumet sex menses Eighteenmonths was the regular term
for censors, according to the Lex Aemilia of ¡¸¡ ncr (Livy, i\..¡.¸).
deceptus . . . censura Cf. Livy, ix..q.6–8 memoriae tamen felicioris ad
posteros nomen Appi, quod viam munivit et aquam in urbem duxit; eaque unus perfecit
quia ob infamem atque invidiosam senatus lectionem verecundia victus collega magistratu se
abdicaverat, Appius iam inde antiquitus insitam pertinaciam familiae gerendo solus censuram
obtinuit. Cf. Diod. xx.¸6.:, Livy, ix.¸¸.¡.
tamquam{ib} idem facturo We cannot keep C’s ibidem ‘at the same
time’, for facturo requires an object. Poleni suggests id idem on the analogy of id
ipsum, but for this there is no parallel.
nomen aquae . . . honorem Cf. 6.¡ gloria perductae pertinuit ad Fulvium.
Appius’ desire that the aqueduct and road should bear his name is an early
example of the combination of civic pride and family ambition that lay behind
so many Roman building projects. For Appius and his family in particular, see
Mommsen (:86¡) ¸o:, ¸o6–8; RE ¸..: .68¸ (M¨ unzer).
tergiversationibus This is the only attested instance of the plural; the
noun is rare in itself: F. uses it at Str. i.:.:o. Elsewhere: Bell. Afr. \iii.¡, Cic.
Mil. ¸¡ (these both with mora), Att. x.¸.:, Sen. Ep. 66.:6 (cf. ¸¸.:: tergiversatur),
Dial. \ii.8.6, Dig. i\.6..¡.pr. (Paulus), xx\i.¡.:.: (Ulpianus); cf. Gell. xi.¸.q
extraxisse censuram Livy, ix.¸¸.¸–¸¡..6 reports an attempt on the
part of the tribune P. Sempronius to compel Appius to relinquish his cen-
sorship when his colleague had done so at the end of the regular term. The
effort was unsuccessful (ix.¸¡..6): approbantibus sex tribunis actionem collegae, tres
appellanti Appio auxilio fuerunt; summaque invidia omnium ordinum solus censuram gessit.
Mommsen (:88¸) ii: ¸¸: suggests that Appius received extraordinary proro-
gations to finish his public works. On the building projects in this censorship,
see Ferenczy (:q6¸).
j.q–j F. uses a highly formulaic style to present data on both the location
of sources and the lengths of the conduits, the latter broken down according
to construction type (¸.:n.): for detailed discussion, see Introd. .¸–¸.
j.q concipitur Appia Appia’s source has not been identified by the pres-
ence of ancient remains. Ashby (:q¸¸) ¸: reasonably suggests that the springs,
well beneath ground level in F.’s day (6¸.¸), have disappeared since ancient
times. The general area is clear enough, a level stretch of marshland between
the Via Praenestina and the Anio River, fed from the slopes of the Alban hills:
see Quilici (:q¸¡) ¸:¸–:6, and note the searches of di Fenizio reported in Pace
(:q8¸) ::8–:q. Earlier scholars resorted to emendation. According to Luini
(:qo¸) .¡¸, (:qo¡) .:¸, Stewechius altered sinistrosus to dextrosus and put the
source at the foot of Rocca Cenci. Lanciani (:88:) .¡¸ emended Praenestina to
Collatina to justify identification with springs in the quarry of La Rustica. If a
textual error exists, however, it ought probably to be sought in the figure of
¸8o paces given for the deverticulum.
concipitur F.’s regular word for capturing water at the source: 6.¸, ¸.6,
8.., :o.¸, ::.¸, :¡.:; cf. ¸¸.: modus concipitur ad capita. Note also the noun concep-
taculum ...:n. and the expression conceptionis modus (66.6, 6¸.¸, etc.: see 6¡.¡n.
capita ductuum). F. uses capere and its compounds for the most part precisely,
although there are some instances of overlap. Capere itself is restricted to the
sense of capacity (.6.¸–¸, .8.:–., ¸¸.., ¸8.¡, ¸q–6¸; cf. noun capacitas .6.:n.) –
except for ¸..8 (quinariae) capiuntur ex Augusta, and8q.¡ ex fontibus, where the sense
is either that of concipere or excipere. Close in meaning is accipere (¸.¸, :¡.. accipit
fontem), which is also used for transfers (:q.¸ modum accepit, .o.¸, 68.¡, 6q.¸)
as well as ‘to receive’ by way of deliveries (q.¸, 8¸.¸); in the latter sense the
word appears twice in legal documents (:o6.:, :o8). Similarly close is excipere,
connoting both ‘collect’ (:¸.:, ¸¸.6, q. ex flumine, :q.: contectis piscinis, .¸.¸ in
castellum, q¸.¸ ([lakes] in quos excipitur) and ‘draw off’ (:q.¸, ¸6.6 excepta). For
recipere ‘take back’ most instances describe what happens at or beyond the
piscinae (:q.¸, .o.:, 66.6, 6¸.6–¸, 6q..), synonymous is .¸.¸ (from a castellum),
while ¸..8 seems to be a variant of accipere or excipere. For deliberate inter-
ruption of flow, either complete or partial, the word is intercipere (¸.:n., :8.¡,
¸¸.:, 6¸.8, ¸¸.¸, 8¸.., :oq..; special instances apply to Tepula (q.:n., :q.¸,
in agro Lucullano Nothing is known of this estate of Lucullus, from
which Agrippa later introduced the Virgo (:o.:). For a separate ager Lucullanus
near Tusculum, see 8.:n. The reference is probably to C. Licinius Lucullus (RE
no.:o¡), whose name would have stuck with the property long after his death.
F.’s appellation probably derives from documents created under Agrippa.
deverticulo sinistrosus These side roads (cf., e.g., §¸, ¸.6, 8..) may
have been quasi-official access roads. The directional references sinistrosus or
dextrosus always apply, as do the numbers on milestones, to one travelling from
Rome (see ¸.6n.).
j.j a capite For this specialised use of caput see TLL ¸: ¡oq–:o. A good
definition appears in a legal context: Dig. xriii..o.:.8 (Ulpianus) caput aquae illud
est, unde aqua nascitur: si ex fonte nascatur, ipse fons: si ex flumine vel lacu, prima incilia
vel principia fossarum, quibus <a>quae ex flumine vel ex lacu in primum rivum compelli
solent. plane si aqua sudoribus manando in aliquem primum locum effluere atque ibi apparere
incipit, eius hoc caput dicemus, ubi primum emergit.
Salinas . . . portamTrigeminam The Porta Trigemina, between the
Aventine and the Tiber, was built about :qo ncr, slightly upstreamfroman ear-
lier opening in the ‘Servian’ Wall, the Porta Minucia (demolished about :qo):
NTD ¸:o; LTUR iii: ¸¸.–¸ (Coarelli); Coarelli (:q88) .¸–¸o. The topograph-
ical reference is thus later than the Appia itself (cf. Livy, i\.:6.., Plaut. Capt.
qo). Salinae, perhaps in origin salt works (cf. Livy, xxx\.:o.:.), is a toponym
of uncertain origin that lingered long after it had any current relevance to the
area of commercial establishments along the river at the foot of the Aventine
(as a topographical reference also in Livy, xxi\.¡¸.:¸, a disastrous fire in .:¸
ncr). See NTD ¸¡:; LTUR i\: ..q (Coarelli); Coarelli (:q88) ::o–::; Evans
(:qq¡) 68–q. Cf. also Etienne (:q8¸).
<ex eo rivus est> sub<t>er<raneus pas>suum B¨ ucheler’s
restoration is based on 6.6, where F. also uses the nominatives rivus subter-
raneus and substructio. Elsewhere he uses ablatives (§8, ¸.8, q.¸, :o.¸, ::.¡, :¡.¡,
:¸.6), hence Giocondo and Poleni opted here for ablatives.
arcuatura In the formulae of these introductory chapters F. regularly
uses opus arcuatum, but always in the ablative (prec. n.), and in other contexts his
usual word is arcus (¸..n.). The word arcuatura is suspicious: not so much because
it is extremely rare, nor yet because F. might have used arcuatio (elsewhere only
in plural: :8.¸n.), but because the single other attestation of the word is in an
inscription set up under Constantine (CIL 6.¸:¸6¡ = ILS ¸o., Appendix B,
no.:¸) – by the curator aquarum et Miniciae, hinting at some possible use of F.’s
text in the fourth century (see Introd. ¸o). The -ura suffix could equally well be
the result of scribal misinterpretation of -um(hence B¨ ucheler’s <opus>arcuatum)
or -io (hence Poleni’s arcuatione; for this error cf. ¸.¸n. collegio] collega C). If the
transmitted reading be kept, however, it is hard to resist the suspicion that
F. intends to indicate that the ‘arching’ in this short stretch differs somehow
from that which he calls opus arcuatum. This may have been part of a second-
century restoration (¸.:): Van Buren (:q¸¸) ¡68; cf. Quilici (:q8q) 6q, who notes
that the continuous arcade of Marcia in the :¡os ncr was a stunning novelty
in Roman architecture. (Altogether implausible is Van Buren’s notion that
earlier structures might have been wooden.) Evans (:qq¡) 66–¸ presents the
tantalising idea that this above-ground stretch would have been unavoidably
convenient to locate a visible monument to Appius’ censorial achievements,
both road and aqueduct.
proximum portam Capenam The Porta Capena, an opening in the
‘Servian’ Wall between the Caelian and Aventine, was the starting-point of the
ancient Via Latina and later of the Via Appia (§:): RE ¸: :¸o6, NTD ¸o:, LTUR
iii: ¸.¸ (Coarelli), S¨ aflund (:q¸.) :qq–.o:. Appia’s short course above ground
was near (proximum), not above, the gate (see :q.qn. supra portam Capenam). No
traces have ever been found, pace Parker (:8¸6) 8–q; see Lanciani (:88:) .¡q.
Noting that the channels identified with Appia are deep below ground in this
area, Ashby (:q¸¸) ¸¸–¡ concluded that F. must be in error. He suggested that
the arches to which F. refers might be those connected with the arch of Lentulus
and Crispinus (. ncr), which stood belowthe Aventine near the Pons Aemilius,
not far from Appia’s terminal point. The remains that have been reported,
however, are so scanty and have so often been confused with drainage systems
that one should not hastily discredit F.’s account. It is likely enough that Appia
ran for a short distance at or slightly above ground level between the Caelian
and the Aventine, although the channel probably went beneath the roadway
and need not have paralleled the line of the wall.
sexaginta C’s lacuna may possibly indicate that the number was orig-
inally not an even 6o. If that is so, however, adjustments have been made in
the earlier figures, for ::,:¸o + 6o = ::,:qo.
j.6 iungitur . . . Gemellorum Cf. 6¸.¸ ad Gemellos tamen, qui locus est
intra Spem veterem, ubi iungitur cum ramo Augustae. The place was probably called
‘Twins’ because of the two channels. Since the junction was subterranean
the appellation, which occurs nowhere but in our text, was almost certainly
restricted to workers of the water system (see note below): P–A .¡6, NTD
:8o–:, LTUR ii: ¸6¸ (Aronen).
ad S[p]em veterem RE Suppl. ¸: :6¸¡, NTD ¸6¸, LTUR i\: ¸¸8
(Coarelli). The site was that of an ancient shrine on the Esquiline dedicated to
Spes, just inside the later Porta Praenestina (now Porta Maggiore). Dion. Hal.
ix..¡.¡ tcpc :c :n, Lìtioc, ítpcv fixes it eight stades (c.: mile) from Rome
(i.e. from the Porta Esquilina). It occupied the highest point on the east side
of the City and was the scene of a battle in ¡¸¸ ncr (Livy, ii.¸:.. ad Spei). The
temple itself had disappeared, probably soon after another temple of Spes was
built in the ForumHolitoriumduring the First Punic War (Cic. Leg. ii..8). Spes
vetus as a topographical reference appears also in S.H.A. Heliogab. :¸ ipse secessit
in hortos Spei veteris; CIL :¸.¸q.q ( =ILS ¸¸¡¸). In addition to the subterranean
junction noted here, the place was a dividing point for waters from high-level
aqueducts (see :q.¸, .o.¸, .:.¸, 8¸.¡).
in supplementumeius additus Cf. ::.. in subsidium . . . salientium dari,
6¸.¸ dabantur in adiutorium Tepulae, ¸8.: in adiutorium aliarum dantur, with other
verbs: q.6 in supplementum Iuliae vindicaverunt, :..: in supplementum Marciae perduxit,
¸..8 inventum in Marciae supplementum. For in +accusative rather than final dative
cf. Str. ii.¸.¡¸ in auxilium ei supervenire, iii.::.¸ in subsidium hosti venisse.
in confinio hortorum The Horti Torquatiani are otherwise unknown,
but ad Spem veterem makes their general location clear: RE 8: .¡88, NTD .o¡,
LTUR iii: 8¸–6 (Mancioli); Colini (:q¸¸). The property probably belonged to
Torquatus Silanus, whose forced suicide in 6¡ cr (Tac. Ann. x\.¸¸.:) might
have opened the way for Nero to build the arcus Neroniani (.o.¸, 8¸.¸): see
Carcopino (:q.¸) ¸.. The name of the adjoining estate is uncertain, since
we cannot precisely locate any of the several horti known to have been in this
vicinity. F. mentions the nearby Epaphroditiani (68.¡) and the Pallantiani (:q.8,
.o.., 6q.¸). In the same area were the Tauriani (which were a cause of the
suicide of Statilius Taurus in ¸¸ cr: Tac. Ann. xii.¸q), and these bordered on
the Calyclani (CIL 6.q¸¸: cippi hi finiunt hortos Calyclan(os) et Taurianos).
ramus Augustae The ramus Augustae (also so called at 6¸.¸) is to be
distinguished from the fons Augustae of the Marcia (:..:, :¡.¸) and the name
Augusta applied to the Alsietina (¡.¸, ::.:): Lanciani (:88:) ¸.¸–¸., Evans
(:qq¡) ¸:. For ramus ‘branch’ cf. ::¸.¡ sublatis ramis, referring to lead pipes. The
term is applied to water far less commonly than it is in English; OLD s.v. .
cites no examples other than the two from F.; cf. Hern´ andez Gonz´ alez (:q8¸)
ab A[ugusto] Perhaps as part of his general repairs (:.¸ and Appendix
B, no.:). Poleni’s restoration, anticipated by Holste (:66o), is probably right,
but it is very tempting to think that the work was at least contemplated by
Agrippa in connexion with the project of bringing the nearby Virgo (:o.:).
Cf. Suet. Aug. ¡..: (a remark of Augustus) a genero suo Agrippa perductis pluribus
aquis, which could embrace the ramus Augustae as well as Julia and Virgo. The
appellation in any case is to be dated after the title Augustus was conferred
in .¸ ncr, thus subsequent to Agrippa’s repairs on the Appia in ¸¸ (q.q). It is
likely that the introduction of an additional supply by the ramus Augustae for the
first time permitted deliveries to Trans Tiberim (¸q..): Evans (:qq¡) ¸:. Such
deliveries may have begun under Agrippa as Taylor (.ooo) :¸6–8 suggests, but
a coincidence with Augustus’ rebuilding of the Pons Aemilius would be slightly
more likely (ibid. :¡q–¸¡).
[cui lo]co . . . Gemellorum The text is very uncertain. For the genitive
with cognomen see §. cui . . . Venocis cognomen datum est, :¸.¸ priori Anioni cognomen
veteris adiectum. Grimal first sensed the need for a formof dare, and fromthe mu-
tilated -denti I have conjectured dederunt (probably misread from having been
abbreviated). The appropriate subject is aquarii (q.6n.), the staff and work-
men, to whose circle the nickname must be attributed (see note above). The
restoration traditional since Poleni involves an ambiguous and uncharacteris-
tic ablative absolute (F. prefers a relative: cf. ¸.¸, :..:) <imposi>to co<g>nomin<e
respon>denti. Dilke (:q¸¸) .¡8 proposes indito, on the grounds that C’s lacuna is
quite short.
j.¡ proxime viam Collatinam The directions seem unusual, for one
would have expected a shorter access road leading to the right off the Via
Collatina. There may have been legal or topographical hindrances, or workers
might have conveniently used an access road leading to Appia’s source (above
§¡). It is curious, but probably without import, that C reads collatia both here
and at :o.¸.
j.8 ductus . . . efficit This arithmetical sense of efficere + accusative
(‘comes or adds up to’, OLD s.v. ¸c) also appears at q.¸, ::.¡, :..¸, :¸.6, 66.¡n.,
passus Here and elsewhere (¸.6, :¸.6) I have altered the transmitted
form of passus, and in one case (:¸.6) I have changed the inflection of the
numeral which follows. The errors may be no more than scribal carelessness,
although they may be related to a systematic change fromnumerals, and passus
might at some stage have been abbreviated.
j.q incipit distribui More specific than §¸, and appended to indicate
that no distributions or reallocations were made at any earlier point within the
City (cf. :q.¸–.:.¸). Poleni’s addition of Appia seems stylistically awkward, but it
does make clear that there was no separate distribution fromthe ramus Augustae
(cf. 6¸.¸, ¸q). Note that F. makes no mention, here or elsewhere, of a particular
kind of terminus, although scholars regularly speak of the terminal castellum,
or castellum divisorium, at which point in most aqueducts the water ceased to
run in open flow and its distribution was effected through closed pipes: Hodge
(:qq.) .¸q–q:. On this omission, see ¸..n., .o..n.
[in] imo Publicii clivo Cf. ...¸ emergit, ut diximus, infra clivum Publicii.
Into C’s short blank I have restored in (easily lost before imo), although the
preposition may not be essential. The Clivus Publicii was built by the plebeian
aediles L. and M. Publicius Malleolus about .¡o ncr (Varr. Ling. \.:¸8, Ovid,
Fast. \..8¸). It seems to have been a major street, ascending the Aventine from
the Forum Boarium, more or less the route of its modern namesake: P–A :.¡,
NTDqo; LTURi: .8¡ (Coarelli). An outside water supply to this particular area
of Rome was presumably a response to noticeable growth in the busy area of
the Forum Boarium: Coarelli (:q88), Evans (:qq¡) 68–¸o. In addition to his
other projects, Appius was involved during his censorship with the nearby Ara
Maxima (Livy, ix..q.q, Macrob. Sat. iii.6.:¸).
ad portamTrigeminam. . . Salinae appellantur The topograph-
ical references are redundant (see §¸) and all or part of this passage may well
derive from an intrusive annotation. The clause qui locus . . . appellantur is com-
mon in documentary texts pertaining to land, e.g. Hyg. Grom. De limitibus
p.¸¡T, p.¸8 Campbell.
6.r Post annos quadraginta .¸. ncr. For possible variance between
the AUC date and the consular year see ¡.:n.
anno . . . primo C’s uno is perhaps a vestige of numerals written out
M’. Curius Dentatus . . . cum L. Papirio Cursore According to
the Fast. Cap. censors were [Pap]irius L. f. M. n. Praetext(atus) (in magistratu
mortuus est) and M’. Curius M’. f. M’. n. Dentatus; consuls were Papirius L. f.
Sp. n. Cursor and Sp. Carvilius C. f. C. n. Maximus (both had been consuls in
.q¸ ncr). F. (or his source) might have confused the cognomen of the consul
with that of the censor. The error is perhaps venial, since Papirius Praetextatus
(RE no.¸.), who had not held a consulship and who died in office, is otherwise
unknown (except for a charming story in Gell. i..¸): MRR i: :q8 n.:. The
consul (RE no.¸¸), on the other hand, and his father (RE no.¸., dictator ¸.¸
and five times consul) were familiar figures in Roman history (cf. Str. ii.¡.:,
censuram . . . gessit The expression occurs twice in Cicero (Brut. :6:,
Fam. iii.:o.¸), once each in Pliny (HN x\ii.¸), Suetonius (Claud. :6.:), Val. Max.
\i.¸.¸, but five times in Livy (i\..¡.8; ix.¸¡.:6, ix.¸¡..6, cited above ¸.¸n.).
M’. Curius Dentatus RE no. q (:qoo) (M¨ unzer).
cum L. Papirio C ( =Peter the Deacon) spells Lucio in full here and at
¸.:, ¸.¸, qq.¡ (cf. also Spurio at 6.:). F. himself would presumably have used the
familiar abbreviations; see also :o:.¡n. Ser<v>io. But all praenomina are written
out in the ninth-century portion of the codex containing Tacitus’ minor works:
R. Kaster (:qq.) ¡¸.
Anionis C’s reading is in the original hand, and although there is some
apparent hesitance inthe script, the fourthletter is o (not e). F. seems consistently
to have used the nominative Anio (¡.¸ etc.) and oblique forms Anion-: Anienis in
:.¸ is taken directly from the text of the S.C., and Anienses (qo.:) is adjectival.
For the variants, see Meister (:q:6) .–.o.
qui nunc vetus dicitur In distinction to the Anio Novus, introduced by
Claudius (:¸.¸). The simple Anienis appears in the S.C. of :: ncr (below :.¸),
and Augustan cippi (:¸.:n.) bear only the name Ani(o). CIL 6..¸¡¡ and .¸¡¸
( =ILS :q¸¡ and :q¸¸) both attest the added veteris.
ex manubiis de Pyrrho captis locavit Cf. De vir. ill. ¸¸.q aquam
Aniensem de manubiis hostium in urbem inductam. Letting the contracts (locare) would
have been an official act of the censor (see ¸.:n.), although the funding came
from Curius’ manubiae, personal share of booty: Aberson (:qq¡), ¸¡–:o:, :8o–
¸, post Shatzman (:q¸.) and Bona (:q6o). As consul in .¸¸ he had defeated
Pyrrhus near Malventum (later Beneventum) and had celebrated a triumph
(Plut. Pyrr. .¸..). For a victorious general to use his manubiae for the public
good was traditional (e.g. Cic. Agr. ...¸, Livy, x.¡6.:¡, Pliny, HN \ii.q¸), but
previous spending had been for dedicating temples or spolia and Curius’ is
the first instance of such funds used on censorial construction. There is also
an irregularity in the year of this censorship, since censors had last been in
office in .¸¸; Strong (:q68) q¸ connects this with the project for a new aque-
duct. More emphatic is Aberson (:qq¡) :q¸–8: ‘L’´ election de Curius ne peut
se comprendre que comme partie int´ egrante d’un programme coh´ erent qui
visait ` a utiliser ces manubiae-l` a [i.e. those of Pyrrhus] pour un ouvrage d’utilit´ e
publique, tout en offrant ` a celui qui les avait r´ ealis´ ees la garantie d’en retirer
le prestige correspondant, comme s’il avait suivi le processus votif habituel.’
Sp. Carvilio L. Papirio consulibus iterum The pair previously
served together in .q¸: MRR i::8o, :q¸.
6.z post biennium After Curius’ term as censor had lapsed (see
¸.¸n.), for post biennium does not imply an extension of the censorial power;
cf. Mommsen (:88¸) ii: 668. The date is probably .¸o, but could possibly be
.6q (see :o.:n. post annum tertium decimum).
de consummando . . . opere Here and perhaps at 88.¡ the verb
consummare seems to convey the sense of a formal completion (but not ¸.¸ or
¸.¡, where it simply = conficere). A public project required an official probatio,
normally by the officials whohadlet the original contracts (e.g. Livy, xxiii.¸o.:¸,
xxxi\.¸¸.¸–¸). Special measures were required in this instance since the censor
Fulvius (qui eam locaverat) was no longer in office. Astin (:qqo) .¸ and n.:¡
suggests that this discussion might have been over a need to use public funds,
for the manubiae may have run out.
†irefent . . . praetor† B¨ ucheler’s referente . . . praetore may be right (cf.
¸.¸ Lepido . . . verba faciente), with the praetor’s name lost in the lacuna and the
meaningless nocumi. Although the text is probably beyond recovery, I should
prefer to think that it contained some reference to the legal procedure involving
the appointment of ad hoc magistrates. Cf. Livy, xxii.¸¸.¸–q in religionem etiam
venit aedem Concordiae . . . locatam ad id tempus non esse: itaque duumviri ad eam rem creati
a M. Aemilio praetore urbano . . . aedem in arce faciendam locaverunt. ab eodem praetore
ex senatus consulto litterae ad consules missae . . . Also Livy, xxiii.¸o.:¡ senatus decrevit
ut Ti. Sempronius, consul designatus, . . . ad populum ferret ut Q. Fabium duumvirum esse
iuberent aedis dedicandae causa.
6.j duumviri <a>quae perducendae The decision to choose special
magistrates, modelled on the procedure for duumviri aedi dedicandae (Livy, ii.¡..¸,
\i.¸.8 etc.) and the like (RE ¸: :8o:), is strikingly different from the circum-
stances which obtained at the introduction of Appia (¸.¸). One might imagine
that Curius himself had insisted on a strict observance of legal proprieties.
<et Q.> Fulvius Flaccus The et may be unnecessary, but cf. 8.: and
:o. passim, and it could have fallen out after -at. I add a praenomen since F.
seldom omits it when introducing an important figure. M¨ unzer (RE no.¸¸)
assumes that this is M. Fulvius Q. f. M. n. Flaccus (cos. .6¡), who was tribune
of the plebs in .¸o. I prefer Q. (the praenomen of the cos. of .6¡ is in some
sources Q.), because the letter q is elsewhere in this text particularly susceptible
to loss (¸.:, :o.:, q6) or misunderstanding (¸.¸).
6.q duumvirum Partitive genitive: cf. Cic. Att. ii.6.:, TLL s.v.
gloria perductae pertinuit ad Fulvium Gloria perhaps, but not the
name of the water. Curius, whose primary role is unambiguous, was a man of
proverbial selflessness (e.g. Pliny, HN x\i.:8¸, x\iii.:8; Val. Max. i\.¸.¸; Front.
Str. i\.¸.:.); this might explain why the aqueduct took its name from the river.
6.j concipitur . . . Tiburtium usum Although the Anio Vetus was a
considerably more ambitious project than the Appia, with somewhat more ad-
vanced engineering, the aqueduct was still primarily an underground channel.
It took its water from the Anio River, a copious supply if irregular in quality
(cf. qo.:), the high elevation of its source enabling it to supply areas of the city
that the Appia could not reach (cf. :8.¡–6). A decision to bring water from
the Anio practically dictated that the aqueduct begin above the cataract at
Tivoli (praeceps Anio: Hor. C. i.¸.:¸). For several reasons the transmitted text is
troublesome. In the first place, this indication of the starting-point for Anio
Vetus differs from F.’s practice elsewhere, especially in that his directions lack
any reference to a major road (Introd. .¸). More serious is the fact that F.’s
statements do not square with the archaeological evidence for the intake of
this aqueduct. Finally, the mention of a derivatory channel for the Tiburtines
seems baffling rather than explanatory. A separate but closely related prob-
lem is that posed by the transmitted figure for the length of the aqueduct
(see §6n.).
concipitur Aprecise location for what F. is describing as the intake of this
aqueduct wouldbe helpful inaddressing textual problems. Canina (:8¸6) \: :¡o
(\i: tav.:¡:) and Lanciani (:88:) .¸6–¸ had proposed sites at San Cosimato,
and despite the data from Reina et al. (:q:¸) 6¡, it was premature of Van
Deman (:q¸¡) ¸. and Ashby (:q¸¸) ¸q to describe as ‘indubitable’ a site near
the bridge at Vicovaro, some eight miles above Tibur. Recent scholars have
paid more attention to sites just slightly further upstream: Panimolle (:q68)
¸o–.; Roncaioli Lamberti (:q86a) ¡:, (:q88) q:–¡; Mari (:qq:a) :68–q. Except
for Roncaioli Lamberti (:q88) qq–:oo, (:qq.) 8¸, who tentatively identifies with
Anio Vetus some remains about : km NE of Marano Equo, there is general
consensus that the starting point of this aqueduct cannot have been more than
say nine or ten miles beyond Tivoli.
supra Tibur ‘Above Tibur’, but just above or at some distance above?
The latter is implied by any interpretation in which F. takes us twenty miles
beyond Tibur (see next note). But supra by itself need not otherwise bear that
meaning, pace Evans (:qq¸) ¡¸:. I take supra Tibur closely with extra portam: above
(i.e., beyond) Tibur, outside the gate.
<via Valeria> vicesimo miliario Although the ablative appears at
::.¸ in contrast to F.’s usual construction ad miliarium (¸.¸, ¸.6, 8.., etc.; cf.
¸.¡), it does so always in conjunction with via, and the ‘twentieth milestone’
clearly cannot be the right distance to the sites proposed for the intake of
Anio Vetus – measured either from Rome, F.’s consistent practice elsewhere,
or from Tibur, for twenty miles would take us well beyond even Anio Novus,
and the earlier aqueduct inferior excipitur (q.). The road distance on the Via
Tiburtina/Valeria from Rome to Tibur is twenty miles (Itin. Anton. ¸oq.:. cf.
Mart. i\.¸¸.¡), that from Tivoli to Vicovaro/San Cosimato is between eight
and ten miles, so Grimal’s tricesimo (ed. p.6q n.:¡) might seemto bring adequate
accuracy (the identical emendation is accepted at ¸.:). Blackman (:q¸q) :¸–:8
explains the ‘.oth mile’ as coincidence; he suggests a lacuna between Tibur
and vicesimo (a line-end in C) and would like the original text to have contained
‘the real distance of the source above Tivoli and the location of the diverticulum
for the Tiburtines’; cf. Evans (:qq¸) ¡¸:. (F., however, seems to indicate that
his conceptio is the point where water is diverted for the Tiburtines: see note
below.) Roncaioli Lamberti (:q88) q:–q is untroubled by no mention of a via,
for she quite erroneously supposes (with Grimal xiii: see Introd. .: n.¸6) that
F.’s directions date from the time of original construction (.¸. ncr!) and that
his point was to indicate the distance beyond Tivoli. In her view, the twenty
miles is taken with supra Tibur; a distance from Rome would have to mean that
F. describes a site at or just outside Tibur, not consistent with archaeological
evidence for the intake further upstream. Ehlers (:q8¸) ¸¸ proposes vicesimo
<octavo>to place the site near the Vicovaro intake. Mari (:qq:a) :6q–¸o would
account for the distance to archaeological remains by adding <via Valeria>and
changing ‘.oth’ to either ‘¸oth’ (or preferably ‘.qth’, for he would place Mile
.q near San Cosimato). He does not explain why F. might in this instance have
failed to mention any deverticulum (cf. ¸.¡, ¸.6, 8.., etc.).
extra portam To be taken closely with the clause which follows (see
next note). Names for the mysterious porta have been variously proposed. Most
scholars, observing two references to Tibur, have felt that the gate was at Tibur.
Cassio (:¸¸6) proposed Baranam, Fea (:8:: ) iii: ..: Varianam, both referring to
ancient Varia (Hor. Epist. i.:¡.¸), modern Vicovaro. Lanciani (:88:) .¸8 and
Ashby (:q¸¸) ¸¸ n.¸ give tentative approval. The gate would then correspond
to modern Porta S. Giovanni, on the road leading to Valle degli Arci. Roncaioli
Lamberti (:q88) q8–q proposes Trebanam for F.’s gate, arguing that F. (or his
source, from the third century ncr) refers to a gate that takes its name from a
road leading from Tibur to Treba (q¸.¸n). As she admits, there is no evidence,
literary or archaeological, in support of this proposal, nor for the exact location
of such a gate. For other names conjectured for a Tiburtine gate see Pace
(:q8¸) .¡o n.¸. Grimal believes that F.’s reference should be to a point of
topography at Rome, but his solutionis completely intolerable. Since he cannot
palaeographically justify the correction extra portam Esquilinam (the starting-
point of Via Tiburtina in F.’s day), he reads Tiburtinam (a gate in the Aurelian
Wall) and is then compelled to delete the whole prepositional phrase as a
‘retouche topographique’ dating from the third century cr or later. If a gate
at Rome is to be sought, Esquilinam might just possibly be right, for it was near
Porta Esquilina that Anio Vetus had its terminus in the city (.:.¸); cf. Ehlers
(:q8¸) ¸¸, Evans (:qq¸) ¡¸o n.:¸. (Prejudiced eyes perhaps could discern il in
what has generally been read as A – I print A, but that too is inexact – for the
unusual majuscules at this point, see Introd. ¡¸.) F.’s reference, it might then
be argued, is similar (but not parallel) to that at ¸.¸. A Roman gate is not really
likely: extra portam should mean (close) outside the gate (cf. .:.¸n. intra portam
Esquilinam), and for ‘counting from’ one would expect a porta.
ubi partem . . . in Tiburtium usum Mari (:qq:a) :6q–¸o rightly
says that F.’s only reason for mentioning a gate here is to indicate as exactly as
possible where the aqueduct delivered part of itself for the use of the Tiburtines.
Cf. 66.. praeter eummodumqui in propriumductumTiburtiumderivatur. For partem[dat]
(a short gap in C), cf. :q.8, .o.¸, and especially .:... Grimal (6q n.:¡) notes that
the Tiburtine channel must have begun at the same point where F. fixes the
conceptio of Anio Vetus (else the quantity diverted would have been reckoned as
erogatio). He then opines that two channels began at the (Vicovaro) intake and
hesitantly attributes to this Tiburtine conduit remains found between this point
and Tivoli: see Ashby (:q¸¸) 6.; cf. Panimolle (:q68) ¸¸–¸; Roncaioli Lamberti
(:q88) q¸. But to suppose F. describes two channels from the intake ignores
his explicit derivatur at 66.. (a separate channel could not thus be described),
and the upstream existence of a totally unrelated conduit for Tibur must be
acknowledged, pace De Kleijn (.oo:) :¸ n.:6. No archaeological evidence is to
hand for a Tiburtine branch of Anio Vetus. For summary of what is known
and what has been proposed, see Evans (:qq¸) ¡¸:–.. In attributing various
remains in and near Tibur it is easy to overlook the fact that Tibur’s own
water supply was sophisticated and carefully managed: see Dessau, CIL :¡,
p..6¸; Ashby (:q¸¸) 6.–¸; Evans (:qq¸). The water will have supplied the
numerous villas near Tibur, and there may have been some connexion with
the irrigation system of the famous orchards (Horace, C. i.¸.:¡) in the valley
below the waterfall.
6.6 longitudinem . . . libramento F. explains that the length of the
conduit (so considerably in excess of the mileage by road) is a consequence
of its libramentum ‘gradient’, the result of the process of levelling (called perli-
bratio by Vitr. \iii.¸ in his description of the technique). See also below :8.:n.
diversa libra, :8.¡n. humiliore derectura. The gradient of a channel must be pre-
cisely calibrated to assure a constant downward flow but one that is nowhere
so precipitous that there is an excessive loss of altitude between source and
destination. Ancient writers call for a gradient of between :/¡ and :/. foot
per :oo feet: Vitr. \iii.6.: solumque rivi libramenta habeat fastigata ne minus in centenos
pedes semipede (transmitted text: <sicilico ne plus> semipede often printed, based
on Pliny, HN xxxi.¸¸ libramentum aquae in centenos pedes sicilii minimum erit). For
aqueduct gradients in general, see Hodge (:qq.) :¸¸–8¡; for those at Rome,
Ashby (:q¸¸) ¸¸–q (based on Reina et al. (:q:¸)).
quadraginta trium milium The transmitted total is suspicious for
being a round number, and the arithmetical ‘confirmation’ (..: + ¡.,¸¸q) is
no guarantee of accuracy, for a scribe or reader could readily have adjusted
one or more of the three numerals for internal consistency. Forty-three miles
approximates the length of the aqueduct’s course between Tivoli and Rome if
one accepts the evidence of Augustan cippi as at least a rough approximation of
the tortuous course below Tivoli: Roncaioli Lamberti (:q88) 8q; Mari (:qq:a)
:6q. Emendation is necessary if one believes that F. is counting fromsome point
beyondTivoli (supra Tibur). Assuming a starting-point at Vicovaro, Ashby (:q¸¸)
¸¸–8suggests a‘boldemendation’ to¸¸,ooo, andBlackman(:q¸q) :8concludes
that Anio Vetus had a total length of approximately 8: km (¸¡,¸oo paces).
Roncaioli Lamberti (:q88) 8q–qo would add to ¡¸ the additional distance of
F.’s ‘.oth mile’, emending to 6¸ (xriii < rxiii). Mari (:qq:a) :¸o–: accepts ¡¸
miles as the total, assuming some :o miles above Tibur and leaving a distance
Rome–Tivoli of ¸¸ miles.
In my view the difficulties are perhaps less serious than scholars have sup-
posed. Ready though one is to find errors in this text, ‘at the .oth milestone’
(i.e. just outside of Tibur) is entirely consistent with a total length of ¡¸,ooo
passus. With ubi partem dat F. is specifying the place which he considers to be
the starting-point – not an intake further upstream, but a point near Tibur
where the channel forked, as it were, with some of the water delivered into a
conduit for the Tiburtines and the rest flowing onward towards Rome. Ap-
proximately the same mileage can be reckoned for Marcia’s course fromTivoli
to Rome: ± ¡¸ miles = 6:,¸:o passus Marcia’s total (¸.8), minus ± :8 miles
in a reasonably straight line from Tivoli (Mile .o) to Marcia’s source (Mile
¸8 on the more direct Via Sublacensis: ¸.6n.). There is no inconsistency with
F.’s statements at :8.¡–6 (where he is not comparing the lengths of any aque-
ducts), for he implies that Marcia lost altitude relative to Claudia because
of its more sinuous and largely underground course. By similarly reckoning
Claudia’s mileage Tivoli–Rome at ± .8 miles = ¡6,¡o6 total (:¡.¡n.), minus
± :8 miles (their sources were proximate: :¡.:, :8.¡), we can see – as we ex-
pect – that length is shorter when altitude is maintained along a more direct
course. On this view, F. neglects entirely the channel of Anio Vetus beyond
Tivoli. He can have done so only if this stretch was for some reason not under
the supervision of the Roman curator, and I hazard the guess that arrange-
ments had been made (no doubt many years earlier) whereby the Tiburtines
assumed responsibility for its maintenance (perhaps in exchange for the water
which they drewoff). Silence is far fromproof, but the highest numbered cippus
known to date, no. q¸., was apparently found at Tivoli not far from no. 8:6
of Marcia: Ashby (:q¸¸) ¸¸. A drawback to my suggestion, of course, is that
concipitur (§¸) must in this case be taken more generally than F. seems elsewhere
to use it (e.g. ¸.¡, ¸.6 etc., apparently with reference to the actual source).
But the verb is not necessarily so limited in its meaning (6¡.¡n.), and it can
easily bear the sense – both here and elsewhere – ‘to have its starting-point’
(cf. TLL ¡: ¸¸.¡¸). It might also be objected that F.’s measurements ad caput
(66..) are taken from the same point as that specified in this chapter. Yet caput
too can be used loosely: cf. 68.¸ caput eius (i.e. of Tepula) observandum est a piscina
Alternatively, we could emend ¡¸ to ¸¸, as Ashby suggested (which would
incorporate in F.’s total the mileage between Tibur and an intake upstream).
But this would require further adjustments in §¸, for instance <via Vale-
ria> tricesimo milliario <deverticulo ∗∗∗ caput eius observandum est [cf. 68.¸]> extra
portam . . . ubi partem dat. . . .
rivus est subterraneus The builders of Anio Vetus deliberately chose
to maintainaltitude by following the contours of the countryside, skirting ridges
and valleys to avoid the more troublesome (and costly) alternative of tunnels
and bridges. The result was a longer course and a lower level at the termi-
nus (:8.¡–6). Because this aqueduct was mostly subterranean, long stretches
have eluded modern searchers, but evidence from the Augustan cippi (:¸.:n.)
indicates that at some points the Anio Vetus must have taken an extremely tor-
tuous course. (Ashby’s Map ¡ shows, for example, how this aqueduct skirted
the valleys near Fosso di Caipoli while the other aqueducts bridged and tun-
nelled.) F. apparently does not reckon short bridges in the upper course. The
most interesting of these (and one which existed in F.’s day) is that at Fosso di
Caipoli, where a second bridge carried the official path running parallel to the
aqueduct: Van Deman (:q¸¡) ¡q and pl. xi; Ashby (:q¸¸) ¸¸–¡. For a similar
apparent oversight in the case of Virgo see :o.¸n.
septuaginta novem Instead of unumdeoctoginta. C regularly has such
forms, e.g. ¸.8, q.¸ viginti octo, :¸.6 quinquaginta octo, 66.6 sexaginta novem, etc.:
K–S i: 6¡o–:.
substructio Traces of this substructure (¸.:n.) have been more or less
satisfactorily identified outside the present Porta Maggiore: Ashby (:q¸¸) 8o–:
(see also below .:.:–¸).
¡.r Post . . . octavo :¡¡ ncr. For the AUC date, see ¡.:n.
Ser. Sulpicio Galba {cum} L. Aurelio Cotta MRR i:¡¸o. Sulpicius’
praenomen is securely attested (Fast. Cap.) and C’s. S. was corrected in Renais-
sance copies. The intrusive cumis perhaps connected with the nominative form
Sulpicius (¸.:n.). Galba (RE Sulpicius no.¸8) and Cotta (RE Aurelius no.q8),
patrician and plebeian respectively, were probably grandsons of the consuls of
.oo. For their roles in partisan politics of this period see M¨ unzer (:q.o) ..¡–¸¸,
Astin (:q6¸) 8o–::¡; Morgan (:q¸8) .¸–8.
vetustate quassati At this date the Appia had been in service for
c.:66 years, the Anio Vetus for c.:.¸ years. There is no record of previ-
ous repairs, although it is hard to believe that both channels had not had
routine if somewhat sporadic upkeep. Here and elsewhere in building con-
texts the word vetustas means ‘wear and tear’ such as would come about
throughnormal use (OLDs.v. ¸): :8.¸n., :.on., :.:.:; see also Thomas–Witschel
(:qq.) :¸.–8. Examples with reficere, e.g.: Nep. Att. .o.¸ cum aedis . . . vetustate
atque incuria detecta prolaberetur, ut . . . Caesar eam reficiendam curaret; Livy, i\..o.¸
aedem . . . vetustate dilapsam refecit; Suet. Cal. .:.: conlapsa vetustate moenia deo-
rumque aedes refectae; Tac. Ann. i.6¸.¸ ruptos vetustate pontes reponeret. Quassatus (OLD
s.v. .–¸) suggests more serious structural problems, akin to those F. outlines
at :.:.:. With reficere, e.g.: Livy, xx\i.¸:.q refectisque quae quassata erant muri (cf.
xxxiii.:¸.:o), xxx\ii.¸:.6 naves in proelio quassatas cum refecisset; Hor. C. i.:.:8 reficit
naves quassas.
privatorum . . . fraudibus The censors of :8¡ had taken action of
some sort against private use of water: Livy, xxxix.¡¡.¡ aquam publicam in pri-
vatum aedificium aut agrum fluentem ademerunt; Plut. Cato ma. :q.: cï :c tcpcpptcv
onucoicv 0ocp otcìcµcv:t, ótn,cv ti, cisic, ioic, sci sntcu,. It was per-
haps in this connexion that Cato spoke de aqua against L. Furius: Serv. ad
Aen. i\..¡¡; Charisius, GLK i: .:6.::; Livy, xxxix.¡..6; cf. Astin (:q¸8) 8¸–¡.
Geißler (:qq8) :q. points out that there is no mention of sanctions, and De
Kleijn (.oo:) q¸ notes that neither source speaks of illegal tapping. Cato and
his colleague may have been calling a halt to excessive private use of public
water in the face of public needs.
interciperentur The verb indicates an interruption of the flow (OLD
s.v. :.d). This can be either complete (as at :oq..) or partial. F. applies the
word most often, as here, in cases of fraud (¸¸.:, 6¸.8, ¸¸.¸), but cf. :8.¡ ne
ab hostibus interciperentur. For a somewhat different sense see q.:n. Tepulae rivum
a senatu The senate’s initiative is an apparent novelty. Repairs to aque-
ducts would presumably have been the responsibility of censors: in :8¡ censors
had let contracts for cleaning drains and building new ones (Livy, xxxix.¡¡.¸).
Both earlier aqueducts had been built by censors, and censors had contem-
plated new construction in :¸q (Livy, xr.¸:.¸). All three parts of the senate’s
mandate in :¡¡ thus encroached on the traditional province of the censors.
Something resembling a public emergency might have existed if problems of
water supply could not await the choice of new censors for :¡..
<Q.> Marcio REno. qo; MRRi: ¡¸:. For Marcius’ praenomen see Pliny,
HN xxxi.¡:, xxx\i.:.:; Plut. Cor. :.:; CIL ¸ p.8¡6; on the omission of Q. see
praetor inter cives This style for the magistrate later known as praetor
urbanus also occurs in the Lex Papiria de sacramentis (Festus, s.v. sacramentum,
p. ¡68 L) praetor qui inter cives ius dicet: see Watson (:q¸¡) 6¸–¸. The reason for
the senate’s choice of a praetor is not at all clear. While his juridical authority
might have facilitated the handling of fraudes privatorum, the office in no way
explains the mandate for repairs. Some have seen here an indication that the
consuls had already been assigned their provinces for :¡¡. They may have been
absent from Rome when the issue was raised in the senate; or, if they were still
in the City, the senate’s choice of a praetor might have been a reflection of
distrust in the consuls (cf. Val. Max. \i.¡..): Morgan (:q¸8) .¸–¸o.
reficiendorum ac vindicandorum Repeats for emphasis the two
points of vetustate quassati and fraudibus interciperentur. Repairs and reclamation
were the primary concerns behind the senate’s decree; cf. Pliny, HN xxx\i.:.:
iussus a senatu aquarum Appiae Anienis Tepulae [a slip: see 8.:] ductus reficere. Vindicatio
(Berger ¸66) is the process by which existing ownership is confirmed, through
adjudication if necessary: Festus, s.v. vindiciae (p.¸:6 L) Cato in ea quam scripsit L.
Furio de aqua . . . praetores secundum populum vindicias dicunt. See also below :.8.:n.
ea spatia vindicarentur.
¡.z incrementumurbis There is no reliable demographic information,
but building projects attest a fairly rapidurbandevelopment fromthe endof the
third century. The years between :8¡ and :¸6, in particular, witnessed a sudden
spurt of public buildings which coincided with censorial efforts directed to the
water supply in :8¡ and :¸q. Morgan (:q¸8) ¸:–¸ argues that the incrementum
urbis to which F. refers was a sizeable and sudden increase in population during
the period after the last censorship (in :¡¸) which had produced an emergency
that could not wait for newcensors in :¡.. Three victorious armies (perhaps as
many as ¸.,ooo Roman citizens) had returned to Italy in :¡6–¸. Some veterans
might have elected to remain in Rome rather than resume farming with a risk
of renewed conscription in what appeared to be a new and unpopular war
against Viriathus. For population growth more generally, see Brunt (:q¸:) ¸8¡;
on the additions to urban water supply as an index of growth, Morley (:qq6)
ampliorem modum aquae Even with the Appia and Anio Vetus re-
stored to their rightful quantities there was pressing need for more water. The
problem was not new. In :¸q the censors M. Aemilius Lepidus and M. Fulvius
Nobilior had let contracts for building an aqueduct and constructing arches,
but M. Licinius Crassus blocked their project (Livy, xr.¸:.¸ duci non est passus,
conventionally taken to mean that he refused to grant right of way through
his property). There is no evidence that construction was ever begun on this
aqueduct, and it has even been suggested that the censorial activity in :¸q was
entirely unconnected with the urban water supply: Pennitz (:qq:) ¸o–:. In any
case, there is no record of any further efforts to increase the water supplies
between :¸q and :¡¡.
eidemmandatum Apparently second in priority was the mandate that
Marcius explore possibilities for augmenting the supply. The charge was left
vague: alias aquas cannot refer to a new aqueduct (in that case F. would have
used the singular aquam). Perhaps the senate had in mind increasing the supply
conveyedinthe older conduits: a nearby spring might be tappedfor the channel
of Appia, or additional river water might be diverted into the Anio aqueduct –
but the verb perducere does not bear this meaning, pace Morgan (:q¸8) ¡6–¸ and
n.:o6 (see ¸.:n. perducta sit). If the emergency was acute, an immediate increase
would have been highly desirable.
ut curaret . . . perducere For curare + infinitive cf. :¸.¸ formas facere
curavimus (the more familiar construction with gerundive occurs at ¸.:, 8.:).
If perduceret is kept, then quatinus must = ut, an unclassical use (cf. curare with
substantive clause Str. i.¡.¸, ii.¸.¸:, iii.8..). Better to contemplate that quatinus
(for the spelling see OLD s.v.) construes with posset, ‘to the extent that he was
able’. Transposing quatinus reveals more clearly howquas is intrusive; cf. Sauppe
(:8¸q) qq¸.
¡.j itaque . . . nomen est This mutilated sentence contains F.’s general
summary of Marcius’ accomplishments; in§¡ he turns todetails whichattended
the introduction of Marcia. My restoration follows Pliny, HN xxxi.:.: iussus
a senatu aquarum . . . ductus reficere, novam a nomine suo appellatam . . . adduxit. The
phrase in urbem perduxit responds to in urbem perducere of the senate’s mandate
and neatly contrasts with in Capitolium perduci / perductam of §¸.
[itaque pri]ores ductus ref[ecit et] The surviving letters clearly re-
fer to the repairs of the earlier aqueducts; the subject of [. . .]duxit can only
be Marcius, and its object must be aquam, the only possible antecedent for
cui. Kunderewicz’ Marcius is awkward, and Grimal’s qui is little better, for it
links this sentence too closely to the preceding. F. deliberately turns from the
senate’s mandate to Marcius’ execution of it. Asyndeton is possible, but C’s
lacuna gives encouragement for the summarising particle itaque: Jordan (:8¸8)
i: ¡6¸ n.88 ‘es fehlt vielleicht itaque, vielleicht Nichts’. B¨ ucheler based res[tituit]
on q.q Agrippa ductus . . . dilapsos restituit. Holste’s ref[ecit] picks up §: reficiendorum
and, more importantly, conforms to ubiquitous use of reficere in such contexts:
:...¸, :.¡.¡., :.¸, :.¸.:, Mon.Anc. .o.. (cited below :.¸n.), CIL 6.:.¡¡ ( =ILS
q8, Appendix B, no.:), etc.; see also §:n. vetustate. Archaeological evidence for
the Appia is too scanty for confirmation, but remains of Anio Vetus show
traces of repairs made in the mid-second century, plausibly ascribed to Mar-
cius’ project: Van Deman (:q¸¡) ¸q–6o, ¸8¸–qo. The repairs consisted of the
addition of reinforcing walls on the interior of the conduit (cf. quassati) and
occasional rebuilding of the upper portion of the channel. The workmanship
is very similar to that on the oldest parts of the Marcia, especially where crude
vaulting was used to replace an earlier pointed roof.
tertiam illa<m pro>prio ri[vo . . . per]duxit C’s t iam is legible
enough, but the textual tatters hereabouts do not inspire full confidence. For
the troublesome illiobrior(um) recent editors have adopted illis uberiorem, in-
dependently conjectured by Jordan (:8¸8) i: ¡6¸ n.88 and Krohn. Marcia’s
more copious supply was exactly what was needed in :¡¡ (cf. ubertas =copia
8¸.¸), but F.’s normal word is amplior (§., ¸¸.¸; cf. amplius 6¸.¸, 66.., etc.);
cf. abundantior ¸..:. Marcia’s quality is elsewhere praised (:..:, :¡.., 8q.¡,
q:.¸); thus some appeal to salubriorem. In this context we rather expect the
announcement that Marcius (:) effected repairs and (.) built a new aqueduct;
cf. Pliny, HN xxxi.:.: iussus a senatu aquarum . . . ductus reficere, novam a nomine suo
appellatam . . . adduxit. Thus I accept Schultz’s tertiam illam (for a/io confusion
cf. §¸ collegio] collega C), but without aquam, for the relative clause leaves the
antecedent in no doubt. From brior(um) I extract <pro>prio ri[vo]: see ¸..6, q:.¸
(cf. 8.., :q.¸, 66..). Confusion of b/p is frequent in this text (e.g. 8¸.¸ duplicata]
publicata C). It may originally have been thought that Marcius could bring more
water through already existing channels, and the decision to build an entirely
new aqueduct will have been an important outcome of the senatorial charge.
For [in urbem per]duxit see 6.:, 8.., Suet. Claud. .o.:; cf. :o.: Romam perduxit, also
in urbem with pervenire :¸.¸, :8.:, ¸..6. Here, in urbem perduxit exactly responds to
in urbem perducere (§.) and neatly contrasts with in Capitolium perduci / perductam
of §¸. See also ¸.:n. perducta sit.
ab auctore Marciae nomen The auctor (as at .¸.: and q¸.¡) is the
figure chiefly responsible for the introduction of the water (cf. ¸.: per quos).
Pliny, HN xxxi.¡: says vocabatur haec quondam Aufeia. The mysterious Aufeia
may be a figment of later Marcian propaganda, for Pliny goes on to say primus
eam in urbem ducere auspicatus est Ancus Marcius, unus e regibus. AQVA MR (MARC,
MRC, MARC) proudly appears on denarii struck by Marcius Philippus c. ¸6 ncr
(BMCR i: ¡8¸ nos.¸8qo–¸; RRC i: ¡¡8–q no.¡.¸), along with representation
of aqueduct arches above which there is an equestrian statue, presumably that
of Q. Marcius Rex which stood behind the temple of Jupiter: CIL ¸, p.8¡6
(a military diploma of 6¡ cr) ex tabula aenea, quae fixa est Romae in Capitolio post
aedem Iovis O.M. in basi Q. Marcii Regis pr(aetoris): Gesche (:q68) .¸–6; Hill (:q8q)
68. The Marcia is frequently mentioned in Augustan writers: Strabo, \.¸.:¸,
Vitr. \iii.¸.:, Tib. iii.6.¸8, Prop. iii...:¡. For Augustus’ personal interest in this
water, see below :..:n.
¡.q legimus apud Fenestellam Fr.:o (Peter HRR ii: 8:). This sentence
is parenthetical, andthere is no reasonto believe that the Augustanannalist was
F.’s sole or even his main source for the events connected with the introduction
of Marcia.
in haec opera Martio decretum De Klein (.oo:) :¸ supposes that
the appropriation was for all three of Marcius’ assignments (§:), but vindicatio
will probably not have been costly and opera properly is limited to building. The
plural, however, implies that the money had been intended to cover whatever
repairs or building might be necessary. Of course, we do not knowwhether the
outlay was a matter of one-time largesse on the senate’s part, for Fenestella’s
figure may have been a cumulative sum. Nowhere else does F. record a sum
appropriated for maintenance or construction (cf. :¸..n.). He may do so here
to emphasise the magnitude of the enterprise – easily the most costly building
project the Roman state had ever undertaken – and the strong senatorial
support behind it. Funds were available, of course, in the booty obtained
at Carthage and Corinth in :¡6 (cf. 6.:). For a discussion of building costs
associated with Roman aqueducts, see Blackman–Hodge (.oo:) 8o–¸, Leveau
(.oo:) q¸–¸.
sestertium milies octingenties Originally written MDCCC and
later transcribed as cardinals. If sestertium was not also abbreviated, this would
be :8o,ooo,ooo sesterces. There is no control for this figure. Crawford (:q8¸)
:¡8 cites this passage as evidence for the time when Romans began to express
monetary sums in sesterces (replacing the earlier asses). Although he is probably
right to do so (as he says, the construction of Marcia was no doubt costed after
its completion in :¡o), F.’s text is not contemporary evidence, for Fenestella
was writing a century or more later.
ad consummandum negotium Marcius’ unfinished negotium em-
braced the specific tasks of repairs and reclamation (§:) and the vaguer man-
date to increase the supply. In addition to straightforward reconstruction of the
earlier aqueducts, Marcius would have needed a certain length of time to de-
termine howbest to execute the mandate for augmentation. When he decided
on a new aqueduct, he would need still more time to arrange for the many
details of its construction. Morgan (:q¸8) ¡¸ has shown that Marcius’ negotium
was to make preliminary arrangements (ut curaret §.) for a new aqueduct; if F.
had meant ‘to finish constructing’ he would have used opus instead of negotium
(as in 6.., :¸..; cf. ¸.¸). From the verb adducere in Pliny, HN xxx\i.:.: intra
praeturae suae tempus adduxit, Morgan (:q¸8) ¡¸–6 also notes that at least Pliny’s
source was aware that not even a Marcius could accomplish both arrangement
and building in less than two years.
spatium praeturae This genitive of ‘office’ or ‘term’ has surprisingly
few parallels, but cf. Cic. Arch. : spatium praeteriti temporis, .q vitae spatium; Val.
Max. \iii.:¸(ext.).: regni spatium; Vell. Pat. ii.¡6.. prorogatae in idemspatiumtemporis
provinciae; Str. i\.¸... unius noctis ad deliberationem spatium.
in annumalterumest prorogatum It is probable that Marcius’ pro-
rogation was specifically and strictly limited ad consummandum negotium. Pliny,
HN xxx\i.:.: writes that Marcius novam . . . intra praeturae suae tempus adduxit
(cf. HN xxxi.¡: in praetura). It is not surprising, in a highly rhetorical passage
(very likely based on later Marcian propaganda), that Pliny fails to mention a
prorogation; but his phrase can easily include a legal extension of the prae-
torship. Pliny stresses the feats of engineering (cuniculis per montes actis), and it is
clear that he believed the Marcia reached the city while Marcius was praetor:
he juxtaposes the lavish accomplishments of Agrippa’s aedileship (eaque omnia
annuo spatio, but see below q.qn. and :o.:n.).
¡.j decemviri . . . dicuntur In :¡¸ the decemviri sacris faciundis consulted
the Sibylline Books in their charge on the occasion of the consul Appius’ initial
defeat in his war against the Salassi: Obseq. .: cum a Salassis illata clades esset
Romanis, decemviri pronuntiaverunt se invenisse in Sibyllinis, quoties bellum Gallis illaturi
essent, sacrificari in eorum finibus oportere. Taken with aliis ex causis, the discovery
has been thought to smack of deliberate fabrication. The decemviral objection
could have been conservative reaction against a gens which had imposed the
authority of the carmina Marciana in .:. ncr (Livy, xx\.:.): see Gag´ e (:q¸¸)
.¸¸–¸, ¸8¸–8. But the objection must have been more pointed, for – even
without delivery to the Capitol – the newaqueduct alone would have conferred
no small gloria upon its auctor. Morgan (:q¸8) ¡q–¸¸ points to a pragmatic
reason why the decemviri might have objected to a water supply for the Capitol:
certainloca publica around the Capitol had beenentrusted to their college (Oros.
\.:8..¸), and availability of water would encourage ‘squatters’ to encroach on
this undeveloped land.
non esse [fas] . . . perduci C’s short space (at the end of a line) is
not in itself sufficient justification for accepting fas: Sch¨ one (:8¸.) .¡q. But the
brilliant conjecture obviates the need to alter perduci and, equally, it eliminates
a’s sed from serious consideration. The position of the negative is decisive: it
is impossible Latin to understand ‘it was not right for the Marcia, but it was
right for the Anio’. B¨ ucheler’s tentative <sed Appiam> seu ought now firmly to
be rejected, pace Astin (:q6: ), Hainzmann (:q¸¸).
Marciam seu potius Anionem A vestige of the more prevalent tradi-
tion is found in the Oxyrhynchus Epitome of Livy, which covers the events of
:¡o (Oxy. Per. ¸¡). Despite physical damage and the likelihood of corruption, the
papyrus clearly reads aqua annio and [in Capi]tolium contra Sibyllae carmina. Tradi-
tions differed over whether the decemviri had found an objection to ‘Marcia’ or
to ‘Anio’, but the latter was more consistently reported. Context makes clear
that F. himself believed they had opposed the Marcia: this was the aqueduct
whose water Marcius victoriously carried to the Capitol. He records the preva-
lent version but does not explain how a Sibylline objection to ‘Anio’ (which
in his account can only mean the Anio Vetus) could have been used to block
delivery of water from the Marcia. The consistency of the tradition suggests
that ‘Anio’ was in fact at the centre of the issue in senatorial debates of :¡¸ and
:¡o. An objection to the Anio Vetus would seem to make little sense: only by
reconstructing its entire course could this aqueduct have been able to deliver
water to the Capitol (:8.6n.). An oracle which apparently banned the Anio
Vetus might have served the decemviri as a basis for arguing that no aqueduct
should be taken to the Capitol. Gag´ e (:q¸¸) ¸88 very tentatively suggests a
connexion with the prophecies of Albunea, the ‘Tiburtine Sibyl’. Oracular
obscurity ought perhaps not to be overlooked. The Sibylline ‘Anio’ might have
been the river itself: the new Marcia drew its water from springs which fed the
river and thus could perhaps have been called ‘Anio water’: Rodgers (:q8.)
M. Lepido pro collegio Verba faciente invites the supposition that
Lepidus raised the issue in the Senate; but did he do so on Marcius’ behalf
(pro collega) or on that of the decemviri (pro collegio)? The change is palaeograph-
ically straightforward, io misread as an open a. The structure of F.’s sentence
as a whole confirms the conjecture: Pighius (:6:¸) ii: ¡¸:. Lepidus led sena-
torial debate in :¡¸; Lentulus did so in :¡o; but on both occasions Marcius’
cause surmounted the opposition. The M. Lepidus who was spokesman for the
decemviri was probably M. Aemilius M’.f. Lepidus (cos. :¸8, RE no.¸o), a senior
member of the Aemilian gens and no doubt a decemvir himself: Bardt (:8¸: )
¸o, Sumner (:q¸¸) ¡¸–8, Morgan (:q¸8) ¡:, Broughton, MRR iii: ¸. Those
who have accepted the transmitted collega identify Lepidus with M. Aemilius
Lepidus Porcina (RE no.8¸) and thereby fix Porcina’s praetorship in :¡¸ (cos.
in :¸¸, he must have been praetor at least by :¡o), despite the technicality that
he would not have been the ‘colleague’ of the pro-praetor Marcius (see MRR
iii: q).
Appio Claudio Q. Caecilio consulibus :¡¸ ncr: MRR i: ¡¸:.
post annum tertium . . . retractatam In :¡o, apparently a last-
minute attempt to block the Capitoline supply when the aqueduct had been
completed and the distributory lines were being laid. Lentulus is probably
L. Cornelius Lentulus Lupus (cos. :¸6, RE no...¡), from whose family had
come L. Cornelius Lentulus, decemvir from .:¸ to :¸¸ (Livy, xx\...., xrii.:o.6).
An identification with L. Cornelius Lentulus (cos. :¸o, RE no.:q.) is unlikely,
and there are no solid grounds for setting this man’s praetorship in :¡o (MRR
i: ¡8: n.:, corrected in iii: 68).
C. Laelio Q. Servilio consulibus :¡o ncr: MRR i: ¡¸q.
gratiamMarcii Regis Inthe footsteps of M¨ unzer (:q.o) ..¡–¸¸, Stuart
(:q¡¸, :q¡¡, :q¡¸) constructed an elaborate nexus of political alliances, making
the effort to block Marcius a scheme of the party led by P. Cornelius Scipio
Aemilianus. On Marcius’ side were ranged Porcina (whose opposition to Scipio
is clear in :¸¸ when he attempted to block Ravilla’s lex tabellaria, which had
Scipio’s support) and Appius Claudius Pulcher (cos. :¡¸, a bitter opponent
of Scipio) as well as Galba and Cotta (coss. :¡¡). Opposing Marcius were
Lentulus (on no better grounds than that he was a Cornelius), Laelius (a warm
friend of Scipio), and of course Scipio himself (whose censorship in :¡. had
supposedly delayed Marcius’ endeavour). But this entire fabric of political
intrigue disintegrates for lack of evidence. In its place we have the senior
decemviri Lepidus andLentulus opposedtoMarcius, acting probably inresponse
to practical interests of that college. On Marcius’ side we have no one in
particular, only enormously successful gratia. There is no evidence at all for
the role Scipio Aemilianus may have taken. The decemviri might have been
overruled on procedural grounds: their discovery was made while consulting
the Sibylline Books aliis ex causis (although an unexpected discovery could have
been presented as more reliable). It may be that Marcius and his supporters
successfully dismissed the objection by arguing that the water in his aqueduct
was not the water forbidden by the Sibyl. Morgan (:q¸8) ¸: tentatively suggests
that some senators were willing to support Marcius because they hoped to take
up residence themselves on the Capitol and thus be able to influence public
assemblies held there.
atque ita in Capitolium For atque ita implying something of an abbre-
viation, ‘and so, to make a long story short’, cf. Str. i.¡.¸, ¸.8, 6.:, q.:, ii.¸.6,
¡.:o, ::.¸: Kortz (:8q¸) .¸–8. The Oxyrhynchus Epitome of Livy (Oxy.Per. ¸¡)
records for the year :¡o what must be the triumphal delivery of water to the
Capitol (see above). Since Marcius’ repairs to Anio Vetus were long since com-
pleted, the aqua Anio to which the epitomator refers is almost certainly the same
‘Anio’ to which the decemviri had objected. The epitomator is unlikely to have
wasted space on a second unsuccessful decemviral objection, so his report of
delivery to the Capitol means that the Marcia was completed in :¡o. Delivery
to the Capitol could most efficiently have been made by means of an extension
running from near the Porta Viminalis via the Quirinal along the ‘Servian’
wall: Bruun (:qq:) :¸¡ n.¸¡; Evans (:qq¡) 86. Of such an extension there is no
surviving evidence. Pipes to the Capitol in :oo ncr (Cic. Rab. Perd. ¸: fistulas
quibus aqua subpeditabatur Iovis Optimi Maximi templis ac sedibus; Florus ii.: sed cum
[Capitolium] abruptis fistulis obsideretur) were almost certainly lead pipes used for
distribution (cf. .o.. in usum urbis fistulis diducuntur). Nor is there evidence to
decide that Marcia’s delivery to the Capitol in the :¡os was by means of a
siphon as suggested by Lanciani (:88:) ¸::, followed by Van Deman (:q¸¡)
:¸q; Ashby (:q¸¸) :¸.; Van Buren (:q¡:–.) 6¸–¸o; DeAngelis d’Ossat (:q¡6);
Smith (:q¸6a) ¸¡; Cattalini (:q86a); Hodge (:qq.) ¡.8 n.¡¸; De Kleijn (.oo:)
:¸; cf. Evans (:qq¡) 86. If a siphon was used, it would be the earliest specimen in
Rome’s water system; on siphons in Roman hydraulics see Hodge (:qq.) :¡¸–
6o. Construction of the Marcia thus lasted from :¡¡–:¡¸ until :¡o. Even after
his praetorship Marcius’ involvement is clearly attested by the reference to his
successful gratia in :¡o. Since Plutarch, Coriol. :.: mentions Mópsici o nocv
sci lctìic, sci Kciv:c, cí tìtïo:cv 0ocp sci sóììio:cv tv Pcu n sc:c-
,c,cv:t, (whence Shakespeare, Coriolanus iii.¸), Morgan (:q¸8) ¸¡ suggests
that Quintus Marcius and a kinsman Publius delivered water to the city in the
capacity of duumviri aquae perducendae (see 6.¸–¡).
¡.6 concipitur . . . ducentos The close proximity of other springs (no-
tably those of Claudia: :¡.:–¸), the marked changes in natural conditions since
ancient times, and the presence of modern water works make difficult a precise
location of the source that F. describes. There is general consensus, however,
that Marcia’s springs are those at Rosoline: Panimolle (:q68) ¸:–qo, Fiore
Cavaliere (:qq¸) ¡6¡–¸.
via Valeria The ¸6th milestone of the Via Valeria was found about
¸oo m above the Anticoli bridge on the Anio: Ashby (:q¸¸) q¸.
euntibus ab urbe Roma With dextrosus, for clarity. The directional
reference is strictly unnecessary (cf. ¸.¡, ::.¸, :¡.:), but the Via Sublacensis
forked to the right at precisely this point with Marcia’s springs off that road to
the left. F. may have wished to avoid ambiguity.
Sublacensi . . . strata est The Via Sublacensis was built by Nero for
his villa at Subiaco (q¸..n.; Pliny, HN iii.:oq, Tac. Ann. xi\.....). Fabretti (in
the :6¸os) saw the ¸8th milestone in situ and on its original base, then in the
river bed: Ashby (:q¸¸) q¸. The water-men must have used the new road
(which ran closer to the spring) to the exclusion of the earlier deverticulum. To
account for the shorter distance, Lanciani (:88:) .¸q–8: suggested that the
older road followed the base of the mountains keeping above the springs while
Nero’s Sublacensis went straight across the plain; cf. Fiore Cavaliere (:qq¸).
F. no doubt took the earlier directions from a source which antedated Nero’s
construction; those for the alternative route may be his own addition.
¡.¡ Marcia’s source was inviting enough for Nero to pollute its potus sacri dur-
ing a stay at his nearby villa: Tac. Ann. xi\....¡ (cited above ¡..n.). ‘The springs
themselves lie under the rocks at the edge of the valley’ (Ashby (:q¸¸) q¸). Leg-
end gave Marcia a more distant source: Strabo, \.¸.:¸ ts ot :n, 1cusivc,
tïvci :c, tn,c, ío:cpc0oi :c0 Mcpsicu 0oc:c, :c0 :nv Pcunv tc:içcv:c,
sci tcpc :cììc toocsiuc0v:c, 0oc:c, and Pliny, HN xxxi.¡: oritur in ultimis
montibus Paelignorum, transit Marsos et Fucinum lacum, Romam non dubie petens; mox
in specus mersa in Tiburtina se aperit, ita novem milibus passuum fornicibus structis per-
ducta. F.’s sentence must have described the situation of Marcia’s spring, but
the transmitted text gives little reason to see this as an outburst of exuber-
ant diction like those of Pliny, HN xxxi.¡:, xxx\i.:.:: Baldwin (:qq¡) ¡q¸,
†fontin . . . petrei[. . . .]† Suggested restorations are in various ways
unsatisfactory. Since fontium has gained a firm foothold, it is well to remark
that we nowhere hear that Marcia had more than one spring: the singular fons
appears at :... and in Tac. Ann. xi\....¡; cf. Pliny, HN xxxi.¡: fons autem ipse
Pitonia. Perhaps fons M[arciae] would be nearer the mark, although this would
make an awkward subject for stat. Grimal’s sub rupibus is not unattractive, as
guesses go. From petrei we should probably extract petrae, since the adjective
petraeus is otherwise unattested. Earlier scholars sought to restore some mention
of masonry structures at the source (not unreasonably: see :..:, :¡.¸, ¸..8; cf.
im[mobilis] . . . perviridi Although immobilis or immota (Dilke (:q¸¸)
.¡8) goes well enough with the notion of a pool, the damage in the first part
of the sentence still leaves room for grave doubts about the transmitted statim.
Since this is a Latin word, it may be no more than erroneous decipherment.
While stagni may be perfectly sound, mo[do] is less certain. F. has a fondness
for genitive + more (or specie): Str. i.:..¸, ii.¸.:, ¸.:¸, see Kortz (:8q¸) ¸¸. C’s
praeviridi ‘conspicuously green’ is a ctcç, and only with much reluctance have
I accepted Poleni’s perviridi ‘intensely green’. The two abbreviated prefixes are
frequently confused, and perviridis itself is extremely rare (OLDs.v.). C’s reading
may be a matter of carelessness. Yet whombetter to accuse of carelessness than
Peter the Deacon? In his Liber de locis sanctis :o (ed. Franceschini, CCSL :¸¸,
q¸–:o¸), autograph ms. Cod. Cas. ¸6:, pp. 6¸–8o, Peter writes of the Red
Sea est limpidus et prelustris, whereas the text of Egeria’s Itinerarium from which
Peter was copying (he had before him what is now Arezzo ¡o¸, the surviving
archetype of this text) very clearly reads p(er)lustris: Devos (:qq:) 8¸–¸.
colore For the indication of colour cf. :¡.:n. Caerulo qui a similitudine ap-
pellatus est; see also Introd. .¡.
¡.8 sexaginta miliumet mille septingentorumdecem Elsewhere
there is no et in numerals, but neither do we find an instance of the thousands
broken as here (perhaps to avoid another awkwardness, e.g. sexaginta unum
milium or unum et sexaginta milium).
et semis If ¡¸,ooo for the length of Anio Vetus is suspicious for being a
round number (6.6n.), the precision of Marcia’s length – down to half a passus –
is even more so. From anything but a mathematical viewpoint the accuracy
of such a measurement is illusory: Blackman (:q¸q) :¸. I wonder if we should
change et semis here to ex eo: we find this phrase before rivo subterraneo at :o.¸,
:¡.¡, :¸.6. At q.¸ and :¡.¡ we have possible textual disturbances, both of which
involve the word sex, a convenient catalyst for dropping the ex eo phrase. Semis
could have been a clever misinterpretation of some illegible series of letters;
it appears just below only because it was needed for the arithmetic. It is also
worth noting that semis is not preceded by et the two times it is written out
elsewhere: ¸8.¸, 8o.. (8o.. and 8:.: have a numeral with S).
longius ab urbe Contrasted with propius urbem (cf. :¡.¡, :¸.6). For the
distance fromthe point near the seventhmilestone see :q.:n. Pliny, HNxxxi.¡:,
in obvious error, reports nine miles on fornicibus structis: Ashby (:q¸¸) qo.
8.r Cn. Servilius . . . Ravilla Servilius (cos. :¡:, RE no. ¡6), Cassius
(cos. :.¸, RE no.¸.). The sobriquet Ravilla is not otherwise attested: Kajanto
(:q6¸) ..8. It probably denotes distinctive eye-colour: Festus, p.¸¡o L raviliae a
ravis oculis, and p.¸¸8 L ravi coloris appellantur qui sunt inter flavos et caesios.
anno . . . M. Plautio Hypsaeo M. Fulvio Flacco consulibus :.¸
ncr: MRR i: ¸:o. For the AUC date see ¡.:n.
quae vocatur Tepula The name perhaps antedates the aqueduct; if so,
it could point to the existence of a cult associated with the unusual warmth of
the water (cf. Albula, Albulae: see :¡..n.). The stem tep- is certainly related to
tepere (§.n.), whether the suffix denotes agent or instrument.
ex agro Lucullano A distinctly different ager Lucullanus from that where
lay the sources of Appia (¸.¡) and Virgo (:o.:). The reference must be to an
estate of C. Licinius Lucullus (RE no.:o¡) near Tusculum (modern Frascati),
conceivably the Tusculan property mentioned by Cicero, Leg. iii.¸o, Fin. iii.¸
(cf. Varro, Rust. i.:¸.¸, Pliny, HNx\iii.¸., Colum. i.¡.6). The belief that this was
rather ager Tusculanus may simply be one of terminology, although confusion
couldhave arisenfromthe proximity of the Crabra, whichbelongedtoTusculani
possessores (q.¡–¸).
Romam et in Capitolium For et = et quidem cf. Str. i.::.¸ adversus
Germanos et Ariovistum. In this case delivery to the Capitol was accomplished
apparently without opposition (¸.¸), although F. perhaps means only to note
that Tepula could supply higher elevations (:8.¸–¡). Conceivably a higher
level than that of Marcia (even if only slightly) might have made a noticeable
adducendam curaverunt ‘Arranged for the introduction’. Curare ap-
pliedtoletting the contracts; the work presumably took longer thanthe censors’
term, hence adducere instead of perducere (cf. ¸.:n. perducta sit).
8.z Tepula concipitur Holste (in :6¡q) was the first to identify Tepula’s
source as the Sorgente Preziosa, located about . km west of Grottaferrata
(Ashby’s Map .). Because of residual volcanic activity, the temperature in
winter measures :6


C while that of the nearby Julia is :o


C: Ashby
(:q¸¸) :¸q–6o; Cattalini (:q86c) 6o.
ad decimum miliarium The number usually follows the word mili-
arium and I suspect an unnoticed transposition. Rocchi (:8q6) :.6 wanted
to emend decimum to XI because he believed F.’s deverticulum was not the via
Cavona at Ciampino (exactly ten miles from Rome) but rather a hypothetical
road leading off the Via Latina between the tenth and eleventh milestones.
In that case F. would very likely have been more specific; cf. ¸.¡ inter miliarium
septimum et octavum.
{euntibus ab Roma} The phrase is no more necessary here (or at q.:
just below) than at ¸.¡, ¸.¸, ::.¡, :¡.:. It could have been added by a reader or
copyist from ¸.6 above. C’s blank space after dextrosus, where nothing seems to
be missing, could be a vestige of an interpolator’s activity. The form ab seems
unusual (:n. mihi ab Nerva), and if these passages are to stand we ought perhaps
to read ab <urbe> Roma.
[rivo] suo Giocondo’s rivo is unobjectionable (cf. q.: Tepulae rivum), nor is
it necessary (with B¨ ucheler) to transpose it to follow suo (which here = proprio;
cf. :q.¸ [Tepula] proprio canali et nomine venit). Grimal ¸:–. n... contemplates
suo <opere>, but opts in the end for suo <iure> (which he translates ‘de fa¸ con
autonome’). Taylor (.ooo) :.: n.6. is unwise to stretch Grimal’s iure to mean
iure territorii (i.e. ‘under the jurisdiction of its destination city’).
perducebatur Note the imperfect. Inthis single instance the verbperduce-
re appears to have no specific connotation of new construction; it still bears its
usual meaning ‘to deliver into use’ (¸.:n. perducta sit). Subsequent to Agrippa’s
projects (q.:–.) the water of Tepula was conveyed with that of Julia and its
identity (in a separate channel only from the latter’s piscina: 68.¸) was purely
nominal. F. gives no length for Tepula: he thinks of it primarily as a distributory
system for other aqueducts (see q.., :q.¸, 68, 8.). Of Tepula’s original conduit
virtually nothing is known. No remains have been discovered between the
source and Capannelle. It might safely be surmised that its route ran close
to that of the later Julia and that it was abandoned or, better, completely
reconstructed to formthe conduit of Julia. The original Tepula must have been
carried along with Marcia to an urban terminus near Porta Viminalis (cf. 8.:
in Capitolium), but the extant remains of the channel atop Marcia’s arcade seem
to be roughly coeval with those of Julia: Van Deman (:q¸¡) :¸¡, :6¸–¡. Ashby
(:q¸¸) :¸:–. calls attention to the fact that Tepula’s channel is narrower than
that of Marcia and notes the deliberate (but in the event structurally unwise)
decision to place it off-centre. My guess would be that the original Tepula was
built directly atop Marcia’s cover-slabs; resulting damage could explain both
Agrippa’s wholesale reconstruction and the unusual manner of superimpos-
ing the higher specus. Cf. Hodge (:qq.) q¡–¸: ‘The Tepula’s specus was placed
in the way described because it had less water to carry. Too large a channel
would make the water stagnate, yet reasonable head-room was required for
the cleansing staff.’
q.r Postea M. Agrippa Postea (cf. .¸.:) outshines other restorations, be-
cause F. is still in effect discussing the Tepula (and explaining the sense of
imperfect perducebatur at 8..). Renaissance readers knew Agrippa’s praenomen
(needed here: see 6.¸n.): it was prominently visible, for example, on the
Pantheon (CIL 6.8q6, conceivably the source of a’s tertium at :o.:).
aedilis post primum consulatum Cf. q8.: M. Agrippa post aedilitatem
quam gessit consularis. Agrippa’s first consulship was in ¸¸ ncr (MRR ii: ¸q6). To
accept the lower office was an extraordinary act to win support for Octavian,
and the munificence of his aedileship was legendary (e.g. Pliny, HN xxx\i.:o¡,
Dio, xrix.¡¸.:–¡); note also Horace, Serm. ii.¸.:8.–6. See Reinhold (:q¸¸)
¡¸–¸.; Shipley (:q¸¸) ¡¸–8; Roddaz (:q8¡) :¡¸–¸¸.
Caesare Augusto.II. L. Volcatio consulibus ¸¸ ncr; MRR ii: ¡:¸–
:¡. The title Augustus is strictly an anachronism, but not unduly distressing in
a consular date.
anno For the AUCdate see ¡.:n. Dio, xr\iii.¸..¸ records the introduction
of Julia during Agrippa’s praetorship in ¡o (MRR ii: ¸8o): sci : c co: c :cú:c
ypcvc :c :t 0ocp :c lcúìicv cvcucoutvcv t, :nv tcìiv ttcyt:túûn. The
majority of scholars incline to accept Dio’s date, supposing that F. wrongly
includes Julia with the extensive water projects of Agrippa’s aedileship (q.qn.):
Shipley (:q¸¸) .¸–8; Ashby (:q¸¸) :6:; Hanslik (:q6:) :.¸¸; Ehlers (:q8¸) ¸8–
q. Others impute an error to Dio, who might have confused praetorship with
aedileship or whose source may have been less reliable: Hainzmann (:q¸¸)
:oq; Roddaz (:q8¡) 6., :¡q. Fence-sitters propose that work on Julia was begun
in ¡o, finished in ¸¸: Gardthausen (:qo¡) ii: 6o8; Robinson (:q8o) ¸6. They
may be right. Steps to develop the new source could have been taken in ¡o
(conceivably in consequence of some scheme of Julius Caesar: see q..n.), with
its water at first conveyed in Tepula’s existing channel (perhaps partly rebuilt
at this time). In ¸¸ there could have been extensive rebuilding of the combined
Tepula-Julia, much of it new construction and with an eye towards improved
{euntibus ab Roma} See 8..n. Again, the words are otiose (here es-
pecially so, after miliarium ab urbe), and it might be reasonable to suppose that
they originated as a supralinear addition which then ousted deverticulo.
milium passuum duum Ashby (:q¸¸) :6. notes that two miles seems
excessive and supposes that F.’s figure is not to be taken as an accurate mea-
surement (cf. similarly round figures at 6.6, ¸.6, ::.¸). There may, however, be
a textual error, especially since the distance on this deverticulum is precisely the
same as on that for Tepula (8..).
alterius <a>quae proprias vires collegit Another, that is, distinct
fromTepula. Agrippa gathered water fromseveral individual veins into a single
supply (cf. 6q.. ex pluribus adquisitionibus constat). These were the propriae vires of
Julia. (Note the abnormal plural of vis = copia, explicable here perhaps because
of the multiple source.) Colligere is similarly used of combining small quantities
of water at :o.: collectam; cf. Vitr. \iii.:.: sin autem non profluent, quaerenda sub terra
sunt capita et colligenda, Pliny, Ep. x.6:.¡ qui [rivi] si diligenter colligantur, augebunt
illud . . . Julia’s source has been located near Ponte degli Squarciarelli about
:¸oo m southeast of Grottaferrata: Ashby (:q¸¸) :6.–¡.
Tepulae rivum intercepit Cf. :q.¸ (Tepula) intercepta . . . rivo Iuliae
accesserat, 68.. (venae) interceptae sunt in Iulia. In these instances the verb intercipere
(¸.:n.) means that Tepula’s flow was cut off from its own original channel and
its water conveyed in that of Julia. Tepula’s spring, therefore, was reduced to
the status of a feeder for Julia.
q.z adquisitaeque . . . nomen Iuliae I prefer adquisita<e>que because
it links the detail of appellation more closely to the names and date of the
introduction (cf. ¸.:, 6.:, 8.: etc.: note especially ¸.¸, :..:). Schultz’s version,
however, is equally close to C’s reading. For the noun adquisitio see :o.6n.
The aqueduct was called Julia until it was divided at the piscina (cf. :q.¸), and
Augustan cippi bear the single name Iul(ia).
ab inventore It is easy to take the phrase as ablative of agent (‘by its
builder’), with the obvious inference that Agrippa named it in honour of Julius
(Octavian); cf. Dio, riii..¸.. lcúìic [sc. :c 2tt:c] co:c ótc :c0 Ao,cúo:cu
tpcoc,cptúoc,. But F. nowhere else tells us who named an aqueduct, though
he often explains the origin of a name (¸.¸ ab auctore, :..: ab inventore, :¡.: a
similitudine; cf. :o.¸ Virgo appellata est, quod . . .). The inventor might have been
none other than Julius Caesar (q.:n.), for a project of the Dictator lay behind
Agrippa’s work on the Portus Julius near Cumae (built during his consulship
in ¸¸): cf. Tortorici (:qqo) .:–6.
ut maneret Tepulae appellatio See :q.¸ Tepula . . . a piscina eiusdem Iu-
liae modum accepit ac proprio canali et nomine venit, and 68.¸ caput ergo eius observandum
est a piscina Iuliae. The nominal Tepula ran in a separate channel from Julia’s
settling-tank to the City (:q.¸–¡) and served thereby as a separate supply in the
scheme of distribution (68.¸, 8..:–.). The water which it conveyed consisted
of supplies received from Julia’s piscina, from Marcia shortly thereafter, and
finally from Anio Novus near the City (68.¡). To rebuild the channel of the
earlier aqueduct was perhaps related to practical matters of engineering and
distribution; to preserve its identity was no doubt meant as a kind of pietas (cf.
¸6.¸ quamvis mutata aqua vetus appellatio mansit).
q.j <rivo subterraneo . . . sex> Only for the unappealing Alsietina
(::.¡) does F. omit the full computation (Table :), and homœoteleuton could
account for the loss. On the presence of C’s strange form sexs and my addition
of ex eo see ¸.8n. et semis. My restoration assumes that :¸,¡.6 is the correct figure
for the total length. Cippus no. ¸o., found near the Abbey of Grottaferrata,
implies approximately :¡ :/. miles from this point to Rome (although exact
lengths cannot be reckoned from the cippi; see :¸.:n.).
a septimo miliario These figures, of course, are identical to those for
Marcia (¸.8; see Table :), for the arches of the earlier aqueduct carried the
channels of Tepula and Julia (:q.¸–¡).
q.q Crabra The source of Crabra (RE ¡: :68o) lies in the Valle della
Molara under Colle Bartolucci, north of the eighteenth mile of Via Latina. Its
water still supplies Frascati: Lanciani (:88:) ¸.:–¸, Ashby (:q¸¸) :6¸–¡.
q.j seu quia . . . sive quia The variation is admitted by post-Augustan
writers (L&S); cf. :8.¡ sive + ablative absolute . . . seu quia.
Tusculanis possessoribus The Tusculans had legal claims on the
Crabra which Agrippa is not likely to have brushed aside (see Cic. Agr.
¸.q ego Tusculanis pro aqua Crabra vectigal pendam, quia mancipio fundum accepi;
cf. Fam. x\i.:8.¸). For irrigation in general, see Hodge (:qq:) .¸o with
haec namque est The postpositive position of namque (frequent from
Livy on: K–S ii: ::¸–:¡) lends emphasis to haec.
tractus eius The word in this same concrete sense of area or district
(OLD s.v. ¸) recurs below .:...
per vicem. . . dispensatam A marble slab from the Augustan period
found on the Aventine (CIL 6.:.6: : Mommsen attributed it to the Crabra)
exemplifies a schedule with details of water-rights with names, hours and
numbers of sluices. Bruun (:qq:) ::. n.68 observes that this is the sole reference
in F. to division of water by any system other than constant flow. For another
epigraphic record of distribution see ::.:n. hortis.
in dies modulosque certos Both by set days and by exact delivery
gauges (¸..n. modulorumque rationes). Note that dispensare is semantically approx-
imate to F.’s usual word erogare (¸..n.).
q.6 aquarii nostri Seems to contrast with Agrippa’s righteousness. Nos-
tri may mean ‘of recent days’ or ‘whom I inherited when I got this job’; cf.
Evans (:qq¡) ¸¸. Perhaps deliberately, F. never precisely defines what he means
by aquarii. There emerge clearly from his usage, however, two overlapping
senses. (:) Staff workmen are apparently responsible for monitoring intake to
guarantee constant level of supply (here, ¸:.:) and for effecting transfers within
the system such as those noted at q:.¸ (and some at least of those involving
pipes, ¸..¸–6, ¸¸.:); cf. ¸.6n. (.) Staff members have a variety of roles in mak-
ing distributions to privati, and thereby ample opportunity for misbehaviour
– very often deriving profit therefrom. The element of profit is explicit here,
::¡.: intolerabilis fraus . . . venalem extrahunt aquam, ::¸.: reditus, and see :oq..n.
Elsewhere it is everywhere implicit, usually by use of fraus with (¸¸.., 8¸..,
::o..) or without the word aquarii (¸¡.¸, ¸..¡, ¸¸.:, :o¸.¡). In most of these
cases there is an unambiguous connexion with privati, who would obviously
have been party to the fraud (cf. ¸.: privatorum fraudibus). Beyond the collec-
tive appellation, however, F. is not very helpful. On implementation of an
imperial beneficium he warns that vilici are subject to fraus and libratores to the
possibility of breaking rules pro gratia personarum (:o¸.¸–¡). But at ::..¸–¡ he
casts his net wider and catches not only the vilicus but the imperial procurator
(:o¸.:n.) and even the illegal beneficiary (accipientis). At the procurator F. levels
a charge of ambitio (::..¸), and in the aquarii he detects the undesirable qual-
ities inertia (8¸..), neglegentia (:o¸.¸) – not perhaps accidentally the same vices
he elsewhere attributes to the curators themselves prior to his appointment
(e.g. :o:.. inertia et segnitia). He may be treading lightly here, not to implicate
overtly any among his senatorial colleagues. No cura, no ordo is always immac-
ulate: cf. Suet. Vit. ¸, c. 6¸ cr: in urbano officio dona atque ornamenta templorum
subripuisse et commutasse quaedam ferebatur, proque auro et argento stagnum et aurichalcum
semper . . . vindicaverunt . . . augerent F. uses vindicare here in
a non-legal sense (OLD s.v..; contrast ¸.:n. vindicandorum), for context makes
plain that his aquarii have given no formal notice to their victims. They would,
on the other hand, have had a ready pretext that they were acting to maintain
the requisite supply for Rome.
hauriebant . . . gratia. Note tense: their practice is a thing of the past.
There is a certain indignant flourish in largiendo . . . gratia, for something like
sibi lucro or vendundi causa might have done as well (cf. :oq.. intercipere solebant
ut . . . venderent).
q.¡ exclusa . . . Tusculanis It hardly matters for sense whether one
reads exclusi Crabram / totam reddidi or exclusa Crabra / tota reddita, but the con-
structions ought in any case to be parallel. The passive is preferable, for in
speaking of his official actions F. seems carefully to avoid using the first-person
singular, using a kind of ‘official’ plural instead (see .¸..n. invenerim): e.g., q:.¸
deprehendimus, ::¸.¡ instituimus, :¸o.. laboravimus; cf. :¸.¸ nostrae sollicitudini, :o:.¡
nobis. Note especially 6¡.:, :o¸.¸, ::8.¸, :¸o.¸, where the imperial agency is
explicit. Baldwin (:qq¡) ¡qq, however, finds here ‘first person verbs contrasting
vigorously with the royal plurals’ F. uses elsewhere.
iussu imperatoris The noun iussus conveys the sense of authority (OLD
s.v.); note q¸.. iussit, also of the princeps.
forsitan non sine admiratione Litotes may indeed express ‘light
irony’: Baldwin (:qq¡) ¸o¸. Hodge (:qq.) ¡¡8 n.:¸ observes that these per-
sons’ surprise implies that they were located at some distance downstream.
q.8 modum . . . notabili siccitate servavit ‘A period or condition
of conspicuous drought’ sometime after F. took office, probably during the
summer of q¸. Note the similar use of the perfect tense at ¸¡.¸ tota deinceps
aestate . . . exploravi (cf. Introd. 6). For words relating to weather, F. normally uses
the plural: :..:, ¸¡.¸ siccitates; :¸.: pluviarum iniuria (but cf. 8q.¡ pluvia inquinatur);
:¸.¸, 8q.¸, qo.. imbres; :.o vi tempestatium (but cf. Str. i.:.. vi tempestatis); :.:.¸,
:.¸.. caloribus; :.:.¸ gelicidiis.
q.q eodem anno F. assigns Agrippa’s repairs to the aedileship of ¸¸
(q.:n.); cf. Pliny, HNxxx\i.:.: Agrippa vero in aedilitate adiecta Virgine aqua [!] ceteris
conrivatis atque emendatis lacus DCC fecit praeterea salientes D, castella CXXX, complura
etiam cultu magnifica, operibus iis signa CCC aerea aut marmorea imposuit, columnas ex
marmore CCCC, eaque omnia annuo spatio. Dio, xrix.¡... dates the restoration of
Marcia to the previous year: c :t A,pittc, :c 0ocp :c Mópsicv cvcuco-
utvcv, tsìitcv gûcpc :cv cyt:cv, sci óvts:noc:c octóv n cistic sci tti
tcììc :n, tcìtc, ttcyt:tuotv. That Agrippa accomplished so much in one
year is no doubt an oversimplification, inferred perhaps from the account in
Agrippa’s own memoirs (Introd. :: n.¡¸). The period of his aedileship, on the
other hand, marked a clear beginning of what was Agrippa’s personal and
lifelong responsibility for Rome’s water supply (q8.:, :o¡.:).
ductus . . . restituit It is difficult to distinguish repairs which may be-
long to Agrippa’s project of ¸¸ from those carried out a few years later by
Augustus (:.¸n.): see Van Deman (:q¸¡) 6o–., :.8–¸o. But at Ponte Lupo
it is possible to see two stages of reconstruction, of which the earlier proba-
bly belongs to Agrippa: Van Deman (:q¸¡) q¸–8, Ashby (:q¸¸) ::8–:q with
figs. :: and :..
singulari cura The sense ‘personal (individual) initiative’ overlaps with
‘unique forethought’ (Bennett). The phrase alludes to Agrippa’s single-handed
management subsequent to his aedileship: cf. q8.: operum suorum et munerum velut
perpetuus curator. Note alliterative chiasmus: singulari cura compluribus salientibus.
compluribus salientibus {aquis} Pliny, HN xxx\i.:.: credits
Agrippa with ¸oo salientes. The word aquis is superfluous: salientes (substantive,
masculine plural) are pipes from which water flows constantly. They need not
have been‘fountains’ (or ‘jets’), for salire canmeansimply to flow(Colum. i.6.::;
Pliny, Ep. ii.:¸..¸): Bruun (:qq:) :o¸–6; Del Chicca (:qq¸) .¡o–6; Blackman–
Hodge (.oo:) ¸. In F. we usually find them called salientes publici (::.., 8¸.¸,
:o¸.¸–¡, :o¡.:–¸) and associated with lacus (¸..n.; cf. also ¸..n. munera).
ro.r tertium consul While a might have got tertium from its prominent
appearance in the inscription on the epistyle of the Pantheon (CIL 6.8q6 =
ILS :.q), it is undoubtedly the ‘right’ reading. F. might have been hesitant
(like Pompey: see Gell. x.:.6), but I should prefer to see C’s tertio as a case of
ignorance or carelessness on the part of a scribe. Agrippa’s second and third
consulships (both times as colleague of Caesar) were in .8 and .¸ ncr (Degrassi
¸: in the latter year Caesar received the title Augustus). It was at this time that
he undertook an impressive series of building projects in the Campus Martius,
including the Pantheon, the completion of the Saepta Julia (started by Julius
Caesar: Cic. Att. i\.:6.¡), and what was to become the first of Rome’s public
baths. The baths were built some years prior to the new aqueduct, such water
as they required initially being supplied by other aqueducts. Virgo furnished
them later, to be sure, but it is misleading to suppose that to supply merely the
baths was its primary purpose (cf. 8¡..–¸). On the baths (Thermae Agrippae):
P–A ¸:8–.o, NTD ¸86–¸, LTUR \: ¡o–. (Ghini); Lloyd (:q¸q), Fagan (:qqq)
C. Sentio <Q.> Lucretio consulibus :q ncr (Degrassi ¡). Despite
omission of an AUC date, post annum tertium decimum brings us also to :q, for
the ‘:¸ years’ are those which intervene (¸.–.o): note the parallel post annum
tertium at ¸.¸ (for the interval between :¡¡ and :¡o).
COMMENTARY :o..–:o.¸
Iuliamdeduxerat It is curious that F. does not use perducere (¸.:n. perducta
sit) for Julia, either here or in chapter q (contrast Virginem . . . perduxit just below),
conceivably because of different circumstances surrounding the introduction
of Julia.
Virginem. . . perduxit Dio, LIV.::.¸ :c :t 0ocp :c lcpûtvicv scìcú-
utvcv :cï, ioici, :tìtoiv toc,c,cv A0,cuo:cv tpcon,cptuot. Virgo shared
with Marcia the praises of poets: Ovid, Fast. i.¡6¡, Ars Am. iii.¸8¸, Pont. i.8.¸¸;
Stat. Silv. i.¸..¸; Mart. \..o.q, \i.¡..:8, \ii.¸..::, xi.¡¸.6. Pliny, HN xxxi.¡.
compares its excellence to that of Marcia: quantum Virgo tactu praestet, tantum
praestet Marcia haustu (Virgo’s softness preferable for bathing, Marcia’s hardness
for drinking). According toDio, ri\.::.¸ Augustus recommendedit tothe popu-
lace when they complained of scarcity and the high price of wine: íscvc:c:c
tgn :cv A,pittcv tpcvtvcnstvci co:t un oi¸ n tc:t co:cu, ótcìtoûci
(cf. Suet. Aug. ¡..:).
in agro Lucullano collectam The same ager Lucullanus where Appia’s
source lay (¸.¡n.). F. uses collectam (cf. q.:n. collegit) because Virgo, like Julia,
consisted of several feeders in addition to the main spring (§6n.).
ro.z die . . . invenitur ‘The day on which it first came forth (?) into
the City is discovered as the ¸th before the Ides of June.’ I take invenitur as
impersonal, with indirect question die quo ( =quo die, transposed because die
construes also with quinto) responderit as well as implied indirect statement (e.g.
invenitur aquam quinto die respondisse). An exact date can be given for the intro-
duction of newer aqueducts (cf. :¸..), perhaps from calendar entries (like that
in Fast. Ost. recording the new Aqua Traiana in :oq cr), although in this case
it may have come from Agrippa’s memoirs (Introd. :: n.¡¸). It is probably no
more than coincidence that q June is the date of the Vestalia, for there is no
apparent connexion with the Vestal Virgins.
responderit The use of respondere (‘declared its presence’?) could be an
extension of a legal use (e.g., with citatus or vocatus); cf. also Cels. i\.::.. sanguis
per menstrua non respondet. But perhaps this is a technical or colloquial use which
F. may have taken directly from his source – if this was Agrippa’s memoirs,
then conceivably a peculiarity of the latter’s style. See also vocaverunt below
ro.j Virgo appellata est Dio states that Agrippa named it Augusta:
ri\.::.¸ (cited above). But this appellation nowhere recurs and Dio could
have confused another Augusta, perhaps the branch of Appia which Agrippa
may have built in conjunction with Virgo (¸.6n.). F.’s explanation (with its
pictorial corroboration) is certainly more plausible than that of Pliny, HN
xxxi.¡.: iuxta est Herculaneus rivus, quem refugiens Virginis nomen obtinuit. Cas-
siodorus, centuries later, links the name to the quality: Var. \ii.6.¸ currit aqua
Virgo sub delectatione purissima, quae ideo sic appellata creditur, quod nullis sordibus
COMMENTARY :o.¡–:o.¸
polluatur. The purity of the supply might have reinforced the name: Svennung
militibus Perhaps military engineers at Agrippa’s disposal, whose prac-
tical experience with locating water would have been invaluable: Ashby (:q¸¸)
¸¡. The diggers who came later might also have been milites.
puella virguncula B¨ ucheler’s omission of puella was inadvertent, and
it is unlikely that either word is a gloss. The unusual attributive noun (cf. q6
servorum opificum) is perhaps intended to stress that this was a very young girl
(puella having long since lost its original diminutive force): cf. P. Watson (:q8¸).
Virguncula seems to have appealed to Silver Latin writers (Petr. :8.¸ and .o.8,
Sen. QNat. i.:¸.q, Curt. \iii.¡..¸, Juv. :¸.¡o), but F.’s attributive use is elsewhere
unattested: Baldwin (:qq¡) ¸o¸, who also notes the alliteration with venas. Very
similar is Pl. Tht. :6:d µó:pcyc, ,upïvc, (where some delete µó:pcyc,); cf.
also homo nemo ( Plaut. Miles ¸¸., Ter. Eun. :o8., Cic. Nat. D. i.¸8). It is tempting,
but unwise, to think in terms of virgula ‘dowsing rod’: Corsetti (:q¸¸) ¸:, Van
Buren (:q6: ) .oo; cf. Hodge (:qq.) ¡oq n..¸, who traces the first appearance
of a virgula furcata to Georg Agricola’s De re metallica (:¸¸o).
quas secuti qui foderent C’s foderant is illogical: while these men had
dug before they found water (vocaverunt), they had not yet dug when they were
following. Better is Professor Reeve’s foderent: ‘those who followed to dig’. But
what or whom did they follow? C’s quas (i.e. venas) is probably right, although I
can find no parallel other than Lucr. \i.8o8 argenti venas aurique secuntur. Yet we
might read either quam (the young girl) or, more likely, quos (the soldiers who
were reconnoitring). For the latter cf. Tac. Hist. \.¸.. secutus Moyses coniectura
herbidi soli largas aquarum venas aperit, where the object of secutus is grex asinorum
vocaverunt Long troublesome (hence a’s invenerunt), but apparently =
evocaverunt, an example of a simplex verb bearing the meaning of one of its
compounds: H–Sz .q8–q; Reynolds (:q6¸) :.6; Goodyear’s note on Tac. Ann.
i.¸¸.¡ (p.8o). I can find no parallel, but Krohn adduced Sen. QNat. iii.::.¸
alias quoque causas intervenire opinatur, quae aliter evocent aquas. Bendz (:q¡¸) .¸ cites
two instances from Str.ii where editors have emended (he thinks unnecessarily)
to evocare: ¸... ad vocandum hostem, ¸..6 vocatis suis. If there is some connexion
with respondere just above, one might note Sen. Ep. .¸.¸ vena [sc. metalli] plenius
responsura fodienti (cf. Pliny, HN xxxi.¡8).
ro.q aedicula . . . ostendit The painting was perhaps an ex voto in a
small shrine erectedat the source. The fons wouldno doubt have beenrespected
for its sanctity (cf. ¡..n.).
ro.j concipitur Virgo Cf. Pliny, HN xxxi.¡. idem et Virginem adduxit
ab octavi lapidis deverticulo duo milia passuum Praenestina via. The discrepancy
is insignificant: Pliny follows the more important Via Praenestina, F. the
Collatina (¸.¸n.), which ran closer to the spring (and thus he needs no dever-
ticulum; cf. :¸.:). The principal source is located about 6oo m to the south
of the railway station of Salone, just to the left of the modern Via Collatina:
see Quilici (:q68) :¸¡–6o. It supplies the modern Acqua Vergine (Fontana di
signino circumiecto Note alliterative chiasmus signino circumiecto conti-
nendarum scaturriginum causa. Opus signinum (cf. Vitr. \iii.6.:¡, Pliny, xxx\.:6¸)
was a waterproof type of concrete used to line the conduits. The large collect-
ing basin F. describes here seems to have been in part preserved as late as the
eighteenth century. Its primary function was to prevent the natural flow of the
water towards the Anio.
scaturriginum The word is rare: Livy, xri\.¸¸.¸ scaturr<ig>ines emicare
coeperunt; Colum. iii.:¸.8 scaturrigo palustris; Pliny, HN xxxi.¡¸.
ro.6 adiuvatur . . . adquisitionibus These tributaries had a total
length of :,¡o¸ paces (§8). They merged with the main channel a short distance
later. For remains, see the map in Quilici (:q68) :¸6.
adquisitionibus The neologism adquisitio occurs seven times in this
text: Espinilla Buis´ an (:qq¸) 6¡¡. Five times F. uses the plural (also §8, 6q..,
¸o.., ¸..¸) as a concrete noun indicating ‘feeders’ or contributory chan-
nels. The other two instances (¸¸.¸, ¸¸.:), singular + genitive, approach our
ro.¡ venit per longitudinem Because compluribus locis seems to apply
only to substructio (outside the City: cf. ¸.¸, ¸.8, :¡.¡, :¸.6), one would expect
that opus arcuatum is to be identified with the arcus Virginis delineated at .... –
especially since ¸oo passus (:,o¸6 m) approximates their length. Remains above
ground, however, total considerably more than the ¸¡o passus F. gives for sub-
structure (these are: at Bocca di Leone ¡¡8 m, at Gottifredi .6o m, at Pietralata
¡¸. m: total :,:6o m). A stretch of some .oo m is known to have existed in
the valle di S. Agnese, and one near Pietralata was carried on arches in F.’s
day: Quilici (:q68) :¡.–¡. It is therefore highly probable that F.’s figures for
both substructures and arches are for the extra-urban course – excluding in the
overall longitudo the arcade which ran fromthe Pincian to the Saepta. Evidence
from the cippi (:¸.:n.) supports this view: those numbered ‘:’ were found at
Villa Medici. A straight course between the source and the urban terminus
might have been shorter by as much as three miles (cf. the length of Appia:
¸.¸). But Virgo followed a more circuitous route, turning sharply northward
from Via Collatina at Portonaccio and running first north, then west, and
finally south (beneath the Pincian): Ashby (:q¸¸) :¸.–¸, Quilici (:q68) :.6–8.
Lanciani (:88:) ¸¸¡ attributed the route to a desire to avoid the engineer-
ing difficulties of a long (c. ¡,.¸o m) and deep (c. ¸¸.¸ m) tunnel, but Quilici
(:q68) :.8 points out that the chosen route had fully comparable engineering
challenges of its own, and he suggests that the planners might have sought
rather to avoid construction through (albeit far beneath) thickly populated
parts of the City. An approach from the north brings the aqueduct to its final
destination in the Campus Martius without traversing built-up areas. Engi-
neering decisions might rather have been influenced by the relative ease of
acquiring right to the necessary land. Grimal q8 n.:¡¸ suggests that Virgo ran
entirely through public lands. One may be sure that Agrippa made appropriate
legal arrangements, whether he acquired the land personally (Dio, ri\.::.¸) or
on the part of the state. Taylor (.ooo) :o¸–6 concludes that Virgo’s circuitous
route was chosen because Agrippa had no recourse when landowners refused
permission (cf. :.¡.¡n.).
decemduum Cf. 6¡.. decem duo, ¸..¸ decem duas. In these three instances
F.’s usage is apparently unique (K–S i: 6¡o), but he has the regular duodecim at
¸6.¡, 6¸.:, 8.... The irregularity could result from writing out numerals.
rr.r Quae ratio F.’s apparent contempt for the quality of Alsietina and
his summary treatment of it are no doubt due to the fact that it was by his
day a very problematic supply, reduced to an unreliable trickle because the
level of Lake Alsietinus (§¸n.) had begun to fall towards the end of the first
century cr: Liberati Silverio (:q86a), Taylor (:qq¸). For this aqueduct F. gives
no date, although one can be determined from the building of the naumachia
(see below).
providentissimum principem As an imperial attribute (cf. 6¡.:n.),
providentia conveys the sense of beneficial concern (and expenditure) for vast
public projects: B´ eranger (:q¸¸) .:o–:¸, J.-P. Martin (:q8.) .¡8–¸:. Baldwin
(:qq¡) ¡q¸ exaggerates somewhat when he suggests on the basis of this passage
that F. is deliberately downplaying the first princeps, but see notes below (nisi
forte, proprio opere).
quae vocatur Augusta Cf. ¡.¸ Alsietina quae eadem vocatur Augusta. AQVA
ALSIETINA has been restored in a fragment of the Severan marble plan
(which depicts its arches): Rodr´ıguez Almeida (:q8: ) tav.¸o; Taylor (.ooo) :q¸
fig..o. The name Augusta occurs in an Augustan inscription found in :88¸
(Appendix B, no. :o). F. may prefer Alsietina (which he uses consistently) to
avoid confusion with the ramus Augustae of Appia (¸.6) or the fons Augustae of
Marcia (:..:, :¡.¸).
nullius gratiae . . . parum salubrem A result of its being surface
rather than spring-fed water. The judgement no doubt applies primarily to its
quality for drinking (see Introd. .¸).
ideoque A favourite conjunction in this work: :¸.., ¸..¸, ¸..6, ::..6,
::q.¸, :.:...
nusquam in usus populi Cf. §. solet tamen . . . in subsidium publicorum
salientium dari.
nisi forte F.’s hesitation is plainly rhetorical. He seems to express regret
that Augustus’ expenditure could not have been made on water of decent
quality. Malissard (:qq¡) .¸¸ notes that Tacitus would not have refrained from
irony. From the rhetorical question Taylor (.ooo) :¸q n.:¡ opines that F. prob-
ably considered the naumachia ‘as much of a planning blunder as a waste of
naumachiae Cf. ...¡ Alsietinae ductus post naumachiam, cuius causa videtur
esse factus, finitur. See Mon. Anc. .¸.: navalis proeli spectaculum populo dedi trans
Tiberim, in quo nunc nemus est Caesarum (cf. Suet. Aug. ¡¸.:; Dio, r\.:o.¸). The
artificial lake was :,8oo feet long, :,.oo broad, and the grand performance was
part of the dedication of the temple of Mars Ultor in . ncr (Vell. Pat. ii.:oo.:;
NTD .¡¸–6, LTUR iii: ¸¸¸). For other aspects of the event, see Syme (:q¸¡)
:¸. The nemus Caesarum, a park surrounding the naumachia Augusti, was laid out
by Augustus in honour of Lucius and Gaius Caesar (Lucius died in ., Gaius
in ¡ cr). The site was across the Tiber from the tip of the Aventine, below the
Janiculum. RE :6..: :q¸:, P–A ¸¸¸–8, NTD .6¸, LTUR iii: ¸¡o (Papi); Taylor
(:qq¸), (.ooo) :8:–q¸.
proprio opere ‘in an entirely separate channel’; cf. <rivo> suo of Tepula
8... The point is somewhat otiose. Perhaps F. will suggest an excessive outlay
for water of limited usefulness.
quod . . . coeperat superesse The pluperfect may indicate (:) that
the naumachia was filled all the time and therefore drained at least a little more
slowly than its inflow; (.) that the naumachia was out of commission by F.’s day.
hortis adiacentibus To be taken closely with et privatorum usibus, and
not as referring to the nemus Caesarum. Horti is best translated ‘properties’ (or
estates): see Shackleton Bailey’s note on Cic. QFr. iii.:.:¡. For the location of
these horti see 8¸n. extra urbem. The forma Mentis of an Augustan inscription
(Appendix B, no.:o) seems to have been a special channel serving exactly the
kind of private interests to which F. refers. The stone was found near Galera
(probably the Careiae of ¸:.:). See Taylor (.ooo) 6¸–¸¸.
rr.z in Transtiberina regione Cf. :8.8n., .o.¸. The phrase is appar-
ently synonymous with the Augustan Regio xi\ (Trans Tiberim), but see 8¸n.
extra urbem.
pontes reficiuntur . . . aquae cessant No aqueduct besides Alsietina
led directly to the right bank until the Aqua Traiana was built in :oq cr. The
plural (aquae) refers to more than a single supply piped across the Tiber bridges,
on which see in general Taylor (.ooo) :¸:–68. If one can trust the transmitted
numbers in chapters ¸q–86, Regio xiiii (Trans Tiberim) received distributions
fromall but Tepula and Julia: see Table :o. The Pons Aemilius was the obvious
crossing for the Appia: Evans (:qq¡) ¸:, Taylor (.ooo) :¸¸. The same served
perhaps also for Anio Vetus and Marcia (Evans (:qq¡) :o¸), though Taylor
COMMENTARY ::.¸–:..:
(.ooo) :¸q–¡6 plausibly argues for their crossing by way of bridges to the
Tiber island. Virgo was carried on the Pons Agrippae: Lloyd (:q¸q), Evans
(:qq¡) :o6, Taylor (.ooo) :¡6–q. Mucci (:q86a) opts for the Pons Aemilius to
transport supplies from Claudia and Anio Novus, but water from these higher
aqueducts could have been carried on any of the Tiber crossings.
in subsidium publicorum salientium dari This use of dare
‘distribute’ (sometimes through interchanges) is more general than that F. uses
to connote a formal grant (¸..n. detur): .:.. in specum Octavianum, 6q.¸ in Tepulam,
8¸.¸ regionibus; note especially 6¸.¸, ¸8.: in adiutorium.
rr.j ex lacu Alsietino So called from ancient Alsium (modern Palo).
This is the modern Lake Martignano. The aqueduct received additional water
from Lacus Sabatinus, modern Bracciano (¸:.:n.).
miliario quinto decimo C’s variant is original, not (as B¨ ucheler
thought) by a posterior manus. The Renaissance copyist ignored it, as have recent
editors. But Ashby (:q¸¸) :8¡ points out that there is no visible deverticulum at
the fourteenth mile, whereas just beyond the fifteenth mile an important road
goes northwards, the distance to the lake being about ¸ :/. miles. I therefore
readily accept fifteen, but I am not quite bold enough to write quinque milium
quingentorum. The transmitted number may indeed be wrong; F.’s deverticulum,
on the other hand, might have been as much as a mile longer (cf. ¸.6n.).
rr.q longitudinem F.’s figure shows that the aqueduct must have run in
a fairly direct line to the City, but its precise course has not been traced. Ashort
section was discovered in :q.¸ on the Janiculum (see :8.8n.). The form duorum
is unusual (elsewhere duum: 8.., q.¸, :o.¸); if this is not merely scribal accident,
it may be related to the omission of a figure for the underground channel. I
have supplied the number .:,8:¡ for consistency, although I am well aware
that F. might simply not have cared enough about this aqueduct to include it.
rz.r Idem Augustus Again, F. omits a date, but the addition of the new
spring was no doubt part of Augustus’ general renovation of the water supply
between :: (the date of the S.C. quoted in chapter :.¸) and ¡ ncr (the inscrip-
tion on Porta S. Lorenzo: Appendix B, no.:). In his Res Gestae (Mon. Anc. .o..),
Augustus singles out for special mention the supplement to Marcia: aquam quae
Marcia appellatur duplicavi, fonte novo in rivum eius immisso. The Marcia was Rome’s
finest water (q:.¸), but there is probably an additional sense of family pride
in this accomplishment: Julius Caesar boasted descent from the Marcii Reges
(Suet. Jul. 6.:), and L. Marcius Philippus (cos. ¸6 ncr; cf. ¸.6n.) was married
to the mother of the princeps (Suet. Aug. 8.¸).
quotiens siccitates egerent auxilio ‘Whenever droughts required a
reinforcement’, i.e. from another source. If Marcia’s supply was adequate, the
fons Augustae was diverted to supplement Claudia (:¡.¸n.; cf. ¸..8). Siccitates as
COMMENTARY :...–:¸.:
personified plural recurs at ¸¡.¸ metu aestatis aut siccitatum. For auxilium in other
contexts see .¸.:n., ::¸..n.
rz.z nascitur The mot juste for springs and rivers (OLD s.v. ¸a): note espe-
cially Vitr. \iii.¸.: fontes dulcis aquae nascuntur; Pliny, HN xxxi.¸¸ nascuntur fontes;
Pliny, Ep. \.6.¡o fons nascitur simulque subductur; Dig. xriii..o.:.8 (cited at ¸.¸n.).
rj.r publicis usibus et privatis voluptatibus Cf. .¸.: publicis pri-
vatisque non solum usibus et auxiliis verum etiam voluptatibus, 8q.: <usibus> ac
voluptatibus nostris [i.e. civium]; cf. also q¸.¸, the contrast in Republican times
between communium utilitatium and privatarum voluptatium. F.’s use of voluptas rep-
resents something of a change from, e.g., Cic. Mur. ¸6 odit populus Romanus
privatam luxuriam, publicam magnificentiam diligit. No longer does the word con-
vey so much a judgement on morals as a comment on the higher standard
of living the state is prepared to provide for the denizens of the metropo-
lis. Among the list of memories which constitute urbis voluptas for Ovid are
stagnaque et euripi Virgineusque liquor (Pont. i.8.¸8–¸q). Imperial times were spe-
cially beneficial for those privati who might reasonably expect a grant to draw
public water (q8..n., :o¸.:n.). Cf. Strabo \.¸.8 (something of an exaggera-
tion) :coc0:cv o to:i :c tioc,c,iucv 0ocp oic :cv oopc,c,ticv co:t
tc:cucu, oic :n, tcìtc, sci :cv otcvcucv ptïv, ctcocv ot cisicv oytocv
otçcutvc, sci oigcvc, sci spcuvcu, tytiv ógûcvcu,, cv tìtio:nv ttiutìticv
ttcinoc:c Mópsc, A,pittc,, tcììcï, sci cììci, óvcûnucoi scounoc, :nv
tcìiv. For voluptas as an element in the Roman bathing habit, see Fagan (:qqq)
altero . . . consulibus Caligula’s second year began on :6 March ¸8
cr. For the consuls of that year see Degrassi ::.
anno urbis conditae The AUCdate should be ¸8q for consistency with
F.’s practice elsewhere (¡.:n.), corresponding to the Varronian ¸q:. The trans-
mitted ¸qo probably results froman error in copying numerals (DCCLXXXX
for DCCLXXXIX) or in writing them out. C’s pattern for ordinals follows
that of cardinals (6.6n. septuaginta novem); cf. ¸.6, :¡.:, :¸.¡ tricesimum octavum,
q.: nono decimo, so one might be tempted to write octogesimo nono or non<o et
oct>agesimo: K–S i: 6¡¸–¡.
duos ductus incohavit Cf. Pliny, HN xxx\i.:.. vicit antecedentes aquarum
ductus novissimum impendium operis incohati a C. Caesare et peracti a Claudio; Suet.
Gaius .: incohavit autem aquaeductum regione Tiburti, Claud. .o.: ductum aquarum a
Gaio incohatum. How much progress was made under Gaius is unclear, and it
might even be questioned whether his initial project had been to build two
distinct aqueducts. Claudia and Anio Novus were closely associated (¸..6,
86.:, q:.¸) and they must always have been popularly considered to be in
essence a single project. Amongst the countless possible reasons for Gaius’
work being unfinished, there might have been disorganisation and difficulties
with redemptores – for these must have been used in the stages of both planning
and building (::6.¡n., ::q.¸n.).
rj.z quod opus Claudius . . . consummavit Speaking of Claudius’
censorship in ¡¸, Tacitus, Ann. xi.:¸.. writes fontisque aquarum Simbruinis collibus
deductos urbi intulit. Against the common view that Claudius actively took up
Gaius’ unfinished project in ¡¸, Furneaux (:qo¸) ii: :8–:q suggests that Claudia
was in some way brought to Rome in ¡¸ (the dedication in ¸. then marking
completion of both Claudia and Anio Novus). Ashby (:q¸¸) :q:–. notes that
fontes aquarumwould appropriately describe Claudia. In any case, interest in the
water supply in ¡¸ was more than coincidental on the part of the traditionalist
who restored the censorship (Suet. Claud. :6.:).
magnificentissime Cf. Suet. Claud. .o.: lapideo opere in urbem perduxit
divisitque in plurimos et ornatissimos lacus. The magnificence of Claudius’ work is
still to be seen in the remnants of the arcade on which the channels of Claudia
and Anio Novus were borne from the seventh milestone to their terminus in
the City (.o.:–.). The series originally numbered no fewer than a thousand
arches and was remarkable for its height (:¸.¸n.) as well as its length. As the
Claudian arches commanded visual admiration, they also invited comparison
with those of the Marcia (upon which the channels of Tepula and Julia were
also carried: :q.¸–¡). Claudius’ engineers could more easily (and more cheaply)
have built with concrete; the arches and the specus of Claudia were built of
ashlar, a conservative style befitting their magnitude. The specus of Anio Novus
was, of course, superimposed upon that of Claudia for reasons of engineering,
but that it was built of concrete is hardly evidence for dating its construction
to a separate, somewhat later, stage: so, tentatively, Ashby (:q¸¸) :q:. The
exterior facing (brick below, opus reticulatum above) has a quality of lightness
which combines with the massive blocks to lend aesthetic appeal. A similar
type of construction had been employed for the channels of Tepula and Julia,
and these might have been the immediate model for the Claudian architects.
Pliny, HN xxx\i.:.. records the sum spent (erogatum in id opus), but the figure
is textually uncertain: ¸¸o,ooo,ooo sest. seems more likely than ¸¸,ooo,ooo.
Despite the enormous expense, the quality of the work (cf. :.o) was apparently
not what it might have been. Claudia allegedly failed after only ten years,
and extensive repairs were proclaimed to have been made at least three times
before F. wrote (Appendix B, nos. ¡, ¸, ::). Expenditure on portions other than
the final arcade might have been considerably less lavish, but the inscriptions
on Porta Maggiore are above all proclamations of the Flavian dynasty.
consummavit dedicavitque Note the identical collocation in CIL
¸.¸¡q and ¸.:¡:q¸.¡¸ (cf. also q.6o¸8). For consummare see 6..n. In Repub-
lican times dedicare was applied almost without exception to objects of religious
importance (TLL s.v., RE ¡: .¸¸6, D–S s.v. dedicatio). With F.’s ‘secular’ use
cf. Asc. Pis. : theatrum dedicavit, Tac. Ann. xi\.¡¸.. gymnasium dedicatum, Pliny,
Ep. i.8.. bibliothecam dedicaturus, x.::6.: opus publicum dedicant. By F.’s day, the
verb dedicare seems to have come to mean formal presentation to the public of
something undertaken by private initiative and with private funds. Observe,
in this instance, that Claudius proclaims the use of his personal resources in
the formula sua impensa (see :.¸n.).
<Fausto> Sulla <Salvio> Othone Degrassi :¡. I base this drastic
change on Tac. Ann. xii.¸..:. For consular dating elsewhere in chapters ¸–
:¸ F. consistently gives two names and omits the conjunction; the traditional
restoration Sulla et <Ti>tian<o> is thus doubly awkward. Salvio could easily
have fallen out because of its similarity to Sulla. For C’s et tian I offer a mechan-
ical explanation: o misread as e, the first stroke of h seen as long i, the second
stroke and o taken together as open a (cf. ¸.¸ collegio] collega C).
anno ¸. cr. For the error which underlies the transmitted octingentesimo
sexto (DCCCVI for DCCCIII: 8o¸ in F.’s scheme, 8o¸ Varronian) see :¸.:n.
Closely related to the dedication of the new aqueducts is the spectacular arch
where the Viae Labicana and Praenestina diverge (Porta Praenestina in the
Aurelian Wall, known since the tenth century as Porta Maggiore: P–A ¡:.–
:¸, Nash ii: ..¸–8, NTD ¸o6–¸, LTUR iii: ¸o¡, ¸:o–::, Pisani Sartorio). The
inscription thereupon (Appendix B, no.¸) can be dated (by Claudius’ trib. pot.
xii) to the periodbetween.¸ January ¸. and.¡ January ¸¸; its words perducendas
curavit relate to Claudius’ role as censor (see above), and the inscription as a
whole leaves no doubt that idemque probavit is to be understood at its end.
Kalendis Augustis Claudius’ sixty-first birthday (Suet. Claud. ..:); cf.
Caetani Lovatelli (:8¸8) :6q–¸o. There may be significance to an event in the
year ¸. (five years after Claudius’ censorship), for Republican censorships had
been a quinquennial magistracy.
rj.j alteri nomen . . . Claudiae datum The inscription on Porta
Maggiore likewise gives pride of place to the Claudia – perhaps for the superior
quality of its water (cf. :¸.:–¸, qo.:–.), but more probably because it bore the
emperor’s name. The appellation Claudia cannot have failed to recall the
ancestor Appius Claudius who had built Rome’s first aqueduct. F. could as
easily have put nomen directly before Claudiae (cf. q..); perhaps he aimed for
greater stylistic variety (alteri nomen . . . datum; altera . . . vocitari coepit).
quae . . . perducebatur Cf. Suet. Claud. .o.: Claudiae aquae gelidos et
uberes fontes, quorum alteri Caeruleo, alteri Curtio et Albudino nomen est. B¨ ucheler’s
<aquae> (after nomen) seems unnecessary. His deletion of quae . . . perducebatur
is superficially attractive (the information is repeated in :¡.: just below), but
the ex fontibus phrase seems to be almost a part of the official name. For the
appositional formcf. ¸..¸ Curtium et Caerulum fontes, Pliny, HN xxx\i.:.. Curtius
atque Caeruleus fontes.
COMMENTARY :¸.¡–:¡..
rj.q {haec . . . Marciae} Marcia andClaudia were fedby neighbouring
springs (:¡.:–¸, :8.¡), and both were highly prized for their quality (8q.¡,
q:.¸–¸), but this is not the place where Claudia’s bonitas should be mentioned,
let alone compared to that of Marcia. Awkward syntax smacks of marginal
annotation (cf. :¡.. tantae bonitatis ut Marciae . . .).
rj.j Anio novus Cf. Pliny, HN xxx\i.:.. et Anien novus; Suet. Claud. .o.:
simulque rivum Anienis novi (both of whom regard the aqueduct almost as part
of Claudia). Statius’ Anio (Silv. i.¸..¸) probably refers to the Novus. Known
cippi for Anio Novus are uninscribed. Anienis N(ovae) appears in an inscription
recording repairs made by Gratian and Valentinian II in ¸8: (CIL 6.¸86¸ =
{alia omnes praecedit} The words are hopelessly intrusive (of which
perhaps alia is a sign: cf. ¡¡.., ¡¸). Like haec . . . Marciae (§¡), they might have
originated as a marginal note – on the level of the aqueduct (cf. :8.¡) or its
copia (cf. ¸..:).
cognomen veteris adiectum See 6.:n. qui nunc vetus dicitur.
rq.r via Sublacensi . . . trecentos The springs of Claudia were only
about :oo paces from those of Marcia (cf. :8.¡, 8q.¡). There is some consensus
that they can be identified with the modern i and ii Serena: Panimolle (:q68)
ex fontibus duobus The two principal springs, originally tapped for
Claudia: :¸.¸n. For the quantity see ¸..:–..
Caerulo The o is a scribal correctionfrome, whichmight prompt Caeruleo,
a formoffered unanimously in other sources (cited §¸n.). But there is no similar
hint there (cerulo) or at ¸..¸ (cerolum), so it is safer to assume that F. used Caeru-
lus. The origin of the name is obvious: for characteristic colour, cf. ¸.¸ colore
Curtio F. might not have known the origin of the name. The gens Curtia
comes to mind, as does a superficial similarity with Lacus Curtius, but there
is no good reason to draw a ‘Sabine’ connexion (Grimal :. n.¡).
rq.z Albudinus Suetonius (Claud. .o) links Albudinus with Curtius (alteri
Curtio et Albudino nomen est). The third spring was probably added at some
later point, perhaps during the repairs recorded by Titus; to it may refer
his reconstructions a capite and the nova forma (Appendix B, no.¸). The name
is otherwise unknown. It need not indicate that the water had ‘une teinte
blanchˆ atre’ (Grimal :. n.¸), since the root may be a pre-Indo-European word
meaning ‘mountain’: Bertoldi (:q¸6); cf. Albula, old name for the Tiber (Varro,
Men. ¡:¸, Virg. Aen. \iii.¸¸., Ovid, Fast. i\.68); Albulae, springs near Tibur
(Vitr. \iii.¸.., Pliny, HN xxxi.:o); Albunea, Sibyl associated with the Anio at
Tibur (Virg. Aen. \ii.8¸, Horace, C. i.¸.:.).
COMMENTARY :¡.¸–:¡.¡
Marciae quoque adiutorio Alongside the verb adiuvare (:o.6, :¡.¸),
F. uses the post-Augustan noun adiutorium ( = auxilium): 6¸.¸, ¸8.: (dari) in
adiutorium + genitive; cf. Str. ii.¸.:: tamquam in adiutorium exierunt.
quotiens opus est Albudinus normally supplied Claudia, but it could
be diverted to Marcia, just as Augusta (§¸n.) could be diverted from Marcia to
Claudia. For the surviving remnants of what was apparently a complex system
of feeders and interchanges, see Ashby (:q¸¸) q6–8.
adiectione sui C’s adlectiones sex reveals two features of earlier transmis-
sion: (:) long i had been used in minuscule script, and (.) numerals had been
written as symbols rather than in words. Whether these were features of C’s
immediate exemplar cannot be known, although it would be easy to credit
Peter the Deacon with the latter change (Introd. ¡¸–¡¡).
rq.j Augustae fons F. makes it clear that Augusta was to feed Claudia
only if its water was not needed for Marcia (cf. :..:, ¸..8). This shows pious
respect for the intentions of its inventor.
quia . . . adparebat See 6¸.q, where F. reports an overflow of ¸oo
quinariae at Marcia’s source.
ita demum Cf. ::..¸ ita demumenim. Ita demumas stipulative inconsecutive
clauses is post-Augustan, frequent in Pliny and especially in legal texts.
manente . . . praesidiario ctcç as a substantive (OLD). F. uses the
noun praesidium ( = auxilium) at 8q.: parum praesidii and ::¸.¸ praesidium aquarum
rq.q longitudinem Poleni’s restorations are self-evident. Homoioteleu-
ton in numerals (xxxx\i, cccc\i) could easily account for the omission of
quadringentorum sex. F.’s total (¡6,¡o6 as emended) does not agree with the a mil-
liario XXXXV of the Claudian inscription (Appendix B, no.¸), nor with Pliny’s
a XXXX lapide (HN xxx\i.:..: this could be an approximation, or the text may
be in error). Albertini (:qo6) ¸o¸–:: supposes that ¡¸,ooo was the distance to
Claudia’s principal springs and that F. includes :,¡o6 paces of an additional
channel for the Albudinus. Yet F. nowhere mentions that Albudinus had a no-
ticeably long channel of its own (contrast :o.8, :..¸), and his remarks suggest
rather that this spring was located quite close to Claudia’s other fontes (and
to that of Marcia: see :¡..n.). The difference between F. (as emended) and
the inscription (obviously rounded to an even mile) is less than ¸ per cent: see
Blackman (:q¸q) :¸–:¡. Nor do we know to what extent Flavian repairs and
reconstructions (e.g. those noted in Appendix B, nos. ¡, ¸, ::) may have altered
the length of the original channel.
in superiori parte Claudia and Anio Novus (:¸.6) followed a more
direct route from the Anio valley (cf. 6.6, ¸.8) and thus had more construction
above ground in their upper stretches (note also :8.¡–6).
COMMENTARY :¸.:–:¸..
propius urbem The comparative occurs in the same formula at ¸.8
(contrastedwithlongius ab urbe) andat :¸.6(note alsoq.¸ proximis urbi locis a septimo
miliario). Cf. Tab. Heracl. (RS .¡, Bruns :¸, FIRA :¸), line .o que viae in urbem
Rom(am) propiusve u(rbem) R(omam) p(assus) m(ille) ubei continente habitabitur sunt erunt.
We find prope urbem at 6q.¸ and ¸o.¸, where there is no such contrast. I emend
C’s prope here, supposing it could have come from propius (-i +abbreviation for
-us misread as -e).
a septimo miliario Substructures (6oq) and arches (6,¡q:) total ¸,:oo
paces, while those for Marcia total exactly ¸,ooo (¸.8, q.¸). The line of the
Claudian aqueducts was straighter than that taken by Marcia’s builders, and
one would therefore expect the total to be less. (Because ¸,:oo paces is the figure
given for Anio Novus as well (:¸.6), it cannot include the branch – Claudia
only – carried on the Neronian arches.) The starting-point can only have been
at the respective piscinae (.o.:), and F. locates the terminus post hortos Pallantianos
(.o..). Measurements of the surviving remains (some ¸oo m substructure,
approximately 8,86¸ m arcade) suggest that the transmitted figures (Table :)
are excessive for both types of construction: Ashby (:q¸¸) ..¡–¸; cf. Blackman
(:q¸q) :¸. The text may have undergone some arithmetical tampering, but to
discredit it completely is hardly wise.
rj.r viaSublacensi . . . inSimbruino F. here describes the AnioNovus
as it was when he took office, before the project to move its intake upstream
(q¸.:–¡nn.). With in Simbruino cf. Tac. Ann. xi.:¸.. Simbruinis collibus. It should
come as no surprise that there are apparently no remains of the original intake
or of the settling tank F. describes.
ex flumine . . . turbida pervenit F. gives further details at qo.:–.,
contrasting the current circumstances with projected improvements (q¸.:–¡).
Pliny, Ep. \iii.:¸.¸ describes an inundation of the Anio which took place a few
years later.
terras cultas . . . soli pinguis . . . ripas solutiores For soil types
and their descriptors in Latin see White (:q¸o) qo, q¸–:o¸. For the adjective
solutior appliedto soil, cf. Colum. ii...q humus, \.¸.: terra (contrastedwithdensior),
Pliny, HN x\iii.¡¸, Sen. Ep. qo..:.
rj.z piscina limaria The adjective is a ctcç (TLL). Cf. :q.:n. (piscinae
near the seventh mile) ubi limum deponunt; Vitr. \iii.6.:¸ limus enim cum (aqua)
habuerit, quo subsidat, limpidior fiet.
specum Aspecus, originally ‘tunnel’, differs fromrivus in that the former is
restricted to a covered (usually underground) channel: cf. Vitr. \iii.:.6, 6.¸. F.’s
use here is consistent with that sense, as at :¸.¸, .:.. specus Octavianus (contrast
§¡ rivus Herculaneus), q:.¸, and presumably :.¡.¡n. He far and away prefers rivus,
the more versatile word (¸.:n.). Legal texts juxtapose both words: :.¸, :.¸.:,
:.q.¡–::; Venafrum edict (CIL :o.¡8¡. = ILS ¸¸¡¸), line q. For a lawyer’s
distinction: Dig. (Ulpianus) rivus est locus per longitudinem depressus,
quo aqua decurrat, cui nomen est ótc :c0 ptïv, specus autem est locus, ex quo despicitur:
inde spectacula sunt dicta.
rj.q rivus Herculaneus Ashby (:q¸¸) .¸¸–8 identifies this source with
the slightly acidulous springs near Mola Nuova. The origin of the name is
unclear (perhaps a local cult of Etruscan origin), but this rivus is in any case
not to be confused with a homonymous branch of Marcia (:q.8) nor with that
which Pliny associates with the source of Virgo (:o.¸n.).
rj.6 passus . . . septingentos As for Claudia (:¡.¡n.), F.’s figure dif-
fers from the a milliario LXII of the Claudian inscription (Appendix B, no.¸),
although in this case he gives a lower rather than a higher figure and the
difference (¸,¸oo paces) is greater. It is clear that his data are for the length
of the conduit before extension (§:n.; cf. q¸.:). Implausible in the extreme is
the theory of Albertini (:qo6) that the inscription underwent a ‘correction’
after F. wrote, pace Gordon (:q8¸) ::8–:q. For an alternative explanation of the
divergent figures, see Rodgers (:q86a).
rj.¡ arcus altissimi F. uses arcus for the physical objects as opposed to
the construction type (see ¸.:n. opere arcuato and cf. :8.¸n. arcuationibus) also at
.o.:, ¸, .... (fornices is the word in Augustan legislation :.¸.:n.). Anio Vetus
was highest in level, therefore in potential for urban delivery: :8.¡, .o.:, q:..
editissimus. Plainly height is a consideration in upkeep (see :....). The height
of the arches varied, of course, with the terrain. F.’s :oq pedes is equivalent
to ¸...6 m; the arches north of the Cassino–Naples railway, estimated by
Lanciani (:88:) ¸¸q at .¸.¡: m, are the highest which survive.
r6–r¡ F.’s commentary deriving from the data presented in chapters ¸–:¸.
Note that some such remarks have been inserted into the context already
r6 Tot aquarum . . . Graecorum A proud statement, surprising for
its exuberance, but which seems to follow naturally after arcus altissimi. Given
their practical value, it is easy to overlook the fact that these aqueducts were
architectural feats for which the Roman world offers few if any parallels.
Strabo, \.¸.8 praises the aqueducts of Rome along with other works of prac-
tical engineering: co:ci [Romans] tpcovcnocv uóìio:c cv cìi,cpnocv
tstïvci [Greeks], o:pcotc, cocv sci ooó:cv tioc,c,n, sci otcvcucv
:cv ouvcutvcv tssìúçtiv :c ìúuc:c :n, tcìtc, ti, :cv Tiµtpiv. Pliny, HN
xxx\i.:.:–¸ includes them among the urbis nostrae miracula, beginning em-
phatically sed dicantur vera aestimatione invicta miracula, and ending grandly quod
si quis diligentius aestumaverit abundantiam aquarum in publico, balineis, piscinis, euripis,
domibus, hortis, suburbanis villis, spatia aquae venientis, exstructos arcus, montes perfossos,
convalles aequatos, fatebitur nil magis mirandum in toto orbe terrarum. F., like Pliny, con-
trasts the wonders of Roman practicality with the useless monuments which
had excited the admiration of centuries. His specific references to pyramids
can be nothing other than a deliberate echo of Pliny, HN xxx\i.¸¸ dicantur obiter
et pyramides in eadem Aegypto, regum pecuniae otiosa ac stulta ostentatio. Among F.’s con-
temporaries, note Pliny, Ep. x.¸¸.¸ (to Trajan, on an aqueduct at Nicomedia)
adfirmo utilitatem operis et pulchritudinem saeculo tuo esse dignissimum; Tac. Ann. ii.6:.:
(on pyramids) eductae certamine et opibus regum. For the tone of this passage and
its literary allusions: Baldwin (:qq¡), ¡qo–:, DeLaine (:qq¸), :.¡–¸, Geißler
(:qq8), .6q.
tot . . . tammultis . . . molibus Note alliteration. Apparent pleonasm
prompted Bergk (:8¸¸) to delete tam multis. But F.’s point is to emphasise both
the number of aqueducts comprising the urban water supply (aquae) and the
number and size of the arches (molibus virtually = arcubus). One need think
only of the five high-level channels running on two arcades from the seventh
milestone: the parade of Claudia-Anio Novus alone originally numbered over
a thousand arches. For similar expressions cf. Cic. Tusc. ii.¡6 quod a tam multis
et quod tot locis perferatur; Livy, xxi\..6.:¸ cum tot ac tam validae eluctandae manus
essent, xx\iii.¡..:q adversus unum tot proeliis et tam diurna ac gravi militia fessum; Val.
Max. \ii.:.: tot partus, tot incunabula, tot viriles togae, tam multae nuptiales faces; Quint.
Inst. xii.:.¸ tam occupatum, tam multiforme, tot ac tam variis adfectibus concisum atque
necessariis molibus pyramidas . . . otiosas Chiasmus. Roman
poets compared their literary opera to the pyramids: Horace, C. iii.¸o.. regalique
situ pyramidum altius, Prop. iii...:q.
compares Unwontedapostrophe contributes innosmall part tothe force
of this sentence, especially since the verb is then followed by another, more
comprehensive, direct object preceded by an alliterative string of modifiers.
opera Graecorum The noun and its attributive genitive are the more
emphatic for delay. Cf. Curt. \.:.¸. (gardens of Babylon) vulgatum Graecorum fab-
ulis miraculum (note also Curt. iii.:.. Marsyas amnis, fabulosis Graecorum carminibus
inclitus). F. has in mind the traditional ‘wonders’ of the ancient world (cf. RE
Suppl. :o, s.v. ‘Weltwunder’), cataloguedfor example by Martial, Spect. :, begin-
ningBarbara pyramidumsileat miracula Memphis; cf. Mart. \iii.¸6.: Regia pyramidum,
Caesar, miracula ride.
r¡.r Non alienummihi visumest The expression recurs at .¸.:n.; cf.
:o..: non est alienum, ::..: non ab re est. For examples of alienum in transitional
use: (with videtur) Caes. BGall. \i.::.:; Nep. Milt. 6.:; Vitr. i\...:,¡, x.:¸..;
Cels.; Pliny, HN iii.6¸, :¸6, xxxii.:¡.; Quint. xii.q.:¡; (with est) Cic. Inv.
i.¸, ii.6¸; Fin. iii.¸:; Vitr. \¸, ¸.¸, :..:, x...::; Colum. xii.:8.:; Quint.
i.:.¸¸, \iii.:..; Suet. Jul. ¡¡.¡, Rhet. .¸.q; (arbitror) Cic. Prov. ¡o; (existimo) Pliny,
Ep. \ii.:..¡; (puto) Cic. Orator .¡; Sen. Suas. \i.:¡; Colum. iii.¸.:; (iudico) Scrib.
longitudines . . . per species operum Not just the total lengths,
but also those of individual sections according to their location and type of
construction (cf. ¸.:). Longitudines rivorum (cf. .¸.:) refers to overall lengths:
note the formula ductus eius habet longitudinem (¸.¸, 6.6, ¸.8, etc.). Per species
operum ‘according to categories or types of construction’ refers to F.’s figures
for rivus subterraneus and opus supra terram, the latter further distinguished as
substructio and opus arcuatum. The quoque has puzzled editors, perhaps because
of an apparent pleonasm in etiam. But quoque seems to emphasise longitudines
(‘the lengths indeed’, as well as other data: cf. §¸ formas quoque ductuum), with
the following etiam almost = non solum. I admit that I have no good parallel
for this use, and Heinrich’s deletion may be a better solution.
Since F. does not indicate that the measurements are his own (cf. 6¡.:–¡), it
is likely that he has taken the information from records at his disposal (Introd.
.o). That for some aqueducts there were such systematic data (at least for
overall lengths, if not necessarily broken down per species operum) can safely be
assumed from the evidence provided by the Augustan cippi which marked the
zone legally reserved as a clearway for the aqueducts (:.¸.:n.). There exist,
it may be remarked, no known cippi for Appia; those for Claudia and Anio
Novus are uninscribed; the series for Virgo postdated the Augustan project
(cf. :o.¸n.). To Fabretti (:68o) Book ii, p.:::–.o goes the credit for being first
not only to have called attention to the spacing of the cippi as an aid in tracing
the aqueducts’ courses but also to suggest that from their data overall lengths
could be determined. Lanciani (:88:); ¸¸q–6o, Ashby (:q¸¸) ¸8, q¡; Blackman
(:q¸q) :¡, and Roncaioli Lamberti (:q88) find them to be a useful adjunct in
calculating the overall lengths. Mari (:qq:a) :6¸–¸: demonstrates how totally
unreliable they are for this purpose.
These cippi (examples in Appendix B, no.¸) bore a sequential numeral and
indicatedthe distance fromthe last preceding marker. Inmost cases the interval
was .¡o pedes, a standard unit of land measurement (. actus = ¸: m); frequent
anomalies are readily explainedby difficulties of terrainor special complexities,
e.g. in areas where buildings or monuments existed. (Because they were set
up subsequent to the initial construction, their positioning has no relationship
whatsoever to the location of putei (8q.¡n.) in preexisting tunnels, although Vitr.
\iii.6.¸ recommends spacing such shafts inter duos actus.) Froman administrative
viewpoint, it was probably not at all incidental that the precise surveys required
to define the clearway also provided an accurate measurement of channel
lengths. It is open to question, of course, whether the figures upon which F.
was drawing hadincorporatedchanges introducedsubsequent to the Augustan
period. F. never mentions these cippi. For his purposes they were irrelevant, for
COMMENTARY :¸..–:¸.¸
directions to the sources were better given, as he does, by reference to highways
and side-roads, and his attention to the clearway (:.6–q) is theoretical rather
than specific to any individual stretch that might be marked with cippi: see
Hodge (:qq.) :o¸–¡.
r¡.z maxima huius officii pars Cf. 6¡.¸ praecipuum officii opus. On the
curator’s overall responsibility for tutela ductuum see ::q–.¡ (especially ::q.:,
where it is called res enixiore cura digna). The praepositus in this case is the curator
himself, whose personal involvement is an important feature of F.’s attitude
towards his office (cf. ..:).
r¡.j nostrae quidem sollicitudini non sufficit An impersonal use
of sufficit (cf. K–S i: 6¸¸) carries the force of generalisation, with which the
transmittedpresent +perfect infinitive, ‘it does not suffice tohave inspected’ (cf.
§. scire oportet), accords better than Giocondo’s perfect; contrast facere curavimus,
applied to a specific action, and cf. 8q.: nec hoc diligentiae principis . . . sufficit
parum . . . contulisse sese credentis, quod . . . adiecit. . . . Nostrae sollicitudini = mihi:
F. is fond of abstract for concrete (cf. H–Sz ¸¡¸–8), though more often in the
nominative (e.g. :o:.¡ fides nostra, ::8.¸n. nostra sedulitas).
singula oculis subiecisse The singula are the physical details of the
aqueduct system (especially the species operum). The tense of the infinitive sug-
gests ‘on (just) one occasion’. F. announces that he had assumed an active
personal role early in office, a grand tour not wholly unlike Agrippa’s personal
inspection of the Roman sewers (Dio, xrix.¡¸.:). Maps will be a safeguard
against faulty memory on F.’s own part as well as ignorance and negligence
on that of his successors (cf. ..¸).
formas Context makes plain that these maps (or diagrams: OLD s.v. :¸)
refer to outlying stretches of the aqueducts. Urban portions of the system
could more conveniently be viewed. While there is no reason to discredit F.’s
statement of personal inspection, there is every likelihood that data for this
project were already available – on the earlier aqueducts, for example, as-
sembled perhaps already under Agrippa (q8) or surely when the Augustan
cippi were set up (§:n.); for the Claudian works at the time of their construc-
tion (cf. :o¸..); cf. Grimal xi–xii, Mari (:qq:a) :6¸ n.¡q. For map-making in
general, and especially the precedent of Agrippa, see Dilke (:q8¸), Nicolet
(:qq:) q8–:.:, Evans (:qq¡) ¸8–6: (the latter specially on the influence of
Agrippa’s map of the world in Porticus Vipsania). This use of the noun forma
occurs in the Lex agraria (RS ., Bruns ::, FIRA 8) lines ¸, ¸8, 8o; cf. Craw-
ford (:qq6) :¸8. It is ubiquitous in gromatical contexts, with which F. was not
unfamiliar (Introd. ¸ n.:¸), for which formae in archives had a special legal
standing: Moatti (:qq¸) ¸:–¡8, Cranach (:qq6) :¸8–8:. On what F.’s formae
may have looked like, we ought not to imagine a contoured tableau but
rather a schematic diagram not unlike the Severan marble plan, on which
aqueduct arches appear, e.g., for the Alsietina (::.:n.) and near the Temple
of Divus Claudius (.o.¡n.). Very probably there were at least some numerical
references (to distances, or perhaps to some of the numbered cippi), implied
certainly in F.’s words valles quantaeque: Hodge (:qq.) :¸., ¡¸¡ n.¸; Mari (:qq:a)
ex quibus adpar<er>et ubi . . . exig <er> ent curam ‘that from
these might be seen where there are valleys and of what size, where rivers
are crossed, and where channels contoured along the mountainsides require
greater and constant attention’. Cf. :.:.: fere aut vetustate aut vi <eae> partes
ductuum laborant quae arcuationibus sustinentur aut montium lateribus adplicatae sunt, et
ex arcuationibus eae quae per flumen traiciuntur. If the present adparet be kept, the
tense of traicerentur is puzzling. A final relative clause follows smoothly after
facere curavimus. For the difficulty with C’s exigant see below.
ubi valles . . . ubi flumina . . . ubi montiumlateribus To be sure,
a tricolon crescens with anaphora, but not quite ‘poetry in prose’ (Baldwin
(:qq¡) ¸o¸), because the device is more deictic than decorative and even the
juxtaposition of vales, streams and mountain slopes does not evoke – in this
context – any sense of natural wonders (as perhaps at q¸.:–¡, quite different
in tone).
valles quantaeque Vitr. \iii.6.¸ makes the point that terrain will dictate
construction type: sin autem non longa erit circumitio, circumductionibus, sin autem valles
erunt perpetuae, in declinato loco cursus dirigentur . . .
specus adpliciti C’s adplicite (cf. :.:.: partes . . . adplicatae) cannot stand,
for elsewhere in this text specus (:¸..n.) is masculine: q:.¸ in alienos . . . specus,
:.¡.¡ specus . . . derecti.
maiorem. . . exig<er>ent curam I take this to refer only to the third
ubi-clause; cf. :.:.: (cited above) where we have distinction between arcuationes
(of two sorts) and hillside channels. Professor Reeve, however, has astutely
observed that maiorem . . . curam might perhaps apply to all three ubi-clauses,
not just the third. We could thus turn C’s exigant into the indicative and take
maiorem . . . curam as an explanatory parenthesis (perhaps asyndetic as probably
at ¡..) or more likely preceded by a lacuna. We should then require at least a
verb after C’s adplicite (to avoid a second ellipsis of essent) as well as something
like <hae partes ductuum> (cf. :.:.:).
†petendi ac muniendi vi† I have failed to derive any help from the re-
lated passage at :.:.: (cited above). If muniendi could be right (though it does not
normally mean reinforcing or repairing: ¸.:n.), the notion behind the trans-
mitted petendi ought to be watchfulness and prevention; cf. the Venafrum edict
(CIL :o.¡8¡..:q = ILS ¸¸¡¸.:q) specus reficiundi aut inspiciendi causa. B¨ ucheler’s
tuendi has an obvious appeal, but he does not account for the initial pe-. The
troublesome vi looks more like a vestige of scribal confusion (perhaps for the
minims in muniendi) than the remains of a noun.
COMMENTARY :¸.¡–:8.¸
r¡.q velut in conspectu . . . tamquam adsistentes Cf. Cic. Amic. ¸
ut tamquam a praesentibus coram haberi sermo videretur. Contrasted with oculis subiecisse
(§¸); cf. Curt. iii...:. qui oculis non subiecere animis concipere non possunt. F. avoids
the form veluti, the final letter of which I think might have been a matter of
scribal carelessness here (before in). On the double apology see ...n. velut.
r8–zz F. turns now to altitudinem cuiusque (¸..n.). He lists first the aqueducts
in order of their respective levels at entering the City (:8), for these dictate to
what heights each can deliver its water. Then he turns (:q–..) to more specific
topographical details to indicate what parts of the City are served by each
r8.r diversa . . . libra ‘At a different level’ or height. Libra = altitudo
(cf. ¸..) in contexts which involve perlibratio (6.6n. libramento; cf. §¡ arte librandi):
Poleni ad loc., citing Turnebus (:¸8o) \.¸: ‘Libra interdum altitudinem et
fastigium significat, propter perlibrationem Aquarum.’ F.’s ordo librae (.¸.:)
refers to the respective elevations of the channels at the point where open flow
in the aqueducts gave way to closed-pipe distribution, for if they descended
without closed pipes the water could not be raised up again.
r8.z s<erv>iunt quaedam altioribus locis C’s siunt has consistently
been misread as fiunt, with fluunt an apparently easy correction. Yet fluere (¡.¸n.
confluunt) is better constructedwitha prepositional phrase (::.:, :¸.¸, 8¸.:, :.q.¡;
cf. ¸.: influunt, ¡.¸ confluunt, :::.. effluere) or used absolutely :¸.:, :o¸.¡, :oq.¸).
For servire ‘to supply, furnish’ +dative see ¸.., 8¸.¸, q:.¸, q., and esp. §8 below.
nam et colles An explanation that ‘even the hills’ had gradually been
raised by rubble from fires seems a trifle out of place here. F. may be thinking
of historical changes which affected distribution (cf. §¡).
r8.j quinque sunt The levels of these five aqueducts (§¡) can still be
compared at Porta Maggiore. The gate itself carries Anio Novus and Claudia,
while slightly tothe northeast the channels of Julia-Tepula-Marcia pass through
the Aurelian Wall.
quarumaltitudo . . . adtollitur A curious expression for ‘the altitude
of which makes it possible for their water to be lifted’. Since urban distribution
was effected by gravitational flow, the height to which water could be delivered
depended on the elevations of the terminal castella.
in omnem partem urbis Cf. Pliny, HN xxx\i.:.. (Claudia and Anio
Novus) ad eam excelsitatem ut omnes urbis montes lavarentur.
pressura coguntur Cf. ¸¸.: ex humiliore [sc. loco], id est minore pressura,
6¸.¸ propter pressuram librae; also 6¸.¸ qui [sc. Appiae ductus] sit depressior, ::..¸ aqua
per brevis angustias expressa. Note also Pliny, Ep. x.6:.¡ fossam . . . altius pressam,
a ditch dug (lit. pushed) deeper. Context alone here makes it clear that F.’s
pressura relates to height, and has therefore no connexion whatsoever with
that ‘water pressure’ with which urban dwellers are familiar in the modern
world, where the force of water issuing from, say, a tap, has been in most
cases artificially regulated and no longer depends solely upon the gravitational
head, or the distance that tap lies beneath the highest level from which the
water is drawn. In F.’s language, by contrast, pressura is nothing more or less
than the hydrostatic pressure generated – in a closed-pipe system – by the
distance an aperture may be below the water’s natural surface level. To use an
example fromthis chapter: Anio Novus is highest in level at the point where the
aqueduct reaches the city (§¸n.); assuming that the tank (or terminal castellum:
¸..n.) into which its water flows is filled to a level absolutely higher (in, say,
metres above sea level) than that of any other aqueduct and that all water from
that tank issues in closed pipes, the Anio Novus water theoretically can be
conveyed to any point, no matter how distant, and then raised to any point up
to the level at which it first issued. Anio Vetus, now, arrives at its terminal castellum
at a much lower point (in m.a.s.l.): it cannot be raised by pressura any higher
than the level in its distributory tank. On the hydraulics, see Hodge (:qq.)
r8.q quae capite . . . libram aequat Cf. ¸.6, :¡.:–¸. The proximity
of the sources prompts F. to explain the noticeably different levels: at Porta
Maggiore the specus of Marcia is about :o m beneath that of Claudia. For capite
(‘the source’), ablative of respect, cf. q¸.¡ (Claudia) dotibus aequatura Marciam.
sed veteres . . . traiciuntur The parenthesis applies to Anio Vetus as
well as Marcia, for F. is contrasting the earlier Anio aqueducts with the newer
(and higher) Claudia and Anio Novus.
humiliore derectura The word derectura is rare and technical (TLL ¸:
:.¸o.:6): it relates to deliberate calibration (note the verb derigere, used at .o.¡,
¸¸.:, ., :.¡.¡; cf. :.6.¡) of the downward course of the rivus from source to
destination. ‘Lower levelling’ describes the loss of altitude inevitable when the
course is longer: see 6.6n.
sive . . . seu See q.¸n. F. seems to have overlooked the additional expense
involved in building bridges and arches in distant places. Taylor (.ooo) .q–¸o
dismisses F.’s first explanation with the comment that the Romans of the third
century did not have the engineering expertise to build arcades, but this fails
to account for the humilior derectura of Marcia vis-` a-vis Claudia, for Marcia’s
builders plainly were able to construct an extensive arcade into Rome.
ad subtile explorata For ad subtile, prepositional phrase used adver-
bially, cf. .8.. in solidum, ¸¡.¸ in tantum. The meaning of explorata is ‘found
reliable by testing’ (experience). F. uses the word twice elsewhere: 6¡.¸ in ex-
ploranda fide aquarum, ¸¡.¸ copiam . . . durantem exploravi. In Pliny’s letters it can
denote responsible decision-making in an engineering context: Ep. x.¡. sane
plane explorandum est diligenter, 6:.¸ et haec et alia multo sagacius conquiret explorabitque
COMMENTARY :8.¸–:8.6
librator; qo.:, q:.:. Cf. TLL ¸..: :¸¡¸.¸¸, OLD adj. ¸. There are some twenty
examples of explorare in F.’s Str., a predictable word in military contexts.
arte librandi See 6.6n. libramento, §:n. diversa libra.
interciperentur The verb intercipere (¸.:n.) is appropriately used of
hostile actions (TLL ¸.:: .:6¸.., .:66.6:).
contra Italicos bella gererentur As Poleni noted, the remark better
applies to the subterranean channel of Anio Vetus (.¸. ncr). Marcia’s arches
near the City would have been vulnerable in any case. (Their water was cut
off in the Gothic siege of ¸¸¸ cr: Procop. BG i.:q.:¸ and :8.) For F.’s use of
the term Italici cf. Str. ii.¸.:6 quorum pars non solum ex diversis gentibus, sed etiam ex
Italicis constabat; novissimos Italicos constituit. DeLaine (:qq¸) :.¡ observes that this
passage reinforces F.’s intent to connect aqueducts with the larger progress of
Roman history, ‘the development of the aqueducts parallels Roman territorial
expansion in Italy’; cf. ¸.: Samnitici belli.
r8.j iam tamen . . . traiciuntur If the text is correct, the subject of
traiciuntur can only be an understood aquae (cf. plural aquas mergebant in §¡),
making this a general statement applying not only to Marcia but to Anio Vetus
as well (but cf. §6n.). Datable repairs prior to F.’s day, meagre as they are,
do not reveal any examples of extensive shortening of either aqueduct: Van
Deman (:q¸¡) 6:–¸, :.8–¸.. An excellent example of the Anio Vetus of what
F. describes here is present in the Hadrianic work carried out in Valle della
Mola di S. Gregorio ¸o.8¸q: Ashby (:q¸¡) 68–¸o; cf. Mari (:qq:a) :6¸. On
replacements and shortening, see Hodge (:qq.) :¡¸–6.
ductus vetustate dilapsus See ¸.:n. Further examples of vetustas with
labi and its compounds: Livy, xxx\..6.6 navigium . . . vetustate dilabentem, Suet.
Aug. ¸o.. aedes sacras vetustate conlapsas aut incendio absumptas refecit, Claud. .¸.¸
templum . . . vetustate conlapsum ut . . . reficeretur, Tac. Hist. i.68.: dilapsis vetustate
moenibus, Ann. i\.¡¸.¡ aedem . . . vetustate dilapsam restaurari.
arcuationibusque Arcuatio (a neologism) occurs four times in our text,
always in the plural: here, §6, :.:.:. The plural is understandable: to describe
discontinuous segments for Marcia at ¸.8 F. resorted to pluribus locis per valles
opere arcuato. Espinilla Buis´ an (:qq¸) 6¡¡–q discusses the difference between
arcuatio and opus arcuatum (¸..n.) as well as the ctcç arcuatura (¸.¸n.).
r8.6 similiter suffecturus . . . erigeretur ‘If it were now raised’ (i.e.
if it had in the past been raised: eriguntur = erecti sunt, ‘are elevated’ rather
than ‘are being raised’), ‘it would furnish’: for future participle thus used in
a condition cf. Str. i.:.q detractaturo pugnam, si intellexisset. The contrary-to-fact
condition makes clear that Anio Vetus had never delivered water to Rome’s
higher elevations (¸.¸n.) and could never do so. Ashby’s proposed intake of
Anio Vetus at Vicovaro (6.¸n.) is .6¡..¡ m.a.s.l.: Reina et al. (:q:¸) i\.:¸, p.¸q.
Like the five higher aqueducts (similiter looks to §¸), it would have been able
COMMENTARY :8.¸–:q.:
to supply more elevated parts of the City if it had not lost altitude by skirting
valleys in its outlying stretches.
{veteris} The transmitted veteris is puzzling. Krohn supposes the abbre-
viation for ter to be a misreading of in: palaeographically clever, but stylistically
awkward in having as antecedent of is the genitives vallium summissarumque re-
gionum. Poleni believed veteris to be an intrusive notation (fromAnio Vetus). The
word more likely conceals a corruption (perhaps by perseveration: cf. vetustate
in §¸ as well as vetus), and one rather expects an adverb such as altius (cf. TLL
¸: ¸8¡.::).
r8.¡ ex urbano agro ‘From territory near the City’ (Evans). For the
sources of Appia and Virgo see ¸.¡n. and :o.¸n. These two had the lowest
fall of all the aqueducts: Appia approximated the Vitruvian minimum (§¡n.)
with c. o.¸ per cent, while Virgo had only c. o.o.¸ per cent: Ashby (:q¸¸) ¸¡,
:¸:; but cf. Taylor (.ooo) .q. The level of Appia may in part be explained
by its age, but Agrippa’s Virgo had been deliberately built to serve low-lying
r8.8 omnibus humilior Alsietina The statement applies to the eleva-
tion of the terminus (about :6 m a. s. l.: see ...¡, ::.:n.). A section of Alsietina
discovered on the Janiculum in :q.¸ is about ¸: m above sea: Van Buren–
Stevens (:q.¸a,b), Nash i: ¸¸–6, Taylor (.ooo) .8.
Transtiberinae regioni et maxime iacentibus locis F. seems to
distinguish between the Augustan Regio xi\ (::..n.) and (other) extremely
low-lying places. He tells us later on that the entire supply of Alsietina extra
urbem consumitur (8¸n.), but that is no reason to suppose that Transtiberina regio
here must refer to the suburbs rather than to Regio xi\ itself. For participle
iacens = humilis cf. Tac. Hist. i.86.. iacentia et plana urbis loca.
rq.r Ex his sex The six to which F. refers are Marcia-Tepula-Julia (§§¸–¸),
Anio Novus-Claudia (.o.:–.), and Anio Vetus (.:.:). There were, however, no
more than five settling-tanks, for Tepula assumed an identity of its own only
after Julia’s piscina (below §¸; cf. 68..–¸). See Map ¸ for the routes of the
aqueducts in this area and tentative locations of the piscinae.
via Latina The topography of the Roman campagna practically dictated
the course of the aqueducts. They followed a prominent ‘finger’ of higher land
extending from the seventh milestone towards the City. By using this natural
feature the builders of Anio Vetus needed substructures for only a fewhundred
metres in the valley close outside Porta Maggiore: Ashby (:q¸¸) 8o–:. The
two arcades, similarly, maintain altitudes which would otherwise have been
contectis piscinis . . . limum deponunt The only one of this series
of settling-tanks which has been discovered is that of Anio Novus (.o.:n.). On
COMMENTARY :q..–:q.¸
settling tanks in general see Hodge (:qq.) ::¸–.¸, .¸¸–q. Blackman–Hodge
(.oo:) :.¸ point out that these piscinae might have functioned more as cross
links and junctions for transferring water than for purposes of purification.
Hodge (:qq.) :.¸: ‘Often we find simply an enclosed vaulted reservoir, and it
is hard to tell whether purification was part of its intended function, since on
going through a tank designed for any function – such as a junction with a
branch line arriving or leaving – the current would slow down and impurities
settle, whether that was the chief purpose intended or not.’ With contectus F.
must mean, in addition to ‘covered’, that these piscinae were underground:
they would have to have been so to accommodate transfers (68.¸–¡) before the
channels were taken up on arcades; cf. §¸n. emergunt.
quasi respirante rivorum cursu For the apology see ...n. velut; for
rivus allowing this metaphor ¸..n. ductus cuiusque.
rq.z modus . . . initur The verb inire is here a mathematical term: to
determine by calculation, often with rationem (OLD s.v. 8). F. uses it only in
the passive: in the present formula (two further instances: see next note), also
.¸..n. ratio eorum initur (conjecture) et computatur, ¸:.: ratio fistularum . . . omni
genere inita. See also 6q..n. mensura in<ven>iri non potuit.
mensuris . . . positis The mensurae (noun always plural in this concrete
sense) are some kind of gauges for measuring the water available for urban
distribution. F. cites their readings in his investigation of the supply: 66.¡ (Anio
Vetus) modus in piscina, qui per mensuras positas initur, 6¸.¸ (Marcia) in piscina mensuris
positis initur, 6q.. (Julia) modus eius manifestis mensuris efficit; cf. ¸..¸ (Claudia) in
piscina ubi indubitatae mensurae sunt. (On his failure to do so for Anio Novus
see ¸¸.¡n.) That he mentions them here as a feature of the overall system
and describes them as positae (cf. ¸:.. [moduli] qui in commentariis . . . principis
positi et confirmati sunt; see also ¸.:n. ponam) strongly suggests that the gauges
antedated his tenure and that their installation was permanent and official.
The importance of these piscinae to F. may explain why he never mentions
terminal castella (¸..n.) within the City; at these there might have been no
similar way of effecting measurements. Because F.’s interest was administrative
rather than technical, we do not know what the gauges were like or how
readings were taken. It is my view that the mensurae were little more than scales
to show the depth of water in the outflow channel, although they might have
been calibrated to indicate the area of cross-section (which could readily be
converted to quinariae: see 6¸.¸): Rodgers (:q86b). It is theoretically possible, on
the other hand, for them to have been more sophisticated devices capable of
creating hydraulic conditions comparable to those in the closed pipes of urban
distribution: Hodge (:q8¡).
rq.j una autememergunt ‘Julia, Marcia, and Tepula emerge together’
(the names in this order to allowthe parenthesis for Tepula). The sense resumes
COMMENTARY :q.¡–:q.¸
with hae tres: ‘These three are taken up on the same arches.’ The channels
of Tepula and Julia run superimposed upon that of Marcia from the point
where all three emerge near Capannelle: Ashby (:q¸¸) :.8–q, :6¸. F. has just
described the general location of the piscinae. He now reveals that those of the
high-level aqueducts are located at the very end of their subterranean course
(§:n. contectis), and he is about to explain that five channels are carried on two
arcades: the first (older) of these carries Marcia-Tepula-Julia, a second (higher)
one carries Claudia-Anio Novus (.o.:). Remedies applied to the transmitted
text have left awkward syntax (and tres earum is unidiomatic). Emergunt here is
consistent with the archaeological evidence and gives a better point to rursus
emergunt in §6.
(quae . . . venit) For Tepula see 8..n. and q..n. Giocondo’s accipit is
unnecessary (even with nunc): note the pluperfect accesserat and cf. .o.¸ modum
quem acceperunt . . . dimittunt.
a piscinis At 6q.. F. places Julia’s reservoir ad sextum ab urbe miliarium (see
note ad loc.). There seems also to have been some kind of side-channel for
Marcia (6¸.¸n.), no doubt related to the fact that it delivered q. quinariae to
Tepula at this point (6¸.¸, 68.¡).
rq.q summus . . . Marciae Withthe masc. summus one must understand
rivus: Professor Reeve prompted me to see that it might as well be printed. As
they leave Porta Maggiore the three channels are now incorporated into the
Aurelian Wall; they crossed the Via Labicana on an impressive Augustan arch
which became Porta Tiburtina, known as the Porta Taurina (from the bull’s
head on the keystone), modern Porta S. Lorenzo (P–A¡:¸, ¸68; Nash i: ¡q–¸o,
NTD ¸oq–:o, LTUR iii: ¸:.–:¸, Pisani Sartorio). For inscriptions on the arch
see Appendix B, nos. :, ..
rq.j quae ad libram . . . deveniunt The three channels went under-
ground when the arches ended (at a point not far from the corner of the Via
Marsale and the Viale Castro Pretorio). Cippus no. ¸ was located between V.
Castro Pretorio and Via Milazzo, and the three specus were discovered running
separately but intersecting each other: Lanciani (:88:) ¸o¡ and, figs. ¡,
a–c; Van Deman (:q¸¡) ::6–..; Ashby (:q¸¸) :¡6–q; Cattalini (:q86a); Evans
(:qq¡) 8¸; Aicher (:qq¸) ¸8–q. Grimal’s continenter seems weak, for a supplement
to the transmitted con- need hardly express the fact that the channels ran close
together: this is clear enough from in eosdem arcus and the plural verbs. The
participial ending is unhelpful, and attempts to build upon B¨ ucheler’s [flu]entes
have had no success: una fluentes Grimal, adfluentes Kunderewicz, infra euntes or
transeuntes Gonz´ alez Rol´ an. What the damaged text presumably contained is
an explicit reference to the engineering requirement imposed by the topog-
raphy: a short subterranean section just outside the gate (ubi rursus emergunt:
§6). Without much confidence I propose con[tine]ntia [subterflu]entes: what the
conduits ran beneath was the area known as continentia aedificia (see ¸..n. extra
urbem, :o¡.:, :.¸.:).
ad libram [collis Vi]minalis The phrase ad libram means ‘at the ele-
vation’ (libra = altitudo ¸.., :8.:n.) or here ‘along the high ground’ (a slightly
different sense at ¸6..n. ad libram conlocatus). Because their libramentum had pre-
served altitude along the entire course (noticeably with the arcade from the
seventh mile), Marcia’s engineers were able to reach the Porta Viminalis, a
higher level than had been possible for their predecessors: Anio Vetus had
been, one might say, ad libram portae Esquilinae (.:.¸n.).
ad Viminalem . . . portam RE 8A: .¸88, P–A ¡:q, NTD ¸:o, LTUR
iii: ¸¸¡ (Coarelli).
rq.6 ubi rursus emergunt Cf. §¸n. emergunt. C’s ibi produces a sentence
far shorter than is F.’s norm. The clause refers to Viminalem . . . portam. Parallel
are 6.¸ extra portam . . . ubi partem [dat], 6¸.¸ ad Gemellos . . . ubi iungitur cum
ramo Augustae, ¸o.¸ in agro . . . Commodi ubi . . . cursum habet, q¸.. super villam . . .
Sublaquensem ubi limpidissima est; cf. :¸.., :q.:, 6q.., ¸..¸ all relating to piscinae. F.
specifies no particular terminus for the main channels of the three aqueducts,
although emergunt may suggest that he considers the intra-urban distribution to
begin at Porta Viminalis (note ...¸ [Appia] emergit . . . infra clivum Publicii and ¸.q
incipit distribui in imo Publicii clivo). Some distribution seems to have taken place
near Porta Viminalis, to which a small circular tower standing until recently in
the Piazza d. Cinquecento was presumably related (although this was hardly
the terminal castellum): Ashby (:q¸¸) :¡q, Nash i: ¸:, Evans (:qq¡) 8¸. Cippus no.
. was found just outside the gate, indicating a distance of forty-five feet from
cippus no. :. The channels turned at this point fromWto WSWand crossed the
agger of the Republican wall near the gate. Inside the wall they turned again at
right angles and followed the wall towards the SE. At this turn was found the
cippus erected by Didius Gallus (Appendix B, no. q); a pair of similar markers
were found ::o m further along. The labels with trium aquarum indicate that all
three aqueducts continued in a SEdirection towards Porta Esquilina, although
no clearly defined terminus has been found. Another branch of the aqueducts
turned NW from inside Porta Viminalis: Julia and Tepula have been identified
with two channels following the agger; a third specus running at a higher level
seems to have been contemporary with the Baths of Diocletian. These three
conduits lead to what were apparently three reservoirs found on the site of
the Ministero delle Finanze. Ashby (:q¸¸) :¡q–¸o believed that the terminal
castella of Marcia-Tepula-Julia lay hereabouts in F.’s day and suggests that the
cippi no longer accurately represented the terminal section. F.’s text, however,
is not necessarily at odds with the archaeological and epigraphic evidence. If
some or all of the channels forked at Porta Viminalis he could fairly, if vaguely,
consider that the point where they emerged marked their ‘end’.
COMMENTARY :q.¸–:q.8
rq.¡ pars Iuliae . . . castellis Caelii montis diffunditur Note the
dative (with compound verb) where preposition +accusative might have been
expected; cf. ¸q..–86.¸ dividebantur in castella. The statement that a branch of
the Julia runs from Spes vetus (¸.6n.) to the Caelian is puzzling, for at ¸6.¸–6
F. says that Marcia and Julia had once supplied the Caelian and Aventine but
had been replaced (in fact if not in name) when Nero brought Claudia on the
arcus Neroniani to the Caelian. At 6q.¸ he records that Julia receives :6. quinariae
from Claudia post hortos Pallantianos, a quantity which he does not reckon in
computing Julia’s losses (6q.6n.). Although the topographical references are
not quite synonymous (see ¸.6n. and :q.8n.), Lanciani (:88:) ¸:¡–:¸ perhaps
rightly identifies this pars Iuliae with the :6. quinariae which Julia receives from
Claudia post hortos Pallantianos (6q.¸n.). This pars Iuliae would thus have been
only nominal, for the water was that of Claudia carried on the Neronian arches
to the Caelian but distributed in the same manner as the Julia, which it had
replaced (¸6.¸ nulla enim castella adiecit, sed isdem usus est, quorum quamvis mutata
aqua vetus appellatio mansit). Against Lanciani’s suggestion it can be objected
that F. nowhere mentions that Claudia delivers water to a nominal Marcia for
delivery to the Caelian and the Aventine. An earlier branch of Julia had in fact
supplied the Caelian, but there is no evidence for its course or for supposing
that it originated at Spes vetus. It is possible that the arch of Dolabella and
Silanus (:o cr: CIL 6.:¸8¡; Nash i: ¡6, ::¸), later built into the arcus Neroniani,
originally carried an aqueduct across the Clivus Scauri (modern Via d. S. Paolo
della Croce). For a branch of Marcia leading to the Caelian see ¸6.¸n. and
rq.8 Marcia . . . post hortos Pallantianos Behind the horti Pallantiani
(RE 8: .¡86, P–A .¸o, NTD .o:, LTUR iii: ¸¸, Mancioli) also lay the terminus
of Claudia-Anio Novus (.o..). The location of this property can be fixed with
some certainty ‘in the triangle formed by Viae Tiburtina vetus, Praenestina-
Labicana and the Aqua Marcia itself: in other words, somewhat south of
Piazza Vittorio Emanuele’: Ashby (:q¸¸) :¸¡. These horti took their name
from Claudius’ freedman Pallas, to whom they had belonged.
in rivumqui vocatur Herculaneus Not to be confused with the rivus
Herculaneus of Anio Novus (:¸.¡) or with that which Pliny, perhaps quite erro-
neously, associates with Virgo (:o.¸n.). The name conceivably is related to a
temple or statue of Hercules Sullanus knownfromregionary catalogues: Nordh
(:q¡q) 8o, P–A .¸6, NTD :88, LTUR iii: .:–. (Palombi). The starting-point
of this rivus Herculaneus has usually been identified with remains (apparently of
a castellum) incorporated into the Aurelian Wall at the fifth tower S of Porta
S. Lorenzo: Piranesi (:¸¸6) i: tav. ¸8; Lanciani (:88:) ¸o¸; Van Deman (:q¸¡)
:¡o–:; Evans (:qq¡) 8¸; for grave doubts, see Ashby (:q¸¸) :¡¡. Note that F.
gives no date for its construction (cf. the specus Octavianus .:..n., and contrast
COMMENTARY :q.q–.o..
Appia’s ramus Augustae ¸.6n.). Its course apparently was that followed later by
the arcus Neroniani of Claudia (.o.¸–¡nn.). Remains of a channel distinctively
built of tufa blocks with a circular hole in the centre have been found belowVia
Statilia and at several further points (in one case two channels superimposed)
in the neighbourhood of Villa Wolkonsky: Van Deman (:q¸¡) :¸q–¡:, ¡o.;
Ashby (:q¸¸) :¸¡–¸; Colini (:q¡¡) 8.–¡, 88; Evans (:qq¡) 8¸–8. Dating and
attribution are far from certain: Lugli (:q¸:) ii: ¸¸6, for example, assigns them
to Appia.
rq.q hic per Caelium I prefer hic to is (cf. ¸.¸, q.¸, :q.¸, .o.¡). C’s se
(coming after deicit) may result from the earlier sui. The stone pipe channel was
discovered near the hospital of S. Salvatore (S. Giovanni) and again on the
Caelian, under the Neronian arches in the Via di S. Stefano Rotondo. It was
below ground level and thus unavailable for use on the hill itself. An Augustan
cippus of Marcia alone (CIL 6.¸:¸6o) found near Piazza di S. Giovanni in
Laterano may belong to this conduit, but the distance indicates that it was not
found in situ and it may as easily have belonged to Marcia’s high-level branch
to the Caelian (¸6.¸n. and 8¸.¡n.): Ashby (:q¸¸) :¸¸.
<f>initur supra portamCapenam Although supra elsewhere means
‘beyond’ (6.¸ supra Tibur, q¸.¸ supra Trebam Augustam), the rivus Herculaneus prob-
ably ended above – not ‘beyond’ (Loeb) – the Porta Capena (¸.¸n.). Juvenal
¸.:: refers to veteres arcus (perhaps those of the aqueduct) madidamque Capenam,
and his scholiast explains madidam ideo, quia supra eam aquaeductus est, quem nunc
appellant arcum stillantem. Martial iii.¡¸.: mentions Capena grandi porta qua pluit
gutta. The leaks might have come from a castellum located above the gate and
supplying the low-lying region along the Via Appia. For discussion of the to-
pographical problems associated with this arcus stillans, see Evans (:qq¡) 8¸–8.
The likelihood of a castellum at this spot is no reason to defend C’s initur, a word
not otherwise used in a sense approximating ‘be available for delivery’.
zo.r Anio Novus et Claudia a piscinis No settling-tank has been dis-
covered for Claudia (¸..¸). That of Anio Novus was identified in :88¡, about
¸o m east of the Villa Bertone near Capannelle. It consisted of two cham-
bers, both of which were filled with calcareous pebbles (and the villa itself was
constructed on an artificial mound formed by deposit cleared from the tank):
Ashby (:q¸¸) ..6; Blackman–Hodge (.oo:) ¸8. It is curious that F. gives no
measurements taken at the piscina of Anio Novus (¸¸.¸–¸n.).
altiores arcus Higher, that is, than those of Marcia-Tepula-Julia (:q.¸).
Cf. :¸.¸ arcus altissimi, where it is only implied that the channel of Anio Novus
ran atop Claudia.
zo.z finiuntur arcus earum post hortos Pallantianos For the horti
Pallantiani see :q.8n. Although F. does not say so here directly, the two
COMMENTARY .o.¸–.o.¡
waters shared a single castellum divisorium (¸..n.) in which their waters were
mixed (¸..6, 86.:). Until :88o this castellum existed in ruins near the ‘three
arches of the railway’, but it has since disappeared: Piranesi (:¸¸6) tav.
¸8 ( = Nash i: ¸¸); Lanciani (:88:) ¸6. and tav. ii, fig.¸; Mucci (:q86b)
inusumurbis fistulis diducuntur F. never mentions a terminal castel-
lum (¸..n.), but rather the place where distribution begins, with diducere (¸.:n.
perducta sit) here, with distribuere (¸.q, .:..), or merely by noting the end of the
conduit (.... arcus Virginis finiuntur, ¡ Alsietinae ductus . . . finitur); cf. :q.¸ pars
Iuliae . . . diffunditur, 8 Marcia partem sui . . . deicit).
fistulis Fistula in general means a water-pipe of lead (TLL 6.:: 8.8):
Varro, Ling. \.:.¸ fistula . . . a qua fusus aquae, Vitr. \iii.6.:o; cf. Hor. Epist.
i.:o..o aqua tendit rumpere plumbum. In F. this use is consistent, although only
once do we find fistula with the adjective plumbea (:o¸.¸). Bruun (:qq:) :.¡–8
suggests that F. may be using the term fistula more loosely here than elsewhere,
and he rightly emphasises that we cannot use this passage as evidence that
all water in the urban distribution system was conveyed through lead pipes
(for open channels, see .¸.¸n. ne rivus convulneretur). He observes that most of
F.’s fistulae are private pipes, or pipes delivering water for private use (.¸..,
¸6.¸, 6¸.6, ::¸..n., etc.); from F. he cites only :o6.¸ ne rivi aut fistulae publicae
lacerentur and perhaps ::o.:n. ex manationibus fistularum as clear instances of lead
pipes used for water mains (add that in the Lex Quinctia, :.q.¡, 6 fistulae are
modified by the genitive aquarum publicarum). Surely, however, ¸..¡–¸ (plerumque
erogant . . . adsidue accipiunt) refers to pipes used in wider distribution than merely
to privati.
zo.j partem . . . transfert For this high-level branch of Claudia see
¸6.6 and 8¸.¸. The precise date of Nero’s work is not known, but it was no
doubt after the fire of 6¡ (cf. Tac. Ann. x\.¸8.. initium in ea parte circi ortum quae
Palatino Caelioque montibus contigua est; cf. ¡¸ with the date: :q July) and probably
was connected, tangentially at least, with his project for the vast Domus Aurea
(cf. Suet. Nero ¸:.:–.): Lanciani (:88:) ¸¸o, Ashby (:q¸¸) .¡q. Not improbably,
indeed, Nero was responsible for the extension to the Palatine (¡n.). The arches
were later called Caelimontani (CIL 6.:.¸q = ILS ¡.¡) because they led to the
Caelian. On this branch in general, see Evans (:q8¸); Mucci (:q86c); Evans
(:qq¡) ::8–.:.
zo.q hi derecti . . . terminantur The course can be traced by exten-
sive remains (most of which are to be attributed to the Severan restoration):
Lanciani (:88:) ¸6¡–¸¡, Van Deman (:q¸¡) .66–¸o, Ashby (:q¸¸) .¡¡–q,
Colini (:q¡¡) 88–:o6; Aicher (:qq¸) 6:–¸. The branch leaves Claudia’s main
conduit at the right-angled turn before Porta Maggiore, spans Via Eleniana,
and turns along Via Statilia for a short distance before crossing the Villa
COMMENTARY .o.¸–.:.:
Wolkonsky (where it follows the line of an ancient road). It continues in Via
Domenico Fontana, along the north side of Piazza S. Giovanni in Laterano,
then to Via di S. Stefano Rotondo, turning at Via della Navicella and proceed-
ing across the Arch of Dolabella and Silanus towards the site of the Temple of
Claudius in the garden of SS. Giovanni e Paolo (Nash i: ¸8–¡6).
templum divi Claudi The Temple of Claudius was begun by Agrip-
pina, almost entirely destroyed by Nero, restored and completed by Ves-
pasian (Suet. Vesp. q.:): P–A :.o–:, Nash i: .¡¸–8, NTD 8¸–8, LTUR i: .¸¸–8
(Buzzetti). Nero’s destruction was perhaps connected with building a distrib-
utory castellum for the new branch aqueduct, but this is nowhere explicit;
cf. Ashby (:q¸¸) .¡¡. In F.’s day the arches ended alongside (iuxta) the temple
(cf. ¸6.6 usque ad templum). The undamaged label AQVEDVCTIVM appears
on the Severan marble plan adjacent to the temple (Pianta 6¸, tav. x\i; Nash i:
.¡¸); for this vulgar form see Jordan (:8¸¸).
zo.j circa ipsum montem The Neronian aqueduct divided near
Piazza della Navicella, with a branch leading to the Aventine (see below). The
greater part of the supply for the Caelian was distributed fromnear the Temple
of Claudius (¸6.6), and from the same point a channel perhaps originally
led to the Stagnum Neronis, where the Colosseum was later to be erected:
Lanciani (:88:) ¸¸o; Evans (:qq¡) ::q. A smaller distribution was effected be-
fore reaching the terminus, near the churchof S. Stefano Rotondo: I. Gismondi
apud Colini (:q¡¡) :o., fig. ¸8.
in Palatium Aventinumque Any branch arcade which may have ex-
tended from the Caelian to the Aventine has entirely disappeared except for
possible vestiges near S. Prisca: Ashby (:q¸¸) .¡8–q. The Palatine was served
by an extension from the Temple of Claudius, arches of which remain in the
valley between Caelian and Palatine (Nash i: ¡6). Van Deman (:q¸¡) ¡:¸ calls
the work Neronian (although Ashby loc. cit. associates it with Domitian’s build-
ing on the Palatine); Evans (:qq¡) :.o, :.. emphasises that delivery of Claudia
to the Palatine was perhaps already under Claudius part of a plan to supply
the imperial residence with Rome’s newest (and highest) water. Severans pro-
claimed that tradition in extensive repairs and restorations recorded in CIL
6.:.¸q ( =ILS ¡.¡: see Appendix B, no.:.). The arches can never have been
high enough to carry water to the Palatine without the use of a siphon (¸.¸n.
atque ita in Capitolium): Ashby (:q¸¸) .¸o.
in regionem Transtiberinam Presumably from the Aventine, by
pipes laid across the Tiber bridges (cf. ::..n.).
zr.r Anio vetus . . . piscinamhabet The site has not been discovered.
This aqueduct followed a subterranean course along the ridge traversed by the
Via Latina, with no single point obviously appropriate for a settling-tank. F.
does not mean to give exact directions, but only to show the general location
of this piscina relative to those of other aqueducts (the point of et ipse). Ashby
(:q¸¸) ¸q inclines to identify the piscina with a castellum viae Latinae contra dracones
(CIL 6..¸¡¸: Appendix B, no.:.), although this may more likely have served
for deliveries ante piscinam (66.¸); cf. Lanciani (:88:) .6o–:.
citra quartum miliarium On the Via Latina. Near this point Anio
Vetus crosses beneath the arches of Marcia-Tepula-Julia, just before the triple
aqueduct intersects with Claudia-Anio Novus. It then turns northward, away
from Via Latina and towards Via Labicana.
in tramite qui . . . traicit ‘On the pathway (or service road?) which
runs between the arches from Via Latina to Via Labicana.’ The transmitted
text presents two related difficulties: (:) What is the subject of traicit? (.) What
lies behind the corruption intra novie? It is well to address the questions in this
order. It has gone unnoticed that inter arcus can help to determine ‘that which
crosses’. Neither arcade could be thus described. The underground channel
of Anio Vetus, which leaves Via Latina hereabouts and parallels the course of
the higher aqueducts, might perhaps be said to run ‘along with’ them (taking
inter less strictly than ‘between’). But surely F.’s point is not merely to delineate
the course of Anio Vetus. Schultz seems to have been first to suggest that the
subject of traicit is a roadway: <ubi> intratur via, quae a Latina . . . traicit. Ashby
((:q¸¸) ¸¸ n.6) considers intra novum et viam quae a <via> Latina. Fabretti long
ago noted traces of a road which left Via Latina at the fourth mile. But even
more appealing is the official service road for the aqueducts, parts of which
have come to light near Porta Furba (Ashby (:q¸¸) :¸6 and fig..¸), where it
clearly ran between the arcades. F. does not elsewhere use trames (deverticula,
for his purposes, are dead-end side roads: ¸.¡, ¸.6, etc.), but the word aptly
fits a special work road giving access to all six aqueducts along this crucial
stretch (cf. :.¡.:–.). Scholars have grasped at the transmitted novie (so close
to the phrase inter arcus) as a reference to Anio Novus. It is not at all clear,
however, that F. would refer to this aqueduct simply as novus (cf. q:.:–., q.,
q¸.:). B¨ ucheler’s version (infra novum qui a <via> Latina) must be rejected for
two reasons. First, as noted already, the phrase inter arcus cannot apply to Anio
Novus. Second, AnioVetus seems tohave turnedtoosoontohave run‘beneath’
Anio Novus at all. Grimal tried to reflect topographical reality by writing intra
Novi <sp>e<cum> q<ua a> via Latina . . . traicit, ‘before reaching the channel
of Anio Novus, at the point where it passes under the arches [of Marcia etc.]
and crosses from Via Latina to Via Labicana’. Apart from palaeographical
insouciance, the impossible use of intra, and the curiosity of taking inter arcus
to mean ‘beneath the arches’, there is no reason whatsoever why F. should in
this context use the word specus (:¸..n.) of any aqueduct other than that of
Anio Vetus itself. If my in tramite requires further justification, I venture the
guess that -ui- may be a vestige of a gloss (via) incorporated by a subsequent
COMMENTARY .:..–.:.¸
zr.z intra secundum miliarium Presumably on the Via Labicana:
Grimal ¸q n.¸8. Lanciani (:88:) .6¡ supposed that F. was measuring along the
specus itself; this would put the junction about a mile outside Porta Maggiore:
Ashby (:q¸¸) 86.
in specum qui vocatur Octavianus Like Marcia’s rivus Herculaneus
(:q.8n.), this branch might have been very old. Remains identified as the specus
Octavianus are scanty and doubtful, so there is no firm archaeological evidence
for its date: Ashby (:q¸¸) 86–¸, pace Van Deman (:q¸¡) 66. Scholars generally
associate it with Augustus or with Agrippa’s work in ¸¸ ncr (perhaps incau-
tiously, for Caesar’s heir did not use the name Octavianus): Lanciani (:88:)
.6¡, Ashby (:q¸¸) ¸¸, Hainzmann (:q¸¸) 86, Robinson (:q8o) ¸q, Roncaioli
Lamberti (:q86) ¡o, Evans (:qq¡) ¸8–q (perhaps ‘little more than a nickname’).
in regionem viae Novae ad hortos Asinianos Neither of F.’s topo-
graphical references can be identified. By Via Nova he can neither mean the
important street of that name built by Caracalla a century later (NTD ¡:¸,
LTUR \: :¡.–¸, Patterson) nor a much earlier Via Nova that ran between
Forum and Palatine (Livy, i.¡:.¡; NTD .68, LTUR iii: ¸¡6–q, Santangeli-
Volpe). Grimal deleted in regionem viae Novae as a late interpolation, but it is
safer either to assume that F.’s via Nova is a street otherwise unknown to us
or that the two words are corrupt. The horti Asiniani (NTD :q¸, LTUR iii: ¸¡,
Chioffi) are perhaps to be identified with the area to which Pliny, HN xxx\i.¸¸
refers as Asinii monumenta. The property has conventionally been located south
of the Caelian, near the later Baths of Caracalla: Grimal (:q¡¸) :68. The
low level of Anio Vetus would indeed suggest that this is the area to which
F. refers.
per illum tractum See q.¸n. omnes villae tractus eius.
zr.j rectus vero ductus The main channel has been traced from just
inside Porta Maggiore to the Esquiline Gate. It did not proceed in a direct line,
but ran NW (perhaps to avoid the valley of the Villa Altieri) and then turned
sharply to cross the agger beneath the present Stazione Termini. Thence it
followed the agger SE to Porta Esquilina: Lanciani (:88:) .6.–¡; Van Deman
(:q¸¡) ¸¸–¸; Ashby (:q¸¸) 8:–6; Roncaioli Lamberti (:q86a); Evans (:qq¡)
¸¸–6. Brunt (:q¸:) ¸8¡ notes that this point of entry and the pattern of dis-
tribution for Anio Vetus show an expansion of the City eastward in the early
third century.
secundum Spem veniens Giocondo’s <veterem> is plausible enough
since veniens follows, but it is not essential. F. uses Spes alone at ¸6.6 and 8¸.¡.
intra portamEsquilinam RE6: 6¸8, P–A¡o¸, NTD¸o.–¸, LTURiii:
¸.6–¸ (Coarelli); cf. NTD .¸ (Arcus Gallieni). Excavations below the church
of S. Vito have revealed remains of the channel and a castellum: Santa Maria
Scrinari (:q¸q). Intra apparently means ‘just inside’; cf 6¸.¸n. intra Spem veterem.
COMMENTARY .:.:–....
in altos rivos The distributory channels were probably ‘deep’ (Ashby
(:q¸¸), Grimal) and not ‘high-lying’ (Bennett, Van Deman (:q¸¡)): cf. Santa
Maria Scrinari (:q¸q) and :8.6n. Bruun (:qq:) ::q notes that rivus means free-
flowchannel (¸.:n. ductus cuiusque), as opposed to pipes (fistulae .o.¸n.), and that
the plural here suggests several such channels since this aqueduct supplies ten
of the fourteen regiones (8o..).
zz.r conceptacula, id est piscinas The latter phrase suspiciously re-
sembles a gloss, and conceptaculum is not otherwise attested in quite the sense
used here (a space to contain water) TLL ¡: .:.::. There seems to be no sig-
nificant relationship to the verb concipere (see concipitur ¸.¡n.; conceptio 66.6n.).
Unlike the six aqueducts F. has so far discussed (cf. :q.:), these three have
no settling-tanks. Virgo and Appia came shorter distances, and neither was
afflicted by the turbidity (:¸.:, qo.:) or the calcareous deposits (:...:) which
plagued the Anio aqueducts. Alsietina was delivered for meaner uses and its
quality was thus less crucial (::.:). A piscina clearly associated with Virgo is
mentioned in CIL 6.¸qo8¸, but its location in the Campus Martius connects it
to public or recreational uses: Lloyd (:q¸q) :qq–.oo. There was a later piscina
for Virgo near the Spanish Steps, possibly Hadrianic in date: Van Deman
(:q¸¡) :¸.–¸.
zz.z arcus Virginis Arches within the City (see :o.¸n.) began in the
triangle formed by the modern streets of Capo le Case, Via Due Macelli,
and San Giuseppe. We can follow their gently curving course from here to
a point very near the terminus: Ashby (:q¸¸) :¸¸–8.. In Via del Nazareno
is an archway which spanned an (unknown) ancient street (Nash i: ¸6): its
inscription (Appendix B, no.6) records that Claudius rebuilt arches destroyed
by Caligula (cf. Suet. Gaius .:; Dio, rix.:o.¸). Amore elaborate archway crossed
the Via Lata (modern Via del Corso), a fragmentary inscription (CIL 6.q.o–¸
= ¸:.o¸–¡) showing that it commemorated Claudius’ British triumph: see
Ashby (:q¸¸) :¸¸–8o, Nash i: :o.–¸. For these arches as shown on coins see
Hill (:q8q) ¸o.
sub hortis Lucullianis It is from F.’s text that we know the location
of this property, on the Pincian, above the modern Piazza di Spagna; they
are called the horti Asiatici in Tac. Ann. xi.:.:, Dio, rxi.¸:.¸. See RE 8: .¡8¸,
P–A .68–q, NTD .oo, LTUR iii: 6¸–¸o (Broise-Jolivet); Kaster (:q¸¡); Broise–
Jolivet (:q8¸, :qq¸, :qq8), Jolivet (:qq¸) .o6–8. Despite the form Lucullano at
¸.¡, 8.:, :o.: (and elsewhere, e.g. Suet. Tib. ¸¸.:), Lucullianis is required here.
Poleni was too cautious in preserving the transmitted Lucilianis (named for an
unknown Lucilius), for the same error occurs in the transmission of Tac. Ann.
xi.¸..: Lucullianos in hortos (Beroaldus: fucilianos M) and ¸¸.: Lucullianis in hortis
(Alciatus: lucilianis M). Note also Plutarch Lucullus ¸q: cí Acuscuììióvci sntci.
COMMENTARY ...¸–.¸.:
finiuntur . . . secundum frontem Saeptorum The location of the
Saepta (NTD ¸¡o–:, LTUR i\: ..8–q) has been firmly established by G. Gatti
(:q¸¡, :q¸¸, :q¸8): Pianta q¸–:o:, tav. xxxi; cf. Nash ii: .q:–.. It lay just to
the east of the Pantheon, where a fragment of its western portico (Porticus
Argonautarum) is still preserved. To the east of the Saepta was the Porticus
Meleagri and to the south was the Diribitorium. The only part of the voting
precinct which could fairly be called its frons is that facing north, along the
modern Via del Seminario. So the aqueduct arches extended at least to Palazzo
Serlupi (where a travertine arch was found in :¸o.) and perhaps as far as the
Piazza del Pantheon. For distribution beyond this point see 8¡..–¸nn.
zz.j rivus Appiae . . . clivumPublicii See ¸.qn. Remains which have
been assigned to Appia are discussed by Van Deman (:q¸¡) .¸–8, Ashby (:q¸¸)
¸.–¸. C’s rivos could be an archaic spelling; more likely it is scribal carelessness.
zz.q Alsietinae ductus . . . factus, finitur Taylor no doubt rightly
concludes from videtur here (along with ::.:n. nisi forte) that in F.’s day water
was no longer flowing into the naumachia.
zj–6j Announced in the prologue as modulorum <rationes>(¸..n.). .¸.: leads
readers to believe that F. is about to turn directly to a presentation of data
(singula subicere) on the supply (copia) and delivery (erogatio) of water, but this he
delays until chapter 6¡. Instead he abruptly announces at .¸.. that first he
will set forth something about the history, nomenclature and calibration of
delivery pipes – including an explanation of irregularities that he has detected.
This matter he covers in chapters .¡–¸¡, with a few additional comments in
¸¸–6. At ¸¸ begins his ‘tabular’ listing (cf. ¸¸.: subieci) of exact measurements
(diameter, circumference, capacity) for twenty-five pipes – ten of which are not
in usu!
zj.r Quoniam. . . persecutus sum Briefly recalls (cf. 6¡.:) the subjects
of chapters ¸–..: auctores and aetates correspond to per quos and quibus consulibus
of the prologue (¸.:), origines to quibus ex locis, longitudines rivorum to quantum
subterraneo rivo . . . opere arcuato (cf. :¸.:), and ordinem librae to altitudines cuiusque.
(Chapters :q–.., on the settling-tanks and terminal courses, are in effect a
continuation of the ordo librae in chapter :8.) The transition is formulaic: cf.
Vitr. i\...6 quorum ingressus persecutus de Ionicis et Corinthiis institutionibus supra dixi,
nunc vero Doricam rationem summamque eius speciem breviter exponam, Colum. ii.::.:
Quoniam quando quidque serendum sit persecuti sumus, nunc quem ad modum quotque operis
singula eorum, quae rettulimus, colenda sint demonstrabimus; \iii.:o.6 Atque ea genera quae
intra septa villae cibantur fere persecuti sumus, nunc de his dicendum est . . .; Hygin. Astron.
i\.:¡.: Praeterea cum reliqua omnia diligentissime persecuti fuerimus, alienum videtur esse
nos non eandem persequi causam.
origines The transmitted ordines is perhaps an error of anticipation (note
longitudines and ordinem).
ordinem librae See :8.:n. libra.
non alienum {autem modi} mihi videtur C’s reading seems to
represent scribal misunderstanding of earlier adjustments. The simple alienum
matches :¸.: non alienum mihi visum est (perhaps the source of a’s conjec-
ture) and :o..: non est alienum. Cf. Str. i.¸.¸ non alienus, ut arbitror, hic locus est
referendi . . .
quanta sit copia F. promises details for the available supply (chapters
6¸–¸6) as well as for the distribution (¸¸–8¸); cf. 6¡.: quem . . . modum quaeque
aqua . . . habere visa sit quantumque erogaverit. See also ¸..nn.
publicis privatisque See ¸..n. publica privataque.
usibus et auxiliis . . . voluptatibus Cf. :¸.:n. publicis usibus et privatis
voluptatibus, :¡..n. adiutorio.
per quot castella . . . erogetur Largely a repetition, in somewhat
different order, of ¸.. quantum extra urbem . . . beneficio principis detur.
per quot castella quibusque regionibus diducatur Cf. ¸q..–86.¸
dividebantur per regiones . . . in castella. For the dative cf. :q.¸n.
zj.z sed rationis existimo For the genitive (of ‘worth’) with existimo see
H–Sz ¸¸, TLLs.v. DeLaine (:qq¸) :.6–8points out that the mathematical chap-
ters are didactic in tone, beginning with definitions (.¡–¸o) and concluding
with ‘a patently philosophical statement on the immutability of measurement’
(¸¡.¸). She compares the brief explanation of gromatical techniques in Colum.
\.:–¸, similarly confined to what the author thinks necessary for his audience.
Cf. ::6.: de qua [sc. tutela] priusquam dicere incipiam pauca . . . explicanda sunt. The
transition here seems awkward and abrupt. F. has not yet mentioned (except
in the prologue: ¸..n.) that he will discuss moduli, and the previous sentence
leads one to expect that he will turn directly to details of copia and erogatio. His
problem, of course, is that he cannot do so until he defines his terms. This
whole section seems awkwardly written: note, e.g., that the word ratio is used
three times in close succession (but see ¸:.:n. ratio fistularum).
quinariarum centenari<ar>umque Only two pipes are named
here, the quinaria because it was the standard (.¸.:, .6..) and the centenaria
because it was in common use for distributory mains (¸..6). Except for the
centenum vicenum, all pipe-names are feminine, in origin adjectives of fistula.
{et} indicare . . . significet Announces chapters .¡–¸o; the regula
proposita, included in this discussion, is summarised again at ¸:.:–¸, where F.
points out that mathematical logic accords with the auctoritas of the imperial
commentarii. The et is uncharacteristic and awkward, and I have found no good
reason to keep it. The intrusion may reflect a confusion between abbreviations
for in and et (see also below).
COMMENTARY .¡.:–.¡..
quae vires Poleni: ‘Virium nomine Frontinus hoc loco intelligit capac-
itatem modulorum.’ But perhaps better vis ‘characteristic’ as F. uses it at Str. horum (sc. stratagems) propria vis in arte sollertiaque posita.
ratio eorum initur et computatur I accept initur, which is regular
with ratio (OLD s.v. 8), nor can I guess how to compute an initium. Note ¸:.:
ratio fistularum . . . per omnes modulos ita se habet . . . et omni genere inita constat sibi
(cf. initur :q..n., 66.¡, 6¸.¸), and for the confusion of initus and initium see TLL
¸: :6¸¸.¸¸.
qua ratione discrepantia invenerim Ratio (.¸.¸, .8.¸, .q.:, etc.)
discovers fraud (¸:.¸–¸¡.¸) on the part of the aquarii (q.6n.); cf. De Laine (:qq¸)
:.¸. F.’s use of the first-person singular is precise and significant. He employs it
as author of the treatise (e.g. ... potui . . . contuli, ¸.: videar . . . ponam, ::.: perspicio,
:¸.: mihi, .¸.: mihi, ¸¸.: subieci, 6¡.: ponam), more or less interchangeably with
the plural (e.g. ¸.¡ legimus, .¸.. proferamus, .6.. dinoscamus, ¸¸ memineramus, ¸¸.:
locuti sumus, ¸¸.¡ ponemus, ::q.: exposuimus . . . sicut promiseram divertemus). When
he speaks of official actions he consistently uses the plural, to emphasise that he
performs his delegatum officium only as agent of the princeps. The singular forms
here belong to a distinct category, that of the rational and attentive official
who assumes responsibility for the accuracy of his data and the conclusions
drawn: the practice is more clearly illustrated in chapters 6¡–¸¸ (see 6¡.:n.
and 6¡.¸n.).
quamemendandi viamsimsecutus In ¸:.¡–¸¡.¸ F. reveals that the
aquarii have introduced irregularities in the case of four pipes. The correction
of which he speaks is a matter of forbidding the use of non-standard pipes or
recomputing the capacity of those that remain in use (¸¸..).
zq.r Aquarum moduli . . . instituti sunt Baltar Veloso (:q8¸) would
insert <olim> after moduli to provide a temporal adverb to balance adhuc. The
perfect tense here is sufficient, accounting for the survival of both calibres.
in Campania . . . in Apulia Discrepancies in nomenclature, measure-
ments and standards are noted for their importance in gromatical writers,
e.g. the Dalmatian versus recalculated to iugera (p.88...–¸. Campbell), or the
Ptolemaic foot and medimnon (p.qo.:–¸ Campbell). The emendation Apulia was
made independently by Ursinus (noted by Holste: see Introd. ¸¸ n.:6¡) and
Scaliger (reported by Poleni).
adhuc observa<n>tur adhuc > citahuc C, perhaps with initial open a
misread as ci, t/d confusion, and the second a originally a correction. I see little
merit to ita hoc: Dilke (:q8¸b) .¸¸. Plural nominatives digiti and unciae require
plural verb.
zq.z digitus : digit = :/:6 pes = :.8¸ cm. F. consistently reckons in
digits: see .6.¸, ¸q–6¸. It seems that for pipe-sizes at least the sextadecimal
system had supplanted the duodecimal.
COMMENTARY .¡.¸–.¸.:
zq.j duplex observatio Simplex is plainly impossible. It is easy enough
to insert a non, but the negative nowhere fits smoothly. To negate est is awkward,
for the verb is understood positively in the first clause. B¨ ucheler negated simplex,
but the two clauses ought to be more closely parallel. The diversitas of the foot
is its division into digiti or unciae, and there are two kinds of digitus (quadratus,
rotundus; cf. .¸.: ab alterutro digitorum).
zq.j maior . . . minor Defines the relationship between the area of a
square and the largest circle that can be inscribed in it. Presumably based on
the Archimedean system (Circ. .). The square digit is larger than the round
by ¸/:¡ (..:¡), so the area of the round digit is .¸86; and ¸/:: (..¸¸) of .¸86
equals ..:¡. And since the area of a circle with a diameter of : digit (cf.
.6.¸) is t/¡, F.’s figure for the ratio of circumference to diameter is equal to
../¸ (¸.:¡.8¸¸) – the upper limit as defined by Archimedes, and that used
as a satisfactory approximation in the works of Hero of Alexandria (fl. mid-
first century cr). Maher–Makowski (.oo:) ¸86–¸ remark that in defining the
sizes of these pipes relative to each other F. ‘used, so far as we know, a truly
general type of fraction for the first time in Roman literature’. They express
F.’s statement in mathematical terms: ‘a
− ¸a
/:¡ = t a
/¡, where a is the
side of the square and diameter of the circle, one digitus, and t a
/¡ + ¸t
/¡¡ = a
’. Solving these equations for t yields t = ../¸. To perform
computations which involve t (see ¸q–6¸n.) F. may have used ¸ ¡:/.88 ( =
¸.:¡.¡): see Blackman–Hodge (.oo:) :..
anguli detrahuntur ‘The corners are taken away.’ C’s reading has
always been interpreted as a simple slip for deteruntur, but it is in fact nearly if
not in fact equally close to detrahuntur, conjectured by Lanciani (:88:) ¸6¸. The
verb detrahere is the standard arithmetical term for subtraction (TLL s.v.). In-
teresting, but irrelevant, is the similarity to AL 6¸oR (Petronius fr..q B¨ ucheler
& M¨ uller) line ¡: (turris) detritis procul angulis rotatur (for the commonplace of the
tower: Lucr. i\.¸¸¸–6¸); see Courtney (:qq:) 6¸.
zj.r ut quidam putant . . . ut alii Hardly a matter of research on F.’s
part, but rather anindicationthat he had made inquiries of his contemporaries,
e.g. members of the staff, persons associated with pipe manufacturing.
per Vitruviumarchitectum Vitruvius as auctor (withplumbarii applying
the scheme he had devised) comes close to representing Vitruvius in a highly
placed role, more directly and officially influential than is usually supposed;
cf. Purcell (:q8¸) :¸6.
exclusis prioribus . . . quinariae The new standard replaced what
was perhaps a confused assortment of earlier adjutages: those based on the
three types specified in .¡ (uncia, digitus quadratus, digitus rotundus) and per-
haps others as well. Adoption of the quinaria standard plainly occurred in the
latter half of the first century ncr. Its use became official as part of Augustus’
reorganisation in :: ncr (qq.¡ modulos . . . constituit).
in usum urbis In the entire discussion that follows F. deals only with
fistular patterns and standards at Rome. There is no evidence at present ad-
equately to determine to what extent, if at all, urban standards were applied
zj.z qui autem Agrippam If Agrippa did not ‘introduce’ the quinaria
standard, its use in his projects from ¸¸ onwards must have gained for it a
certain currency. Old-fashioned pipes were very small, because the supply in
those days was itself small (cf. q8.. iam copia permittente under Agrippa). F. does
not explicitly refute this explanation, but he might well have suspected the
implication that such pipes had been used widely enough to make them an
unofficial standard (cf. q¡.¸–6). If five of the antique adjutages approximated a
quinaria in area of cross-section, each would have had a diameter of ¸/:. uncia
(roughly one cm).
dicunt quod . . . coacti sint The subjunctive is used in the quod clause
because of implied indirect statement; cf. below ab eo quod plumbea lammina . . .
modulum efficiat.
velut puncta Despite F.’s use of puncta for (small?) illicit pipes of his own
day (::¸.:n.), it is not at all clear whether he refers to the same thing here.
(If so, why would he not take this opportunity, even in passing, to condemn
qui Vitruvium et plumbarios Vitruvius \iii.6.¡ lists the weights of
ten pipes and explains their nomenclature: e latitudine autem lamnarum, quot
digitos habuerint antequam in rotunditatem flectantur, magnitudinem ita nomina concepe-
runt fistulae; namque quae lamna fuerit digitorum quinquaginta, cum fistula perficietur
ex ea lamna vocabitur quinquagenaria similiterque reliqua. Pliny, HN xxxi.¸8 follows
the same scheme: denaria appellatur cuius lamnae latitudo antequam curvetur digi-
torum decem est, dimidioque eius quinaria. F. dismisses this explanation with his
remark in §¸: the shaping of a lead sheet will reduce the inner circumfer-
ence and increase the outer one. Yet on the surface the Vitruvian nomen-
clature is simpler than F.’s definition (with pipes named for the number of
quarter digits of diameter); it accounts for methods of manufacturing, and
it allows one to ignore the irregularities of determining the ‘diameter’ of an
ovoid opening (on which see Bruun (:qq:) ¸¸). The two systems may some-
how be related to the distinction between smaller and larger sizes (below
.8–¸o). F.’s concern is not so much with pipes as with the bronze calices
(¸6.¸–¸) used to gauge official grants (:o¸.¡). For the construction of lead
pipes, their shapes and their seams, see Fahlbusch (:q8.); Cochet and Hansen
COMMENTARY .6.:–.6.¸
z6.r Omnis . . . modulus colligitur Cf. ¸:.: Ratio fistularum . . . per
omnes modulos . . . omni genere inita constat sibi. For colligere ‘reckon’ see OLD s.v. :.,
TLL ¸: :6:q.6¡.
capacitas Cf. capit (§§¸–¸, ¸q–6¸), ¸¸.. quot capient; see ¸.¡n. Capacity (of
pipe or conduit) is the means by which ‘quantity’ (modus, hence modulus ‘pipe’)
is expressed. Capacity is equivalent to a cross-sectional area of the conduit
in which water flows (cf. 6¸.¸), in the case of pipes the area of the lumen or
opening (.q.:). What F. means in this sentence is that the capacity of any
given pipe can be reckoned (and expressed) mathematically in terms of one
variable only: diameter, or circumference, or area. He goes on (§.) to explain
that ‘the quinaria’ (a unit of capacity) is a convenient standard, one that can and
should be used consistently for both gaugings for the aqueducts (6¸–¸¸) and
the accounts of distribution (¸8–86). See further .q.:n. area id est luminis, also
Appendix C.
z6.z Differentiam . . . dinoscamus F. defines the obsolete adjutages
relative to the standard quinaria (see Table ¸). For the ut clause with comparative
cf. :¸.¸ ut facilius dinoscerentur.
substantia quinariae ‘The standard of the quinaria’, or ‘the quinaria as
a standard’, i.e. the basic unit (that which ‘stands under’) of capacity – defined
in chapter ¸q fistula quinaria . . . capit quinariam unam. Substantia is not elsewhere
attested in this sense (OLD s.v. ¸, which mysteriously attributes to this passage
the additional anomaly of an ablative form in -e).
quinariae The fistula quinaria, definedat .¸.¡ a diametro quinque quadrantum.
A diameter of five quarter-digits = circumference in digits of ¸ + ::/:. +
¸/.88 (Poleni’s emendation: see ¸qn.) = capacity : quinaria. F. does not express
the quinaria in terms of square digits, but he would have reckoned it as follows:
(../¸) × ([:/.] × ¸/¡)
= :...8 (or :...¸ using t= ¸.:¡:¸q).
z6.j unciae . . . trientem digiti Expresses the diameter of uncia-pipe
(:/:. pes) in digits (:/:6 pes): (: + :/¸ digits) ×(:/:6 digits/foot) = :/:. foot.
capit plus quam <quinaria> Despite the simple correction, the
printed version is a trifle awkward (even as a rough calculation): what F. wants
to say is ‘its capacity is greater than a quinaria by more than one-eighth of a
quinaria’. A quinariae octava (sc. pars) is one-eighth (a sescuncia) of a quinaria. F. can
relate the area of one circle to another, or one pipe to another, without using
t: he simply divides the square of the radius of one by the square of the radius
of the other, in this case the uncia-pipe (radius ./¸ dig.) and the quinaria-pipe
(radius ¸/8 dig.). (¡/q) / (.¸/6¡) = .¸6/..¸ = : ¸:/..¸ = :.¸¸¸¸¸¸+,
which F. first calls ‘more than :/8’, then expresses more precisely as l/8 +
¸/.88 + (./¸ × :/.88) = :.¸¸¸¸.. Maher–Makowski (.oo:) ¸8¸–8 observe
that F.’s result is ‘astonishingly accurate – to less than four parts in one hundred
thousand’, even though we do not know how he performed his calculation.
COMMENTARY .6.¡–.¸..
In my calculations (Table ¸) I have expanded F.’s digits to scripula (: ∋ =
:/.88 dig.) because the scripulum is the lowest denominator F. uses in chapters
¸q–6¸. Here, for instance, the same two pipes relate as ¡¡6¡ ∋/¸.¡oo ∋ , which,
following F., I reduce to his same fractions (such astonishing accuracy in the
transmission of fractions is by no means, however, consistently to be found in
z6.q digitus quadratus in rotundum redactus A ‘square digit re-
shaped to a circle’ means a circle with an area of one square digit. We do not
know whether F. solved this problem by deriving diameter from area, which
involves extracting a square root in the formula d = .

(a/t), or by relating
the area of one pipe to that of the other (as we saw him do above .¡.¸n.). His
diameter of : digit +: :/. twelfths of a digit +:/¸. (:.¸888+) would give us
by modern calculations :.:.8: using t = ../¸, or :.:¸8q with t = ¸.:¡:¸q.
F.’s ratio for the area of the quinaria to the dig. qdr. is :o/:. (.8¸¸¸): Maher–
Makowski (.oo:) ¸88–q. A modern reckoning of the area of the quinaria in
square digits using the two figures for t is given above (§.n. quinariae), and
from these we arrive at the ratio quinaria/dig.qdr. of .8:¡ and .8:¸ respectively.
By reckoning with scripula quadrata and ignoring t I achieve the proportion
.68q6/¸.¡oo or .8¸o: (Table ¸).
z6.j digitus rotundus The area of F.’s digitus rotundus is t/¡ square
digits, expressed as a decimal .¸86 (or .¸8¸ with t = ¸.:¡:¸q). This relates to
the quinaria as .o¸¸6t/¸.¡oot (Table ¸) or .6¡, whereas F.’s fractions (¸/:. +
:/.¡ + :/¸.) represent .6¸8+: Maher–Makowski (.oo:) ¸8q.
z¡.z est unum cum ipsa<e> multiplicantur The combined emen-
dation est unum is virtually assured, as it comes between duobus generibus (§:) and
alterum genus est (.8.:). Most since B¨ ucheler have accepted it, although Krohn
prints C’s et una followed by a lacuna. Yet B¨ ucheler, and even Krohn, accepted
the vulgate singular multiplicatur in preference to C’s plural: they did so taking
quinaria as antecedent of C’s ipsa. Grimal justified multiplicatur on the ‘authority’
of MV and kept et una as well, construing apparently una cum ipsa (ablative); he
translates, however, ‘le premier, lorsque la quinaria elle-mˆ eme est multipli´ ee’.
Pace takes moduli to be subject of the transmitted multiplicantur (‘e divengono
multipli della quinaria stessa’). This is comfortable only so long as we keep
et una. Once that has been relinquished, cum loses its prepositional option and
we expect it to be followed by a nominative. Better to change ipsa to ipsae, keep-
ing C’s plural verb, to match plures quinariae includuntur and adiectionemquinariarum
(cf. §¸ excipiuntur and ¸¡.¡ quinariarum multiplicatio).
eodem lumine For lumen the ‘orifice’ of the pipe, see below .q.:n.
in quibus ‘In the case of these pipes’. Antecedent is moduli (§:) as rightly
translated by Pace (‘in questi moduli’); cf. ¸:.¡ <in>pluribus consentiant, in quattuor
COMMENTARY .¸.¸–.q.:
modulis novaverunt. For a neuter indefinite (‘in which case’, Bennett and other
translators) Latin would require a singular (e.g. in quo).
z¡.j plures quinariae impetratae Pipes of which the capacity is a
multiple of the quinaria were usedonly toleadfromanaqueduct or mainconduit
to a castellum privatum (¸..n., :o¸.¡n.). The singuli are private individuals (cf.
q¡..n.). For impetrare / impetratio, the formal process whereby a privatus receives
water as a beneficium Caesaris (¸..n.), see :o¸.:–¡.
ne rivus . . . convulneretur The same verb at ::¸.¸ (fistulas) convulneratas
(cf. :o6.¸ ne aut rivi aut fistulae publicae . . . lacerentur, ::¡.¸ [castellum] foratum vitiatur).
The compound is rare and post-Augustan: there seems no intensifying force
to the prefix here although there may be at Str. ii.¸.¸: (cf. TLL ¡: 8qo.¸.).
z8.r alterum genus est The other manner of increasing pipe-sizes in-
volved two methods, both of which are based on the area. First is the scheme of
adding quarter-digits to the diameter (.8..–¸): this applies to pipes up through
the vicenaria (¡o–6). For larger pipes the increases are by area of square digits
(.q.:–.), used from the vicenum quinum to the centenum vicenum (¡¸–6¸).
ad quinariarum necessitatem ‘According to the necessity (for a
requisite number) of quinariae’ (.¸..–¸); for this sense of necessitas see OLD
s.v. 6.
z8.z nec iam in solidum Added for clarity. A senaria, larger than a
quinaria by only :/¡ digit in diameter, does not increase the capacity in solidum
(OLD s.v. solidus qb), i.e. by a whole quinaria. On the figures for the capacity of
the senaria see ¡on.
z8.j et deinceps . . . quadrantibus diametro adiectis Although F.
does not make the point explictly, linear increments to the diameter increase
the capacity quadratically: Blackman–Hodge (.oo:) ::–:..
zq.r area, id est lumine Although F. has already used both area (.6.:)
andlumen (.¸.., twice) without definition, the explanatory parenthesis is not out
of place. Latin has no word for ‘cross-section’. While area was the geometrical
term(cf. ¸o.., 6¸.¸, Colum. \.:.¡ modus omnis areae pedali mensura conprehenditur qui
est digitorum XVI, and see OLD s.v. 6, TLL .: ¡qq..¸), lumen occurs in technical
contexts involving pipes (¸6.¡, :o¸.¡, Vitr. \iii.6.¡ lumen fistularum: see OLD s.v.
8b, TLL ¸..: :8:¡..¸); cf. Callebat (:q¸¸) ¸.¸. Neither sermonis egestas, however,
nor F.’s desire for clarity requires the same definition in the following sentence
(§.); thus I delete id est luminis as an intrusive notation – along with its participial
phrase in rotundum coacti. Degering (:qo.) .¸ brought an improvement of sorts
with coactos (the twenty-five square digits are in circular shape), but moduli or
fistulae are by F.’s definition round (.¸.., .6.:) and redactus (cf. .6.¡) would have
been expected rather than coactus (‘combined’ at .¸..).
COMMENTARY .q..–¸:..
zq.z per incrementum C originally had incrementor(um) but the same
hand has added u above -or(um), apparently an instantaneous correction (the
original perhaps anticipating digitorum quadratorum). B¨ ucheler’s emendation is
unnecessary, although it would bring the construction close to .¸.¸ et dein-
ceps simili incremento usque ad vicenariam (cf. adiectione singulorum quadrantum). The
amount of the increment can be understood from the nomenclature (quinorum,
incidentally, would be inapplicable beyond :oo).
jo in vicenaria fistula . . . exiguo minus viginti F. ends with the
vicenaria not only because it stands on the borderline, but because discussing it
here permits him unobtrusively to recapitulate both systems of nomenclature.
It is no accident that he ends with exiguo minus viginti, because it is by pointing to
an illicit diminution of the vicenaria, considerably more than could be described
as exiguo minus, that he begins his demonstration of malfeasance on the part of
the aquarii (¸:–¡).
jo.z in <an>tecedentibus modulis One could perhaps emend to
(dative) antecedentibus without the preposition, but cf. ¸¡.¡ in amplioribus modulis
servare . . . regulam debet.
exiguo minus viginti The capacity of the vicenaria according to the
method of gauging smaller pipes (see ¡6.:n.) nearly coincides with the reck-
oning for larger ones, its area being just under .o square digits: t(:o/¡)
jr Here begin F.’s comments which make clear why he has spent so long
defining the principles of mathematics and nomenclature that govern pipe
sizes. Only now does he bring an administrative concern to the forefront, for
he wants to focus sharply on the unmethodical (and, to his mind, unprincipled)
practices of the aquarii (q.6n.). To this section (¸:–¸¡.¸) in particular remarks
on F.’s rhetoric are especially apt: DeLaine (:qq¸) :.6–¸. He begins and ends
with an emphasis on ratio (¸:.:) and regula (¸¡.¡) as he moves from accusation
(¸:.¡ sed aquarii), through demonstration (¸¸.. cum ratione . . . re quoque ipsa), to
condemnation (¸¡.¸ fraudem, the last word).
jr.r omni genere inita constat sibi The ratio fistularum which F. has
just completed, we learn, has been the first of two strands of argument which
he brings to support his charge of fraus. (Note F.’s stress on ratio at ...¸.)
jr.z convenit et cumis modulis qui . . . confirmati sunt The sec-
ond strand of F.’s argument is the auctoritas of the imperial registry, wherein are
set down (positi: see ¸.:n. ponam) exactly the same calibrations he has patiently
explained on mathematical grounds. With confirmati sunt cf. .¸.. modulorum
per quos mensura constituta est and the more pointedly legalistic qq.¡ (Augustus)
modulos . . . constituit.
COMMENTARY ¸:.¡–¸...
commentariis . . . principis The imperial commentarii to which F.
refers are a set of records which were begun under Agrippa (q8..) and formed
an important part of the Augustan reorganisation in :: ncr (qq.¸). These
contained data both for the available supply (¸¸.¸, 6¡.:) and for the water
delivered (6¡.:, q8.., qq.¸); on these data F. draws for chapters 6¸–¸¸ and
¸8–86 respectively. And, as the present passage makes clear, they contained
a record of standard pipe-sizes (.¸.:n., qq.¡). The style a commentariis aquarum
appears in CIL 6.8¡8¸ (ILS :6oq), but details of record-keeping are unknown.
For the existence of similar imperial registers and their authority, see Sic. Flacc.
Cond. agr., ed. Campbell (.ooo) p.:.o, .8–¸. illa tantum fides videatur, quae aereis
tabulis manifestata est. quod si quis contra dicat, sanctuarium Caesaris respici solet. omnium
enim agrorum et divisorum et assignatorum formas, sed et divisionum et <assignationum>
commentarios, et principatus in sanctuario habet. qualescumque enim formae fuerint, si
ambigatur de earum fide, ad sanctuarium principis revertendum erit.
{invictissimi et piissimi} The superlative epithets are both out of
place, pace Baldwin (:qq¡) ¸o6. They are anachronistic as well (although Pliny,
Pan. 8.. speaks of Trajan quite plainly as invictus imperator), and neither be-
came standard in imperial titulature until late in the second century: TLL
¸..: :86.¡6; OLD s.v. pius ¡. On the other hand, they are entirely character-
istic of Peter the Deacon. Note especially Peter’s interpolations in his copy of
Vegetius (Introd. ¡¡): i..8.. ipsos <invictissimos atque excellentissimos> progenuere
Romanos, imperator invict<issim>e, i\.pr.¡ imperator <excellentissime>. With-
out the ‘improvements’ we have chiastic alliteration: commentariis principis positi
et confirmati.
jr.q in quattuor modulis novaverunt F. here uses novare intransitively
with in +ablative, ‘made changes in the case of four pipes’. Contrast ¸¡.. and
¸¸.: with accusative (quos aquarii novaverunt, -ant).
jz.r in duodenaria quidem Giocondo’s genitive is convenient for error
and usus, but in + ablative ‘in the case of’ follows more easily from ¸:.¡ (cf.
.¸.¸, ¸o.:, ¸¸.:, ¸¡.¡). Omission of in occurs at ¸:.¡ and ¸..¸; here perhaps
C’s et is a corruption of in.
jz.z diametro adiecerunt The addition of a semuncia and a sicilicum
gives the unofficial pipe a diameter of ¸¸/¡8 digits. This essentially squares
with figures transmitted at ¡¡...
capacitati quinariae †ebesem† The capacity of the official pipe is
:866.¡t scripula quadrata or ¸.¸6 quinariae (see Table ¡). With a larger diameter
the unofficial pipe has :q¡¡8:tscr. quadr. or 6.oo.¸ quin. (Table ¡). The capacity
has thus been increased by ..¡.¸ quinaria; hence Poleni’s quinariae quadrantem.
It is probable, however, that F.’s figure was more precise (note the fraction of a
scripulum at .6.¸). If F. reckoned the larger pipe at 6 quin. (¡¡..) the difference
COMMENTARY ¸..¡–¸¸.:
becomes ..¡, and he may have written capacitati quinariae <sextantem semunciam
duellam scripulum et scripuli> besem (:/6 + :/.¡ + :/¸6 + :/.88 + ./86¡ =
o..¡.). Tannery (:88q) ¸:–. proposes <sextantem semunciam sicilium> et besem,
which comes to ..¸..
jz.q vicenariam exiguiorem With a diameter of ¸ digits the vicenaria
will have a capacity of ¸:8¡oot scr. quadr. or :6 quinariae. Reducing the dia-
meter by :/. digit will give a capacity of ¡:qqo¡t scr. quadr. or :..q6 quinariae.
The difference (¸.o¡) closely approximates F.’s ¸ + :/.¡ (¸.o¡:66). Modern
writers would probably express the reduction as :8.¸¸ per cent. – The com-
parative of exiguus is very rare, only three instances: Dig. xxix.¸.:..¸ exiguioris
vocis, xxxii.:.¸¸.pr. exiguiores nummi, xriii.:..:.:¸ statio exiguior facta.
jz.j–6 plerumque erogant . . . adsidue accipiunt The subject is
aquarii (as consistently from ¸:.¡–¸¸.¸) and the verbs should be active. The ad-
verbs indicate that these pipes were in regular use and for water mains (.o..n.).
For the number XX appearing on surviving lead pipes and the uncertainties
of identification with F.’s vicenaria see Bruun (:qq:) ¸¡.
jz.¡ centenariae adiciunt According to data in chapter 6. (Table ¡),
the difference between the diameter of the standard pipe (:: + 8:/.88 or
∋ ¸.¡q) and that of the unofficial one (:. or ∋ ¸¡¸6) would be exactly .o¸
scripula, whereas F.’s ./¸ (besem) + :/.¡ (semunciam) comes to .o¡ scripula, no
worrisome inaccuracy. The capacity of the unofficial pipe was increased by
:o + .o¡/.88 quinariae (Table ¡), the fractional part of which coincidentally
perhaps reduces to the same two fractions, ./¸ (besem) +:/.¡ (semunciam). The
transmitted one-half (semissem) is far too low (perhaps a misinterpretation of
symbols or a confusion with the identical fraction earlier in the same sentence).
In modern terms the increase in capacity amounts to :o ∋ .o¡ / 8: ∋ :¸o =
∋ ¸o8¡ / ∋ .¸¸.8 = :¸... per cent.
jz.8 centenum vicenum . . . adiciunt In chapter 6¸ the official pipe
has a diameter of :. + :o./.88 or ∋ ¸¸¸8. To this has been added ¸ + ¸/:.
+ :/.¡ + :/¡8 (or ∋ :o¸o) giving the unofficial pipe a diameter of ∋ ¡6o8,
an exact :6 digits. Thus modified, the :.o-pipe has a capacity of :6¸ ::/:.
quinariae, larger by 66 :/6 than the official standard of q¸ ¸/¡. Calculate the
increase as 66.:¸ / q¸.¸¸ = 6¸.6q per cent.
jj.r intercipiuntur . . . octoginta sex F. bases this reckoning on the
fact that according to the official standard (¡6.:) the capacity of five .o-pipes
of :6 quinariae each (total 8o quinariae) approximates that of the :oo-pipe (8:
:¸o/.88), and six .o-pipes (q6 quin.) that of the :.o-pipe (q¸ ¸/¡ quin.). By
diminishing the size of the .o-pipe (used for delivery) and increasing the size
of the :oo- and :.o-pipes (used for receiving), the watermen have ‘stolen’ from
COMMENTARY ¸¸..–¸¡.¸
official capacities at both ends. As F. explains in §¸, from their five smaller
.o-pipes they deliver :¸ quinariae each (total 6¸), although they have received
in an enlarged :oo-pipe not the official 8:.¸ but a total of slightly more than
q. quinariae. The difference between q. and 6¸ represents their gain of .¸
quinariae. Likewise, from the enlarged :.o-pipe they receive approximately
:6¡ quinariae, but deliver from six .o-pipes only ¸8, gaining in this case 86
jj.z ratione . . . re quoque ipsa On these two strands of F.’s argument
see ¸:.:nn.
jj.j aeque certum est The transmitted aeque certum est brings ex
centenaria . . . ad artiorem numerum into balance with ex vicenaria . . . non plus
quam tredecim. Schultz fretted that this text did not mention the centenum vicenum
alongside the centenaria in the first part of this sentence, and B¨ ucheler’s solution
eque <centenum vicenum> has prevailed despite the unusually strange form eque
(preposition + enclitic). It is altogether easier to defer until the parenthesis
item . . . nonaginta octo any mention of the :.o-pipe.
jq.z consentiunt et rationi et commentariis See ¸:.:nn. on the two
strands of F.’s argument.
jq.j omnia . . . quae mensura continentur Mensura comprehendere
and mensura continere are technical terms (OLD s.v. comprehendo ¸, s.v. contineo 8),
essentially synonymous and meaning ‘embrace’ or ‘enclose’. Cf., e.g., the open-
ing sentence of F.’s De agrorum qualitate (p.: Thulin, p.. Campbell): Agrorum
qualitates sunt tres: una agri divisi et adsignati, altera mensura per extremitatem conpre-
hensi, tertia arcifini qui nulla mensura continetur. An instructive discussion is that of
Hinrichs (:qq.).
jq.q sextarii ratio . . . ad cyathos respondet One might keep the
spelling cyatus (cf. OLD s.v.), though C’s authority is worthless. A cyathus was
:/:. of a sextarius, and a sextarius was :/:6 of a modius. One modius (c. 8.¸ litres) =
:6 sextarii = :q. cyathi: Dilke (:q8q) .¸. The scheme for dry measurements
would have been familiar to any class of Roman society, but F. has perhaps in
mind the strictly methodical measurements expected of the officials in charge
of the annona (see below, ¸q–6¸n.).
jq.j in erogatorio modulo . . . in accepto Adjective erogatorius (TLL
¸..: ¸qq.¡¸) is a ctcç, as is also the long accepted emendation acceptorio (sc.
modulo). F., however, would likely have varied the construction in his ‘mini-
peroration’. I take accepto here as ‘on the credit side’, a substantive use of
neuter participle as at 6¡.. plus in distributione quam in accepto computabatur. For
<ve>ro adversative adverb in second position cf. .:.¸, ¸¡.¡, ¸¸, q:.¸, q¸.¡, q¸.¸,
::..¡, ::8.¸, :.¡.¸, :¸o.¸.
non errorem esse sed fraudem The emphatic conclusion follows
from the two-fold argument based on ratio and auctoritas (¸:.:–¸). While F. was
no doubt justified in accusing the watermen of fraus (cf. q.6n., ¸¸.., ::o..), it
seems at least possible that there was error as well. Modifications to the duodenaria
and the vicenaria produce pipes with an exact number of quinariae (6 and :¸
respectively), and they might thus originally have been used to deliver multiples
of the quinaria (see .¸..–¸). With the larger unofficial pipes, the diameters are
even digits (:. and :6 respectively), which suggests a deliberate difference in
manufacturing. Note that the adapted :.o-pipe is nearly equivalent to two of
the official :oo-pipes (6¸..). Evenif modifications had beenauthorised, the new
pipes might not have had new names and the changes might not have been
reflected in the records. Bruun (:qq:) ¸¸ nn.¸¡–¸ comments upon certain
possibilities for honest confusion, and at ¸¸–8 he calls attention to possible
differences on the part of manufacturers. F. could have downplayed any such
elements for rhetorical effect.
jj–j6 Memineramus . . . coartari It might have been more natural
for the text to continue directly from the end of chapter ¸¡ to the list of pipes
in chapter ¸¸ (cf. ¸¸.:–.). This could result either from a situation where these
two chapters were a later insertion, or fromone in which the preceding passage
had been reworked for rhetorical effect. They contain, in any case, comments
of a technical nature which prepare the reader for some points F. makes later
on. Note :o¸.¡ the juxtaposition of mensurarum modum and positionis notitiam, and
especially the importance of the calix (cf. ::...–::¸.¡). Although ¸¸–¸6.. relate
to general hydraulic theory, context (especially use of modulus) suggests that F.
has in mind particularly issues that arise in distributions to privati.
jj Memineramus The indicative is entirely normal (cf. :¸.:, ¸¸.:, ¸..8,
¸¸.: etc.), while editors’ meminerimus gives an uncharacteristic hortatory sub-
junctive. The sense of memini (pluperfect with preterite force) here comes close
to ‘we observed, we came to know from observation’.
ex [altiore loc]o . . . deperdere These comments address a situation
in which water arrives at a distributory tank (castellum: ¸..n.). F. conflates two
distinct hydraulic principles at work in closed pipes, probably because the
ancients were unable to make the technical distinction. (:) Volume is affected
by the head of water (i.e. the pressure resulting from the difference of altitude
between intake and efflux): ex altiore loco, ex humiliore. (.) Resistance to flow will
depend on the length of the course: intra breve spatium, longius ducitur. For the
two together cf. Pliny, Ep. x.6:.¡ fossam longius ducere et altius pressam. See Lewis
(.ooo) ¸¡8.
respondere modulo suo . . . exuberare Here respondere seems to
mean ‘discharge in the expected quantity’, literally to make return to its
own pipe-capacity (but see 6¸.¡n. mensura respondet). Exuberare is apparently
COMMENTARY ¸6.:–¸6..
also intransitive and with its familiar meaning ‘to flow copiously, overflow’
(OLD s.v., TLL ¸..: .oq6.:, ¸¸). Examples to support taking it transitively
(with e.g. modulum understood as object) are not readily to be found: TLL
.oq6.8¸ offers nothing satisfactory. See, further, ¸¸.6n. exuperare (conjecture) . . .
in castellum The principle would be the same for all tanks, but F. prob-
ably has in mind castella privata (¸..n.). Public deliveries would very likely be
altered less often, and grants to individuals required watchful regulation (:o¸.¡,
::.–:¡) to maintain a reasonably uniform distribution and to safeguard the
supply needed for public purposes.
secundum hanc rationem ‘Accordingly’ disappoints in its vagueness
(cf. the more precise use of ratio at .8.¸, ¸o.:, ¸:.:, ¸¸.., ¸¡.¸, ¸¸.: etc.). One
would expect F. to have specified how the delivery was to be adjusted. It is
possible that something has dropped from the text after deperdere.
oneranda<m> . . . relevanda<m> Aquam is understood with the
gerundives, and the ablative erogatione is best taken as ‘in the way that delivery
is effected’ (it can hardly be modal). The metaphor in onerare and relevare is
not at all clear, although it looks to be related to the verbal sense in pressura
(:8.¸n.). Even less clear is how such a ‘burdening’ or ‘relieving’ was to be
accomplished. Certainly not by an adjustment in fees (:o¸.:n. impetrare), pace
White (:q8¡) .:¡–:¸. Perhaps somehow by means of positio (see next note),
perhaps also by making use of valves; at the present state of our understanding
speculation alone is possible: Rodgers (:qq:) :8.
j6.r sed et positio . . . exiguum sumit A further comment on hy-
draulic principles, which complements the preceding by reference to water
leaving a distributory tank (or a rivus).
positio ‘Setting’, but of what? A masculine noun, so much is clear from
conlocatus etc. Giocondo’s <calicis> has gone unquestioned, because of the ap-
pearance of that word in §¸ as well as at :o¸.¡ (with positio), ::... (with positi),
and especially ::¸.:–., where F. speaks of calices placed inequitably. Despite
the change of subject (from aqua in ¸¸), mention of a calix is still premature: F.
must first finish his description of hydraulic circumstances in which the calix
could be a useful regulatory device. Better would be <moduli>, which follows
naturally from ¸¸ modulo suo. But that word is not essential here, either, and
awkwardness is considerably lessened by not starting a new paragraph with
sed et positio. When F. mentions the calix in §¸, moreover, he introduces it as a
special kind of modulus.
j6.z in rectum et ad libram conlocatus Poleni’s explanation and
illustration of this passage are exemplary. All references are to the water flow
(ad cursumaquae). In rectum, seenas it were fromabove (inplan), qo

to direction of flow; ad libram, seen in section, qo

perpendicular to height/level
COMMENTARY ¸6.¸–¸6.¡
of flow. If angles are not exactly qo

, the terms are respectively:
obpositus >qo

, pointing downstream
amplius rapit
devexus >qo

, pointing downwards

conversus <qo

, pointing upstream
exiguum sumit
supinus <qo

, pointing upwards

A good diagram in Landels (:q¸8) ¸o ( = Hodge (:qq.) .q8).
ad libram Here and at :.¡.¸ the phrase means ‘on a level’, ‘all’altezza’:
Panciera (:q¸8).
ad cursumaquae Despite the rule ‘only froma castellum’ (S.C. of :o6..),
F. speaks here of pipes placed not only in tanks but also in channels with free
flow; cf. §¸ rivo vel castello (but see also ::¸..n. cursus aquae). Bruun (:qq:) ::8–.¸
discusses the evidence for such channels in urban distribution alongside fistulae;
cf. Hodge (:qq.) .q8, ¡6: n.¸6.
ad haustum pronior If the phrase is not a gloss, Krohn’s transposi-
tion is essential, for pronior cannot possibly mean ‘less favourably positioned’
(Herschel / Bennett).
j6.j est autem calix ‘There exists, moreover, the calix.’ F. turns to an
existing device that addresses to some extent potential irregularities of delivery
that he has just outlined. Purpose and use of these ‘delivery necks’ is not in
doubt, but the origin of the name is unknown. One thinks naturally of some
connexion with calix drinking cup, e.g. a similarity of shape if the pipes were
flanged: Pace (:q8¸) ¸¸–8, Hodge (:qq.) .q¸. No surviving pipes conforming
to F.’s description have been found in Rome, and therefore the device remains
a puzzle. Although F. is not himself responsible for their existence (§¸ excogitatus
videtur, ::... calices . . . inveni), his interest in them may be in part prescriptive:
Bruun (:qq:) ¡:–¡. Any relationship with the calix (as transmitted) at :.q.:: is
rivo vel castello inditur Indere +dative ‘insert’: TLL ¸.:: :.6..¸.. The
transmitted induitur can hardly be defended, pace Dilke (:q8¸b) .¸¸ (comparing
Colum. \iii.::.¡), especially since indo, induo, and induco are so readily confused.
For inducitur see Fea (:8¸.) ..¸, ¸¡¸.
huic fistulae adplicantur Cf. ::¸.¸ in quorumdam [sc. privatorum] fistulis
ne calices quidempositi fuerunt. The calix, inother words, derives its true importance
as a technical device for regulating distribution to privati.
j6.q longitudinis . . . impe<t>rata fuerit A calix is to be at least :.
digits in length, its area of cross-section depending on the grant (cf. :o¸.¡–¸,
::.). Subject of habere debet is calix, with digitos longitudinis (genitive of definition:
COMMENTARY ¸6.¡–¸¸..
.6.¸–6, .q.. etc.) and lumen (id est capacitate<m>) as direct objects. For
impe<t>rata cf., e.g., .¸.¸, :o¸.:–¸; the same error occurs at :.8... On the
formality of impetratio see :o¸.:n.
lumen id est capacitate<m> Ablatives are impossible, unless one
reads quantum impetratum: cf. Rodgers (:q8¸) :¸¡, Ehlers (:q8¸) 86. Area and
lumen are virtually synonymous (the latter used in the sense of ‘orifice’): see
.q.:n. For capacitas, expressed in terms of the cross-sectional area/lumen of a
pipe, see .6.:n. For modus ‘quantity’ see ¸..n. pro suo modo, and for modulus
‘calibrated pipe’ see ¸..n. modulorumque rationes.
j6.j rigor aeris For other uses of rigor +genitive of metal cf. Lucr. i.¡q.
rigor auri solvitur aestu, tum glacies aeris flamma devincta liquescit, Virgil, G. i.:¡¸ ferri
rigor (also Manil. ii.¸8o; cf. Sil. Pun. i\.6¸o). On the strength and durability of
bronze, e.g. Hor. C. iii.¸o.: monumentum aere perennius, Ov. Tr. i.¸.¸¸ pectus mihi
firmius aere. For differing proportions of the alloy in various usages see Fassitelli
(:q¸.) .¸–¸:; for valves his studies showed copper ¸¸.¸¸ per cent, tin ¸.88 per
cent, lead :8.¸q per cent, iron o.¡o per cent.
non temere potest laxari vel coartari Cf. ::¸.¡ hae fistulae solutae
vocantur et, ut aquario libuit, laxantur vel coartantur. The adverb temere, as usual, is
used in litotes.
j¡.r subieci The word is normal for introducing verbatim quotations
(:o¸.¸n.), so here one might reasonably suppose that F. is reproducing
a list – perhaps that in the imperial registers to which he refers above
in usu quindecim tantum frequentes Not in use, and so noted in
the listing (¸q–6¸), were the septenaria, duodenaria (cf. ¸..:), and vicenum quinum,
and the remaining intermediate sizes from tricenum quinum to nonagenum quinum.
Vitruvius’ list of ten sizes (\iii.6.¡) does not include F.’s senaria, sexagenaria,
septuagenaria, nonagenaria, and centenum vicenum.
emendatis quattuor Cf. ¸¡. . exceptis his quattuor quos aquarii novaverunt.
For the ablative absolute without pronoun but followed by relative clause cf.
:o8 (S.C.) exceptis quae, ::..: explicitis quae: K–S ii.:: ¸¸¸–¡. (Similar are demon-
strative pronouns omitted in other constructions, e.g. qq.¸ uterentur [ii] qui, :o..:
subiungere [eos] qui, :¸o.. ignorarentur [ii] qui.) The four fistulae are the :.-, .o-,
:oo- and :.o-pipes (see ¸:.¡–¸¡..). F.’s ‘correction’ (cf. .¸.. quam emendandi
viam sim secutus) seems to consist of little more than calling attention to their
j¡.z fistulae omnes F. normally places the adjective omnis before the
noun (cf. : omnis res, q.¸ omnes villae, :8.¸ omnem partem, ¸:.:. omni genere, etc.), so
perhaps here its position is for emphasis. Cf. ¸6.. corruptelas denique omnes, :...¸
ante praeparatis omnibus.
ad quinarias quot capient computari In other words, the real ca-
pacity of such anomalous pipes should be determined (and officially accounted
for) rather than assumed from their nomenclature.
j8 {Ed diametri . . . sextam} I transcribe without editorial improve-
ments the text of C. As Schultz observed, the entire chapter is a repetition of
.6.¸–¸ (the initial Ed from habet). It is possible that the passage was repeated
through a copyist’s error (Rodgers (:q¸8) :o.). But context rather suggests that
it originated as a reader’s marginal note: someone noticed (perhaps after read-
ing chapter ¸¸ about fistulae omnes, even those not in use) that F. had neglected
to include the obsolete pipes he had mentioned earlier. The annotator’s mark
is apparent in the simpleminded digitus quadratus in longitudine et latitudine aequalis
jq–6j For each pipe F. gives figures for diameter, circumference, and ca-
pacity (in quinariae), the latter of which is an expression of the cross-sectional
area (cf. .6.:). The nomenclature of the smaller pipes is based on their diam-
eter (.¸.¸), that of the larger ones on the area of their cross-section (.q.:–.).
His syntax is that of earlier chapters; cf. .6.¸ digitus rotundus habet diametri
digitum unum; capit quinariae septuncem semunciam sextulam. Significantly, in the
‘tabular’ format adopted by F. here (Introd. .¡), there is consistent ellipsis of
habet (noted by Dederich); the recurrent capit, by contrast, serves for emphasis,
rather like bold-face type. The figures were probably transcribed from im-
perial records (¸:..n.). For a summary, with metric conversions, see Table ¡.
Similar tables appear in Landels (:q¸8) ¸¸ ( =Hodge (:qq.) .q¸), Fahlbusch
(:q8.) :¡:, Pace (:q8¸) 86; cf. Blackman–Hodge (.oo:) ::. For the fractions, see
Table ..
The data in these chapters might seem pedantic, save for F.’s wish to avoid
ambiguity of any sort in light of detected misuse (¸:.¡–¸¡.¸) and his sense of the
curator’s ultimate responsibility for accuracy and exactness in official grants
(:o¸.¡). One can compare, for example, the activity of mensores frumentarii, and
especially the responsibility assumed by the praefectus annonae as attested on a
bronze weight found at Ostia (NSA:q¸¸, .¡8 no.:¡ = AE:q¡o, ¸8 = AE:q¸¸,
:6¸): Imp. Caesar Nerva Traianus . . . p(ondera) fecit exacta cura M. Rutili Lupi prae(fecti)
annonae. Bruun (:qq:) ¸:–8, in his detailed study of fistulae at Rome, observes
how difficult it is to accommodate F.’s ‘official’ listing to any, let alone all or
most, of the surviving pipes. He insists that one must abandonthe straightjacket
of F.’s text to analyse fistular evidence in broader terms, e.g. of manufacturing
patterns and processes. Such mathematical accuracy as F.’s text represents can
never have been attained by manufacturers. Why, then, the list – especially
since ten of the twenty-five are not in use (¸¸.:)? With grants to privati in the
hands of the emperor (qq.¸, :o¸.:) could not the practice have been simplified?
‘One would even think’ writes Bruun (¸¸) ‘that such a table would restrict the
choice of the Emperor in the matter! Could not a practice have equally well
existed, in which the Emperor decided fromcase to case on howmuch water to
award to the applicant, or alternatively that the standard table had been more
manageable, with e.g. a “small,” a “medium” and a “large” pipe to choose
from . . .?.’ Indeed. But Romans appreciated detail and – decisive point – the
table had been given legal authority by none other than the first princeps: qq.¡
(Augustus) modulos etiam, de quibus dictum est, constituit.
Transmitted figures are at times badly corrupt, but F.’s earlier definitions
(.8–¸o) make arithmetical restoration reasonably certain. Blackman–Hodge
(.oo:) :¸ are grossly unfair when they write, ‘It is almost certain that Frontinus
did not have personally the skills to deal with almost any of the technical issues’
of arithmetical computation.
(:) In chapters ¸q–¡¸, the diameter (d) is established by definition (.8.:–¸),
and t = ../¸ (.¡.¸n.).
perimeter p = td
capacity c = a/q = :6d
a = area of cross-section, t(d/.)
; q = area of the fistula quinaria (.8.:−.),
which is t(¸/8)
. Although tis not involved, the result may occasionally differ
by : scripulum depending on how one rounds off the final fraction. Noticeable
differences result by taking t = ¸.:¡:¸q (.¡.¸n.).
(.) In chapters ¡¸–6¸, the area (a) is established by definition (.q.:–.).
diameter d = .

perimeter p = t d = .

t (a/t)
capacity c = a/q = 6¡a/.¸
Blackman–Hodge (.oo:) :. note that extraction of square roots (necessary to
reckon diameter) is a task that must have involved ‘substantial labour under-
taken by the staff of Frontinus’ predecessors’.
(¸) In chapter ¡6 the vicenaria is a borderline case, approximately the same
whether reckoned by diameter or area (¸o.:–.).
To simplify the apparatus I have sometimes noted only the correct reading
and the scholar who proposed it (e.g., in ch.¸q, S = = – ∋ III Polenus), ignoring
completely what one finds in C. The transmitted readings are primarily of
palaeographical interest, and they can in any case be studied more closely in
jq capit quinariam unam Appendix C contains Professor Bruun’s
discussion of various attempts, none successful, to determine the value of
‘a quinaria’ expressed in modern terms of volume and time.
qo capit quinariam With an area of ¡66¸6t scripula quadrata, the senaria
has :.¡¡ quinariae, or a trifle more than : :.6/.88. Poleni’s <quincuncem
sicilicum>, based on .8.., is accurate enough. There is no need for B¨ ucheler’s
<VII> here (technically closer, but the normal pattern seems to be dropping
a remainder or rounding to the next lower ∋ ), nor for Krohn’s <scripulum> at
qq.z {alia} apud aquarios Cf. ¸¸..n. Note not only the imperfect (the
pipe has been ‘retired’) but that the formula is modified from capit quinarias
(¸q–¡., etc.) to capacitatis quinarias (so also ¡6.., 6..., 6¸..). To B¨ ucheler we
owe the suppression of this alia and that below in chapter ¡¸. Here it might
just be defensible, but it smacks of intrusive notation and nowhere recurs in
similar instances below.
q6.r Fistula vicenaria Poleni adjusted the transmitted figures for the
vicenaria, making them consistent with F.’s statements at ¸o.:–. (diameter of
¸ digits, area ‘just under’ .o square digits) and ¸¸.¸ (imperial records show a
capacity of :6 quinariae); see also ¸..¡n. B¨ ucheler, on the other hand, adhered
more closely to C’s readings in this chapter and, again for consistency, made
textual adjustments at ¸..¡. At issue is whether in this tabulation the .o-pipe
belongs with the smaller pipes (.8.:–¡) or the larger ones (.q.:–.), for with
the vicenaria the two systems practically coincide (¸o.:–.).
:st system .nd system
Diameter :¡¡o scr. :¡¸¸ scr.
Perimeter ¡¸.6 scr. ¡¸6¸ scr.
Area :6.q.¸¸ scr.
:6¸888o scr.
Capacity :6 quin. :6 8¡/.88 quin.
¸:8¡oot with t=../¸
The difference (roughly . per cent) is likely to have been negligible, particu-
larly in reckoning the approximate equivalents of larger pipes in terms of the
vicenaria (see ¸¸.:n.). I followCin accepting the larger figures here, defining the
vicenaria with an area of .o square digits, but I do not think that mathematical
consistency requires adjustments in the earlier chapters.
q6.z apud aquarios Cf. ¸..¡–¸. The unofficial vicenaria had a smaller
diameter and diminished capacity (¸¸.¸). Context makes F.’s calculations rea-
sonably clear at ¸..¡. The alternatives are:
Diameter :.q6 scr. (¡ :/. dig.) :.q8 (¡ :/. + ∋ .)
Area ¡:qqo¡ t scr.
¡.:.o: t scr.
Capacity :. .¸6/.88 quin. :¸ quin.
To insist upon a difference of ∋ . (.:.8 mm.) in diameter would be pedantic.
For consistency with ¸..¡ Poleni decided that the capacity here should be XII
COMMENTARY 6...–6¡.:
deuncem semunciam; Krohn’s <duodec>im <S = = – £> is palaeographically
6z.z apud aquarios Cf. ¸..¸ with note.
6j.r diametri Krohn–Lindow proposed ∋ VI<I>, giving a diameter of
:. :o¸/.88, which is in fact closer to the theoretical :..¸¸q digits (∋ ¸¸¸q). But
the figures at ¸..8 support the transmitted ∋ VI: :. :o./.88 + ¸ :86/.88 =
6j.z apud aquarios Cf. ¸..8 with note. According to Dessau CIL :¸
p.q::, some surviving pipes bear the legend CXX, with an equivalent diameter
of .q6 mm.
qui modus This instance of modus is very close to ‘area’ (see ¸..n.). The
‘unofficial’ :.o-pipe with :6¸.8¡ quinariae approximates the capacity of two
official :oo-pipes (8:.¡¸ × . = :6..q). See also ¸¡.¸n.
6q–¡6 C begins a new book at this point (Introd. ¸6). The next thirteen
chapters respond to <quem modum quaeque aqua habere visa sit quantumque erogaverit,
quaeque> erogationes habiles factae sint of the prologue (¸.:n.) – unless one takes
the transmitted erogationes . . . factae sint at that point to be instead a general
announcement of chapters ¸¸–86. By taking his own measurements of the
water available, F. investigates the reliability and accuracy of figures which he
has found in the imperial registers. From the survey it emerges that there is an
unaccounted ‘surplus’ of some :o,ooo quinariae, enough for him to claim that
his administrative investigation has yielded the equivalent of an entire new
source (8¸.., :o¸.¸). DeLaine (:qq¸) :.¸–8 observes that these chapters are
in some ways a well ordered continuation of the mathematical and statistical
material of which F. has treated already (.¸–6¸). In the previous section he
has laid blame on the aquarii for irregularities involving pipe-sizes that he has
detected and corrected (emendare: .¸.., ¸¸.:). In what follows he will address
irregularities that point to similar misbehaviour. His evidence now will consist
of differences between deliveries actually made at the present time (greater
than the total officially available) as well as measurements of his own which
indicate that even the total shown in the records as available is well short of
reality. The demonstration culminates in a rhetorical flourish (¸6.:–.). But
DeLaine is right to point out that the wearisome statistical account has a
rhetorical aim of its own: ‘to impress the reader with the author’s learning,
and to manufacture an air of mystery [I should have said ‘authority’] about
the subject’.
6q.r Persecutus . . . nunc ponam Cf. above .¸.:n. Quoniam. . . persecutus
sum, ¸.:n. ponam.
principum commentariis For these records, see ¸:..n.: they con-
tained figures for both the total available supply (in universo, in accepto) and
for delivery (in erogatione, in distributione; cf. ¸¸.¸). F.’s own measurements aside,
the data in chapters 6¡–86 reflect official records at the time he took office.
He notes at 88.¡ that amended figures cannot be given until reforms and
improvements are completed.
usque ad nostram curam Cf. :, :o..:, :¸, and esp. 8¸.: ad Nervam
imperatorem usque.
ipsi . . . invenerimus The plural is ‘official’ (q.¸n. exclusa, .¸..n. in-
venerim); cf. ad nostram curamimmediately preceding. The action here is explicitly
credited to imperial initiative. But in giving the details of his investigation F.
deliberately turns to the first person singular (see §¸n.).
providentia {optimi} diligentissimi{que} Nervae principis
Cf. 8¸.. providentia diligentissimi principis (name omitted there because preceded
by ad Nervam imperatorem usque). The word-order Nervae principis is that of C
(with no signes de renvoi, a slip in Kunderewicz’ apparatus). Cf. 8¸.: and 88.:,
where both times the title follows the proper name; contrast Pliny, Pan. 8.¸
imperator Nerva, perhaps for distinctive emphasis. (Divus Nerva at :o..¡, ::8.¸,
Pliny, Pan. ¸8.6, 8q.:, qo.6 is a quite different style.) Meant here is Nerva (not
Trajan), whose initiative for administrative reform is unambiguous at ::8.¸.
The emperor’s name has not appeared since chapter :: its recurrence here
reinforces the point in F.’s prologue that he was working at that prince’s be-
hest. Providentia optumi principis occurs in a document of Claudius’ reign (CIL
:o.:¡o: = ILS 6o¡¸). For providentissimus applied to Augustus see ::.:n. Pliny,
Ep. \iii.:¸.. speaks of Trajan as providentissimus imperator (the context is a Tiber
flood), and the superlative epithet appears on that sovereign’s coins: RIC ii:
nos..8, ¸¸8, ¸:¡, 66¸. Superlative adjectives as imperial attributes were be-
coming standard by F.’s time, but were not yet fossilised in imperial titulature.
Optimus was a favourite, understandably, for Trajan: Pliny, Pan. ..¸, 88.¡, and
passim, as well as in coins from :o¡ (RIC ii: .¸¸). The appearance of the rather
less decorative providentia diligentissimi principis later (8¸..) raises a suspicion that
our text might have been tampered with here (cf. ¸:..n. {invictissimi et piissimi}
principis). I keep the second adjective because diligentia is so recurrent a theme
in this work (see :n. diligentiam) and a second echo from the prologue would be
welcome at the beginning of a major new section – one that could be taken up
again in deliberate repetition still further on at 8¸..; cf. Baldwin (:qq¡) ¡88.
6q.z Fuerunt . . . in universo A total of :.,¸¸¸ available quinariae ap-
peared in the records, less by :,.6¸ than the total of :¡,o:8 recorded for dis-
tribution (of which the latter does not include transfers from one aqueduct to
another: see ¸8.:n. and Table ¸).
in universo . . . <in> accepto After in commentariis a second preposi-
tional phrase with in sounds a bit awkward, especially since in universo here
seems = in universum. Perhaps the construction was chosen to achieve a
COMMENTARY 6¡.¸–6¡.¡
chiasmus with anaphora: in universo, in erogatione, in distributione, in accepto (on
the last item see ¸¡.¸n.).
6q.j cum . . . crederem The singular should not be pressed to mean
that the investigation was F.’s own idea (although it no doubt was). In the
chapters that follow F. uses the first person singular to indicate his personal
responsibility for the accuracy of his measurements (.¸..n.): note inveni (6¡.¡,
6¸.¸, 66.., 6¸..), invenissem(66.¸), invenerim(¸...), egi (¸o.¸), repperi (¸¸..), exploravi
(¸¡.¸); cf. me (¸..¸, ¸¸.¸). But the data become ‘official’ (§:n.) and he turns to
the plural (see q.¸n.): nostra (¸o.¡, ¸..¸, ¸¸.¸–6), egimus (¸..¡).
in exploranda fide aquarum atque copia For explorare ‘test, prove’
by experience: ¸¡.¸ copiam . . . durantem exploravi, Str. iii.¸.¡ explorata eius (sc.
comitis) fide, 6.6 quendam exploratae sollertiae submisit. See also :8.¡n explorata. Fides
atque copia is practically a hendiadys, ‘reliable supply’. For this sense of fides see
nonmediocriter Cf. :.8.¸ poena non mediocris. Litotes is regular withboth
adjective andadverb, ¸q times inCiceroalone (oratorical works :¸, rhetorical q,
philosophical :¸, epistles .o).
ad scrutandum quemadmodum amplius erogaretur F. usually
ignores any vestigial financial sense in erogare / erogatio (¸..n.), but he appar-
ently recalls it here as it introduces his metaphor (in patrimonio, ut ita dicam). The
apology may be simply for the metaphor, but since patrimonium could apply to
the personal resources of the emperor (despite long traditions of patrimonium
publicum in, e.g., Cicero: OLD s.v.) F. might have been cautious in his use of the
word; see ::8.¸ proximis temporibus in Domitiani loculos conversum. For an interest-
ing parallel to F.’s administrative action see Sic. Flac. De condicionibus agrorum
ed. Campbell (.ooo) p.:.8.¸¡–6 cum data assignata nos computaremus et excederent
centuriae modum, reversi ad originem primam assignationis invenimus postea adiecta esse
nomina . . .
6q.q capita ductuum metiri Of this transitive verb, here with direct
object, F. uses the perfect participle absolutely at 6¸.. ad capita mensus inveni
and ¸¸.. mensus ad caput repperi. F. takes measurements at the sources whenever
possible, or explains why he cannot do so. He seems to have believed that the
official figures (the origins of which he does not know: cf. ¸¡..n.) were supposed
to reflect the quantity ad caput (6¸.., 66.., 6¸.., 6¸.6, etc.) or ubi aqua concipitur
(cf. ¸.¡n.). He speaks of the conceptionis modus given in the commentarii (66.6, 6¸.¸,
¸:.:, ¸¸.¡; cf. ¸..6) and of concepta commentariorum (6¸.6, ¸¸.:); note also the
phrase in accepto (§., ¸¡.¸). It is likely that one or both words were actually used
inthe records. But suchfigures inthe commentarii were nodoubt meant primarily
as a control of the available quantity (cf. ¸¸.¸). F. seems a bit harsh in dismissing
them as incompetent approximations (¸¡..), for earlier data might have been
taken either from gauges at the piscinae (:q..n.: note ...: conceptacula and ¸¸.:
COMMENTARY 6¸.:–6¸.¸
concipitur understood with in piscinis) or at terminal castella (¸..n.). Ashby (:q¸¸)
¸o–: notes that there is far less discrepancy between the earlier figures and F.’s
own measurements at the piscinae than with his measurements at the sources.
Blackman–Hodge (.oo:) :o.–¡ observe that F. might have failed to account
adequately for accidental losses inhis zeal totrack downfraudulent ones. It may
be that the records showed a larger quantity delivered than supplied simply
because their erogatio included deliveries made earlier than the point they took
as their conceptio. Confusion might have ensued if there was need to convert
grants made by pipe-size into quinariae units (:o¸.¡n.). It is also conceivable that
official figures for availability were set deliberately low to ensure an adequate
supply for public uses (a variation perhaps of the principle enunciated by Vitr.
\iii.6.:–.). Changes in record-keeping with some resultant disarray (¸¸..n.)
may, however, provide an equally reasonable explanation. Transmitted figures
in chapters ¸8–86 are too unreliable even to determine the deliveries extra urbem
(although the extra / intra distinction might suggest that records were based on
measurements taken at the urban castella). The total extra-urban delivery of
¡,o6¸ (¸8..) is of course considerably larger than the difference of :,.6¸ (§.)
between official deliveries and the supply, but the similarity of the numbers
may be more than coincidental.
circiter . . . decem milibus Cf. ¸¡.¡ decem milia quinariarum intercidisse.
The total of F.’s measurements is .¡,¡:¸ (Table ¸), larger than the recorded
:.,¸¸¸ by ::,6¸8 (or a difference of :.,o¸o if one includes the ¸q. quinariae of
6j.r in commentariis adscriptus est Formulaic in chapters 66–¸:
(66.:, 6¸.:, etc.); cf. §¡ in commentariis adscribitur.
octingentarum <quadraginta> unius The number 8¡: is needed
to square with the figures in §¸ (:,8.¸ less q8¡) and §¡ (¸o¡ plus :¸¸), as well
as to provide for the total of :.,¸¸¸ (6¡.. and Table ¸).
6j.j ad Gemellos . . . Augustae I print intra (cf. .:.¸n. intra portam
Esquilinam) and take it to mean within – or possibly this side of – the open
area known as Spes Vetus. F. usually says ad Spem (¸.6, :q.¸, .o.¸), but cf. .:.¸
secundum Spem veniens.
altitudinem aquae The measurements give an area of 8 ¸/¡ square
feet, or .,.¡o square digits (cf. .¡..). This is the equivalent of .. centenariae (each
of which has a cross-section of :oo square digits: .q.:–.) plus one quadragenaria
(¡o square digits). To compute the quinariae F. multiplies by .. the capacity of
a centenaria (8: :¸o/.88: 6..:) and adds that of the quadragenaria (¸: :68/.88:
¸o). The result is :,8.¡ :¡8/.88 (:,8.¡.¸:), rounded to :,8.¸.
<in> commentariis habet The subject of habet is the water (in this
case Appia). The phrase is exactly paralleled at 6q.6 (cf. ¸... and ¸¸.:), and
COMMENTARY 6¸.¡–6¸.6
nowhere do we find commentarii habent. C’s habent is accidental, perhaps from
the preceding fiunt and efficiunt.
6j.q erogabat F. generally uses the imperfect tense (e.g. 66.¸n., 6¸.¸,
¸..6, 8o.:, 8¸.:), for his own reforms have rendered these figures obsolete,
particularly for distribution (see 8¸–8, esp. 88.¡). Instances of the present,
however, appear at 6q.¸n., ¸o.¡n., ¸:..n. and ¸..¸n.
mensura respondet Cited in OLD s.v. respondeo ::, of which no defini-
tion is really appropriate (‘to be obtained as a result of, and in proportion to,
someone’s exertions, reward the efforts of’). Is the subject modus, as with ad-
scribitur (cf. §:)? If so, we should expect dative mensurae. If the subject is mensura,
can respondet mean ‘shows, reveals’? We get little help from ¸¸ (aquam) respondere
modulo suo.
6j.j intercidit ‘To be lost, disappear’, inter +cadere: usedbelowat 66.¸, ¸,
6¸.8, ¸¡.¡; cf. Livy, xx\ii..¡.8 fraude amotas [sc. claves] magis ratus quam neglegentia
cumsit depressio<r> F. might have spelt (conjunction) quom(the same
issue arises just below in §¸, where C has quam for preposition cum), but the
error is probably one of scribal misinterpretation at some stage in the tradition.
For depressus = humilis cf. Str. i.¸..¡ depresso loco.
manationes The word is extremely rare: F. uses it always in the plural
and in a concrete sense; see TLL 8: .¸:.¸; cf. Espinilla Buis´ an (:qq¸) 6¡6.
Here perhaps the sense is ‘water from (undetected) leaks’, but elsewhere (::o.:,
:...:) more generally ‘discharge’. Since leaks would have been for the most
part evident in one way or another and therefore likely to have been repaired,
it is possible that F.’s term manationes also embraces some kind of deliberate
discharge (by valves, for instance): see further ::o.:n. For manare applied to
water in normal flow, see TLL 8: ¸.o.¡¡.
quas esse . . . ex ea manat The text is uncertain. F. seems to say
that the appearance of water at several points in the city can reasonably be
attributed to losses from Appia’s subterranean channel (cf. ¸.¸); it could have
been ‘observed’ as ground water or in underground cisterns (¡.:n. hauriebant).
Probata could just possibly be genuine (‘of good quality’), although editors have
always found it troublesome; note Vitr. \iii.¡.: fontes probatissimi, aqua erit . . .
probata (cf. Colum. i.¸.¸). The end of the sentence is abrupt and awkward.
There may be a lacuna, but what it contained I could not guess. I delete id
quod ex ea manat as a gloss on the rare word manationes, else for id quod we should
have wanted something like Dederich’s quae (sc. aqua) ex eo (sc. ductu) manat.
6j.6 deprehendimus F. uses this verb when he has detected illegal uses
of water (see ¸o.¡n. comprehendimus), as has happenedfrommost of the aqueducts
(66.¸, 6¸.8, 6q.6, ¸..¸, ¸¸.¸). He mentions such misdemeanours to justify the
COMMENTARY 6¸.¸–66.¡
accuracy of his calculations and to reveal the thoroughness of his scrutiny. He
is firm in blaming the aquarii (¸¸.., 8¸.:), but with private individuals he deals
more gently (¸¸.¸, 88.., :¸o.:–¡).
6j.¡ propter pressuramlibrae Appia is spared the problem of illegal
taps because its conduit is so far underground. Pressura librae is a compressed
expression meaning ‘low level’: pressura (:8.¸n., ¸¸) is the force which creates
gravitational flow to places lower, libra (:8.:n.) the level above which water
cannot be raised. The absolute level of the source is technically irrelevant (as
F. must have known: cf. :8.6n.); he mentions it here to underscore the overall
low level of this aqueduct (cf. :8.¸, ¸q.:).
infra terram . . . pedibus quinquaginta While not securely iden-
tified (¸.¡n.), Appia’s source is without question subterranean. Even if one
therefore accepts cum sit infra terram, there remains no little uncertainty with
pedibus quinquaginta. Kortz (:8q¸) ¡¸ noted the unexpected ablative where ac-
cusative would be normal (cf. :¸.¸ sublevati . . . pedes centum novem). And fifty
feet (c. :¸ m) seems remarkably deep (cf. ¸..n. venas): perhaps both words are
66.r in commentariis modus The official figure probably included
the :6¡ quinariae which Anio Vetus received (before reaching its piscina) from
Marcia (6¸.¸). Grimal 8¸ n.¸¸ believes that the transfer took place near Tibur
(6¸.¸n.) and suggests that F. either ignored the figure for the sake of simplicity
or that Anio distributed the water soon after it was received and thus the figure
could be eliminated.
mille quingentarum quadraginta unius The transmitted text has
quadringentarum, but a figure of :,¸¡: is needed for the arithmetical computation
in §. (¡,¸q8 less .,8¸¸), as well as for the grand total of :.,¸¸¸ (6¡.. and Table
¸). Poleni maintained :,¡¡: by altering .,8¸¸ (§.) to .,q¸¸ and 6q (§6) to :6q;
he kept the total :.,¸¸¸ by following early editors in giving Virgo ¸¸. (¸o.:).
in proprium ductum Tiburtium derivatur See 6.¸n.
66.j eroga<ba>ntur Imperfect tense is confirmed by veniret in depen-
dent clause (see 6¸.¡n.).
66.q modus . . . efficit Cf. 6¸.6 mensurae ad caput actae efficiunt, 6q..
modus . . . mensuris efficit, ¸o.¸ mensuram egi quae efficit; see also :q..nn. modus earum
mensuris ibidem positis initur. On efficere see ¸.8n.
per mensuras positas If the gauges at the settling-tanks were already in
place (see :q..n.), F. can quote their measurements without risking a charge of
inaccuracy: note especially 6q.. manifestis mensuris and ¸..¸ indubitatae mensurae.
He also uses the passive construction modus initur (:q.., 6¸.¸) or efficit (6¸.¸,
6q..) in contrast to first-person verbs (see 6¡.¸n.).
COMMENTARY 66.6–6¸.¸
66.6 amplius . . . minus To the :,¸¡8 quinariae delivered after the tank
F. adds the .6. delivered before (§¸) to get a total of :,6:o, which is 6q more
than the :,¸¡: shown in the records. It is surprising that he does not make
this clear (as at 6¸.6–8), and it is possible that a phrase has dropped out after
octo (e.g. summa quae erogabantur aut ante piscinam aut post piscinam quinariarum mille
sexcentae decem). Of the .,¸6. measured at the piscina only :,¸¡8 are accounted
for in delivery, leaving :,o:¡. The total loss is thus :,¸¸¡ (§¸) +:,o:¡ = .,¸88.
66.¡ errore mensurae F. apparently refers to his own measurement at
the source (ad caput inveni §.; cf. ¸..¸n. mensurae ad caput actae). Elsewhere he puts
great trust in the gauges at the piscina (:q..n.).
6¡.z mensus For the participle used absolutely (as also at ¸¸..) see 6¡.¡n.
capita ductuum metiri.
6¡.j in adiutorium Tepulae Context makes clear that Marcia deliv-
ered water to Tepula before its own piscina (:q.:), but this seems to have been
only a short distance beyond the piscina of Julia (68.¡ statim). The junctions
and interchanges must have been fairly complex, especially at the piscinae near
the seventh mile (:q.:n.), but we are seriously handicapped by the lack of
archaeological material. On junctions in general, see Hodge (:qq.) :.o–:.
item <in> Anionem The construction is normal (cf. :¡.¸, 6q.¸) and
loss of in very easy after item. a has Anioni – an obvious emendation, probably
to parallel Tepulae (mistaken as dative). Anio Vetus is meant, for Marcia’s level
rules out Anio Novus (cf. :8.6, q:.:). Grimal (8¸ n.¸6) thinks that the transfer
took place at Grotte Sconce; cf. Ashby (:q¸¸) 66 andsee q:.¸n. Context suggests
rather a spot shortly before Marcia’s piscina.
6¡.j cumeo . . . in arcus excipitur The text is uncertain. Elsewhere F.
simply states a measurement takenat the piscina (66.¡, 6q.., ¸..¸), but here there
is apparently a reason for mentioning an additional quantity not measured
there but which is carried on arches (beyond the piscina: :q.¸, thus Krohn’s
citra); note below §6 in arcus recipitur and §¸ ex piscina in arcus recipiuntur. I print
B¨ ucheler’s text, which can be translated ‘along with that (quantity) which is
carried around the settling-tank and is taken up on arches in the same channel
(as that which has passed through the tank)’. For (modus) ductus thus used see
¸.:n. perducta sit. But if this is right, there still remains a difficulty with circa,
for it is not clear why some of the water would be diverted around the tank.
Perhaps a diversion was connected with the distribution to Tepula (§¸); some
adjustments might have been required because of the levels (Marcia was lower
than Tepula after the aqueducts left the piscinae: :q.¸n.). I see no way to take
circa as ‘at, near’ (cf. ¸:.: circa Careias), nor ductus as genitive of the noun (¸..n.
ductus cuiusque).
COMMENTARY 6¸.6–68.¡
6¡.6 mensurae ad caput actae Cf. ¸..¸n. mensurae ad caput actae.
6¡.¡ <modum> significari Exactly paralleled by 66.6 above; cf. ¸..¸
(quinariae) quas significavi.
6¡.8 duo milia quingentae The transmitted quingentae is not quite ac-
curate. The loss before the piscina (:,¸q¸: §6) and that beyond (:,:o¡: §¸) total
.,¡qq – either of the lower numbers might be off by one. At an earlier stage
the text very likely had numerals, from which a single stroke might easily have
been lost here: thus Poleni and B¨ ucheler would have duo milia ID.
6¡.q manifestum est et ex . . . eo The word hoc is a variant on eo
which has made its way into the text; without it the construction is perfectly
normal: cf. 6¸.¸, ¸¸.6. The et before ex were perhaps better deleted, but I take
it as = etiam ‘from the very fact that . . .’.
praeter eam mensuram quam comprehendisse nos capacitate
ductus posuimus ‘Besides that measurement which we set down as having
ourselves determined by the capacity of the channel’ (in direct statement: nos
eam mensuram capacitate ductus comprehendimus); cf. §. ad caput mensus inveni. Mensura
here is the result of action (a measurement) rather than the action itself (as at §6
mensurae ad caput actae: see ¸..¸n.). We have mensuramcapacitate comprehendere where
we might have expected capacitatem mensura comprehendere. F. seems to employ
mensura and comprehendere in two different yet quite similar constructions: (:)
as here, with mensuram as direct object, and (.) another with mensura ablative
(quinarias) quas mensura comprehendimus (¸o.¡n.: conjecture). For the former we
have the example of comprehensam a nobis mensuram (¸¸.6n.). For the latter cf.
Colum. \.:.¡ modus . . . mensura conprehenditur (cited at .q.:n.), De agr.qual. p...¸–
¡ Campbell, p...:. ager mensura comprehensus, and especially p...:¸–:6 mensores,
quamvis extremum mensura comprehenderint . . .
effunduntur The verb does not necessarily imply waste, only the over-
flow at the source. Cf. ¸..¸–8, where F. uses the synonymous superfluunt. Cf.
exuberare (¸¸n.) and ex(s)uperare (¸¸.6n., conjecture).
68.z huius aquae . . . in Iulia See 8..n., q.:n., :q.¸n. With the imper-
fect constabat compare perducebatur (8..). Note the contrast betwen fontes (¡.:n.)
and venae (¸..n.).
68.q ad hortos Epaphroditianos The horti Epaphroditiani (RE \: .¸:o,
NTD :q8–q, LTUR iii: 6o, Mancioli) were named for Nero’s freedman Epa-
phroditus (Tac. Ann. x\.¸¸.:, Suet. Ner. ¡q.¸). Since Claudia and Anio Novus
ended post hortos Pallantianos (.o..), the horti Epaphroditiani lay closer to Porta
Maggiore. Both properties were created from the horti Tauriani: Grimal (:q¸q)
and (:q¡¸) :¸q.
COMMENTARY 6q..–¸o..
6q.z mensura in<ven>iri non potuit F. regularly uses the passive of
invenire in this way: 6¸.. inveniri mensura non potuit, ¸o..n. mensura . . . inveniri non
potuit, ¸:.: modus nec . . . certus inveniri potuit, ¸¡.: copia . . . mensuris inventa sit (cf. also
¸..¸ minus quam in piscina invenitur). (For invenire active see 6¡.¡n., 6¸.¸ altitudinem,
66.., 6¸.. quinarias, etc.) F. employs the mathematical term iniri (aside from
.¸..n. and ¸:.:) only in the formulaic modus . . . mensuris positis initur (:q..n.,
66.¡, 6¸.¸).
[ad] sextum . . . miliarium At :q.: F. says that the piscinae of six
aqueducts are located intra septimum miliarium. Giocondo’s ad is the easiest sup-
plement for the lacuna (cf. e.g. ¸o.¸), but the number could be erroneous.
manifestis mensuris Note alliteration, for effect (cf. ¸..¸ indubitatae
mensurae). On these gauges see :q..n.
6q.j accipit . . . ex Claudia Lanciani (:88:) ¸:¡–:¸ is perhaps right
in explaining that these :6. quinariae comprise the pars Iuliae which was taken
from Spes Vetus to the Caelian (but see :q.¸n.). If so, they would have been
only a part of the water carried on the Neronian arches, perhaps calculated
only by the record of Julia’s nominal deliveries on the Caelian (¸6.¸).
post hortos Pallantianos See :q.8n.
6q.q est omne Iuliae in acceptis = Summa quinariarum quas Iulia accepit.
The awkwarduse of singular neuter adjective withgenitive may be for variation
(commoner, and more elegant, is the plural, e.g. Str. i.¸.:o in profunda silvarum).
6q.j dat . . . erogat The present is usedperhaps for variation(see 6¸.¡n.);
cf. 6¸.¸ erogabantur . . . et dabantur in adiutorium Tepulae.
6q.6 minus quam in piscina F. computes the loss from the measure-
ment at the piscina (:,.o6 less qq¸ delivered = .:¸); he does not include the
:6. quinariae received later from Claudia (§¸).
quas ipsas Ipsas is emphatic. Note the tense of usurpabant.
¡o.r sexcentarum quinquaginta duarum The official figure for
Virgo’s supply (6¸.) differs most astonishingly from the record of its deliv-
ery (.,¸o¡: §¡); it apparently delivered only .oo quinariae outside the city (8¡.:,
cf. 6¡.¡n.). Ashby (:q¸¸) ¸: n.: suggests that 6¸. may be corrupt; but the arith-
metic works to give the total of :.,¸¸¸ (6¡.. and Table ¸). Virgo had no piscina
(...:) and there could be no reliable measurement taken at the source (§.).
The official figure might have been derived from an inaccurate calculation
(¸¡..n.), or conceivably it represented deliberate exclusion of all but certain
categories (e.g. to privati: cf. 8¡.:–¸).
¡o.z quoniam . . . intrat For Virgo’s source see :o.¸–6nn. Measure-
ments could not be taken because the supply was one of feeders (as for Julia
6q..; cf. 6¸.., 68..); lenior contrasts with velociorem (§¸n.).
COMMENTARY ¸o.¸–¸:.:
¡o.j prope urbem . . . Commodi It is a little strange to read prope
urbem, for the source was at the eighth milestone (:o.¸) and F.’s measurement at
the seventh. The phrase may be for directional clarity (as at 6q.. sed ad sextum
ab urbe miliarium, or ¸..¸, but see 8..n. and q.:n.), or Poleni may be correct in
suspecting an error in the transmitted septimum. The source was in agro Lucullano
(:o.:) and Ashby (:q¸¸) :6¸ n.q is probably right to assume that the reference
here is to the same ager, now the property of one Ceionius Commodus, but he
slips in identifying himas L. Ceionius Commodus (RE no.¸), cos. :¸6, who was
the adopted son of Hadrian and father of Lucius Verus. F. probably means L.
Ceionius Commodus (RE no.¸), cos. ord. of ¸8 cr, or his son of the same name
(RE no.6), cos. ord. in :o6 (the latter accepted by Kuhne, KP ii: :qo¸ no.¸).
velociorem iam cursum At Virgo’s source the water merely filled a
collecting basin(:o.¸) andenteredthe conduit too slowly to provide anaccurate
measurement. But approximately a mile further on the channel had enough
fall to give the water a certain velocity. The situation here is the reverse of that
whichF. observes at the intake of AnioNovus (¸¸.6vis aquae rapacior). His method
of measurement, it seems, requires a certain velocity – but not too much.
The computation of Virgo’s .,¸o¡ quinariae is almost certainly determined
by the area of the channel (as at 6¸.¸). Agreement between his calculations
and the record for delivery must have been entirely coincidental.
¡o.q adprobatio nostra Note the objective force of the personal adjec-
erogat If the present surprises (see 6¸.¡n.), perhaps F. uses it here because
Virgo has yielded no instances of fraud that will alter erogatio in future.
quas mensura comprehendimus ‘Which we determined by mea-
surement’ (cf. §¸ mensuram egi and see 6¸.qn. praeter eam mensuram). The trans-
mitted deprehendimus is anomalous: in this text F. uses this verb only for detecting
abuses (¸..¸, 6¸.6, 6¸.8, 6q.6, ¸..¸, ¸¸.¸, ¸¸.., q:.¸, ::..: and ¡). It might be
defended by here supposing he means that he has detected an error in the
earlier reckoning (cf. 66.¸).
¡r.r nec in re praesenti F.’s remarks have become less opaque since
Liberati Silverio (:q86a) brought it to our attention that the level of Lacus
Alsietinus began to fall near the end of the first century; cf. Taylor (.ooo)
circa Careias For the locative force of circa ( =ad) cf. .o.¸ circa ipsum
montem, q¸.¸n. circa ipsum oppidum, Str. i.¡.¸ circa Abydon, ¸..¸ circa Amphipolim,
ii.¡.6 circa Aquas Sextias. Careiae is the deserted village of S. Maria di Galera
(RE ¸: :¸88), fifteen miles from Rome on the Via Clodia (Tab. Peut. ¸.¡, Itin.
Anton. ¸oo..).
ex Sabatino The Lacus Sabatinus is the modern Lago di Bracciano
(RE :A: :¸¸¸). Lack of remains makes it impossible to determine whether the
COMMENTARY ¸:..–¸..¡
supplemental water was taken from the Arrone (the outlet stream of the lake)
or from the lake itself, but the elevations in the area establish that the junction
was not far from modern Osteria Nuova (::.¸n.). The inscription of the Forma
Mentis (Appendix B, no.:o) was found nearby, as a cover slab on a branch of
the modern Acqua Paola.
<tantum accipiat> . . . temperaverunt Quantum seems to require
some such supplement, although F. is not overly fond of correlatives. He cannot
accurately determine the quantity (or does not care to: ::.:n.), perhaps because
he might have been unable to take a measurement at the junction of the two
branches. There is no settling-tank (...:) from which to gauge urban delivery.
For temperare ‘to regulate, control’ see ¸¡.¡ beneficia sua principes secundum modum
<in> commentariis adscriptum temperant; cf. Pliny, Ep. x.6:.¡ cataractis aquae cursum
¡r.z Alsietina erogat The present tense here is consistent with that
which F. uses for this aqueduct in his report of deliveries (8¸n.). The quan-
tity is only that delivered outside the city (8¸); it does not include any for the
naumachia, perhaps by this date defunct, and its adjacent park (::.:n. coeperat
¡z.r abundantior aliis Both Marcia (with ¡,6qo quinariae) and Anio
Novus (¡,¸¸8) surpass Claudia, according to F.’s own measurements (Table
¸). But with the overflow of :,6oo quinariae (§¸) Claudia’s total comes to 6,.o¸,
greater than Marcia by more than :,.oo quinariae (even reckoning Marcia’s
own overflow: 6¸.q).
¡z.j minus inveniatur To arrive at the difference of :,.q¸ quinariae F.
merely subtracts the measurement at the tank from that at the source (¡,6o¸
less ¸,¸:.). But quam revera esse debeat is not strictly fair, since some of the loss was
due to official grants (ex beneficiis). F. does not specify how much was delivered
and how much stolen; but with plurimum he certainly implies that most of the
loss was illegal.
¡z.q circa erogationem For circa ( =de) cf. ¸¸.: circa ea, :o¸.. circa ius,
::¸.: circa conlocandos calices, Str. i.6.¸ circa praedam, ii.q.8 circa gratulationem. See
also ¸6.¡n. circa montem Caelium et Aventinum.
ad illas saltem The three-fold comparison is continued in §¸:
ad commentariorum fidem commentariorum ratio
ad eas quas ad caput egimus mensuras mensurae ad caput actae
ad illas . . . ad piscinam in piscina (invenitur)
Editors have dealt variously with the textual disturbance, usually by contriving
a relative clause for illas to balance that for eas. B¨ ucheler’s (mensurae) quae . . .
positae has the merit of conforming to F.’s practice elsewhere (cf. :q.., 66.¡,
COMMENTARY ¸..¸–¸¸.¸
6¸.¸). As a prepositional phrase, post tot iniurias ( =postquam rivi tot iniuriis affecti
sunt) seems intolerably odd: Kortz (:8q¸) 8¡. These words might be explained
if positae had become post and was apparently in need of an object.
¡z.j erogantur The present (rather than imperfect: 6¸.¡n.) may be
found here because F. is thinking throughout §§¡–¸ in terms of the calcula-
tions rather than records for erogatio; cf. §¡ adparet, convenit, §¸ dat, demonstraverunt,
mensurae ad caput actae ‘Measurements taken, or performed, at the
source’: here mensura indicates the action of measuring rather than its result
(‘a measurement’ or ‘reading’: see 6¸.qn.). Actae is exactly parallel to 6¸.6
mensurae ad caput actae and consistent with the idiom: ¸o.¸ mensuram egi, ¸..¡
egimus mensuras, ¸¡.: actis mensuris, Agrim. p...:¸ Campbell mensurae aguntur (cf.
...q mensurarum actu), :..¸o mensura agenda. See TLL 8: ¸¸8.6¸ and ¸¸q.¸¸.
¡z.6 in urbe miscebatur Cf. 86.: Claudia et Anio novus extra urbem proprio
quaeque rivo erogabantur, intra urbem confundebantur, and q:.¸ Claudiam, quae per multa
milia passuum proprio ducta rivo Romae demum cum Anione permixta in hoc tempus perde-
bat proprietatem. F. seems mildly distressed that the spring water of Claudia is
thoughtlessly mingled with river water. The imperfect perhaps hints at a plan
to segregate, never in fact carried out so far as we can tell from archaeological
evidence. The two aqueducts shared a single terminal castellum (.o..n.), prob-
ably for reasons of both economy and efficiency. With F.’s cynical explanation
here compare q:.¸, where he blames the imperitia of the watermen for allowing
Anio Novus to pollute lower-level aqueducts.
¡z.8 capiuntur enim We can keep quem inventumif we understand its an-
tecedent to be fons Augustae (:¡.¸; cf. 8q.¡ capiuntur ex fontibus), and this should be
possible – between fontium and fontibus – without writing ex <fonte> Augusta<e>.
¡j.r in commentariis habere ponebatur Tense contrasts with F.’s
use elsewhere (in commentariis habet 6¸.¸, 6q.6, ¸..., in piscina habere 6q.6). Perhaps
the reason is that there are plans afoot for major changes to Anio Novus
(q¸.:–¡) which will have rendered obsolete the present data in the records for
this aqueduct.
¡j.z mensus ad caput repperi For this absolute use of the participle
(also at 6¸..) see 6¡.¡n. capita ductuum metiri. F. seems to have repperi for variation,
for his usual word is invenire (6¡.¸n.).
¡j.j quarum adquisitionem non avide me amplecti Cf. ¸..¸ me
adquisitionummensuris blandiri (+dative, ‘exaggerate’: OLDs.v. ¸). For avide amplecti
cf. Livy, xxx\i.¸..8 aviditate plura amplectendi; Pliny, Ep. ii.:¸.: et tu occasiones
obligandi me avidissime amplecteris, et ego nemini libentius debeo. In the present context
this conveys something like ‘I grasp with (perhaps unwarranted) eagerness.’
COMMENTARY ¸¸.¡–¸¸.6
in erogatione . . . continetur Although contineri is usually constructed
with the ablative alone (cf. .q.:, ¸¡.¸, ¸¸.:; see TLL ¡: ¸o¸.:¸), here the prepo-
sitional phrase is locative: in erogatione . . . commentariorumstands for in commentariis
in erogatione (as at 6¡..; cf. 68.¸). It is exactly parallel to in conceptis commentariorum
(6¸.6, ¸¸..).
¡j.q quattuor milia ducentae {undecim} Afigure of ¡,.oo quinariae
distributed is needed to arrive at a total erogatio of :¡,o:8 (6¡.., ¸8.:, ¸q.:: Table
¸), and the loss of cum in the following phrase suggests a scribal mistake at a
stage after numerals had been written out. The change to ¡,.oo, however,
necessitates a further change in §¸ (see below). It is curious that F. gives no
measurements at the piscina of Anio Novus, and he seems to be going out of
his way to explain his conclusions when it would have been simpler to report
a reading. Perhaps he aims at variety (cf. ¸o.¡, argument from delivery data),
but there remains a possibility that something has fallen out here.
¡j.j quingentasXXXVIII The transmittedtext reads quin. viginti septem,
which squares with the ¡,.:: transmitted in §¡ (¡,¸¸8 less ¡,.::). But if ¡,.::
is an error (see above), the figure needed here is ¸¸8 (¡,¸¸8 less ¡,.oo). It
is entirely possible that both figures for this chapter are correct and that the
error occurs elsewhere (the data in chapters ¸q–86 are too seriously mangled to
provide a check: see also Tables 8 and q); Blackman–Hodge (.oo:) ::6 remark
on editors’ readiness to see problems only in totals.
<sed et> Giocondo printed sed alone, but B
’s reading is better. F.’s
practice in this text is as follows: non tantum . . . sed etiam ¸¸, ¸¸.¸; non solum . . .
sed etiam qo.., ::¡.¸ (cf. :::..): the only time there is no et(iam) is ::q.¸ non
solum . . . sed et; nec tantum . . . sed, explained perhaps by the close proximity.
deprehendi<mus> The plural is not perhaps necessary, but it fits better
between nostras and nobis and matches deprehendimus in the parallel instances at
6¸.6, 6¸.8, 6q.6, ¸..¸.
¡j.6 exuperare . . . mensuram For the idiom mensuram comprehendere
see 6¸.qn. praeter eam. Transmitted exuberare is troublesome, for the verb is not
otherwise transitive in the sense required here, ‘to exceed, surpass’ (OLD s.v. ¸,
TLL ¸..: .o¸6.¸¸, cf. ¸¸n.). For ex(s)uperare see OLD s.v. ., TLL ¸..: :q¸¸.6:,
but with no good parallels. On the confusion of the two verbs see TLL s.vv.
and cf. Veg. i\.¸o.:.
¡j.6 vis aquae rapacior . . . velocitate Cf. Hero Dioptr. ¸: tiotvci
ot ypn c:i co:t to:iv c0:cpst, tpc, :c tti,vcvci, tcocv ycpn,tï 0ocp
n tn,n, :c toptïv :cv c,scv :c0 ptúuc:c, . . . óììc sci :c :óyc, co-
:c0. :cyu:tpc, utv ,cp c0on, :n, púotc, tìtcv ttiycpn,tï :c 0ocp,
µpcou:tpc, ot utïcv. But on whether F. had technical (as opposed to empirical)
COMMENTARY ¸¡.:–¸¡.¸
understanding see Blackman–Hodge (.oo:) :8. Conditions at the spot where
F. measured were significantly different from those elsewhere (even of Anio
Vetus, whose caput – for him – may not have been the intake from the river:
see 6.¸n.). F.’s explanation makes clear that he recognised not only that there
existed a relationship between velocity and quantity (cf. ¸o.¸) but also that his
systemof measurement was incapable of taking velocity into account. Baldwin
(:qq¡) ¸o¡ observes that rapax is a ‘mot juste of water and other powerful ele-
ments’. This usage is especially frequent in poetry: to Sen. Med. ¸¸¸ rapax vis
ignium, which he cites, add Ennius Ann. ix, fr.¸o. rapax unda, Lucr. i.:¸ fluvios
rapaces (cf. Virg. G. iii.:¡.), \.¸q¸ rapax vis solis equorum; Ovid, Ars Am. i.¸88
ventus (cf. Val. Flac. Arg. i\.6o6), Met. \iii.8¸¸ ignis; Lucan, \.¸q¸ turbo (cf. Val.
Flac. Arg. i\..6. turbo); Sen. Phoen. ¸: torrens; Her. Oet. :.: flamma; cf. also QNat.
\.:6.¸ violenta vis et in unam partem rapax.
¡q.r Non dubito Chapters ¸¡–6 conclude the section begun in chapter
6¡. This sentence deliberately recalls 6¡.. Fuerunt ergo in commentariis in universo
quinariarum decem duo milia septingentae quinquaginta quinque, in erogatione decem quat-
tuor milia decem et octo: plus in distributione quam <in> accepto computabatur quinariis
mille ducentis sexaginta tribus. F. has by now demonstrated the inaccuracy of the
records and can offer three explanations for the discrepancy: (:) the earlier
measurements were careless or unmethodical (§.); (.) the aquarii had unoffi-
cially allowed concessions to private citizens (¸¸..); and (¸) landowners had
illegally tapped the aqueducts (¸¸.¸). He has also detected a less serious but
noteworthy imprecision in the records of distribution (¸6.¸; cf. ¸¸..).
admiraturos C’s adnotaturos is entirely too weak after non dubito, despite
adnotare + quod at :o¡.¸ crediderim adnotandum quod senatus . . . vetuerit and :o6.¸
dignum adnotatione [C: admiratione a] est quod . . . [S.C.] non permittit. Mere ob-
servation, even acknowledgement, is flaccid in sense. Required here is ‘some
will be surprised’ (Herschel), for after his exhaustive treatment F. not unrea-
sonably expects his readers to appreciate the admiratio that had prompted his
own investigation (6¡.¸).
¡q.z parumdiligenter . . . fecerunt aestimationem The vagueness
of qui ab initio reveals that F. is ignorant of when (and how) the official data were
collected (see 6¡.¡n.). In parum diligenter we hear the purposeful reproach of the
administrator, and facere aestimationempointedly contrasts with F.’s personal (and
supposedly more reliable) mensurae.
¡q.j tota deinceps aestate . . . exploravi Probably during the first
summer F. was in office (q¸ cr): see q.8n. For the ablative to indicate extent of
time cf. :...¸ quam paucissimis diebus rivi cessent: K–S ii.:: ¸6o. On explorare see
:8.¡n. explorata, 6¡.¸n. exploranda fide.
COMMENTARY ¸¡.¡–¸6.:
¡q.q quaecumque tamen est causa Despite the firmness of his state-
ment in §., F. leaves room for other explanations beyond the one he has
eliminated in §¸.
decem milia quinariarum See 6¡.¡n.
¡j.r sequens diversitas See Table ¸.
¡j.z fraus aquariorum F. sets forth the details of this abuse in a later
section (::.–:¸). For other reprehensible actions on the part of the aquarii see
q.6, ¸:.¡–¸¡.., ::o.., ::¸.¡. For the possibility of exaggeration on F.’s part see
Blackman–Hodge (.oo:) :o.–¡.
¡j.j sed et plerique possessorum Note that F. elsewhere sharply
distinguishes between fraud of the aquarii and impotentia (:.on.) of private indi-
viduals: see 6¸.6n. and cf. :.¡.¡, :.6.:, :.8.:. The text is sound at least for the
main verb and its object (formas rivorum perforant). The possessores are occupants (if
F. uses the word loosely, sometimes perhaps technically domini, owners), proba-
bly of suburban villas and of agricultural land along the route of the aqueduct.
Cf. Pliny, HN xxxi.¡. utriusque (Marcia and Virgo) iam pridem urbi periit voluptas,
ambitione avaritiaque in villas ac suburbana detorquentibus publicam salutem.
†e quorum agris aqua circumducitur† The sense here ought to be
something like Bennett’s ‘past whose fields the aqueducts run’, but it is hard
to see how deletion of e can help (Mommsen (:8¸o) ¸o¸), or how circumducere
(a ctcç in F.) can bear this meaning.
formas rivorum Although forma, a technical term for the conduit or
channel of an aqueduct, is attested prior to F. (e.g., Appendix B, no.:o), he
prefers specus (canalis only twice: :q.¸ and 6¸.¸) or the looser ductus and rivus
(¸.:nn.). He uses forma in this sense only here and at :.6.¡ (both times plural).
The use became common in later Latin (see TLL 6: :o¸8.6q): note, e.g., Fav. 6
forma structili and per formam where Vitr. \iii.6.: had written per canales structiles
and canalibus.
†vel ad oritorum† The difficulties are probably connected with prob-
lems earlier in the sentence: note especially the intrusive unde and the apparent
repetition of ex ductibus publicis in privatorum usus. It seems unlikely that vel ad is
meant to vary the construction of dative hominibus, although vel may be inten-
sifying. My decipherment of C prompts me to think that F. might not have
mentioned horti at all. To delete this phrase would leave a pointed contrast
between public conduits and private individuals, but I am still uneasy about the
metaphor in itinera suspendant. That landowners and others near the City had a
major interest in water for commercial purposes is often neglected: see Leveau
(.oo:) 86.
¡6.r nec plura nec melius dici possunt Virtually = nec plura dicenda
sunt nec melius dicere possum. Such an introduction, followed by the title of a
specific work, offers the tantalising possibility that F. quotes directly from
Caelius’ contio. It would have been otiose to interrupt his indignation with
some explicit announcement of verbatim citation: Rodgers (:q8.a). On this
view, nec plura anticipates the succinctness of the quotation, nec melius conveys its
contemporary validity. Uncertainties must remain on the question. Vocabulary
and tone may be no more than allusive.
a Caelio Rufo Caelius (RE ¸::.66, no.¸¸) delivered the speech during
his curule aedileship in ¸o ncr (MRR ii: .¡8). He mentions the controversy
briefly in a letter to Cicero written in February of that year: Fam. \iii.6.¡ nisi ego
cum tabernariis et aquariis pugnarem, veternus civitatem occupasset. The traditional view
is that aquarii had taken bribes from tavern-keepers, but Shackleton Bailey (ad
loc.) points out that Caelius’ concern with tabernarii might have been a discrete
issue over weights and measures.
in ea contione Caelius’ is one of only three known instances of aediles
speaking at contiones: Pina Pola (:q8q) ¡8. The other two are C. Iulius Caesar
Strabo qo ncr (Cic. Brut. ¸o¸; MRRii: .6), P. Clodius ¸6 ncr (Cic. QFr. ii.¸..–¡,
Vat. ¡o; Dio, xxxix.:8–:q; MRR ii: .o8).
¡6.z quae nunc nos . . . simili licentia usurpata Nunc, nos, simili are
all emphatic, conventionally taken to be F. speaking of his own discoveries in
comparison with those of Caelius. If this is so, we ought not to be left in doubt
over what is the antecedent of quae omnia usurpata. Easiest is the preceding vitia
(note leviora ceteris vitia in §¸), but vitia cannot be usurpata (OLD s.v. gives no
meaning even remotely applicable). Del Chicca (:qq¸) :o¸ n.:¸ argues that
the quae is a generalising plural which means vitia; this forces her to translate
usurpata as practised (‘praticati’), a sense that word cannot bear – despite what
I myself said on an earlier occasion: Rodgers (:q8.b) ¸¸6. Baldwin (:qq¡) ¡q¡
proposes that usurpata might be construed as ablative (absolute?) with licentia:
in that case vitia could be antecedent to quae, but we are still uncomfortable
with licentia having two modifiers when usurpata (esse) is the natural way to
read this sentence. If this is a direct quotation, syntactical precision is not
essential, for context makes clear what F. is talking about: illicit practices,
indeed, no matter how one interprets quae. With usurpata, however, we might
readily imagine iura. For iura usurpare I find no precise parallel (more than one
party, of course, is involved), but singular ius is well attested, in contexts which
are appropriately judicial: Livy, i.:¸.q, iii.¸:.¸, xxxi\.6..:¸, xrii.¸¡.:¸; Dig.
\, Nos and simili may be F. in reference to Caelius, but they could
equally well be Caelius in reference to one of his predecessors (e.g. Marcius
Rex, praetor in :¡¡ ncr, or Cato the Elder, censor in :8¡, on both of whom see
¸.:nn.). Nunc, equally, could be applicable to F. or to Caelius. So too, for that
matter, couldthe word-order withalliterative emphasis conveyedinquae nunc nos
per offensas ‘by taking action, even at the cost of personal offences’
(cf. OLD s.v. ¸). Del Chicca (:qq¸) :o¸ n.:¡ rightly declares that ‘by pointing
out flagrant transgressions’ or ‘par des mesures de rigueur’ (Grimal) will not
do here (tautology), and that the phrase bears exactly the same sense as at
:¸on. officii fidem etiam per offensas tueri; cf. TLL q..: ¡q¡.¸q. Roman magistrates,
‘old boys’ so to say, were called upon to, and often did, turn a blind eye to
reprehensible actions on the part of their fellows. By doing so they acquired
gratia, to which offensa here is largely antonymous.
inriguos agros . . . denique omnes Del Chicca (:qq¸) is willing to
attribute to Caelius only this part of the sentence, in which she notes a well
ordered crescendo of indignation: inriguos agros followed by a tricolon: tabernas,
cenacula, corruptelas.
inriguos agros The extra-urban use of water (cf. ¸¸..–¸) contrasts to
that within the City (tabernas, etc.). F. does not elsewhere use inriguus; cf. q¸.¸
agri vero qui aqua publica contra legem essent inrigati publicabantur.
tabernas . . . etiam Of locations within the City, Del Chicca (:qq¸) :o¡
observes that there is a gradatio fromstreet level to upper stories. For postponed
etiamemphatic, see TLL ¸..: q.¸.¸¸. If this passage can be attributed to Caelius
there is an interesting connexion with the aedile’s letter to Cicero linking
tabernarii and aquarii (cited above).
cenacula Possibly ‘upper rooms’: Varro, Ling. \.:6. ubi cenabant cenaculum
vocitabant, ut etiam nunc Lanuvi apud aedem Iunonis et in cetero Latio ac Faleris et Cordubae
dicuntur. posteaquamin superiore parte cenitare coeperunt, superioris domus universa cenacula
dicta. Del Chicca (:qq¸) adduces epigraphic evidence for collocation of tabernae
and cenacula (citing TLL ¸: ¸8o.66).
corruptelas denique omnes The concrete use of corruptela is weakly
attested: Plautus, Truc. 6¸o (cf. OLDs.v., TLL ¡: :o6¸.¸¸). Close are Plaut. Poen.
8¸o (apud lenonem) illic hominumcorruptelae fiunt, Livy, xxxix.:o.6 corruptelarumomnis
generis eam officinam esse. The position of omnis is uncharacteristic of F., except
perhaps for emphasis (see ¸¸..n. fistulae omnes).
perpetuis salientibus instructas Cf. the anonymous De dubiis no-
minibus GLK p.¸qo..: salientes ‘aquarum’ generis masculini, ut Caelius ‘perpetuum
salientem’. Cf. also q.qn. (Agrippa) compluribus salientibus {aquis} instruxit urbem.
Because of F.’s usage there, Del Chicca and Baldwin hesitate in attributing
these words to Caelius. It may be true that locum instruere + ablative is more
frequently attested in F.’s day than in Caelius’ (TLL ¸.:: .o:¸.8:). Cicero more
often uses the construction without the ablative (ibid. .o:8..¸), but very fre-
quently indeed he pairs it with ornare: of which note esp. ii Verr. ii.8¡ domum eius
exornatam et instructam fere iam iste reddiderat nudam atque inanem; Nat. D. ii.q¸ domi-
ciliis quae essent ornata signis atque picturis instructaque rebus his omnibus quibus abundant
i qui beati putantur; Tusc. iii.¡¡ (citing Ennius) vidi ego te [sc. Priami domum] . . .
COMMENTARY ¸6.¸–¸6.6
auro ebore instructam regifice. In Vitr. ii.8.:. we read (colonus) ad eum fontem propter
bonitatem aquae quaestus causa tabernam omnibus copiis instruxit eamque exercendo eos
barbaros allectabat. Del Chicca herself perceptively notes (:qq¸) that if instructas
is attributed to Caelius we have alliterative chiasmus: inriguos . . . instructas.
invenimus Can be either present or perfect (rhythm is indecisive).
¡6.j etiam si inter Giocondo printed etiamsi inter (from B
), but Reeve
(:q8: ) was the first to point out that this is in fact C’s reading (not etiam sunt). I
have taken the liberty of moving Grimal’s section-number (he began §¡ with
inter ea, the division marked in C).
¡6.q circa . . . Caelium et Aventinum Here circa seems to mean ‘in
the case of’; cf. ¸..¡n. circa erogationem.
¡6.j utebantur Marcia et Iulia At :q.¸ F. mentions a channel of Julia
fromSpes Vetus tothe Caelian; at :q.q he specifies that Marcia’s rivus Herculaneus
didnot serve the Caelian. Ashby (:q¸¸) :¸¸–6supposes that a high-level conduit
of Marcia once ran to the Caelian, supported on the Arch of Dolabella and
Silanus. Nero used that same arch for his arcus Neroniani, in the construction of
which he might have destroyed the high-level branch of Marcia. This branch
could have extended to the Aventine as well, either on arches or perhaps more
likely by a siphon. For the problem of identifying actual remains see Ashby
(:q¸¸) :¸6, Evans (:qq¡) 88. F. nowhere says that Julia supplied the Aventine,
but it might have done so if the high-level Marcia also carried Julia.
¡6.6 Nero . . . perduxit For Claudia’s arcus Neroniani, see .o.¸–¸nn.
Evans (:qq¡) :.o suggests that Nero might have been executing a plan al-
ready envisioned by Claudius to furnish water of good quality from the higher
Claudia – as good as Marcia – to regions hitherto deprived of that benefit or
inadequately supplied. For Nero to bring it on all the way to the Palatine would
make sense in the context of his building projects.
ad Spem exceptam B¨ ucheler’s correction is supported by .o.¸ partem
tamen sui Claudia prius in arcus qui vocantur Neroniani ad Spem veterem transfert; note
the parallel in :q.¸n. pars Iuliae ad Spem veterem excepta (which may in fact be Julia
in name only). Poleni’s altius has a superficial closeness to the paradosis; but
although at :8.¡ Claudia is said to be higher than Julia, this hardly explains
why F. should say it was raised higher on arches. Grimal 86 n.8¸ and (:q¸¸)
6.¸ would read a s<pe>cu{s}: the branch to the Caelian was taken directly
from Claudia’s channel and not from its terminal castellum: a trivial detail, and
irrelevant, since F. nowhere mentions whether branch aqueducts begin at such
tanks. B¨ ucheler’s restoration may be redundant (cf. .o.¸–¡), but it makes more
sense for F. to indicate the starting-point and the terminus (cf., e.g., ¸.8, :q.8,
...., 8¸.¡).
COMMENTARY ¸6.¸–¸8.:
non ampliatae sed omissae Dependence on a single supply involved
the danger of interruptions by repair work: see 8¸.¡–¸nn.
¡6.¡ vetus appellatio mansit Cf. Agrippa’s careful preservation of
Tepula’s name (q..n.).
¡¡.r–z Satis iam. . . ; superest ut . . . digeramus Cf. Cic. Leg.Man.
.o quoniam de genere belli dixi, nunc de magnitudine pauca dicam; .¸ satis mihi multa
verba fecisse videar, quare esset hoc bellum genere ipso necessarium, magnitudine periculosum:
restat ut de imperatore ad id bellum deligendo ac tantis rebus praeficiendo dicendum esse
Satis . . . dictum est Summarises 6¡–¸6 (cf. ¸..). The nova quaedam
adquisitio of which F. speaks is the maior copia discovered by his own measure-
ments (¸¡.:–¡); cf. 8¸.. quasi nova inventione fontium.
¡¡.z confertam. . . in massaminvenimus Contrasts with per nomina
aquarum . . . et per regiones urbis digeramus. It is hard to know exactly what F.
means by confertam in massam ‘lumped together’ (I can find no parallel for the
figurative use). Of figures for erogatio which were on record (6¡.., ¸¸.:; note esp.
¸¸.¸), there were totals for each aqueduct (6¸.¡, etc.) and distinction was made
between deliveries before or after the piscinae (66.¸, 66.6, etc.). F. has already
noticed the inaccuracy of distributions made under the wrong name (¸6.¸).
But the process of erogatio seems in part to have been based on data recording
the available supply (¸¸.¸, ¸¡.¸), and perhaps running totals of delivery were
kept in this connexion (cf.:oq.:). One might imagine that there were separate
(possibly chronological) listings of grants and deliveries. What F. apparently
missed was a presentation of these data in a form which would be useful for
administrative purposes. In digeramus, then, we should probably see F.’s own
initiative. The tabulationwhichfollows (¸8–86) was presumably drawnup soon
after he took office (cf. the ‘maps’ of construction types he mentions at :¸.¸–¡).
The imperfect tense is used throughout (¸8.¸, ¸q.., 8o.:, etc.; but see ¸q..n.,
8¸n.) because F. anticipates that the figures will be obsolete for his readers: see
¡8–86 Only a few of the figures in these chapters are demonstrably correct;
for the rest too little is known to showwhere the errors lie. It is very possible that
in some cases there were arithmetical tamperings prior to C (see ¸¸.¡–¸nn.).
Tables ¸–q set forth transmitted figures and acceptable emendations.
¡8.r quinariae DCCLXXI The definitioninthe relative clause supports
Poleni’s drastic emendation, but the corruption is still obscure. Perhaps quia
unus is a mangled form of quinariae and a transposition has further muddled
the error in numerals. The ¸¸: quinariae ‘delivered’ (i.e. transferred) fromone
aqueduct to another are:
COMMENTARY ¸8..–¸8.¸
Marcia to Tepula (6¸.¸, 68.¡) q.
Marcia to Anio (Vetus) (6¸.¸) :6¡
Julia to Tepula (68.¡, 6q.¸) :qo
Claudia to Julia (6q.¸) :6.
Anio Novus to Tepula (68.¡) :6¸
in adiutorium aliarum Cf. 6¸.¸ dabantur in adiutorium Tepulae.
¡8.z ex his dividuntur Table 6 lists the figures for the various categories
as a percentage of total delivery.
dividuntur The present (which follows from fit etc. in §:) is a vestige of
the original version (see ¸¸..n.); contrast distribuebantur and erogabantur in §¸.
nomine Caesaris For the meaning of this category, see ¸..n. In extra-
urban data, :,¸:8 quinariae distributed nomine Caesaris amounts to ¡. per cent
of the total ¡,o6¸: Evans (:qq¡) :¡o and Table 6. If one takes the figures for
high-level aqueducts only, the distributions nomine Caesaris amount to 6q per
cent: Marcia .6:/.6: (8:.:), Tepula ¸8/::¡ (8..:), Julia 8¸/.o6 (8¸.:), Claudia
.¡6/68¸ and Anio Novus ¸.8/¸.8 (86..). Cf. Taylor (.ooo) 8¡, who notes that
6q per cent is also the proportion for the combined Claudia-Anio Novus alone.
There are arithmetical inconsistencies in the data for intra-urban distribution
(between totals given at the outset and the sums from individual aqueducts
in chapters ¸q–86), but from §¸ we can reckon :,¸o¸ nomine Caesaris to be
:¸ per cent of the total q,q¸¸ (Table q). Overall, then, distribution nomine
Caesaris amounted to ¸,¡.¸ quinariae, or .¡ per cent of the total of :¡,o:8.
¡8.j reliquae <quinariae> I accept Schultz’s transposition and addi-
tion of quinariae, for these chapters are highly formulaic. The disorder may
also explain the loss of novem. The total of q,q¸¸ combines with the total ¡,o6¸
extra urbem to give the grand total :¡,o:8 (¸8.:). The subcategories total q,q¸¸:
nomine Caesaris :,¸o¸ (neglecting the semis!), privatis ¸,8¡¸, usibus publicis ¡,¡o:.
But when the latter is further subdivided the total is short by :oo quinariae:
castris .¸q, operibus publicis .¸o:, muneribus ¸86, lacibus :,¸¸¸ ( =¡,¸o:). Still, the
proportions can be adequately determined: Evans (:qq¡) :¡o–:. From §. we
have .,¸¡¸ for privati out of ¡,o6¸ extra urbem, or ¸8 per cent; within the city pri-
vati receive ¸,8¡¸ out of q,q¸¸, or ¸q per cent. Overall, grants to privati amount
to 6,:q. out of :¡,o:8, or ¡¡ per cent.
castella The transmitted total agrees precisely with figures for the indi-
vidual aqueducts (Table ¸). Castella were an essential feature of the hydraulic
engineering: see ¸..n.
castris The number is uncertain. C’s reading is perhaps in anticipation
of ducenta. It is not clear what F. means by castra. The numbers alone indicate
more than the castra praetoria, which was in F.’s day the barracks for the cohortes
COMMENTARY ¸8.¡–¸q.:
urbanae (RE Suppl. :o: :o.6) as well as for the praetorian guard, and might in
any case have received its water nomine Caesaris. The term castra might have
been applied to the stationes of the cohortes vigilum, on which see Rainbird (:q86);
cf. Blackman–Hodge (.oo:) ::q. Similar terminology might have been in use
for collegia who provided public services (cf. TLL ¸: ¸6:.::): castra is thus used in
the Notitia regionum for lecticarii, silicarii and the like, but their ‘military’ organi-
sation (cf. RE :.: :oq¸) may have postdated F. See Bruun (:qq:) .¸o–¸, Evans
(:qq¡) :o.
operibus publicis I have accepted, without much certainty, Poleni’s
adjustments (Table ¡) to give a total of q¸. The figure for Marcia I find most
worrisome: if only ¡: quinariae were delivered, :¸ may be too large a number
for opera publica.
muneribus A total of ¸q munera (¸..n.) is reasonable but not compelling
(see Table ¸).
lacibus Here Poleni’s emendation is virtually certain (..6 for Claudia-
Anio Novus at 86.¸). Figures for individual aqueducts (¸q–86) show that each
basin averaged between ..o¡ (Virgo) and ..¡¸ quinariae (Appia and Tepula).
The uniformity was calculated by Lanciani (:88:) ¸¸q–8o, discussed by Bruun
(:qq:) :o., Taylor (.ooo) ¡6–8.
¡8.q haec ipsa dispensatio I.e. the apportionment just given (§§.–¸).
For dispensatio cf. Livy, \ii..:.¸ solutionem alieni aeris in publicam curam verterunt
quinqueviris creatis quos mensarios ab dispensatione pecuniae appellarunt; x.::.q caritas
annonae . . . ad inopiae ultimum foret . . ., ni eius viri cura . . . tum in annonae dispensatione
praeparando ac convehendo frumento fuisset.
per nomina aquarum et regiones urbis Chapters which follow
(¸q–86) present the data per nomina aquarum, but there is no separate regionary
tabulation (see Table :o).
¡q.r <es>se posuimus See 6¡.., ¸8.:. F. is not fond of compounds of
ponere (¸.:n. ponam), and seponere is not apt.
dantur The present tense may be a slip (see ¸¸..n., ¸8..), although be-
cause of the technical circumstances (6¸.¸) F. can hardly have been including
Appia’s extra-urban deliveries among those capable of augmentation (88.¡).
†humilior . . . metitoribus† The sense should be akin to F.’s expla-
nation at 6¸.¸. I have kept the obelus where Krohn placed it. Although the
word humilior is almost certainly right (for the adjective used adverbially cf.
¸o.. lenior intrat, q. inferior excipitur), Dederich’s humilior <ori>tur, based on hap-
lography, is not perhaps so certain as Grimal thought (8¸ n.qo), for oritur seems
to require some fairly explicit reference to the source (:¸.¡, q¸.¸, syn. nasci-
tur :...; cf. .¸.:, :.6.: and ::q.., :.o, :...:) whereas humilior applies to the
level of the conduit (:8.¡, ¸¸). C’s dot between the a and m may indicate an
indecipherable letter, and the ti of metitoribus has a look of scribal uncertainty
(B¨ ucheler’s collator read meatoribus; one could also see mentoribus). The word
metitor (TLL 8: 888.¸o) has no sense in the context; but neither has petitoribus,
proposed independently by Willenb¨ ucher (:q.¸) :.¡6 and Kunderewicz, for
these cannot be the same as the petitores of :oq.:. Grimal’s quam ut adeatur rivus
boldly supplies the sense desired, but it can be justified neither stylistically nor
¡q.z per regiones secundamIIX VIIII XI XII XIII XIIII There is
no control for the transmitted numbers. For a scheme of regional distribution
according to C’s figures, see Table :o. Observe that C exhibits both ordinals
and numerals (8o.., 8:.., 8..., 8¸.., 8¡..) and with some inconsistencies, e.g.
IIX here but VIII at 8o.., 8:.., 8¸..).
8r.r privatis Here and at 86.. I follow Poleni in suspecting a lacuna
(cf. the same omission for Julia at 8¸..). In all three cases, of course, it is entirely
possible to fix the lacuna differently – here, for example, nomine Caesaris quinariae
, privatis quinariae> CCLXI S.
8q.z per regiones Agrippa owned property in all three of these regiones.
VII and IX were the site of several of his building projects. Water to Regio
XIV (Trans Tiberim) was piped across the river on the Pons Agrippae (see
operibus publicis Even subtracting the ¡6o quinariae for the Euripus,
the remainder is noticeably larger than other apportionments for opera publica.
Grimal (:q¡¸) :8: rightly stresses the importance of water in Agrippa’s urban
projects, a remarkable adaptation of the Hellenistic concept of baths and parks;
see, in general, Shipley (:q¸¸), Evans (:q8.), Roddaz (:q8¡) .8.–q¸, De Kleijn
(.oo:) 8:–..
8q.j Euripo. . . nomendedit Althoughthere were other euripi inRome,
the one in Agrippa’s park was the most familiar: Ovid, Pont. i.8.¸¸ stagnaque
et euripi Virgineusque liquor (cf. Tr. iii.:....), Strabo xiii.:.:q. Word-order here
is inverted to allow for the parenthesis. F.’s phrase means that this euripus was
called euripus aquae Virginis, or possibly just ‘Virgo’. It flowed from the stagnum
(Tac. Ann. x\.¸¸..) west and north towards the Tiber. Lloyd (:q¸q) thinks that
this euripus brought water to Regio XIV (Trans Tiberim); but conventional
thinking is that it emptied into the Tiber near Ponte Vittorio Emanuele: Bruun
(:qq:) ::¸ n.¸, Coarelli in LTUR ii: .¸¸–q.
8j extra urbem Note :8.8 Transtiberinae regioni et maxime iacentibus locis
servit. Taylor (.ooo) :q6–.oo plausibly argues that the deliveries were made
only after the water had been channelled through the regionary city (rather
than before, to properties high on the Janiculum).
COMMENTARY 86.:–8¸.¡
consumitur F. employs an unexpected present tense for Alsietina at
¸:..n., and Taylor (.ooo) :¸¡–¸ is right to observe that F. uses it because
he is not including Alsietina’s supply amongst those subject to improvement
based on his fresh measurements (¸:.:, 88.¡).
86.r proprio quaeque rivo See ¸..6, q:.¸.
8¡–qj Chapters 8¸–q¸ are not specifically announced in the prologue, but
they follow as editorial comments on what possibilities lie ahead for Rome’s
urban water needs and improvements. DeLaine (:qq¸) :.8 remarks on F.’s
rhetorical purpose, amounting almost to panegyric, and concluding (q¸) with
a ‘solemn recital of the emperor’s titles’.
8¡.r Haec copia . . . discribebatur Cf. 6¡.: quem modum quaeque aqua
. . . usque ad nostram curam habere visa sit quantumque erogaverit (note also imperfect
computabatur at 6¡..).
ad Nervam imperatorem usque For the anastrophe cf. :q.¸ ad
Viminalem usque portam, below §¡ in Aventinum usque, :o..: ad nos usque.
8¡.z intercipiebatur For the verb see ¸.¡n., ¸.:n. TLL (¸.:: .:6¡.q) cites
only one other instance of C’s spelling -cap- (fifth-century Cod. Iust.), for which
there is no justification here (pace B¨ ucheler, who introduced it).
inertia Rather than to the aquarii, as normally, F. may here be applying
the undesirable quality to senatorial curatores: cf. :o:...
quasi nova inventione fontium adcrevit Cf. 6¡.: deinde quem ipsi
scrupulosa inquisitione . . . invenerimus. The newsupply is the maior copia discovered
by F.’s own measurements (¸¡.:–¡, ¸¸.:, :o¸.¸).
8¡.j prope duplicata ubertas F. has computed the available supply
at .¡,¡:¸ quinariae (6¡.¡n. and Table ¸), whereas imperial records had shown
only :.,¸¸¸ (6¡..). Duplicata was conjectured independently by Holste and Fea
(:8¸.) ¸¸, ¸¡¸.
sedula . . . partitione Kunderewicz read parutione in C, but the ti (al-
though resembling u) is the same as found elsewhere (e.g. C’s partienda ¸8.¡).
There is a legalistic flavour to partitio here: Hern´ andez Gonz´ alez (:q8¸) .6¸
cites Cic. Caecin. :¸, Leg. ii.¡q–¸o (cf. Off. ii.¡o).
Claudia per arcus Neronianos See .o.¸–¡n.
celeberrimi colles sitirent Note the personification.
8¡.q Marcia reddita . . . in Aventinum Marcia’s ‘restoration’ seems
to apply only to the Caelian (which it had it once served): ¸6.¸n. Amplum opus
and a starting-point at Spes Vetus both suggest a high-level arcade rather
like the arcus Neroniani. No remains of such a branch have been identified:
Ashby (:q¸¸) :¸¸–6; Colini (:q¡¡) 8¸–8; Evans (:qq¡) 8q. Despite nunc and
F.’s present tense, the project might have been only a plan (cf. q. placuit, q¸..
COMMENTARY 8¸.¸–88.:
iussit), subsequently modified into a major pipe line or abandoned altogether
(cf. 88.¡n.).
8¡.j omni parte . . . binos salientes C has punctuation after veteres,
preserved by edd., but Ashby (:q¸¸) ¡¸ n.¸ rightly observes that plerique should
be taken primarily with veteres. Bruun (:qq:) :o¸ points out that data in chapters
¸8–86 yield a total of ¸q: lacus (¸..n.), each lacus supplied by water froma single
aqueduct. It thus seems that F.’s statement is anticipatory. New basins (88.:,
:o¸.¸) would have been designed in line with the policy enunciated here, and
F.’s point is that existing lacus could / would be in many cases improved by
the addition of a second supply. The phrase omni parte urbis reflects the sedula
partitio which aimed to provide reliable delivery throughout the City. The Fasti
Ostienses (Smallwood ..) report that with his new aqueduct in :oq cr Trajan
brought aquam . . . tota urbe salientem, in some cases perhaps supplementing
existing fountains: Bloch (:q¡¡).
88.r sentit hanc curam {imperatoris piissimi} Nervae principis
sui The redundancy of imperial titulature is probably owing to Peter the
Deacon (Introd. ¡¡ and ¸:..n.). We still need a subject for the active verb
sentire, for which regina et domina orbis will do well enough despite its flourish:
Baldwin (:qq¡) ¸o¡ cites Cic. Off. iii..8 (iustitia) omnium et domina ac regina virtutum
and notes two ‘Horatianisms’ (Epist. i.¸.¡¡ regia Roma, C. i\.:¡.¡¡ dominaeque
Romae); he points also to Flavian examples (Mart. \ii.¸o.: fons dominae . . .
regina loci, Stat. ii...:. Appia . . . regina viarum). ‘Frontinus might be regarded
as writing in the poetic idiom of his own day.’ Add Cicero, Tusc. ii.¡¸ domina
omnium et regina ratio, but especially Livy, xxx\iii.¸:.¡ sub umbra Scipionis civitatem
dominam orbis terrarum latere. The flourish indeed is what prompted the following
{quae terrarumdea . . . nihil secundum} The words resemble too
closely for accident the opening lines of Martial xii.8 (a poemwhich apparently
celebrates the accessionof TrajaninJanuary q8): Terrarumdea gentiumque Roma, /
cui par est nihil et nihil secundum. To Lipsius (:¸q8) i.. the solutionwas clear: ‘vereor
ut allitum adscriptumve aliena manu sit, et gravis atque eruditus reliquus
Frontini stilus nonprobat aut amat lasciviampoetarum.’ Dederich(:8¸q) :o8–q
concurs, noting that consistit ‘schmeckt nach sp¨ atlatein’. Kappelmacher (:q:6)
proposes rather that Martial alluded to Frontinus. Martial might have done, of
course, or conversely F. might have alluded to Martial. This is a case where our
acquaintance with Peter the Deacon allows us to forgo the dilemma (including
the matter of dates and locations: when did Martial write the poem? when did
he return to Spain?). Not only would Peter have been familiar with this poem
of Martial (it was widely known in the Middle Ages), but his training in the
abbey school of Monte Cassino embraced rhetorical studies that perpetuated
to some extent the florid prose of late Antiquity (e.g., the anonymous panegyrist
of .q: cr, Pan. lat. ::(¸).:..: ipsa etiam gentium domina Roma); cf. Willard (:q.q),
Bloch (:q¸.).
et magis sentiet<ur> salubritas The future sentiet nicely resumes
the sentit in initial position at the beginning of the sentence: elsewhere F. uses
futures to project, sometimes more subtly than others, a confidence in the
new regime. But nominative salubritas will not do if the verb stays the same
(an accusative would have been expected to balance curam). Poleni rightly
explains ‘salubritatem Vrbis sentiendam ex aucto numero Castellorum . . .’
Yet he failed to see that C’s text in fact is missing nothing more than a mark
of abbreviation to turn the sentiet from active to passive. When the subject
changes to salubritas, however, we need a defining genitive – for which eiusdem
aeternae urbis fits the bill. That collocation too was a commonplace (e.g. Livy,
\.¸.:o, Tib. ii.¸..¸, Ovid, Fasti iii.¸.), as was the urbs / orbs conjunction (e.g.
Cicero, Cat. :.q, etc.). See Gernentz (:q:8), Lugli (:q¸.) i: :oq–:., Mellor (:q8: ),
Arbagi (:q8¸), Edwards (:qq6) 86–8, q¸–:o.. I am still a little uncomfortable
with the position of in dies, which usually modifies a comparative or a verb
such as crescere, and might thus seem to fit more smoothly with magis. Yet
orbis in dies (– ∪ – ∪ x) produces a trochee-cretic (preferable to the hexameter
ending of et domina orbis), and et magis sentietur quite naturally invites a reader
to understand that the adverbial phrase continues its force into the second
88.j ne pereuntes quidem aquae otiosae sunt The general sense
is perhaps clear enough: see below (q¡.¸–¡, ::o–::) on aqua caduca. While the
sentence that follows is troublesome, Krohn’s revision is altogether too drastic:
ablatae causae gravioris caeli, munda viarum facies, purior spiritus, quique apud veteres
se<mper> urbi infamis aer fuit est remotus.
alia munditiarum facies, purior spiritus est ‘An entirely new
(fresh) guise, or appearance, of cleanliness, a cleaner air’. Asyndeton is no-
ticeable (hence Poleni’s iam), but not intolerable. Both facies and spiritus are
welcome novelties, while the second part of the sentence speaks of being rid of
some longstanding nuisance. There is uncertainty in taking alia in this sense
(why not, e.g., nova?), and that word elsewhere in our text seems sometimes to
signal the presence of annotation or interpolation (:¸.¸, ¡¡.., ¡¸..).
et remotae sunt causae . . . quibus . . . fuit Since causae and quibus
construe so well, Renaissance conjecture changed C’s est remotus (after fuit) to
sunt remotae. We should like a form of esse with the two earlier nominatives,
and the passive participle of removere belongs with causae. Words not seldom
rearrange themselves for better or worse.
gravioris caeli ‘unhealthy climate’ (OLD s.v. gravis 6 b). The compar-
ative bears this sense in Cels. i.¸.:, Colum. i.¡.¸, Sen. QNat. iii..¸.:o, Suet.
Tib. ¸6.: (for grave caelum as simply bad weather cf. Sen. Ep. ¸¸.:, QNat. i...¡).
COMMENTARY 88.¡–8q..
The noun-phrase gravitas caeli is not infrequent: Cic. Att. xi....., Vitr. i.¡.6,
Livy, xxiii.¸¡.::, Colum. i.¸.¡, Sen. Dial. xii.¸.8, Tac. Ann. ii.8¸.¡. On bad
climate and poor health cf. Cic. Div. i.:¸o pingue et concretum esse caelum, ut eius
adspiratio gravis et pestilens futura sit; Livy, \.:¸.¡ ex intemperie caeli . . . gravis pestilens-
que omnibus animalibus aetas. A similar concern for public health is reflected
in Dig. xriii..¸.:.. (Ulpianus) Curabit autem praetor per haec interdicta ut cloacae
et purgentur et reficiantur, quorum utrumque et ad salubritatem civitatium et ad tute-
lam pertinet: nam et caelum pestilens et ruinas minantur immunditiae cloacarum, si non
apud veteres The very early existence of a temple to Febris is by itself
good attestation of what F. means: Val. Max. ii.¸.6, Pliny, HN ii.:6; NTD :¡q–
¸o, LTUR ii: .¡¡ (Coarelli). On sanitation and hygiene in general, see :n.
saepe urbis infamis aer fuit Krohn conjectured se<mper>, but B
saepe is right: we should have heard elsewhere and often had the atmosphere
of the Romans’ urban space been constantly in ill repute.
88.q non praeterit me Krohn observed that this section might fit more
aptly at the end of chapter ¸¸. Yet in announcing the projects currently under
way, F. takes care to explain that the delivery figures he has given are those on
record when he took office (6¡.:). Ordinatio is used for delivery: cf. q. ordinari,
::..: ordinationem. The incrementumis that mentioned just above (§.n.). Nowhere
does F. speak of plans for an entirely new aqueduct, although the possibility
might have been contemplated in connexion with his review of the available
supply, and the decision to build the Aqua Traiana (completed in :oq cr)
must have been made not long after F. wrote. The scorn he shows for the
Alsietina (::.:, etc.) may in part reflect options being considered to deliver a
more wholesome supply to Trans Tiberim.
haec . . . adiunxerimus New tables to supersede those in chapters ¸8–
86. If such materials were compiled (under F. or his successors), they postdated
the publication of this booklet and have not survived.
8q.r sufficit parum praesidii <usibus> ac voluptatibus Cf. :¸.:
parum et publicis usibus et privatis voluptatibus . . . sufficere, .¸.: (copia) quae publicis
privatisque non solum usibus et auxiliis verum etiam voluptatibus sufficit.
tantamcopiam. . . sincerioremiocundioremque Chapters 8¸–8
have dealt with the quantity of the supply, in 8q–q¸ F. addresses improvements
in its quality.
8q.z operae pretium est A very common turn of phrase, found from
Plautus onward (in poetry appearing mainly in satire) and found frequently in
Livy. Its use here is mildly, but not specially, dignified; cf. Pliny, Pan. 6¸.¡ operae
pretium est adnotare, 86.: operae pretium est referre.
COMMENTARY 8q.¸–8q.¡
8q.j vel exigui imbres The solutions outlined by F. either were not
entirely successful, or they were not permanently effective. Cassiodorus,
Var. \ii.6 writes that Virgo alone is spared this detriment: nam cum aliae pluviarum
nimietate terrena commixtione violentur, haec aerem perpetue serenum purissime labens unda
8q.q aut quia . . . obtecta sint ‘nor because [waters] ought to suffer
this adverse effect – which are drawn from springs (especially Marcia and
Claudia), whose good quality at the intake is impaired either not at all or only
slightly by rainfall – provided that the shafts [along their course] are built up
and covered over’. Having eliminated the possibility of a general problem to
which all the waters (universis) are susceptible, F. now exculpates aqueducts fed
by springs. In qo he contrasts water taken directly from the Anio River, and
so his fontes here must refer primarily, perhaps exclusively, to those in the Anio
valley: those of Marcia (¸.¸, :...) and of Claudia (:¸.¸, :¡.:, :¸.¡). (Despite ¸.¸
and :o.¡, F. does not regularly use fons when speaking of the other aqueducts;
cf. 68...)
{ac reliquae} The passage in primis . . . reliquae is parenthetical. F. goes
on specifically to mention Claudia and Marcia for their excellence (q:.¸, ¸).
But if these two are special (in primis), why then should we have ‘and the rest’?
I suspect that a reader mistook the parenthesis to apply to fontibus (rather
than to quae capiuntur, i.e. the aqueducts) and then missed, if not the Caerulus
and Curtius (:¸.¸, :¡.:, ¸..¸), at least the fons Augustae (:..:, :¡.¸, ¸..8) and
the Albudinus (:¡..). Note that I have detected a reader’s intrusions when F.
discusses these same springs at :¸.¸–¸.
putea exstructa et obtecta F. uses the less usual neuter plural putea
(sing. puteus in the Lex Quinctia at :.q.::); see K–S i: ¡¸6–¸. Some shafts had
been made to facilitate initial tunnelling, others positioned at regular intervals
for inspection and cleaning (Vitr. \iii.6.¸). Shapes varied, sometimes round
and sometimes rectangular or square; workers and repair men descended
by foot-holes: Hodge (:qq.) :o.–¸. By exstructa F. probably means that the
openings were to be built up enough above ground level so that there was no
danger from run off of surface water and debris (also to avoid the misfortune
of accidental falls); see Hodge (:qq.) ¸.–¸. Covers might have been stone
slabs: Ashby (:q¸¸) 8¸, :¸: n.:; Hodge (:qq.) :o¸, :.¸. The piles of deposit
(cf. :...:) extracted from these putei have enabled modern scholars (beginning
with Lanciani) to trace much of the underground course of the aqueducts. The
traditional interpretation of this passage is that F. recommends constructing
covered well-basins at the springs: thus Grimal’s p<l>utea, but cf. Richmond
(:q¡6) .:q suggesting <puteis> p<l>utea. Some sort of masonry structures for
catchment purposes presumably existed in any case: note the signino circumiecto
for Virgo (:o.¸) and the piscina limaria of Anio Novus (:¸..); see also Ashby
(:q¸¸) q¸–6, :6.–¸.
COMMENTARY qo.:–q:..
inquinatur Cf. q:.¸ imperitia aquariorum . . . aquas inquinabat. The verb is
not often used of water: Varro, Rust. iii.¸.. aqua et inquinatur facilius et bibitur
inutilius; Hygin. Fab. :qq.. Circe . . . medicamentis aquam inquinavit.
qo.r sumuntur . . . turbantur Cf. :¸.: (Anio Novus) excipitur ex flumine,
quod cumterras cultas circa se habeat soli pinguis et inde ripas solutiores, etiamsine pluviarum
iniuria limosum et turbulentum fluit. Pliny, Ep. \iii.:¸ describes an autumnal flood
of the Anio. Blackman–Hodge (.oo:) ¸6–¸ point out that microscopic clay
particles will stay in suspension indefinitely and turbidity of this sort can only
be remedied by chemical means.
purissimo . . . lacu According to Pliny, HN iii.:oq the lakes
were actually three in number: lacus tris amoenitate nobilis; cf. q¸.¸ lacuum.
On the magnificent dams that created these artificial lakes, see Smith
mob<i>lit[ate] Earlier suggestions are largely inadequate: Rodgers
(:q8¸) :¸¸. What one expects here is a single word or a prepositional phrase
meaning either ‘in its intervening course’ (between the lake and the intake
of the aqueduct) or ‘by the effect of its current’. For mobilitate cf. ¸¸.6 quod
vis aquae rapacior . . . velocitate ipsa ampliat modum. At Str. i.6.. F. speaks of the
rapiditas of a river, and he is modestly fond of such abstract nouns (e.g. sancti-
tas ¡.., diversitas .¡.¸, ubertas 8¸.¸, salubritas 88.:, sinceritas qo.., proprietas q:.¸,
opacitas q¸.¸, maturitas :o¸.:, sedulitas ::8.¸, subtilitas ::q.¸, unitas :.¸.., aequitas
qo.z quo tempore exi<g>it<ur> . . . {exigitur} Cf. :...¸ aestate . . .
tempore quo praecipue desideratur (sc. usus aquarum). In light of C’s faulty division
(sinceritas. Exigitur), it seems that exiit has replaced exigitur and that the latter’s
appearance at the end of the sentence reflects incorporation of a variant read-
ing. Rhythm is better with a]quarum sinceritas (– – – –

x), variant of a double
sinceritas Not elsewhere applied to water. The word itself is rare: once
in Columella (i\..6.. plant growth), twice in the Elder Pliny, HN x\... pura
sinceritas (of equipment for olive-pressing), xxxi.qo salinarum sinceritas; elsewhere
usage tends to be metaphorical (e.g. Sen. Dial. \ii.:¸.:, Pliny, Ep. iii.::.6, and
often in Val. Max. ii.6.8, :o(ext.).., etc.).
qr.r libra sit inferior For the relative level of Anio Vetus see :8.6.
qr.z editissimus . . . abundans For the height see :¸.¸, :8.¡, .o.:; for
the quantity ¸¸.:–6. – In Str. F. is fond of editus: i.¸.:: editum collem, ii...., ¸
locum editum, ..¸ ex edito, ¸.:¸ in loco edito, :¸ in saltu editiore.
qr.z defectioni aliarum The word defectio in the sense of ‘shortage’
(of water) is very rare: aside from this passage TLL ¸: .qo.¡¸ cites only Sen.
QNat. i\...:6 (of the Nile).
COMMENTARY q:.¸–q:.¸
qr.j imperitia vero aquariorum Imperitia is ablative (note 8¸..
fraudibus and inertia, and cf. §¡ per imprudentiam); the subject of inquinabat is
Anio Novus (as in §. vitiabat). At ¸..¡ F. charges fraus; cf. ¸..6 miscebatur . . . ut
confusione facta et conceptio earum et erogatio esset obscurior.
deducentium in alienos eum specus An example of what F. de-
scribes has been found at Grotte Sconce (not far below Tivoli), where there
is a large castellum from which water from Anio Novus could be supplied to
each of the lower aqueducts (Claudia, Marcia, Anio Vetus): see Van Deman
(:q¸¡) ¸o:–¸, Ashby (:q¸¸) .¸¸–q, Hodge (:qq.) :.o–:, Aicher (:qq¸) :¸6–q,
Chanson (.ooo) ¸¸.
maxime Claudiam For Claudia’s supply see ¸..:n.; this included the
water from the fons Augustae (:..:) usually diverted into Claudia (¸..8). There
was also a surplus at Claudia’s source (¸..¸–8).
in hoc tempus perdebat The imperfect implies a projected change.
But the single terminal castellum (.o..n.) which served both Claudia and Anio
Novus seems never, in fact, to have been altered (¸..6n.).
qr.q adeoque . . . partientium ‘So far was it (the water of Anio
Novus) from being a help to the aqueducts into which it was turned that
some of these were then called upon (for unworthy purposes) through the
thoughtlessness of those who arranged for distributions.’ A further result of
the pollution from Anio Novus is that water in general is distributed unwisely,
without regard for its quality. I take succurrebat{ur} to be active, with Anio
Novus as its subject (cf. §. defectioni aliarum succurrit), which makes obvenien-
tibus (‘waters which come in contact with it’) less troublesome. Succurrebatur
can be defended as an impersonal passive, but the type of error is frequent
in this text. I print ut {in}dignum because uti is not F.’s usual form and it is
easy to imagine a mechanical slip (perhaps because of the non or through
qr.j Marciam ipsam . . . servientem Baldwin (:qq¡) ¸o¸ remarks
that in this sentence F. ‘deftly combines high style with sardonic moralising’.
Note that ipsam conveys emphatic indignation; cf. q. in primis Marcia potui tota
et rigore et splendore C’s splendore is an autocorrection. Inappropriate
placement of et before gratissimam justifies Krohn’s transposition. Editors have
apparently taken the ï before rigore as a mistake for initial f in frigore, but it could
as easily be a misreading of abbreviated et, and rigor itself is entirely suitable in
context (OLD s.v. .c); cf. Pliny, Ep. \iii.8.¡ (on Clitumnus) rigor aquae certaverit
nivibus. See also q¸.¸n. frigidissimus simul ac splendidissimus.
relatu quoque foedis ministeriis Lipsius (:¸¸¸) i.:. ‘Ministeria enim
haec ad libidinem intellegit.’ Almost certainly F. has in mind purposes more
despicable than street cleaning or sewer flushing, for which, along with baths
and fullers’ shops, aqua caduca was used (q¡.¡, :::..). Note the more dignified
sordidiora ministeria just below(q.). The semantic range of foedus is wide enoughto
embrace a multitude of uses, perhaps chief among themlatrines: Taylor (.ooo)
:¸¸; onforicae ingeneral D–S¸..: q8¸–q: s.v. ‘latrina’ (Th´ edenat), Hodge (:qq.)
.¸o–.. The discovery (deprehendimus: cf. ¸o.¡n.) may, on the other hand, involve
water used by privati, perhaps for less than respectable uses akin to those in
¸6..; cf. Tac. Hist. ii.q¸.: on Vitellius’ soldiers entertained per inlecebras urbis
et inhonesta dictu. Private bathing establishments also come to mind, and the
prurient may wish to consult Cameron (:q¸¸).
qz pluribus ex causis . . . minus salubris Of the ‘several reasons’
for assigning Anio Vetus to more mundane uses the most important was the
judgement of its low quality (‘the further downstream water is received, the
less wholesome it is’), and – save for Alsietina (::.:) – Anio Vetus was the least
appealing. Improvements could perhaps have been made at the intake, but
these might have affected the supply for Tibur (6.¸, 66..): Evans (:qq¡) 8:.
They could in any case have proved to be impractical, for the aqueduct was
very old (cf. :8.6) and there were good uses to which its water could be put.
sordidiora . . . ministeria On the basis of this passage Lanciani (:88:)
.6q identifies the rivus Octavianus of Anio Vetus which served the horti Asini-
ani (.:..) with the Aqua Damnata listed in the Curiosum Urbis (ed. Valentini-
Zucchetti, p.:¸¸.:.). Evans (:qq¡) 8: very plausibly suggests that F.’s ‘meaner
uses’ might have included watering troughs for animals.
qj.r Nec satis fuit principi nostro . . . Anionis quoque . . .
vidit ‘Not only . . .but also . . .’ The fuit + perfect infinitive contrasts to
the kind of generalising present sufficit seen at 8q.: (see :¸.¸n.). The prince is
Nerva, as consistently above (8¸.:, 88.:, 8q.:).
qj.z <aquam> repeti Heinrich’s addition is preferable to Schultz’s ad-
justment limpidissimus (sc. Anio); cf. qo.: limpidae. It is possible that the trans-
mitted text could stand, although I find the ellipsis of aqua to be more difficult
here than elsewhere (cf. q.., q¡.¸, :::.:, etc.).
repeti . . . iussit On the archaeological evidence for the newintake, see
Van Deman (:q¸¡) .¸¡–¸, Ashby (:q¸¸) .¸¸–6, Smith (:q¸o). For postponed
iussit cf. Str. i.::..: incendi eas [sc. naves] priusquam iniret proelium iussit.
villam Neronianam Sublaquensem Cf. Tac. Ann. xi\..... discum-
bentis Neronis apud Simbruina stagna <in villa> cui Sublaqueum nomen est and Pliny,
HN iii.:oq Anio in monte Trebanorum ortus tris lacus amoenitate nobilis qui nomen
dedere Sublaqueo defert in Tiberim. These references to Sublaqueum(modern Subi-
aco) suggest that there was no settlement there prior to Nero. F.’s text is the
only secure attestation of Nero’s villa; at ¸.6n. he records Nero’s part in con-
structing the Via Sublacensis. On the archaeology of Nero’s villa, see Tomei
COMMENTARY q¸.¸–q¸.¡
(:q8¡), Mari (:qq¸) ¡q–¸., Fiore Cavaliere–Mari–Luttazzi (:qqq), Mari–Fiore
Cavaliere (.oo:).
qj.j Trebam Augustam An ancient community of the Aequi, modern
Trevi Laziale (RE 6A: ..¸o). The epithet Augusta is not otherwise attested.
It may possibly be a gloss (referring to what is modern Aosta, a community
located below Subiaco).
paucis . . . cultis Contrast :¸.:, where the river below Subiaco terras
cultas circa se habeat soli pinguis et inde ripas solutiores. There seems little point to
a superlative paucissimis. Rubens (:6o8) ii.¸¸ proposes oppidum, which I accept
with some reservation (he adds, ‘vel possit etiam legi “paucis circa ipsum
opido”, id est, opido paucis comica phrasi’).
lacuum . . . in quos F. elsewhere uses the singular (qo.:, q¸..) in ref-
erence to the single lake from which the water was drawn. All three pools,
however, served as settling reservoirs.
velut defaecatur The verb is rare, even in its literal use (with vinum:
Colum. xii..8.¸, ¸¸.:, Pliny, HN x\iii..¸., Serv. ad Georg. i.¸¡:). Servius ap-
plies it also to honey (ad Georg. i\.:o., Aen. i.¡¸.). Other uses are more dis-
tinctly metaphorical (e.g. Plaut. Aulul. ¸q defaecato animo, Pseud. ¸6o defaecatum
nemorum opacitate inumbratus Although all three words have a
poetic flavour, they are familiar in post-Augustan prose (OLDs.vv.); cf. Baldwin
(:qq¡) ¸o¡. F. himself writes lucum densissimae opacitatis (Str. i.::.:o); cf. Plin. HN
xxxi.¸q frigori et opacitas necessaria; Pliny, Ep. \iii.:¸.¸ (of the Anio) nemora quibus
inumbratur. Of the poets, note Virg. Ecl. :.¸. fontis sacros frigus captabis opacum,
Ovid, Her. .:.:¸¸ in umbroso fonte (for other examples see TLL 6: :o.¸). Compare
also Curt. Ruf. iii.¡.q (of the Cydnus) non spatio aquarum, sed liquore memorabilis:
quippe leni tractu e fontibus labens puro solo excipitur, nec torrentes incurrunt, qui placide
manantis alveum turbent. itaque incorruptus idemque frigidissimus, quippe multa riparum
amoenitate inumbratus.
frigidissimus simul ac splendidissimus Cf. q:.¸n. et rigore et splen-
dore. Athough rigidus ‘cold’ is applied to water by Martial and Statius (see OLD
s.v. .c), the superlative of the adjective in this sense is unattested. There is a
slight possibility that these superlatives are ‘improvements’ on the part of Peter
the Deacon, who was intimately familiar with the words frigidas atque perspicuas
used by Gregory the Great in his life of St Benedict (Dial. ii.:) to describe the
waters of the Anio at Subiaco (see Introd. ¡¸ n.:¡¡).
qj.q copia vero superatura F. had reckoned ¡,6qo quinariae at Marcia’s
source (6¸..) and ¡,¸¸8 at the intake of Anio Novus (¸¸..). The renovations
will increase the supply of the latter even further, but for this F. has as yet no
data (cf. 88.¡).
COMMENTARY q¸.¡–q¡..
novumauctorem. . . titulo No inscriptionhas beenfoundthat records
these improvements to Anio Novus. An obvious location would have been on
Porta Maggiore, the same arch that proclaims Claudius as the original auctor
(Appendix B, no.¸). But all available space there had been taken by inscriptions
of Vespasian and Titus, the purpose of which had been to advertise the Flavian
{Traianum} Work can scarcely have begun at the time of Nerva’s death,
and credit for completion would fall to Trajan. The new auctor (¸.:), however,
was really Nerva (8q.:, q¸.:), not Trajan, and F. will have appreciated that pietas
required Nerva’s heir to record that fact. If Trajan took credit for himself, he
would have done so as Nerva’s successor: e.g., CIL :o.68.¡ = ILS .8o Imp.
Caesar Nerva Aug. . . . sua pecunia incohavit, imp. Caes. Nerva divi Nerv[ae f.] Trai[a]nus
Aug. . . . consummavit; CIL :o.6q.6–.8 = ILS .8¸ Imp. Caesar divi Nervae f. Nerva
Traianus Augustus . . . incohatam a divo Nerva patre suo peragendam curavit. To have
inserted Trajan’s name would not have been beyond the capacity of Peter the
Deacon: Bloch (:q8¡) 66–¸q. Note also how Peter interpolated Veg.
<Theodosi> imperator and that he was aware of F.’s relationship with Trajan
(Introd. ¡¸ n.:¡¸).
qq–rjo The remainder of the booklet addresses legal matters, announced
in the prologue (¸..) as ius <ducendarum> tuendarumque aquarum and attendant
poenae lege, senatus consulto et mandatis principum inrogatae. Although to some extent
the ius aquae ducendae overlaps with tutela ductuum, chapters q¡–::¸ focus on the
former (including Republicanpractices andthe organisationof the cura aquarum
under Augustus), while chapters ::6–¸o describe standing crews for upkeep
of the aqueducts and the importance of the clearway along its entire route.
By quoting verbatim and at length F. has quite accidentally preserved legal
documents of great interest to posterity, whether or not it was his intention to
parade his legal knowledge (or his antiquarianism, for some points have more
rhetorical value than contemporary relevance); cf. DeLaine (:qq¸) :.8–q.
qq.r ius ducendae tuendaeque Cf. q¸.: ne quis violare ductus aquamve
non concessam derivare auderet. From an administrator’s viewpoint the two facets
were interrelated. Ius ducendae could be usurped by tapping into the supply;
tutela included watch over such illicit actions as would damage the conduits.
Cf. Sen. Dial. x.:q.: (on the responsibility of the grain prefect) ut incorruptum a
fraude advehentium et neglegentia frumentum transfundatur in horrea, ne concepto humore
vitietur et concalescat.
qq.z altius repeto The adverb altius is frequent when repeto is used in a
context of research (OLD s.v. ¸). Discussion of obsolete practices is perhaps
partly due to F.’s personal antiquarianism and his interest in legal matters,
but Republican austerity highlights his recurrent emphasis on the justice and
generosity that a princeps displays towards his subjects (cf., e.g., :, ::.:, :¸.:,
88.., 8q.:, qq.¸, :o¸.¸, :o¸.¸, :oq.¸, :¸o.¸). Cf. Sall. Cat. ¸.q res ipsa hortari
videtur, quoniam de moribus civitatis tempus admonuit, supra repetere ac paucis instituta
maiorum domi militiaeque, quo modo rem publicam habuerint quantamque reliquerint, ut
paulatim inmutata ex pulcherruma <atque optuma> pessuma ac flagitiosissuma facta sit,
leges de singulis datas ‘Regulations established concerning individual
persons’. Bruns ::: categorises this and the quotation in q¸.¸–6 below as leges
dictae (as opposed to statutes, leges rogatae), hence Crook, in Crawford (:qq6)
¸.q, tentatively proposed that we read datas here. Similarly, Mateo (:qq6) .8q–
q8 argues that F. is speaking of leges censoriae issued for construction and/or
maintenance of the water-system. General prohibitions could form part of
such leges: Dig. xxxix.¡.:¸.pr. (Alfenus) Caesar cum insulae Cretae cotorias locaret,
legem ita dixerat ‘ne quis praeter redemptorem post idus Martias cotem ex insula Creta fodito
neue eximito neue auellito’ (cf. also q¸.¸ below in isdem legibus adiectum est). B¨ ucheler
conjectured <a>qui<s> lata<s>, but it is hard to credit a corruption of aquis.
One could just as easily read dictas: for the interchange d/p cf. 8¸.¸ duplicata]
publicata C, and for a/ic see :.q.::n. coclea. F. seems to be addressing the ius
ducendae aquae, the first category mentioned in §: (ad cohibendos . . . privatos). For
singuli = privati cf. .¸.¸ ex quo [sc. castello] singuli suum modum recipiunt. It may
be noted, however, that in q6 we have tutelam singularum aquarum locari solitam
invenio, which might lend support to B¨ ucheler’s conjecture.
inveni<o> F. uses the present tense (and first-person singular) for refer-
ence to documents or sources: the closest parallels (cf. OLD s.v. invenio ¡) are
q¸.. permissum invenio, q6 tutelam . . . locari solitam invenio, :oq.¡ constitutum invenio,
but note also legimus (¸.¡, q¸..) and invenitur (:o..).
qq.j eroga<ba>tur et cautum ita fuit Note the contrast between
tenses. For the perfect rather than imperfect of esse as auxiliary to empha-
sise past time (as opposed to present perfect), cf. ::..¸ [fistulae] subiectae fuerunt,
::¸.¸ [calices] positi fuerunt, and see K–S ii.:: :6¡, H–Sz ¸.o–..
ne quis privatus . . . accidit RS ¡¸, in which Crook (in Crawford
(:qq6)) compares the concernwithaqua caduca withthe Lex Coloniae Genetivae
(RS .¸, Bruns .8, FIRA .:), c.:oo si quis colon(us) aquam in privatum caducam ducere
volet isque at IIvir(um) adierit postulabit<q>ue, uti ad decurion(es) referat, tum is IIvir,
a quo ita postulatum erit, ad decuriones, cum non minus (quadraginta) aderunt, referto. si
decuriones, m(aior) p(ars) qui tum atfuerint, aquam caducam in privatum duci censuerint,
ita ea aqua utatur, quoe [sic, em. to quot] sine privati{m} iniuria fiat, i(us) potest(as)que
e(sto). Persons by definition not privati, i.e. magistrates and officials of the res
publica, presumably would include those whom F. calls principes civitatis (below,
COMMENTARY q¡.¡–q¡.¸
aliamducit<o> The antecedent of aliamcan of course be nothing other
than aqua, expressed no doubt earlier in the sentence or in a preceding clause of
the law. Whether F.’s lex is a statute (rogata) or a lex dicta (see §.n.), the imperative
is appropriate. C’s subjunctive is either a slip on F.’s part or an assimilation to
the sentence as a whole: Crook in Crawford (:qq6).
haec enim sunt verba Indicates verbatim citation; cf. :o¸.¸ haec verba.
caducam vocamus Cf. ::o.:; Varro, Rust. iii.¸.. gives the same defini-
tion: quae abundat. Besides the use of accidere in the quotation here, note that
cadere is used alongside ire and fluere in the Lex Quinctia (:.q.¡). On this over-
flow water, see Bruun (:qq:) ¸¸ n.¡¡ and ::o; on sewers and drains, Hodge
(:qq.) ¸¸¸–¸.
qq.q et haec ipsa . . . dabatur Some privati did receive public water, but
only for baths and fulleries (commercial uses that responded to public needs).
Agrippa’s baths (:o.:n.) were the first to be public in the sense that after his
death they were owned by the state. Other bathing establishments (cf. :o¸..
balneis quae publice lavarent) were privately owned, although subject to regulation
by the aediles (cf. Sen. Ep. 86.:o); Agrippa had personally subsidised gratuita
balinea during his aedileship: Pliny, HN xxx\i.:.:, on which see Fagan (:qq¸).
balnearum aut fullonicarum . . . vectigalis A charge for water
delivered to baths is reflected in the prescription of Vitruvius \iii.6..: in balneas
vectigal quotannis populo praestantes (Callebat: praestant vulg.).
qq.j aliquid B¨ ucheler suggested transposing the sentence ex quo . . . per-
tineret (q¸.¸) to precede aliquid – in other words, before F. mentions individual
persons. But the comment fits better in its transmitted context, an editorial
remark on the entire Republican practice, before F. turns from ius ducendae to
in domos principum civitatis Contrasts with the law cited in §¸ and
the general statement of §¡: the emphasis is on in domos (cf. Strabo \.¸.8 cited
above :¸.:n.). Both dabatur (¸..n. detur) and concedentibus (cf. q¸.:, :o¸..) imply
a formal and legal procedure – that in fact which is outlined in q¸.:–.. F.
leaves it unclear whether these parties also paid a vectigal (cf. Vitr. \iii.6..); but
the status of the recipients and the nature of later imperial beneficia (:o¸.:n.)
lead one to suppose that these grants were made without charge (as a token
of recompense for service to the state). For principes civitatis, members of the
elite classes, see chapter :n. That things had not changed in any remarkable
way under the principate is demonstrated by studies of fistular stamps, e.g. Eck
(:q8.c), Bruun (:qq:), De Kleijn (.oo:).
concedentibus reliquis The reliqui are quite possibly other principes
civitatis, for it is unlikely that they are members of the citizen-body as a whole
(by means, e.g., of a lex rogata). For Mommsen’s suggestion, see q¸.8n.
qj.r dandae vendendaeve The words are not specially contrasted. The
question is which officials were authorised for distribution in general, either of
water for which general permission was granted (dandae) or of that for which
a specific charge was levied (cf. q¡.¡ vectigalis). The ius dandae ‘right to give’ lay
with magistrate; ius ducendae with the recipient.
in . . . legibus . . . variatur Perhaps because different magistrates
issued such leges under different circumstances (q¡..n.). For comparison we
have two texts from Roman coloniae, in each of which the matter was under
the care of the decuriones. At Urso we find si quis colon(us) aquam in privatum
caducam ducere volet . . . (quoted q¡.¸n.); at Venafrum, CIL :o.¡8¡..¸¸–¡¸ ( =
ILS ¸¸¡¸.¸¸–¡¸), we have quaeque aqua in oppidum Venafranorum it fluit ducitur, eam
aquam distribuere discribere vendundi causa aut ei rei vectigal inponere constituere, IIviro
IIviris praefec(to) praefectis eius coloniae ex maioris partis decurionum decreto . . . legemque
ei dicere ex decreto decurionum . . . ius potestatemve esse placet.
qj.z ab illis potissimum F. seems to be drawing a general rule: when
censors were in office they had the ius dandae vendendaeve aquae, but when there
were no censors the same potestas fell to the aediles. Exceptions no doubt existed
(hence potissimum), either because of the legal definition (§:) or perhaps because
in the later Republic censorial responsibilities sharply diminished: Mommsen
(:88¸) ii: ¡.6–¸. There were in fact censors in office when Caelius as curule
aedile in ¸o ncr took action against tabernarii and aquarii (see ¸6.:n.). The
ambiguity between censors and aediles recurs below (q6, q¸.:). For the legal
responsibilities of censors in water management, see Kunkel–Wittman (:qq¸)
cum autem non erant For the particle used for noting distinction cf.
.¸.. qui autem Agrippam, :oo.. cum autem in urbe: Rodgers (:q8¸) :¸¸.
q6 Tutelam . . . locari solitam Contracts were let for maintenance
of individual aqueducts (cf. q¡..n.), at first no doubt occasionally as need for
repairs became apparent (cf. ¸.:n.), but later perhaps at regular intervals. A
comprehensive system seems not to have existed before Agrippa. Locatio was
the only available procedure before the formationof Agrippa’s familia (q8.¸), for
public slaves were not employed in this work: Eder (:q8o) q6–¸. For some work
contractors were still employed in F.’s day (::q.¸), despite the existence of the
second familia Caesaris (::6.¡). In Republican times contracts for maintenance
as well as for building aqueducts would normally have been let by censors
(¸.:n.), and thus to them fell the responsibility for approving the work (operum
probandorum curam; cf. 6..n.). Compare Cic. Leg. iii.¸: Censores . . . urbis templa vias
aquas aerari vectigalia tuento; cf. Mommsen (:88¸) ii: ¡¸.. For aediles see q¸..n.
singularum aquarum Thus apparently F. distinguishes the Republi-
can administration (when the concern was for single waters as need arose)
from the system-wide cura established by Agrippa (see q8–q). Augustan
documents use the collective plural aquae (:oo.:, :o8, :.¸.:, :.q.¡; cf. :o¡.: pub-
licorum salientium, :o6.: rivis publicis), and note especially the comprehensive list
at :.¸.
positamque redemptoribus . . . deferrent F. presumably refers to
the stipulations in a lex locationis: RE :¸: q¸¸; Mommsen (:88¸) ii: ¡¸o; cf. Bruns
¸¸o–¸. Contractors (redemptores) provided their own slave workers (servi opifices),
of whom specified numbers were required for the extra-urban courses of the
aqueducts and for work within the City. The names of such workers were to be
entered in the public records according to the locality to which they had been
assigned. F.’s use of urbs here probably reflects legal usage (¸..n., cf. :o¡.:n.),
and regiones (which cannot refer to the Augustan wards) may relate in general
to the vici mentioned at q¸.8 below. The procedure for registration reveals a
state control over private contractors beyond that of the standard probatio. For
a system in Roman Egypt which involved a uvnucvtïcv, written agreement
with copy deposited in the city archive, see Aubert (:qq¡) :: and n.¸¸.
servorum opificum The attributive noun is unusual (cf. :o.. puella vir-
operum probandorum curam Contracts were let for a stated period
of time, at the end of which the work was inspected and final payment made.
Normally the officials who had let the contract would perform the probatio
(6..n.). The point here is that legal provision had been made for probatio in
those cases where the contractors’ term extended beyond the magistracy of
the officials who had let the contract. For some general discussion of contracts
for building projects see Martin (:q86), Rainer (:qq.).
interdum . . . quaestoribus Quaestors may have been involved be-
cause of financial aspects of probatio. But the quaestors’ role is apparently a
distinct exception: Hirschfeld (:qo¸) .¸¡. Aside from the unknown quaestors
to whom F. refers here, we have only an ambiguous reference to P. Vatinius,
quaestor in 6¸ ncr: Cic. Vat. :. in eo magistratu cum tibi magno clamore aquaria
provincia sorte obtigisset (where it is not at all clear that the provincia aquaria is that
of inspecting contractors’ repair work): Kunkel–Wittman (:qq¸) ¡88.
C. Licinio et <Q.> Fabio consulibus ::6 ncr. For the loss of Q. after
et see 6.¸n. C has consule after Licinio and censoribus after Fabio; this looks to
be no more than normal scribal misunderstanding (or, perhaps, meddling).
It happens, however, that Licinius and Fabius were censors in :o8 ncr and
Broughton MRR i: ¸¸: n.: thinks that F. more probably refers to their cen-
sorship. If Pithoeus’ S.C. is correct for the transmitted eo, on the other hand,
consular dating is virtually certain. Furthermore, such dating would make log-
ical sense for archival purposes: Coudry (:qq¡) ¸o cites Cic. Att. xiii.¸¸.¸ ex eo
libro in quo sunt senatus consulta Cn. Cornelio L. [Mummio] coss. Krohn proposed a
lacuna after censoribus, which he took to be the first word of a quotation from
the S.C. (The Loeb edition has gone completely astray at p.¡.¸ n.¸.)
COMMENTARY q¸..–q¸.¡
q¡.z quod durasse The tense makes clear that the practice had since
been abandoned. Parsimony might have been originally at issue, for Regio
XI was served by Appia alone prior to the introduction of Claudia and Anio
Novus (:o¡.¡, Table :o). Humphrey (:q86) ¸¡, 6¡¸ n.qq observes that control
made good sense applied to large euripi dug under Caesar to separate audience
from action (Pliny, HN \iii..:, Suet. Jul. ¸q.., Dion. Hal. Ant. Rom. iii.68..).
Water needs might have been smaller at the Circus after these euripi were
removed under Nero (Pliny, HN \iii..:). With more copious supplies available
in the mid-first century, the restriction could have been lifted, although water
remained an important feature of the central barrier: Humphrey (:q86) :o:,
sub Augusto The phrase is in an odd position, and could be a gloss
(derived from qq.¸–¡). Alternatively, it might be moved to follow etiam.
apud Ateium Capitonem A distinguished lawyer: Tac. Ann. iii.¸o.¸
humani divinique iuris sciens; Gell. x..o.. publici privatique iuris peritissimus. Note
also his obituary in Tac. Ann. iii.¸¸ Capito Ateius, principem in civitate locum studiis
civilibus adsecutus . . . illa aetas [sc. Augusti] duo pacis decora simul tulit: sed Labeo
incorrupta libertate, et ob id fama celebratior, Capitonis obsequium dominantibus magis
probabatur. Curator aquarum:¸–.¸ cr (:o...–¸n.). The work to which F. refers
was probably the Coniectanea (Gell. i\.:¡.:: S–H ii: ¸¸¸).
q¡.j agri . . . publicabantur Our ancient sources explicitly report no
cases of property confiscated as a penalty for using public water, although there
were instances of water recovered from privati (¸.:n.). Publicatio was entirely
distinct from expropriation for public utility (pace Berger 66.): Scapini (:qq8)
.8–¸o. The word inrigati may hint at a dichotomy between agricultural and
domestic use (cf. ¸6.. inriguos agros, also Livy, xxxix.¡¡.¡ aquam publicam in
privatum aedificium aut agrum fluentem).
q¡.q mancipi . . . multa dicebatur There was a fine for the contractor,
even if it was established that someone had broken the law without his knowl-
edge. The manceps can hardly be other than one of the redemptores of chapter
q6, ‘dem
Ubernehmer der Instandhaltung’ Mommsen (:88q) 8.¡ n.¸, with
<constaret> as printed; cf. De Kleijn (.oo:) q8 n.:¸. It is hard to make C’s cum
eo mean ‘with his complicity’ (Grimal), and this sense (a fine if he knew) is
considerably weaker than that derived from clam eo (a fine even if he did not
know). Mommsen (:88¸) ii: ¡66 n.¸ had earlier proposed si con<staret sciente>
eo quem . . . , with which etiam is awkward: would there be two persons fined?
Or would the manceps be responsible for the actions of ‘anyone’? It is possible
that cum eo is a corruption of constaret: a fine even for the contractor as well,
merely by establishing that someone had broken a law – with or without his
COMMENTARY q¸.¸–q¸.8
multa dicebatur A fine was imposed by the appropriate magistrate: for
this judicial sense of dicere ‘pronounce, declare’ note :.q.q multae dictio (cf. iuris
dictio) and see OLD s.v. :oa. For multa (distinguished from poena ¸..n.): Varro,
Ling. \.:¸¸ multa <e>a pecunia quae a magistratu dicta, ut exigi posset ob peccatum; cf.
Cic. Leg. iii.6; Geißler (:qq8) .¸¸–q. Note further Dig. r.:6.:¸:.: (Ulpianus)
inter ‘multam’ autem et ‘poenam’ multum interest, cum poena generale sit nomen omnium
delictorum coercitio, multa specialis peccati, cuius animadversio hodie pecuniaria est; cf.
Mommsen (:8qq) :¡ n.., Leibs (:q68) :8¸.
q¡.j–6 ne quis aquamoletato dolo malo ubi publice saliet. si quis
oletarit, sestertiorum decem milia multa{tum} esto An uneasi-
ness remains that we should perhaps read multatus esto (resuming ne quis oletato
and si quis oletarit); alternatively perhaps an impersonal multatum could stand,
for which we could understand or supply a dative. The verb multare is regularly
constructed with ablative of forfeit (TLL , OLD s.vv.), but C’s sestertiorum decem
milia could have been mistakenly expanded from numeral and abbreviation. –
J. Crook in Crawford (:qq6) ¸¸o notes that a sum stated in sesterces would
date this enactment after c. :¡: ncr; cf. Crawford (:q8¸) :¡¸–8. More likely, he
thinks, the figure has been updated by F. or his source. For reprehensible acts
of this sort in general cf. Dig. xr\ii.::.:.: (Paulus) veluti si quis fimo corrupto aliquem
perfuderit, caeno luto oblinierit, aquas spurcaverit, fistulas lacus quidve aliud ad iniuriam
publicam contaminaverit, in quos graviter animadverti solet. For an instance involving
water, Bruns ::o (CIL :...¡.6): lex rivi Ul[. . .] si quis in eo mixerit spurcit(iam)
fecerit, in temp(lum) Iovis d(omestici?) X d(ato). Del(atoris) pars dim(idia) esto nesi
l. p. v.
q¡.¡ {oletato . . . facito} Cf. Paul. Fest. p...: L oletum stercus humanum.
The archaic verb oletare ‘befoul’ is a ctc¸ and may have required a gloss in
F.’s day (note a similar gloss in q¡.¸ id est quae . . . eam nos . . . vocamus), but the
form with imperatives is awkward and deletion probably right.
q¡.8 aediles . . . iubebantur The aediles’ involvement is part of their
general cura urbis. F.’s use of iubere here is consistent with usage of that verb
which conveys a legal authority (OLD s.v. ¸, 6). Note, in particular, the formula
Velitis iubeatis ut . . . at the head of a rogatio legis (Gell. \.:q.q): see Crawford
(:qq6) :, :o with further examples. Cf. below :o:.¸n. senatus iusserat.
per vicos . . . binos praeficere Tothe arbitriumof these neighbourhood
appointees would come matters involving any irregularity of the deliveries to
public basins (including, perhaps, water quality). Mommsen (:88¸) ii: ¡¸¸ n..
sees them as local deputies who could allow deliveries in domos principum civitatis
(q¡.¸n., where the reliqui would be other members of the vicus). It is entirely
unclear whether F.’s text here relates inany way tothe fragmentary LexSulpicia
of uncertain date, RS ¡.; see Crook (:q86), Mateo (:qq6). There is a similarity,
COMMENTARY q8.:–q8..
at least superficial, to the custodes mentioned by Tacitus, Ann. x\.¡¸.¡ in the
sequel to Nero’s fire.
in publico saliret Salire applied to water is a technical usage, referring
to continuous and unimpeded flow (cf. ¸6.., :o¡..). Pipes which fed the lacus
are called salientes (q.qn., ::.., 8¸.¸, :o¸–¡).
q8.r primus M. Agrippa . . . velut perpetuus curator Agrippa’s
personal commitment (cf. q.qn. singulari cura) dated at least from his aedileship
in¸¸ ncr (q.:n.), a markedchange fromwhat might have beenlargely desultory
management of Rome’s water-system. Along with his rebuilding and newcon-
struction, Agrippa assumed an ongoing responsibility for both maintenance
and distribution. He created, as it were, the cura aquarum – a standing office,
in contrast to annual magistracies or ad hoc commissions. The nature of the
aedileship itself might in part have helped to define the new initiative (cf. q¸..,
q¸.8), and personal munificence was entirely consistent with Roman traditions
(¸.¸n., 6.:n.). But the novelty is better viewed as an instance of Agrippa’s role
in the constitutional drama of these decades: Roddaz (:q8¡) :¡¸–¸¸, Tortorici
operum suorum et munerum F. has in mind the accomplishments
outlined in chapters 8–:o, but his words could no doubt be applied more
generally. Funds for the projects had come primarily fromAgrippa’s own purse
(Dio, xrix.¡...,.¡¸.:, ri\.::.¸), a substantial fortune derived by inheritance
from Atticus and augmented by his share of war booty: Roddaz (:q8¡) .¸¸–
¡¡. Bearing the costs of maintenance (§¸) was a further, if less conspicuous,
contribution to the public good. The word munera here can bear its familiar
meaning (see ¸..n.), pace Grimal. To call Agrippa curator implies that both opera
and munera are some sort of permanent public structures.
q8.z iam copia permittente Introduction of the Julia and the Virgo
had roughly doubled Rome’s water supply, but no less important was the
conscious planning that made possible wider and more equitable division
within the City: Evans (:q8.).
discripsit quid aquarum . . . daretur The signal achievement was
system-wide coherence (contrast the emphasis on aquae publicae in :oo.:, :o¡.:,
:.q.¡ with that on singulae aquae in Republican times (q6). An orderly appor-
tionment of all the available water according to recognised needs and priorities
was perhaps Agrippa’s most important innovation. These three categories rep-
resent long established customs. (:) Publica opera (¸..n.) might have embraced
baths and fulleries (q¡.¡, :o¸..). (.) Lacus (¸..n.) continued to supply the per-
sonal and domestic needs of the bulk of the population (q¸.¸–8). (¸) From any
available excess grants could be made to privati – although not necessarily re-
stricted in domos (q¡.¸); cf. Jansen (.ooo) :.., who notes that privati used water
for industrial properties, both outside the City and within.
COMMENTARY q8.¸–qq..
There is a marked resemblance to the threefold division sketched by Vit-
ruvius \iii.6.:–.: balneae, lacus et salientes, privatae domus: on the Vitruvian
‘canon’ see Hodge (:qq.) .8o, Evans (:qq¡) ¸. Geißler (:qq8) :¸¸ comments,
‘Diese enge Zusammenarbeit von Agrippa und Vitruv legt nahe, daß das von
Vitruv dargestellte Verteilungsmodell dem damaligen System Agrippas
entsprach und nicht nur ein theoretisches Modell war.’
F. implies that the categories were definedby Agrippa; it follows that Agrippa
might have regularised a procedure for implementing grants. On what au-
thority he acted one can only speculate (qq..n.). It was a striking feature of
Agrippa’s cura that grants to privati not only were multiplied, but they were
now recognised as an integral part of the distribution; cf., despite the probable
exaggeration, Strabo, \.¸.8 (cited above :¸.:n.).
q8.j familiampropriamaquarum Routine upkeep of the entire sys-
tem was handled by a standing crew of Agrippa’s own slaves detailed to these
tasks. Contrast the public contracts for tutela of individual aqueducts (q6). For
private familiae working alongside public crews in cases of fire in pre-Augustan
times cf. Dig. i.:¸ (Paulus) apud vetustiores incendiis arcendis triumviri praeerant,
qui ab eo quod excubias agebant nocturni dicti sunt: interveniebant nonnumquam et aediles et
tribuni plebis. erat autem familia publica circa portam et muros disposita, unde si opus esset
evocabatur: fuerant et privatae familiae, quae incendia vel mercede vel gratia extinguerent.
deinde divus Augustus maluit per se huic rei consuli.
qq.r hanc Augustus . . . publicavit In F.’s day the familia publica still
existed (::6.¸), its expenses paid from the aerarium (::8.:). In effect, Augustus
inheritedthe financial responsibility borne by Agrippa. Making the crewpublic
was a formality in the process of creating the new cura aquarum. The princeps
is unlikely to have thrust an obligation upon the state without making some
provision to ensure the means for supporting this familia, and special vectigalia
to offset the expenditures presumably date from this period (::8.:n.).
qq.z post eum . . . consulibus :: ncr, Degrassi ¸. Agrippa died in :.
ncr (Dio, ri\..8.¸).
in re, quae . . . certo iure eguit C’s eguisse perhaps arose because of
the following senatus consulta (likely originally abbreviated); there is no need for
a (concessive) subjunctive. Despite its lack of coherence (q8..n.), the Repub-
lican administration had not wanted a legal basis. The res here is Agrippa’s
cura (q8.:n.), now become Augustus’ responsibility: usque in id tempus embraces
only the period subsequent to ¸¸ ncr. By accepting his gifts (including mainte-
nance), the state hadtacitly acceptedAgrippa’s right to regulate andadminister
the entire system. His authority was unofficial because it was neither that of a
regular magistracy nor conveyed by special legal authority: as velut curator (q8.:)
Agrippa had only quasi potestas; cf. Eck (:qq.). The translations of Bennett and
Grimal mistakenly represent the res as including the Republican administra-
tion, and they ignore quasi in the interpretation of potestas. It is interesting to
note that Agrippa shared censoria potestas with Augustus in .8, and that Augustus
again had censorial powers in :: ncr: Jones (:q6ob).
senatus consulta The text of one S.C. (:o:.¡) implies that a senatorial
commission had been charged with inspecting the water system prior to the
senatus consulta passed in this year. The aim, of course, was to perpetuate the
cura of Agrippa, but there were constitutional formalities to be observed. One
supposes that Augustus had requested this commission and that his suggestions
underlay its recommendations, embodied in the series of senatorial resolutions.
F. quotes from six consulta, but there were others (see :oo.:n., :o¡..n., :.¸.:n.).
The Augustan legislation was something like a charter for the curatorial office,
and documents from that era would probably have been more or less readily
available to any curator who held the post. It is not dissimilar that the com-
mentarii of the Secular Games begin by quoting senatus consulta relating to the
organisation of that event in :¸ ncr (Pighi (:q6¸) :o¸–¸o): cf. Moretti (:q8¸),
Scheid (:qq¡) :8o.
lex promulgata Context initially suggests :: ncr; but since F. elsewhere
uses the singular (also at ¸.., :o¸.:) and quotes only the Lex Quinctia of q ncr
(chapter :.q), it is probably this law of which he speaks here. Promulgatio was
public announcement prior to and distinct from rogatio: Berger 6¸¸, Crawford
(:qq6) q–:o. Atwo-year interval between the senatorial resolutions and Quinc-
tius’ rogatio need not have any special significance. F.’s point is that the legal
basis of the curator’s position had been established by the three judicial for-
malities of senatus consulta, lex and principis edictum (cf. ¸.. lege, senatus consulto et
mandatis principum). Precise distinction in force between a senatus consultum and
a lex is difficult to determine in this era, and the lex might have been primarily
‘a means of publicizing decisions’: Martin (:q8¸) ..¸ (but see :.8.¸n.).
qq.j edicto complexus est A confirmation, no doubt, of grants made
by Agrippa, but on new terms. Presumably the edict implemented policies
formulated in senatus consulta (e.g. :o6, :o8), in which some regulations might
have been either entirely new or more consistently applied.
ex commentariis Agrippae This is the sole reference to Agrippa’s
‘notes’ on the Roman water-system, which formed the nucleus of the commen-
tarii principum (¸:..n.). Being primarily administrative records, they should be
distinguished from his memoirs (Introd. :: n.¡¸).
tota re in sua beneficia translata Tota res at first sight suggests an
overall responsibility akin to Agrippa’s cura, but context limits its meaning
here to the matter of distribution (especially grants to privati). Of course a
primary responsibility of the princeps was to ensure the supply for public
needs, but once this had been assured (cf. :o¡) distribution of any remaining
water was left to Augustus’ discretion. The category nomine Caesaris (¸..)
likely dates from this period. Most important, all grants to privati were here-
after issued as imperial beneficia. Behind Agrippa’s apportionment had been
a kind of personal authority (§.n.); the senate now recognised this author-
ity as a beneficium belonging to Augustus. Overall responsibility and control
remained entirely in the hands of one individual. Certain administrative el-
ements were entrusted to curators, but there is no evidence that the latter
themselves ever made grants on the emperor’s behalf (:o6.:n., :.q.::n.). Im-
perial control over grants to privati remained the rule: cf. Dig. xriii..o.:.¡.
(Ulpianus) idque [sc. ius ducendae aquae] a principe conceditur; alii nulli competit ius
aquae dandae.
qq.q modulos . . . constituit See above .¸.:–¸:.¸ and¸¸–6¸. Pipe-sizes
had probably never before been officially regulated, and the fistula quinaria was
now declared the standard (cf. :o6..n.). Bruun (:qq:) ¸¸–8 feels that there was
no need for standardisation of all delivery pipes, and that the emperor could
have defined his beneficia in other ways. Yet the verb constituere (Berger ¡oq)
implies an authority by which these pipes can now represent a legitima mensura
rei continendae exercendaeque Perhaps to be taken generally, ‘for
the maintenance and operation of the whole system’ (Herschel/Bennett). But
context invites more specific legal senses, ‘to keep matters within defined
boundaries and to administer the laws if those boundaries are breached’.
It is, after all, the juridical powers of curators with which F. deals from
chapter q¡ onward; cf. q¡.: ad cohibendos . . . privatos, ad ipsorum ductuum . . .
tutelam and :¸o.¸–¡ admoniti, executio legis. In this respect they held, in the
sphere of public water, most of the competency which had fallen to Republican
curatorem fecit Appointment was left to Augustus (:oo.:n.). Contem-
porary documents use the plural curatores (also at :o6.:, :.¸.¸, :.q.q–::, while
the singular at :.q.¸ has a special point), but the consular chairman readily
eclipsed his fellows: see :, :o..:, :o¸.:, ::q.¸. Antecedents for a three-man
board were very ancient, e.g. the tresviri agris dandis assignandis first attested in
¡6¸ ncr (Livy, iii.:.6; in .:8 the board consisted of one consular and two non-
consulars: Livy, xxi..¸.¸–¸). Cf. Eck (:qq.) on collegia of other curatorial boards
created under Augustus.
MessalamCorvinum REValerius no. .6:, PIR\ qo. Messala’s nobility
might have lent high prestige to the post, but the choice no doubt also reflected
the close relationship between Messala and Agrippa. The two had shared the
house of Mark Antony (Dio, riii..¸.¸), and Grimal (:q¡¸) :¸.–¡ opines that
Messala might have possessed the properties of Lucullus (¸.¡n., 8.:n.) which
figured so prominently in Agrippa’s projects.
adiutores F. nowhere else mentions these junior curatores, who are to be
sharply distinguished from more general ‘underlings’ (see ..:n.). Beyond this
initial pair we know that Didius Gallus (:o..¸) is named with two colleagues
(Appendix B, no.q). Aboard of three senators appears on a lead pipe of Domi-
tianic date (CIL :¸.¸.8:a = ILS 868.), although it cannot be established that
their cura was that of the water administration (see :o..:¸n.). There is no ev-
idence, one way or another, to suppose that senatorial adiutores ceased to be
chosen subsequent to the appearance of an imperial procurator (:o¸..n.). Of
the manner of their appointment, the nature of their duties, and the length of
their service we are completely ignorant.
Postumius Sulpicius RE no.8o, PIR r 66¸; cf. Syme (:q86) ..q–¸o.
Perhaps a nephew of Messala (whose sister Valeria had married a Sulpicius)
and brother of the poetess Sulpicia. Butrica (:qq¸) ¸:–¸ proposes him to be
the poet Lygdamus, author of iii.:–6 in the Corpus Tibullianum. Born in ¡¸ ncr
(C.T. iii.¸.:8; cf. Ovid, Tr. i\.:o.6), he would have been old enough to have
held the praetorship by :: ncr. Beyond this we know nothing.
Lucius Cominius RE no.q, PIR
c :.6..
pedarius I.e., who had not held the praetorship. Tacitus, Ann. iii.6¸..
distinguishes three categories in the Senate: consulares, praetura functi and pedarii
senatores. On pedarii see Taylor–Scott (:q6q), Korpanty (:q¸q), Talbert (:q8¡)
¸:6, Kunkel–Wittman (:qq¸) ¸:6.
qq.j insignia . . . quasi magistratibus Although their powers in-
volved some which had fallen to Republican magistrates, especially to censors
(e.g. :o¡.., :o6.:, :.¸.¸, :.q.¸), these curators were senatorial agents of an
‘imperial executive’ and not technically magistrates in their own right: see,
in general, Eck (:q¸q); Bruun (:qq:) :¸q–8¸. A magistrate’s insignia consisted
of the toga praetexta, the sella curulis and – for certain magistracies in certain
circumstances – the fasces (with lictors): see Mommsen (:88¸) i: ¸¸.–¡¸¸; for
Augustan curators, ibid. ii::o¡q–¸o. Cf. Palma (:q8o) .o:, Wanscher (:q8o) esp.
:..–¡, Sch¨ afer (:q8q) esp. .:¡. By association F. may be including the privilege
of apparitores specified in the S.C. below (:oo.:), for censors had scribae (Pliny,
HN xxxiii.:¸, Dig. i......¸) and praecones (Varro, Ling. \i.86, Livy, xxix.¸¸.8),
while curatores viarum had lictors outside of Rome (Dio, ri\.8.¡): Cohen (:q8¡)
¸¸ n.66. For management of the water system, curatores aquarum apparently
had all censorial powers except, of course, for (:) building and major repairs
and (.) the ius dandae vendendae aquae: Mommsen (:88¸) ii: :o¡¡–¡, :o¸¸–¡;
Ashby (:q¸¸) :.; Hainzmann (:q¸¸) 6¸; D’Amato (:q86) :8¸; Geißler (:qq8)
qq.j infra scriptum est Cf. :.8.¡ subscripsi verba legis. Elsewhere when
he quotes from documents (except for :o¸.¸ haec verba) F. uses the word subicere
roo.r Quod . . . I. C. The formulaic heading of a senatus consultum. The
consuls relate the subject (V.F., verba fecerunt), which is expressed either by a
prepositional phrase (de . . .) or in indirect statement. The question is then
raised and the resolution itself introduced, according to Probus, Q(uid) D(e)
E(a) R(e) F(ieri) P(laceret) D(e) E(a) R(e) V(niversi) I(ta) C(ensuere). F.’s text nowhere
preserves the letter for universi; note here the order D. E. R. Q. and the omission
of D. E. R. in :o¡ and :.¸, all of which may be due to a scribe rather than to F.
On the formal patterns of senatus consulta in general, see Mommsen (:88¸) iii:
:oo8, Daube (:q¸6) ¸8–86, Volterra (:q6q) :o¸¡–¸, Sherk (:q6q) ¸–:¸, (:q¸o)
¸q–6¸, Talbert (:q8¡) ¸o¸–¸. F. may have truncated this and others of the S.C.
which he cites, and it may be noted that he nowhere includes the formulaic
heading in which were specified the date and place and the names of senators
who finalised the wording (qui scribundo adfuerunt).
de iis . . . ordinandis Here (as in :o¡.:, :o6.:, :o8) the relatio statement
alone expresses the motivation for the S.C.; for a slightly different form, see
:.¸.:n. cum . . . corrumpantur.
ex consensu senatus . . . nominati Cf. :o¡.. quos Caesar Augustus
ex senatus auctoritate nominavit. The phrases ex senatus auctoritate and ex consensu
senatus imply a previous resolution (which F. does not cite) that must in effect
have created the cura aquarum, or at least defined the bounds within which
Augustus couldmake appointments (qq..n.): Mommsen(:88¸) ii: :o¸¡ n.:. Part
of the prestige which attached to this office derives from the fact that curatores
aquarum were specifically appointed (nominati) by the princeps, in contrast to
being chosen by lot as had been the case with praetorian commissioners who
distributed grain (Dio, ri\.:.¡ cíptïoûci, :¸.: ìcycv:t,) and probably with
curatores viarum(:o:.:n.). (Sortitionwas later usedinselectingcuratores alvei Tiberis:
Dio, r\ii.:¡.8). The present text gives no hint that the Senate is confirming the
appointments: Mommsen (:88¸) ii: :o¡8 n...
ornandis The transmitted ordinandis anticipates ordini. Ornare fits better
with insignia and personnel; cf. Cic. Agr. ..¸. [XVviros] ornat apparitoribus, scribis
librariis, praeconibus, architectis.
extra urbem lictores A symbol of judicial authority, but allowed only
outside the pomerium. The concession of fasces and lictors to curatores aquarum
differs fromRepublican practice, wherein censors lacked this right: Mommsen
(:88¸) i: ¸86. The restriction extra urbem reflects the traditional separation of
imperium between domi and militiae: cf. ibid. i: ¸¸q–8o.
servos publicos Public slaves served as aides to magistrates and as as-
sistants in public cults: Mommsen (:88¸) i: ¸.¸–.6, Halkin (:8q¸) ¸q–8¸, Eder
(:q8o) ¸q–¸6. Cohen (:q8¡) ¸o–¸ points out that these persons because of
servile status could not be employed for ‘contacts and negotiations with the
society of free citizens’. They are not to be confused with members of the
public work-gang (qq.:), and there are no grounds for Grimal’s view that
the three servi are identical to the architectus, the scriba and the librarius. The na-
ture of their duties for the curators is unclear. They might have been couriers
(cf. Plut. Galba 8.¸); or, if they served only outside the City (§.n.), they might
have been concerned with transportation.
architectos Professional experts or engineers, used as technical advisers
(cf. ::q.¸n.). Note that Pliny, Ep. x.¸q.¡, ¡:.¸ askedfor anarchitectus inconnexion
with public building projects. – Cohen (:q8¡) ¸¸ n.¸o cites Cic. ii Verr. iii.:¸¡
for the usual order (and grades) of these apparitores: scribae, accensi, lictores, viatores,
praecones. On the ‘office staff’ in general, see Bruun (:qq:) :q¸–¡.
scribas et librarios Clerical assistants for routine administration and
record-keeping. The curator’s staff may at the outset have had some respon-
sibility for maintaining records of beneficia. Asyndeton would be appropriate
here (K–S ii..: :¡q), but not necessary (cf. TLL ¸..: :¸¡¸..¸). The copula et
(and that in praeconesque just below) might result from literary tradition. Any
possible distinction is vague (cf. Festus, p.¡¡6 L scribas proprio nomine antiqui et
librarios et poetas vocabant; at nunc dicuntur scribae equidem librari, qui rationes publicas
scribunt in tabulis), with librarii sometimes lower in rank. Frequent attestations of
scribae librarii might suggest a separate (intermediate?) rank of clerical servants.
accensos praeconesque Personal aides and public heralds are a regu-
lar part of amagistrate’s apparatus. AccordingtoCohen(:q8¡) ¸8–q the accensus
differed from other apparitores insofar as this was a distinctly lower grade, the
relationship with the office-holder was more personal, and apparitorial status
did not survive the term of service.
totidem habere . . . per quos frumentum plebei datur A clear
precedent for the curatores aquarum was ready to hand in the agency comprising
senators of praetorian rank which in .. ncr had been put in charge of the
distribution of free grain (frumentationes), its organisation somewhat modified in
:8 (Dio, ri\.:.¡, :¸.:). Their standard title is praefecti frumenti dandi ex s.c. (e.g.
CIL q.¸¸o6 = ILS q¸.), but for the periphrasis cf. qui aquis publicis praeessent
just above and see :o:.:n. The precedential relevance of this board lies in
the fact that its creation marked the first of several stages in the evolution
of a new system of civil service under imperial direction. See van Berchem
(:q¸q), Pavis d’Escurac (:q¸6), Pflaum (:q¸8), Rickman (:q8o), Eck (:q86,
roo.z in urbe For the distinction extra urbem / in urbe see ¸..n. extra urbem.
ceteris apparitoribus Because their purpose was so closely related to
public duties, apparitores were required to be citizens (either freeborn or liberti).
It is unlikely that the word is used loosely here (cf. Diz. Epig. :: ¸.¸), and so servi
publici would by definition be excluded. On the ordo called apparitores and the
offices available to its members, see Purcell (:q8¸), Cohen (:q8¡); cf. Badian
roo.j eos . . . ad aerarium deferrent The normal procedure for
enrolling names on the public payroll. Again, only apparitores are involved,
for servi publici would not be salaried. Like magistrates, the curators pre-
sumably chose their own personnel, and terms of service were probably
praetores Subject of darent et adtribuerent. C’s praetoris is more likely a
matter of scribal carelessness (after iis) than a transmitted nominative plu-
ral in a consonantal-stem noun (on which see K–S i: ¸¸.–¸). At this period
two praetors were in charge of the aerarium Saturni (Dio, riii...:, ¸...; Suet.
Aug. ¸6).
mercedem cibaria Remuneration for apparitores is called merces in the
Lex Coloniae Genetivae (RS .¸, Bruns .8, FIRA .:), ch. 6.–¸. Cibaria, payment
for official services (lit. ‘food allowance’), is virtually synonymous with merces
(RE ¸: .¸¸¸; cf. Pavis D’Escurac (:q¸6) :68 n.:8) and I take it to apply only
to apparitores. Public slaves by definition were provided for at state expense.
Funds for their commoda (::8.:n.) were also paid from the aerarium, but special
authorisation was hardly needed. I cannot accept cibaria annua in this passage
as ‘explicit testimony’ that they were salaried: so Eder (:q8o) :o8–q.
praefecti frumento dando For these commissioners, whose title may
not yet have fossilised, see :o:n. Perhaps praefectus here (+ dative) should be
taken generally as ‘those in charge of’ (rather than ‘the prefects of’).
dare deferreque solent Salaries for attendants were to be the same
as those paid to personnel of the grain distributors. (Krohn was right to un-
derstand suis apparitoribus, but the words need not be inserted.) The sense of
dare deferreque is not entirely clear, but it can mean neither ‘deposit’ (cf. TLL
¸: ¸:¸.¸¸) nor ‘deliver’ (i.e., as the fixed ration of the frumentatio calculated by
a ‘moveable scale’ adjusted to the market value of grain: Grimal q. n.::¸).
I take deferre to be closely related to the preceding deferrent and delati: by en-
rolling their personnel the prefects authorise their salaries. This may be what
is meant in the Lex Cornelia de xx quaestoribus (RS :¡, Bruns :., FIRA :o): (the
magistrate?) [ad] q(uaestorem) urb(anum) . . . eam mercedem deferto, quaestorque . . . eam
pequniam . . . solvito; cf. Crawford (:qq6) .qq, who takes deferre mercedem as ‘to
register’. Precise amounts of the salaries are not stated here (as they are, e.g.,
in Lex Coloniae Genetivae), but it may be supposed that these had been offi-
cially set in legislation creating the board of praefecti frum. dandi. There remains,
admittedly, an uncertainty in the case of personnel not explicitly paralleled in
the earlier agency (lictores and architecti).
isque . . . capere liceret Sine fraude sua facere is ubiquitous in legal texts
(cf. :.q.q), but there canbe nodoubt that capere is requiredhere: cf. LexCornelia
de xx quaestoribus (RS :¡, Bruns :., FIRA :o), lines ¸–6 olleisque hominibus eam
pequniam capere liceto; Lex Coloniae Genetivae (RS .¸, Bruns .8, FIRA .:) 6¸.8
(apparitores) mercedem pro eo kaperent, itque iis s(ine) f(raude) s(ua) c(apere) l(iceto).
COMMENTARY :oo.¡–:o:.:
The error very likely arose from misunderstanding an abbreviated K(apere).
To accept facere pecuniam requires alternative absurdities: the antecedent of is
would be either praetores aerarii (who may ‘disburse that money’) or praefecti frum.
dandi (‘establish that sum’).
roo.q quae . . . opus essent Cf. Venafrum edict, line .6 quaeque ea[ru]m
rerum cuius faciendae reficiendae causa opus erunt; contrast :.¸ below ceteraque quibus
ad eam rem opus esset.
roo.q cos. ambo [alte]rve, si is videbitur The formula usually ap-
pears in the order alter ambove, frequent in Cicero’s Philippics, e.g. \.¸¸ C. Pansa
A. Hirtius consules, alter ambove, si eis videretur (\ii.::, \iii.¸¸, xi\.¸¸, etc.), cf. Livy,
xxx..¸.. non videri sibi absente consulum altero ambobusve eam rem agi satis ex dignitate
populi Romani esse. The order here may be owing to an uncertain expansion;
for an abbreviated form see, e.g. S.C. de Asclepiade Clazomenio sociisque
of ¸8 ncr uteique . . . co(n)s(ules) a(lter) a(mbove) s(ei) e(is) v(ideretur) . . . curar-
ent (ctc, :t . . . 0tc:ci, c t:tpc, n óugc:tpci, tcv co:cï, gcivn:ci . . .
ror.r itemque Poleni was first to notice that this sentence forms part of
the S.C. The entire section is woefully obscure, and the text is almost certainly
more corrupt than editors have admitted. Little is sure beyond the fact that
the new curatores aquarum are somehow likened to existing agencies.
viarum curatores Although the post of curator viarum is attested in Re-
publican times (e.g. CIL 6.:¸o¸ = ILS ¸8q.), the reference here is to an
office created under Augustus (cf. Suet. Aug. ¸¸), probably in .o ncr: Dio,
ri\.8.¡ :c:t ot co:c, :t tpco:ó:n, :cv ttpi :nv Pcunv cocv cíptûti, . . .
sci coctcicu, ts :cv to:pc:n,nsc:cv, pcµocúyci, oúc ypcutvcu,, tpcot-
:c¸t. Persons called viarum curator (sometimes ex S.C.) are also known from in-
scriptions. But Augustus’ interest in roads was not confined to Rome and its
environs (cf. Dio, riii....:; Suet. Aug. ¸o.:; Mon. Anc. .o), and the province of
these officials is not at all clear. See, in general, Culham Ertman (:q¸6) .:–¡;
Eck (:q¸q) ¸¸–¡¡, (:qq.).
† . . . frumentique† The reference can hardly be to anything other than
the senatorial boardresponsible for distributionof grainrations (:oo.:n.), called
praefecti frumento dando only a few lines above. There are Augustan inscriptions
which mention a frumenti curator ex s.c. (CIL 6.:¡6o = ILS 88¸) or cur. fru. (CIL
6.:¡8o = ILS qo¸), plausibly taken to be examples of an initial hesitation over
the formal title (cf. :oo.¸n.).
parte quarta anni The grain prefects were four in number, chosen
annually, and serving in succession (ts oicocyn, Dio, ri\.:¸.:); each would
thus have had a quarter-year term. The number of curatores viarum in :: ncr is
not known. They need not have been merely two.
fungerentur . . . vac<ar>ent Syntax requires at least some adjustment
in the tenses, and the style of this document as a whole (as of S.C. in general)
supports the imperfect subjunctives. Loss of an abbreviation stroke could have
produced vacent from vacarent. C’s indicative fungebantur is more difficult, but
hardly adequate to justify Grimal’s deletion of qui parte . . . ministerio as a gloss
which has replaced an entire clause.
iudiciis vac<ar>ent Vacare + ablative (be free from; cf. vacatio militiae)
is altogether more plausible than vacare + dative (be available for). Exemption
fromjudicial duty is alsomore appropriate inthe context of the curators’ ornatio;
cf. the S.C. de Cyrenaeis of ¡ ncr, which similarly exempts certain categories
of person (including those n tt ópyn, n tt t¸cuoic, :t:c,utvc,, n ttio:ó:n,
spi:npicu n ttiutìn:n, oti:cut:pic,). The most convincing interpretation is
that proposed by Hirschfeld (:qo¸) .o6: the other officials are excused from
iudicia for that portion of the year (qua parte anni) when they superintend their
special agencies; so the curatores aquarumwhile on ‘active duty’ are to be likewise
excused. Amatucci (:q.¡) :8¸–8 proposes cum viarum <curatores> curatoresque
frumenti quota parte anni fung{eb}antur ministerio iudiciis <vacent privatis publicisque>,
ut curatores aquarum iudiciis vac<ar>ent. . . . With no hint to suggest otherwise, I
should have thought the exemption to be a standing one. (For the use of plural
curatores here see qq.¸n. eis.) Others have taken the meaning to be that the
curatores aquarum were to hear cases within their jurisdiction for certain periods
of the year (usually three months, based on the transmitted quarta parte anni).
To assume from the text as it stands that the curators’ service was limited to
certain parts of the year is entirely unwarranted.
ror.z apparitores et ministeria On the metonymy ministeria / ministri,
see TLL 8: :o:.. Context suggests that both words refer to the administrative
personnel (:oo.:). If it is meaningful at all, F.’s distinction is then probably
between apparitores, applied here inexactly to the symbols of curatorial status
(lictores, accensi, praecones), and ministeria, helpers in the routines of office (architecti,
scribae, librarii, servi publici); cf. Bruun (:qq:) :q¸–¡. There would be little point to
a distinction between free persons and slaves (cf. :oo..n.). Ministeria, however,
is used in chapter ::¸.: of the familia of workers (cf. q6 and Pliny, Ep. x.¸:..,
also Str. i.:.¸ per speciem servitutis ac ministerii); and although not mentioned in
the S.C., they may have come to F.’s mind at this point because they too had
ceased to be closely accountable to the curator.
tamen esse curatorum videntur desisse Their positions, it seems,
had either become complete sinecures or their duties had in effect been inter-
mingled with the imperial administration. That they still existed seems clear:
else F. had been yet more indignant. The present tense reveals a situation which
obtained when F. came to office, perhaps a gradual development to which he
gives no clear starting-point. What we should make of plural curatorum is not
COMMENTARY :o:.¸–:o..:
at all obvious: it could suggest a lengthy series of inept predecessors (cf. :¸o.:
neglegentia longi temporis) or it could be rather a generalising defamation. See also
below :o..:6n.
inertia ac segnitia non agentium officium For the qualities, Syme
(:q¸8a) :: ‘that discretion which men called “quies” if they approved, otherwise
“inertia” or “segnitia”’. F.’s remark could easily be taken as a withering rebuke,
directed at men of highest rank and smug in the implicit contrast to his own
sollicitudo and diligentia. Inertia ac segnitia leaves no room for excuses on the part
of senatorial curators (e.g., deference to a greater efficiency on the part of
imperial administrators, or to greater familiarity with technical details on the
part of specialised workers). Yet F.’s attitude elsewhere (e.g. :oq.¸, :¸o.:) is
one of leniency – at least towards fellow senators. It might perhaps be better
to imagine that here he means their inertia ac segnitia were responses to the
political climate that characterised the later regime of Domitian (cf. ::8.¸);
note specially Tac. Agr. ¸.:, looking backwards, subit quippe ipsius inertiae dulcedo,
et invisa primo desidia postremo amatur. See also below :o¸.:n. libertum Caesaris.
ror.j egressis . . . urbem Transitive egredi is rare (OLD s.v. ¡, TLL ¸..:
.8¸.¸¸); withurbemlimitedto Val. Max. i.:.., ix.6(ext.)..; Dig.,;
xr...:¸.pr. (cf. Livy, xx\.8.6 se urbe egressos, Gell. \.¸.¡ egrederetur extra urbem). The
extra urbem/in urbe distinctionwithinthe S.C. itself (:oo.:–.) involves traditional
customs, so F. here may mean the urbs as bounded by the pomerium: De Kleijn
(.oo:) ¸¡.
senatus . . . iusserat Cf. :o¸.¸ senatus consulto facere curator iubetur. For
the use of iubere, perhaps more appropriate for leges than for senatus consulta, see
q¸.8n. iubebantur.
ror.q nobis . . . pro lictoribus erit A statement of personal assurance
(not unlike F.’s scorn for an impressive tombstone: Introd. ¸) and one which
emphasises F.’s view of the close relationship that should exist between the
curator and his princeps. Republican censors had no lictors, nor presumably
had M. Agrippa.
circumeuntibus rivos For the compound verb transitive cf. :o¸.¡ ductus
circumeundi, Str. i...: maximam partem munimentorum circumierunt, ii.¸.¸,:o etc.
roz.r subiungere Although F. uses subicere and subiungere more or less
synonymously (.¸.:, :o¸.:) ‘go on to say, list’, here we might drawa distinction,
for what follows is a list, not a verbatim document (¸¸.:n., :o¸.¸n.).
huic officio . . . praefuerint F’s antiquarian bent may account for
reciting the names of his predecessors, but the list as a whole underscores his
comment in the prologue that the cura aquarum had been the responsibility of
leading senators (per principes semper civitatis nostrae viros). The pattern of tenure
varied with changes in the principate. Appointments at first went to men of
COMMENTARY :o...–:o..¸
forensic eminence and were apparently for life. (A term of such length was
remarkable in itself, given the traditional pattern of Roman magistracies: it
must certainly have reflected the perennial service of M. Agrippa.) Later, be-
ginning perhaps as early as Claudius (see :o..¸n.), the tenure was shortened
and the office was gradually merged into the senatorial cursus: Ashby (:q¸¸)
.o. What may have begun as a prestigious (qq.¡n.), if not necessarily onerous,
post apparently loses some of its lustre by the time of Claudius (:o¸..) if not
earlier. Speaking of this and other imperial curae, Syme (:q86) ..: writes, ‘A
suspicion arises that some posts came close to the ideal of “something to live
for and nothing to do”. Inertia or even absence might not be detrimental;’ cf.
idem (:q8.) :q¡, Bruun (:qq:) :8¸–¸. Table :: sets forth the transmitted data,
with prosopographical references for individual curators. For the list in gen-
eral, see Cantarelli (:qo:), Ashby (:q¸¸) :¸–.o, Rodgers (:q8.b), Bruun (:qq:)
roz.z Planco et Silio consulibus :¸ cr (Degrassi ¸). (For C’s nomina-
tive forms of consuls’ names, here and elsewhere, see ¸.:n.) Alongside jum-
bled data in Jerome’s Chronicle, this sentence has been taken as evidence that
Messala Corvinus died in :¸ cr: RE Valerius no..6: (Hanslik), defended by
Jeffreys (:q8¸). References in Ovid (Tr. i\.¡..¸, Pont. i.¸..q) can be interpreted
to indicate an earlier date, prior to Ovid’s exile, probably in 8 cr: Syme (:q¸8)
:.¸–¸, (:q86) .oo–.6. In the latter case it would be necessary to assume a
vacancy in the office of curator from 8 to :¸, or to postulate a lacuna in F.’s
text which would have contained a date and the name of Messala’s successor.
Ateius Capito RE no.8, PIR
\ :.¸q. Suffect consul in ¸ cr (Degrassi 6).
An eminent jurist (cited by F. at q¸..). In :¸ cr he was appointed to what
seems to have been a commission which led to forming the board of curatores
alvei Tiberis: Tac. Ann. i.¸6.. remedium coercendi fluminis Ateio Capitoni et L. Arruntio
mandatum; cf. Dio, r\ii.:¡.8; Mommsen (:88¸) ii: :o¡6. His death fell in ..
(Ann. iii.¸¸.:). Capito might have been chosen curator aquarum because of a
relationship to Agrippa. Ateii Capitones were related to Quintilius Varus,
Agrippa’s son-in-law: Reinhold (:q¸.). See q¸..n., Horsfall (:q¸¡), Bauman
(:q8q) ¸q–6., Bruun (:qq:) :¸¸.
roz.j [C. Asinio Pollione] C. Antistio Vetere consulibus .¸ cr
(Degrassi q). C’s short erasure indicates some disturbance in the exemplar.
The identical praenomen is perhaps enough to account for the omission.
Tarius Rufus Possibly – but by no means certainly – L. Tarius Rufus
(RE no.¸, PIR + :¡), infima natalium humilitate consulatum militari industria meri-
tus (Pliny, HN x\iii.¸¸). Relic of Actium (Dio, r.:¡.:–.) and suffect consul in
:6 ncr (AE :q¸6, :8; Degrassi ¡), his career is otherwise obscure, although
Augustus’ generosity brought him financial success. This Tarius would have
been about eighty years old at the time of his appointment as curator, his death
COMMENTARY :o..¡–:o..¸
soon afterwards a plausible explanation for the quick succession of Nerva in
the following year. One can do no more than speculate on the appointment of
so senior a consular and one without the forensic preeminence of his predeces-
sors: Tiberius might have chosen him because of some (unknown) relationship
with Agrippa. Possibly the curator was another Tarius, perhaps son of the
consul of :6 ncr (Dessau PIR + :¡), although the son, detected in attempted
parricide, was banished by family council at which Augustus was present (Sen.
Clem. i.:¸.¡). Syme (:q¸8) :.¸ suggests that the order in F.’s text may be in-
correct and that a sexagenarian Tarius could have served from 8 to :¸ (see
above). In later works, Syme tentatively nominates as curator .¸–¡ C. Vibius
Rufus (cos. suff. :6 cr): Syme (:q8: ), (:q86) ..¸–¸. He cautiously proposes
two lacunae in F.’s text: Messallae successit <Furio et Nonio consulibus Tarius Rufus,
Tario> Planco et Silio consulibus Ateius Capito, Capitoni <C. Asinio Pollione> et C. An-
tistio Vetere consulibus <Vibius Rufus, Vibio> Servio Cornelio Cethego L. Visellio Varrone
consulibus M. Cocceius Nerva, a solution admittedly ‘complicated, hazardous, and
roz.q Ser<v>io Cornelio Cethego L. Visellio Varrone consulibus
.¡ cr (Degrassi q). Perhaps one should attribute the expansion Serio to mistaken
fussing on the part of Peter the Deacon (cf. 6.:n. Lucio). C has Ser. alone as
praenomen at ¸.: and §¸ below.
M. Cocceius Nerva RE no.:¡, PIR
c :..¸. Consul in .: or .. cr:
Syme (:q8: ) ¸¸:–6. An intimate of Tiberius: see Tac. Ann. i\.¸8.: (in the year
.6) profectio arto comitatu fuit: unus senator consulatu functus, Cocceius Nerva, cui legum
peritia; \i..6.: Cocceius Nerva continuus principi, omnis divini humanique iuris sciens . . .
moriendi consilium cepit. The suicide came in ¸¸ (cf. Dio, r\iii..:.¡). This name
on a lead pipe found on the Esquiline (CIL :¸.¸¡¸¸) and on another from the
Viminal (BCAR :q8¸–8, :.¡ no. ¡¸) may refer to the curator, or his son, or
perhaps even the emperor: Bruun (:qq:) 6¸.
divi Nervae avus, scientia. . . iuris inlustris Headof what was later
the Proculian school: Bauman (:q8q) 68–¸.. F.’s biographical note is unique
in this list, plainly because of his own relationship to the jurist’s grandson.
roz.j Fabio Persico L. Vitellio consulibus ¸¡ cr (Degrassi :o).
C. Octavius Laenas RE no.6¡, PIR
o ¡¡. Consul in ¸¸ cr (Degrassi
:o), but otherwise unknown, presumably a novus homo. He might have been
the Laenas whose daughter was Nerva’s mother (ILS .8: Sergiae Laenatis f.
Plautillae matri imp. Nervae Caesaris Aug.): Bruun (:qq:) :¸¸ n..¡. If so, the choice
might have been made because of the relationship to his predecessor. And if
the identification is accurate, each of Nerva’s grandfathers would have been
curator aquarum, an additional reason for that emperor’s taking a special interest
in this facet of urban administration.
COMMENTARY :o..6–:o..¸
roz.6 Aquila Iuliano et Nonio Asprenate consulibus ¸8 cr
(Degrassi ::).
M. Porcius Cato REno.¸¸, PIR
r 8¸6. Suffect consul in¸6(CIL:..¡¡o¸
= AE :q¸6, ¸88), and not necessarily fromthe ancient consular family. He was
among the praetura functi who came eventually to bad ends for having accused
Titius Sabinus in .8 (Tac. Ann. i\.68..).
roz.¡ huic successit . . . Ser. Asinio Celere [Sex.] Nonio Quintil-
iano consulibus Nowthat the Fast. Ost. have revealed these consuls’ names
and their date (suffecti from : July ¸8 cr: Degrassi ::), some of the difficulty
in this passage has been removed: see H¨ ulsen (:q.o) ¸o6; Bruun (:qq:) :¸¸. It
would seem that the death of Porcius Cato could be assigned with confidence
to the early months of ¸8. The successor took office in the same year: hence
the anomalous dating by suffecti. There remains the problem of post quem, a
phrase (as Poleni noted) which repeats the sense of huic successit but is out of
place in the syntactical formula of the chapter as a whole. The pattern is set by
Messalae successit in §., and F. normally proceeds with juxtaposition of nomina-
tive and dative (e.g. Capito Capitoni); huic successit intervenes at §¸ because of the
parenthesis and here, I think, because of the irregularity of two appointments
in the same year. F. uses post quem in §:¸ to mark the end of the series. To
date without comment by suffecti is perhaps too abrupt, and there is thus some
merit in attempts to wrest a temporal indication from the transmitted reading.
I suggest that post quem is not a corruption but a reader’s gloss which has ousted
a short phrase such as eodem anno.
A. Didius Gallus RE no.6, PIR
n ¸o. Suffect consul ¸q cr, proconsul
of Asia ¡q/¸o. Attested as curator aquarum on a cippus for Marcia-Tepula-Julia
(Appendix B , no. q) – but not, alas, in the extant portion of an inscription from
Olympia (CIL ¸.¸.¡¸ = ILS q¸o). For a useful summary of Didius’ career,
see Petersen–Vidman (:q¸¸) 6¸¸–6q; Birley (:q8: ) ¡¡–q; Vogel-Weidemann
(:q8.) ¸¡8–6.; and for concerns about his termas curator Syme (:q86) ..:–.,
Bruun(:qq:) :¸8–6o. Appointment in¸8(as transmitted) presents aremarkable
anomaly in light of the discovery that Didius’ consulship did not come until
¸q (AE :q¸¸, :¸8); the post of curator aquarum is otherwise without exception
reservedto consulars. Asevere shortage of available consulars inthe emergency
circumstances of Cato’s downfall could be adduced to account for the choice
of Didius (perhaps already designate for ¸q). But in the mid-¡os Didius was
dispatched on military service to the Bosporan kingdom (Tac. Ann. xii.:¸; cf.
BMCBosporos, ¸.–¸) and was away fromRome for perhaps two or three years.
To explain this absence one must imagine Didius ‘on leave’ fromhis curatorial
duties, a practice not otherwise attested. Both irregularities could be removed
by assuming a lacuna in F.’s text (a copyist’s eye jumped from consulibus to
consulibus). Didius could plausibly have been appointed curator aquarum upon
COMMENTARY :o..8–:o..:.
his return from the Bosporus (perhaps in ¡6); his tenure ended in ¡q (when
he might have become proconsul of Asia). He is the first known curator who
did not retain the position for life. The name of any hypothetical predecessor
has been irretrievably lost, but an appealing possibility is Asinius Gallus (RE
no.:¸, PIR
c :..8: note the textual proximity of the name Ser. Asinius Gallus,
suff. ¸8, and the similarity of asinius to adidius). Yet this Asinius is not attested
in the consulship: Bruun (:qq:) :¸¸ n.q.
roz.8 Q. Veranio et Pompeio Longo ¡q cr (Degrassi :¡). Pompeius’
cognomen is given as Gallus in Solinus, i..q, and apparently as Longinus in
a contemporary inscription (AE :q.8, q8). Perhaps we should read Long<in>o
Cn. Domitius Afer REno.:¡, PIR
n:.6. Suffect consul ¸q cr (colleague
of Didius Gallus). The foremost orator of his day (Tac. Dial. :¸.¸, Quint.
x.:.::8), and with remarkable wealth frombrick production: Bruun (:qq:) :6..
It has always been assumed that he held the office until his death in ¸q (Tac.
Ann. xi\.:q). Another lacuna, however, is possible here: consider Afro Nerone
Claudio < II et L. Pisone consulibus *** Nerone Claudio> IIII et Cosso Cossi f. consulibus
L. Piso. The interval between Piso’s consulship in ¸¸ and his appointment as
curator exactly equals Piso’s own termas curator (6o–¸): roomfor an unknown
consular who may, in ¸¸, have been the first of Nero’s appointees.
roz.q Nerone Claudio Caesare .IIII. et Cosso Cossi f. consulibus
6o cr (Degrassi :6).
L. Piso RE Calpurnius no.¸q, PIR
c .q¡. Consul ordinarius in ¸¸ with
Nero II as colleague (Degrassi :6), proconsul of Africa in 6q/¸o. Under Nero
in 6. he was one of three consulars in charge of vectigalia publica (Tac. Ann.
x\.:8.¸). Perhaps he was the L. Piso whose name appears on a pipe from the
Lateran (CIL :¸.¸¸:¸): Bruun (:qq:) 6¸. For the prominence of the Calpurnii
Pisones, see RE ¸: :¸8¸ (Groag). Verginio Rufo et Memmio Regulo consulibus 6¸ cr
(Degrassi :¸).
Petronius Turpilianus RE no.¸¸, PIR r .¸¸. Consul ordinarius in 6:
cr (Degrassi :¸), favoured under Nero but executed in 68.
roz.rr Crasso Frugi et Laecanio Basso consulibus 6¡ cr (Degrassi
P. Marius RE no.¸¸, PIR
x .q¡. Consul ordinarius in 6. cr (Degrassi
:¸, but without the cognomen Celsus), otherwise unknown. See Eck (:q¸¸)
roz.rz Luccio Telesino et Suetonio Paulino consulibus 66 cr
(Degrassi :8).
COMMENTARY :o..:¸–:o..:6
Fonteius Agrippa RE no.:6, PIR
r ¡66. Suffect consul in ¸8 (Degrassi
:6), proconsul of Asia 68/q: Vogel-Weidemann (:q8.) ¡6¸–¸.
roz.rj Silio et Galerio Trachalo consulibus 68 cr (Degrassi :8).
Vibius Crispus RE no..8, PIR \ ¸¸q. Suffect consul under Nero in an
unknown year, suff. II ¸¡, suff. III c. 8¸ (AE :q¸:, ¡¸6), proconsul of Africa
perhaps ¸:/.: Vogel-Weidemann (:q8.) .:q. The gentilicium is transmitted as
Albius, but the error is palaeographically straightforward: u mistaken for open
a, tall i for l. An Albius Crispus is otherwise unknown. This may be the Q.
Vibius Crispus whose name is on a pipe found on Via Latina (CIL :¸.¸¸6¡).
His complete name, L. Iunius (Q.) Vibius Crispus, is now known from Fast.
Ost. (AE :q68, 6), the L. Iunius presumably by testamentary adoption. Highly
favoured under the Flavians, he was cos. III, probably in 8¸: Bruun (:qq:)
roz.rq Vespasiano .III. et Cocceio Nerva consulibus ¸: cr (De-
grassi .o).
Pompeius Silvanus REno.::6a, PIR
r 6¸¡. Suffect consul in¡¸ cr (De-
grassi :.), proconsul of Africa ¸¸–6: Vogel-Weidemann (:q8.) :6o–¸. Closely
linked with his successor Tampius Flavianus: both were cos. II together, prob-
ably in ¸6 (CIL ¡..¸6o). Tacitus, Hist. ii.86.¸ scornfully refers to both men as
divites senes. His full name, M. Pompeius Silvanus Staberius Flavinus, is known
from an inscription from Africa (AE :q68, ¸¡q); Bruun (:qq:) :66–¸.
roz.rj <Domitiano .II.> Valerio Messalino consulibus ¸¸ cr
(Degrassi .:).
Tampius Flavianus RE no.:a, PIR + ¸. Suffect consul under Claudius,
in an unknown year, proconsul of Africa ¸:/.?: Vogel-Weidemann (:q8.)
:¸o–¡. The post curator aquarum has been restored for him on an inscription
from Fundi (CIL :o.6..¸ = ILS q8¸; AE :q66, 68); Bruun (:qq:) :6¸–8.
roz.r6 Vespasiano .V. Tito .III. consulibus ¸¡ cr (Degrassi .:).
Acilius Aviola RE no..., PIR
\ ¡q. Consul ordinarius in ¸¡ cr (Tac.
Ann. xii.6¡.:, Degrassi :¸), proconsul of Asia 6¸/6: Vogel-Weidemann (:q8.)
The transmitted text would lead one to believe that Acilius served as curator
from ¸¡ until q¸, presumably his death: Bruun (:qq:) :68. Idle and aged, he
could thus conveniently bear the blame for the deplorable legacy inherited
by F. Alternatively, the phrase post quem could indicate that the post had been
vacant for something like two decades: a similar lapse has been proposed for
officers administering the grain supply: Rickman (:q8o) :q.–¸; see also Bruun
(:qq:) :¸q, who opines that F. is hardly likely to have deliberately omittednames
of senatorial curators for the sake of reputations tarnished by service under
Domitian. DeLaine (:qq¸) :¸¸, on the other hand, notes that silence regarding
Domitian and Domitianic personnel might have been very wise if this chapter
hadever beenpart of a public oration. No reasoncomes to mindwhy Vespasian
should so abruptly have abandoned the pattern of short-termcurators. Athird
possibility cannot be excluded, a somewhat longer lacuna in F.’s listing here.
Unless perhaps the frequent iteration of imperial consulships under Flavian
emperors might somehow have confused the manuscript tradition, there is
no obvious explanation for such a gap at this point. Epigraphical evidence,
however, provides the names of consulars who could plausibly have held this
office under Domitian:
:. L. Funisulanus Vettonianus: RE no.., PIR
r ¸¸o; curator aquarum in an
inscription fromForlimpopoli (CIL ::.¸¸: = AE:q¡6, .o¸), praetor c. 6o,
consul c. ¸8, governor of Dalmatia in early 8os, of Pannonia c. 8¡–¸, of
Moesia Superior 86; proconsul of Africa c. q:: see Bruun (:qq:) :6q–¸:.
.. L. Neratius Marcellus: PIR
x ¸¸ (Vidman); curat. aquar. Vrbis in CIL q..¡¸6
( =ILS :o¸.), consul in q¸, governor of Britain c.:o:–¡, cos. II :.q: Bruun
(:qq:) :¸.–¸.
Were we not restricted by F.’s testimony as to Acilius’ tenure, Funisulanus could
readily be placed as curator aquarum in the late 8os after service in Moesia or in
the qos after his proconsulship (most have regarded himas an senatorial adiutor
of praetorian status in the ¸os: see qq.¡n.). Scholars waver between putting
Neratius’ curatorship before his post in Britain (thus also before Frontinus) or
after that post. Certainty eludes us.
Fistular evidence has brought forward two further candidates of senatorial
:. M. Arrecinus Clemens: RE no.., PIR
\ :o¸.. A Domitianic pipe (CIL
:¸.¸.¸8) reads sub cura M. Arricini Clementis; cos.suff. ¸¸, cos. II ord. 8¸.
See Bruun (:qq:) .¸8–q.
.. C. Laecanius Bassus Caecina Paetus: RE no.6, PIR
c :o¡. A lead pipe
on which Domitian is styled Germanicus reads sub cura Cae[cin(ae)] Paeti
et Articulei Paeti et Nini Hastae (CIL :¸.¸.8:a = ILS 868.). Caecina Paetus
can be identified as cos. suff. ¸o/¸:, proconsul of Asia 8o/8:. The accom-
panying senators, it has been suggested, could be adiutores of the pattern
under Messala (qq.¡n.). See Bruun (:qq:) .¸q–¡o.
The cura in these cases cannot be that of the familiar fistular pattern, where the
name is that of an imperial procurator: Dressel, CIL :¸, p.qo¸; Bruun (:qq:)
.:o–:.; see :o¸.:n. If one can accept a lacuna in F.’s text, there could be reason
to imagine that Caecina Paetus might have been curator aquarum in 8¸/¡ after
the proconsulship of Asia, Arrecinus Clemens in the mid-8os after his second
consulship. Some have postulated a discrete cura for these senators: Passerini
(:q¡o) :6:–.; Eck (:q¸¡) .o8–q. In light of an important fistular inscription sub
COMMENTARY :o..:¸–:o¸.¡
cura Q. Verani, however, Bruun (:qq:) .¸¸–¡:, has conclusively shown that the
senatorial and/or consular names on lead pipes may be related to some other
urban cura entirely, for Veranius (cos. ¡q) was curator aedium sacrarum et operum
locorumque publicorum: see Gordon (:q¸.).
roz.r¡ imperatore Nerva .III. et Verginio Rufo .III. consulibus
q¸ cr (Degrassi .q).
roj.r subiungam ‘Go on to say, list’; cf. .¸.: singula subicere (see also
roj.z ius ducendae aquae Cf. q¡.: ad cohibendos intra modum impetrati
beneficii privatos. Emphasis laid on water rights for privati is hardly surprising.
Nearly half of the official delivery was distributed to privati (¸8..–¸, Table 6),
and fraudulent practices testify to still greater demand.
ne quis . . . ducat For the process of impetratio see :o¸.:.
roj.j quem adquiri diximus The water ‘discovered’ by F.’s own mea-
surements (6¸–¸¸); cf. ¸¸.: velut nova quadam adquisitione, 8¸.. quasi nova inventione
possit ad novos . . . pertinere We have already heard of new appor-
tionment to both salientes (8¸.¸) and beneficia (88..). The point of ita enim efficiemus
is that these gains will be lost without careful and constant monitoring.
roj.q in utroque Drawing water (:) without a grant (sine litteris Caesaris)
and (.) in excess of the granted amount (amplius quam impetravit).
in castellis et salientibus publicis The S.C. addresses salientes (q.qn.)
only, but water for these came through castella (¸..n.) and it was also fromcastella
(¸..n.) that privati drew any water for which they had a formal grant (:o6).
sine intermissione diebus <et noctibus> Cf. :o¡.. uti salientes publici
quam adsiduissime interdiu et noctu aquam in usum populi funderent; ::¸.¸ quo efficiebatur
ut exiguus modus ad usus publicos perveniret. The foremost concern was to assure an
adequate quantity of water for public uses. Night-time demands will not have
been high, but a greater overflow in these hours helped to maintain sanitary
standards and scoured the drains (:::..). Continuous flow (¡.¸n. confluunt) was
a feature of the entire system, for the aqueducts could not simply be shut off:
Ashby (:q¸¸) ¸6, Eck (:q8¸a) 8¸. Water-division of the sort F. mentions for
the Crabra (q.¸n.) is another matter entirely. Bruun (:qq:) ::o–:¡ attempts to
retain the transmitted text, suggesting that conservation could be effected by
flow restriction at night made possible by taps, about which the Romans were
very knowledgeable: Kretzschmer (:q6o), Fassitelli (:q¸.), Hodge (:qq.) ¸..–
¸6. Scribal omission of et noctibus is a simple case of homœoteleuton (easier than
noctibusque); diebus alone (plural, without modifier or preposition) cannot bear
the sense of interdiu ‘in the daytime’ (which need not necessarily be contrasted
with noctu). Bruun, however, is correct in his observation that practices will
have changed between Augustan legislation and F.’s time. Nor is it impossible,
even at the date of the present S.C., that means were to hand for cutting off
flow (at night or according to other temporal or local schemes) when circum-
stances urged – and permitted – maintaining reserves, e.g. in case of seasonal
fluctuation or essential repair work (:...¸).
quod . . . facere curator iubetur The immediate relevance of the
S.C. to F.’s own tenure can be questioned, since (as he notes at :o¡.¸)
the introduction of new waters under Claudius had significantly augmented
the Augustan supply: Bruun (:qq:) :o6, Evans (:qq¡) ¸¸. Yet F.’s point is to
emphasise the curator’s responsibility (:o¡..), the second rather than the first
item in the S.C.
roq.r aedificia urbi coniuncta The urbs proper was the area within
the pomerium; built-up areas adjacent to but extra urbem were known as aedificia
coniuncta (or continentia: see TLL :: q.:.¸., ¡: ¸:o..o.); cf. :.¸.:, :.q.¡. The
distinction was convenient (cf. Dion. Hal. Ant. Rom. i\.:¸.¸ :c ttpi :nv tcìiv
ciscúutvc ycpic), but apparently also a legal nicety: Tab. Heracl. (RS .¡,
Bruns :¸, FIRA :¸) line .o quae viae in urbem Rom(am) propiusve u(rbem) R(omam)
p(assus) m(ille) ubei continente habitabitur sunt erunt; Dig. iii.¸ (Paulus) in urbe et
continentibus aedificiis, xx...¡.: (Neratius). See further Dig. r.:6.... (Paulus) ‘urbis’
appellatio muris, ‘Romae’ autem continentibus aedificiis finitur, quod latius patet; :6.8¸.pr.
(Marcellus) ut Alfenus ait, ‘urbs’ est ‘Roma’, quae muro cingeretur, ‘Roma’ est etiam qua
continentia aedificia essent: nam Romam non muro tenus existimari ex consuetudine cotidiana
posse intellegi, cum diceremus Romam nos ire, etiamsi extra urbem habitaremus; :6.:¸q.:
(Ulpianus) aedificia ‘Romae’ fieri etiam ea videntur, quae in continentibus Romae aedificiis
fiant; :6.:¡¸.pr. (Terentius Clemens) qui in continentibus urbis nati sunt, ‘Romae’ nati
intelleguntur; xxxiii.q.¡.¡ (Paulus) si autem extra urbem, Romae tamen sit, sed et si in
hortis sit urbi iunctis, idem erit dicendum.
M. Agrippa See q.qn., q8.:n.
quos nunc esse ‘Which are now in existence’. A precise number is
not needed. The prefatory reference to Agrippa is no more than a polite
acknowledgement of his accomplishments (cf. :.¸ Augustus . . . pollicitus senatui
est), and the commission’s report was not limited to salientes built by Agrippa
(which numbered :o¸ according to Pliny, HN xxx\i.:.:).
quibus negotiuma senatu The number and rank of this commission’s
members are not specified, but the high standing of Agrippa and prestige of
Messala Corvinus would suggest a consular board. Note that consulars served
on similar commissions (albeit in emergency circumstances) relating to the cura
annonae in 6 cr (Dio, r\..6.) and the cura alvei Tiberis in :¸ (Tac. Ann. i.¸6..). The
two-fold mandate gave this board considerable latitude. (:) ut inspicerent aquas
publicas implies a comprehensive reviewof the water-system, but their attention
COMMENTARY :o...–:o¸.:
may naturally have focused on administrative policies rather than technical
matters. (.) The inventory (inirentque numerum; for the phrase cf. Tac. Ann. ii.8¸..)
of salientes will in effect have been a means of determining the adequacy of the
existing supply and the system of distribution. Special emphasis on public
basins reflects the long established concern that free water be available in this
way to the entire populace (cf. q¡.¸, q¸.¸, q¸.8).
roq.z quos . . . nominavit See :oo.:n. S.C. is best deleted (in origin
perhaps a marginal note; cf. qq.¡). Kornemann, RE ¡: :¸¸¸.¸o saw in it a ref-
erence to an earlier resolution that created the post of curator aquarum, with the
senate’s auctoritas (consensus in :oo.:) applying only to the manner of appoint-
ment. On the issue of curatores aquarum deriving authority ex senatus consulto, see
Bruun (:qq:) :8o–..
quam adsiduissime Quam is specially emphatic, for the superlative
alone is rare enough: Cic. Brut. ¸:6, cf. Suet. Aug. ¸:.:.
roq.q id factum existimo F.’s explanation does not wholly satisfy. Had
the quantity for public fountains been judged inadequate, the pattern of dis-
tribution (cf. q8..n.) could in theory have been shifted without requiring an
increase in the overall supply. Losers might predictably have been privati (sen-
ators, no doubt), and on this point a cautious conservatism may have been
felt in order. The senate seems to have been affirming the priority of salientes
in the process of ratifying or redefining Agrippa’s apportionment (qq.¸n.). For
possible readjustments later on, note :¸.: (Gaius’ initiative) cum parum et publicis
usibus et privatis voluptatibus septem ductus aquarum sufficere viderentur.
Claudia et Anio novus The dramatic impact of these two new aque-
ducts can be seen from a glance at the numerical data (¸.–¸, 86; see Table ¸):
official figures credit them with 6,::8 quinariae, ¡q.¸ per cent of the total concep-
tio; ¸,6.¸ quinariae, ¡:.¸ per cent of erogatio (neglecting in both cases Alsietina).
The combined length of their conduits totalled more than :oo Roman miles,
some two-thirds again that of the earlier works (Table :).
roj.r volet . . . debebit Perhaps the future tense signifies modifications
of present or earlier practice. In any case F.’s discussion is plainly prescriptive.
Note also §¸ debet and the jussive subjunctives in §§¸–6 cogitet, intendat, habeat,
permittatur. See further ::..6n. diligentiae adiciendum est.
impetrare Impetrare ‘to ask for and get’ (cf. Berger ¡q¡) describes the
procedure whereby a privatus acquired by personal request the ius ducendae
aquae (:o¸..). Delivery of public water to individuals was never recognised
as a universal right, only a favour on the part of the state. The favour was
called beneficium (¸..n.), and power to confer it lay with the princeps (qq.¸n.). A
grant was contingent, of course, upon the availability of water in the particular
locality for which petition was made (:oq.:n.). The favour depended further,
and no doubt far more importantly, on the ease with which the petitor (:oq.:)
had access to the princeps: Eck (:q8.c) :q¸–.o8; Bruun (:qq:) :¡¸–q. It ought
perhaps to be stressed that F.’s concern here is with privati alone, with no
distinction made between the many uses to which those persons might have
put the water (¸..n., :o¸.¸). Was any kind of regular payment required from
those privati who received their right to water by means of imperial beneficia?
Many scholars assume that the water thus granted was subject to a vectigal (as
had once been the case with water for baths and fulleries: q¡.¡), the income
thereby produced helping to defray the costs of operating the public system.
Such a concept is indeed set forth by Vitruvius, \iii.6.. uti qui privatim ducent
in domos vectigalibus tueantur per publicanos aquarum ductus, and such practices are
attestedelsewhere (e.g. at Venafrum: CIL:o.¡8¡..¸8 = ILS¸¸¡¸.¸8). At Rome,
however, there is no clear evidence to support this view, for the vectigalia F.
discusses at ::8.:–¸ are at best ambiguous. Again, widespread willingness to
pay bribes (:oq.., ::.–:¸) is no proof that a charge was imposed for delivering
water that had been officially granted (the abuse reveals opportunism on the
part of the staff: cf. q.6, ::¸.¡). Neither F.’s silence nor the administrative
vocabulary (impetrare, beneficium, petitores) can constitute a solid basis for arguing
that deliveries were made without charge. Yet, given the political and economic
viewpoint of the principate, it would be surprising indeed to find privati (even,
as they might be, principes civitatis) expected in this way to subsidise the routine
maintenance of the public water-system, even if the cost would be trifling in
comparison to their wealth. Names of lesser persons and of lower status appear
on surviving pipes, so the opportunity for impetratio/beneficiumseems not to have
been limited, at all times at least, to the elite. Water to be used for commercial
purposes also required an official grant (:o¸.¡), and such users might have
been expected to pay a fee – whether or not this was categorised as a vectigal.
(Of course, we should beware of assuming anything about this latter practice
solely on the grounds that it is normal in modern times.) No need, however, to
doubt that privati of any status bore most of the expenses attending the initial
delivery, as well as those for maintaining castella privata (:o6.:). And perhaps
the process of petitio and impetratio involved a nominal payment to the imperial
bureaucracy, or (in some cases, and perhaps more likely in all) a bribe to a well
placed courtier; cf. Rickman (:q8o) :qo–:.
The majority of scholars opine that privati paid for their water rights (usually
on the basis of F. q¡.¡, :oq.., ::8.:–¸, or practice at Urso and Venafrum):
Hainzmann (:q¸¸) 6¸, Millar (:q¸¸) :q¸, Rickman (:q8o) :q:, Bruun (:qq:)
.oq, Evans (:qq¡) q, Petrucci (:qq6) :q¸–¡, Taylor (.ooo) q¸–¸, Blackman–
Hodge (.oo:) ::6 (cf. ¸¸n. onerandam). Favouring the idea that beneficia were
free of charge are Grimal (ed. p.q¸ n.:¸.), Eck (:q8¸a) ¸q–8o, Geißler (:qq8)
¸8–8o, :¸8 and n.:8. and Bruun (.ooo) ¸8q (the latter a modified view). De
Kleijn (.oo:) q¸–:oo is uncommitted; cf. also Virlouvet (:qq¸) .o¸n.
a principe epistulam Cf. :o¸.. litteris Caesaris. In this document was
specified the name of the grantee (:oq.:), the location of the property, the
castellum from which water could be drawn (:oq.6), and the capacity of the
granted pipe (§¡n.). The details of petitio are unclear, although records needed
to be searched to assure that there was in fact water available and in the right
place (:oq.:). Although the letter no doubt followed a certain formula, it was
issued not by an imperial bureaucrat acting on his behalf, but by the emperor
himself. And it was delivered to the petitioner, not forwarded directly to the
curator (as might happen with modern chancery memoranda). The strong
personal element of Roman bureaucracy is doubly apparent.
ad curatorem . . . curator deinde The curator’s involvement strikes
one as mildly inefficient. It might have been newly devised as a safeguard
against fraud (on the part of the procurator no less than the vilici: see ::..¸–¡).
It is equally possible, however, that this was a role of the original curators
(cf. :o6.:) which F. will rescue from desuetude.
maturitatem ‘Speed’ or ‘promptness’ per se is not the only issue here;
implicit is an underlying contrast with inertia, not unlikely an indirect com-
ment on F.’s predecessors. Upon receiving the epistula principis, the curator is
to verify the document and arrange for implementing the grant with ‘due and
timely consideration’. Gellius, x.::.¸ (explaining Augustus’ ott0ot µpcotc,)
gives a definition that would have pleased F. no end: ut ad rem agendam simul
adhiberetur et industriae celeritas et diligentiae tarditas, ex quibus duobus contrariis fit
<ad>procuratorem. . . scribere The next step is to send written no-
tice of the new grant to the procurator, who is responsible for official markings
on certain pipes (§§¡–¸, ::..¸). The transmitted text cannot stand. New pro-
cedures or not, it passes belief that the curator ‘appointed in writing’ (scribere +
accusative; cf. OLD s.v. 8) an imperial freedman as procurator: the emperor
himself or one of his senior bureaucrats would have made that choice. On the
other hand, F. envisions a close relationship between princeps and curator, a
circumstance that might have allowed the latter some expression of personal
preference in the choice of any ‘assistant’. Malissard (:qq¸) ¸66 is mistaken
to say of the curator, ‘d’ordre s´ enatorial, il nomme un procurateur d’ordre
´ equestre [!] qui est de nouveau enti` erement plac´ e sous ses ordres et ne d´ epend
que de lui’, and ibid. n.¸¸ ‘le procurateur cr´ e´ e sous Claude (:o¸..) a donc
perdu ind´ ependance et r´ eel pouvoir’. It might be noted, however, that while
procurators were still imperial freedmen when F. was writing (.n.), fistular
evidence reveals an equestrian with this title under Trajan (between :oq and
::¸, one C. Pomponius Hyllus: CIL :¸.¸.qq–¸q:, PIR
r ¸¡.): for discussion
of the transition and scholars’ views on the change see Bruun (:qq:) .:8–.o.
eiusdem officii I.e., aquarum? And perhaps officium here = ‘sphere
of responsibility’. This puzzling locution may indicate that more than one
procurator might be involved; cf. Bruun (:qq:) .:¸–:¸. The phrase ought
perhaps also to mean more than ‘the appropriate procurator’. Both in nomen-
clature and by rank F. sees the procurator as clearly subordinate to the curator,
and eiusdem officii may lay stress on the fact that there is to be a single adminis-
tration regulated by the curator as the emperor’s deputy (cf. : mihi . . . aquarum
iniunctum officium; 6¡.¸ praecipuum officii opus; :¸o.¡ officii fidem).
libertum Caesaris A number of liberti Caesaris are known from lead
pipes on which the legend sub cura is followed by a name in the genitive. The
earliest of these is one Gnesius (under Nero), who appears in CIL :¸.¸.¸:; he
is nowhere specifically called procurator aquarum; on this Gnesius see nowBruun
(:qqq). Another, a certain Bucolas, appears on two pipes: CIL :¸.¸.¸q ( =ILS
86¸q): Imp. Domitiani Caesaris Aug. Germanici sub cura Bucolae l(iberti) proc(uratoris)
Ti. Claudius Philetaerus fec(it), also CIL :¸.¸.8o without l(iberti) and with the
maker named Fortunatus lib. This man’s career appears in another inscription:
CIL ::.¸6:. ( = ILS :¸6¸) Ti. Claudius Aug. lib. Bucolas praegustator, triclinarc(ha),
proc(urator) a munerib(us), proc. aquar(um), proc. castrensis. For Bucolas (PIR
c 8:q)
see Bruun (:qq:) .:¸–:¸ with bibliography. Because Bucolas in this inscription
is called procurator aquarum (the only instance of this title in the first century),
it has been assumed that his (unspecified) procuratorial role on the fistula is
specifically proc. aquarum; from this instance it is argued that other Augusti liberti
on pipes are likewise proc. aquarum. Nine such men occur under Domitian alone,
which is a puzzle in itself: see Eck (:q¸8) ¸8¸. It is probably too simple to view
them all as incumbents of a single office called proc. aquarum: see Bruun (:qq:)
But as proof that there really was an ‘aquarum office’ staffed by impe-
rial freedmen, Bruun (:qq:) :q¡ (following Boulvert (:q¸oa) :¡8 n.¸¸:) cites
three imperial freedmen (Aug. lib.) with titles a commentari(i)s aquarum (CIL
6.8¡8¸), tabul(arius) aquarum (CIL 6.8¡88), tab(ularius) rat(ionis) aquarior(um) (CIL
:o.:¸¡¸ = ILS :6o8); ontabularii andcommentarii ingeneral, see Boulvert (:q¸oa)
¡.o–¸. Dating of such material is unhelpfully vague (see also below::¸.:n.). By
procurator eiusdem officii could F. be referring to a man in charge of such a bureau,
head of a corps responsible for clerical duties and other chores of implement-
ing imperial beneficia? The apparent turnover of proc. aquarum under Domitian
might reflect instability of a kind occasioned by administrative changes being
made during this period: in such circumstances, perhaps, F. might have had
additional reason to deplore inertia ac segnitia of senatorial curators (:o:..).
roj.z procuratorem. . . videtur admovisse F.’s uncertainty may be
not so much over when a procurator first appeared as why one was thought
necessary. Duties of an initial procurator are only to be surmised: responsibility
for the familia Caesaris (::6.¡), more immediate supervision of routine upkeep,
and perhaps financial matters. It is not clear to what extent – if at all – the role
COMMENTARY :o¸.¸–:o¸.¡
of a procurator overlapped with that of the curator. Admovisse might suggest
that the post was intended to be complementary, to spare the senatorial offi-
cial, as it were, the additional burdens associated with the two new Claudian
aqueducts. Consistent with his own imperial responsibility (cf. ::6.¡), Claudius
would have fully respected the existing administrative institution (despite the
fact that the nature of the senatorial cura was apparently undergoing some
change in this period: see :o..¸n.). Inefficiencies of such a dual administration
might no doubt have been apparent soon enough, and at :o:.. and ::8.¸ F.
might be hinting that consular curators had not been loath to surrender their
duties to an imperial ‘adjunct’. The institution of a procurator aquarum has long
been viewed as part of the emperors’ gradual assumption of more adminis-
trative control: Hirschfeld (:qo¸) .¸¸, Boulvert (:q¸oa) :¡8–q, Palma (:q8o)
.o.. F.’s text thus represents something of a ‘senatorial reaction’: Grimal (ed.
p.xv), Pflaum(:q¸o) ¸¸, Boulvert (:q¸oa) :¡8, D’Amato (:q86) :8.. Against the
view that emperors deliberately attempted to limit senatorial power by use of
freedmen and equestrians see Brunt (:q8¸), Eck (:q8¸b); cf. Alf ¨ oldy (:q8: ) and,
overall, Bruun (:qq:) .o¸–q.
postquam Anionem novam et Claudiam induxit See :o¡.¡n.
roj.j vilicis quoque notumfieri debet Cf. ::..¸ vilicus . . . omni carebit
excusatione. It is unclear why F. uses the plural here, for there is unlikely to have
been more than one working crew superintendent (::¸.:n.) involved. Perhaps
one should read vilico and then defendat. As Heinrich saw, quoque belongs with
vilicis. For the word-order notum fieri, cf. Pliny, HN xxx\ii.¡8 verum hoc quoque
notum fieri oportet.
roj.q calicem . . . signari The procurator’s responsibility for properly
marking the calix (¸6.¸–¸nn.) is noted below::..¸–¸. (It is important to observe
that F. here discusses marks on the calix only: for separate marks on fistulae see
::..6). Frompresent context it is clear that the marking is a means of certifying
that the size of the modulus installed corrresponds to that which was granted:
see ::..¸n. legitimam mensuram. Since grants were made to persons – either
individuals or partnerships (:oq.¡–¸) – it has been a convenient assumption
that F. speaks of markings that show(:) a pipe-size, probably a numeral, and/or
(.) the name of a recipient. Because no identifiable calices have been found, we
are ignorant of actual practice – except for what F. tells us about detecting calices
larger than granted (::..¸) and ones without marking at all (::..¡). For want
of calices, however, scholars have looked to markings on lead pipes in hopes of
identifying patterns that conform to F.’s hints. In a recent and thorough study
of such fistular material, Bruun observes that numbers on surviving pipes are
in many cases quite obviously unrelated to size and argues that the names on
the pipes may in many cases, but not necessarily always, be those of recipients
of water grants: Bruun (:qq:) ¡:–6.. Whether F. speaks descriptively, of marks
in use, or prescriptively for his own underlings and successors, it remains true
that he ‘did not state explicitly what the content of lead pipe marks had to
be’: DeKleijn (.oo:) :¸8. Bruun (:qq:) ¸q–6. speaks of practical difficulties
which would handicap any process of updating marks subsequent to an initial
installation (cf. ::¡.:n. foramen novum).
eius moduli qui fuerit impetratus That the calix conforms to pipe-
size (eius moduli) is evidence that grants were issued or deliveries were made
(cf. ¸..¡–¸) by pipes rather than by quinariae, and to a limited extent we
find apparent confirmation in surviving pipes: Dressel, CIL :¸ p.q::; Bruun
(:qq:) ¸¡–¸. Only a few of F.’s official pipes have even multiples of quinariae
(Table ¡), and computation of total deliveries, expressed in quinariae, must have
been somewhat complex (cf. ¸¸.¸, ¸¸..). This may partly explain the discrep-
ancy noted at 6¡...
adhibitis libratoribus These ‘levellers’ are hydraulic experts who han-
dle the technical matters of delivery, presumably the whole process of laying the
water line froma distributory castellum to a final delivery point. In all likelihood
they were members of the regular staff (among the opifices at ::¸.:); cf. ::q.¸n.
They receive their orders from the procurator but, F. insists, they should be
monitored by the foreman (vilicus), who would also be held responsible.
mensurarum quas supra diximus Chapters ¸¸–6¸.
positionis notitiamhabeat For what F. means by ‘setting’ see ¸6.:–.,
maioris luminis . . . minoris . . . calicem probare For lumen, the
‘orifice’ of the pipe, see .q.:n. The marking of which F. is speaking here is one
which certifies the size of the conduit: see further §¸n. eiusdem luminis quo calix
signatus est.
pro gratia personarum Perhaps gratia here refers to a bribe, although
more likely it is that normal ‘influence’ which brings pressure to bear without
the need for money changing hands. It is a little hard to imagine a grand per-
sonage stooping to the lowly libratores when arrangements might conveniently
be made at a higher level. For gradations of social distinction and power cf.
Dig. iii.¸.:¡.pr., Serv. Aen. i.66¸ pro qualitate personarum; Dig. xriii :¸.:.¸ secundum
qualitatem personae; Dig. xx\ii.¸.¸¸.¡, xr\iii.:8.¸.pr., Serv. Aen. i.¸.q, \i...6 ex
personae qualitate; Dig. xri\...:¡.pr., xr\ii.q.¡.: condicio personarum.
roj.j plumbeam fistulam F. uses the adjective no doubt to contrast
with the bronze of the calix (¸6.¸–¸); he uses fistula alone at ::¸.¸–¡ in quorumdam
fistulis ne calices quidem positi fuerunt. hae fistulae solutae vocantur.
eiusdem luminis quo calix signatus est In ::..6 F. recommends
the additional step that pipes be marked as well as the calix. The marking on
the calix (and by implication on the pipes) is one which indicates its size or
calibre (lumen); we have no hint here whatsoever of a requirement that either
COMMENTARY :o6.:–:o6..
calices or fistulae be marked with the names of the individuals receiving their
water. Names on existing pipes, therefore, need not reflect any regulation,
existing in F.’s day or prescribed by him, and we are not therefore bound to
an interpretation which insists that they are those of recipients.
per pedes quinquaginta The precaution is explained below (::..¸).
Fifty feet (:¡.8 m) is the length similarly specified in the Venafrum edict CIL
:o.¡8¡..¡¡ = ILS ¸¸¡¸) line ¡¡. This would be five lengths of lead pipe, each
:o feet (..q¸ m), according to the standard prescribed by Vitr. \iii.6.¡ (cf. Fav.¸,
Pallad. ix.:.) and Pliny, HN xxxi.¸8.
quod subiectum est F. uses subicere to introduce verbatim quotations
from documents (:o¸.¸, ::o.., :.¡.¡, :.6.6); cf. qq.¸ infra scriptum est, :.8.¡
subscripsi, :o¸.¸ haec verba (also q¡.¸). For the same use elsewhere, Pliny, Ep.
x.¸6.¸, ¸8.¡, ¸q.¸, q..:, ::¡.¸, Suet. Gram. .¸.:, Gell. iii.¸.:q, i\.¸.¡. See also
:o..:n. subiungere.
ro6.r aquae ducendae ius The first attestation of this phrase according
to Capogrossi Colognesi (:q66) 8: n.:¸o; see further ¸..n.
intra extra urbem Deliberately more general than the formulae at
:o¡.:, :.¸.:, :.q.¡. B¨ ucheler wanted to be rid of extra, perhaps on the grounds
that castella were only for urban distribution (chapters ¸8–86).
castella privati facere possent Under no circumstances are privati to
drawwater frompublic conduits: they receive it instead fromdistributory tanks
(castella). F. does not regularly distinguish between various types of castella: (:)
those that modern scholars refer to as ‘terminal’ or ‘distributory’ (¸..n.), (.)
those that are part of the main systemof urban distribution (cf. :q.¸, ¸6.¸, ::¸.¸,
::8.¡) and (¸) those which served as the final delivery tanks for consumers (cf.
.¸.¸, ::¡). The last are meant inthe present passage. They might fairly be called
castella privata (¸..n.), although they were not necessarily the private property
of the persons they served. Supervision surely, and perhaps maintenance as
well, were in the hands of the water-board (q8.¸, :o¸.¡, :oq.6, ::o.:, :::..,
::8.¡). Grimal q¡ n.:.q, it may be noted, supposes that castella privata will only
be constructed if a public tank is at too great a distance. Distance, however,
is but one of several hydraulic conditions that need to be taken into account
aquam . . . communem Cf. .¸.¸ plures quinariae impetratae . . . una fistula
excipiuntur in castellum, ex quo singuli suum modum recipiunt.
quam . . . accepissent a curatoribus Curators only implemented
imperial beneficia (qq.¸n., :o¸.:); they did not themselves make grants to privati
(cf. :.q.::n.).
ro6.z quam quinariam It is of course possible that quinariam has re-
placed some phrase of more general definition, e.g. ‘one of that same size
which had been granted’ (cf. :o¸.. ne quis amplius quam impetravit ducat). When
COMMENTARY :o6.¸–:o8
F. writes, grants could be of more than one pipe-size (:o¸.¡n.), but this might
not always have been the case. Augustus’ beneficia might deliberately have been
limited to quinaria pipes (cf. :o¡.¡): Bruun (:qq:) :::–:..
ro6.j ne . . . lacerentur Cf. .¸.¸ ne rivus saepius convulneretur. Beyond
incidental damage to the conduits, uncontrolled tapping – especially of
pipes – would seriously affect the entire system of urban distribution
(cf. ::¸..–¸).
aut rivi aut fistulae publicae Bruun (:qq:) :.¸ notes that fistulae here
are explicitly called public (see .o..n.) (as rivi are more frequently): it is F.’s
only explicit indication that lead pipes were used for water mains (cf. ::o.:n.).
ro¡.r ius . . . sequitur Cf. Dig. xriii..o.:.¡¸ (cited at :oq.6 below).
Grants were not a praedial servitude running with the land; they were strictly
personal, in personam (for the custom governing a societas see :oq.¸), and
could be transferred neither by testamentary disposition (heredem) nor by sale
(emptorem). The ius impetratae aquae could be acquired solely as a beneficium by the
process of impetratio (:o¸.:n.). For some apparent instances of circumventing
this requirement, see Eck (:q8.c) .o8 n.¸¸.
ro¡.z antiquitus concedebatur For ‘baths open to the public’ see
q¡.¡n. The imperfect tense is important. Although there is no indication of
when the change took place, bathing establishments no longer enjoyed perpet-
ual water-rights. Nunc . . . beneficium (§¸) contrasts the contemporary practice:
every grant is subject to renewal (cf. :oq.¸n.). Whether the new rule applied as
well to grants made haustus nomine (:o8) is not clear; the position of the parenthe-
sis suggests that it might have (the citation would normally follow immediately
upon subieci: cf. qq.¸, :o¸.¸, ::o.., etc.).
ro¡.j omnis aquae cum possessore instauratur beneficium Cf.
ro8 intra [extra]que It is hard to determine what difference, if any, can
be drawn in legal texts between the enclitics -que and -ve: Crawford (:qq6) :q.
adtributio aquarum Adtributio, distribution of what is public property,
is synonymous with adsignatio (Berger ¸¸:–¸). But distribution of public water is
unlike that of ager publicus by which public land becomes private property. Water
which flowed perennially was by definition public, but it could be granted or
assigned for particular purposes or to individual persons.
haustus nomine Understand either essent datae or essent, for it is not
clear whether ius hauriendae aquae, like ius ducendae, could be granted (datum).
Like iter, actus, aquae ductus, haustus (‘hauling up’, i.e. by mechanical means) was
an ancient praedial servitude: Dig. xriii..o.:.6 (Ulpianus) ceterum sunt quaedam
(sc. aquae), quae, etsi perennes sunt, duci tamen non possunt, ut puta puteales et quae ita
COMMENTARY :oq.:–:oq..
sunt summersae ut defluere extra terram et usui esse non possint. sed huiusmodi aquis, quae
duci non possint, haustus servitus imponi potest. Since this S.C. deals with adtributio
(applying to distribution of public property: cf. :.q.¡ data vel adtributa), it seems
likely that some water rights might have been granted – or perpetuated, if they
represented a preexisting state – in this special way. Observe that sumere haurire
appears among certain exceptions at the end of the Lex Quinctia (:.q.::n.).
Haustus nomine, which we find here, seems to be effectively, though not perhaps
technically, the same as haustus iure (or iure hauriendae aquae); as such it may have
enjoyed the same perpetuity as would a praedial servitude (but cf. :o¸..n.).
Note the mention of id solum in quod accepissent aquam: in the case of haustus, no
change of ‘title’ – even a new ownership, dominium – would alter the right (for
it belonged to the land, not the owner or occupant).
idemdomini . . . idsolum Unless this S.C. has beentruncated, those to
whomthe ius ducendae aquae applies here are legal landowners (domini). Contrast
:oq.¡ ii ad quos res pertinerent, who could be either domini or possessores.
roq.r Cum vacare . . . coeperunt aquae, adnuntiatur Vacare, of
property ‘to be without owner’, appears first in Livy (OLD s.v. . b). Adnuntiare
appears only in mid century (TLL :: ¸8¸), and its usage seems to connote a
formal or official report: Sen. Dial \ii..8.: iam funesta domus est nec adnuntiatum
malum; Pliny, HN \ii.:¸¸ adnuntiavere exanimatum illum; Curt. x.8.:: adnuntiatur
equites frumentum retinuisse. The word is rare until it blossoms in Christian writers.
F. Str. i.q.. propere enim adnuntiari iussit hostem adesse, Stat. Theb. \ii.¡¸¸ rumor
plures adnuntiat hostis; Suet. Aug. ¡q.¸, Vit. q.:. Water became ‘vacant’ under
circumstances outlined at :o¸.: (cf. also :oq.¸), but F. specifies neither who is
to report this fact nor to whom it is to be reported. Perhaps the responsibility
was made incumbent upon the recipients (or their heirs) in the process of
impetratio. Had it been a duty of water-men to report such changes, F. might
have noted their shortcomings: the aquarii, who had profited from an interval
between grants, would predictably have delayed the report or overlooked it
in commentarios . . . petitoribus Records of deliveries were kept
in the imperial bureaux (¸:..n.; cf. ¸¸.¸, 6¡.:–., ¸¸.¸). On the persons who
sought beneficia see :o¸.:n. Qui respiciuntur here implies that the lists were of
aquae vacuae, although there may also have been ‘waiting lists’, for in some
areas demand might have exceeded the official supply (cf. ¸¸.¸, ¸¡.¡).
ex vacuis dari A formal grant. For this general, but official, use of dare
see ¸..n. detur.
roq.z has aquarii . . . intercipere solebant For intercipere ‘to shut off,
divert’ see ¸.:n. (often involving fraud, e.g., ¸¸.:, 8¸..). Neither verb intercidere
is apt: inter + cadere is intransitive (6¸.¸, 66.¸, etc.), inter + caedere ‘to cut to
pieces’ is absurd. It is uncharacteristic for F. to leave inexplicit the subject(s)
COMMENTARY :oq.¸–:oq.¡
of solebant and venderent. We can easily spare aquas here, and aquarii makes the
ideal subject. The most cogent argument is his use of venderent, because desire
for lucre is these men’s trademark: q.6 largiendo compendi sui gratia, ::¡.: venalem
extrahunt aquam, ::¸.: aquariorum tollendus est reditus.
medio tempore Between expiration of a former grant and a new im-
petratio. Although a new owner or tenant had no legal assurance of the right
to water, there is an implied presumption that an imperial beneficium would
be forthcoming (cf. :o¸.¸ instauratur, :oq.¸ renovaretur, ::¡.: translata in novum
possessorem). Renewal was neither instantaneous nor automatic: routine for-
malities took time (§¸), and there might conceivably have been others whose
applications had priority (§:).
venderent Contrasts with dari (§:): the aquarii had neither ius dandae nor
vendendae (cf. :o¸), and their ‘sale’ was a matter of taking money for not shutting
off the water. Willingness to make temporary payment for the convenience
of uninterrupted delivery is entirely understandable: a sympathetic attitude
underlies Nerva’s introduction of thirty days’ grace. Others who could not
look forward to a beneficium might have taken advantage of an offer on the part
of the aquarii to acquire the use of water – quite unofficially, of course, and
therefore not necessarily limited to the interval between grants (cf. ::¡.:).
aut aliis etiam Cf. ¸6.. cenacula etiam. Either F. is very careless or there is
special emphasis (indignation?) conveyed by etiam postponed to final position.
roq.j humanius . . . indulgeri Humanitas and indulgentia (cf. :¸o.¸),
particularly towards the senatorial class, were general attitudes projected by
the regime. Generosity derives in part from the discovery that the water sup-
ply was larger than had formerly been thought: renewal could now be vir-
tually guaranteed, and a larger number of petitores could be accommodated
ii ad quos res pertineret <***> Jordans (:8¡¡) ¡q supplied petere
possent beneficii instaurationem – no more than exempli gratia – which derives from
:o¸.¸ (although that sentence applies to public baths). Poleni thought the sense
wanted was an opportunity to make arrangements as needed, but without the
implied likelihood of a newgrant. Perhaps the specified length of time is aimed
at increasing bureaucratic efficiency (cf. :o¸.:). This, no less than a policy of
routine renewal, would severely limit the possibilities for impropriety on the
part of the aquarii. I should like to suppose that this sentence ended with
optimism and encouragement: <novum beneficium impetrarent>.
roq.q in praedia sociorum Cf. §¸ modus praediis adsignatus. F. does not
mean that the concession was made in praedia in the sense of a perpetual grant
to run with the land, but that the property would continue to have the benefit
so long as there remained any member of the original partnership.
nihil constitutum invenio Contrasts with §¸ observatur ac iure cautum.
COMMENTARY :oq.¸–::o..
roq.j impetraverant The pluperfect better matches datum erat.
modus . . . flueret For the meaning of modus see ¸..n., for fluere ¡.¸n.
roq.6 epistula principis continebit The future tense is prescriptive
(:o:.¡n., ::..6n.).
alio . . . aut ex alio castello A beneficium is defined in personam, and
so strictly that it is further limited to a single location. There is nowhere
a suggestion that any grant is in praedium (even that once to baths, etc. is
now obsolete: :o¸.¸). Cf. Dig. xriii..o.:.¡¸ (Ulpianus) et datur interdum praediis,
interdum personis. quod praediis datur, persona extincta non extinguitur: quod datur personis,
cum personis amittitur ideoque neque ad alium dominum praediorum neque ad heredem vel
qualemcumque successorem transit.
mandatis prohibetur That both destination (praedia) and source (castel-
lum) are subject to regulation is clear from the senatus consulta (:o8 solum in quod
accepissent, :o6.: ex castellis ducerent). But mandata (¸..n., ::o..n.) contain more
specific and detailed requirements.
rro.r caducae Cf. q¡.¸n. The definition given here, which does not men-
tion overflow from lacus, may imply that grants were limited to certain kinds
of aqua caduca. Possibly some of this water was still furnished for commercial
uses, as in the old days (q¡.¸–¡), although very likely under entirely different
regulations (cf. :o¸.¸). In any case, the aqua caduca available for grants must
have been reasonably reliable – even if perhaps only at night or for hours of
lighter use.
id est . . . fistularum Corradinus de Allio (:¸¡.): ‘in Codice nullo
reperies effluunt, fictitium et inane. Facile enim & hic subintelligitur effluunt,
labuntur, veniunt, & his similia’. The ellipsis is also eased by the verbal force in
the noun manationes.
ex castellis ‘Leaks’ at Porta Capena (:q.qn.) and those mentioned by
Martial i\.:8.: (qua vicina pluit Vipsanis porta columnis) are possible examples of
what F. describes here. Overflow from delivery tanks (castella) may also have
served to standardise the discharge (¸¸n.).
manationibus fistularum See 66.¸n. manationes. Not, probably, leak-
age due to wear or damage (for which repairs would have been expected), but
rather some kind of deliberate discharge directly from the pipes. Possibly we
could see these pipes as major water mains: Bruun (:qq:) :.¸.
parcissime tribui The simplex verb probably = adtribui (:o8n. adtribu-
tio). For the adverb cf. Suet. Aug. .¸.¸ (coronas) quam parcissime et sine ambitione ac
saepe etiam caligatis tribuit, ¡o.¸ civitates Romanas parcissime dedit.
rro.z fraudibus aquariorum Perhaps the abusive puncta (::¸.:–¸) were
an illicit extension of beneficia granted ex manationibus fistularum.
COMMENTARY :::.:–::..¸
ex capite mandatorum Caput = ‘chapter’ tells us at least that these
mandata (¸..n.) were issued in an systematic form, very likely as a liber: Finley
(:q¸¡) :¸¸; Marotta (:qq:) :¸, .6 n.¸¸, ¸¸–6. The citation strongly suggests,
although F. does not say so directly, that these mandata are specifically addressed
to the curator aquarum.
quod subieci C’s gap is presumably a vestige of editorial emphasis, al-
though B¨ ucheler may have been right in thinking that a heading has been
rrr.r meo . . . aut priorumprincipum Note the use of the first person:
mandata are personal instructions on the part of the emperor. The identity of
the princeps is uncertain, except that he must be post-Tiberian. Since we know
too little about these mandata, we can only speculate at what period they may
have become ‘codified’ and whether, like the praetor’s edict, they developed
into a fixed text (which could more readily be updated).
rrr.z salubritatem. . . utilitatem See :n. salubritatem. The advantages
of public health and sanitation are mentioned at 88.¸n.; cf. Scobie (:q86)
¡o¸–.¡, esp. ¡:¸, where he estimates that a city of Rome’s size would have
produced ¡o,ooo–¸o,ooo kg of human waste daily. Note also Dig. xriii..¸.:..
(cited 88.¸n.).
rrz.r Explicitis quae . . . pertinebant Note omission of noun or pro-
noun in the ablative absolute, antecedent to relative clause (cf. ¸¸.:n. emen-
datis quattuor). Similar is the omission of a noun with ex iis quibus later in the
non ab re est Cf. :¸.:n. non alienum mihi visum est. For non ab re est similarly
used in transitional passages, see Pliny, HN xxxi.¡¸, Suet. Aug. q¡.:; (with duco)
Livy, \iii.::.:, (with iudico) Pliny, HN xx\ii.:¡¸.
in ipso actu ‘In the mere performance of my duty’, i.e. without making
a special effort to uncover misdemeanours. Ipso is mildly emphatic (as e.g. :q.q
ipsius montis, q. ipsius urbis, :.:.¡ in ipso alveo); we find in actu (with esse, mori) in
a sense approximating ‘on the job, at one’s work’ in Sen. Dial. \iii.:.¡, ¸.¸,
Ep. 8.:. By its position the prepositional phrase is adverbial to deprehendimus.
It makes little sense to take it with circumscribi constitutiones: one might catch
persons, or persons might be caught (cf. Sen. Ep. :o:.¡), ‘in the act’ of doing
something, but regulations can be neither ‘thwarted in the act’ nor ‘in the act
of being thwarted’ – unless we specify by whose act.
rrz.z calices . . . positos See ¸6.:–¸, :o¸.¡n. positionis notitiam.
rrz.j legitimammensuram ‘Legal calibre’ means in compliance with
the standard sizes F. has outlined in chapters ¸q–6¸. For the legitimacy and
authority of these pipe-sizes see ¸:.. in commentariis principis positi et confirmati
COMMENTARY ::..¡–::..6
sunt, qq.¡ (Augustus) modulos constituit. There is no indication in F.’s use of the
word legitimus that markings are required by some regulation (see :o¸.¸n.).
ambitio procuratoris For the role of the procurator see :o¸.¡. For the
vice cf. ::¸.¡ ambitio aut neglegentia praepositorum; Suet. Gram. q.¸ de iniuriis quas
professores neglegentia aut ambitione parentum acciperent; Pliny, Ep. i.8.:¸, :¡.¸ ambitioni
dicam an dignitati (the political dangers of a senatorial career, in contrast to
equestrian honesta quies: Sherwin-White ad loc.). It is interesting that F. explicitly
says which persons might be involved. Elsewhere much of the fraus he reports is
left vague, or by implication confined to lower levels. The procurator, a libertus
Caesaris, is not of low rank: whose favour might he be currying?
rrz.q culpa omnium The recipient because he has no beneficium, the
vilicus because he overlooks impropriety amongst his underlings (:o¸.¸–¡), other
staff members (e.g. libratores, cf. :o¸.¡) because they sought safety – or a share
of profit – in collusion. But F.’s omnium seemingly includes the procurator as
well (and perhaps even the curator), whose responsibility it was to keep watch
for precisely such collusion.
maxime accipientis A recipient who has bypassed the procurator will
have an unmarked calix; yet if calices and/or fistulae were unmarked the culprits
remained safely anonymous – unless a scrupulosa inquisitio were to be under-
taken: cf. De Kleijn (.oo:) :o¸–8.
deinde vilici Aubert (:qq¸) :¸¸–6 proposes that a vilicus such as this
holds that rank simultaneously (:) in the familia aquarum (::¸.:) and (.) in a
plumbers’ workshop owned by someone of the senatorial elite. The idea is
highly implausible: De Kleijn (.oo:) :oq, :.o–¸¸.
rrz.j legitimae mensurae . . . legitimum spatium The mensurae
are defined pipe-sizes (§¸n.); the spatium is ¸o feet, specified in the S.C. (:o¸.¸,
:o6..). For legitimus see Berger ¸¡¸.
subiectae fuerunt For the syntagm see q¡.¸n. cautum fuit.
acciderat The pluperfect tense suggests that F. has seen to having all
such practices discontinued.
per brevis angustias expressa Attaching a larger pipe (a diffuser)
within a short distance of the castellum will produce the so-called Venturi effect,
increasing the quantity because the velocity is greater than in the case of a
delivery pipe which matches the calix in size. See Hodge (:qq:) .q6 and ¡q6
n.¸:, with reference to Fox–McDonald (:q8¸) ¸8:–¡ (the latter use F.’s very
circumstances to set forth a problem and demonstrate its solution).
rrz.6 diligentiae adiciendum est With this construction F. resumes
his prescriptive voice (:o¸.:n.); cf. ::¡.. emendandum, ::¸.: tollendus, but note as
well §¸ carebit, ::¸.: observari oportet. The bureaucratic tone is emphasised by
the word diligentia (:n.).
COMMENTARY ::..¸–::¡.:
fistulae quoque . . . signentur Marked, presumably, in some way so
as to be recognisably in conformity with the size of the calix. Surviving pipes
do not as a rule bear an indication of their gauge, although some (with very
few exceptions only those bearing the name of the emperor) are stamped with
the name of a procurator or some other official: Bruun (:qq:) .:o–:.. On
the showing of surviving fistulae Bruun (:qq:) ¸q–¡¡ inclines to view F.’s whole
discussion of markings as more prescriptive than descriptive: see also :o¸.¡n.
rrz.¡ ita demum enim See :¡.¸n. ita demum.
rrj.r ad lineam ordinentur Thereby assuring a uniform head, assum-
ing that all discharge pipes are completely submerged. F. grasps the principle
clearly enough, although his explanation is no more than empirical. Apipe set
lower would have a greater, one set higher a smaller head; if both were the
same size and positioned at the same level, they would discharge the same
quantity. Grimal’s explanation of this passage (ed. q¸–6 n.:¸¸) is inept, for
which Hodge (:qq.) ¡6: n.¸6 rightly chastises him. Panciera (:q¸8) implausi-
bly takes ad lineam to mean ‘horizontal’ ( = ad libram of ¸¸), with inferior and
superior tilted downwards and upwards respectively.
alterius . . . alterius Sc. privati (so Poleni); cf. in quorundam fistulis just
below §¸.
rrj.z cursus aquae . . . rapitur Cursus at ¸6..n. seems perhaps to be
used of water in a free flow channel, whereas here the focus is on conditions at
a castellum (::...). F. means that water flowis more rapid, and therefore greater
in quantity, through the pipe which is lower.
rrj.q fistulae solutae Cf. ¸6.¸: a calix, because it is bronze, non temere
potest laxari vel coartari. The aquarii cannot readily have adjusted the size of
actual pipes, but they could perhaps have regulated the size of the aperture by
means of a valve. For the term solutus cf. Sic. Flacc. p.:o¡.:q Campbell hi agri
a quibusdam soluti appellantur: soluti autem non sunt quorum fines deprehendi possunt et
rrq.r translata in novum possessorem aqua The combination of
translata and possessor assures that F. is speaking of one and the same property, a
change which nonetheless carried the explicit requirement of an entirely new
beneficium (:o¸.:–¸, :oq.:–¸). Unless the new grant is of a different size, such a
transfer should necessitate virtually no alterations: both calix and fistulae could
remain in place. Such transfers must have been very common, but under these
circumstances any markings other than those indicating calibre would have
been continuously obsolete: see :o¸.¸n., Bruun (:qq:) ¸q–6o. New grants to
new properties are a different matter, but in such cases a wholly new conduit
would have to be laid out.
COMMENTARY ::¡..–::¸.¡
foramen novum . . . imponunt, vetus relinquunt In the process
of implementing any new grant, staff members may have felt authorised to
introduce a new foramen. There may have been a ‘loophole’ in the formula
of the epistola principis (:o¸.:n.); it is perhaps unlikely that anyone specified –
as a precaution against impropriety on the part of the staff – that in case
of renewals any existing connexions as well as pipes might continue in
quo venalem extrahunt On aquarii and their money q.6n., :oq..n. As
to how and to whom the aquarii sold water our ignorance is complete: Bruun
(:qq:) :o8–q.
rrq.z in primis . . . emendandum curatori For the bureaucratic
language (‘officialese’) cf. ::..6 diligentiae adiciendum est, also Dig. xxx\.:o.¸.¸ hoc
autem diligentissime praetori examinandum est.
rrj.r tollendus est reditus quem vocant puncta One must remove
‘that source of income whichderives fromwhat they call “punctures”’. Baldwin
(:qq¡) ¸o¸ calls this ‘gangster slang’. The practice, F. suggests, is widespread
and well organised. Its origin might have been related to distribution of aqua
caduca (::o..n.), and deliveries to commercial establishments may have met a
need not otherwise readily addressed within the system. (There is no apparent
relationship to the term puncta, a pipe, at .¸..).
rrj.z latentes sub silice As a precaution against damage, but therefore
subject to greater abuse because taps would not be visible.
rrj.j a punctis ‘The borer, puncturer’, a staff specialist. The title was
presumably an informal one (created on analogy with ab epistulis, a libellis, etc.).
In any case it has been ‘retired’ (note the imperfect).
negotiationibus Not abstract ‘business activities’, but apparently a con-
crete use meaning either negotiatores or, perhaps better because of in transitu,
‘places of business’. Parallels to the latter use elude me, although business and
its location are often linked, e.g. Dig. \ii.:..¸.: si dominus solitus fuit tabernis ad
merces suas uti vel ad negotationem, xi\.¸.::.¸ ante tabernam scilicet vel ante eum locum
in quo negotiatio exercetur.
convulneratas Cf. .¸.¸n. ne rivus saepius convulneretur, Str. ii.¸.¸: (Hispani)
convulnerant confunduntque nihil tale exspectantes.
exiguus modus Noticeable no doubt in a sluggish flow at points of
discharge and in a volume of overflow at public basins smaller than expected
(cf. 8¸.¸, :::..).
rrj.q quantum ex hoc modo aquae servatum sit First be it noted
that ex hoc modo does not mean hoc modo (‘in this manner’ Loeb, ‘de la sorte’
Bud´ e), nor is it the equivalent of fraude huiusmodi. The prepositional phrase
COMMENTARY ::6.:–::6.¸
means rather ‘from this supply’, i.e. from the modus (¸..n. pro suo modo) just
mentioned, the one that had been reduced to exiguity. So much is clear
from word-order. Now, what has happened to this water? Has it been ‘saved’
(C reads servatum), or somehow ‘stolen’ (as most nowadays would have it)? Of
course, it had been stolen; what concerns F. here is what has now happened to it
(perfect tense). There is a straightforward answer. It has been reclaimed to the
public good: OLD s.v. servo :o, ‘to recover, regain (money lent); to make good,
recoup (losses, expenses)’; cf. Berger ¸o.. The saving is not so much a matter
of conservation as preservation. Recovery, as always, is more important to F.
than loss. Had F. been assessing a loss, he would have used the imperfect (note
efficiebatur, praebuisse, etc.): the perfect servatum sit exactly matches redactum est.
(For those who will disagree, I note that F.’s regular word would perhaps have
been not serivare but derivare: q.8, :¡.¸, 66.., ¸¸.., q¸.:.)
aliquantum plumbi ‘A considerable amount of lead’ may be slightly
disappointing after quantum aquae, which suggests that a precise measurement
could be given. But for F. to be more exact in his estimate he would have
needed to determine the capacity of each and every one of the illicit pipes;
he is satisfied to say that a noticeable amount of water must have been saved
because a sizeable amount of lead has been retrieved.
rr6.r familia F. is about to explain that there are in fact two familiae, his-
torically distinct and separately maintained. The singular, however, might have
a specific point. In practice the two gangs may have been so closely associated
that their supervision could readily have been combined in the interests of
efficiency. The familia publica was under at least the nominal control of the sen-
atorial curator (:o:..n.), while an imperial official presumably was responsible
for the familia Caesaris (:o¸..n.). The latter conceivably had superintended the
work of both crews. The present passage suggests that F. himself had assumed
responsibility for the combined workforce: note especially the singular used
in §¡ quid esset actura (sc. familia) dictaremus. This implies an administrative re-
form, instituted no doubt under Nerva and entirely consistent with F.’s attitude
towards senatorial responsibility.
rr6.z familae sunt duae C’s blank space may be insignificant (Introd.
¸q), but some short phrase (e.g. ‘to speak more precisely’) would signal a paren-
thesis and lessen the abruptness of the plural; cf. Dilke (:q¸¸) .¡8. Grimal’s
<aquariorum>, however, is otiose.
rr6.j publica . . . publicatam See above q8.¸–qq.:. In all likelihood
the familia publica continued to be employed only on the pre-Claudian aque-
ducts. Servi publici are attested for the Anio Vetus (CIL 6..¸¡¸–.¸¡¸). Lack of
epigraphic evidence has been taken to indicate that this familia disappeared
entirely sometime in the second century cr: Eder (:q8o) :6¡.
COMMENTARY ::6.¸–::¸.:
circiter F. seems untroubled by an approximate figure. The number
might have fluctuated, perhaps inpart because of procedures involvedinthe ac-
quisition of servi publici and deputing themto the familia aquarum(more complex
than those for the familia Caesaris). Possibly .¡o was the number at Agrippa’s
death. Since F. makes no explicit reference here to inadequacies of the cura-
torial administration (cf. :o:..), it is likely that curators were only marginally
responsible for ensuring the complement of the familia publica.
rr6.q Claudius . . . constituit C. ¸. cr (:¸..n.). The familia Caesaris
seems to have been established to maintain the new aqueducts; as such it
might have been intended to parallel – not to supplant – the existing familia
publica. Claudius’ action reveals a sense of continuing responsibility for projects
undertaken on imperial initiative. The model, like that for his creation of
the procurator (:o¸..n.), was the Agrippan scheme perpetuated by Augustus
(q8–q). Not to press the tense of perduceret, it is possible that Claudius had a
crewof imperial slaves working alongside redemptores in construction: cf. Martin
(:q8q) 6:. Keeping them on as a second familia would make sense, for these
workers would have acquired useful experience with aqueducts.
rr¡.r aliquot ministeriorum species Walser (:q8q) :¸¸–6 thinks that
F.’s listing might be hierarchical (vilici at the top, alii opifices at the bottom); he
cites Veg. iii.8.¸ idoneos tamen tribuni et probatissimos eligunt qui circumeant vigilias
et renuntient si qua emerserit culpa, quos circumitores appellabant; nunc militiae factus est
gradus et circitores vocantur. Evidence, however, is entirely lacking: Aubert (:qq¡)
¸.¸–¡. Persons with specialised duties might have been trained in service
or, less likely, recruited for their technical expertise. F. has little interest in
these lowly workers, and one can only surmise what their actual duties and
responsibilities might have been. Bruun (:qq:) :q: collects the epigraphical
evidence for workers of the sort that F. mentions here. Vilici and castellarii seem
to have been associated with a particular aqueduct. Urban workers may have
been assigned to a specific regio, e.g. Barnaeus de familia publica VIII of CIL 6..¸¡.
(althoughit is far fromclear that the familia publica of this inscriptionrefers to the
water staff). But such epigraphical material is not susceptible of close dating,
and relating it to F.’s text should be done with caution: see Bruun (:qq:) :qo–..
vilicos The vilici bear some responsibility for the positioning of delivery-
pipes (:o¸.¸; ::..¡, ¸), presumably because they were foremen of the working
crews. Avilicus aquae Claudiae (CIL 6.8¡q¸ = ILS :6:.) and a vilicus aquae Marciae
(CIL6.8¡q6 = ¸¸¸.q) are knownfrominscriptions; cf. CIL6.¸¸¸¸. vilicus aquae
and CIL 6.¸¸¸¸¸ ( = ILS :6::) servo vilico Cae. aquario.
castellarios Presumably in charge of major castella (¸..n.), especially
those with a standing crew (::¸.¸). An imperial slave is called castellarius aquae
Claudiae in CIL 6.8¡q¡ ( = ILS :6:¸); public slaves are attested for the Anio
Vetus (e.g. CIL 6..¸¡¡ = ILS :q¸¡).
COMMENTARY ::¸..–::¸.¸
circitores Watchmen or inspectors, no doubt primarily employed in out-
lying areas (§¸). No fewer than twenty-one circitores are mentioned in a late
inscription from Tibur (CIL :¡.¸¡6q), which by general agreement pertains
somehow to the urban aqueducts.
silicarios The word is rare, but attested epigraphically: Baldwin (:qq¡)
¸o¸. Perhaps paving crews were needed for frequent and extensive work in
the City where pipes ran sub silice (::¸..); cf. Lanciani (:88:) ¸¡:. Silex was, of
course, the usual term for pavement, but the word applied to any hard stone
used in building. Silicarii is thus better taken as ‘masons’ or ‘stone-workers’. It
should embrace those who worked with concrete (cf. Vitr. \iii.6.:¡ caementum
de silice frangatur) as well as a variety of other building materials. Their tasks
would have involved structural repairs (cf. :.:.:, :...:–.).
tectores ‘Plasterers’ would have applied or repaired the layer of wa-
terproof stucco (opus signinum) that lined the channels; cf. :...: tectoria
aliosque opifices Among them libratores (:o¸.¡) and architecti (::q.¸).
rr¡.z ex his Despite its long tradition (cf. q6), the distinction extra urbem /
in urbe was essentially a matter of practical administration: cf. :o¸.¡.
ad ea quae non sunt magnae molitionis Routine repairs (cf. ::q..
quibus ante succurri debet quam magno auxilio egere incipiant) and regular maintenance
(outlined in :.o–¸) – in short, those tasks which could reasonably be accom-
plished per domesticos artifices (::q.¸). F. uses molitio in the sense ‘effort, work’ (OLD
s.v. ¡): cf. Str. i.:.8 nec ignoraret [Germanos] maiore bellum molitione inituros, si adventum
tanti ducis praesensissent; Sen. Controv. ii.¸.:q potuit enim uxor etiam non indicante marito
tam magni consili molitionem deprehendere; Sen. QNat. \i...¸ iam intellegetis nugatoria
esse nos et imbecilla corpuscula, fluida, non magna molitione perdenda; \ii.:6.: nec magna
molitione detrahenda est auctoritas Ephoro: historicus est.
rr¡.j castellorumet munerumstationes Context suggests that both
munera and castella (¸..nn.) could serve a similar function in the distributory
scheme (cf. 8¸.¸–¸). Stationes may imply a permanent workplace, perhaps in
combination with living quarters. The sense of ‘headquarters’ sometimes ap-
plied is here not appropriate (cf. ::q.¸n. stationis).
opera quaeque urgebunt Opera here are not structures but ‘tasks of
ordinary maintenance’, i.e. jobs needing to be done (cf. ::q.. opera subinde
nascuntur). The future tense is prescriptive (:o¸.:n., ::..6n.). The expression
opus urgeo seems otherwise confined to poetry: Tib. i.q.8 et durum terrae rusticus
urget opus, Ovid, Ars Am. ii.¸¸o timor urget opus, Fast. i\.8¸¸ hoc Celer urget opus
(contrast opus nominative at Am. iii.:.¸o a tergo grandius urget opus, Fast. i\.q¡8 me
grandius urget opus).
ad subitos casus One thinks naturally of fires (cf. Suet. Claud. .¸.. ad
arcendos incendiorum casus), a constant threat against which it was an important
precaution to ensure an adequate supply of water (Tac. Ann. x\.¡¸.¡). But on
the whole it is unlikely that delivery could have been modified in any very
effective way in response to an emergency, for the volume and availability of
additional water would always have been subject to the overall limitations of
the hydraulic system. F. more probably refers to unforeseen breakdowns (the
sense of casus at 8¸.¸). A shortage in one area could be alleviated by diverting
surplus from another, exactly as sometimes happened in the case of scheduled
repairs (::.., 8¸.¸–¸).
abundantium Bruun (:qq:) ::¸ suggests that these waters are in storage
tanks (perhaps F.’s castra); see ¸8.¸n. castris.
rr¡.q tam amplum . . . conprehenderetur In a single sentence F.
diagnoses an administrative malady, explains its cause, and reveals his remedy.
tamamplumnumerum Aworkforce of some ¸oo slaves requires more
thanperfunctory supervision. F. is not directly concerned withthe economic ef-
ficiency of maintaining these crews. Bruun(:qq:) :qo thinks there may be some,
perhaps impractical, significance to the fact that ¸oo workers /:¡ regiones =
¸o men per ward; better the idea of a less rigid system: Hirschfeld (:qo¸) .¸¸
n.¡, Ashby (:q¸¸) .¡.
praepositorum The plural may follow from utriusque familiae (a general
superintendent for each gang). Or it may be deliberately vague. The vices of
ambitio and neglegentia suit well enough F.’s view of the procurator and the vilici
(:o¸.¸, ::..¸–¡), but the latter charge could be levelled even at the curatores
(:o:.., :¸o.:).
in privata opera If there remained sufficient staff for maintenance, both
slave gangs might fairly have been called upon in times of public emergency,
and their special skills might have been welcomed in certain kinds of projects.
But it was quite another matter for the praepositi to reap personal profit from
hiring them out as a labour force for private contractors. It was perhaps in-
evitable that the situation would recur. Strikingly similar terms are used by
Cassiodorus Var. iii.¸:: Mancipia vero formarum servitio principum provisione deputata
in privatorum cognovimus transisse dominium. . . . et quoniam malarum rerum emendatio
nos delectat . . . Iohannem virum spectabilem electum nostra iustitia ad haec quae supra
memoravimus direximus inquirenda.
revocare . . . <statuimus> It might just be possible to construe the
infinitive revocare with the verb instituimus later in the sentence (cf. TLL ¸.::
:qq¸...): ‘I resolved to bring back to some discipline and to the service of
the state’ (Bennett). But ministeria here are persons (as in §:, cf. :o:..), and
the word is direct object of instituimus which means ‘organised’ in the latter
part of the sentence. It is hardly possible to take the main verb in two quite
different senses. Either we must emend revocare to revocavimus (so K¨ uhne’s ‘habe
ich wieder . . . gebracht’) or supply an appropriate verb to go with the infinitive.
My statuimus has no elegance, but I was loath to lose the emphatic juxtaposition
diduci revocare.
publica ministeria Contrasts with privata opera, but not applied solely
to the familia publica.
instituimus . . . dictaremus The plural is ‘official’ (q.¸n., .¸..n.).
The singular subject of esset actura and egisset can only be (utraque) familia, a clear
implication that F. himself is in charge of the entire work force (::6.:n.). The
bureaucratic innovation is one of common sense (cf. :...¸n.), comparable to
the improvements envisioned for record-keeping (¸¸.:–., 88.¡), and perhaps
as well to the preparation of maps for use in a central office (:¸.¸–¡).
rr8.r commoda Distinction between the two familiae is inherent in the
ownership of slaves and in the separate history of each gang (::6.¸–¡). Their
remuneration (commoda) would embrace the basic expenses for food, clothing,
and housing to which slaves were entitled, although it is not clear whether these
obligations were met wholly or partly in kind; cf. Boulvert (:q¸¡a) ::¡–:6. The
term may well include some direct payments in cash (cf. Sen. Ep. 8o.¸ servus
est: quinque modios accipit et quinque denarios). In light of the special nature and
status of both public and imperial slaves, it is reasonable to suppose that they
received a salary of sorts, but direct evidence of the practice is nowhere to be
found (§¸n.; cf. :oo.¸n. mercedem cibaria).
quod impendium exoneratur F. raises financial matters only paren-
thetically and from an administrator’s viewpoint; his main focus seems to be
on preserving the integrity of the curator’s role. On the other hand, because of
F.’s membership on a senatorial commission in precisely this same era (Pliny,
Pan. 6...) we know that he was neither ignorant of nor unconcerned with
economic and fiscal realities. Perpetual revenues for the water-department de-
vised to offset routine expenses (chief of which was upkeep of the familia) dated
no doubt from the establishment of the cura aquarum: see qq.:n. For a different
and broader application, but something quite similar and exactly contempo-
rary, note Pliny, Ep. \ii.:8 (cf. x.8), a letter with interesting information about
the Nerva-Trajan alimentary benefactions, on which see, e.g., Duncan-Jones
(:q6¡), Woolf (:qqo).
vectigalium reditu ad ius aquarum pertinentium ‘By income
from fees accruing to the state which belong to (are attached to?) the right
of the public water supply’ (for this plural aquae see :n. aquarum iniunctum
officium). There is a shade of redundancy, for vectigal alone can mean ‘income’
(OLDs.v.), but F. distinguishes revenue actually collected (reditus) fromrates due
(vectigalia). These vectigalia are not necessarily payments for water-rights (see
:o¸.:n. impetrare), and they have no relation, at least none explicit, to the vecti-
galis aqua furnished in Republican times to certain commercial establishments
There exist certain loca or aedificia, by definition public because subject to
vectigal, which, as F. explains (§.), are adjacent to ductus, castella, munera and lacus.
It is revenue from these properties that ought not, we are told, to be lost to the
aerarium – and as a corollary that ought to be credited as it were to a ‘water-
gang account’. Here vectigalia ‘pertain to’ ius aquarum, whereas below (§¸) it is
loca that ‘pertain to’ a vectigal. Why are these loca subject to a vectigal (singular)?
And what does F. mean by vectigalia ad ius aquarum pertinentia? To answer the first
question we need to address the second as well. An administrative guideline
comes to the aid: Dig. r.:o.¸.: (Ulpianus) curabit igitur praeses provinciae, si qui publici
sunt (sc. fines) a privatis separare et publicos potius reditus augere. si qua loca publica vel
aedificia in usus privatorum intervenerit, aestimare utrumne vindicanda in publicum sint an
vectigal eis satius sit imponi, et id quod utilius rei publicae intellexerit sequi. By analogy to
such a context we realise that F.’s loca aedificia are public property in immediate
proximity to ductus etc. It seems to me entirely straightforward to connect them
directly with the locus circa rivos, specus . . . castella, lacus . . . terminatus of the Lex
Quinctia (:.q.¸, q). Within this reserved zone some activities and buildings are
indeed permitted (:.q.q–:o; cf. :.¸.:), a circumstance that would qualify these
loca as subject to a vectigal. But is this to a specific vectigal? And how, then, does
the vectigal pertain to ius aquarum? The possessores of these public spaces owe the
state a regular fee merely because the land is not their own; they hold it by
lease: cf. Campbell (.ooo) ¸6o–:. Such persons might just possibly have a right
to water, a ius ducendae aquae (perhaps of long standing, e.g. for caduca q¡.¡),
although F. need not be talking about any such right here: rather he goes out
of his way to make clear (:o¸.¸, ::o.:) that this water, like all to privati, is not
a ius at all, but a beneficium principis. Any sense of ius aquarum must account for
F.’s use of the plural aquarum (not sing. aquae) in the present passage and must
square with his implication that this revenue should be earmarked for support
of the familia publica (sc. aquarum; cf. q8.¸ Agrippa’s familia propria aquarum). I
confess that I know of no parallel for my interpretation that ius aquarum is a
compressed expression which conveys the notion that money in the treasury
could be called iuris aquarum (‘in the water fund’), especially since it derived
as income from loca publici iuris. (F. might have tangled us up more than he
intended by an unhappy application of pertinere ad, one of his favourite and
rather careless expressions.)
rr8.z ea constant ex locis aedificiisve ‘These (vectigalia) are due from
places and buildings . . .’ For this use of constare as a technical term (understand
populo Romano) ‘stand to (the name of) the people’ see Crawford (:qq6) ¸8:
(in §¸ just below constaret bears its usual meaning). The loca and aedificia would
have been called vectigalia loca (OLD s.v. :).
rr8.j quem reditum Bruun (:qq:) .oq, (.ooo) ¸8q n.6. points out that
HS .¸o,ooo is a ridiculously small sum of money; cf. De Kleijn (.oo:) 8o n.:6.
F. addresses a general principle, and his text gives no warrant for assuming that
the expenses of the familia publica either equalled or were intended to be covered
by the total available revenue. The entire income might not necessarily have
been applied to the costs of the familia; some, for example, may have covered
expenses of the curator’s office and staff (:oo.¸–¡). It is equally facile to reckon
the annual cost per slave at approximately :,ooo sesterces (revenue of .¸o,ooo
divided by .¡o slaves), although this figure seems to be within reasonable
limits (as can be seen by a comparison to the legionary’s ¸oo denarii): Eder
(:q8o) :o8–:o.
alienatumac vagum If Poleni’s conjecture is accepted, alienatum seems
to mean that the revenue is lost fromcontrol (i.e., no longer reserved for water-
related expenses) and vagum that it no longer is credited to a special account.
I have found no parallel for F.’s collocation of words, nor for either adjective
applied to money. Cf., nonetheless, Suet. Vesp. :.: incertum diu et quasi vagum
imperium suscepit.
in Domitiani loculos conversum ‘Domitian’s purse’ should, of
course, be the fisc (§¡n.), whereas populo restituit means to the aerarium. The
former manoeuvre, shocking to F., was perhaps part of a more general admin-
istrative reorganisation under Domitian: Dilke (:q8¸a) :¸., Griffin (.ooo) ¸¡;
cf. Southern (:qq¸) 6¸–¡, a misinterpretation of this passage. F.’s point may
be subtle and double: Domitian’s were the loculi of a bad prince (cf. our pe-
jorative ‘purse-strings’); his control had therefore been truly personal and not
‘imperial’, i.e. exercised openly and for the good of the state: contrast iustitia
divi Nervae. For a somewhat similar use of loculus, Suet. Galba :..¸ peculiaribus lo-
culis suis (cf. Dig. xxxiii.8..¸.: ratio loculorum). For convertere of transferring funds,
note Tac. Hist. i.8q.: conversa in militum usum pecunia, qo.: reliquias Neronianarum
sectionum nondum in fiscum conversas; cf. Dig. xx.6.q.¸ stipulati sunt ne ea summa in alios
usus converteretur, r.8...¡ sin autem frumentaria pecunia in alios usus quam quibus desti-
nata est conversa fuerit, veluti in opus balneorumpublicorum(also x\.:.¸.6, xxi\.:.¸¸.pr.,
iustitia divi Nervae Note that the iustitia of the late Nerva finds its
complement, or better its perpetuation, in nostra sedulitas (the same phrase in
Str. nostra sedulitas impendet operam; cf. : fides sedula). On the abstraction
cf. H–Sz ¸¡6–¸. For iustitia as an ‘imperial virtue’ see Lichocka (:q¸¡).
ad certamregulam Cf. .¸..n. proposita regula. F.’s initiative was to verify
the public interests (as they relate to the water-system); such action was akin to
vindicatio, which was within the curator’s competence. Application of his ‘rule’
would have fallen to the prefects of the aerarium.
rr8.q ex fisco The imperial treasury, distinguished from the emperor’s
patrimonium. By analogy with the fiscus frumentarius, a specialised account for
which the first clear evidence is Flavian (e.g. CIL 6.¸¡¡, 6¸¡ = ILS :¸¡o, :¸¡oa;
COMMENTARY ::q.:–::q..
6.8¡¸¡–¸ = ILS :¸¡:–¡), there might have existed a similar, but no doubt less
complex, accounting system for expenditures pertaining to the water-system.
Jones (:q¸o) .¸ points out that the separate payments go back to the time
of Claudius, whose slaves would have been paid from the emperor’s private
fiscus (in the Julio-Claudian sense). He is certainly right to assume that F. is
distinguishing only between whether funds are disbursed by the prefects of the
aerarium or the a rationibus. Indeed, sharp distinction between fiscus and aerarium
is first clearly attested under Trajan: Tac. Ann. ii.¡¸.., \i...:, \i.:¸... Evenmore
important is the point made by Bruun (:qq:) .oq that this passage should not
be taken as evidence of ‘a takeover by the Emperor resulting in a loss by a
presumed “senatorial” administration of Rome, but should be seen rather as
one step in the slow process which formed the Roman empire and gradually
strengthened the Emperor’s influence in almost every sector of public life’.
unde . . . erogantur Materials for upkeep and repairs were the tradi-
tional responsibility of the princeps (:.¸n.).
omne plumbum Pipes supplying the public uses and those nomine
Caesaris. It is not clear whether pipes delivering water to privati were tech-
nically part of the public system (as in a sense were castella privata: :o6.:n.). On
use of lead in general and on the possibility of imperial interests in the lead
industry, see Bruun (:qq:) :q8, ¸¸:–6¸.
rrq.r ad tutelam ductuum sicut promiseram At ::6.:. For the plu-
perfect cf. : sicut . . . institueram.
rem. . . indicium Cf. :¸.. cummaxima huius officii pars in tutela eorumsit, scire
praepositum oportet quae maiora impendia exigant. The parenthesis here is primarily
for literary embellishment, and vel praecipuum is a pardonable exaggeration.
Krohn (pref. p.v) was needlessly troubled by F.’s assertion that tutela – rather
than the ductus themselves – is the best testimony of Roman greatness (cf. Vitr. ut maiestas imperii publicorum aedificiorum egregias haberet auctoritates). Reliable
functioning of the vast and impressive system was what rightfully attracted
respect (:6n.), and this could be maintained only by unrelenting attention to
overall upkeep. Cf. Dig. xriii..:.¡.pr. (Venuleius) non tam necessariam refectionem
itinerum quam rivorum esse, quando non refectis rivis omnis usus aquae auferretur et homines
siti necarentur; et sane aqua pervenire nisi refecto rivo non potest, at non refecto itinere difficultas
tantum eundi agendique fieret, quae temporibus aestivis levior esset. The overlapping
moral overtones of tutela should not be overlooked, for a strong personal sense
of pride and responsibility underlies this comment.
enixiore cura Cf. Pliny, HN ix.¸. enixioris operae; Sen. Ben. \i.:¸.: opera
enixior, Str. ii.¸.¸o promptam et enixam operam exhiberet.
rrq.z opera subinde nascuntur Opera ‘jobs’ (::¸.¸n.); cf. :.o.: nascuntur
opera. These are numerous (multa) and sometimes extensive (ampla), but routine
in nature and expected as a matter of course (subinde). Proper attention at
this stage is essential to avoid more severe damage and a greater likelihood of
breakdown. B¨ ucheler takes the word here to mean ‘structures’ (cf. :6), which
for him is an easier antecedent for quibus succurri, hence his dilabuntur (and
labascunt, which he credits to Kiessling).
prudenti temperamento ‘sensible moderation’ or ‘careful restraint’.
Cf. Colum. xii.:6.: quadam moderatione temperamentoque opus est; Pliny, HN xii.::¸
incidentis manus libratur artifici temperamento, ne quid ultra corticem violet; Pliny, Pan.
:o.¸ at quo, di boni, temperamento potestatem tuam fortunamque moderatus es; Tac. Ann.
iii.:..: die senatus Caesar orationem habuit meditato temperamento.
sustinenda Sustinere here = ‘put off, delay’ (OLD s.v. 8 b), not ‘kept up’
as interpreted by Dilke (:q8¸b) .¸¸–6. Krohn’s festinanda is totally unwarranted
(despite a superficial similarity to :.:..), for the point here is judgement rather
than haste. F. might also have had in mind the seasonal considerations to which
he points in :....–:.¸...
non semper . . . credendum est Cf. Vitr. \¸ ceteri architecti rogant
et ambiunt ut architectent; mihi autem a praeceptoribus est traditum: rogatum, non rogantem
oportere suscipere curam, quod ingenuus color movetur pudore petendo rem suspiciosam. nam
beneficium dantes, non accipientes ambiuntur.
rrq.j non solum . . . instructus F.’s ideal administrator (cf. esp. ..:)
will have personal experience or understanding of technical matters for which
he is ultimately responsible, enough at least for an intelligent assessment of
the problems to be addressed (cf. :¸..–¡) and to enable him to control the
advice or performance of his subordinates. The range implicit in proprio usu
is breadth rather than depth: in practical terms the kind of information F.
himself outlines below (:.o–:.¡.¸). These chapters may strike an educated
modern reader as rather elementary (a few basic observations presented with
wholesome common sense), but neither the knowledge nor the attitude ought
too readily to be assumed for the senatorial aristocracy of F.’s day. The whole
is a good deal more practical than what we read in Vitruvius’ preface.
suae tantum stationis architectis Cf. Pliny, Ep. x.¸q.¡–¸ cogor petere
a te non solum ob theatrum verum etiam ob haec balinea mittas architectum dispecturum –
precisely because he has just reportedadvice fromanarchitectus aemulus. Trajan’s
response (x.¡o.¸) architecti tibi deesse non possunt misses the point, as Sherwin-
White observes. Presumably ‘engineers in his own bureau’ are those who
formed part of the curator’s permanent staff (:oo.:). The exhortation to seek
opinions independent of ‘in house’ technicians may be no more than a point
of general wisdom: it cannot hurt to get a different viewpoint. But context sug-
gests that F. may have had grounds for distrust. If these architecti were standing
bureaucrats without personal attachment to an individual curator (:o:..n.),
they might have fallen prey to fraudulent interests (on the part of the work-
ing crew, for example, or from outside contractors). Nor is it inconceivable
that staff architecti were themselves independent redemptores who could profit
from administrative decisions made in their own favour.
stationis Superintendency of the workers and other managerial aspects
of Agrippa’s cura presuppose the establishment of a permanent ‘bureau’, but it
is not clear that F.’s statio refers to a physical location or ‘headquarters’. Bruun
(:q8q) argues that it has the looser sense of ‘administration’; cf. Bruun (:qq:)
:q¸–6. In later years we hear of a statio aquarum connected with the temple of
Juturna: Coarelli (:q8: ) ¡¸, Burgess (:qqq). Coarelli (:qq¸) .¡¸–¸o maintains
that this same temple served as headquarters from the very beginning of the
cura aquarum in :: ncr.
fidem. . . subtilitatem Reliability (fides) is equally if not more important
than expertise (subtilitas). Qualities sought in outside advisers are of course
among the reasons why the curator should avail himself of the resources they
offer. F. underscores the need for responsible standards.
ut aestimet The informed decision will be that of the curator alone: im-
mediate action or cautious postponement, outside contractors or staff workers.
quae repraesentanda, quae differenda sint Examples of reprae-
sentare ‘do at once’ are cited in OLD s.v. :a. Note especially the contrast with
differre: Sen. Ep. q¸.:, to which add Dig. xxx\ii.:o.¸.¡ and esp. :o.¸.¸ an ex-
pediat pupillo repraesentari cognitionem an potius differri in tempus pubertatis (cf. Dig.
xr.¸.¸¸.pr., ¡:.: mora fieri vs repraesentari).
redemptores . . . domesticos artifices The latter are presumably
the work force of the two familiae (::6..–::¸.:). Contractors (redemptores) would
be called upon for projects too specialised or too large for the domestic staff. An
excellent example is that of L. Paquedius Festus, who rebuilt Claudia’s channel
sub monte Aeflano in 88 cr (CIL :¡.¸¸¸o = ILS ¸¸:., cited in Appendix B,
no.::). He is called redemptor operum Caesar(is) et puplicorum, but this credential
(akin to ‘by appointment to . . .’) sheds less light than one could wish on
the details of administrative procedure. What official(s) selected a redemptor for
a particular project? Who let the contract and performed the probatio? F.’s
context is no more helpful, for he stops short of revealing the extent of the
curator’s role. Responsibility for repairs to all aqueducts seems, however, to
have been assumed by the emperor (:.¸n.), and expenses of this sort were
paid from the fisc (::8.¡). It is therefore probable that contracts were handled
solely by persons in the imperial service. Adiligent curator might have defined
the projects for which redemptores were needed (and could have furnished at
least general specifications), and inevitably he was familiar with the finished
work (cf. :.o, :.¸.¸). But F.’s silence as to curatorial probatio (cf. q6) and iudicatio
suggests that these functions were exercised by another party. Nominally at
least this could be no other than the emperor; in practice it will have been
an imperial agent. For discussion Eck (:q8.b) 6q, Martin (:q8q), Bruun (:qq:)
COMMENTARY :.o–:.:.:
rzo ex his causis Instead of <h>is (for the transmitted ι) one might con-
sider quattuor (i\), or quattuor ex causis: cf. q. pluribus ex causis, :...: duplici ex causa
(TLL ¸: 6¸o.¸q).
vetustate corrumpitur For the verb corrumpere cf. ¸.:n., :...:, :.¸.:
(TLL ¡: :o¸...:); in combination with vetustas (¸.:n., TLL ¡: :o¸...:) cf. Dig.
\ii.:.¸ villam vetustate corruptam (cf. xxx...¸¸.¸), xr\iii.¡.¸.: statuas Caesaris
vetustate corruptas reficit; see also Thomas–Witschel (:qq.) :¡o–q.
quid aut Krohn’s transposition is no doubt the easiest solution to the
obvious problem of arrangement here, but it is not wholly satisfying. Repairs
are required in case of structural damage, for which F. lists four causes. Since
he goes on to speak of vetustas and vis (:.:.:), one would have expected a se-
quential listing of these two closely related causes, both of which are natural in
origin. Impotentia possessorum (:.6.:) and shoddy workmanship (cf. :.¸.¸), on the
other hand, are instances where human agency aggravates and accelerates the
processes of nature. The transmitted order may, of course, be what F. intended:
vetustas and impotentia possessorum require an administrator’s more constant vigi-
lance, while weather-related phenomena and unpredictable structural failures
must be addressed only when they occur. In that case, there is merit to the
version proposed by Sauppe (:8¸q) qq¸ aut quid impotentia possessorum aut vetustate
impotentia possessorum Impotentia (OLD s.v. .) is a loose term for a
variety of illegal activities: the same charge recurs at :.6.:.
vi tempestatium ‘Storms’: the collocation is ubiquitous: Cicero, Phil.
\.8, Inv. ii.q8, Nat.D. iii.¸6, 8q, Off. ii.:¸, QFr. i.:.¸; Caes. BGall \.:o..,
BCiv ii.:¡.¡, iii..6.¸; Front. Str. i.:..; Dig. ii.::...¸, \ii.:.¸, xix...:¸..,
xxi\.¸.¸.:., xxxix....¡.q; Livy, xx\.¸.:o, .¸.::, xxx.¸q.:, Per..:; Pomp. Mela
iii.¡¸; Pliny, HN x\i.:¸o, x\ii...6, xxx\.8q; Sic. Flacc. p.::6..¸ Campbell;
Suet. Jul. .¸.., Vesp. ¸.¡; Tac. Hist. ii.8.:; Val. Max. i.:(ext.).:, i\.8(ext.)...
Heavy rainfall poses the threat of erosion where conduits run along hillsides,
and swollen streams would weaken the bridges spanning the valleys. Seismic
damage fromearthtremors (more frequent thanmajor quakes) wouldprobably
fall into the category of vetustas.
saepius accidit in recentibus The almost incredible testimony of
Flavian inscriptions on Porta Maggiore (Appendix B, nos. ¡, ¸) might suggest
that F. has in mind the two post-Augustan aqueducts (chapters :¸–:¸ above):
Thomas–Witschel (:qq.) :¡6–¸, Evans (:qq¡) ::6–:¸. But it is better to take
recentes more loosely, in reference to repairs and reconstructions throughout
the system (:8.¸n.); cf. Van Deman (:q¸¡) ..:. Faulty workmanship is another
manifestation of administrative incompetence (cf. :¸.., ::¸.¡, ::q..–¸).
rzr.r aut vetustate aut vi Giocondo’s <tempestatium> is not needed,
despite or rather because of its appearance in the preceding sentence.
COMMENTARY :.:..–:...:
Cf. Dig. xix...¸o.¡ colonus villam hac lege acceperat, ut incorruptam redderet praeter
vim et vetustatem. W¨ olfflin (:88:) .8o notes vis vetustas as an alliterative
<eae> partes ductuum . . . traiciuntur Cf. :¸.¸ ubi valles quantaeque,
ubi flumina traicerentur, ubi montium lateribus specus adpliciti maiorem . . . exig<er>ent
curam. B¨ ucheler’s eae may not be essential, but it certainly helps the reader who
will encounter et . . . eae quae later in the sentence.
rzr.z explicanda ‘Finish off’ or ‘be done with’ (OLD s.v. q, TLL ¸..:
:¸¸:.¡:); cf. ::..: explicitis, but contrast ::6.: explicanda ‘reveal, explain’ (OLD
s.v. ¸).
rzr.j minus iniuriae Physical damage is foremost inF.’s mind, but iniuria
is used of illegal taps at ¸..: (and perhaps also at ¸..¡).
subterranea . . . exposita Subterranean sections were better insulated
from the effects of thermal stress. In the Mediterranean climate these would
seldom have been extreme, but they were nonetheless potentially significant
(for even small leaks could never be tolerated). The older, stone-built channels
were most liable to suffer damage, for tightness depended on very fine jointing
with only a thin key of adhesive material (mortar) to seal vertical slots: Ashby
(:q¸¸) ¡¡. Innewer work, the tensile properties of concrete couldcompensate to
some extent for seasonal expansion and contraction. The danger was greatest
in the unbroken arcades just outside the City (:.¡.:n.).
nec gelicidiis nec caloribus The two words are contrasted in Colum.
iii.:.¸ (on vines) quarum inter caligines uvae deflorescunt et mox gelicidiis ac pruinis, ut
aliarum caloribus mitescunt. Except for Varro, Rust. i.¸¸.. and Vitr. \ii.:.6 gelicidii
pruinam, the word gelicidium (syn. pruina) occurs only in the plural and its ap-
pearances are limited to Cato (once), Varro (once), Vitruvius (six times) and
Columella (thirteen times).
rzz.r limo concrescente . . . in crustamindurescit The Anio water
(both from springs and the river) was notorious for hardness, and even in
constant flow the mineral content would precipitate to form a calcareous
deposit (limus) within the conduit. Even if this lining may not significantly have
increased resistance to the flow of water, over time the size of the channel was
greatly reduced (as can be observed in numerous remains where the opening
has been nearly or completely choked). Periodic removal of this deposit was
required (ideally on a regular basis). From tunnelled sections this process was
facilitated by the presence of shafts (putei: 8q.¡n.), from which the debris could
be removed at convenient intervals. It was by following piles of such deposits
that Lanciani began to trace the extra-urban courses of the aqueducts: Ashby
(:q¸¸) xi, ¡¡. On the deposit of sinter (calciumcarbonate CaCO
) see Borgioli–
Terzano (:q86), Hodge (:qq.) q8–:o¸, ..¸–¸.. For the same phenomenon in
COMMENTARY :....–:...¸
pipes, which would have prevented lead contamination, see Hodge (:q8: ),
Fahlbusch (:q8., :qq:), Grewe–Blackman (.oo:).
tectoria corrumpuntur Cracks could develop in the hydraulic cement
(:o.¸n.) used to line the channels, and resulting leaks would lead to more ex-
tensive damage (an excellent instance of the need for prompt repair: ::q..).
Latera rivorum et substructiones describe underground channels and those on sub-
structures, the latter here perhaps including opus arcuatum as well (¸.:n.); see
§.n. Cracks were caused by stress of various sorts: erosion and landslides might
have been to blame in hillside sections (cf. :¸.¸, :.:.:), and underground por-
tions suffered damage from tree-roots (:.6.¸). Weakened supports beneath the
conduit could lead to settling.
rzz.z pilae quoque . . . labuntur A parenthetical remark, introduced
perhaps at the mention of major structural damage: F. remembers that water
is heavy. Ashlar piers occur at various places (:¸.:n.), some of themin locations
difficult of access. But tam magno onere suggests that F. may have particularly in
mind the arcades near the City (see :¸.¸, :.¡.:n.), which were subjected to
noticeably greater weight where channels ran superimposed (see 8..n.).
rzz.j <eo>tempore quopraecipue desideratur Cf. qo.. quo tempore
gratior aquarum sinceritas exigitur (‘at which time’). Here we want ‘at a (better the)
time when’. Outside of poetry C’s word-order tempore quo is to be found almost
invariably with a preceding demonstrative (an exception is Livy, ii.¸..q): Str.
ii.¸.:¡ eo tempore quo . . . oppugnabantur, iii...¡ eo tempore quo . . . exploraverunt.
ante praeparatis omnibus The force of the prefix prae- seems to have
weakened, perhaps in part influenced by adverbial constructions indicating the
length of time prior. The combination of praeparo with ante is common in Livy
and occasional in later authors (Livy, x.¡:.q, xx\i..o.:o, xxx..o.¸, xr\.¸..8;
Colum. i\.¸o.:; Curt. \iii.:¡.¸:; Pliny, HN xxiii.¸, xxi\.:86), often with an
additional adverb (iam Livy, i.¡¸.q, i\.¸¸.¡, \ii..¸.6, xxi\...::, xxxiii.¡¸.:o,
xxx\.:¸.:; diu Livy, xr.:¸.:¸; multo Livy, xxxix.¸:.8; tanto Quint. ii.¡..8). For a
specific temporal reference note Celsus, \ii..6..b ante aliquot diebus victu corpus
praeparandum est (cf. Pliny, HN xx\.¸q), Colum. \.q.¸ praeparantur anno ante. In
Cicero the use of ante with compounds of prae- is for the most part limited to
ante praecipio (Inv. i.¸8, Tusc. ii.6¸; cf. Rhet. Her. iii.:6) and ante praedico (Phil. \i.¸,
Inv. ii.:¸., Sen. ¡q, Div. i.:.8), but note Orator :¸¸ ante praemuniat, ad Q. Fr. i.:.:.
ante praescripsi.
quam paucissimis diebus For the ablative indicating extent of time
see ¸¡.¸n. tota aestate exploravi.
neminem fugit The unwanted scenario is not perhaps as unlikely as it
sounds, especially if different crews were assigned jobs without some system-
wide coordination (cf. ::¸.¡).
civitati Still ‘the citizen body’ and not yet ‘city’; here it is virtually = in
usum populi (::.:, :o¡..).
rzj.r maxime structura constant Repairs to the exterior of the chan-
nels as well as to substructures and arches primarily involved the use of con-
crete: for this sense of structura cf. Vitr. ii.8.:; Espinilla Buis´ an (:qq¸) 6¡¡. The
versatility and strength of concrete made it an ideal material for reconstruc-
tions of every sort, including buttressing and reinforcing the square-cut ashlar
of earlier work. On concrete in general, see Malinowski (:q¸q, :q8.), Rakob
(:q8¸), Lamprecht (:q8¸, :qq6), Grewe–Blackman (.oo:).
et suis temporibus et fidelem Concrete work is best done in tem-
perate seasons and its quality is subject to standards which can be monitored.
Both timing and control are the responsibility of an administrator. For fidelis
‘of good quality, durable’ (OLD s.v. ¸c) note, e.g., Pliny, HN xiii.¸¡ materies . . .
fidelis ad vetustatem; cf. fides ‘reliability’ in §¸.
rzj.z idoneum structurae tempus F.’s remark is explanatory, under-
scoring the technical reason for doing work in spring or autumn (by contrast,
convenience and demand are at issue in :...¸). Administrative ignorance on
this point could lead to work of inferior quality.
<et humorem> ex commodo conbibat Sch¨ one (:8¸.) .¡q–¸o ar-
gued that conbibere could be used absolutely, but not one of his parallels holds
up (TLL s.v.). His ex commodo, on the other hand, has better support (cf. Colum.
\i...:¡, xii.:q.¸). Since et / ex are prone to confusion, the phrase (rather than,
say, commode) might explain the omission of et humorem.
non minus . . . sol acrior quam gelatio Five winter months are
excluded (obviously on account of cold), but F. also excludes midsummer (nimiis
caloribus). Thus it makes for good sense (and rhetoric) to say ‘intense heat no
less than cold’. Dederich’s transposition is both inelegant and unnecessary.
sol acrior Sol by itself can mean excessive heat (OLD s.v. ¡b). For acrior as
its attributive cf. Colum. xi...¸¡ priusquam sol acrior exurat terram, Horace, Serm.
i.6.:.¸ ubi me fessum sol acrior ire lavatum / admonuit.
gelatio The word is very rare (OLD, TLL s.vv.). Closest to F. is Pliny, HN
x\ii.¡6 (on preparation of marl, a soil-additive) glaebis excitatur lapidum modo, sole
et gelatione ita solvitur, ut tenuissimas bratteas faciat. Cf. also Colum. iii.::.¸ (even
the most intractable soil) tempestatibus et gelu nec minus aestivis putrescere caloribus ac
praecipit From prae + capit: concrete sets prematurely, or too quickly.
Keuchen (:66:), cited by Poleni: ‘Venuste praecipit, ut apud Poetam si lac prae-
ceperit aestas’ (Virg. Ecl. ¸.q8). Cf. Vitr. \ii.¸.¸ (on plaster) uno tenore perduci, uti
ne praecipiendo non patiatur uno tenore opus inarescere; cf. ii.¸.. ducendi [sc. lateres]
autem sunt per vernum tempus et autumnale, ut uno tenore siccescant. qui enim per solstitium
parantur, ideo vitiosi fiunt quod summum corium sol acriter cum praecoquit efficit ut videatur
COMMENTARY :.¸.¸–:.¡..
aridum, interior autem sit non siccus; et cum postea siccescendo se contrahit, perrumpit ea
quae erant arida.
rzj.j opus . . . aquae obstaturum A general truth (cf. Seneca,
QNat. iii.¸o.6), but specially applicable to the hydraulic concrete (opus signinum:
:o.¸n.) used as a waterproof lining in the masonry channels.
diligentiorem poscit curam Cf. :n. intentiorem exigat curam. F.’s poscere
here is unique, for which he elsewhere employs a favourite verb exigere (6.6, ¸..,
:¸..–¸, :8.6, ¸6.¸, qo.., ::¸..); the variation presumably is sought because he
will use exigenda est later in this sentence.
fides . . . eius Sc. operis. Not ‘honesty’ (Bennett), but ‘reliable quality’
(i.e. durability) of the work itself (OLD s.v. qb); cf. §: fidelem.
secundumlegemnotamomnibus . . . a paucis observatam Note
the indignant commonplace, marked with chiasmus. The lex here could con-
ceivably be statutory regulation of some sort (cf. Tac. Ann. i\.6¸.:, a S.C. in
consequence of the disaster at Fidenae). More likely would be a lex locationis
(q6n.), which was comprised in fact of the dictated terms spoken (dicta) by the
magistrate letting the contract. More likely still is that this lex is a clause in
such a contract, one included so regularly in contracts for all kinds of build-
ing projects that it could be ‘known to all’. For such leges, often pronounced
by censors (hence leges censoriae) and their essentially tralaticious character, see
Magdelain (:q¸8) ¸.–¸6, ¸:–., esp. ¸¸ where he calls attention to a lex at
Puteoli concerning undertakers (AE :q¸:, 88): ‘elle contient un v´ eritable code
de la profession, affich´ e aux yeux de tous’.
rzq.r proximos ductus Context eliminates all but the high-level aque-
ducts: Marcia-Tepula-Julia (:q.¸) and Claudia-Anio Novus (.o.:). Note that
for four of these F. has already drawn specific attention to the arcades propius
urbem (¸.8, q.¸, :¡.¡, :¸.6).
a sexto miliario Editors have accepted septimo, based on the phrase a
septimo miliario at ¸.8, etc.; and the error would be simple for a numeral. But
at issue here is the stretch after the settling-tanks, and these were located intra
septimum miliarium (:q.:); cf. :q.¸ a piscinis in arcus recipiuntur and .o.: a piscinis
in altiores arcus recipiuntur. At 6q.. the transmitted text places the piscina of Julia
<ad> sextum ab urbe miliarium.
lapide quadrato Ashlar masonry was used for both arcades, that of
Marcia and that of the much later Claudia-Anio Novus (see :¸..n.). To exam-
ples of quadratus ‘dressed’ (stone) cited in OLDs.v. ., add Curt. \.:.¸¸ pilas lapide
quadrato, Pliny, Ep. x.¸¸...
rzq.z maiore{m} parte{m} aquarum urb{i}s destituet<ur>
For the emendation see Rodgers (:q8¸) :¸¸–6. For ablative with destituere cf. Str.
iii.¡.¸, q.¸, ::.¸. The point is that interruptions here would deprive the City
COMMENTARY :.¡.¸–:.¡.¡
of ‘the greater part of its water supply’. That these five aqueducts supplied
more than the others is apparent both from F.’s own calculations (chapters
6¸–¸¸: Table ¸) and from the figures for erogatio (chapters ¸q–86: Table ¡). As
it happens, these five are the highest in level and thus are capable of being
delivered to all parts of the City (:8..). But it is most unlikely that F. would
have thought in these terms or described the pattern of their potential delivery
as maior pars urbis.
rzq.j huius difficultati<s: o>pus Editors have wavered between gen-
itive and dative, either of which construes with remedia. Huius could be an error
for his, but on the other hand the transmitted difficultatibus results, I think, from
misreading difficultatis opus. Poleni saw that a noun is needed for both inchoatum
and deficientis.
<o>pus inchoatum A temporary structure (cf. TLL ¸: q¸..¸¸) built up
to the level necessary to divert the water through troughs. On the difficulties
which such a structure would entail see Blackman–Hodge (.oo:) :o¸–8.
plumbatis canalibus The adjective (participle of plumbare) is very rare:
Bruun (:qq:) :q8. This seems to be the only use of it meaning covered (i.e.,
lined) with lead. Cf. Pliny, HN x.q¸, xii.8¸ plumbatis sagittis; Val. Max. iii.¸..
plumbatis tabulis; Solin. iii.¸ iacula.
rzq.q fere omnes specus The omission of Virgo in the S.C. (:.¸) is
certainly deliberate, else it would have been easier to say aquarum publicarum
(cf. :.¸.:, :.q.¡). It may be that Virgo (built only eight years earlier and not
plagued by calcareous incrustations) had at this date no need of repairs. It
is entirely plausible that Virgo differed from the older aque