B BFH3

Hffnhffi

ii$i

'iSMirTflH

WKjiM

(Entire af Stforal Arts Stbrarg

,

Greneral.....Hurndl
i

Ma^..2.i9 2

no,

574-1

Bennetts latin Series

The

Latin Language
A HISTORICAL OUTLINE
OF ITS

SOUNDS, INFLECTIONS, AND SYNTAX

BY

CHARLES

E.

BENNETT
IN

GOLDWIN SMITH PROFESSOR OF LATIN

CORNELL UNIVERSITY

BOSTON UNIVERSITY COLLEGE OF LIBERAL ARTS

LIBRARY

Allyn and
Boston
Neto gorfc

Bacon
Chicago

V-\

vV

o t

k^

COPYRIGHT,

1907,

BY

CHARLES

E.

BENNETT.

ADO

PA

FoU
PREFACE.
This book
is

a revision of
in

my

Appendix

to

Bennetts Latin
originally pre-

Grammar, published

1895.

That book was

pared as a series of lectures to advanced students on subjects

not covered in any Latin
\

Grammar

published in America.
to

The

title

"Appendix," however, was misleading and gave

many

a wrong impression of the purpose and scope of the book, which

was

in reality written

long in advance of the publication of

my
and

Latin

Grammar and
title
is

entirely without reference to that work,

to the views discussed

The new

more appropriate
;

the facts brought out

hence the change.

In the revision some dozen pages of old matter have been

omitted, while nearly forty pages of

1(

new matter have been

intro-

duced

but the general plan and scope of the book are un-

changed.
I

am

indebted to Professor

J.

C. Rolfe, of the University of

Pennsylvania, and to Professor Charles L. Durham, of Cornell

University, for valuable suggestions

made

while the book was

passing through the press.

C. E. B.

Ithaca, March,

1907.

111

Digitized by the Internet Archive

in

2011 with funding from

Boston Library Consortium

Member

Libraries

http://www.archive.org/details/latinlanguagehisOObenn

TABLE OF CONTENTS.
to*

CHAPTER

I.

Origin of the Latin Alphabet

Changes

in the

Form

of the Letters

Later Additions to the Alphabet

New

Characters proposed by Claudius

......... ....... ........ .......

CHAPTER
II.

THE ALPHABET.

PAGE
I

Sources of Information

The Vowels
a
e
* i

.......... ........... ...

.
. .
. . . .

PRONUNCIATION.

4
6
6

for

....

7
7

y
The Diphthongs
ae
oe

8
8

au
eu
w?

............ ............ ............ ............

.

8
9 10
10
10

The Consonants. The Semivowels

.11
Ir

J
v

12 18

The Liquids
/

18

r
v

18

vi

TABLE OF CONTENTS.
PAGE

The Nasals

18 18

m
n
nf, ns

n- adulter inum

......... ..........
.
.

20 20 20
2I

gn
The
Spirants
.
.

.22
22 22
23

/
s

The Mutes The Voiceless Mutes

t

........

23

23

23

24

k,q

.25
26 26

p
The Voiced Mutes
b

26 26
26
.

g
Distinction between Guttural and Palatal
.

27 27

The

Aspirates, ph, ch, th

Development of ph to/ The Double Consonants

........
.

.29
30
30

x
2

30
.

Doubled Consonants
Division of

31

Words

into Syllables

31

CHAPTER

III.

HIDDEN QUANTITY.
Methods of Determining Hidden Quantity General Principles of Hidden Quantity Vowels before ns, nf Vowels before gn, gm Vowels before nt, nd, ss Pontem, Fontem, Montem, Frontem, Frondem Hidden Quantity in Declension
.
.

Superlatives

....... ......... ....... ...........

. .
.

36 40

.40
40

.42 -43
46 49 50

Numerals

TABLE OF CONTENTS.
Pronouns
Conjugation

vii
PAGE

c
. .

.50
.

Root Forms

co
c?
ca

Verbal Endings

Compounds
Inchoatives

........... ...........
. .
.

54 54
55

Irregular Verbs

Word Formation
List of Chief
List of

Words containing Disputed Words

.

Long Vowel
.

before

Two

Consonants

56 66

CHAPTER
ACCENT.

IV.

Accent Defined

Character of the Latin Accent

Changes

in the Latin

Special Peculiarities

........... ........ ........ ..........

Accent

73
74

74 76

CHAPTER

V.

Standard of Spelling

Quom,

volt,

volnus,

etc.

Assimilation of the Final Consonant of Prepositions in

Compounds oi jacio List of Words of Doubtful

...........78 ..........
. . .
. .

ORTHOGRAPHY.

77

...

Compounds

79

82
83

or Varied Spelling

CHAPTER

VI.

THE LATIN SOUNDS.

The Vowels
Ablaut
Ablaut-Series
^-Series
-

.90
90
91

^-Series

a-Series

d-Series
5-Series

........... ........... ........... ... ....... ...........

92
93 93

94
94

^-Series

94

viii

TABLE OF CONTENTS.
PAGE

Vowel Changes
a

94
94
95

a
e

96

96
i

h
o o

97 97 98

u u
at
oi
ei

98

ui

............ ...........
.
.

98

98
99 100
100

au
eu,

100
101
101

ou

Shortening of

Long Diphthongs

Re-composition and De-composition

Shortening of Long Vowels

Compensatory Lengthening Assimilation of Vowels

.

Parasitic

Vowels

Syncope

Apocope

The Consonants The Mutes The Gutturals and Palatals The Dentals The Labials The Indo-European Aspirates in Latin
.

......... ........... .......

. . .

...... ..... ....... .......

.
. .

102

102
103

.104
104
104 104
105
105
105

106
106
.

107

bh

107

dh

107

gh

108 108
109 109

The Spirants The Liquids As Consonants As Sonants The Nasals As Consonants

As Sonants

........... ..........
.
. . . .

The Semivowels, J, v

.110 in .111 .111 .112

TABLE OF CONTENTS.
Consonant Changes . Initial Combinations
In the Interior of Words
Assimilation
.

ix
PAGE

113 113

114
. . . . . . .

115

Partial Assimilation

.116
1

Metathesis

16

Other Changes

117 118

At the End of Words

Disappearance of Syllables by Dissimilation
.

.119

CHAPTER
Declension of Nouns and Adjectives
^-Sterns
. .

VII.

0-Stems
Consonant Stems
/-Stems

....... ..........
.

INFLECTIONS.

120

.120
124 128

Stem Formation of Consonant Stems

.130
.131
133

Consonant Stems that have Partially adapted themselves

/-Stems
/-

to

2 -stems

134
"

and /-Stems

136

7?-Stems

Stems ending in a Diphthong

Formation of the Comparative and Superlative

Numerals
Cardinals

............
.
. .
.

.......
. .
.

136 138
139

140 140 143

143

Ordinals
Distributives

Multiplicatives

144
144

Pronouns
First

Person
Reflexive

Second Person

The

Possessives

Demonstratives

........... ........... ........... ..........

. .
.

144
145

146
146 147
147 149

Hie
Is
Iste,

Die, Ipse
Relative, Interrogative,

150

The

and Indefinite Pronouns

151

Pronominal Adjectives

152

TABLE OF CONTENTS.
PAGE

Conjugation
Introductory
. .

152
.
. .

.152
.

Formation of Present Stem

Unthematic Presents

........
. .

153
153
I

Thematic Presents
Root-Class

55

155

Reduplicating Class
T-Class
. .

156
. .

.156
156

N-Class

NO-Class
SCO-Class
JO-Class

.156
157

Tense Formation

in the Indicative

.......
. .

157

158 158
159 159

The Imperfect The Future The Perfect

Reduplication
.

159

Stem Formation

The

Primitive Perfect
-si
. .

.160 .160
.

Perfect in

161

Perfect in -vl Perfect in -ui

.161 .162
162

Inflection of the Perfect

The Pluperfect The Future Perfect

164
.
.

164 164
164
165

The

Optative

Present
Aorist

...
.
.

The

Subjunctive
^f-Subjunctives

..........
.
.

166

.
.

^-Subjunctives

.166 .166
167 167

The Imperative
Active
Passive
.

.168
168

The

Personal Endings
Active
Passive
. .
.

168
.
.
.

The

Infinitive

.169 .170
.

Active
Passive
.

171
I

7I

The

Participles

.172

Gerund and Supine

*73

TABLE OF CONTENTS.
CHAPTER
VIII.

xi

ADVERBS AND PREPOSITIONS.

PAGE

Adverbs
Accusatives
Ablatives
.

174
.
-

.174 .174
175

Locatives

Instrumentals
Prepositions
.

...
.

175 175

Origin of Prepositions
List of Prepositions

.........
.
.

175

176

CHAPTER
SYNTAX.

IX.

The Cases Names

181
of the Cases
.
.
.

181

Review of Case Theories The Localistic Theory The Logical Theory The Grammatical Theory Subsequent Views
.

183
183
. .

.184
184

.185
.

185 The Accusative 186 With Passives Used as Middles Produced Of Result 187 Of Person Affected, and of Result Produced Dependent upon .187 the Same Verb
.
. .

Synecdochical or Greek Accusative

187
.

In Exclamations

As Subject of the Infinitive Id genus, muliebre seats, etc.

.......
.
.

.188
188

.188
189

Original Force of the Accusative

The Dative
Original Force

.........
'favor,' 'help,' etc
,
.

.191
191 191

Dative of Indirect Object

With Verbs Signifying With Compounds

Dative of Reference
Ethical Dative

192 192
193

193 193

Dative of Agency Dative of Purpose

193

Xll

TABLE OF CONTENTS.
PAGE

The

Genitive

194

Original Force

194
194
195 195 195

With Nouns
Genitive of Quality

Genitive with Adjectives

Genitive with Verbs

The Ablative

197

Syncretism in the Ablative

198 198 198 198

199

Genuine Ablative Uses

Separation

Source

Comparison
Instrumental Uses
.

199 199

Accompaniment
Association

200 200 200

201 201
.

Attendant Circumstance

Manner
Accordance

Means

Way

by Which

202
203

Cause

Degree of Difference
Price

203 203

Quality
Specification

204
205
205

Ablative Absolute

Locative Uses

206
206 207 208

Place Relations

Refert and Interest

Time Relations
Locative of the Goal
Surviving Locative

208 208
209

Forms

The Moods The Subjunctive

Original Force of the Subjunctive

210
211

Original Force of the Optative

Classification of Subjunctive

212
2I 5

Uses

Subjunctive in Principal Clauses

Original Uses
Volitive Subjunctive

215

215
215

Optative Subjunctive

217

TABLE OF CONTENTS.
Subjunctive of Contingent Futurity

Xlll

PAGE

Derived Uses

.......
.
.

218 218 218

220
221

Extensions of the Jussive and Prohibitive

Extensions of the Deliberative

Extensions of the Optative Extensions of

Futurity

the

Subjunctive

of

Conti ngent
.

222 223 223

Subjunctive in Dependent Clauses

Parataxis

and Hypotaxis

Subjunctive of Purpose
Clauses of Characteristic Clauses of Result

223
225

227

Causal Clauses

228 228 229 229

240
241 241
241

Temporal Clauses
Substantive Clauses
.

Developed from the Volitive Developed from the Optative

Of Result

....

Indirect Questions

Conditional Sentences
Clauses with quamvis
Provisos

243 243

LIST OF ABBREVIATIONS

AND

SIGNS.

Archiv

Wolfflin's

Archiv fur Lateinische Lexikographie und Grammatik.

Leipzig, 1884-1907.
Berlin,
1

Vols. I.-XV.

= CIG. = CIL. =
CIA.
Gr. or

Corpus Inscriptiotium Atticarum. Corpus Inscriptionum Graecarum.

Corpus Inscriptiotium Latinarum.

873

ff.

Berlin, 1828
Berlin, 1863

ff.

ff.

E. L. D.

Lewis, Elementary Latin Dictionary.

the author's Latin

New

York, 1895.

Grammar =

Grammar.

Boston, 1895.

Grober's Grundriss

Grober's Grundriss der Romanischen Philologie.

Strassburg, 1888

ff.

Keil

Grajnmatici Latini, ed. Keil.

Wbrterbztch

Leipzig, 1855

ff.

Korting,

Korting, Lateinisch-Romanisches

Wbrterbuch.

2d edition.

Paderborn, 1901.

Marx

Marx, Hulfsbuchlein fur die Ausspr ache der Vokale in positions langen

Silben.

3d

edition.

Berlin, 1901.

References by are to the Latin Language

itself.

Words marked with a

star are hypothetical forms.

{e.g. a, e)

Vowels printed without the macron

are short

for greater precision

these are sometimes printed with a breve

{e.g. a, e).

xiv

CHAPTER

I.

THE ALPHABET.
The Latin alphabet is a development of that type of the Greek alphabet known as the Chalcidian. In the widest sense
1.
i.

the term
bets
;

'

Chalcidian

'

is

applied to
it

all

the non-Ionic

Greek alphacolonies,

in a

narrower sense

designates the special alphabet of the

Chalcidian colonies of lower Italy and Sicily.

settled originally

These

from Chalcis

in

Euboea, date from very early

times.

Cumae,

in fact, is said to

have been founded as far back

as 1050 b.c.

But most of the Chalcidian settlements do not

B.C.

antedate the eighth century

It

was probably from the Camthat sometime in the sixth

panian colonies of
century
B.C.

Cumae and Neapolis

the Chalcidian alphabet was introduced into Latium.

Special peculiarities of this alphabet are the following

2.

The

character

H was

lacking,

X was used as
was C.
viz. ?,

x,

and

Y (v)

as ch.

Lambda, which
]/,

in Ionic

had the form

(Attic
l~)

A, took in Chal-

cidian the form

while

Gamma

Besides K,
called Koppa.

another character for the -sound existed,

For Rho, R was employed as well as

of that letter.

P, the ordinary Attic

form

In conformity with

its

Chalcidian origin the earliest Latin alpha:

bet consisted of the following twenty-one characters

A B C

(=)
3.

DEFI(Z)HIKIfMNOr?R^TVX.
Of these
characters,
\i

subsequently became

L.

C in course
in

of time

came
:

to

be used for K, which then disappeared except

a few words

new
C.

For the ^-sound a was invented, by appending a tag to the older But permanent traces of the original value of C as g,
character, G,

Kalendae, Kaeso, Karthagd.

THE ALPHABET.
in the abbreviations C. for

remained

Gains and Cn. for Gnaeus.

The new character G took the place hitherto occupied by I, which now disappeared. These changes are ascribed, with
some degree of probability, to Appius Claudius, Censor 312 B.C. P was at first open as in Greek, but subsequently became P. The Greek alphabet had no character to represent the sound of/, but the Greek Digamma (F) represented a closely related sound, v. This F, combined with H (apparently to indicate the
voiceless character of the sound, as opposed to that of the

Greek

Digamma), was introduced

designate
(=fefaced,
the
i.e.

into the early Italian alphabets to

sound
fecit),

of /.
in the

An

example

is

FHEFHAKED
alone.

earliest extant Latin inscription,

CIL.
4.

xiv. 4123.

Later, the
letters

H was discarded and F used

(0),

The Greek

(<),

and

Y V

(X),

being

aspirates, represented sounds

which did not originally

exist in
intro-

the Latin language.

These characters were accordingly

duced as numerals,
quently

as 100,

as 1000,

as 50.

Subse-

became G, and
became
first 01,

finally C.

This

last

form resulted
initial letter of

perhaps from associating the character with the

centum.

and

later

M, a change facilitated

probably by association with the

initial letter of nulle.

The

half of

viz.

D,

was used

to designate 500.

(50)

became successively ^, _L, and L. 5. In Cicero's day Y and Z were introduced

tion of

for the translitera-

Greek words containing

u,

v or

Previously Greek v had

ss (medial),

been transliterated by
Olumpio, sona

and

by

s (initial),

as,

(luvrj), atticisso (olttlkl^o)).

The Emperor Claudius proposed

characters, J to represent v
{i.e.

the introduction of three

new

our w), 3 (Antisigma) for ps,

i,

and

(-

to represent the
etc.

middle sound between u and

as seen

in optumus, optimns,

some

inscriptions of

These characters were employed in Claudius's reign, but gained no further

xi. 14.

recognition.

See Tacitus, Ann.

THE ALPHABET.
On
the alphabet in general, see

3
Geschichte
des

Kirchhoff, Studien zur

Oxford, 1894.

Griechischen Alphabets.

4th

ed., Berlin, 1887.

Lindsay, Latin Language.

Clarendon Press.

p.

ff.

Encyclopaedia Britamiica, Article Alphabet.

Johnson's Encyclopaedia, Article Alphabet.

Sommer, Handbuch der Lateinischen Laut- und Eormenlehre.

2.

p. 25

ff.

In writing

in the

Grammar

to represent the Latin i-cdn-

sonans, reference has been

ations.

had mainly

to practical consider-

Typographical distinction of the vowel and consonant

i is

sounds of
apart.

absolutely essential to enable the pupil to

i is

tell

them

Where

written for both sounds there

is

nothing to
;

show the student

Gaius
is Ga-i-its.

that

iam

is

jam;

that etiam

is

et-i-am

or that

Moreover,

it is still

usual to distinguish befor the former,

tween the vowel and consonant

u,

by writing u
are

and v
Zusatz

for the latter.

The two

cases

perfectly parallel.
p. 8,

See Deecke, Erlauterungen zur lateinischen Schulgrammatik,

2.

CHAPTER

II.

PRONUNCIATION.
3.

Sources of Information.

Our

sources of knowledge con-

cerning the ancient pronunciation of Latin are the following

Much has been by the Roman grammarians on the subject pronunciation, far more
a)

Statements of Roman writers.

left

of

commonly supposed. The remains of the grammatical writers as collected and edited by Keil under the title
in fact than
is

Gra?nmatici Latini (Leipzig, 1855-1880)

fill

eight large quarto

volumes.

These writers cover the

less

entire field of

grammar, and

most of them devote more or

space to a systematic consider-

ation of the sounds of the letters.

this subject

As

representative writers on
(fl.

may be
350

cited:

Terentianus Maurus
Litteris,

185 a.d.),
;

author of a work entitled dc

Victorinus
tury a.d.
of
;

Syllabis, Metris

Marius
cen-

(fl.

a.d.)

Martianus Capella (fourth or

;

fifth

not in Keil's collection)

Priscian

(fl.

500

a.d.),

author

the

Institutionum

Grammaticamim Libri

xviii.

Even the

classical writers

have often contributed valuable

bits of infor-

mation, notably Varro in his de Lingua Latina, Cicero in his

rhetorical works, Quintilian in his Institutio Oratoria,

and Aulus

Gellius in his
b)

Nodes

Atticae.
is

second important source of evidence

found

in inscrip-

tions.

The

total

body

of

these

is

very great.

The
of

Co?'pus

Ifiscriptionum Latinarum, in process of publication since 1863,

consists already of fifteen large folio volumes,

some

them

in

several parts,
close

and

is

not yet completed.

These

inscriptions dis-

many

peculiarities of orthography

which are exceedingly

spellings as vrps,

instructive for the pronunciation.

4

Thus such

SOURCES OF INFORMATION.

pleps, by the side of vrbs, plebs, clearly indicate the assimilation

of b to p before
s.

Even the blunders

of the stone-cutters often

give us valuable clues, as, for example, the spelling

for

acletarvm
See

athletarvm, which shows that the

th

was

practically a /
c.

otherwise

we could not account

for

its

confusion with

3 J
c)

Greek transliterations of Latin words constitute a third

source of knowledge.
historians of

Not only Greek

affairs),

writers (especially the

Roman

but also Greek inscriptions, afford

us abundant evidence of this kind.

(Cicero)
Atoviia

Thus the Greek

Kt/cepwv

furnishes

support for the /-sound of Latin

c\

while

and OuaAevna bear similarly upon the w-sound

of Latin v.

The

inscriptions are naturally

much more

trustworthy guides in

this matter

than our texts of the Greek authors, for we can never

in the

be certain that the Mss. have not undergone alterations

process of transmission to modern times.

d)

The Romance Languages

also, within

limits,

may be

uti-

lized in determining the

riss

sounds of Latin.
I.,

See Grober's GrundStrassburg, 1888

;

der Roma7iischen Philologie, Vol.

W.

Meyer-Liibke,
Leipzig, 1890.
e)

Grammatik der Romaiiischen Sprachen, Vol. L,

as analyzed

The sound-changes of Latin

itself,

by etymologi-

Modern scholars, particularly in the last fifty years, have done much to promote the scientific study of Latin sounds and forms, and, while much remains to be done, the ultimate solution of many problems has already been reached. As representative works in this field may be cited
cal investigation.
:

Brugmann, K. Brugmann, K.
Sprachen.

Grundriss der Vergleichenden Grammatik der IndogermaVol.

I.,

nischen Sprachen.

2d

ed.

Strassburg, 1897.

Kurze

Vergleichende

Grammatik der Indogermanischen

Muller's Handbuch der Klassischen

Strassburg, 1902.

Stolz, F.

Iatei?iische

Grammatik
Vol.
II.,

in

Altertumszvissenschaft.

3d

ed.

Munich, 1900.

6
Stolz, F.

PR ON UN CIA TION.
Lautlehre der Lateinischen Sprache.
Leipzig, 1894.

Lindsay,
Giles, P.

W. M.

The Latin Language.

Oxford, 1894.
Philology for Classical Students.

A
F.

Short

Manual of Comparative

2d ed.

London, 1901.

Sommer,
Riemann,
Vol.
I.

Handbuch der Lateinischen Laut- und Formenlehre.

Goelzer, H. Gramtnaire Comparee du Grec

Heidel-

berg, 1902.
O., et
et

du Latin.

Paris, 1897.

Henry, V.

Grammaire Comparee du Grec

et

Latin.

5th ed.

Paris, 1894.

As special works on pronunciation alone may be

Seelmann, E.
Roby, H.J.
Ellis,

cited

Die Aussprache

des

Latein.

Heilbronn, 1885.

The most

important work on the subject yet published.

./

Latin Grammar.

Vol

I.,

4th ed.

pp. xxx-xc.

London, 1881.

Alexander.

The Quantitative Pronunciation of Latin.

London,

1874.

discussion of special problems.

'

See also the chapter on

Pronunciation' in the work of Lindsay

above

cited.

SUMMARY OF THE EVIDENCE.

THE VOWELS.
4.

A.

The consensus

of

the

Romance languages

indicates

clearly that a

was pronounced substantially as

that a
it

in English father.

In the absence of any specific evidence to the contrary, we

safely

may
in

believe

had the same sound

less prolonged.

qualitatively;

quantity, of course,
5.

was

E.

Long

was probably

close,

i.e.

spoken with the vocal

Short

organs (more particularly the tongue and hard palate) nearer

together than in the utterance of short
e.
<?,

on the other
rela~e

hand, was open,

i.e.

spoken with the tongue and hard palate

tively further apart.

These differences

in the

pronunciation of

and

are confirmed

by the testimony
vi.

of the

grammarians,
iv.

e.g.

Marius Victorinus (Keil,

33. 3)

Servius (Keil,

421. 17)

Pompeius

(Keil,

v.

102.

4).

The Romance

languages also,

though they have

lost the original quantitative distinctions of

THE VOWELS.

the Latin, have preserved with great fidelity the qualitative distinctions of the close

and open

e.
~e

See

36. 5.

It is to

be noted

that the relation between Latin

and

stands in marked contrast

rj

with the relation existing between Greek

and
was

e.

In Greek
It

it

was the long

ual

<?-sound

(rj)

that

was open

close.

should
unus-

further be observed that in our normal English speech

it is

and

difficult to

pronounce a pure
e*~,

e.

We

regularly

add an
etc.

z-sound and pronounce a diphthong,

6.
I.

e.g. in

fatal, paper,

1.

Long

was

relatively closer than short

in

/,

as

shown

by the fact that 1 appears unchanged

from the Latin, while open character of
of e for
2.
1 i 1

Romance words descended

e.

regularly appears as

This relatively

is

also indicated
e.g.

by the occasional occurrence

tempestatebvs

in

Latin inscriptions,
b,

(= ibus).
1

Before the labials p,

/, m,

an earlier u changed to Republican period.

in

many words
is

at about the close of the

This
:

confined regularly to unaccented syllables.

recupero
lubldo

Examples are

recipero
libido

pontufex

pontifex

lacruma
maxunius, optumus,
etc.

lacrima

maximus, optimus.

Quintilian,

i.

7. 21, tells

us that Julius Caesar was said to have

been the
tween

first

to introduce

the

new orthography.
it

In

i.

4.

Quintilian further states that the sound

i

was intermediate bewill

and

u.

The Emperor
secure
this

Claudius,
for

be remembered,
character

endeavored
(h)
to

to

recognition

special

represent

intermediate
of

sound, which

probably
This view

was approximately that

gains support from the

French

u,

German
e.g.

u.

occasional employment of

for i in

words
CIL.

of the
ix.

category under discussion,

;

contybernalis
the sound of
it.

2608

illacrymant.

This

y had

See below under y.

8
7.

PR ON UNCIA TION.
0.

Long

was

close, i.e.

nearer the ^-sound

short o was
clearly indi-

relatively open, that

is,

nearer the tf-sound.

This

is

cated by the descriptions of the sound as given by the

Roman
vi.

grammarians,

e.g.

Terentianus Maurus (Keil,

vi.
2>2>-

vi.

329. 130-134);
421.

Marius Victorinus (Keil,

17-19)
;

3~8)

Servius (Keil,

it is

further confirmed by the testimony of the

e (see

Romance

languages, which, as in case of

above), have faithfully preo,

served the qualitative character of Latin o and

lost the original quantitative distinction.

while they have

See

36. 5.

Short
top,

should never be pronounced like English o in

not, etc.

hot,

rock,

English o in these words really has a short

English obey and

tf-sound.

Latin o was a genuine #-sound.

it.

melody well exemplify

8.

TJ.

Short u was relatively more open than

u, as is

shown by
also

the frequency with which Latin inscriptions show o for u, as

eroditvs, secondvs, nomero.

The Romance languages

;

have o for Latin

9.

u, as Italian

sovra {super)

ove (ubi),

etc.

Y.

In conformity with
of

its

origin, Latin
u,

y (= Greek
il.

v; see

1.5) had the sound

xii. 10.

French

German

Cf. Quintilian,

27,

who mentions the sound

as different

from any existent

12.

in native Latin words.

See Blass, Pronunciation of Greek,

THE DIPHTHONGS.
10.

AE.

1.

The

original
till

form of

this

diphthong was

at,

spelling which prevailed

in the

about 100
i.

B.C., e.g. aidilis,

qvairatis
a

Scipio inscriptions (CIL.

at in

32. 34).

The sound was

genuine diphthong (that of

English
Cf.

aisle),

and continued
Terentius

such throughout the classical period.

transliteration of Latin words,

the use of at in Greek

Kataap.

e.g. irpatrop,

Scaurus

(first half of

second century a.d.) bears testimony to the

vii. 16. 9),

diphthongal character of the sound, when he says (Keil,

THE DIPHTHONGS.
a propos of the orthography, that ae
than
ai,
is

a more accurate designation

d'-sound.

as the second element

is

an

He

thus clearly
i.e.

indicates that there

that the

was a second element

in the

combination,

sound was diphthongal.

This difference between ai

and

ae,

though a real and perceptible one, was probably not

very great.
2.

By

the fourth century a.d., however, ae

had
It

altered

its

character and had

become a monophthong.

This change had

originated

begun
general
is

in the first century a.d.

or even earlier.

probably in the rustic and provincial speech, but did not become
till late.

Conclusive evidence of the new pronunciation

found

in the frequent

occurrence in inscriptions of such

spell-

hec (= haec), qvestor, etc. does not become frequent till after 300
ings as Cesar,

But
a.d.

this

orthography

See Seelmann,

Aus sprache
11.

des Latein, p.

224

f.

OE.

The

earlier

form of

oe

was

oi.

But

oi regularly de-

veloped to

u, e.g. utilis for earlier oitilis ;

unus for oinos.

later oe, e.g.

In a

few words

oi resisted this
etc.

change and became

moenia

(yet munio), foedus,

The sound was

a genuine diphthong

throughout the classical period.

traces of a

In the vulgar language we find

in the third

monophthongal pronunciation

and fourth

centuries a.d., a change which ultimately

evidence tends to show that

ae,

oe,

became prevalent. The and e in the late centuries

which gives us the key
Mss. of the

became extremely
Latin writers.

similar in sound, a fact

to the hopeless confusion of spelling in our mediaeval

Thus we
hollow

find caelum written as coelum, a spelling

its

doubtless suggested in part by

fancied derivation from the

Greek
and

koiAos

'

'

cena,

'

dinner,' appears variously as caena

coena, the latter spelling being perhaps a result of association

'

with Greek koivos

common,
is

'

i.e.

'

the

common meal

'

tie,

the

asseverative particle,

often written nae, probably another inCf. vat

'

stance of Greek influence.

verily.'

Other instances of

PR ON UN CIA TJON.
;

confusion are cerimonia for caerimonia

terium (Gr.
Koi/jLrjTrjptov) ;

cetrieferium for coerrie;

;

moestusiox maestus ; foe?nina for femina

koX erepoi)

caeteri for ceteri

(probably owing to the influence of Gr.

;

coelebs for caelebs

coecus for caecus.

Some

of these false forms

are unfortunately
12.

still

printed in our texts of the classical writers.

like

AU
Cf.

was a true diphthong, pronounced

Eng. ow in
as

how.

Greek

transliterations of Latin proper

names such

IlaovAAu/77 (Paulina), ^aoo-rivos (Fausfinus).

13.

EU
u.

appears in Latin in only a few words, and in these

Primitive Latin

is

of

secondary origin.

eu

early

became
:

ou,

whence
seu, heu.

The chief Latin words that have eu are ceu, neu, The combination appears also in numerous proper
e.g.

names borrowed from the Greek,

Europa, Teucer.
i.e.

In

all

these the sound was that of a genuine diphthong,

an

<?-sound

quickly followed by an ^-sound, both being uttered under one

stress.

14.

UI appears

to

have been a genuine diphthong in

In the
first

cut, huic,

and hut
in his

(the interjection).

two of these words ui was

tells

certainly of secondary origin.

Quintilian

us

(i. 7.

27) that

boyhood (about 50 a.d.) quoi was still in use, and that its pronunciation was substantially identical with that of qui (the
Norn.).

Some

scholars have accordingly inferred that qui and cut

distinct,

were simply graphically

being alike in pronunciation.

Consistently with this view they regard the u in cui as

= v,

and

mark

the

long, viz. cui.

But

if

the facts were thus,

we should

expect

cut,

when

resolved into two syllables in verse by metrical

be an iambus (y _). Such is not the case. On the other hand, we find it appearing as a pyrrhic (w w), and that,
license, to
too, at just

about the time when,

if

we may credit
is

Quintilian, cui

began

to supersede quoi, viz.

soon after 50 a.d.

Apparently the

earliest instance of the resolution

mentioned

in Seneca, Troades

THE CONSONANTS.
852 cuicumque (about 55
a.d.).

II

Subsequently, in Martial and

Juvenal, such resolutions are frequent.

See Neue, Formenlehre

der Lateinischen Sprache, 3d ed.,

(e.g.

ii.

p. 454.

Very

late writers
it

Prudentius, 400 a.d., Venantius Fortunatus, 600 a.d.),

is true,

sometimes have

cui in verse, but there is apparently

no

trace of any such resolution in the early centuries of the Empire.

Another argument

may be found
and cui
is

in the verse treatment of huic.

The

id in both huic

obviously of the same nature, and

those

who

write cut also write huic.

hv-,
it

But

if

huic were huic,

then hu- must stand for

since the

word

is

a monosyllable.

But

if it

begins with hv-,

would not permit

elision of a preced-

ing vowel in poetry.

But

elision does occur before huic.

There-

fore huic begins with hu-.

That being the case and the word

i

being monosyllabic, ui goes together to form a diphthong, the being short, as in

all

diphthongs ending in
to

i.

The
to

of huic

and

cui would therefore

seem

have been short, and

It

have blended

with the u to produce a diphthong.

must, of course, be con-

ceded that the pronunciation of cui could not have been widely
different

from qui

yet

it

must have been

sufficiently so to

keep

the two words distinctly separate in

Roman

speech, a view which

receives the very strongest confirmation in the fact that the

modern

Italian has chi as the

descendant of Latin

qui,

but cui

(with diphthongal ui) as the descendant of Latin cui.

THE CONSONANTS.

The Semivowels,
15.
like
J.
1.

/,

v.

J (Seelmann,
(I)

Aussprache des Latein,

:

p.

231

fT.)

was

our y in yes.

Evidences

a)

single character
i

sufficed with the

Romans

to indicate

both the vowel

and the consonant / (i consonans). This would indicate a close proximity in sound between i and/, a proximity
if

manifestly existing

Latin

j was

English y.

C/. } for

example,

PR ON UN CIA TION.

English

New

York with a hypothetical

i

New

I-07'k.

In any

English word the vowel

may

easily

be made

to pass into the

semivowel y by energetically stressing either the preceding or the

following vowel.
b)

The Roman grammarians nowhere suggest any

essential

difference in sound between the vowel and consonant functions of

the character, as they almost certainly would have done, had the

consonant been other than the corresponding semivowel.

On
is

the

other hand, the grammarians repeatedly suggest a close proximity

in the pronunciation of i

and

/.

Thus Nigidius Figulus

6) as
is

cited

by Gellius {Nodes
conception that
I

Atticae, xix. 14.

warning against the

a vowel.

in Iam,

Iecvr, Iocvs

Such a

warning can have no meaning whatever, except upon the assumption that the

sound of

j was

very close to that of

i.

1,

i.e.

was the

semivowel y.
c)

Cf. Quintilian,
i,

4. 10.

In the poets,

when followed by another

vowel, often

becomes consonantal, uniting with the preceding consonant to make position e.g. abietis, parieie?n, ariete become abjetis, par;

jetem, arjete.

In these cases the consonant sound can have been

of the of

none other than that

(trisyllabic),

semivowel

y.

Cf.

also

nunciam

compounded

nunc and jam; etiam, compounded

as the

of et

and jam.
1

d) Greek transliterations of Latin words employ

nearest equivalent of Latin/,
2.
e.g.

'IouAios

In the

last

centuries

of

the Empire,

(=Julius). j seems

to

have

progressed, at least in the vulgar speech, to a genuine spirant,

probably similar in sound to that of

z in the

English word azure.

Thus

in late inscriptions (from the third century on)

we

find such

spellings as

Zksu (= fesu), zunior (= junior), sustus {= Justus),

Cf.

Giove
16.
It

(z=zJove).

Seelmann, Aussprache des Latein,

p. 239.

V.

Fis a

labial semivowel, with the

}

sound of English w.
i.

corresponds to the vowel u just as/ corresponds to the vowel

THE CONSONANTS.
The evidence
a)

single character (V) sufficed with the

Romans

to indi-

cate the vowel u (u vocalis)

and the consonant u

(u consonans).
v,

This indicates a close proximity in sound between u and

proximity which manifestly existed,
if

Latin v was English w.

For the vowel u naturally passes into

w before a vowel whenever

is

either the preceding or following syllable

energetically stressed.

For example, tenuia easily becomes

be so read
b)
first

tenvia,

and must repeatedly

to the close of the

in verse.
(at least

The Roman grammarians

down

century a.d.) nowhere suggest any essential difference in

sound between the vowel and consonant functions of the character V,

no more than

in the case of the analogous

I,

I.

On the other

hand, just as in the case of

v were very similar.

they repeatedly suggest that u and

Figulus, cited above in con-

Thus Nigidius

nection with the discussion of j, observes in the

(Gellius, xix. 14. 6) that initial V in

same passage
is

Valerivs, Volvsivs,

not a

vowel, an observation which would be pointless unless the sound

of v

had been

closely similar to that of u>

i.

i.e.

had been

that of w.

Quintilian in
c)

4.

10 gives, a similar warning.

The same Nigidius Figulus (Gellius, x. 4. 4) says that in pronouncing vos we thrust out the edges of our lips, which conforms
physiologically to the pronunciation of v as English w.

d)

The Greek

ordinarily transliterates Latin v

by means

of ov,

as OwXeptos {Valerius), OvoXctkol (Volsci), AtovCa {Livid).

e)

7 and v often interchange in the

(e.g.

same words.

Thus

early

Latin la-ru-a

Plautus, Captivi, 598) appears later as a dis-

syllable, larva.

Similarly mi-lu-os appears later as milvus.

e.g.

In

verse, silva occurs repeatedly as si-lu-a,

Horace, Odes,

i.

23. 4.

On
etc.

the other hand, tenuis, puella,

etc.,

often appear as tenvis,pvella,

This interchange

is

conceivable only upon the supposition

akin.
vii.

that the vowel

and consonant sounds were closely

first

Cf. also

Velius Longus (close of the

century a.d.) in Keil,

75. 10,

'

PR ON UNCIA TION.
'

to the effect that a-cu-am,

I shall sharpen,'

and aquam, 'water

(where qu

is

simply the traditional inconsistent spelling for qv),

Caesellius (see Seelmann,
is

were

liable to confusion in his day.

p.
;

Anssprache des Latein,

234) cannot say whether tenuis

while in the

a dis-

syllable or a trisyllable

Romance languages we
Old French
teneve (rep-

sometimes find doublets pointing to parallel Latin forms, one with

u
vocalis,

another with u consonans,

a

e.g.

resenting
ten-vis).

Latin te-nu-is) and

tenve

(representing a Latin

Italian soave points to the existence of a Latin su-a-vis

(i.e.

by the

side of sua-

sva-) vis.

Cf.

Seelmann,

p. 234.

f) The phonetic changes incident point in the direction of the w-sound

(root fav-)

to word-formation

also

of
;

we getfau-tor

(for *fav-tor)

Thus from faved from lavd (root lav-) we

v.

get lau-tus (for */av-tus).

rally

In such cases the semivowel v natu-

becomes the vowel u and combines with the preceding

vowel to form a diphthong.

Had

v been a spirant, either labioit

dental, like our English v, or bilabial,

would naturally have

Cf, for example,

become/before tin the foregoing examples.

our English haf
to (colloquial) for

hav(e)

to.

The evidence given under f)

roborative testimony.

holds, of course, only for the

;

formative period of the language

but
is

it

is

valuable as cor-

For Latin v

all

the

more
if it

likely to

have been a semivowel

in the historical period,

was such

immediately anterior to that period.

g)

The

contracted verb-forms, such as amasti for amavisti,

delesti for de/evistz, audisti for

audwisti, com??iossem for co77uno-

vissem, all point to a semi-vocalic

easily disappears

sound for
in

v,

since this sound

syllable.

between vowels

an unstressed

Cf

English Hawarden, pronounced Harde7i ;

tord.

toward, pronounced

h)

Several anecdotes found

among

ancient writers give fur-

ther confirmation of the similarity in sound of u

and

v.

Thus

Cicero (de Divinatione,

ii.

84) relates that,

when Marcus Crassus

THE CONSONANTS.
was preparing
to set sail

from Brundisium on
figs

his ill-fated expedi-

tion to the East,

he heard a vender of

on the street cry out

Cauneas, really the

name

of a variety of figs, but

which Cicero
viz.

suggests was intended by the gods as a warning to Crassus,

cavie) n(e) eds, don't go.
2.

While the above evidence may be accepted as

fairly con-

clusive for the pronunciation of Lat. v as

in the best period, to

indications

are not wanting that v

had begun

change

to a

spirant sound before the period of the decline.

The

earliest

testimony on this point

first

is

that of Velius
of v as
vii.

Longus

(close of the
asfiirdtio,

century

a.d.),

who speaks

having a certain
v irom

e.g. in

valente,

primifwo (Keil,

58. 17).

This reference to
its

asplrdtio hints at the

development of

earlier value as

a bilabial
spirant,

{i.e.

produced by the two

lips)

semivowel to a bilabial
is

somewhat
{i.e.

similar to our English v, except that our v

lip).

labio-dental

produced by the teeth and lower

confused with

This

view

is

confirmed by the fact that, beginning with the second

a.d.,

century

we note

that v

is

b,

which had also

become
a.d.

a bilabial spirant at this period.

its

This confusion, which

increases as time goes on, reaches

height in the third century

;

Examples are

biginti

(=

vigintt)

vene (=bene)

Favio

{= Fabio).
3.

Some

scholars have sought further confirmation of the

spirant character for the period referred to (100 a.d.

and

afterv.

wards) in the use of Greek

/3

as a transliteration of Latin
ft

Beginning with about 100 a.d. we find

frequently employed in

Greek
e.g.

inscriptions in place of earlier ov for such transliterations,

;

Kovfievros {conventus)

fiipva

(vernd)

KaA/8eivos (Calvinus).
ft

Similarly our text of Plutarch (about 100 a.d.) usually has

in

Latin words

(e.g.

BaXepios, BeVov?
ov.

= Venus)
it is

where

earlier
(cf.

Greek

writers mostly

employed

Now

believed
(3

Blass, Pro-

nunciation of Greek, p. 109) that

of the

Greek

at this time (beginning

second century a.d.) had become a bilabial spirant. How-

PR
may
for v

NUNCIA TION.
would be gained from that
v.

ever this

be, little support

fact

for the pronunciation of Latin

For while

it

is

true that the

a.d., yet the

use of

(3

assumes great frequency from ioo

still

earlier spelling ov

remains the predominant one.

Eckinger,

Orthographie Lateinischer W'orter in

p. 87,

Griechischen Inschriften,
f3

gives 234 instances of ov as against 100 of

in

Greek

inscriptions of the second century a.d., while often the

inscription exhibits both spellings.

same
in-

Moreover, occasional
last years of the
first

stances of

(3

=v

occur as early as the

Republic,
B.C.,

Eckinger, p. 87, cites five examples from the

century

and twenty one from the

either ov or

first

century a.d.

The

facts

seem

to

indicate that the Latin sound

ft
;

was not adequately represented by

that Latin
at a time

consequently no permanent equivalent was ever

is,

adopted.

It

therefore, perfectly conceivable

(3,

should have been transliterated by Greek

the latter sound

even

when
In

had not progressed

to its spirant

stage.

fact, it is quite possible that

the confusion in Latin

itself,

which

resulted in writing b for

v,

may have

contributed to the increas(3

ing frequency in the employment of

as against earlier ov in

Greek

transliterations

of Latin

words.

The two phenomena

connection suggested

coincide so accurately in time that the

becomes extremely probable.

Even
(as
it

if

Greek

(3

had by 100

a.d.

become

a bilabial spirant

certainly did ultimately), yet this

would not necessarily

v.

prove anything for the pronunciation of Latin

spirant
is

For the

bilabial

very easily confused with the semivowel.

Thus

the

dialectal pronunciation of

German Wein, Winter

with an

initial

bilabial spirant easily deceives

to

American and English

travellers,

whom

this

sound

is

not familiar, and produces the impression

that an English
therefore,
is

is

pronounced.

The evidence
it

of the Greek,

purely negative, and while

seems probable, as

already indicated, that Latin v at about the beginning of the

second century a.d. had begun to become a bilabial spirant, this

THE CONSONANTS.

I?

conclusion rests upon other grounds than the evidence of Greek

transliterations.
4.

Gothic and Anglo-Saxon loan-words have been thought

by-

some

to confirm the Z/-sound of Latin v, but without reason.

it is

Gothic and Anglo-Saxon w,

true,

appears regularly as the

e.g.

representative of v in words borrowed from the Latin,

wein, 'wine' (Lat. vinuni)
;

Gothic

aiwaggeli, 'gospel' (Lat. evangeliutri)\,

-wic,

Anglo-Saxon weall, 'wall' (Lat. vallum);

vicus).

'town' (Lat.

But here again

it

is

not only possible but extremely

probable that the Gothic and Anglo-Saxon gave only an approxi-

mate representation

of the Latin sound.

Gothic could hardly

have borrowed from the Latin before the fourth century, Anglo-

Saxon not before the

this period Latin v
5.

fifth,

and

it

has been shown above that at

bilabial spirant.

had already become a

Others have cited Claudius's attempted introduction of J

for v as
a.d.),
it is

an indication that

v,

as early as Claudius's
stage.

day (50
Claudius,

had progressed beyond the semivocalic

did not suggest a
it is

urged, while suggesting the employment of a

(z')
t

new

character
for i conu, at

for

u cdnsonans

new character
to
i.

sonans (/).
this time,

Hence
v, it is

claimed that the relation of v to

different

must have been

from that of/

As/ was

a semivowel, a spirant.

claimed, could have been nothing less than

But these conclusions would be valid only upon the

Neither of these
is

assumption that Claudius was a competent phonetic observer,

and was not acting from mere caprice.

assumptions would be safe.
cation that v

Moreover, there
its

no other indi-

had progressed beyond

value as a semivowel

as early as Claudius's day.

6.

It

may be added

in conclusion that the

development of

Latin v was not complete even

when

the sound

had passed from

Later
still

that of a semivowel to that of a bilabial spirant.

(fifth

century a.d.

?)

it

became a

labio-dental spirant (Eng. v),

and with that value passed into the Romance.

NUNCIA TION.
Liquids,
/,

The
17.
its

r.

L seems

to

have been pronounced

differently, according to

position in a word.

No

fewer than three different sounds of

the letter were recognized by Pliny the Elder, as cited by Priscian

(Keil,
ii.

29. 9), viz. 1) an exilis sonus, as in the

second

/ of ilk,

Metellus ; 2) a pinguis sonus, after a consonant or at the end of a

word or

syllable, as in clarus, sol, silva

3) a medius sonus, viz.

when

inital,

as in lectus.

Just what the differences were which

were involved in these three modes of articulation cannot now be

determined.
exilis

Lindsay {Latin Language,

p.

90) thinks that Pliny's

/,

sonus and medius sonus were our normal English

as

is

the

case in the Italian descendants of the Latin words cited by Pliny.

The pinguis
this
e.g.

sonus,

Lindsay suggests, consisted

itself, e.g.
l

in

an

/-glide pre-

ceding or following the /

a lter cParus.

The

basis for
I

view he finds in the Romance development of this

clarus

pingue;
;

becomes

Italian chiaro

fiunien becomes fiume

alter

becomes French
18.

autre.

was

trilled

with the tip of the tongue, as

is

clearly
f.)

described by Terentianus

Maurus
15).
(ix.

(Keil, vi. 332. 238

and

Marius

Victorinus (Keil, vi. 34.

The name
letter.

littera canina,

given to r as early as Lucilius

29, M.), agrees excellently

with the enunciation attributed to the

The
19.

Nasals, m,

n.

M.

Initial

and medial
final

m probably had the sound of normal

m, the true pronunciation can prob-

English m.

As regards

ably never be satisfactorily determined.

When the following word

uttered.
Cf.

began with a vowel,

Quintilian,
ix. 4.

final
'

was only imperfectly

is final

40

When m

and comes
it

in contact with

the

initial

vowel of

the following word so that

written, yet
it is

can pass over to

the latter, though

it is

only slightly uttered, as in

9
,

THE CONSONANTS.
multum
as
it
ille,

quantum
For
it

erat, so as to give the

sound of a new
is

letter,

were.
is

does not absolutely vanish, but

obscured,

and
In

a sort of sign that the two vowels do not

become merged.'

ix. 4.

39 Quintilian

tells

us that Cato the Elder wrote diee for

diem, evidently in recognition of the vanishing value of the final

nasal. Velius

Longus
sound

also tells us (Keil,

vii.

80, 12

ff.)

that Verrius

Flaccus,

who lived under Augustus, proposed a mutilated M, viz. W

of final
p.

to indicate the

m before an initial vowel.

in question as a
'

Seelmann

{Aussprdche des Latein,

356), following the above statement

bilabial nasal
Cf.

of Quintilian, defines the

sound

spirant with partial

closure.''

This seems a just statement.

also Lindsay, Latin Language, p. 62.

Evidently the sound must

did not interfere with the
initial

have been quite inconsiderable, as

slurring of final syllables in
is

it

-m

with a following

vowel, as

abundantly shown

in poetry

by the frequency
ff.,

of elision. Ellis

(Quantitative Pronicnciation of Lati?i, p. 60

especially p. 65)

interprets the testimony of Quintilian above cited to

final

mean

that

was not omitted

{iieque eximitur),

but was inaudible

scholar also
of the

(obscuratur) before an

initial

vowel.

The same
In case this

maintains that every final

initial

was inaudible, irrespective

initial

sound of the following word.

Ellis (pp. 55, 65)
;

sound

was a consonant,
doubled

holds that the consonant was

thinks,
is

in pronunciation
etc.

e.g.

quorum pars, he

was pro-

nounced quoruppars,

This view, however,

its

based on the

improbable assumption that the Italian with

mai), ovvero (for o vero),
tion of Latin final m.
etc.,

gia??imai (for gia

gives the clue to the pronunciait is

Latin inscriptions,

true, in the earliest

times show that final

m was frequently

omitted in writing.

Thus

the Scipio inscriptions, the earliest of which

may antedate 250 B.C.,

show
in

omitted before consonants as well as before vowels, but

of the classical period final
;

good inscriptions
this source.

was not

omitted with any frequency

hence no argument can be drawn

from

20
20. N.
initial,
i.

PRONUNCIATION.
iVwas
the dental nasal, as

was the

labial.

When

n could hardly have differed materially from English n in

situation.

the

same

The same

is

true also of n in the interior of

t,

a word
vowel.

when followed by
; i.e.

other dental sounds (as

d, s, n)

or a

Before the gutturals, n took on the sound of ng in sing,

eg. in ango, uncus

n here became the guttural nasal, a sound

is

as different from dental n as

m, and quite as

much
it

entitled to

representation by a separate character.

Nigidius Figulus recogn-adulterinum

nized the individuality of the sound

(Gellius,
xix. 14.
ii.

in calling

7).

Certain

Roman

writers,

according to

Priscian (Keil,

30. 13), followed the analogy of the Greek,

e.g.

and

used

g (=

y nasal) for the n-adultei'mw?!,

(as distinguished

Agckises, agceps,

aggulus.

The Greek

phoneticians gave y in such situations the

name Agma
sound,
2.
e.g.

from Gamma), and their

Roman

successors sometimes employed the same designation for the

Priscian in the passage just cited.

The vowel

before
37.

nf, ns, as

is

well known, was regularly

in

long in Latin.

See

Some have assumed,

in

consequence,

that a nasal vowel

was pronounced

such cases, particularly

Johannes Schmidt {Zur Geschichte des Indogermanischen Vokalismtis, I. p.

98

ff.).

The

chief basis of this hypothesis was found

e.g.

in the

omission of n before s in inscriptions,

cosol

(for consul)^

cesor, trasitv. Adjectives in -ensimus and adverbs in -tens were

also often written -esimus,
-ies, e.g.

viclsimus or vicensimus
vii.

vlcies

or viciens.

Velius Longus (Keil,

foresia,

78-79)

tells

us that Cicero
as Megalesia,
lost.
a-

pronounced forensia as
transliterations of Latin
e.g. KA.77/A77?

and Megalensia

while in adjectives in -osus the n was permanently

Greek
vo- (vs),

words also frequently show

K^o-wpivos.

for

(Clemens)

But

all

this
s

evidence

may

indicate nothing

more than that n before

was unstable and

inclined

to

disappear.

Whatever conclusion be drawn with

regard to the nasalization of the vowel before ns would seem to

hold also for the vowel before n

when followed by other

dentals,

THE CONSONANTS.
viz.

21

before nt and nd.

For here,
if

too, the

n shows quite as strong

a tendency to disappear,
inscriptions,
e.g.

secvdo

we may judge by the testimony of = secundo) testameto ( = testa;

mento).

No

instance of the disappearance of n before /occurs

prior to the fourth century a.d.,

is

and even then the phenomenon

of extremely rare occurrence, being confined to four instances,

all of

which are

in the

word

inferus.

See the discussion of Seelmann,

pp. 283-290.
3.

Aussprache

des

Latein,

It

should be added that the omission of the nasal occurs

sporadically in case of

m when

followed by labial sounds, as

so also in

Decebris

(=

Decembris);

Capanum (= Campamtin)\

case of n-adulterinum
rant)
;

before gutturals, as iquirant

(=

inqui-

pricipis

(=

principis).

sion

is,

accordingly, a
nasals

The phenomenon under discusgeneral one, and may be stated thus

:

The Latin

(labial),

n (dental), and n-adulterinum

labial, dental,

(guttural), exhibit a

tendency to disappear before

respectively.

and guttural sounds

4.
i.e.

GN.

Many

scholars hold that

-f-

gn was pronounced
for this

as ngn,

as n-adulterinum

n.

The evidence
;

view

lies

mainly
i,

in the fact that e before

gn not infrequently changes

see

to

e.g.

digmts for *deg-nus (from % dec-nus

ular law that this

73).

Now

it is

a reg-

change takes place before

See Brugmann, Grundriss,
2
.,

ng, e.g.

tingd for

*tengo (73)

hence the inference that gn

I

in

such cases was

p.

pronounced ngn.

122
p.

Sommer,
241.

Handbuch der
the
as ngn,

Lateinische7i Laut-

und Formenlehre,
it

But

Roman grammarians nowhere

and
in

attest this pronunciation of

is

gn

view of their silence

doubtful whether the

are sufficiently

considerations urged by

Brugmann and Sommer

weighty to warrant the adoption of their view.

5.

Besides the three nasals already considered (m, n, and np.

adultermum), Seelmann {Aussprache des Latein,

another,

270) recognizes

midway

in

sound between

and

n.

The evidence

22
for the existence

PR ON UN CIA TION.
of
this

sound he
vi.

finds in the statement of

ff.)

Marius Victorinus (Keil,

16.

to the effect that

11)

such
in

an intermediate sound (neither

antiquity.

nor

was recognized
where

Marius Victorinus compares the sound

adfx(3v,

in question

with the sound of the Greek nasal in

likewise,

he observes, neither
nunciation.
as

nor

fi

accurately designates

the pro-

Seelmann suggests that such inscriptional forms

qvamta, tamta, damdvm, semtentiam on the one hand, and

Decenbris, senper, ponpa, inconparabilis on the other, support

by

their vacillating spelling the theory

propounded.

The

facts,

however, do not seem sufficiently clear to warrant a posi-

tive conclusion in this matter.

The
21.
F.

Spirants, /,

s,

h.
it

F
it

is

the labial spirant.

In the earlier period

is

probable that
origin of /,

was
in

bilabial.

This theory accords with the

is

which

most cases

the descendant of an origi-

nal bh

agrees also with such spellings as comflvont, com-

vallem
lish

of the

Minucii inscription,

CIL.

i.

199
it

(122 B.C.).
is

Subsequently

f became

a labio-dental spirant as

in

Eng-

and

in

most modern European languages.

is

At

just

what

time this change took place

uncertain.

It

was complete by
from the

the close of the second century a.d.,

as appears
vi.

testimony of Terentianus Maurus (Keil,

22.
S.

332. 227).

S was

a voiceless dental spirant, like English s in sin.

Some
Latin

scholars have thought that intervocalic s was voiced in

(i.e.

sounded

like

English s in

these),

but there

is

no valid

support for this view, nor do the

hint at

Roman grammarians anywhere

for the letter.
j-

more than a

single

sound

The Gothic
s,

in

loan-words transliterates intervocalic Latin

sented a voiceless sound in Gothic,
e.g.

by

which repre-

Kaisar

(Lat. Caesar).

The Gothic possessed

also a character for the voiced vf-sound

THE CONSONANTS.
{i.e.

23
of
it,

z),

and would undoubtedly have made use

had the

Latin intervocalic s been voiced.

23.
lish h.

H.

H was
As

a guttural spirant

and was voiceless

itself in

like

Eng-

The same

uncertainty manifested
is

the employ-

ment

of initial h, as

noticeable

among
it

the lower classes in

England.

a result of this uncertainty, words etymologically

entitled to initial

h frequently dropped

in the to

speech of the

less

cultivated, while other

words acquired an h

which they were

not historically entitled.

represent the correct

Thus harena, haruspex, hirundo, holus, spelling but these same words were fre;

quently pronounced arena, aruspex,

in that

etc.

and appear repeatedly

Occasionally

form

in

our Mss. of the classical authors.

its initial
is

a word permanently lost

h even in the speech of the

educated.

A
On

case in point

anser,

which comes from an Indo-

Eur. word with initial gh, and should appear in Latin as hanser

(97.

3).

the other

hand

erus, unior, iwierus are the cor-

rect forms, but these

were frequently supplanted by herns, humor,

fully conscious of their defects in

humerus.

The Romans were

and Catullus

this particular,

in his 84th

poem humorously refers

to

one Arrius, who said hinsidias for msidias, and Hio?iios for

Ionios.

Intervocalic h easily vanished between like vowels, as

is

shown

by such contractions

as liemo for *ne-hemo

;

firendd for prehe?ido

praeda for *prae-heda

etc.

THE MUTES.

The
24.
/

Voiceless Mutes,
in

/,

c,

k, q,p.

T.
i

T was

pronounced as

English
is

satin.

In English,
i.e.

before

followed by another vowel

regularly assibilated,
'

acquires an .r/z-sound, as, for example, in the word rational ; but

Latin

was always a pure

/ in the classical period.

(

Cf.

such

Greek

transliterations as OvaXevria

Valentid).

In late imperial

24

PR ONUNCIA TION.
ti

times (not before the fourth century)

when followed by

a vowel

begins to show traces of assibilation.

Inscriptions of this period

;

exhibit such forms as Voconsivs (for Vocontius) septus)

.

sepsies (for

Probably

this

orthography was not exact, as the sound

;

was rather that

of our English sh

but the Latin had no more

:

accurate designation.

The

phonetics of the change are as follows

An

original Vocontius, for example,

i

became

first

Vocontyus,

i.e.

the vowel

(very likely under the influence of extra stress upon

the preceding syllable)

stage this semivowel

become the semivowel

spirant, the
Vocont-chus.

y.

In the next

became a

sound represented by
this,

German

palatal ch, viz.

From

the transition

to the assibilated pronunciation

was easy and

natural.

25.

C.

i.

was always pronounced

like k.

This

is

abun-

dantly proved by the evidence.

a)

Thus

C and k interchange in certain words, e.g.

Caelius, Calendae,

Carthago.
b)

We
:
'

have the express testimony

k, it

of Quintilian

(i. 7.

10),
it

who

says
a,

As regards

should not be used.

'

Some

write

before

but c has the same sound before all vowels.

c)

In Greek transliterations of Latin words we always have

a, o, v,

k,

not only before

but also before

e,

1,

where

if

anywhere we

should have expected the >r-sound of

are
:

c to

have

arisen.

Examples

VLtKepwi/,

Katonp.

d) Gothic and German loan-words borrowed from Latin (probably in the early centuries of the Christian era) show k for Latin
c in all situations, e.g.

Gothic lukarn
;

= Lat. liicerna)
German
Keller

karkara

(=

Lat. career)
;

Kaisar (= Caesar)
cistd).

(=

cella-

riuni)
e)

Kiste

(=

The Old Umbrian

>l

of the Iguvine Tables uses in

for c (an j-like

its

en-

choric alphabet

for c
/).

and d

sound developed

from
is

before

and

The New Umbrian

of the
for
c,

same

tables
(or

written in Latin characters, and uses

but

S)

THE CONSONANTS.
for the
.r-like
it

25'

sound represented

in

Old Umbrian by d.

This

makes

clear that at the time the

New Umbrian

tablets

were

written, Latin c before e

Otherwise the

New

had not yet become assibilated. Umbrian would not have resorted to the use
and
i

of a special character (S' or

S) to designate
1893.

this sound.

See

Jones, Classical Review, No.

1,

The

exact date of the

New

written

Umbrian many years before

tablets
is

not certain, but they can hardly have been

the beginning of the Christian era.

/)
for
c,

No
as

Latin grammarian ever mentions more than one sound

certainly

some one
i.

would have done had

had an j-sound
the change of

before e and

In paradigms like died,

d'teis, dicit,

sound, had

it

occurred, would have been too striking to escape

comment.
g)
tions)

Pulcher (originally pulcer, and often so written in inscrip-

shows by

its

aspirated c

(i.e.

eh) that e

must have been

'hard.'

Similarly anceps, with

its
s.

n-adulterinam, shows that c

could not have had the sound of

not have become guttural, as
2.

Otherwise the nasal would

it

we

are assured
fifth

did.
c

Beginning with the fourth or

century a.d.,

before

followed by a vowel becomes assibilated, exactly as explained

above

in the case of

/.

Inscriptions of this period exhibit such

;

forms as felissiosa (=feliciosa)

Marziae (= Marciae).
c

The
s,

phonetics of the change are precisely analogous to those already

described under
e.g.

/.

Later

still,

every

before e or

became

paze (for pace) in an inscription of the seventh century a.d.

3.

This development of

ti

and

ci

(before vowels) to the

same

sibilant

sound led naturally

in

mediaeval times to the greatest

confusion of orthography in our Mss. of the Latin writers.

condicio appears frequently as conditio
;

Thus
;

suspicio as suspitio

nego-

tinm as negociu?7i

t \

convicium as convitium.

In the case of some

of these words, the false forms

have not yet been entirely elimi-

nated from our texts of the classic writers.

4.

and

are simply superfluous duplicates of

c,

as

was

26
recognized by the
(Keil,
vi.

PR ONUNCIA TION.
Romans
f.).

themselves.

Cf. Terentianus

Maurus

331. 204

26.

P.

P was apparently a plain

English

p and

presents no

peculiarities.

The Voiced Mutes,

27.

b, d, g.

B.

was

like

English b except before

and

t,

where

it

had the sound

of p.

This was simply the result of the natural

Inscriptions
e.g.

assimilation of the voiced sound to the voiceless.

show repeated instances

of

the phonetic spelling,

pleps,

apsens, optinvit, opsides, but ordinarily such words

made a
Quin-

concession to the etymology, and were written with

tilian
(i.

b.

7.

7) prescribes

the use of b:

'When

pronounce

obtinuit our rule of writing requires that the second letter

be

but the ear catches p.

28.

D.

D was like English d.

like
e.g.

Late

in imperial

times

di,

when
dy-)

followed by a vowel,

sound somewhat
peculiarity,

became (through the medium The Romance languages our/.

Italian

of

retain this

French journee,

giorno,

from Latin

diurnus.
Inscriptions

show

that final

d had

a tendency to

become

t,

e.g.

aput, haut, at, quit, for apud, haud, ad, quid.

Mss. also exhibit

the

same
G.

spelling.

29.

G had

the sound of English

in get.

That before

and
a)

i it

did not have the sound of

in gem,

seems clear from the

following evidence

The Roman grammarians

give but a single sound for the

like

letter.

Had g before
in a

been pronounced

our

j,

the altera-

tion

of sounds

paradigm
to

like lego, tegis, or leges, iegum,

would not have

b)

failed

elicit

comment.

In the Greek transliteration of Latin words

;

is

always rep-

resented by y

e.g.

TeXVios (Geilius).

THE CONSONANTS.
30.
'

2J

Distinction between 'Guttural' and 'Palatal.'

'

Guttural

and
'

'

Palatal

'

are
'

not

interchangeable

terms.

Strictly speaking,

Guttural

applies to the c (k)

'

and g sounds

produced

in the throat, while

Palatal

'

applies to those pro-

duced against the hard

guttural

palate.

The

guttural or palatal character

depends upon the following vowel.

^-sound
is
;

Before

a, o,

or u the c or
kill,

before

e or i it is palatal.
;

Cf. English

gill with call, gall.

Latin k (used only before a

;

see

1.3) was,
a,

accordingly, always guttural

the

same was the case with

while
vowel.

and g varied

in character

according to the following

ch, th.
its

The
31.
1.

Aspirates, ph,

The

Latin originally had no aspirates of

own, and

was not concerned with the representation

the

of these sounds until

<j>,

Romans began
c,

to

borrow Greek words containing

explained in the

%,

or

6.

These Greek

letters (as

Grammar, 2.3) were

It is

equivalent to /,

or / with a following /z-sound. 1

not surx>

prising, therefore, that at first the

Romans rendered

<,

by

/,

c,

respectively.
e.g.

This

is

regular in early inscriptions (down to

about 100 b.c),

Corintvs, Delpis, Aciles.

In the Captivi of
. .

Plautus, verse 274, the evident

that the th
little
2.

pun on Thalem,
a
/,

tale?ito,

shows
can be

was

felt as substantially
t is

and

in fact there

doubt that

w hat Plautus
T

actually wrote.
b.

Beginning, however, with about 100

c, Greek
in

<f>,

^, 9

came

to

be represented with increasing frequency

this

Latin by ph,

ch, th,

and by Cicero's day

had become the standard orthography.

The

multitude of Greek words employed in Latin at that time,

along with the constantly increasing attention paid by educated

Romans

to the

Greek language and

to

Greek culture generally,

naturally led to this striving for greater exactness.

1

Initial

and

final/,

c,

and

/,

in stressed syllables, in English are also uttered

with aspiration, though

lock, pot.

we do

not indicate this in writing.

Examples are:

top,

28
3.

PR ON UN CIA TION.
As
a result

we

notice the aspirates gaining a foothold in cere.g.

tain genuine Latin words,

(after

pulcher, originally pulcer; Gracchus

Bacchus

BctK^o?),

originally

Graccus ; Cethegus,
in

origi-

nally Cetegus.

An English
The
s

analogy

is

seen

such words as island,

rhyme.

Island comes from the Anglo-Saxon igland, Middle

English Hand.

was introduced

at a comparatively recent
isle

date as a result of associating

insula).

Hand with French

(from Latin

rime,

'

Rhyme comes from Anglo-Saxon rim, Middle English number.' The spelling rhyme is due to the influence of
pvQp,6i)
,

rhythm (Greek

with which rime was associated in the

tells

folk consciousness.
self,

Cicero {Orator, 48. 160)

how he him-

in deference to

popular usage, was forced to abandon the

pronunciation pulcer, triumpds, Cetegus, Kartagd, in favor of the

aspirated forms, pulcher, triumphds,
etc.

But he adds that he

refused to pronounce an aspirate in sepulcrum, corona, lacrima,

and some other words, where apparently a popular tendency

existed in favor of ch, ph,
th,

as against the genuine Latin p,

c, t.

Catullus, in the epigram already cited

(Carmen

84),

humorously

alludes to Arrius's pronunciation of conwioda as chonunoda.

In Bosphorus (BoWopos) the

for a tenuis
4.
;

Romans

introduced an aspirate

yet the spelling Bosporits also occurs.

of ph the Latin aspirates retained their

With the exception

still

original

character throughout the history of the language.

A
is

proof that th was

an aspirate

in the time of the

Empire

seen in the spelling acletarvm for athletaru??i, and aclheticvm

for athleticu7?i, in an inscription of about

360
(

a.d.

(Wilmanns, No.
;

2639).

Cf. also

CIL.

viii.

5352,

termas

Inscriptiones Hispaniae Chris tianae,

= thermas) Huebner, i42,aetereas = aethereas)',

(
\.

and the variant Chyesten

This orthography
that th
of c
is

for Thyesten in Horace, Odes,

16. 17.

capable of explanation only on the ground

/ {viz. t
-+-

was
t,

still
cf.

very close to

h).

For the confusion

and

the occasional English pronunciation of at least as

is

ac

least.

There

not the slightest indication that Latin

th,

either

THE CONSONANTS.
in the flourishing period of the

29
its

language or in

decline,

had a

spirant sound

like

our English
t as
;

th in this or thin.

The Romance
th, e.g.

languages regularly have

the descendant of Latin

Italian teatro (Latin theatruni)

cattolico (catho/icus).

Similarly ch
else the sim-

must have always been either a genuine aspirate or

ple

mute

c,

as

shown by the

Italian in such

words as carta

(Lat.

charta), coro (Lat. chorus).

5.

As regards ph,

the aspirate seems in late imperial times (not

before the fourth century a.d.) to have developed into the spirant
/.

Some have thought

that this change occurred

<f>,

much

earlier,

basing their opinion upon the fact that Greek

larly represented in Latin
literate

which was reguto trans-

by ph, was always employed

for representing/.

Latin/

But

<f>

was simply the nearest equivalent that the

Quintilian
(i. 4.

Greek alphabet possessed

14)

shows that the two sounds were quite

the Latin word Fundanius.

different,

by

his account of

the Greek witness mentioned by Cicero

who could

not pronounce

This seems to show that the Greeks,

not having the sound of Latin

/(a

bilabial spirant), chose

<j>

(a

bilabial aspirate) as the nearest equivalent, very

much

as Slavs

and Lithuanians to-day reproduce the

of

modern languages

by pn
In the speech of the educated classes at Rome, ph seems to

have followed the history of $

in

Greek.

The

latter

sound,
did not

according to Blass (Pronunciation of Greek,

28),

become the equivalent of/ before the

tions of this

third century a.d., a view

substantiated for Latin by the interchange

and the following

:

centuries.

of/ and ph in inscripThe phonetics of the

i.e.

change are as follows

-f

First,
;

we have/

+ h,
is
;

the labial

mute

a guttural spirant

secondly, the h

assimilated from the

finally,

guttural spirant to the labial,

lated

/ (i.e. pf)

the

is

assimiorigi-

to/

giving^",

which

is

then simplified

to/

Thus an

nal Philippus becomes successively Pfilippus, Ffilippus, Filippus.

Cf.

German Pfah

(the

name

of the district about Heidelberg).

30

PRONUNCIA TION.
Palatiiun,

The mediaeval Latin designation of this was Phalatium, German Pfalz, but dialectically
Falz.

whence

often

pronounced

The Double Consonants,

32. X.

x,

z.

is

always equivalent to

cs,

never to

gz, as

it

some-

times

is

in English.
s,

This conclusion follows from the voiceless

before which a guttural was necessarily

character of Latin
assimilated.
33. Z.
is

The

value of z

is

somewhat uncertain.

The

character

confined exclusively to foreign words, chiefly Greek.

first

Though

introduced in the

Latin alphabet,

it

was early dropped (see

1.3),

its

place being taken by g.

Long

afterwards,

parently about Cicero's time,

ap

it

was again introduced

for the

more accurate transcription

Greek.

of in

words borrowed from the

Prior to this time the Latin

had

transliterated

Greek
e.g.

when
of

initial
;

by

s,

and by

ss in

the interior of words,

.

sona

(== &0V9)

atticissd

(=

olttlkl^o))

But with the increasing use

Rome, a more accurate designation of the sound was felt to be necessary, and accordingly the Greek character itself was introduced. Cf. the care exercised at the same period in
Greek
at

designating the aspirate in Greek loan-words.

The pronunciation
ciation of

of z in Latin

must have followed the pronun-

Greek

for the

corresponding period.

As regards

while
fifth

it

almost certainly had the sound of zd in the Attic of the

B.C., it is likely

by the beginning of the Macedonian period (approximately 300 B.C.) it had become a simple though probably somewhat proz sound (as in English gaze),
century
that

longed; for

it still

'made

position,' as

though a double consonant.

31.

See Blass, Pronunciation of Greek,

bly attached to

The same sound proba-

Roman z. For

while certain
ds, their

Roman grammarians
z.

explain z as equivalent to sd or

statements are probably

It

but the echo of Greek discussions concerning the sound of

is

worthy of note that one

Roman grammarian,

Velius Longus, a

DIVISION OF WORDS INTO SYILABLES.

most competent witness on phonetic questions,
that z
is

specifically denies
it is

the equivalent of sd, and asserts that

all,

not a double
(Keil,

consonant at
vii.

but has the same quality throughout.

50. 9.)

Doubled Consonants.
34.

When

the mutes were doubled

(//,

dd; pp, bb

; cc,

gg) there

were two
/

distinct

consonant articulations.
definite

Thus

in initio, the first

was uttered with a

muscular
;

effort,

involving closure of

the organs in the /-position

then after a

momentary pause a
same
Such
tt,

second muscular effort followed, with the organs in the

position.

See Seelmann, Anssprache des Latein,

but pronounce only a single

p.

no.
e.g.

doubled consonants do not occur in English.

pp,
cc, etc.,
t,

We

often write
c,

p, or

ut(f)er,

up(p)er,

etc.

But

in Italian

and several other modern languages

e.g.

these doubled consonants are frequent,

cappello.

Italian bocca, conobbi,

The same double

of doubled liquids

articulation
rr),

is

probably to be assumed in case

nri),

(//,

doubled nasals (mm,

possible that in

and doubled

spirants

(ff, ss),

though

it is

some words where

these combinations followed a long vowel they merely indicated a

liquid or spirant that

was prolonged

in utterance, as, for example,

vallum, uttus.

Division of
35. of

Words into Syllables.

The
of

principles given in the

Grammar ( 4) for the division

;

words into syllables are the traditional ones

yet the validity of

some

them

is

open
4.

to question,

particularly of the principle

of consonants as

embodied under

'
:

Such combinations

can begin a word are joined

port of this principle

to the following vowel.'

In sup-

may be

cited the testimony of the

Roman

grammarians, who practically agree

above, and some of

in prescribing the rule given

whom

even include such combinations of

32

PR ONUN CIA TION.

e.g. pt, ct,

consonants as can begin a word in Greek,

for instance Caesellius, cited

bd.

See
;

by Cassiodorus
351. 879).

(Keil,

vii.

205. 1)

Terentianus Maurus (Keil,

vi.

On

the other

hand

it

may be urged that the

is

principle laid

down

by the Roman grammarians

merely an echo of rules maintained

language.
Cf., for

by Greek scholars

for their
iii.

own
p.

example,

Bekker, Anecdota Graeca,

p. 63,

1127; Theodosius (ed. Gottling),

where the same laws for syllable division may be found.

such irresponsible borrowing

We have already seen indications of

in the case of the testimony of the

grammarians concerning the

Moreover, we find Quintilian
of division,
e.g.

pronunciation of
(i.

0.

See

^^.

7.

9) advocating

an etymological principle

haru-spex, abs-temius.

When we come
lowed

to

examine the mode of dividing words

fol-

in our best Latin inscriptions, the evidence is strikingly at

variance with the traditional rule which prescribes joining as

many consonants
80 per cent of
of a line,
all

as possible with the following vowel.

In about

the cases in which words are divided at the end

is

one of the consonants

joined with the preceding

vowel,
rule.

evidently
Even greater

a systematic violation of the grammarians'

is

the proportion of violations of the rule in

i.e.

those words which exhibit interpunctuation in inscriptions,

separation of the syllables by dots,

e.g.

eges-tas; vic-to-ri;

op ta
Vol.

'TVS.

For a

full

presentation of the epigraphic evidence

bearing upon this point, see Dennison, in Classical Philology,

I. p.

47
is

f.

There
question.

also evidence of a phonetic nature bearing

upon

this

A
is

syllable containing a short vowel followed

by two
But
in such

consonants

phonetically long, as recognized by

all

our gram-

mars and demonstrated

words as

in every line of Latin poetry.

;

open syllables containing a short vowel are short

doctus, minister, hospes,
{i.e.

and

if

we

divide according to the

grammarians' rule

do-etus, minister, ho-spes),

we

get pre-

DIVISION OF WORDS INTO SYLLABLES.

cisely these
syllables.

33
i.e.

open

syllables containing a short vowel,

this utterance, there
is

short

For with

no more reason why

?ii-

the do- of dd-ctus should be long than the do- of do-cet; or the
of minister

any more than the

clear that the

ni- of

mim-mus. In both cases we

i.e.

have open syllables containing a short vowel,

short syllables.

Hence

it is

Romans
{i.e.

in actual utterance

must have

joined one of a group of consonants to a preceding short vowel.

This gives a closed syllable

a syllable ending in a consonant),

and
is

it is

a fundamental phonetic principle that a closed syllable

long.

These principles

also

throw light on the nature of comis

mon

syllables.

A common

syllable
/ or

one containing a short

t/,

vowel followed by a mute with

In verse such a syllable

r {pi, d,

pr, cr, tr ;

etc.).

may be

either long or short.

But natuthis

rally a difference of pronunciation

must have accompanied

variation of quantity.

In a word

like

patron, for example, when

/

the

first

syllable

was used as long the

;

was joined with the a

first

{pdt-reni), thus closing the syllable

but when the

syllable

was used

as short, the /

was joined with the r (pd-trem), thus

found also
in

leaving the syllable open.

Evidence contradicting the grammarians' rule

the division of words in examples cited

is

by ancient writers on

Latin prosody.
into its to the

When
feet,

these writers separate a verse of poetry

component

they divide the syllables not according

grammarians'

rule,

but according to the principle exconsiderations,

te

plained above as

demanded by phonetic
om nes
in tenti

e.g.

Conticu ere

que ora

nebant

Turnus ut infractos ad verso Marte Latinos

Ut

bel

li

sig

num Lau renti Turnus

ab arce.
p. 268.

See especially Hale, Har~oai'd Studies, Vol. VII.

The

rule of the
It is

grammarians, therefore, seems thoroughly

contradicted by the testimony of inscriptions,

discredited.

by considerations

of phonetics,

and by

syllabification followed

in

wnsb] the writers on

is

prosody.

It

should

res

unpetent phonetic
rather
set

-.

speed]
rules

upon the
up

tra-

ditional

which
t\

die
es

Greek grammarians
."

for their

the

way, which were

Latin.

no

more
like

Greek that
their
pti

Very

....
is

is

t'.-.

med
.

b] the ancients

them3

selves

more probable ths

it
:
.

ere s

copyists

rules intended to furnish

stands
S]

practical use

sionof

Abies

where two
folio
r
s

consonants are involved

s

may be

formulated as

In case

ch combinations of consonants, a mute 4- /or

.,

is is

joined to the folk

except
cas
e

when a long
is

syllable

needed, in which latu

the

mute

joined to the precee ing

;-i,-but
/

eL Thus regularly Jw~iris

poetry the
-

rv/a-

-hen in

"able is
first

used as long.

In prepositional comin a mute,

pounds, also, wht^:

member ends
the mute
5

and whose
t

tnd begins with /or

/ ; :

r,

is

always joined to the pre:.

the t tetei " ^

it Le is t

ys

tit..
- :

<mjv.

In

all

other combinations of consonants, the

a*-gt>,

consonant

is

joined to the preceding vowel, as a/- fits.

kns-fes, jSc-^ms, mi*i$-&iy

m^-M*ss M?M-sfr*m.
^httiii he

This principle

;r
s

;us

hett-iris thi:

i:~ie-d in tt: it t::i.:::i

was undoubtedly the case.

Thus x& mist have been

~v T

pro-

nounced m
,

:~xms as Igf-sms; so, also, very likely after a

.

thrush

is

:rvi;^s thai

I,:
the J
re

..:'.i

sitt

is

:: t. i;

tei .:.i..~"

T.ririr

As regards
1

die rule of the ancient grammarians laid

_

down

in

-;

_
:

::
:

:i: etfett thttrtepisitiiiiil ::=it:-tis

:

set i

met
.

if~

1 1 ns

tie it: nen :

e~ ten ;
-

- c

:he zexi Letter

::
1

the ttziptiztiii t
lltS

::

:1

.Il_

:m:
-

:.

r; zyzza^ies

-.:

_l:

t
^
; :e

'-.-'

rxs.-^T

^
=

:i
:

:f

ii

izr.i =:
-:

-i- ~trt
-

-.:

--

- ::
2

^
i;
;

--"7

:t ::r ill =.----.-

c iiiii.ir:
.

rz:t~T i= iIjti:
-

~i:
------

-;uz.i=

:i

:;;

:;_-

CHAPTER
A

III.

HIDDEN QUANTITY.
36.

hidden quantity

is

the quantity of a vowel before two

is

consonants.

Such a quantity
of a

called hidden, as distinguished

from the quantity

the metrical

vowel before a single consonant, where

of the

employment

word

at

once indicates whether

the vowel

is

long or short.
is

The

quantity of a vowel before a

it

mute with /or r

in verse used

hidden unless the syllable containing

appear

as short.
of determining

The methods
ing
i.
x

hidden quantity are the follow-

Express testimony of ancient

159,
is

Roman
;

writers,

e.g.

Cicero,

Orator, 48.

where the principle

for the length of vowels

before
cae,
ii.

nf,

ns

laid

down
ix.

(see 37)
xii. 3.

Aulus Gellius, Nodes Atti-

17;

iv.

17;

6;

Nearly every

Roman gram-

marian furnishes some

little

testimony of this kind, and though

some

of

them belong

to a

comparatively late period, their eviearlier usage,

dence often preserves the tradition of

entitled to weight.
/ 2.

and hence

is

The

versification

of the earlier

Roman
a
short.

dramatists, especially
liquid never

'Plautus

and Terence, with

whom
is

mute before a

lengthens a syllable whose vowel

Hence, before a mute

followed by a liquid, the quantity of the vowel always appears in

these writers, being the

same

as the quantity of the syllable, just

as in case of a vowel followed by a single consonant.

Furthermore, Plautus and Terence not infrequently employ as

short
1

many syllables which

in classical poetry
is

would be invariably

The

material here presented

based chiefly upon Marx's Hulfsbiichlein,

cited below, p. 39.

36

METHODS OF DETERMINING HIDDEN QUANTITY.

long by position.
:

37

Examples are the following juvenilis, Plautus, Mostettaria 30; Curculio 38; voluntas, Trinummus 1166 PseuAmphivoliiptas, Mostettaria 249, 294 Stichus 59 dolus 537
;
; ; ;

truo 939, and elsewhere.

fact that the
to
'

These cases are

to

be explained by the

vowel was short and the following consonants failed

make position.' In some instances,

e.g.

it

must be confessed, even long vowels are

mis,

used as short,
pultabo, 368.

bonis

Plautus,

But these cases are of a

the examples cited.

Trinummus 822, forts peculiar sort and may

be explained on metrical grounds, or by the iambic nature of

the
\yx>rds, as in

Cf. 87. 3.
first

'$. Inscriptions.

Since

the middle of the

century
e, 0,

B.C.

the apex (or point) appears added to the vowels a,

dicate their length.

u to

in-

Long i was designated

and hence called
i

originally

by / (rising
ei ; later,

above the other

1

letters

longa)

and by
2311
;

took the apex.

xi.

Examples are traxi, CIL.

vi.

x.

priscvs,
vi.
i.

CIL.

1940; olla, CIL.

10006; quInqve, CIL.

i.

3539;

mIllia,

Monumentum Ancyranum,
e,

16

fecei, CIL.

551,

Before the employment of the apex the length of the vowel in

case of a,

u was indicated by doubling the vowel,

;

e.g.

paastores,
in this

CIL.

i.

551

peqvlatvv, CIL.

i.

202

is

never doubled

manner.

This peculiarity belongs to the period from 130 to 70 B.C.

thoroughly consistent use of these methods of designating

is

the vowel quantities

inscriptions.

found,

it

must be admitted,
rule, in

in but

few

Of the vowels contained

marked, as a
official inscriptions of

in syllables long

by

posi-

tion only a portion are

tion.

any single

inscrip-

Certain

the late republican and early

this,

imperial period form an exception to

reliable markings,
e.g.

and exhibit very

full

and

the speech of the

p.

Emperor Claudius

(Bois-

sieu, Inscriptions de

Lyon,

136) and the

Monumentum
This

Ancyra-

num, containing the Res Gestae Dlvl August!.

a great
e.g.

latter,

among

number

of correct markings, contains also

some

false ones,

clvpei, svmma.

Such errors

also occur occasionally elsewhere.

38
4.

HIDDEN QUANTITY.
Greek transcriptions of Latin words.
vowels
e

This method
0.

is

most

fruitfully applied in case of the

and

The employment
of the Latin

of

Greek

or

rj,

o or

w makes the quantity

vowel

certain,

wherever faith may be reposed

in the

accuracy of the

transcription.

Thus we may
;

write Esquiliae in view of 'HctkvAivos,

;

Strabo, v. 234, 237

OveaovTLCDv,

Vergilius, after OvepyiAios

lxviii.

Vesontio, after

Dio Cassius,

24.

The quantity of i may also often be determined by Greek transThus et before two consonants regularly points to literations.
Latin Latin
J, e.g.
z,

Beaj/avios,

CIG. 5709,

Vipsanius

Greek

points to

e.g.

"larpos

= Ister.
much
greater weight in such mat-

Inscriptions are naturally of

ters
5.

than are our texts of the Greek writers.

The vocalism of the particularly the Spanish and

be remembered that Latin
open.

Romance languages. These languages,

Italian, treated
e, i, 0,

Cf

3. c).

u with great
It will

regularity according to the natural length of the vowel.

~~e

and

were close

Latin e and o

Now

the

Romance languages have

;

not preserved the

original quantity of Latin vowels

for both the long

and the short

;

vowels of the Latin have become half-long in

have very faithfully preserved their quality.

as a close e in Italian

Romance but they Thus Latin e appears

an open
;

and Spanish

Latin

e as

or as

ie.

Latin o appears as a close o in Italian and Spanish

Latin o as an

open

or as uo (ue).
;

Similarly Latin

~i

remained
close

/,

but

became
:

a close e

Latin u remained u, but u

Latin.
mensis.
honestus.

became

0.

Examples

Italian.

mese (with close

onesto (with

e).

open

e)

monstrare.
doctus.
dixT.
dictus.
dilxl.
'

mostrare (with close

dotto
dissi.

0).

(with open 6).

detto

(with close

e).

-dussi.
-dotto (with close 0).

ductus.

METHODS OF DETERMINING HIDDEN QUANTITY.

The Romance

39

languages, however, authorize conclusions only

with reference to the popular language as opposed to that of the

better educated classes.

In the popular speech the tendency was

rather toward the shortening of long vowels than toward the

lengthening of short ones.

Hence where

the

Romance languages
it is

point to a long vowel in the popular language,

that the vowel

safe to

assume
the

was long

in the literary language.

When, on
if

other hand, the

Romance languages

point to a short vowel, this

testimony

is

not necessarily conclusive, particularly

other facts

point clearly in the opposite direction.

Again, the

Romance languages

authorize conclusions only in

case of words inherited from the Latin.

Many Romance words

All

represent mediaeval borrowing by the learned class, as Italian

rigido,
cibo,

metro,

tenebre, pustula,

lubrico.
it is

such words
decide

retain the Latin vocalism.

In some cases

difficult to

whether a word has descended by the popular or the learned

channel,
e.g.

luxus, arna.

by the methods above enumerated, there nevertheless remain some words whose vowel
With
all

the assistance furnished

quantity cannot be determined.

It is

customary to regard

all

such vowels as short, until they are proved to be long.

The

following are the most important works of reference on

this subject

Marx, Hulfsbuchlein fur

langen Silben.

die Aussprache lateinischer Vokale in PositionsBerlin, 1901.

3d ed.

work valuable
its

for its collection

of evidence, but frequently untrustworthy in

conclusions.
p.

Seelmann, Die Aussprache

Grober,

des latein.

Heilbronn, 1885.

69

ff.

Vulg'drlateinische Substrata

Romanischer Worter, a
2d ed.
p.

series of articles

in Wolfflin's

Archiv fur Lateinische lexikographie,

Oxford, 1894.

vols. i-vi.

Korting, Lateinisch-Romanisches Worterbuch.

Lindsay, The latin Language.
d'Ovidio, in Grober's
1888,
i.

Paderborn,
ff.

90 1.

133

Grundriss der Romanischen

Philologie.

Strassburg,

p.

497

ff.

Meyer-Lubke, Grammaiik der Romanischen Sprachen.

Leipzig, 1890.

40

HIDDEN QUANTITY.
et

Christiansen, De Apicibus

I longis.

Husum,

1889.

Eckinger,

Orthographie Lateinischer Worter in Griechischen Inschriften.

Munich, 1891.

Heraeus,

Beitr'dge

zur Bestimmung der Qtmntitat in Positionslangen Silben

xiv. pp.

in Wolfflin's

Archiv fur lateinische Lexikographie, Vol.

393

ff.;

449

ff.

Further literature up to 190 1

is

cited

by Marx,

p.

xiv

f.

G ENTERAL PRINCIPLES FOR THE DETERMINATION OF HIDDEN QUANTITY.

Vowels before
37.
ns,

nf
e.g.

vowel

is

always long before ns and nf,

consul, in:

felix.

This principle rests upon the following evidence

a) Cicero, Orator, 159, expressly states that in

con

compounds

of

and

in,

the vowel was pronounced long

when followed by
an
;

/or s.
o)

Before ns the vowel

xii.

is

often

marked

in inscriptions with
vi.

apex, as CIL.

3102 Censor; CIL.

;

1527

d.

64 consto

CIL.
e.g.

xi. 1 1

18

mensvm
1 1

the apex occurs less frequently before nf,

CIL.

xi.

18 conficivnt.
e.g.

But
iii.

longa occurs repeatedly

;

before both ns and nf,

CIL.

67 Inspexi
;

vi.

647 Instrvx;

ervnt; CIL.
CIL.
c)

ii.

4510 Inferioris

CIL.

xiv.

1738 Infanti

4294 Inferri. Greek transliterations

x.

of

Latin words often indicate a

long vowel before

(

ns,

as

Kprjo-K^vs

(=

Crescens)

II/ooi;Sr;vs

= Prudens)
Vowels before
38.

gn, gm.
all

Until recently the doctrine was current that

vowels

are long in Latin before gn.

In the Appendix to

my Latin Gram-

mar, I showed that this general principle was altogether too sweeping and that at most

we could go no

farther than to recognize

with Priscian the length of the vowel before the suffixes -gnus,
-gna,

-gnum and

in

such other individual words as

may be

sup-

VOWELS BEFORE
ported by specific evidence.

-GN, -GM.

Admitting the validity of Priscian's

testimony for the length of the vowel before -gnus, -gna, -gnum,
I

showed

that there

was certainly no evidence

to support the
is

doctrine of

Marx

(see his Hulfsbilchlein, p. 1) that the vowel

long before gn in gignd, agnbscd, agnatus, cognbsco, cognatus,

ignarus, ignavus, igndrb, ignoscb,
in these latter
etc.

Marx

holds that the vowel

forms was long as the result of compensatory

lengthening, ignarus being for *in-gnarus, cogndscd for *con-gnosco.

But no such theory of compensatory lengthening

Marx's appeal
syllable before
(p.

is

tenable.

1)

to the fact that Plautus always uses the

gn as

long,
'

is

of

no weight, since we should

'

naturally expect
larly

gn

to

make

position

in Latin just as yv regu-

does in Greek.
in recent years to re-

But there has been a growing tendency

ject

even Priscian's testimony in favor of the length of the vowel

before the suffixes -gnus, -gna, -gnum.

Keil, Vol.
ii.

The passage
vel
"

is

found in

p.

82

"

Gnus" quoque
mallgnus"
while

gna'

1 1

'

vel

"gnum

"

terminantia
"

longam habent vocalem paenultimam, ut " regnum,"

" benignus" "
" abiegnus,"

sfagnum"

"privignus"

" Pelignus-."

Some

scholars, as Havet, regard this statement as

an interpolation.
of the passage,

Others,
its

admitting

the genuineness
(Classical Review,

impugn
ff.)

correctness.

Buck

Vol. xv. p. 311,

has discussed the question here at issue with

great thoroughness and candor, and urges (p. 312) against the

long vowel before -gnus, -gna, -gnum

in

" (1) the fact that, except

words with an original long vowel, the Romance languages

point to a short vowel before

gn

(2) the fact that the Celtic

and

Germanic words borrowed from Latin signum

vowel
;

also point to a short

(3) the total

absence on inscriptions of the apex or Ilonga

in the case of the great majority of

like

words with gn, some of them,

this

magnus, of so frequent occurrence that

;

absence can

hardly be accidental

(4) the citation of dignitas as

i.

an anapaest

by Diomedes

(Keil, Vol.

p. 470),

who has

in

mind only vowel

42

HIDDEN QUANTITY.
Nevertheless certain words
localities,

quantity, not syllabic quantity. "

of this class

seem occasionally

(in special

perhaps,
as

or in special social strata) to have

had a long vowel before gn,

seen in sIgnum,CIL.
vi.

vi.

10234; SEiGNVM,xiv. 4270; sIgnificabo,

16664

826.

vi. 3541 Ignis, xi. But these Buck regards as abnormal and exceptional pro;

dIgni,

x.

5676

privIgno,

nunciations.

Buck's argument

is

a very strong one, and his conIt

clusions deserve at least provisional acceptance.

should be

noted, however, that three words,

regnum, stagnum, abiegnus,

being derived from stems with a long vowel, were legitimately

entitled to their long quantity

and always retained

it.

39.

Before

gm

the vowel

is

long in fiigmentum (see CIL.

{cf.

viii.

1344, pIgmen[t)

and

in

segmentiMn

Greek

or^y/xcVra),

but

there

is

no evidence warranting the formulation of a broad rule

all

embracing

vowels before gm, as

is

done by Marx

(p.

1).

Marx

appeals to the analogy of gn in support of his attitude;

but apart from the dangers of this kind of reasoning, we have

already seen that the case for vowel length before gn
is

of the

weakest possible kind, so

the analogy, there
is

that,

even

if

we admit

the validity of

nothing to indicate regular vowel length

before gm.

Vowels before
40.
1.

nt,

nd,

ss.

All vowels are regularly short before nt

and nd,

e.g.

amandus, montis, amant, monent.

2.

Exceptions:
a) Before nt the
a) quintus.
ft)

vowel

is

long in

the following contracted words

contio (for coventio),

j'entacuhim (for *jyuntaculuni), j~entatid (for *Jejuntatio), nuntius (for *noventius).

y) Greek proper

names

in -us,

Gen.

-untis, e.g.

Se/mus,

Selinuntis (Greek, ^eAivowros).

VOWELS BEFORE
8)

-NT, -ND.
Gen.
-otitis,

43
e.g.

Greek proper names

in -an.

Xeno-

phon, Xenophontis (Greek, Hevo^wvros)

b)

Before

nd the vowel

is

long in

a) the following contracts

and compounds

prendo

(for

prehendo), nondum

{71011 -f-

duni) , vendo (venumd~o),

nundinus (noveni
cini {Tunis).
ft)

dies), quindecim (qidnque), unde-

some Greek names,

(-covSas).

e.g.

Charondas,

Ep annuo nd as
lies in

3.

The evidence

for the short

vowel before nt

the fact

that, while in the

cliens,

Nominatives of such words as

cl~eniens, crescens,

fons, gens, parens, pons, praesens, the long quantity of the

is

vowel
vowel

assured either by the presence of the apex, or by a long

in

Greek
and
(i.e.

transcriptions, in the oblique cases the apex

is

lacking,
K\rj/ir)s

in

Greek transcriptions the vowel

CIA.
iii.

is

short, e.g.

K\r}/jL7]vs),

1094,

but
;

KXrj/xevros,

CIG.

CIG. Addenda, 1829 c. crescens, CIL. xii. 4030, but crescenti, CIL. vi. 9059; Kp-rjcrKrjvs, CIG. 6012, c. but KprjaKevTL, CIG. Addenda, 1994, f. Ilpatcr^s (i.e. Upaicrrjvs;),
3757
;

KXtJ/jlcvti,

CIA.

iii.

147, but Upataevrt, Upataevra,

Even where

a vowel

is

CIG. 3175, 3991. naturally long, it sometimes becomes

; cf.

shortened before

nt, e.g. in

linteum from linum

Greek Xivnov,

CIG. 8695. For the vowel before nd the evidence

the

is

not so

full.
iv.

We

find
;

Greek transcriptions

KaAeVSais, Lydus, de Mens.

53, 57

<ovSavtos (i.e.
ix. p.

Fundanius), Bulletin de Correspondance Hellenique,

439.

4.

Vowels are

also regularly short before ss, according to the

i.

express testimony of Quintilian,

7.

20.

But see

47.

1.

PONTEM, FONTEM, MONTEM, FRONTEM, FRONDEM.

41.

slight uncertainty

exists

as

to

the quantity of

the

vowel before nt in the oblique cases oi/ons, mons, pons, from

44
(frontis)
;

HIDDEN QUANTITY.
and before nd
:

in frons (frondis).

Three

sets of facts

are to be considered
a)

The analogy

of other words in -ns (Gen.

-ntis).

Such words,

so far as they are

genuine Latin words, have, without exception,

See

a short vowel before nt in the oblique cases.

b)

40.
is

The testimony

of the

Romance

languages.
:

This

as fol-

lows for the different words under discussion

fons.

The Romance languages seem to point to an antecedent Thus the Italian fonte has close o so the Profo?itis,fonti, etc.
;

vencal fon.

Spanish
ii.

alone with

its

fuente points to fbntem

(Grober, Archiv,
buch).
frons (-ndis).
to

p.

426

Korting, Lat.-Romanisches Wdrter-

The Romance languages

ii.

all

agree in pointing

frondem (Grober, Archiv,

frons (-ntis).

p.

426

Korting, Wdrterbuch).
0,

Provencal fro?i and Italian fronte, with close

point to frontem.

So the other Romance languages, except

to frontem.

Spanish,
Archiv,
ii.

which has fruente, pointing

p.

(Grober,

426

Korting, Wdrterbuch^)
point unanimously to montem

mons.

The Romance languages

ii.

(Grober, Archiv,
pons.

p.

426

Korting, Wdrterbuch).
point to

Provencal pon and Italian potite with close

pontem ; so the other Romance languages, except Spanish, which

has puente, pointing to pontem.
If

mere numerical preponderance were

all

decisive,

we might

at

once conclude that

these words went back to Latin forms

with o in the oblique cases, and might explain Spanish fruente,

fuente, puente (which should be fronte, fonte, ponte, to represent

Latin

o)

as exceptions to the prevailing law of development.

glance at certain facts, however, in Italian and Provengal,

suggests another conclusion.

We

find

it

to

be a regular law
(i.e.

in

these languages that an original open Latin

36. 5),
/,

short

o,

see

when followed by m, n, or -f another consonant, becomes close. Thus Latin tondet with open 0, becomes Italian

VOWELS BEFORE
tonde,

-NT, -ND.

45
risponde ;

with close

o.

Similarly

respondet becomes
all

rhombus becomes rombo ; pol(y)pus becomes polpo,

0.

with close

Just what has brought about this change

in

is

not certain.

D'Ovidio
i.

Grober's Grundriss der Rotnanischen Philologie,

it

p.

522, thinks

was the analogy

of

words

in on

+ consonant,
had dee.g.

om
(

consonant,

and

ol-\-

consonant, in which close

veloped regularly from

an earlier u (see
;

36. 5),

rompe

= rumpii)

onda {= undo)

dolce

(=

dulcis).

In accordance

with this principle, whose operation

is

certain, Latin

fontem,

frondem, frontem,

mo litem, pontem, would (assuming

become
o,

these to be

the original forms) regularly

in Italian

fonte, fronde,

fronte, monte, ponte, with close

exactly as

we

find them.

The
words

admission of a long
is,

in the oblique cases of these Latin

therefore, not necessary in order to account for Italian

and

Provencal close o in their

Romance descendants. In
all of

fact,

when
to

we consider Spanish fuente,fruente, puente,

Latin
o, it

which point

seems more reasonable to regard Spanish monte and fronde (which point to 0) as the exceptions. Grober, who {Archiv,
vi.

p.

389) expresses himself in favor of assuming an original

etc.,

fontem,

in these

words, suggests that Spanish monte, fronde,

,

are loan-words, while fuente fruente , puente represent an original

inheritance.
Briefly, then,

a fair interpretation
to

of

the

evidence of the

Romance languages seems

warrant the belief that the oblique

cases of the words under discussion

came

into the
;

Romance

lan-

guages from the Latin with a (short) open

Provencal this open

that in Italian

and

subsequently became close in accordance

with a regular law of wide operation. Spanish regularly developed

the open
(viz. in

to ue in those

words which
;

it

inherited from Latin

fuente, fruente, puente)

while Spanish monte and fronde

are probably loan-words from Italian.

c)

The third

bit of

evidence comes from Greek transliterations

in

of Latin

words as found

Greek inscriptions and Greek authors.

46
Thus we
in

HIDDEN QUANTITY.
find (bovrrjios
(

=
iii.

Fonteius) in Plutarch and

Appian

also
iii.

an inscription,

CIG.

5837, b (59
;

a.d.)

<poimvos,

CIA.

Qpoi/Ttlvos, CIA. iii. 1177 1 154 (between 150 and 200 a.d.) (about 220 a.d.) QpovTw, CIA. iii. 1113, 21, 26 (before 161
;

a.d.),

and

in texts; all of

which point

to Latin

Fran to, and

Fronfinus, and indirectly to front-em.

as MovTavos,
Ptol.
iii.

Latin Montanus appears

;

CIG. Addenda, 4805 b

passim
9
;
;

and we

find rpi^ovriov,
is

11, 12, et

irovrep,

(=

Latin pdnteni)

the text

in Plutarch,

Numa,

7tovti<i

(=

pontifex), in Dionysius,
iii.

Dio

Cassius,
<f>u<es,

and Zosimus

7rovri<e|, in

Lydus, de Mens.
irovTifyiKa, in

21

wovtl-

in Plutarch,

Nnma,
in

and

an inscription in
a.

Kaibel's Sylloge

Epigrammatum, Addenda, 888

any

The Gieek
is

never shows an w
in

of these words, either in inscriptions or

Mss.

The evidence

furnished by that language therefore

unanimous

in favor of o for the Latin.

Nor can
all

recognition be

refused the inscriptions above cited on the ground that they are
late.

As

the annexed dates show, they

belong to the good

period of the language.

We

thus have the strongest possible grounds for writing fontis,

etc.

frondis,

favors this

The analogy of other words in -ns (Gen. -ntis) view the Romance languages favor it, and the testi;

mony
and

of Latin

words
apex
2915,

in

Greek
it.

dress, as exhibited both in texts

in inscriptions, favors

In fact, the evidence

(for

is

complete.

The

isolated
v.

in
is

Front

frontem,

as the context
of the stone-

shows), CIL.

certainly a

mere blunder

cutter, as is often the case in other words,

even in carefully cut

Apicibus et

inscriptions (see 36. 3).

p. 57, cites

Christiansen,

De

I Longis,

thirteen such instances for vowels before nt

Hidden Quantity
42.
1.

in Declension.
(e.g.

It

is
;

maintained by some scholars

Marx,

FTulfs-

buchlein, p. 2

Lane, Harvard Studies,

i.

p.

89) that the ending

-nm

in the Genitive

Plural of nouns of the First and Second

-UM IN DEUM, NUMMUM,

The

ETC.

47

Declensions has u in such forms as Aeneadum, deum,

also in nostrum and vestrum.

nummum ;

facts in evidence are the

following
d)

On
final

early Latin coins prior to the First Punic

War, we
final

find

the

of

many

Genitives Plural omitted,

e.g.

Romano,

Corano.

Coins of the same date regularly retain

e.g.

of

the Nominative or Accusative Singular,

Volcanom, propom
i.

(=

probuni).

This has led

Mommsen

(CIL.

p.

9) to infer
in

that there
instances.

was a difference

in the quantity of the

the two

As

the o of the Nominative and Accusative Singular

was

short,

Mommsen

thought that in the Genitive Plural

it

must
ex;

be long.

But the material with which

Mommsen

deals

is

tremely scanty.

Genitive Plural forms occur in some

number

but only a few Nominative and Accusative forms are found,

viz.

Volcanom, propom. Again, Romanom (CIL. i. 1) and Aeserninom (i. 20) show that Genitives sometimes retained the m. Mommsen attempts to solve this difficulty by taking Romanom and Aeserninom as the Nominative Singular Neuter of the Adjecbut that is awkward. The natural inference must be that tive
;

there was no system in the omission of final

on these

coins.

The

coins represent no dialect

;

in fact

they represent widely

if

separated localities

hence

it is

no wonder

the final

m (always

weak) was sometimes written, sometimes omitted.

inscriptions, the oldest of

In the Scipio

which may date within a quarter of a

find final

century of these coins,

we

freely omitted in the

It is,

Accusative and Nominative Singular just as elsewhere.

therefore, extremely unlikely that

Mommsen 's
x. x.

hypothesis con-

cerning the coins

b)

is

correct.

An inscription
i.

of

Nuceria (CIL.

1081) has dvvmviratvs,

which Schmitz (Rheinisches Museum,

Studies,
PI. of

p.

89) regard as evidence

no) and Lane {Harvard that the // of duum (Gen.

dud) was long.

But even conceding the correctness of the

it

apex in this isolated instance,

remains to be shown that the

48
duumof

HIDDEN QUANTITY.
duumvir and duumviratus
is

in origin a Genitive.

Such

an etymology would involve the assumption that the duum- of

the Genitive Plural, duumvirum,

became transferred
etc.

to the other

cases, replacing duo in earlier duoviri,

is

Such an assumption
likely that

extremely improbable.

It is

much more

duumvir
In the

and triumvir are formed

extremely awkward, and
vir,

after the analogy of centumvir.

singular especially such forms as duovir, tresvir would have been

it

seems probable that the singular duumNucerian inscription,

triumvir were, for that reason historically anterior to duumviri,

triumviri.

The apex

in the

if

this etymol-

ogy be
as
is

correct,

would then be simply a blunder of the engraver,

altogether probable.

The evidence

in favor of

-um

in these

Genitives must, therefore, be regarded as of no weight, especially

in

view of the regular shortening of vowels before

Certainly
if

final

-m

in

Latin.

-um did by any

possibility exist in the days

of Augustus, the u
tilian
(i.

6.

18),

had become shortened by 90 a.d. For Quinas noted by Lane (p. 90), shows that to his ear

nummum,
2.

Genitive Plural, was nowise different from

nummum,

Accusative Singular.

Words

in

-er

of

the Second Declension, and words of

the Third Declension in -er and -x, have in oblique cases the

same quantity
dgri
;

of

the vowel
;

as
;

in the

Nominative,
;

e.g.

dger,
;

frater,

fratris

acer,

acris

pax, pads

tenax,

tenacis

fax, facts ; rex, regis; nix, nivis ; comix, cornicis; calix, calicis ; fel,fellis; os, ossis ; plebs, plebis.

Thus sometimes

the

Nomi-

native gives the clue to the hidden quantity in the oblique cases
(as dger, dgri)
;

sometimes the oblique cases give the clue to

the hidden quantity of the Nominative (as cornicis, comix).

3.

Words

of the

Third Declension ending

in -ns (Gen. -ntis)

uniformly have a short vowel in the oblique cases, as already

explained in

40. 3.

Greek words

in -as

(Gen.

-antis),

e.g.

Aias, Aiantis ; gigas, gigantis, have the

original

same quantity

as in the

(Aids; Atavros

yiyas, ytyavTos).

So, also, contracted

ADJECTIVES, NUMERALS.
Greek names
of cities in -ovs, -ovvtos,
-wv, -wvtos, eg.
e.g.

49
;

Selinus, Selinuntis
.

and

proper names in

Xenophon, Xe7iophontis

Acheron
two or

(not a contract form) has Acheruntis.

4.

In

all

words

of the

Third Declension ending

-ns

in

more consonants (excepting

and
is

-x

preceded by a vowel),

the hidden vowel before the ending

short, e.g. urbs, sors, drx.

Exceptions to this principle are plebs and compounds of uncia

ending

in -unx, e.g. deunx\ deuncis

is

quincunx, quincuncis.

Be-

fore -x the vowel

sometimes long, sometimes short, as already

explained in

2,

above.

Comparison of Adjectives.
43.

In the terminations -issimus, -errimus,

is

-illimus,

the hidden

vowel

short,

e.g.
i

carissimus, acerrimus, facillimus.

Apparent

traces of a long

in the
/

termination -issimus are found in inscrip-

tional forms with

is
;

longa.

piIssimus besides this

The word of most frequent occurrence we find a few other words, carIssimo,
<?..

CIL.vi.5325; dvlcIssimo,

vi.

16926; fortIssimo,

vi.

1132.

But

many

of these inscriptions

belong to the
i

last centuries of the

Empire, when the use of

untrustworthy guide, as

longa had

become an extremely

may be

seen by palpable errors.

As

regards the frequent occurrence of piIssimae, piIssimo, these

may perhaps be explained on

used
to indicate not

the theory that

longa was here

in

merely

1,

but also the/ which developed

/'s, i.e.

pronunciation between the two

use of
i

pijissimo.
ix.

Cf. the similar

longa in words like PompeIivs, CIL.

3748.

At
and

all

events, in the absence of the apex in these superlatives,

in

view of the absolute silence of the grammarians,

to attach great weight to the occurrence of

it

seems unwise
i

the

longa alone.

Against

i,

Lindsay {Latin Language,

p.

405) urges the occur-

rence of late spellings like merentessemo, karessemo, CIL.

ii.

2997.

Cf. 6.

1.

50

HIDDEN QUANTITY.
Numerals.

44.

As

separate words are to be noted

a)
b)

quattuor, but quartus (see 53 under area).

qifinque

and

its

derivatives, all of

which have

1,

as quin-

decim, quintus, quingenti, qiiinquaginta.

c)

the derivatives of unus

undecim, undeviginti,

etc,

d) milk, millia, and millesimus.

Pronouns.
45.
2.
1.

Nos, vos ; but noster, vester ;

nostri, vestri, etc.

Hunc

and hanc have a short vowel.

have
z.

3.
4.
5.

Ille, ipse, iste

The suffix -cunque has u. Compounds retain the quantity

of the elements of

which

they are compounded, as

qiiisquis, cujusque.

CONJUGATION.
Root Forms.
46.
1.
e.g.

Presents formed by means of the infix n have a short

vowel,
jug-).

/undo

(root fud-)

frdngo (root frag-)

e.g.

jungo (root
;

Before a labial, n becomes m,

/tf-).

runipo (root rup-)

lambo (root
ative

Care should be taken not

like

to confuse deriv-

and contract Presents

v~endo, pi'endd,

with genuine

nasal formations.
2.

In most Presents the hidden vowel

is

short,

e.g.

necto, serpd,
:

verfo.

But the following exceptions are to be noted

:

a) First conjugation
tracto.

jingo (for jurigo), narrd, drno, purgo,

b)
c)

Second Conjugation
Third Conjugation
:

ardeo.
all

verbs in

-sco (r),

except compesco,

disco, posed, vescor.

d) Fourth Conjugation

nutrio, ordior.

: :

CONJUGATION.
3.

51

The

quantity of the vowel in the Present regularly remains

it

unchanged (when

becomes hidden) throughout the

e.g.

entire con-

jugation of the verb,

ardeo

ardere
gerere
scribere

arsi
gessT
scrlpsi

ar suras
gestus
scriptus
victiirus

gero
scribo

vivo

vivere

vixl

ftgo

figere

fixT

fix us

Thus
veixit.

inscriptions give fIxa, scrIptvm, conscreiptvm, vIxit,

But the following exceptions to

noted
a)
dico

this general principle are to

be

dicere
diicere

dlxT
dihxT
cessi

d ictus
ductus
cessurus
is

duco
cedo

cedere

The

short vowel of the Perfect Participles dictus and ductus

ix.

assured by the statement of Aulus Gellius {Nodes Atticae,

6)

and by the testimony

b)

of the

Romance

languages. (See
is

52. s.vv.)

The

short vowel of the Present

lengthened in the Perfect

hidden, in the following

Indicative

and Perfect

Participle,

if

verbs
ago
cingo

agere
cingere

egi

actus
cinctus
delictus

cinxT
deliqui distlnxi

delinquo
distinguo

delinquere
distinguere

distlnctus

emo
exstinguo fingo

emere
exstinguere

eml
exstlnxT

emptus
exstlnctus

fingere

finxi
fregi

flctus

frango
ftingor

fr angere
fungi
jungere
legere

frdctus

functus sum
jilnxi
legi

jungo
lego

junctus
lectus

pango
pingo

pangere
pingere

pepigl

pdctus
piclus

pinxl

BOSTON UNIVERSITY

cou
jJBRARY

52
pungo
rego

HIDDEN QUANTITY.
pungere
regere

pupugT
rexT

ptinctus
7'ectus

r din quo
sancio
struo

relinquere
sancire
struere

r ellqui

relictus

sdnxi
striixi
tetigi

sanctus
structus
tdctus
tectus

tango
tego

tangere
tegere

texi

tinguo traho

tinguere trahere

finxi

finctus

trdxi

trdctus

ungo

ungere
in

iinxl

unctus

So also
4.

compounds and
list is

derivatives of these verbs.

The

evidence for the long vowel in the Perfect Participles of

the foregoing

found

a) In the statements of Gellius,

ix.

who

testifies

{Nodes

Atticae,

6) to the quantity of the vowels of actus, lectus, unctus,

and

in

xii. 3.

4 to that of

structus.
:

b)

In the testimony of inscriptions, which show the following

vi.

Actis CIL.
tions de

1377
p.

redacta vi. 701

;

exactvs Boissieu,
4104 exstInctos
; ;

Inscripv.

Lyon,

136
vi.
;

cInctvs CIL.
;

x.

defvnctis CIL.
vi.

1326; dIlectvs

6319 lectvs

xi.
;

1826

25617;
e.

INFRACTA
recte
xii.
iii.

ix.

60
;

IVNCTA

X.

l888
;

SEIVNCTVM

vi.

1527

38;

ranum
v.

2494 tra[cta (not

;

tector

vi.

coemto Monumentum Ancy5205 certain) CIL. vi. 1527 e. 14; sancta

actus, tactus,fractus,

2681
c)

Oscan saa(n)htom (= sanctoni). In the retention of a in compounds of

;

pactus, tractus
that the a
tion, as for

{e.g.

coaches, attactus, refractus, etc.),

which shows
this situa-

was long; short a would have become

example
in confeclus for

e in

an original * confactus ; acceptus

*e~raptus.

for

an original *accdptus ; ereptus for

d) For cinctus, delictus, distinctus, exstinctus,

pTinctus, relictus, tinctus, the long

fictus,

pictus,

vowel

is

assured by the evidence

of the
5.

Romance,

e.g.

Italian cinto, delitto,fitto, relitto, Unto.

The evidence

for the quantity of the

vowel

in the Perfects

of the foregoing

list is

found

CONJUGATION.
a) In inscriptional markings, as

53

C0NivNXiT(Wilmanns, Inscript.
1793);

Latinae

104);
x.

texit (CIL.
231
1,

x.

rexit (CIL.

v.

875);

traxi (CIL.
b)

18).
ii.

In Priscian's statement (Keil,

~e.

466) that rexi and text

have
c)

In the testimony of the

exstmxl,

Romance

languages, which point to

chixl, dist'inx'i,

finx'i, plnxl, struxi, tinxi, unx'i.

d)

The long a

in sanxi rests

upon no

specific evidence, but

may
first

perhaps be safely inferred after the analogy of sanctus.

(e.g.

Until recently the principle was maintained

edition) that
all

by Marx in his
b, d,

monosyllabic stems ending in

or

g had
this

the hidden vowel long in the Perfect Indicative and Perfect Participle

wherever euphonic changes occurred.

According to

theory

we should have

e.g.

seindo, scindere, scidl, scissus ; niergd,

first laid

merge?-e,

mersi, niersus.

This principle was

i.

down by
apply to
entirely
specific

Lachmann

(on Lucretius,

805) for Perfect Participles alone,

to

and was subsequently assumed by other scholars

the Perfect Indicative as well; but this position
is

now

abandoned.
evidence.

Each long vowel must be supported by

In the 3d edition of his Hulfsbilchlein

the principle that
all

(p. 1),

Marx

lays

down
net.

vowels are long

in

Latin before nx and

These combinations occur almost exclusively

on pp.

in the verbs given

Whether the general principle is sound, may be 51, 52. questioned. For example, we have no definite evidence in favor
of the long

vowel before nx in anxius, lanx, or phalanx..

Verbal Endings.
47.
1
.

The hidden vowel

is

short before ss ( 40. 4)

and

st in the

terminations of the verb,

e.g.

fuissem, amaz'isse ;

ftiisfi, fiiistis.

This

is

shown not only by the historical

dedisfi,

origin of these formations,

but by such metrical usage as Plautus, Amphitruo, 761, dedisse;

Menaechmi, 687,

where

iss

and

ist

are treated as short

54

HIDDEX QUANTITY.
*
'

y lables by neglect of position (see

36. 2).

Contracted forms

are. of course, an exception to the above principle, as amassc.

com:~\
2.

:~:sse,

audlsset, amasfi, nestis.

:

Formations of the type

.

dixit \ioi dixisft), accestis (for ac-

..-.-.-

jusfi [for jussisti). trajre,

sun

etc..

have

the

same

quantity- as the regular forms.

Compounds.
48.

Marx

yj>.

S) holds that the

is

vowel of a monosyllabic prepo-

sition, if

hidden,

long in composition

when

the preposition loses

a final consonant.

Thus he maintains a long vowel for the initial

*dd-scdnde~)
;

syllable of ascendd (for

di-stinguo

(*dis-stingud^

suspicw (for *sub-spicio).

But

this principle rests

:

upon an unS9.

tenable theory of compensator}- lengthening

see

Inchoatives.
49. Inchoatives in -see, -scor
e.g.

have a long vowel before

-sc,

labdsco, floresco, nitesce, treniisco, ddipiscor.

Gellius [Nodes
this class as havall.

Atticde.

vL 15^ mentions a number of words of

ing a long vowel, and implies that this was generally true of

The Romance languages show

and
:
:

that -see
is

and

-iseo (-iscor)

had

1.
-.

But the hidden vowel

short in compesco, disco, posed,

'.

Irregular Verbs.
50.
e.g.
1.

The
i

root vowel of esse

essem.

is

short under

all

circumstances,

est, estis, esfo,

2.

Edd,

eat.'

has a long

e in the

forms

Is.

est

estis,

essem. esse,

estur, essetur.

Cf.

Donatus on Terence. Andria, 81; Servius on

Virgil, Aeneid, v. 7S5.

Marx

(p. 9) lays
ii

down

the principle that in

i

compounds
st,

of
in

eo,

forms containing

have the second

long before

as

e.g.

: ;

WORD FORMA TIGX.

interiisti.

r:

This theory rests solely upon the occurrence of este-

rieisti in

CIL.

i.

1202.
i.

But

ei occurs
(

elsewhere in inscriptions,

incorrectly written for

eg. pare>~teis

'= parentis), CIL. L 1009

probable that esteri-

f aceivndae
kcstti is

= fanundae).

It is altogether

another instance of the same sort

Word
51.
1.

Formation.

Substantives in -abrum, -atrum, -atrum, derived from

.

verbs, have ~
2.

ftdbrum, lavacrum. aratrum.

The
have
'

derivative endings -ettus (a,

e

um\

-Ulus (a, um), reguwis.

larly

and

/,

but the following have a long vowel,

hillae,
ovillus,

catella,

httle

chain/ angulfla, BoriUae,

stiBa,

suillus, villa.
3.

The vowel
:

is

short

in -em:

::us,

-e minus),

-urnus

(-umius, -urn'inu:
:r
:

nus, taberna, Saturnus.

In vernus

:he r

is
is

not a part of the

suffix.
(-ester, -estris, -estz-

4.

The vowel

short in the endings -estus

cus, -es:~

:mm),

-ust.

.jelestis.

domesticus,

:-.

pesfas, capistrum, venustus.

In

semestris, Justus, the long

vowel

belongs to the stem.

:

The vowel

is

short in the endings -unculus, -undo, -ercuiks,

-useulus. e.g. ratiuncula, paterculus, maiusculus,

homuncib ; pTus-

eulus (from plus) naturally has u.

6.

In compounds, the connecting vowel

i is short, eg.

nari-

fragus. lectisternium.

56
52.

HIDDEN QUANTITY.
List of the

Most Important Words containing a Long Vowel before Two Consonants.

1

B.

abiegnus

see 38, end.

:

bardus,

'

stupid
:

'

from bard.
Plutarch,

acatalectus

Gr. aKardXriKTOS.

Bedriacum
Otho, 8,

JSrjTpiaKov,

actio, dctito, actor: see ago.

11.

actutum

like actus.
:

bellua
bestia,

for bes-lua.

ademptio

see adimo.
:

Bestia

Brjcrrias

Plutarch,
;

adz mo, ademptus

afficto
:

like emo.

like jllctus.
:

Marius, 9 ; Cicero, 3 mance would point to

biformis
3. &).
:

the

Ro-

e.

Africa, Afri

from Afer.
:

see

forma.

ago, agere, egl, actus

Alcestis
:

see 46.

bilibris

like libra.
:

Gr. "AXurjcms.

bimestris
bovillus
:

from mensis.
Gr.

Ale do

Gr. 'AXtj/ctc^.
:

from bovinus.
:

aliorsum
allptes
:

for *alio-vorsum.

biibrestis

(3ov(3pr]<TTis.

Gr. aXeiTTTrjs.
:

burrus

Amazon

Gr. Afxafav.
'

bustum

u u

ace. to the

Romance.

ace. to the

Romance

cf.

ambus tus : see uro. combiistus and ustus. Amsdnctus : see sdnctus. Biithrotum : BovOpurov. angullla : I ace. to the Romance. Aquillius : AqvIllivs CIL. vi. 12264. C. ardtrum : see 51. I.
:

drdeo, -ere, arsi, arstlrus: like aridus,

catalectus
catella
: e. :

Gr. KardXr]KTos.
; catella,
(

drdus.

from catena

bitch,'

Arginussae

Gr. 'ApyivovaaaL.

has

dspemor
dthla
:

from d and spemor.

r

calillus

from catinus.
is

Gr. affkov.
:

cetra

better orthography

caetra

athletes

Gr. ad\y]T qs.

:

see 61.
chirurgtis
also
:

dtramentum
atrium
CIL.
axilla
1
: :

like dter.

from
Priscian,

dter

ATRIVM,

cicatrix

Gr. x lP 0V py^a in Plautus, Amphitruo,

2.

vi.

10025.
iii.

446
36.
ciccus,

see 36.
:
1

-um

ace. to the
list
:

Romance.

The

following classes of words are omitted from this

a) Most derivatives and compounds.

b)
c)

All words containing ns or nf.

Inchoatives in -dsco,

-esco, -iscd.

d)
<?)

Some rare Greek loan-words and proper names. Nouns and adjectives in -x, whose Genitive (ace.

to 42. 2)

shows the

preceding vowel to be long.

WORD- LIST.
Cincius
:

57
CIG. Addenda, 4342,
point
fin.

clNClA, CIL.

vi.

14817

et

Kpeicrireivov,
d. 4.
i

passim.
cingo, cingere, clnxT, cinctus
:

The Romance would

but see 36. 5

T in the

to

Perfect
to

and Perfect

Participle ace.

crusta,

crustum

in

CIL.

i.

1199;
form
384)
;

the

Romance
:

see

Korting
cInctvs,

the

Romance

points both to crus-

{Worterbucli)
Grundriss,
i.

d'Ovidio (Grower's

tum and
with u.

also to a collateral

p. 501 f.);

Grober (Arckiv,

vi.

4104 see 53 s. v. clandestlnus : from cla??i and des{i)tus

x.
;

CIL.

Korting
Ctesiphon,
cucullus,
to
'

{Wbrierbucli).
-otitis :

Gr.

-Qjv, -Covtos.

from desino
i.e.

(cf.
;

positus from pond,

'

hood

'

the

po-sino)

hence

secretly put

aside.'

cldtra, cldtri

'

Gr. nXr/dpa. Gr. K\vTcu/xri<TTpa.

one with u, antwo forms, other with u see Grober (Archiv, vi. 384); Korting ( Worteri. 555;
;

Romance

points

Clytemestra

bucJi)

cucullus,

'

cuckoo,' has u.
ix.

Cnbssus

Gr. ^.vwcrcros.
:

ciinctus

CVNCTI, CIL.

60.

edgo, cogere, coegl, codctus

see ago.

ciistos :

KovcrribdTjs,
i.

Lydus, de Magis-

combiiro, combitrere, combussi, combiistus

:

tratibus,

46

ace. to the

Ro-

see tiro

and bustum.

Marx

mance.
Cyclops
:

also

marks the

long, regarding

Gr. KvkX&iJ/.

combiiro as for co-amburo, and com-

paring cogito (for co-agito).

comedo, comestus
:

D.
2.

cf.

edo

see 50.
:

deligo, -ere, delegi, delectus

like lego.

cbmo, cbmere, compsi, comptus

to the

ace.

delinquo, -ere, dellquT, delictus: face,

to the

Romance.
com: :

compingo, compingere, compegT,

delilbrum
1

Romance. u in Plautus, Poenulus, :

see 36.
2.

pdctus
conjiinx

see 46. 3. b).

175

CONIVNX, CIL.

vi.

6592

et

demo, demere, dempsi, demptus ; like

emo.

passim; but conjux has

tango.
contio
: :
:

u.
:

contingo, -ere, contigT, contdctus

like

deunx
died,
3.

from de and
:

iincia.

dextdns
for co-ventio
;

from de
dixi,

+ sextans.
dictus
:

40. 2. a).

dicere,

corolla

from corona.
a in Plautus, Amphitruo, 707;
2.

a).

see 46. Certain of the Romance lan(Fr.

dit ;

crdbro

guages
etc.)

Old

Ital.

ditto,

see 36.

point to

collateral

dictus,

crdstinus
cresco
:

from

eras.
xii.
;

which Osthoff {Morphologische Un-

crescens, CIL.
Gr. KprjtrKrjvs

4030

et

tersuchungen,

iv.

74)

thinks

be-

passim;
to the

also ace.

longed to the colloquial language.

Romance.
Kprjaa-a.
:

But possibly those Romance languages which point to

Mostellaria,
1

Cressa

have simto

cribrum
55
;

in Plautus,
2.

ply adapted the Participle

the

see 36.
:

vowel of the Present and the Perx.

crisp us

Creispinvs, CIL.

3514.

fect.

See Grober {Archiv,

vi.

385).

58
dicterium
Diespiter
digladior
: :

HIDDEN QUANTITY.
Gr. deiKrrjpiov.
for dies for dis

exstinguo,

-ere,

exsfinxT,

exstinctus

and pater.

extInctos,

CIL.

vi.

25617
3.

cf.

+ gladior
;

by comby com-

distinguo

see 46. 3.
:

pensatory lengthening
digredior
:

see 89.

extraordindrius

from Jr^.

for dis

+ gradior
;

pensatory lengthening

see 89.
fdstus, a,
:

dilemma

Gr. StA^^ct.
like lego.
:

um ;

cf.

fas.
v.

diligd, -ere, dilexi, dilectus

favilla

favIlla, CIL.
also

3143.

The
f.

dlrigd, -ere, direxi, directus

like rego.

Romance
fello
:

seems

to point to

dirimo, -ere, diremi, diremptus ; like

from same root

as

femina ;

Gr.

emo.
distinguo, -ere, distinxT, disfinctus
ace. to the
:

dijXvs.

festlvus

from festus.
holiday

Romance
(

see d'Ovidio
i.
;

festus
(

from the same root asferiae

'
'

(Grower's

Grundriss,

p.

502)
cf.

*fes-iae),
i.,

FESTVS in

Korting
doldbra

Wbrterbuch)
3. b.

ex-

CIL.
25th.

Fasti Praenestini for April

in the

stingud ; see 46.

: cf.

So also

proper name
3179; Festi,
5353;
Gr.
fre-

51. I. duco, ducere, duxt, ductus

3.

Festus: Festvs, CIL.

:

xii.

see 46.

v.

2627;

Festae,

iii.

a)

perdvxit, CIL.

xii.

2346

et

3>77<rros,

CIA.

iii.

635

and

passim.
E. ebrius
:

quently.
e,

The Romance

points to

indicating that e of the classical

;

period ultimately became reduced

in
;

e regularly

Plautus,
2.

e.g.

see 36.

5.
:

Trinummus, 812
eclipsis :

see 36.

figo, figere, fixT, fixus

fIxa,
vi.

Monu1

Gr.
' :

e/cXei'/'ts.
<?j/,

mentum Ancyrantim,
<?jj<?,

18;

ace.

edo,

'

eat

^j^w,

^^.

See
:

to the

Romance.

50. 2.

fingo, finger e, finxT,

effregl,

fictus

T ace. to

effringo, -ere,

effr actus

like

frango.
emo, emere, eml, emptus
:

see 46.

3.

Romance see 53 s. v. firmus : fIrmvm, CIL. iv. 175 passim ; the Romance points to
the
;

et
t,

*)
emungo,
to
-ere,

showing that
-,

T of the classical

pe-

emuncttis
;

il

ace.

riod
36. 5-

had become reduced;

:

see

the

Romance

see
i.

d'Ovidio

(Grober's Grundriss,
erigo, -ere, erexT, erectus

p. 515).

Jldbrum
fiTgo,

see 51.
-fiixz,
1

I.
:

like rego.

-ere,
i.

-fiictus

AFLEICTA,
also

esca

e ace. to the

Romance.

CIL.

1
T.

75

the

Romance

Esquiliae, Esquilinus : Gr. 'HctkvXIvos,

in Strabo, v. 234, 237.

points to
fidsculus
:

from fids,
ace. to the
;
it

Etruscus :
existimo
:

cf.

Etruria ; Gr.

'ErpoOcr/cos.
;

fluctus

Romance.

from ex and aestimo STIMAVERVNT, CIL. V. 505O. exordium : from ordior.

EXl- find, -ere,fiuxi

is

probably long in

fluxi in view oifiuxzts.

fiuxus

ace. to the

Romance.

WORD-LIST.
forma
:

59
: 1

see

Donatus
(pwpp.7)

on
in

Terence,

hibisctim

ace. to the

Romance.

Greek in- hillae : from hira. shows 0. hirsutus : like hlrtus. Romance also scriptions the HTrtius and hirtus : T formula : from forma. mance. frango, -ere, fregl, fractus : see 46.
Phormio, 28;
;

ace. to the

Ro-

3- *)

hisco

see 49.
:

fngeo,

-ere,

fnxi

46. 3.
T ace. to the

HTspellum
v. 227.

cf.

Gr.

Et'<r7r^X\oi',

Strabo,

frigo, -ere, frixT,

frictus :

fructus

Hispo, HTspulla : like Hispellum. Romance. hor nus : from hora ? : u ace. to the Romance. Old French froit points to a collateral horsum : for * ho-vorsum.
;

fructus

see Osthoff, Geschichte des

hydrops

like
:

Gen. hydropis;
Gr.

42. 2.

Perfects, p. 523.

Hymettus

Gr. 'TfMijTTos.
:

fruor, frui, fructus

the

sum

ii

ace. to

Hypermestra

'T-rreppiTia-Tpa.

Romance.
:
:

frustrd

FRYSTRA, CIL.

vi.

20370.
Tgnis
:

frustum

ace. to the

Romance.

Ignis, CIL.
:

xi.

826.

fungor, fungi, futictus

FVNCTIS, CIL.
xii.

sum

DE-

illorsum
illustris :

for *illo-vorsum.

V.

I326;

FVNCTO,

from
-ere,

//?.*.
i*

3176 et passim. furtim : from fur.

furtivus
:

Illyria

Eillvrico, CIL.

2
.

p. 77.
:

impingo
infestus

impegT, impdetus

see

from fur.
from fur.
ace. to the

46.3.^).:

furtum
fustis
:

infesti, CIL. v. 2627

cf.

Romance.

manifestus.
InstTnctus
:

see distinguo.
intellec-

G.

intellego, intellegere, intellexT,

geographia
Georgius
georgicus
gtisco
:
: :

Gr. yewypacpia.

tus

like lego,
:

Gr. Teupyios. Gr. yewpyiubs.

intervdllum

from vdllus.
in

introrsum

for *intro-vorsum.
:
ii

49.
:

involucrum
from Gr. yX&aaa.
267
Iolcus
;

Plautus,

Captivi,

glossarium
glossema
gryllus
:

36. 2.

from Gr. yXwaarj/xa.

ace. to the

Gr. 'I\k6s.
:

Romance.
2.

islorsum

for *isto-vorsum.

gryps

like

Gen. grypis ; 42.

J.

H.
hdctenus
:

jentdculum
jentdtio
:
:

see 40. 2. a),

2.

like hdc.
: :

see 40.

#).

Hellesponttis

Gr. 'EWtjo-ttovtos.

jugldns

from /iw- and gldns.

junxi, junctus; see 46.

Herciddnum
xii.

Hercvlaniae,

CIL.
Casiii.

jungo,

-ere,

1357

'HpKovXdveov, Dio
23;

3- *)

sius, lxvi.

'RpK\av6s, CIA.

jurgo

for jurigo,
:

from jus.

1197.

Jiistinidnus

ixova. Justus.

6o
justitium
Justus
:

HIDDEN QUANTITY.
:

from jus.
:

lubricus

u in
853
;

Plautus, Miles
36. 2.
:

Glo-

from jus
v.

also IVSTO,

CIL.

ii.

riosus,

210;
jiixtd,
with.'

5919.
:

luceo, -ere, luxl

see 46. 3.

juxtim

from jugis

'

joined

lucta

ii

ace. to the

Romance.
:
;

liictor liictus

:
:

like lucta.

from
vi.
;

lugeo
e.

also

L.
labor,
Idbi,

CIL.

1527
-

66
2.

lvctvm, lvctv, CIL.

lapsus

sum
xi.

see

46.

v.

337

x 4041.
expiation

dIlatsam, CIL.
'

3123.
;

lugeo, lugere, liixT : see 46. 3.

la-

labrum,

bowl
lip,'
ii :

'

for
a.

lavabrum

lustrum,

'

'

lvstrvm, Monii.
l

brum,
labrusca

has

umentum Ancyranum,
Romance.
ii.

3, 5,

ace. to the

3, 6,

10

lustrum,

haunt,'

has

laevorsum

for *laevo-vorsum.

u.

lamna : syncopated for lamina. lardum : syncopated for Idridum.

LArt-, CIL.
x.

lustro

like lustrum,
:

luxuria
luxus
:

see lux us.

u ace. to the Romance. Idrs, Idrtis : 633. lycurgus Latin : Gr. AvKovpyos. larva : like larua, the early
form,
e.g.

Plautus, Amphitruo, 777;

Captivi, 598.

M.
; cf.

Idtrina

for

lavdtrina
;

Plautus,

mdlle

for

*mag(e) {magis)
:

velle.
i.

Curculio, 580
latro,
'

36. 2. in

manifestus
p.

[mani] festvm, CIL.

;

bark'

oblatrdtricem,

Plautus,
36. 2.

Miles

Gloriosus,

681

319; very uncertain. Mdnlius : from Mdnius

Manlia, CIL.
51. 51.
1. I.

v.

615

Manlio, Manliae,

lavabrum : see lavderum : see

ix.

3942.
:

manilpretium

in Plaut.
:

Men.

544.

lego, -ere, legl, lectus :

see 46. 3.

Mdrcellus, Mdrcella

from Marcus

lemma

Gr. \9iwxa.
:

lemniscus

Gr. XthivLgkos.

MArcella, CIL. xii. 3188. Mdrcius : from Marcus ; Marcivs,

CIL.
v.
1

Lemnos :
lentlscus
:

Gr. Aijfivos.
T ace. to the

555
137.

et

passim

~M.dapia.oi>,

Romance.
816
;

CIG.
2802

libra

in Plautus, Pseudolus,

Marcus: Maarco, CIL.

; ;

36. 2.
libro
:
:

like libra.

i. 1006; xiv. MARCI, Boissieu, Inscriptions MdapKos, CIG. de Lyon, p. 143

llctor

lIctor,
;

CIL.

often

LICTOR,
v.

699 and Ephemeris Epivi.

887

et passim.
:

Mars, Mdrtis
809

Martis, Monumeniv.

graphica,
(

51

Xei'/crwp,

Eckinger

tum Ancyranum,
et passim.

21

CIL.

x.

Orthographie Lateinischer Worter

p. 43).

in Griechischen Inschriften,

MdrsT :
mass a
:

like
:

limpidus

ace. to the

Romance.
:

Mdrlidlis

like

Mars. Mars,
:

lingo, ere, linxi, linclus

T ace. to the

Gr. p.aa.

Romance.

mdlrimonium

from mater.

WORD-LIST.
matrix
:

61
;

from mater.

nondum
MATRONA,
iii.

matrona : from CIL. v. 5249.

maxilla
:

mater ;

: from ndn and dum dvm, CIL. x. 4041. 6.

:

NON-

nongenti

for *no(v)engenti.

according to Priscian,

nonne

from ndn.
:

36. (Keil).

nonnulll See Cramer, Aneciii.

from ndn and nullus.

maza

Gr. /xafa.

Norba :
w^y^J
:
<?

Gr. Nwp/3?/.
ace. to the

dota Oxoniensia,

293.

Romance.
:

mercennarius
Metrodorus
metropolis
:

for *merced-ndrius.

nilbo, -ere, niipsi, niipta

Gr. MTjrpodtopos.
fxrjTpbiro\LS.

nudiustertius, quartus,

etc. :

see 46. 3. see 86.

Gr.
:

nullus

from
x.

;z<?

and

z7//.y /

NVLLVM,

mille, millia

mIllia,
i.

Monumentum
mIlliens,
iii.

CIL.

4787.
:

Ancyranum,
34 ; milvus
miicro
1

16

nundinae, niindinum

for

*no(v)en;

ace. to the

Romance.

dinae ; noundinae in early Latin

as in the early Latin miluos.

:

Nvndinvs, CIL.
nuntius
:

xii.

3650.

Mdstelldria
:
ii

from monstrum.
Frag.
13
(ed.

for
').

*nove-ntius?

('news-

in Atta,
;

bringer
niintio
: :

Ribbeck)
miilleus
:
ii

36. 2.

like nuntius.

ace. to the

mullus: u ace. to
miiscerda
:
:

Romance. the Romance.

nuptiae

like niipta.
:

mlsquam
niitrio
: :

like itsquam.

from
from

mics.
mils,

like niitrix.
ii

musculus

nutrlx

in Plautus, Curculio, 643

muscus : u ace. to the Romance. mustela : from mils.

Mycalessus
:

nutrlcatus, Miles
niitricant,

Gloriosus,
Gloriosus,

656
715
;

Miles
O.

Gr. Mu/caX^ao-os.

36. 2.

N.
nanciscor
:

obliviscor

see 49

OBLlvIsCEMVR,

see 49.

CIL.
olla

vi.
:

6250.
Gr. Olvwrpia.
;

Ndrnia
narro
nascor
ttim
:

Umbrian Nahar- ( = a). narrem, Boissieu, InscripLyon,

p. 136.

Oenotria
:

for

aulula

olla, CIL.

vi. 1 0006

tions de
;

et passim.

49

nascerer, Monumenii.

Onchestus

Gr.
:

"OyxW"

*-

Ancyranum,
xii.
:

44

Nascenti-

Opus, -imtis
ore a
:

Gr. 'Ottovvtos.

bvs, CIL.

3702.

ace. to the
:

Romance.
-

nasturciwn
tiefastus
:

from ndsus.
neglexT,

orchestra

Gr. 6pxv a"rP a

from nefds.
neglectus ;

ordior
see

like ordd.

neglego,

-ere,

ordo
de

ordinis, Boissieu, Inscriptions

Lyon,
;

p.

nequidquam (nequicquani)
quid.
nltor, nitT,

from Abl.

xii.

orno

3312 ornare, CIL.

ace. to the

136; CIL. ix. 5177; Romance.

xii.

4333

et

mxus sum

nolle:
(for

by contraction from
*ne-velle ; 73. 3).

see 46. 3. *novelle

passim,

ornamentum

ornamentis, CIL. 3203 et passim ; cf. orno.

:

xii.

62
oscen
:

HIDDEN QUANTITY.
from
:

os.

pigmentum:
1344
;

pIgment-,

CIL.

viii.

oscillum
oscito :

from osculum.
os.

face, to the

Romance.
:

from
:

pingo, pingere,plnxi, pirtus

Jingo, which
is

see under

osculor

from
from

os.

precisely parallel.

Ostia

from
:

os ; Gr. 'ft<m'a.

pistillum, plstor, pistus (from pinso),

ostium

wtjria, scholion to
;

pistrlnum, plstrilla
v.

pIstvs,

CIL.
is

Aristophanes, Plulus, 330

ostivm,

6998.

The Romance evidence

is

CIL.
ovillus
:

vi.

4710

ostio,
v. 14.

Monumen-

conflicting, but

favorable to

T.

tum Ancyranum,
Oxus:

Pistdria

like plstor.
:

from ovlnus.

plebisatum

plebl scltum, and

better

Gr. *fios, in Strabo.

P.

so written.

plebs
v.

like genitive plebis ; PLEPS,

;

CIL.

6797
:

xii.

4333.

patiscor, paclscl, partus

sum

see 49. plectrum

Gr. TrXrJKrpov.
:
:

palimpsestus
paluster
:

Gr. iraXlpj^-qiiTos.

PlTsthenes

Gr. nAetcrflei^s.

from palus.
:

plostellum

from plaustrum.

pango, pangere, pepigT, partus

the

plusculum

from plus.

compdrttcs, po stria, -is : Gr. TroirjTpia, TroiTjTpls. compounds impdrtus, Polla : = Paulla ; PoLLA, CIL. xii. see 46. 4. r). point to d paradFgmd : Gr. Tra.padet.yfAa.. cf. Pollio. 3471
; ;

pdscd, pdscere, pdvT, pdstus

pdstillus
:

see 49.

pollingJ, -ere, llnxl, linctus

like lingo.

like pdscd.

pollinctor
Pollio
:

like pollinctus.

pdstio

: ;

Yike pdstus.

from Paullus ; Pollio, CIL.

pastor
i.

like pdsttis
;

PAASTORES, CIL.
x.

551

pastoris, CIL.

827.
36.

22840 et passim ; IIa>\AiW in Plutarch, Dio Cassius, and elsewhere.

vi.

pdxillus : ace. to Priscian,

iii.

polluceo, -ere, -uxi

46. 3.

pegma

Gr.
: :

7rr)yfj.a.

Polymestor

Gr. Uo\vfj,r]<rTU)p.
:

pentdthlum
peremptd/is

Gr. a.d\ov.

porrigo, -ere, porrexT, porrectus

rego.

like

from peremptus (emo).

:

pergo,pergere, perrexl, perrectus

rego.

like

praelustris

like lux.
:

prdgmaticus
:

Gr.

7rpay/j.o,TiK6s.

perlclitor

like perTculum.

Praxiteles: Gr.
:

Upd^LTiXrjs (7rpais).

perimo,
emo.

-ere,

peremT, pere??iptus

like

prendo

for

pre-hendo.
:

primordium
:

from ordior.

Permessus
perrepto
persolla,
:

Gr.

TiepfxTjaa-ds.
r<?//<?

prlnceps

from primus and capio.

:

from

(repd).

principalis

from prlnceps.

for person(u)la,

from /^r- prmcipdtus : from prlnceps. principium : from prlnceps.

Frlscidnus
:

Pessinus, -untis

Gr.

Ilecro'iJ'oOi'Tos.

from prlscus.
:

Phoenissa
pictor
:

like

Phoenix,

prisons and Prlscus

xi.

Priscvs,
ix.

CIL.

like pirtus (pingo)

:

1940

pictura

like pictus.

4354. Ilpei<rKOS CIG. 4494 et passim.

;

PrIscvs, CIL.

WORD-LIST.
pristinus
:

63
-sco

like priscus.
:

other

formations have practi:

procrdstino
Procrustes
:

from

eras.

cally invariably e before sc

quievi

and quietus also point to quiesco ; qviescere is found CIL. vi. 25531. profestus : from festus. promo, -ere, prompsi, promptus : see quincunx : from quinque and uncia. quindecim : from quinque and decern ; 46.3face, to the Romance. prorsum, prorsus : for *pro-vorsum, quingentl, quingenT, quingenties : from -sus. quinque. prosperus : from pro *spere ? (' accordQuTnqudtrils : from quinque ; d in ing to expectation'). Plautus, Afiles Gloriosus, prostibulum : from pro and stabulum. 691 2. Publicius, Publicola; from piiblicus. 36. Poplicola is another word, viz. from quinque : QVlNQVE, CIL. vi. 3539 et form of populus, passim ; i ace. to the Romance. poplus, early
Gr. HpoKpoti<TT7]s.
;

'people.'
piiblicus
:

quinqudgintd
piibes ; pvblic<5r[vm,
;

: :

from quinque.

from
1377

quinquennium
quinquies
quintana
: :

from quinque.

CIL.

vi.

in Plautus,
;

Miles

from quinque.
from quintus.
:

Gloriosus, 102, 103

et

Captivi, 334

passim

36. 2

u also

ace. to

Quintilidnus
Quintilis
:

from quintus.

the Romance.
Publilius
Pfiblius
: : :

from quintus.
:

like Publius.

Quintilius
LIO,

from quintus
iii.

QvlNCTlqvIniii.

like piiblicus.

CIL.

384.

pulvillus

from pulvimis ; pvlvIllvs,

Fasti Cap.,
a.

quintus,

Quintus,

Quinctius

CIL.
pungo,
piinctus

i.

297.
:

TVM,
u
ace.
tos,

Monumentum Ancyranum,
CIG.
for

-ere, puptigi,

punctus

I; ilonga occurs repeatedly; Koetv-

to the
:

Romance.
see pungo.
for *purigo

2003

ace.

to

the

Romance.
{purus)
:

purgo

also

quippe

ace. to the

Romance.
:

quorsum, quorsus
*qud-vorsus.

*quid (Abl.) and -pe. : for *quo-vorsum,

purgamentum
purgatio
:

from purgo.
ace. to the

from purgo.
from pus
Q.
;

pustula

R.
ralius
:

il

Romance.
quartus
;

for rdr{ii)lus
:

from rdrus.
ipsa).

rdstrum
iii.

from rddo.

quartvs, CIL.

Monumentum Ancyranum,
et passim.

4959 ; redpse : for re edpse (Abl. of iii. 22 rede, rector : like rectus.
rectus
:

see rego.
-ere,

qtidrtanus

like quartus.
:

redigo,

redegi,

redddus

like

quartdrius
quiesco
:

like quartus.
to

ago.
like
;

Gellius, Nodes redimo, -ere, redemi, redemptus : some persons proemo ; 'PedijvrrTa, CIG. 9811 nounced quiesco in his day but DEMPTA, CIL. vi. 22251.
ace.

Atticae,

vii.

15,

RE-

64
redemptio, redemptor
:

HIDDEN QUANTITY.
from redimo.
screiptvm, CIL.
to the
i.

206. 87
x.

109

regnum
regno
:

see 38, end.

CONSCRiPTis, CIL.
;

3903

i ace.

like

regnum.
:

regndtor, regnatrix
rego, -ere,
b).

from regno.

tor

scrlptos

Romance Umbrian (Nom. Plu.).

screih-

rexT, rectus: see

46. 3.

scrlptio, scriptito,

scriptor, scriptura

see scribo.
-ere,

relinquo,
46. 3.

rehquT,

rehctus

see

segmen
segnis

like
:

segmentwn.

*)
-T :

segmentum
see 49.
:

remimscor,
r<?^0,

see 39. segnis in a

Herculanean
selectus

repere, repsT,

reptum

see 46.

papyrus.
seligo,

3-

seligere,

selegl,

like

restinguo,

-ere,

reslinxi,

restinctus

lego.

see distinguo.

Selinus, -iintis

Gr. ^eKivovvTos.
*ses-mestris,

rlxa

face, to the Romance.

: :

semestris

for

*sexmes-

rqscidus

from

rar.
vi.

tris ; see 89.

Roscius

Roscio, CIL.
Plutarch,
15.

2060, 5
3
;

semuncia
septunx
:

from semi- and uncia,

for sesqui-

"Puxtkios,

Cicero,

from uncia.
:

Pompey,
sychius.

sescuncia

and uncia.
for sesqui-

rostrum: from rodo ; pQarpov,

Rostra
:

He-

sescuplex,
-plex.

sescuphis

and

from rostrum.
:

Sesostris

Secrwa-rpts.

Roxdna

Gr. 'Pw^dj'Tj.

sesqui-

=
:

semisque-.
for semis tertius.

rilcto : ace. to

the
p.

Romance
u.

(Grober,

sestertius

Archiv,
also to a

v.

370), which points

Sestius

Gr. Htjcttios, in Cic. adAtt, 2 et passim ;

1irj<TTia,

form with

vii. 17.
iii.

CIA.

ructus

like ructo.

1450.
:

rursum,
riiscum
rusticus
ix.

nlrsus

for

*re-vorsum,

Sestos, Sestii

Gr. S^crros, S^crrtoi.

i.

*re-vorsus.
: :

Stgnia

Seig[nia, CIL.
:

n.

ace. to the

Romance.

signum and signtcm

xiv.
;

seignvm, CIL.

4012

from rus ; Rvsticvs, CIL. u ace. to the Romance. ;

tions de
slgjiifico,

4270 sIgna, Boissieu, InscripLyon, p. 606 see ^S.

;

signo
:

like signum.

sinciput
sancio,
sancire,

for

semi

caput,

i.e.

sinci-

sdnxi, sdnctus

see

put
tion

for *senciput,
;

by

vowel assimila-

46.3.*). sarculum : like

Sdrsina
sceptrum
seised
: : :

see 90.
:

sdrio.

sinistrorsus

for *sinistro-vorsus.

Sassinas in an inscription.
Gr. aKijirrpov.
;

sistrum

Gr. creiaTpov.

sobrius : o in Plautus, Miles Gloriosus,

d [esc] Iscentem, Monumentum Ancyranum, v. 28.

see 49
:

812

36. 2.
:

Socrates

Gr. 'ZtoKpdrrjs.
:

scrtdo, -ere, scrips!, scriptus

see 46.
1
;

solstitium

from

sol.

scrIptvm, CIL.

vi.

201

con-

Sdphron

Gr. adxppuv.

WORD-LIST.
sospes
:

65
:

Gr.

2cD(T7rij,

CIA.

iii.

1161

el

Telmessus

Gr.

Te\/xr)<T<r6s.

passim,
sospita, sospito
:

Temnos
like sospes.

Gr.
:

T?;/ij'6s.
Tepfj.r)<T<r6s.

7'ermessus

Gr.

sldgno

like stdgnum.
:

teruncius

from uncia.
Gr. dearpov. Gr. Qpy<r<ra.

stdgnum

see 38, end.

thedtrum
Thressa
Tillius
:
:

stilla : i ace. to

the
:

Romance.
like stilla.
:

stillicidium, stillo

tIllivs, CIL.
tinxi,

vi.
:

2043.
see
46.

stringo, -ere, strinxi, strictus

ace. to the

strinxi tingo, -ere,

3-

tinctus

Romance.
:

*)
:

struo, -ere, striixi, structus

3. b)
;

see 46.

traclim
trdcto
:

like tractus.

u
:

also ace. to the

cf.

Romance.
strVctok,
xii.

like trdctus.
-ere, trdxi, tractus :

striictor

like structus ;

traho,
3-

see 46.

CIL.
3-4-

x.

708

ace. to Gellius,

*).

Trapezius,
: : :

-untis

Gr.

Tpa7ref"o0y,

structiira

like structus.
like lux.

-oOl'TOS.

sublustris

triform is
trilibris
tristis :
;

fro

m forma.
;

substructio
suesco
:

like structus.

like libra.

as in suevT, suetus.
:

trIstior, CIG. 6268

i also

sugo, -ere, suxi, suetus

ii

see 46. 3

ace. to the

Romance.

ace. to the
:

Romance.
46.

trulla

for

truella.

The Romance
Romance.

suTllus

from suinus.
also ace. to the
:
:

also points to u.
trilcta :
ii

silmo, -ere,

ic

sump si, sump/us : see Romance.

ace. to the
:

tubilustrium

like lustrtim.

sumptus
surculus
surgo,
rego.

from sumo.
from surus.
surrectus
:

U.
like
iillus
:
;

-ere, surrexl,

from
Gr.
like
:

icntis;

VLLA, CIL.
vi.

ii.

1473
:
:

vlli, CIL.
d:\4vrj.
iiligo.

10230.

silrsum
sutrlna

for *su-vorsum.

ulna
iilva

: :

like sutor.
:
ii

Siitrium

in Plautus, Casina,

324

uncia

like iinus. like iinctus (ungo).

:

36. 2.
syllepsis
:

undid :
Gr.
<rvX\r}\f/is.

itndecim, iindecimus
decern,

from iinus and

T.
tango, -ere,
*).
tetigl,

iindeviginfi,
:

etc. :

like iinus.
:

tdctus

see 46. 3.

ungo, -ere, tinxi, iinctus

3-

see

46.

*).
ii

tdctio : like tdctus.

r^, -ere, ussi, iistus ;

Participle
36.
for the

in the Perfect

Tartessus
tdxillus
:

Gr. Tapr77(7(r6s.
iii.

ace.

to

the

Romance
s.

ace. to Priscian,
:

Tecmessa
tectum
:

Gr. T^K/xrjacra.

iispiam

: : ii

in ussi, see 53 like usque.

like usque.

V.

from /^.
/&w, tectus
:

iisquam
see 46.
3.

&0 J,

-<?r<f,

b)

usque

ace. to the

Romance.

66
ustnna
usiirpo
: :

HIDDEN QUANTITY.
like ustus.
usii

victus
also
villa
:

from

vivo.

The
vi.

Romance
;

rapid?
V.

shows

i.

vIlla, CIL.
points to
:

9834

the Ro-

mance
CIL.
ii.

i.

vallum,

vdllus
;

Vallari,

vlndeniia

from vinum and demo.

:

4509
xix.

also

Vallivs, Vallia, CIL.

Vipsdnius

vIpsanI, CIL.
vi.

vi.

12782

4039.
:

vIpsania, CIL.

8877

BeL^dvios,

vdlldris

see vallum,

valid

see vallum,
:

CIG. 5709. VTpstanus : vIpstanvs, CIL.

vi.

vasculum
vastus
to a
:

like vds.

the Teutonic languages point

and frequently 5837, b; CIA.

viscus
:

Ovei\pTavov
iii.

2039 CIG.

621.
vi.
:

long root vowel.

'
'
:

vIsceris, CIL.

1975.
;

Gr. Ovtjktls. Vectis, Isle of Wight grandis. vegrandis : from ve- and

vivo, -ere, vixT,

victum
xiv.
;

VEIXIT,

CIL.
3449

see 46. 3 vIxiT, 2485 ;

vi.

Veldbrum: a in Plautus, 483 36. 2.

;

Curculio,
originally
51. I.

CIL.
cited

ii.

vIctvro, CIL.
in

12,562;

f3eii;iT

an

inscription

Vendfrum: the suffix is the same as -dbrum ; see

vendo
: :

by Eckinger Worter in

(Ortkograp/iie

lat.

Griech. Inschrif-

from venum and

do.

ten, p. 43).

vermis

from ver.
:

Vopiscus

Gr. Qvo7T?<tkos

vopIsco,

vestibulum

ve -f stabulum ? Cf. pro-

CIL.

x.

4872.

stibulum.

vestigium

ve

+ steigh-?
Ovrjcrrlvot..

X.

Ve stint

Gr.
:

vexillum

vexillo, CIL.

Xenophon,
3167; CIG. 4483,
xii.

-ontis

Gr. ^evocpQv, -Qvtos.

Byzantine Gr.

pftiWa
;

Z.
also
ace.

ov7)i\\aTi(Cb)<TLv

to
zoster
;

Priscian,

iii.

36.

Gr. facrTtip.

53.

Words whose Hidden Quantities are

:

in Dispute.
-xi

agmen

a Marx

see 39.
etc. :

that
;

Perfects in

have a long
state that

agndtus, agnotus,

a Marx

see

vowel before the x only when the

vowel
is

38. agnus : a
allicio
:

he does not
to

ace. to

many
in

see 38.

every
e

is

long before

-xi.

More-

some

scholars

mark the
allexT,

of

over,

little

weight

is
;

be attached
(ix.

the Perfect long

pellexl ;

illexT,

to this testimony

for in the para-

and

likewise

in

-spexT

graph immediately preceding

27)
Priscian
his authority to

{aspexT, conspexT, etc.}, JlexT, pext,

vexi.

lends the weight of

This marking rests upon a

ix.

such forms as traxi,

statement of Priscian in

28.

But

mansi, duxT, which certainly had

a

Priscian in this passage simply says

long vowel in the best period.

DISPUTED WORDS.
Osthoff
p.

6/
vi.

{Geschichte

des

Perfects,

corvinvs,

2041
vi.

orfito,
;

vi.

227) and Brugmann (Grundriss der Vergleichenden Grammatik, ii.

p.
1 1

82) support e in Perfects

of

3203 ; Narbonensis, XII. 3163; HGRT[0S, vi. 9493; COHORTxii.

353 bone,
;

Cordiae,

22,915

Nar-

this type

by arguments drawn from

;

[is, vi.

2993
;

Fort [is Fortvnae,

vi.

comparative grammar

but the evi-

vi

9493

fortvnata,
justify

7527.

dence does not warrant a positive

conclusion in their favor.
"

these
ings

sporadic

inscriptional

Yet mark-

hardly

our

assuming
servus,
;

alliu??i

a Marx; see
:

88. I.

arvum,
vtrgo,
i.

drvalis,

libertus,

amygdalum

y Marx,

without citation
240)

etc.,

for the classical speech

to area.

of evidence, Grober (Arc/ziv,

and the same applies

p.

See

and Korting (Worterbuch) give_y. anxius : a Marx, Brugmann, Sommer, and others ; see 46. 5. end.
Appulus, Appalia
:

Seelmann, Aussprache des Latein,

91.
I.

Arriins

Marx. Apulus,

Marx; see 88. : A arvum, arvalis : see area.

ascendd, ascribo,
etc. :

Apulia are the better spelling. aprugnus : ii ace. to many; see

area
:

a Marx

see

38.

48.
ascia
:

this

word occurs with the apex

in

a Marx
:

see 89.

(arcae)
whether
fies

Boissieu,
it

Inscriptions
is

Asclepiades

Marx.
exi

de Lyon, p. 279, but

this

doubtful

Asculum
Lewis

Marx.
:

single

instance justi-

aspicid, -ere, -exT, ectus

;

Marx and
if

our recognizing the a as long.

root arc-,
'

see above under allicio.

for

The

hold, confine,'
is
;

originally a short vowel, as

had assus : a Marx and Lewis, as shown *drsus, which is improbable.

*arceo
Osthoff,

See

by coerced
position
it is

(for

* co-arced}

Geschichte des Perfects, p.

would have retained the a

;

in

com-

545astus, astutus
:

see 72. Nevertheless undeniable that a tendency ex-

a Marx, as

if for

*ax-

tus, etc. ; see 89.

isted in certain localities to lengthen

axis

ii

Marx, without warrant

i.

Cha-

the short vowel before r

sonant.
sulted in

a con-

risius (Keil,

11. 22)

and Diome-

In some words this repermanent lengthening of

des (Keil,
short a.
balbuttio
:
ii

i.

428) both testify to a

see 88.

short vowels in the classical speech,

e.g.

Marx

I.

forma, qudrtus (cf. quat- barritus : a Marx; see 88. I. tuor), orca, and probably in or do, benignus : i Marx and others
in
ordior,

see

orno.

In

case

of

other

38.
benignitas
:

words
local

we

simply

meet
of

isolated

Marx and

others;

see

manifestations

the
vi.

ten-

38.
bes, bessis
:

dency, eg. in arvali, CIL.

LIBERTIS, CIL.
X.

913

e in oblique cases

Marx

Henzen, 6490; 2150; vIrtvtis,

3523 vIrgo,

SERVILIO,

but in view of Quintilian's state-

CIL.
vi.

vi.

ment

(i.

7.

20)

that ss

was not

CIL.

449;

written after a long vowel in the

6S

HIDDEX QCAXTITY.
it

post-Ciceronian period,

is

much
fol-

cogndtiis,

cognomen, cognosco, and other

more probable

that the

word

lowed the analogy of

:
.

as, assis.

Ost-

words beginning with cogit- : the here is regarded as long by many;

but the evidence
is
;

hoff, Geschichte des Perfects, p. 545.

.:...::
.:

not sufficient to
see 38.

Marx

see SS.
if

I.

warrant this view

cabaUiis

a Marx, as

verb as for *co-amb-uro ; cf. cogito assumed i for *co-agito. warrant. no which there is : e Marx, after the analogy Probi confestim Camillns : T ace. to Appetu&x which latter is somemanifestos, of i ace. to MarKeil, iv. p. 197) what uncertain. tianus Capella (p. 66. 4, ed. EysMarx, on the conjungo, conjunx : senhardt).

from

an

a diminutive *ca6dnus for

comburo

Marx, who explains the

capesso

e ace. to Osthoff {Geschichte

basis of conivgi,

CIL.

v.

1066

vi.

des Perfects, p.

221),

who

regards

9914, which are too improbable to

capesso, facesso, lacesso, as originally aorists of the

Hcessit, etc.
ii.

merit acceptance.
conspicio, -ere, -exT, -ectus
:

same type

as habesso,
.

exi

Marx
alli-

Br ugmann

( Grim -

and Lewis
cio.

see above under

p. 1203), taking a different

view
t

of the formation, regards the

short.

as

cunctor
this

u Marx, whose treatment of

is

word
:

unintelligible.

cardials

possibly a,
'

if

from the same

'

da mm a

Marx;

see SS.

I.

root as car-ex,

sedge

(lit.

'

rough

despicio, -ere, -exi, -ectus : -exi

Marx and

plant' ?).

Lewis (E.L.D.)
:

see under allicio.

carrus, carruca

a Marx

see 88. see ^&.

:

1.

digitus: f

Marx and
discribo,
dis-

others; see 38.

disto,

Cassandra
::

Cass-

Marx

I.

discidium,

distinguo,

r~.

-ere,

cinxT,

ductus
i

Lewis

distringo ;

Marx and Lewis

48.

(E.L.D.) regards the

'

as short in

(E.L.D.); see
disco
:

cinxi ; likewise in -stinxl, -stinctus ;

tinxl, tinctus,

Marx, on the theory of comlengthening

;

and in

pitixT, finxT.

pensatory
* di-dc-sco)

{disco

for

The Romance languages seem

fect

to

see S9.
;

point to T in the Perfect and PerParticiple of all these words,

Italian
finsi,
cinsi,

distingud,-ere,-inxi,-inctus

see cingo.
dis-

For distinguo, see above under

cidium.

eg.

cinto ;
etc.

stinsi,

;:::to;
tions,

finto,

Inscrip-

duumvir :
(E.L.D.)
;

tl

Marx

and
1.

Lewis

moreover,

give

extInctos,

see 42.

cLvcrvs.

Dyrrhachium : y Marx, who cites the Grundriss, i. p. 501 f. Korting, modern name Durazzo. Worterbiuh, and Frohde in Bez- enormis : Marx and Lewis (E.L.D.)
See d'Ovidio in Grober's
;

zenberger s Beitrage, xvi.

classis
:

p.

193.

a Marx, on the basis of an

see norma. Erinnys : T Marx

cf.

88.
:

I.

assumed

etymological

connection

exstinguo, -ere, -Inxi, -Tnctus

tinguo.

see dis-

with ctarus.

DISPUTED WORDS.
/astigium
:

69
see

a Marx, on the theory of

lengthening
a
;

'semi-literary';

36.
iii.
.

fin.
;

compensatory
89.

see

C/

Grober
(

{Archiv,

139)

Korting
disdain
' :

WbrterbucK)
related.

Marx

fastus,

Marx, on the

compares
cus
hispidus
T

kirtus, with

which hir-

theory of compensatory lengthening

;

may be
:

89.
:

fesfinus, fesfino

Lewis and Marx,

of

cites
is

the

Marx and Lewis. Marx Romance, but the word

fin.

on

the

theory

compensatory

probably 'literary' in the Ro;

lengthening, as though for fendt- ;

see 89.
/estiica,fistuca
:

mance
(

see 36. 5

Korting
the
i

Worterbuch)
Ictus

regards

as

and

Marx, on the

short.
ictus
:

theory of compensatory lengthening

(see 89), as though for /erst-.
Jingo, -ere, finxi, /ictus
flecto,
-ere,
;

Lewis
1.

but the

Romance
others

points to
F/ignus
:
1

see ci?igo.

ace. to

Marx and
in

flexi

JlexT

Lewis

and

see 38.

Marx

see under allicio.

:

immo : Tmmo Marx,

and Immo, CIL.

view of Tmus

forsit, forsitan

Marx

writes fcrsit

iii.
1.

774.

The Ro-exl

and forsitan on the basis of the Romance. But Korting ( WorterbucJi)

mance
inspicio,

points to

-ere, -exT, -ectus :

;

Marx

interprets

the

evidence
0.

of

and Lewis
jubed,
-ere,

see allicio.
jussi,

the

Romance

as pointing to
:

jussus

jussus

fortasse, fortassis

a Marx, who

cites

Lewis.

The
is

only authority for u

ivssvs,
is

nothing valid in support.

in jussus

CIL.

vi.

77.

/ragmen
/rendo,
sus
:

a Marx and

many

others

But the apex here

weight.
at least

entitled to

no

see 39.
-ere,

The same

inscription has

/rendu!, /resus, or /res-

-essus
:

Marx
;

98. 2.

one other error in the use of the apex, viz. Annivs. In favor
of jussi

u Marx see 88. 1 garrio, garrulus : a Marx, who con/uttilis

we

find rvss[iT,
;

CIL.

xii.

nects with Gr. yapvio

see 88.

I.

and iov1930; ivssiT, iv. 25531 sit, CIL. i. 547 a, et passim in

inscriptions

Garumna : u Marx on
Gr.
(Fr.

the basis of

of

the

ante -classical

Tapovvas

but the

Romance

period.

The

simplest solution of
is

Garonne) points
:

to u.

the

difficulties

to recognize

an

gigno

i ace.
;

to

Marx and many

88. I.

ante-classical

jitsf,

which
i.

is

well
21,

others

see 38.
:

attested

by Quintilian in

7.

gluttio, gluttus

u Marx; see
;

and a

classical jussT.

The

shorten-

grallae

a Marx
:

see 88.
;

I.

ing occurs in accordance with the

I.

hallucinor
helluo
: :

Marx Marx; see

a

see 88.
88.
I. i is

principle explained in SS.

1.

In

view of Quintilian's additional

doubt-

state-

hircus
ful,

the quantity of the the

ment

that jussi

was the orthography

ss

as

which judgment

Romance words upon is based may be

of his day. and that

was not
(i.

writ-

ten after a long vowel

7.

20) this

70
is

HIDDEN QUANTITY.
almost a necessary conclusion.
in

pare nequeo.
to
e.

The Romance

points

The apex

CIL.

xii.

1930

is

then

a blunder, a result of the confusion

of just and jussT.
schichte

norma

Lewis and Marx, who con:

See Osthoff, Gep.

nects with Gr. yvdipifws.

des

Perfects,

532
ii.

ff.

nuncupo
ntisqtiam

u Marx and Lewis, who

Brugmann,
Fronde,
xvi. p.

Grundriss,

n 82;

connect with nomen.

:

Bezzenberger's
184.

Beitrage,

u Lewis
Marx,

see

usquam.

ostrum
;

who

connects with

Juppiter
lascivus

ii

Marx

see 88.

1.

austrum.
ace.
38.

: d Marx, on the basis of an Paelignus : t assumed etymology, which connects others see the word with the root las(ldr-) of Uaikivoi.
;

to

Marx

and

Gr. texts accent

larua.
llbertus
libertds
:

pannus
e e

d Marx
;

cf.

88. I.
:

Lewis Lewis

see area.
see area.

pecto, -ere, pexT,

pexus

pexl Marx,

: :

and Lewis
others;
pellicio
:

see under allicio.

lignum
litter a
:

i ace. to

Marx and
I

see allicio.
:

see 38.
i

perspicio
;

see aspicio.

Marx
:

see 88.

pestis

Marx, in accordance with an

see 89.
to

malignus

i ace. to

Marx and

others

untenable theory of compensatory

see 38.

lengthening

Matrona
Messalla

a Marx, without
a Marx
see 88.
:

citation

pignus

ace.

Marx and
see 88.
I.

others

of evidence.
:
;

I.

see 38. pilleus : i Marx

:

mingo,
end.

-ere,

minxi, mictum
;

minxi pingo

see cingo.
:

ace. to

Marx and Lewis

see 46, plango, -ere, planxT, planctus

pldnxi,
46,

pldnctus ace. to
:

many

see

mis ceo, mis cere, mis cm, mixtus

niixtus
ace.
to

i in

end.
:

Marx and Lewis. planctus

i

ace. to

many
:

see 46,

The Romance
Archiv,
bucli)
iv.

points to

(Grober,

end.
e

117; Korting, Worter- plector, 'be punished'

Marx, who

mitto,

compares ttXtJctctw. Ro- posca : Marx, who compares po-cnmance points to i a few suspicious lum ; but the root had also a reinstances of i longa occur, e.g. Diduced form po- ( 69) cf. Gr. mIssis, CIL. iii. p. 862 (shown by irorbv. Osthoff, Geschichte des Perfects, p. posed : Marx, on the theory of committere, mist, missus: the
; ;

526, to

be probably a blunder)
x.

pensatory
*porsco)
postulo
:
;

lengthening
see 89.

(posco

for

mIssione,
1585.

7890
:

remIssa,

xi.

Marx, as in the case of

I ace. to

Narbo, Narbonensis

a Marx
but

see

posed,

under area.
nescio, nescius
:

privignus
e

Marx and

others;

Lewis

com-

see 38.

DISPUTED WORDS.
propugndculum
others
;

71
see 89.
e.

ii

ace. to

Marx and

*terstd)

The Romance

see 38.

points to
testis, testor,
etc. :

pugna pugnax pugno pugnus : u ace. see 3S. to Marx and others pulmo : ii Lewis. The Romance points
;

testdmentum, testimonium,

Marx, on the theory of comlengthening

see 89.
testa.

pensatory
*terstis)
;

{testis

for

to u.

quousque

Lewis u ; see usque,

-exi
;

testudo

: :

e
1

Marx, as in
ace. to

respicio, -ere,

-ectus

-exi

Marx tignum

Marx and

others

and Lewis
Sallustius
:

see allicio.

a Marx.
others
;

see 38. tinguo, -ere, see

cingo.
torreo,
-ere,

tinxl,

tinctus

see

sagmen
39-

d Marx and
:

torrul,

tostus

tostus

salignus

Marx and
:

others

see

Marx, on the theory of compensatory lengthening {tostus

for
*tors-

38.

Sarmdtae, Sarmdtia
sescentl :

a Marx,

who

tus)

see
0.

89.

The

Romance

compares the form Sauromdtae.

ses-

points to

See d'Ovidio in Gr'ober's

i.

Marx and Lewis, on the

;

Grundriss,
(

theory of compensatory lengthening

see

Korting 520 ; WorterbucJi), Grober {Archiv, vi.

p.

89.

Marx compares
,

Sestius
is

129).
tressis : e

(for Sextius)

but e in that word

Marx

see

bes, bessis.

exceptional.

See Frohde, Bezzento

TuscT:

berger's Beit?'dge, xvi. 204.

sordes

ace.

Korting
basis of

Worterthe Rocites,

u Marx, on the theory of compensatory lengthening ( TuscT for * Tursci) see 89. The Ro;

buc)i),

on

the

mance

points to u.
ii

mance, but the only word he is Italian sorde, which is very

'literary'
;

Tusculu?n:

Marx

see Tusci.
etc. : ii

likely

ultra, ulterior, ultimus,

Lewis,

see 36,

5,

end.

on the basis of an alleged apex in

VLTRA,
Boissieu,
p.

Sphinx
spinter

:
:

i
1

Marx.
Marx.

Inscriptions

de

lyon,
basis

136.

But the apex does See

Lindsay,

stannum
the
stella
:
'

d Marx, on the
to

of

not

occur

there.

by-form,' stdgnum.
stela ace.

latin Language,

p. 595.

The Ro-

the

Romance
two

mance
urceus :
the

points to u.
ii

probably

the

form with

/'s

had

e.
:

strenna

suppartcm

Marx see 88. I. : ii Marx cf. 88.

; ;

I.
:
ii

Marx, who cites drca ; but Romance points to u. Marx iirna : ii Marx and Lewis. compares iirindtor ; but urna is
to

suspicio, -ere, -exi, -ectus ; suspiro

be

referred

to

the

root

arc-,

Marx
taxo
:

see 48.

On

suspexi, see

weak form
ur{c)na.

urc- ( 100. 2),

whence
if

allicio.

The
ii
;

Italian

urna,

a Marx.

genuine Latin
point
to
{testa

inheritance,

would

testa : e

Marx, on the theory of comlengthening

for

but
;

it

is

probably

pensatory

purely literary

36. 5, fin.

72

HIDDEN QUANTITY.
Lewis
;

tiro, -ere, ussi, iistus : ussi

but

vi.

2086

INViCTAi,

vi.

353.

But
to
;

Priscian (Keil,

i.

466. 6) gives ussi.

with a single exception no one of

these inscriptions can be

See under jubeo.

urtica
:

shown

u Marx
tiro,

and Lewis.
:

Marx
Lewis

antedate the

third century a.d.

compares

and
vexi,
;

I quite

agree with Christiansen

veho, -ere, vexi, vectus

see under allicio.

vescus
:

49) in the view that in the classical period

longis, p.

{de Apicibus et

Marx, on the basis of the

the
it

questionable etymology ve
victor, victus, victoria, etc.
:

esca.

i was short later, apparently, was lengthened.

;

T Lewis,

vincio,

-ire,

vinxl,

vinctus

vinxT,

on the
tional

basis

of repeated inscrip-

vinctus, ace. to

Marx and
;

others.

markings, such as vIctor,

viscum; T Lewis
points to
i.

but the

Romance

CIL. vi. 10056; 10115 ; 1058; VlCTORINVS, vi. IO58; VlCTORIAM,

CHAPTER
ACCENT.
See Brugmann, Grundriss,'\ 2 pp. 971
.

IV.

ff. ;

Stolz, Lateinische Grammatik?

ff.
;

pp. 98

ff

Lateinische Lautlehre, pp. 95

ff.;

Seelmann, Atissprache

des

Lindsay, Latin Language, pp. 148 ff.; Sommer, 15 Handbuch der Lateinischen Laut- und Formenlehre, pp. 94 ff.
Latein, pp.

54.

Accent
as

in general

is

the prominence of one special syllable

syllables of the

of a word

compared with the other

same word.
ways.

This prominence

may

manifest

itself in

three

different

Thus
1.

syllable
it

may be made prominent by

The
English and

'stressing'

it, i.e.

by

uttering

with a more energetic expulsory act on the part of the

lungs {stress accent).

this nature.
2.

German accent
it

are of

syllable

may be made prominent by uttering

Sanskrit accent were of this kind.

at a

higher

pitch than the other syllables of the

same word {musical

prominent,

accent).

The Greek and

3.

syllable

may be

quantitatively

i.e.

its

time

may be

greater than that of the other syllables of the

same word.

No

language was ever accented essentially on the quantitative

;

principle alone

but traces of the operation of

this principle are

noticeable at one stage of Latin accentuation.

Neither stress accent nor musical accent prevails alone in any

language.

As a

rule the

one constitutes the essential accentual

is

principle of a language, while the other

subordinate.
;

Thus

in

English

we

notice chiefly the stress accent

but the rise and

fall

of pitch also exists as a feature of the spoken language.

73

74
55.
i.

ACCENT.
The
character of the Latin accent seems to have varied at
the language.
Originally
it

different periods of

seems

to

have
accent
Latin

been a
rested

stress accent.

In the prehistoric period In

this stress

upon

the initial syllable of the word.

this respect

represents a deviation from the accentuation of the Indo-European

parent-speech.

In the parent-speech the accent was free,

syllable of a polysyllabic
{i.e.

i.e. it

might

rest

upon any

word.

Evidences
initial

of this prehistoric Latin accent

syllable) are seen in the

the stress accent on the

weakening of unaccented vowels and

in

the loss of unaccented syllables.

a) Vowel- weakening
:

Thus

exerced for *ex-arced; conficid for *cdn-

facid; existumo for *exaistumo ; ininiicus for *in-am~icus ; contuberrialis for % cdntabernalis ;
cecidi for *cecaidi

(caedo)

con-

cludo for *cdn-claudo

b) Syllable-loss
:

Manlius

for

Manilius.
;

reppuli for *re-pepuli

surpui for *sur-rapui

un-deci?n for *uno-decem.

2.

In course of time another factor seems to have become

viz. quantity.

operative in Latin accentuation,

Apparently a long

came to assume such prominence as to receive a secondary Thus peperci became peperci ; inimicus became ininiicus ; stress. exutuntamus became existumamus. Where the penult was short,
penult
the

preceding syllable

seems to have received the secondary

;

accent, as ex'istunio for existumo

conficiuntTor conficiunt.

Ultiinitial

mately

this

secondary accent prevailed over the primary

accent, and thus established the traditional accentuation of the

historical period, the so-called
'

Three

Syllable Law,'

by which the

accent

is

restricted to the last three syllables of a word, resting

if

upon the penult

Yet the
first

that

is

long, otherwise

upon the antepenult.

have always retained
regularly retained
in

syllable of Latin

words seems
;

to
is

a certain degree of prominence

for

it

Romance, while unaccented

frequently vanish.
3. It

syllables in the interior of a

word

has just been stated that in the prehistoric period the

CHANGES IN THE LATIN ACCENT.

Latin accent was a stress accent.
the classical period
is

J$

The
in

nature of the accent in

a matter of controversy.
their

The

ablest investi-

gators often differ diametrically

interpretation of

the

evidence bearing on
ogists
still

this point,

most of our leading German

philol-

holding that the Latin accent of the Ciceronian age was

stressed, while

French
it

scholars,

on the other hand, are inclined

This latter view has been

to

maintain that

was musical.

made

extremely probable by the discussion of Vendryes, Recherches sur

Vhistoire et les effets de Vintensite initiale en Latin.
Paris, 1902.

See also Johnson, in Transactions American Philological Association,

1904, pp. 65-76.

Still,

even those who advocate the theory of a musical accent

classical

for the

speech,

admit that by the fourth and

fifth

centuries of the Christian era the stress accent had

lished.
4.

become

estab-

Even were we
stress accent,

to
it

admit that the accent of the

classical

age

was a

would be clear that the Latin of that time

was not

as strongly stressed as English

is

and German,

for

example.

One
its

reason for this

found
in

in the accentuation of the

Romance
accent in

languages.

These,

the main,

retain

the

Latin

original position, but they generally agree in

showing a much

slighter

degree of

stress

on the accented
is is

syllable than exists in

English or German.
poetry.

More weighty
stress

the evidence of Latin

Here

the quantitative principle

the fundamental basis

of the verse.

decided

accent would have conflicted with

this to the extent

of obscuring the metrical character of the verse.

Moreover, we often find Latin words containing an unbroken suecession of long syllables,
is

e.g.

edicebatur.

A
Cf.

strong stress accent

inconsistent with such conditions, as

may be

seen from

the

strongly stressed

modern

languages.

Eng. inevitable with

Latin
5.

inev'itabile.

Attention has been called in the

final

Grammar,

6, 4,

to cases
to stand

where, by the loss of a

vowel, the accent has

come

76

ACCENT.
last

upon the

syllable of certain words.

Other instances of the

same

sort are disturbat for disturbavit ; niuriit for

is

munwit.

The

principle
nostras,

stated by Priscian
etc.,

(xv.

17-18).

Arpinas,

Satnriis,

CampHns,

are also cited

last

by the grammarians
as

as hav-

ing an accent

upon the

syllable,
etc.

though for Arpinatis,

Samnatis, nostratis, Campanus,

iv.

See, for example, Priscian

22.

Such forms
etc.

as benefacit, satisfacit, are

properly written

benefacit,
6.

Various Latin grammarians

who
1-3

support the theory of the

(e.g.
;

existence of a musical accent in Latin

Gellius,
vii.

Nigidius

Figulus,

in

Nodes
14

Atticae,

xiii.

26.

Audacis Excerpta, Keil,

2.

357.

ff. ;

Priscian, de Accentu,
(

5) recognize

an acute

'

and a circumflex

"

),

and

lay

down

specific rules for their

all

employment.

According to them, the acute stood upon

short

vowels, as nux, bene, veterem, and

penult, as fegibus.
It also

upon a long vowel

in the ante-

stood

upon

a long vowel of the penult

If the ultima

in case the ultima was long, as regis.

was

short,

long penult took the circumflex, as rege.

The

circumflex also

stood upon long vowels of monosyllabic words, z&flos.

But

it is

more than probable that these rules are merely an echo of the principles of Greek accentuation, just as the rules given for
syllable-division

by certain Latin

grammarians were probably

rules.

merely a learned fiction in imitation of the Greek

See

35-

CHAPTER

V.

ORTHOGRAPHY.
See BRAMBACH, Die Neugestaltung der Lateinischen Orthographie, Leipzig,
1868,
t

and the same author's Hulfsbuchlein fur Lateinische Rechtschreied., Leipzig,

bung 3d

1884

Georges, Lexikon der Lateinischen Wort-

formen, Leipzig, 1890.

56.

The orthography

of Latin words naturally varied at different

periods,

and even within one and the same period there was not

infrequently considerable discrepancy between different writers.

During the

classical era

relatively slight attention

was paid to

the study of the language,

and

as a result

we

notice the absence

of any recognized standard of spelling such as prevails in modern

languages.

This lack of a recognized norm compels us to resort

to other sources of information in order to determine the best

spelling for a given era.

Our manuscripts of

the Latin writers

unfortunately have been so altered in the course of transmission

from the

past, that they

seldom furnish trustworthy evidence.

A
the

few of the oldest give valuable indications of the contemporary

spelling;

but more often the Mss. have

full

been adapted

to

standards of a later age, and are

encies of the Decline.

of the errors

and

inconsist-

On

the whole, carefully cut official inscrip-

tions furnish the safest reliance.

The testimony given by

first

these

is

supplemented

for

the

post-Augustan era by the statements of

century a.d., devoted
questions.

grammarians, who, beginning with the

much

systematic

attention

to

orthographic

Many

points belonging here have already been anticipated in connection with the discussion of Pronunciation.
classes of

The
:

following special

words

call for further

consideration
77

78
57.
i.

OR THO GRAPH Y.
Words of
volnus,
-vos,

the type mentioned in Gr. 9.

voltus,

4, viz.

quom,
-quos,

volt,

volgus;
-uos,

Nouns and Adjectives

in

-quom;
;

-vom
;

-uom; and verbs in

-quont,

-quontur
spelling

-vont, -vontur

-uont,

-uontur.

This was the original

and continued

to

be the regular orthography down to

After that
it

about the beginning of the Augustan Age.

retained, particularly in special

was

still

words as an archaic reminiscence.

But as a

rule,

beginning about the 8th century of the city (BrugI

;

mann, Grundriss,
Language,
p.

2
.

662

Stolz, Lat.

Gr.

46
53

Lindsay, Latin
the following

299

Bersu,
:

Die Gutturalen,

p.

ff.),

changes took place

a) vol

+a

mute or a nasal became

vul,
for

eg.

vultus,

vulnus,
e.g.

But proper names show a preference

Vole anus,
b)
-vos,
Volsei,
etc.

the early form,

-vom,

-vont,

-vontur

became

-vus,

-vum,

-vunt,

-vuntur, eg. saevus, saevum, solvunt, solvuntur.

e)

-uos,

-uom, -uont, -uontur became -uus, -uum, -uunt, -uuntur,

eg. perpetuus, perpetuum, acuunt, acuuntur.

d)

-quos,

-quom,

-quont,

-quontur

developed

somewhat

at

variance with

the foregoing classes.

-cuntur, yielding,

They

first

became
;

-cus,

-cum,
quoni)
2.

-cunt,
;

eg. ecus (for equos)

;

cum

(for

relincunt (for relinquonl)

secuntur (for sequontur) .

during the Augustan Age, and
this class

This spelling established

itself

continued to be the standard orthography in words of

until shortly after the close of the first century a.d.,
1

when

-cus,

-cum,

-cunt,

-cuntur

became

-quus,

-quum, -quunt, -quuntur.

This change was the result of analogy.

for
1

Thus

in a

word

like ecus,

example, the preponderance of forms containing qu (equl, equo,

Examples are anticvm, CIL.
f.;

vi.

615. 4 b);
iii.

coevs, CIL.
a.
2.

vi.

8753

f.;

9264

Propincvs, CIL.

vi.

2408. 3;

5274

Cf. Gr. UpoirivKos

CIG. 6430.

Manuscripts also preserve numerous traces of such spellings.

Virgil's Acneid, see Bersu,

For examples occurring in the Palatine codex of

p. 88, N.


WORDS IN
equls, etc.) in

-QUOS, -QUOM, -QUONT, ETC.

79

time naturally produced the change from ecus to

Similarly, in the verb such forms

equus; and from ecum to equum.

as

relincunt, secuntur

ultimately

became

relinquunt, sequuntur

owing
quit,
3.

to the influence of the forms containing

etc.).

qu

(relinquis, relin-

relinquimus; sequitur, sequimur,

It is interesting to

note that the conjunction

cum remained
a paradigm

unaffected by this tendency.

Not forming part of

intact.
Its

containing

qu-ioxms,

it

remained
turn also

association
its

and
form
in

frequent collocation with

tended to preserve

unchanged.
texts,

The form quum, though

occasionally found

still

does not appear in Latin inscriptions or Mss. prior to the

6th century a.d. (Bersu,

4.

Die Gutturalen,
original
first

p.

44

n.).

What

has been said of forms in original -quont, -quontur,

to

applies

similarly

forms in

-(n)guont, -(n)guontur.

Thus an exstinguont became

exstingunt,

then later (after

analogy of the other forms of the same tense) exstinguunt ; so

exstinguontur developed through the
exstinguuntur.

medium

of exstinguntur to

58.

Assimilation
1

of

Compounds.
a)

Before
c,
f,

the

Final Consonant of Prepositions in

In compounds of ad, the preposition appears,

1)

regularly as ac-,
regularly as ad-,

e.g.

accipio.

2)

Before Before

e.g. adfej'd, adfiii.

3)

g,

regularly as ad-,

e.g.

adgredior; but as

ag- in aggei'd.

4)

Before

regularly as ad-,

e.g.

adloquor; but as

al-

in alligo, usually in allatus,

and often

in allectus.

5)
6)

Before

n, regularly as ad-, e.g. adnitor. p,

Before

regularly as ap-,
e.g.

e.g.

appello

but some-

times as ad-,
1

adpefo, adporfo.

On

this topic, see particularly the illuminating paper by

Buck

in the

Classical Review, Vol. XIII., pp.

156

fF.

Buck's results have materially

modified the position taken in the Appendix to

my

Latin Grammar.

80
7)

OR THO GRAPHY.
Before
r,

regularly as ad-,

e.g.

adrado, adrepo; but

sometimes as
8)

ar-, e.g. arripio, arrigo.

e.g.

Before

s,

regularly as ad-,

adsero, adsisto;
in assido.

but

as as- in assiduus,

and often

9)

Before

t,

regularly as at-,
e.g.

e.g.

attineo; but

some-

times as ad-,
10)

adtingo.
e.g.

Before

q,

regularly as ad-,
sc, st,

adquiro.

11)

Before gn, sp,

times ad-,

we

find

sometimes

a-,

some-

e.g.

agnosco, adgriosco; asfiiro, adspiro.

etc., is

Here

the spelling adgn-, adsp-,

purely etythe
actual

mological,

and
;

does

not

indicate

utterance

the

disappeared in these consonant

groups in accordance with the principle explained

in 105.
1.

12)

In

all

other cases ad was retained both in spelling

and pronunciation.
b)

In compounds of com-, the preposition appears

1)

Before

b,

p,

m, as com-,

e.g.

combibo, comporfo,

commoror.
2)

Before

c,

q,

d,

t,

n;

f,
;

j,

v,

as con-, e.g.

concilio,

conqiuro,

conger
;

co?ido,

contero,

con-

riascor ;

confero, co~nse?v

conjungo, convinco.
collatus.

3)

Before Before

1,

as con- or col-, e.g. conlatus or

4)
5)

r,

regularly as cor-,

e.g.
its

corrumpo, corripio.
(see 105. 1), e.g.

Before gn, con- dropped

cogridsco.

6)

For the origin of

see 89.
1
;

co-

in coniibiiim,

cdfiived,

etc.,

3.
(

c)

The

Preposition ex

= ecs) before
s to

f,

lost the c ( 105. 1)

effero,

and then assimilated

(cf. diffe?'d

f, e.g.

for e(c)sfero

for *disfero).
loss of the
is

Another form sometimes

s,

arises

by the

e.g.

ecfero,

ecfatus, etc.

This orthography
period.

found mainly in the archaic

ORTHOGRAPHY OF PREPOSITIONAL COMPOUNDS.

d)

The
i)

Preposition in appears,

e.g.

Before

1,

regularly as in-,
regularly as in-,
p,

inlatus.

2)

Before

r,

e.g.

inrumpo.
;

3)

Before m,

and b

as im-, e.g. imbibo ; imporfo

immor talis.
4)

In

all

other cases in- was both written and pro-

nounced.
e)

The
1)

Preposition ob
Is regularly assimilated to 00, of-, og-, op-, before
c,
f,

g,

and p respectively,
the

e.g.

occurro,

offendo,

oggero, oppono.

2)

Elsewhere

is

regularly retained in writing

s

and

in pronunciation, except that before

p.

and

t,

b had the sound of

Plautus, Terence,
this situation
;

See

27.

Our Mss. of
in

and Lucretius often have op(i.

but Quintilian

7.

7)

assures us
ob.

that for his time

good usage demanded

f) The Preposition per sometimes appears as pel before 1, e.g.pellicid. Elsewhere r is retained ; ptjero probably does not contain the preposition per.

g)

The
1)

Preposition sub
Is

regularly
c,
f,

changed
g,

to

sue-,

suf-,

sug-,

sup-

before

and p

respectively,

e.g.

succurro,

suffectus, suggestus, supplex.

2)

Before m, appears regularly as sum-,

Preposition trans

e.g.

summoved.

h)

The
1)

Is regularly retained before

r, t,

vowels and

b, c, f, g, p,

v, e.g. transed, transfero, transport!), transversus.

2)

Becomes
Becomes
trans-

tran-, often before

s,

and always before

n

sc-, e.g. tran-sero, tran-scribd.

3)

tra-,

before

j,

d,

1,

ra,

105. 2),

e.g. traicio,
is

traduco, trano.

Yet before these sounds

often restored byre-composition ( 87. 3).

82
59.

OR THO GRAPHY.
Seelmann (Aussprache des Latein,
adf-, adr-, ads-, inl-,

p.

61

f.)

thinks that
prepositional

such spellings as

inr-,

in the

compounds above
tion.

considered,

indicated the actual

pronuncia-

This pronunciation, however, he considers to have been

a faulty one, emanating from half educated persons striving for

special correctness.

Terentius Scaurus, Priscian, and Appendix

declare the

Probi

all

expressly

etymological spelling to be in-

correct in the type of words under discussion.

In

accordance

with

this, in

the Appendix to

my Latin Grammar,

the etymological

spelling was

rejected and the assimilated spelling was

recomof the

mended

as representing the actual speech of the

Romans

best period. conclusions,

The

investigations of

Buck no longer

authorize those

at least

not as

thoroughgoing principle.
is

In
in

many compounds,
yet others
is
it

the assimilated form

practically
it
is

unknown
regular.

the best period of the language.

In others

In

occurs occasionally.

But

in all cases the

orthography

probably to be regarded as indicating the actual pronunciation. 1

60.

Compounds
is is

of jacio.

As indicated
etc.

in Gr. 9. 3, these are

better written inicio, abicio,

That a

was pronounced
that the
verse.
first

after

the preposition,
of these words

made probable by the fact commonly used as long in

syllable

Possibly the

analogy of acid, deicio, reicio (where a

j would

naturally be pro-

nounced, even

if

not written) led to the omission of

in other

compounds
jacio, see

also.

For further discussion of the

compounds of
ff. ;

Mather,

Harvard

Studies, Vol. VI, pp. 53

ff.

Exon,

Hermathena, Vol. XIII, pp. 129

1

In the Appendix to

my

Latin

Grammar

it

was suggested that even

etc.*)

in the

case of unassimilated spellings (adf- ads-, adg-,

in pronunciation;
i.e.

there was assimilation

that adf-

was pronounced

aff-; ads-, ass-, etc.

WORDS OF DOUBTFUL OR VARIED SPELLING.

61.

83

List of

the Most Important Words of Doubtful or

Varied Spelling. 1

A.
abicio
:

arcesso

in early Latin also accerso.

rather than abjicio ; 60.

:

AreopagTta and Arlopagita.

ad in composition
adicio
:

58.
;

Areus pagus and Arius pagus ;

60.

CA

rather than adjicio

:

Alexandria,
artus, artdre
:

adolescens

see adulescens.

not arctus, arctare.

Adria

see Hadria.
:

arundo
{Neugeslalspelling
for

not hartmdo. not autor.

:

adulescens

Brambach

auctor

tung, p. 52) restricts this

to the noun,

auctoritas

not autor itas.

:

'young man,' and

of
adolesco

aurichalcu?7i

better than orichalcum.

the participle
adolescens.

writes

autumnus

not auctumnus.
B.

adulescentia,

adulescentulus

like

adulescens.

bacca

early baca
: :

1.

Aedui

see HaeduT.
:

balbutio
ballista

not balbuttio.
preferable to balista.
:

aeneus, aenus

better than aheneus,

ahenus.

balneum, balneae
:

balineum occurs

agnd sco and adgnosco

58. a).

in early Latin.

Alexandre a
for the

this is the correct

form behia
Later

also early Latin, bellua.

:

Ciceronian period.
is

beneficium
beneficus
:

rather than benificium.

Alexandria

found.

rather than benificus.

:

alioquT and alioquin.

benevolentia
;

rather than benivolentia.

allium: early a Hum

allec
:

88.

I.

benevolus

:
:

rather than benivolus.

bybliotheca also occurs.
bipertitus
fin.
:

not alec.
:

bibliotheca
3.

ancora

not anchor a ; 31. antemna : also antenna.

Antiochea, Antiochia
dria, Alexandria.
:

bipartitus

and
:
:

87. I.

Bosphorus

31. 3

like

Alexan-

bracchium
Britajmia,

brachium
etc. :
:

also occurs.

better than Britt-.

dnulus

not annulus.

Brundisuim

not Brundusium.

Apenninus and Appenninus. Apuleius and Appuleius : cf.

Apulia, Apulus.
arbor
:

Z%. I.

C.
caecus
:
:

not coecus ;

11.

arbos

is

archaic and poetic.

caelebs

not coelebs ; II.

1 This list in the main follows that given in Brambach's Hulfsbuchlein fur Lateinische Rechtschreibung, a book unfortunately much antiquated. The The standard whole subject of Latin orthography calls for new treatment.

followed in this
first

list is

the usage of the early Empire,

correct form
is

roughly speaking, the

to the
list.

century a.d.

The

given

first.

Words belonging

classes treated in 57-60 are, for the most part, omitted from the

84

ORTHOGRAPHY.
ae,

caelum and derivatives have

oe-; ii.

not

coniibium

: :

not conniibium

89.

I.

convicium

not convitium

25.

3.

caementum : not cementum ; io. 2. cottidie and cotldie : not quotidie. cothurnus and coturnus : 31. 3. caenum : see coenum. caerimonia and caeremonia : not ^<?rz- culleus, culleum : early culeus, ciileum ;

monia
caespes
:

io. 2.
/

88. 1.

not cespes

io. 2.
2.

#z

archaic
3.

quom ; never quum

caestus

not cestus ; io. not cetra

;

see 57.

caetra

io. 2.

cumba

also cymba.

cupressus : Camena : not Camoena ; II. cur : quor Augustan prewas the ra^ caussa
.

not cypressus.
is

ante-classical.

form
r <?w#
/

98. 2.
/

not >#

11.

D.
Cerialia.

Ceredlis

and
:

Cerialis ;

damma:
;

early

dama ;

88.

1.

ceterl : not caeterl ; 10. 2.

Cethegus

Cetegus

is

pre-Ciceronian

Ddnuvius : not Danubius. Cf.% 16. 2. Dareus : better than the later form
Darius.
Decelea
:

3i- 3-

circumeo and
claudo
:

circtieo.
is

better than the later form

cludo

rare

and the
;

result
2.

Decelia.
defatlgo,

of 'De-composition'
clip e us
:

see 87.
clupeus,

defatlgatio

also defet-.
;

better
;

than
6. 2.

the

deicio

rather than dejicio

see 60.

early spelling

delectus,

'choosing'; also

dilecttis.
cf.

Clytemestra
coclea

not Clytemnestra.

delenio

better than dellnio ;

:

and cochlea ; 31. 3. coenum : this (and not caenum)

probably the correct spelling.

deprehendo
is

also the contracted

90. form

deprendo.
derigo
:

com- in composition: 58. &). comissdrT and cdmisdrz.

the original form.

ever,

which is probably Brambach, howrecognizes two independent

also dirigo,

comminus
con- in

not cominus.
:

verbs

derigo

(de

-f-

rego) ,

'

to

comprehendo

better than comprendo.

58. b).
die-)
:

move
dirigo

in a particular direction,
(dis
-\-

and
in

compounds: conditio {con and root

ditio,

rego),

'to

move
I.

not con-

different directions.'

detrecto

also detracto ; 87.

conecto
etc.

and derivatives
:

not connecto,

dexter, dextra,

dextrum

also dextera,

dexterum
rather than conjicio ; 60.
coicio also occurs.

regularly dextera

when

conicio

used as a substantive.
dicio
:

form
:

not ditio
: :

25. 3.

conltor

not connltor.

dinosco
disicio

earlier dignosco.

conived : not connived,

rather than disjicio ; 60.

conjunx
contio

better than conjux.

coventio)
:

Duilius or Duillius.
concio

(for

not

dumtaxat
dipondius

not duntaxat ; 87.

earlier

I.

25. 3-

dupondius ;

6. 2.

WORDS OF DOUBTFUL OR VARIED SPELLING.

E.
eculus
:

85

of Caesar, and the Celtic point to

cf^^j.d).
;

Gendva.
60.

Held

rather than ejicio

:

genetlvus

not genltivus.
not genltrlx.

elleborus epistula
: :

better than helleborus.

genetrtx
glaeba
:

rather than epistola.

not gleba.
:

Erlnys
erus,

not Erinnys.
erills :

gndtus, gndta
etc.

this is the early form,

;

era,

not herns,

used also in poetry

ndta.

later

ndtus,

23.

Esqulllae,
llae, etc.

Esqullinus

not Exqul-

gratis
is

and grains.

The

latter

form

archaic.

Euander

not Evander.

exedra and exhedra.

existlmdtlo,

H.
:

existlmo

existumdtlo,
spelling
;

Hadrla,

etc. :

not Adrla,

etc. ;

23.

existumo are the early

6.2.
exsanguis, exsclndo, exscribo, exsilium,
exspecto,

Haedui : rather than AeduT. Hallcamdsus : not Halicarnassus.

hallucinor better than kdluclnor
88.
I
;

; cf.

ex

and other compounds of with words having initial s :

etc.

also dl-, all- ; 23.

better than exanguis, excindo, ex-

amnion : better than Amnion ; 23. harena : not arena ; 23.

haruspex
:

pecto,

rather than aruspex

28.

23.

haud : sometimes haut ;

F.

haveo and aveo

:

23.

faenerdtor, faenero
etc. ;

not fenerdtor,

hedera
helluo,

not edera ; 23.

helludtlo
1.
:

10. 2.
:

early

heluo,

etc.;

faenum
faenus
:

not fenum, nor

see faenerdtor.
etc. :

foenum

88.

Henna
fecundtis,

better than
:

Enna ;
23.

23.

Heraclea
etc.,

later Heraclla.

not foecundus,

her Cisco and erased :

heri
:

also here (a different formation).

etc. :

femina

not foemina
;

II.

Hlber, Hlberes,

not Iber,

etc.

fetus: not foetus

1 1.

finitimus
foetidus
;

earlier

-umus ;
:

6. 2.

23. hlems : possibly

HTlotae
:

also hlemps.

not fetidus ; II.

20. 2.
I.

not Helotae.
better than Ister
;

forensia and foresid

futtilis
:

Ulster
holltor,

23.

early fu tills ; 88.

holltorlum

see holus.
;

holus

rather than olus

23.

gaesw?i

not gesum
:

10. 2.

I.

garrulus

not gdrulus.
of the

Geneva

ace. to the evidence

Imb- in compounds: 58. d") 3). lm?n- in compounds: 58. d~) 3).

Romance
Arckiv,
ii.

(see Grober in Wolfnin's

Immo

not Tmo.
:

437)

but the best Mss.

Imp- in compounds

58.

d)

3).

86
inclitus

ORTHOGRAPHY.
and inclutus
and
:

not inclytus,

litter a;

better than liter a

littus.

88. I.

incoho and inchoo.

ingrdlis
inicio
inl- in
:

lltus
;

rather than
:

ingrdtils

cf.

gratis,

loquela

not loquella.

rather than injicio ; 60.

compounds: compounds:
:

58.

d)

1).

M.
maereo,
etc.;
I.

in primis, or imprimis: 58. d) 3).

z;/r- in
58.
:

maestus,

11.
:

etc.:

not

moereo,

d) 2).
6. 2.

intellegentia, intellego

see 87.

malevolentia

not malivolentia.

intimus

earlier

intumus ;
J-

malevolus : not malivolus.

mancipium
6^ 2.

earlier

?nancupium;
manufestus

jilcundus

not jocundus, since the

'

manifestos:
6. 2.

earlier

word

is

derived from juvo,

is

please

';

the form jocundus

false association

the result of
'jest.'

with jocus,
10. 2.

Judaea: not
jiiniperus
:

Jild'e'a ;

not jump ir us. the regular classical form.

: earlier manupretium 6.2. maritimus : earlier jnaritiwius ; 6. 2. Mauretania : also Mauritania,

manipretium

Juppiter
Jupiter
88.

maximus

earlier

mdxumus ;

6. 2.

was

the

early

spelling;

Megalensia and Megalesia

; 20. 2.

1.

mercennarius

not mercenarius.
I

K.
Kaeso and
Kaletidae
:

Mess alia
7#z7/ /

early Messdla ; 88.

Caeso.

plural

millia

(Afonumentum

better than Calendae.

in
legal

Ancyranuni) and zr/m (the usual

form)

kalumnia:
calumnia.

expressions for

minimus

6.
2.

also

minumus ;
and

6. 2.

Karthago and Carthago.

monumentum

monit?ientum

lacrima:

earlier
6.

lacruma
2
;

(archaic

muccus : earlier muctis ; multa : not mulcta.

multo
:

88. 1.

dacrumd) ; nor lachryma ;

lamina and lamna.

not lachrima

see multa.
:

31. 3.
;

milraena
syncopated

not miirena

10. 2.

lagoena: not lagena

II.

murra and myrrha.

N.
navus
:

lai?i7?iina, also

lanterna : better than laterna.

earlier
;

gndvus.
2.

Ldrentia (in Acca Z.)

lautus
:

not Laurentia.

ne, 'verily'

not w<7^/ 10.

:

better than lotus,

;

neglego, neglegentia
6. 2.

87. I.
:

legitimus : earlier legitut?ius

libet,

negotium, negotiator
etc.;

not negocium,

libens, libido: earlier lubet, etc.;

25. 3.

6.2.
lis:

but

stlis

in the legal phrase

;

stliti-

: not naenia ; 10. 2. nequicquam and nequiqua?n.

nenia

bus judicandis

104. I. b).

novicius

not novitius ; 25.

3.

WORDS OF DOUBTFUL OR VARIED SPELLING.

nunquam and numquam.
nuntio,

87

pomerium
etc.;

not pomoerium.
not Pontinus.
6. 2.

nuntius

not

nuncio,

Pomptlnus

25. 3-

ponlifex : earlier pontufex ; Porsenna and Porsena ; also,

0.

ace. to

obicio

:
:

rather than objicio ; 60.

Brambach, Porsinna and Porsina. prehendo and prendo.

oboedio

not
:

be did ; II.

prehim
proicio
0/$-/
58.

not praelum
:

10. 2.
;

obscenus

not obscaemts ; nor obscoe;

proelium
:

not praelium

II.

nus ;
<?&-

10. 2

II.

rather than projicio ; 60.

:

in

compounds: not
:

pro7nunturiur7i
ttiritim.

better than

promon-

e)2).

obsonium
obsdndre
:

also

opsonium

(oipooviop*).

proscaenium :

10.
2.

not

proscenium

see obsonium.

obstipescd : earlier obstupescd ; 6. 2.

proximus :

earlier

proxwmis ; 6.2.

obtemperd, obtined,
58. e) 2).

obtull

not

<?/V- /

Publicola: on the early forms Poplicola, Puplicola, see publicus.

dpi lid : better than tipilio.

opp- in

pilblicus (from pilbes,

1).
6. 2.

'

youth,'
')
:

'

able-

compounds 58. e) optimus : earlier optumus ;

;

bodied men,'
(early Latin)
lus ;

'

citizens

poplicus

is

from poplus
is

= popu-

Orcus

not Orchtis ; 31.

P.

3.

puplicus

the result of the

contamination of publicus and poplicus.

paelex

not pellex ;

10. 2. 10. 2.

pulcher: early Latin pulcer ; 31.

3.

PaelignT : not PelignT ;

paenitet
:

not poenitet ; 11.

2.

Q-

paenula : not penula ; IO. Parnasus ; not Parnassus.

parricida,
88.
1. etc. ;

quamquam and quanquam.

quattuor
:

not quatzwr.

earlier

pdricJda

querela: better than querella.

quicumqtce

better than quicunque.

Paul/us and Paulus.

quicqtiam and quidquam.

paulus

preferable to paullus.

quicquid and quidquid.

Quinctus,
QuTnctTlius

pedetentim and pedetemptim. pedisequus

p?jero
: :

Quinctius,
:

QmnclTlis,

not pedissequus.
;

these

are

the forms

not pejtiro

perjiiro

is

prob-

for the

Republican period;

under
etc.

ably a different word.

percontor, etc.: not percunctor,
etc.

the Empire, QuTnlus, Quintilis,

quom:

57.

perjiirus and pejilrus : cf. pejero.

quor : see cur.

quotiens and quoties.
1.

pessimus
pilleus,

earlier

pessumus

6. 2.

etc. :
:

early pTleus,

etc. ;

88.

plaustrum
plebs
Pollio
:
:

not plo strum.

R.

not pleps; 58. <?) 2). better than Polio.

raeda: better than re da ;

10.
2.

not rh-;

88
Raetia, Raett
:

ORTHOGRAPHY.
not Rhaelia,
:

etc.

.y?^"-

in

compounds/

58.

")

i).

reccidi (Perf. of recido)

not recidi.

sulpur and
31. 4-

sulphur:

not

sulfur;

recipero

:
:

earlier recupero ; 6. 2.

Regium
reicio
religio
:

not Rhegium.
;

summ 6o.

in

rather than rejicio

:

.//- in

compounds: compounds:
:

58.

")

2).

58. ) 1).

in poetry also relligio.

: :

suscenseo
susplcio
:

rather than succenseo.

reliquiae

in poetry also relliquiae. early Latin relicuos ;

:

not suspitio

20. 3. I. 5.

reliquus

57.

Syria

earlier

Suria

repperl (Perf. of reperio) reppull (Perf. of repelld)

:

not reperi. not repull.

taeter
:

T.
not teter ; 10.
2.

reprehends or reprendo.
res publico,
:

not respublica.
:

rettull (Perf. of refero)

not retull.

tanquam and tamquam.

te?nperi

rotundus

in

Lucretius
90.
S.

sometimes

(Adv.)

not temporT.

rutundus ;

tentdre

and temptdre.
:

Thalia

Thalea
:

is

pre-Augustan.
is

thesaurus
;

thensaurus

archaic.

saeculum
saepio
saeta
:

not seculum

Thrdx and Thraex

10. 2.

(6/)|).

saepes: not sepes ; 10.

:

tingo
2.

also tinguo.
:

totiens

also to ties.
:

see saepes.

not $/#/ 10.

:

trdjectus
2.

not trdnsjectus ; 58.^) 3).

:

Sallustius sdrio
:

not Salustius.

trans- in composition
trdnsicio

58. ^).
:

better than sarrio.

: :

and trdicio

rather

than

satura
scaena

also later satira ; not satyra.

trdnsjicio, trdjicio ; 60.

not scena
:

trdnsndre and trdndre

;

10. 2.

58. ^).

sepulcrum
31. 3sescenti
:

not

sepulchrum

Treveri
;
cf.

rather than Treviri.

:

tribunuius
rather than sexcenti.
tripartltus

not tribumtius
tripertitus
:

25. 3-

and

87. I.

setius

not secius.
:

tropaeum and trophaeum.

tiis :

singilldtim

not singuldtim.
;

rather than

Z#.y.

soldcium
sollemnis
stellio :

not solatium

25. 3.

not sollennis.
;

U.
I.

early stelio
:

88.

ubicumque
Ulixes
: :

better than ubicunque.

stlllicidium
stilus
:

not stilicidium.

not Ulysses, not humerus ; 23.

not stylus.
etc. :
: :

umerus
etc. ;

stuppa,

early stupa,

%'&'&.

1.

zlmidus,
etc. ;

umor,
23.

etc. :

not hzlmidus,

suddela
subicio

not suddella.
rather than subjicio
:

60.

subtemen
succ- in

not subtegmen.
58. -) 1).
; 88.
I.

unguo and ungo. unquam and umquam.

urged : not urgueo.

compounds:

succus

rather than sucus

Suebi

not Suevl ; 16.

2.

utcumque : better than utcunque. utrimque : not utrinque.

WORDS OF DOUBTFUL OR VARIED

V.
valetudo
: :

SPELLING.

89

vinculum and vinclum : 91. vinolentus and vinulentus.

Volcdnus : 57. a).
Volsci: 57. a).

not valitudo.
in poetry often vemens.

vehemens
Vergiliae,

Vergilius,

Verginius

not

Volsiniensis

57. a).

Virg-.

versus (versum)

early Latin vors-.

vertex: early Latin vortex,

verto
:

Voltumus : 57. a). Vortumnus : under the Empire Vertummis : cf. verto.
vu/gus
:

also

early Latin vorto.

:

earlier volgus ; 57. a).

vester

early Latin voster.

:

vulnus :
vulpes
:

earlier volnus ;

57 a),

vicesimus

commoner than vigesimus

victuma
:

earlier volpes ; 57. a).

earlier voltur ; 5 7. a)

sometimes also vicensimus.

victima
vilicus
: :

vultur
vultus

: :

earlier

6. 2.

earlier voltus ; 57. a).

not vlllicus.

'

CHAPTER

VI.

THE LATIN SOUNDS.

THE VOWELS}
Ablaut.
62.
i.

The Indo-European parent-speech, from which

the Greek,

Latin, Sanskrit, Avestan, Slavic, Teutonic, Celtic, Armenian,

and

Albanian languages are descended, had a vowel system of considerable regularity.

By

variation of the the root vowel, each

in three

monosyllabic root
different forms.

was regularly capable of appearing

Thus the Indo-European root

and another form

gen-, 'bring forth,'

had

also

a form gon-,

gn-.
'

The

different
;

phases in which a root appears are designated as

the general

grades'

while

phenomenon of

variation

is

called Ablaut or

Vowel

Gradation.

The

different phases

of a root taken together form

an

'ablaut-series.'

While ultimate conclusions have not yet been

it is

reached on the subject, yet

series as belonging to the

usual to recognize six such ablaut-

Indo-European parent-speech.

Of the
by a

three grades belonging to each series, two are characterized

fuller

vocalism than the third

'

these fuller phases of the root are

is

called

strong

'

grades

the third by contrast

gon-, cited

called the

'

weak

grade.
1

Thus gen-

and

above, represent the strong

Lindsay, Latin Language,
;

See Brugmann, Grundriss 2 Vol.

,

I,

78-549

Grammati&, 4-45 Lateinische Latitlehre, pp. 112-229; Sommer, Handbuch der Lateinischen Laut- und Formenlehre (pp. 34-336), to which work I am under the greatest obligations for the
chap, iv;
Stolz, Lateinische

material here presented.

2

While roots are usually monosyllabic, yet some

disyllabic roots are also

to

be recognized.

90

: ::

ABLAUT.
grades
;

91
loss of the e, is the
its

gn-,

which has been weakened by the

weak grade.

The

first
it

of the two strong grades gives

name

to

the series in which

2.

occurs.
as follows
:

The

six

Indo-European ablaut-series are

Series.
a-Series

Weak
1 (

Grade.

Strong Grades.
bhdbho-

8l
e.g. bfol e.g.

O
e-Series
I e.g.

dhi-

I e.g.

dhe-

dho-

*
6-Series
(

5
pp-

6
po-

{ I

e.g.

\ I
{ I
{

e.g.

po-

Vowel vanishes
e.g.

*
ag-

a-Series

\
{

q-

Vowel vanishes
e.g.

^
\
{

6
pot-

S-Series

pt-

P<*-

[
(

drk-

derk-

dork-

Vowel vanishes
e.g.

6
{ \
od-

8-Series

\
3.

Of

these six ablaut- series,

~e-,

it

will

be noticed that three are

long- vowel series (the a-,

series (the a-, e-,

and
.

o- series),

and three short-vowel

series often

and

o- series)

But the short-vowel

have, in addition to the forms given in the foregoing table, socalled

tog'

protracted forms
e-

'

of the root
'

e.g.

from the root

'

teg-

of the
;

series

comes the
sed-, the
'

protracted form

teg-

in fegula,
'seat.'

'tile'

from the root

protracted form' sed- in sedes,

is

63.

The

origin of this variation in the form of roots

attrib-

uted with great probability to accentual conditions prevailing in

the parent-speech.

Some
;

uncertainty

still

prevails concerning

details in the various series

but for practical purposes the above

(see

scheme

is
;

sufficiently

accurate

Brugmann,
p.

Grundriss2
Stolz,

i.

534
1

ff.

Lindsay, Latin Language,

253

ff. ;

Lat. Gr.,

9 represents an obscure short vowel, which developed variously in the

different

Indo-European languages,

as

a,

e, i, 0.

92

THE LATIN SOUNDS.

ff. ;

15

Lateinische Lautlehre, p. 157; Johnson's

Cyclopaedia,)

Article Ablaut).

Of

the different Indo-European languages

some

have preserved the Indo-European Ablaut with great

this is

fidelity;

notably the case with Greek and Teutonic.

In other lan-

guages the Ablaut has become

the latter class. the other

much obscured

Latin belongs to

Most Latin

roots appear in only a single grade,

in the course of

two grades having disappeared

the

development of the language.

gradation are preserved.

Yet some examples of the original

will

These

be considered according to

the different ablaut-series in which they occur.

- Series.

64.

The ^-series is by far the

:

best represented of any in Latin

it

embraces three sub-types

a)

The

e or o

is

followed
e.g.

by some consonant which

doc-,

is

not
(for

a nasal or a liquid,
*di-dc-sco)
*si-sd-o)
;

root dc-, dec-,

;

seen in disco

dec-et;

doc-eo
'

root sd-, sed-, sod-, seen in siao (for

'

sed-eo ; sod-alis,
('

seat-mate,'

table companion,' 'crony.'

The

root es-

to

be

')

has only the

is

weak grade and one of

etc.;

the

strong grades.

The weak grade

etc.

seen in s-im; s-unt,

the

strong grade in es-t; esse,

b)

The

e or o is followed

by a

liquid or nasal.

By

the loss of the

vocalic, de-

e in the

weak grade the

liquid or nasal often

becomes

veloping according to the principles explained in 100, 102.

Thus from the Indo-European root

gnatus
(for

gn-, gen-, gon-, the Latin has

;

gn-tus

;
;

see 102. 2), and gen-us

gi-gn-o, however,

no form with gon-

has been preserved

the

shows us another form of

weak grade.

From

the root
for

mn-, men-, mon-, the Latin has

^me-men-1, and mon-eo.

mens

(for *mn-t(i)s),

memini

Com-

pare also ex-cel-lo,

tors-).

col-lis (root eel-, col-); terra, ex-torris (root ters-,

;

Occasionally the liquid precedes

e.g.

from the root pre-,

;

prec-,proc-,we get posed (for *prc-sco, *porc-sco

prec-or, proc-us, 'suitor.'

100. 2

105. 1),

ABLA UT.
c)

93
i

The
Thus

e or o of the strong

grades was originally followed by

or u; in the
u.

weak grade the

e,

as usual, disappeared, leaving i or

originally
/

ei

oi

u
But, of these diphthongs, ei

eu

ou
1,

became

while the others

became

u,
:

except that oi (oe) has been retained in a few words.

root fid-, feid-, foid-,
(earlier foid-us)
;

Examples

seen in fid-es ; fido (for feid-o)

foed-us

root due-, deuc-, douc-, seen in duc-em, duco

(for earlier *deuc-o).

By disappearance of
develops from

the

e,

o of the strong grades,

e.g.

sometimes

in

the

weak grade,

mag-is,

ma(g)-Jes-tas,

ma(g)jns (for

-jos).

For protracted forms of the root

in the ^-series, see 62. 3.

Further examples of Ablaut in the ^-series are given in Stolz,

Lat.

Grammatik?

pp. 34

if.;

Lat

Lautlehre, pp. 157

ff.

Lind-

say, Lat.

Language,

p. 255.

^-Series.

65.

No

root shows

all

three grades in Latin;

9,

the obscure
1

vowel, develops regularly as a, but often appears secondarily as

in
1

accordance with

71. 2.

The

root dhp-, dhe-, dho-,

;

place,'

put/ shows the weak grade in con-ditus (for *con-da-tus

71.2),

etc.,

and one of the strong grades

the

in sacer-do-s

fanum

(for *fds-

num) shows
grade.

weak grade ;

fes-tus, the corresponding strong

Cf. also rd-tus, re-ri; sd-tus, se-men.

^-Series.

66.

One form
'

of the strong grade

is

seen in dg-o, the

'

pro-

(62.3) in ambages. The a may combine with i to produce the diphthong ai. An instance of this is seen in
tracted form

maes-tus,

weak grade

mis- in mis-er.

94

THE LATIN SOUNDS.

^-Series.

67.

The obscure vowel

s fa-men,

develops as

a.

The weak grade

is

seen in fa-teor ; the corresponding strong grade in fa-ri, fa?na.

Cf. also std-tus
;

Stator ; rad-ere and rod-ere exhibit the

^-Series.

two strong grades.

68.

Examples of

this

scantily

represented

ablaut-series

are

fo-dere, od-ium.
in fod-i, odi.

Of these

roots, protracted

forms ( 62. 3) appear

^-Series.
69.

The obscure vowel

appears as

a.

The weak grade

is

seen in damns, ddtus ; the corresponding strong grade in donnm,

dos.

Cf. also cd-tus, cos (for *cots).

70.
suffixes

Vowel gradation appears not only

and
in case-endings.

in roots, but also in

Thus

in

nouns of the second de0,

clension the suffix varies between e and

the two strong grades

of the ^-series.

The

suffix e is

seen in the vocative hort-e, and

originally existed in the locative horfi,

which

is

for *hort-e-i ; see

o, e.g.

126.

The

other cases originally had the suffix

Cf. also

hortus, hor-

tum, for a primitive hort-o-s, hort-o-m.

nouns of the type

of genus, generis, originally *gen-os, *gen-es-is, where again the

suffixes -es, -os

show us the two strong grades of the

^-series.

In case-endings

we have an

interesting illustration

of vowel

variation in the genitive ending,

e.g.

which appears
) /

as

-s, -es,

and

-os

familia-s ( 113); ped-is (for *J>ed-es

senafu-os (early Latin).

Vowel Changes.
w
a.

71.

Indo-European a 1

in syllables

which were accented

at the

time of the early Latin accentuation (see

in Latin
;

55) remains unchanged

at that period,

in syllables
:

which were unaccented

develops as follows
1

Including the a arising from Indo-European

? ( 62. 2, footnote).

VOWEL CHANGES.
1.

95

Before

two consonants (not a mute

final)

and a
e,

liquid)

and
for

before

r (not

a regularly becomes
for

e.g.

acceptus

*accaptus ;
i??ipertio

particeps

*pdrticaps

confectus

for
;

*confactus ;
reddere
for

for

*impartio ;

peperci

for

*peparci

*reddare.
2.

Before
i,

a
e.g.

single

consonant in the interior of a word a

for

becomes
*cecadi
;

adigo

*adago

tetig'i

for *tetagj;

cecidi

for

concino for * cone and ;

insitus

for

*insatus ;

redditus

for *reddatus.
3.
e.g.

Before
exsulto

/+ a
for

consonant (but not before

inculco
for

//),

a becomes
insulsus

u,
for

*exsalto;

*incalco

*lnsa/sus.
4.

Before
in

labials,

a becomes the sound which was represented

period,

by u

the earlier
;

and

later

by

(see 6. 2),
;
1

e.g.

occupo for *occapo

contubernalis

for

* contabernalis

mancufollows

pium
the

(later

mancipiuni) for *mancapium.

But when
z,

labial

the preceding appears always as

e.g.

accipio for

*dccapio.
5.

Before ng, a becomes

for

(through the
for

medium
;

of e),

e.g.

attingo

*attango

confringo

*confrango

compingo for

*compango.
6.

After
;

in

open

syllables

a becomes

e,

e.g.

variegd for

*variagd
7.

hie to for *kz'ato (cf. hiasco).

Short a before / in open syllables becomes

a)
b)
u, if
1,

the /
the /

is is

guttural,
palatal,
1

e.g.

exsulo for *exsalo.

if

e.g.

exsilium for *exsa/ium.

a.

72.
e.g.

a regularly remains unchanged in Latin

for *contactus.

in all situations,

mater ; contactus
By
guttural
i.

/ is

meant

before a,

0,

u, or a

consonant

by palatal

/,

before e or

9<5

THE LATIN SOUNDS.

e.

73.

i.

is

regularly retained in Latin:

r,

a)
b)
c)

Before

e.g.fero, confero, sceleris.

When

final, e.g. horte, age, agite.

e.g.

Usually before two consonants,

auspex.

scelestus, obsessus,

2.

becomes
Before

i:

a)

single

consonant

in

syllables

which were
(
;

unaccented by the early accentuation

colligo

55),
obsideo

e.g.

for
;

*collego ;

militis

for

*milefes

for

*6bsedeo
syllables

protinus for *protenus.

before
r,

But in unaccented
according to

the

is

retained,

73. 1. a,

eg. generis.
771

b)

Sometimes before n or
for *se77ip/ex

+a

consonant,

e.g.

simplex
*v~ige7iti

(from sem-, 'one'),

;

vigi7iti for

tinguo

for

*tenguo

qifinque

for

*quenque

(earlier
i,

*penque).
lig-nu77i

Before gn

original e also
dig7ius
for

becomes

e.g

for */eg-nu77t ;

*degnus (from *dec-

nus;
3.

94. 3). v,
e.g.

becomes o before

novos for an original *nevos

(Gr.
4.

ve/ros).

sve-

becomes

first

svo-

and then

so- ( 103. 5), e.g. Indo-

European
change of
socrus,
5.
'

*svesor to
s

*svosor,

whence
;

*sosor,

soror

(for

the

to

r,

see 98. 1)

*svecrus to *svocrus, whence

mother-in-law.'

becomes

before guttural /
e.g.

(i.e.

I followed

by

a,

o,

u
for

or

a consonant,

oliva for *e/aiva (Gr.

iXatpa);

volvo

*velvo (cf. Gr. peXvw).

e.

74.

is

regularly retained
die.

in

Latin

in

all

situations,

e.g.

rectus, correctus, correxi,

VOWEL CHANGES.
1,

97

1.

75.

1.

In unaccented syllables not

1),

final

becomes

before

a secondary r ( 98.

^
ri

cineris for *ciniris, genitive of a/fir.

So also in an accented
2.

syllable in sero for *si-so) *si-ro.

Before a consonant,
*crino

develops to r (see
first

100), then to
cerrio

er, e.g.

(Gr. KptVw)

becomes

*crno,

and then

so

*/m
3.

(Gr. rpis)
2

became
,

*lrs, later ter(s).

<?..

Final
;

becomes

w#/r

for

*mari ; ante
final
2

for *##'

(Gr. olvtl)
e.g.

sedi/e for *sedili ;

but sometimes

disappears.

animal

(for
^

*animalT)

calcar (for *calcart).

4.

Long

regularly remained

unchanged

in Latin.

6.

76.

1.

became u

in

accented syllables
(

a)

Before

n-adulteri?ium
;

20.

1),

<?..

uncus

for *oncos

(Gr. oyxos)
)

unguis for *onguis

<?..

(cf.

Gr.

oW)culpa for
//;

Before /-fa consonant,

*solcos
A7^to.

multa

for

molta ; sulcus for

(Gr. 6A.koY); pulcer for earlier polcer ;

But

this

change does not take place before

hence
2.
(cf.

^///x, mollis.

o also regularly

becomes before
for

e.g.

umbo

for

*ombo

Gr. o/x^aA.os)

numerus
etc.)

*nomeros

(cf.

Gr.

vo/xos).

few

exceptions (domus,
3.

remain unexplained.
vor-, vos-, vot-

About 150

B.C.

earlier

became

ver-,

ves-,
etc.

vet-, e.g.

versus, verto, vertex, vester, veto, for earlier vorsus,

4.

In unaccented open syllables Indo-European o seems to have

:

become
a)
1,

e.g.

novitas for *nevo-tas; armiger for

'

armo-ger ;

z-lico

for *in slloco,

on the spot
e,

'

indigena for *indogena.

;

b)

After

this o

became

e.g.

pietas for *pio-tas

societas for

* socio- tas.

98
c)

the latin sounds.

Before guttural
/ (see

71.

7,

footnote),

became
see 6.

u, e.g.

sedulo for *se dolo.

d)

Before

labials

this

became u
;

(later

1,

2), e.g.

Crassupes for * Crass opes

5.
e.g.

aurufex

for

*aurofex.
u,

In closed

syllables,

originally

unaccented o becomes
(cf.

onustus for *bnostos; euntis for *eontis

e.g.

Gr. idvros).

So also
;

before a consonant in final syllables,

filius for earlier filios

donum
-quom;

for

*donom
-vom

opus for

*oJ>os.

Final syllables

in

-quos,

-vos,

-uos, -uom, etc., retained

the
also

to

a con-

siderably later period; see

57.

1.

was

regularly re-

tained before
6.

r,

e.g.

temporis.
e, e.g.

Final o

became
1.

sequere for *sequeso.

For the rhota-

cism, see 98.

o.

77.
e.g.

regularly remains
vie tores, lice to.

unchanged

in Latin in all situations,

donum,

NX

u.

78.

u before

labials

became

about the close of the Republic

for earlier
syl-

(see 6. 2),
lacubus.
lables,

e.g.

lacrima for earlier lacruma; lacibus

This change regularly took place in unaccented

but by analogy
it

affected

some accented

syllables also,

e.g. libet

for lubet ; libens for lubens.

u.
79.

is

regularly retained in

all

situations, e.g.fumus, conjunc-

tum,

etc,

ai.

80.

1.

In syllables which, under the early accentuation (see

55))

were accented, original ai was retained, becoming, about

ae,

100
into

B.C.,

which,

in

turn, late
;

in imperial times,

developed
ai
arose

monophthongal sound

see

10.

2.

But

VOWEL CHANGES.
secondarily in Latin in a few words,
e.g.

99
aid,

maior?

Maius,

etc.

pronounced
2.

?naijor, aijo, etc.

In syllables which, under the

early accentuation (
1,

55),

were unaccented, original ai became regularly

*inquairo ;

e.g.

inquiro for
etc.,

existumo for *exaislumd

virfufi,

militi,

for

*virtutai, etc.; niensis, porfis, etc., for niensais, etc.

oi.

In Accented
81.
1. e.g.

Syllables.

oi appears in the oldest

monuments of

the Latin lan-

guage,

oinom.

But

it

early

began

to take the
B.C.

form
it

oe,

e.g.

coeravere.
develop
*loidos.

Somewhere between 200 and 100

;

began to
lildus

to u, e.g. utilis for *oitilis

unus

for oinos ;

for

This

change was complete

for a long

by

100

B.C.,

though

tendency existed
oi in formulas,
2.

time after that to use the original

e.g.

coiravervnt, loidos.

Yet

oe (even after the

:

change of oe to u) appears even

in

a few words
a)

As a

result of contraction, e.g.

coetus

for

coitus ;

coepi

for * coepi.

b) In the following special words

poena,

Poenus,

coenum,

foedus,

'

XxesXyj'foedus,

'

ugly,' foetor.

Yet by the side of

poena we have punio ; by the side of Poenus, Punicus

along with coenum? cunio.
is

Soramer suggests that the law

this

When

the labials/,^ or the labio- velar qu began

a word, the following oe was retained, except

followed.
c)

when

Moenia survived
differentiate

as

an archaism.

The form served

to

moenia and munia, which were originally the

same word.
1

This

is

probably the correct spelling for this class of words, not major, ajo,

Mdjus,
2

etc.

coenum

is

for earlier

*quoinom

( 103).

100

THE LATIN SOUNDS.

d) oi in early Latin appears before

in

auoi(J)os;

see

198- 3initial
v,

3.

After
;

oi

became

1,

e.g.

vicus for

*voicos

(Gr.

foiKos)

vinum

for

*voinom (Gr.

foivot).

In Unaccented
4.

Syllables.
ei,

Here

oi,

through the intervening stage of

(cf.

became

i,

e.g.

horti through horlei, from *hortoi

horteis,

from *hortois (Gr. ^dpToi?).

in

Gr. x P T01) j hortis, through Vestiges of the early form

are

preserved

poploe (=populi)

and

oloes

(=

oil is,

Hits),

mentioned by Festus.
ei.

82.

1.

Indo-European

ei is

preserved in the
e.g.
i.

earliest

monuAbout

ments of the Latin language,

200
B.C.
it

deivos, deicerent.

began to pass into

i

This circumstance led to the

e.g.

writing of ei for original

in

some words,

faxseis, for /axis

ei

peilum for pilum.

inal ei

In inscriptions the spelling

(both for origto the time

and

for 1)

was commonly current even down

of Caesar.
2.

After

/,

ei

became

~e,

e.g. levis for

*lewis (Gr. Xet/ros)

levi

for *leivt
3.

(from lino).

ei arose secondarily in

some words,

e.g.

eius,peior, Po?npeius,

the correct spelling, instead of the traditional ptjor, Pompejus.

These were pronounced

eijus, peijor, etc.

ui.

83.

This diphthong undergoes no changes

see 14.

au.
84.
1.

au

is

regularly retained in syllables which, under the

(
55),

early accentuation

took the accent,

e.g.

aurora, claudo.
to

In the speech of

common

life this

au had a tendency
permanently

become

an open o

(later close),

and in some words

itself

this colloquial pro-

nunciation even established

in the literary Ian-

VOWEL CHANGES.
guage.

01

Examples are
etc.

Clodius for Claudius ; plodo, in explodo,

implodo,
2.

In

syllables

which, under the early accentuation

u,

55),

remained unaccented, au regularly became

e.g.

includo for

Hnclaudo

defrudo for *defraudo.

eu and ou.
85.
1.

Primitive Latin eu and ou are nowhere preserved in the

existing

monuments of the Latin language,

dueo, luceo.

eu

first

became ou

(seen in early Latin dotted for *deueo) , and subsequently developed

to u,
2.
e.g.

Original ou

became u

directly.
e.g.

In a few instances we have eu arising secondarily,

neu,

ceu, seu.

Long Diphthongs.
86.
first

The name/

long diphthong

'

is

given to diphthongs whose

At,
far as
oi, ei, eu,

element consisted of a long vowel.

These, so

au, ou, ex-

isted in the parent- speech.

they were inherited

first

by the Latin, more commonly shortened the

element, after

which they developed according

to the
etc.

principles already laid

down
for

for original at,

ei,

oi,

au, eu, ou,

Examples are hortts

;

*hor tot's

81.

4),
(cf.

from original *hortois

Gr.

dative singular,
;

portae, from *portai

for *nocteu,

xW)

aur^ ra

f r * aurora

noctu

from *nocfeu.

So also probably dius in nudiustertius,

In the dative
-oi (Gr.
final

dius here being for *dieus, from original *dietts.

singular of ^-sterns, the
-a)).

Indo-European termination was

In

Latin this generally

became

-0,

by

loss

of the

element of the diphthong, but in our

earliest

Latin inscription
-oi,

(CIL.

xiv.

4123) we have perhaps a dative in

viz.

nvmasioi,

from

-oi.

In the parent-speech, these long diphthongs frequently

lost the

second element.
Latin
f'elo,

Thus
*feilo
;

ei,

eu gave

e.

Traces of

this are

seen in
;

for

rem

(earlier

*rem)

from *reim

diem

(earlier

*diem) from *dieum*

102

THE LATIN SOUNDS.

Re-composition and De-composition.

87.

i.

The

principles laid

down

in the foregoing sections for

the change of vowels and diphthongs in the

(originally)

unac-

cented syllables of compounds often seem to be violated.

appeto, expeto,
intellegd,

Thus

neglego occur

where the law demands

irregularities

*appito, *expito, negligo, intelligo.

are in reality not
result of

These apparent

due

to

any violation of the

i.e.

law, but are the

'Re-composition/
felt

the identity of the simple verb

it

was so keenly

that the language restored

in the

compound,
intel-

thus replacing the regular *appito, intelligo,

lego, etc.

etc.,

with appeto,

Other instances of the same kind are exaequo, conclaurevoco,

colloco,

sus,

exquaero,

interrogo,

where

phonetic

laws

would

demand

*exiqud,

conclusus,

exquiro,

*revico,

*collico,

*interrigo (

76. 4).

Many compound words

2.

are also naturally

to.

much

later

than the

operation of the laws above referred

Sometimes the form taken by a verb

e.g.

in composition occurs

instead of the original form,

etc. ;
'

cludo for claudo, after includo,

etc.

plied for pleco after implied,

This process

may be

called

De-composition.'
3.

Re-composition and De-composition manifest themselves

not only in connection with vocalic changes, but also in connection with

many

of the consonantal changes enumerated in the

Cf. e.g.

following sections.

trdnsduco as an illustration of Reis

composition.
Cf.

The phonetic form

(the

traduco, which also occurs.

also

sescenti

phonetic form;

105. 1), but sexcenti

(Re- composition).

Shortening of Long Vowels.

88.
1.

group of some twenty words exhibits shortening of

fol-

an accented long vowel, with compensatory doubling of the

lowing consonant,
viz.

Juppiter (for earlier Jupiter)

cuppa,

littera,

muccus, succus,

hallucinari, parricida,

bacca, gluttus,

gluttire,

VOWEL CHANGES.

103

bucca, da??ima, mutfire, stuppa, futtilis, Messalla, braccae, puppa,

allium,
stellio,

strenna, helluo, adieus, pilleus.

texts,

Many

of these

words often appear in Mss.,

a single consonant
;

and

inscriptions, written with

that

represents

the earlier spelling.

The

orthography of the Augustan Age has two consonants.

2.

The vowel was

regularly shortened in final syllables in

and

/; also in the original -or, -dr,

and

-er of Passive forms

and

in the
3.

Nominative endings

-ter, -tor, -sor, -or, -dl, -dr.

Words of

original iambic form, e.g.

mihi,

tibi,

sWi,

mo do,
giv'

citd, cedo,

often suffered
tibi,

permanent shortening of the ultima,

etc.

ing mVii,
('

modo, cedo,
')

The name

of

'

Breves Breviantes

shorts shortening
4.

has been given to this process.

is

In the interior of words a long vowel

e.g.

often shortened
;

before a vowel,

pled, taceo,

from
rel

*pled, *taceo
/

deorsu??i

from
(cf.

*deorsum
desum,

; fide'i

from Jidel ;

from ra

deesse,

deeram

d'efia).

Compensatory Lengthening.
89.
1.

In accented

syllables,

an

s before

a voiced consonant

is

often dropped with lengthening of a preceding short vowel,

sido
for

e.g.

*si-sd-o;

querela

for

*queresla;

egenus for *egesnos.

Often the consonantal group contains other consonants before

the
s,

which

first
;

disappear (in accordance with

for

105.
;

1), e.g.
fe??io

dla for *acsla

remus

*retsmos ; scdla for *scantsla

1. b. 2).
it

for

*tens?no ; cotiubo for co-snubo ( 104.

This lengthening

of the short vowel in compensation, as

were, for an omitted

consonant,
2.

is

designated 'compensatory lengthening.'

-11s

short vowel followed by

at

the end of a

word

is

lengthened with disappearance of the

3.

n, e.g.

equds for *equons.

Compensatory lengthening

is

also

claimed by

many

scholars
net,

for those cases in

which a long vowel has developed before

ncs

(J.e.nx),

e.g.

functus, junxi (cf-jungo); and where n discoiiived for *con-coiiived.

appears before

c, e.g.

104

THE LATIN SOUNDS.

Assimilation of Vowels.

90.

Vowels are occasionally assimilated

e.g.

to

each other in suc-

cessive syllables,

nihil for *nehil; nisi for *nesi ; soboles for

suboles
for

rutundus (chiefly in poetry) for ro fundus ;

;

tugurium

*tegurium {tego)

purpura
for

for -n-opcpvpa

and

in reduplicated

perfects, e.g.
for

momordt
etc.

memordi ;
is

toiondi for tetondi ; pupugi

pepugi

Assimilation

mainly restricted to short vowels,

but possibly we should recognize the assimilation of a long vowel

in filius,
lit.

'suckling,'

for

*fe-lius, root

dliei-

(see
;

86);

in

suspicio for *suspecio (protracted

form of root

spec-)

subtilis for

*subtelis {tela).

Parasitic Vowels.
91.
sitic

In the immediate environment of a liquid or nasal, a para-

vowel sometimes develops.

-do-,

Thus, especially in the suffixes

-tlo-, -bio-,

which become
;

-tulo-, -bulo-, -culo-, e.g. in vilulus,

stabulum, saeculmn
the
colloquial

yet the original forms continued in use in

e.g.
;

language and in poetry,

saeclum, vinclum.

Further examples are famulus (for */amlos)

(early Latin)
;

populus for poplus

e.g.

and

several
)

words borrowed from the Greek,

Aesculapius CAcrK\r)7n6s)

mina

(/avS)

drachuma

(Sporty/,??).

Syncope.
92.

In early Latin a short vowel following an accented syllable

Illustrations of this are
:

was often dropped.

auceps for *aviceps ;

;

auspex for *avispex ; ardor for *aridor ; reddo for re-d(i)do

aetas for aevitas ; prudens for *prov(i)dens ; valde for valid e ;

officina

for

* op(J)Jicina

anceps

for

amb{f)-ceps.

Syncope
etc.,

in

final syllables is

seen in ager for *agr(o)s, *agrs, *agr,

etc. ;

and

acer for acris, *acrs, *acr,

see ioo.

Apocope.
93.
i.

Final e and

often disappear,

e.g.

nee (for neque), ac

tot (for

(for atque), et (for *eti ; Gr. en),

aut

(for *auti); quot,

THE MUTES.
*quoti, *toti ;
e.g.
cf.

105
;

toti-dem)

ob for *obi

and

in neuter /-stems

animal

for

*animali; calcar
-1

for *calcari. for

But dissyllabic

z-stems change
2.

to

-e,

e.g.

mare

*mari.
;

Final o disappears in ab, for an original *apo (Gr. euro)

for *j?/^<?
{cf.

and sub
96.
1.

Gr.

t7rd).

On

the change

of/

to

<,

see

r.#

CONSONANTS.
The Mutes.

The Palatal and Guttural Mutes,

94.
1.

c, q,

g.

There are three

series of
'

k and ^--sounds in Indo-EuroPalatals/

'

pean, designated respectively as

Velars.'

Velars,'

and

Labio-

The

Palatals

were formed by approximating the tongue

to the roof of the mouth.

as k,

They developed

in
;

most languages
in Sanskrit

(in Latin regularly as k (V), g, rarely as q

s,

and

Slavic as sibilants,

sk, etc.)

The
in
all

Velars were formed further

back in the
k, g.

throat,

and develop

languages as plain gutturals,

labialization,
i.e.

The Labio-Velars develop with

w-sound
after
gic.

they have

parasitic

the k or g.

Latin represents these

sounds respectively by qu and

2.

Examples of the
:

different Gutturals are

Palatals

centwn, decern, dicere, socer ; ager, ago, genu, genus,

argentum.
Velars
:

qu

for c appears in queror, queo, but never

; ;

gu

for g.

cruor, cavere, canere

:

augeo, grus, gelu, lego.

Labio-Velars

quis, qui, etc.

sequor ; -que ; -linqiw

stinguo,

unguen.

The

labial

element

is

sometimes entirely

lost so that

qu
;

appears as
cf.

c, e.g.

stercus
(cf.

(cf. sterquiliniuni), socius (for *soquius

;

sequor); arcus
1

arquitenens)

-lictus (cf -linquo).

277-532

When
Latin

See in general
iv.;

Brugmann,

Grtmdriss2

Lindsay,

Language, chap,
Latitlehre,

Stolz, Lateinische
;

Grammatik?, 42-69; Lateinische

L.atei?tischen

pp. 232 291

Sommer, Handbtuh der

Laut- und

Fon?ienlehre, pp. 169-336.

106
initial,

THE LATIN SOUNDS.

gu
{i.e.

gv)

loses

the

g and becomes
as

v, e.g.

(g)venlre,

{g)vivos,
3.

(g)vorare.

-en-

and -cm- occasionally develop

salic-)
;

gn and gm,

e.g. salig-

nus from salix (root

*sec-mentum (sec-o).

dignus for *dec-nus ; segmentum for

The Dental Mutes,

95.
-tlo-, t
1.

t,

d.

regularly appears as
c,

t,

but in the Indo-European

suffix

became

e.g.

piaclum (whence piaculum)

;

for *piatlom;
this

saeclum (saeculuni) for *saetlom

-do-cro-,

vinclum,
see
e.g.

etc.

Sometimes

subsequently

(by

dissimilation;

no)

developed to
for *lavaclom,

when
;

a preceding syllable
in

had

/,

lavacrum

*lavatlom

quadraginfa, quadringenfi,

has not

developed

from

quadr-

probably

represents

different

word

see

183- 132.

is

regularly retained, but

becomes

/ in a
;

few words,

e.g.

lacrnma for dacruma (preserved in Ennius)

dingua (helped perhaps by association
with

lingua for early

in the folk-consciousness
;

lingere,
;

'lick')

solium for *sod-ium (Ablaut of sed-

see

64. a)

levir for *devir (dialectal (?) for *laivir ; Gr. 8d(f)vjp).

The Labial Mutes, p and

96.
1.

b.

regularly remains

unchanged; but in the prepositions

earlier p.

ab, ob, sub, b has

developed from an

The

original

forms of these words were *apo (Gr.

tion to Gr. e7n
final
;

airo),

*op-i (in Ablaut relaviro).

cf.

64. a)

*supo

{cf.

Gr.

By

loss of the
;

vowel these became *ap, *op, *sup

to

{cf.

sup-er, supra)

ap-

and op- are probably

be recognized in aperio and operio

but

before voiced consonants the

b

of ap, op, and sup regularly became

by

partial

assimilation,

e.g.

ab duce, ob

delicto,

sub

decessu,

whence the forms with

the
initial

b ultimately
for

became predominant.
;

In bibo
cf.

may be

an original p by assimilation
*quequo, then coquo.

Skr.

pibami.
que
;

By

assimilation also, an original *penque

first

became quin-

and *pequo became

THE MUTES.
2.
b, as

107
b, is

the descendant of Indo-European

in

by no means

a frequent sound

Latin, particularly initial b.

Examples are

baculum, balbus, brevis ; lubricus, labrum.

On

the late develop-

ment of

intervocalic b to a spirant, see 16. 2.

The Indo-European Aspirates

97.

in Latin.

In the Indo-European parent-speech the aspirates were

i.e.

almost exclusively voiced,

labio-velar)
;

bh, dh,

gh (both

palatal, velar,

and
as-

ph,

th,

ch were extremely rare.

:

These voiced

pirates developed in Latin as follows

1.

Indo-European bh became
a)

/at

the beginning of words, e.g.fdgus (for *bhagos ; Gr.

;

<f>r)y6s)

fd-ri (root bha;

Gr.
;

ffyrffiC)

fu-i (root bhu-

Gr.
b)

<f>vw)

fer-o (root bher-

Gr.

^w).
(for

b in the interior of words,

a/xcfMo)
;

e.g.
;

ambo

*ambho ; Gr.
;

orbus

(root

orbh-

Gr. 6p<avos)

mor-bus

(suffix -bho-).
2.

Indo-European dh became
a)

f'at

the beginning of words,

;

e.g.

fumus
;

(for

*dhumos;
;

Gr. Qv/xos)

femina

(root dhei-

Gr. Qr)-\v%)

forum

(root dhor-).
b)

Usually

^
;
'

in

the interior of words, Gr. /xeWos

'

e.g.
;

medius (for
aedes,
')
;
'

*??iedhios

cf.

for */ue#ios)

fire-

place,'

hearth
;

(root

tf/V/A- /

Gr. aWw,

burn

viduus

(root vidh-)
<:)
<

but

in the interior of words, if

tains
r,

an environing
;

syllable con;

e.g.

Tiber
;

(root oudh;

Gr.

ovOap)

rubro-

(root rudhro-

Gr. ipv$po<s)

and

in the suffixes -bro-

(for -dhro-; Gr. -Opo-), e.g. eri-brum.

/ in

Similarly before

the

Indo-European
b, e.g.

suffix -dhlo-

(Gr. -OXo-), dh
for

becomes

stabulum (with -bulum

-blum

see

90-


108
3.

THE LATIN SOUNDS.

Indo-European
gh.

Here we must

distinguish

palatal,

velar,

and

labio-velar gh.

A. Palatal gh.
a) h,

This became
initial
e.g.

when
words,
(root

or between vowels in the interior of

;

hiems (root ghim- ; Gr. ^ci/aw)

;

holus

ghol)

vehd (root
initial

vegh-)

anser (root

gha?is-)

has lost the

see
e.g.

23.
Jingo
(root

b)

before

and

after

consonants
;

dheigh-, with the infix n)

c)

gramen

(root ghra-).

/before

u, e.g.

fu-ndo (root gheu-).

B.

Velar gh.
a) Velar gh becomes regularly h, but
hostis (for *ghostis)
;

before

r,
;

e.g.

pre-hendo (root ghend-)

gra-

dior (for *ghrad-).

C. Labio-velar gh becomes,
1) f,

when
infix

initial, e.g.
;z,
<?.-.

formus

(for

*ghormos).
(s)m'gh-,

2)

gu

after
/z).

ninguit

(root

with

3)

z>

between
snigh-).

vowels,

e.g.

nivis,

nivi,

etc.

(root

The
98.
1.

Spirants,

j-,

/, h.

is

the most important

of the spirants, as

regards

phonetic changes.
vowels
('

An
{cf.

original s regularly

became r between
(cf. ges-si,

Rhotacisni''^), e.g. ger-o for *ges-o

ges-tus)

dirimo for *dis-emd

temples')
;

distingiw)

temporis for *tempos-is

{cf.

portarum

for

*portasom.

This

change took place

It

within the historical period of the language.

had been conBut the

summated before
cite

the close of the fourth century B.C.

earlier forms,

grammarians retained the tradition of the

such words as arbosem, pignosa,
etc.

and often
s to r

This change of
larva (root las-).

sometimes seems to occur before

v, e.g.

But

THE
this is

LIQUIDS.
is

IO9
secondary, having develis

only apparent
u,
cf.

v in such cases
the

oped from
*las-u-a);
fus-cus)
;

so

that

rhotacism
;

regular

lar-u-a
;

(for
cf.

Lar-es (for Lases)

(for

fur-u-os (for *ftis-u-os

;

Mener-u-a

*Menes-u-a)

la-ru-a and Mine-ru-a

are both found in Plautus.

2.
it is

Wherever

appears between vowels in the classical language

usually a result of the reduction of ss after a long vowel or

e.g.

a diphthong,

mm for missi

(i.e.

*mtt-si)

suasi for suassi

{i.e.

*suadsi)

haesi (for haes-si); causa for caussa.

The forms
Quintilian,
later;
i.

with double ss were current in Cicero's day

7.

{cf.

20), and occur occasionally

in inscriptions

much
e.g.

after short

vowels

ss was,

of course, always retained,

fissus, scisstis, etc.

3.

In a few cases intervocalic s appears to have resisted rhotae.g.

cism,

basium, miser, caesaries.

Possibly the s was retained in

i.e.

miser and caesaries as a result of dissimilation ( no),

order to avoid % mirer, *caera?'ies.
4.

in

By

analogy, the r resulting from rhotacism sometimes crept

e.g.

into the Nominative from the oblique cases,

honor
etc.).

(originally

honos) after honoris, honori (originally *honosis,

5.

Compounds, of
quasi,

course, often

show

intervocalic s after the

e.g.

analogy of the simple words of which they are compounded,

nisi,
6.

po situs

(after situs)

desilio, desino, etc.

h, see 23.

For the omission of the spirant

The

Liquids,

/,

r.

The Liquids as Consonants.

99.
ties.
1.

As consonants, the Latin

liquids exhibit few peculiariis

Their most important feature

tendency toward dissimir,

lation, as a result

of which / changes to

or r to

/,

to avoid the
in

repetition of / or r in successive syllables.

Examples are seen

-tlo- ;

the suffixes, -ari-, -cro- for

-ali-,

-do- (from

see 95. 1),

110
e.g.

THE LATIN SOUNDS.

exemplaris (to avoid * exemplalis)
is
\

lucrum (to avoid *luclum).

So caeruleus

for *caeluleics

{caelum), 'sky-blue.'

Sometimes r
r's in
;

disappears as a result of the tendency to avoid two

succes-

sive syllables, e.g. praestigiae for praestrigiae (praestringo)

sempi-

ternus for *sempe(r)-ternus.

The Liquids as Sonants.

100.

In

the

Indo-European parent-speech, whenever roots

el,

which, in their strong grades, contained

ol,

er,

or,

became

reduced

to the

weak grade (see

e or 0)

64. b), the / or r (by the

disappearance of the

became

sonant,
/,

i.e.

endowed with
These
:

vocalic character, usually indicated

by

r.

English has these

etc.

sounds in &?// (written

bottle)

centr (written centre),

Indo-European sonant
1.

liquids

developed in Latin as follows

ol,

/developed regularly
tollo,
i.e.

as

which often became ul (

tel-)
;

76.

1,

b),

e.g.

*tol-no
;

(for *tl-no, root

;

pulsus (for an

Indo-Eur. *pl-tos
Eur. *cl-tos
;

root pel-)

-cultus in oc-cultus (for

an Indoal,
e.g.

root eel-).

Before

vowels, /

developed as

palea for *pl-ea.

Sometimes the sonant

was long
(i.e.

in quantity

and then devel-

oped
vel- ;

as la or
cf.

al,
;

e.g.

lana
(i.e.

* viand) for *vl-nd, from root

;

vel-lus

latus

* flatus

104.

a), from root tel-;

falx
2.
cf.

(for */bc; root flee-, mflec-to), 'the curving tool.'

r developed regularly as or,

Tretpw,

e.g.

porta (for *pr-fd, root per;

Gr.

for *7rep-iw)

cord-is (for *crd-)

cornu (for crn-).

e.g.

In some words this or seems to have developed to ur,

(*crz'os, root cerv1

curvus
cer-,

cf.

cerv-ix)

curtus (for *cr-tos, root

cut

'

cf Gr.

Ketpo), for *Kepia))

e.g.

Before vowels r developed as ar,

card,

'

flesh

'

(for *cr-o,

from root

cer-, 'cut').
/,

Like the sonant

tity.

the sonant r was sometimes long in quane.g.

It

then developed as ra or ar,

;

stratus (for *sff-tos

root ster- in sterno)

cratis (for

*cr-tis);

armus

(ioxfrnos).

THE
3.

LIQUIDS.
itself.

Ill
This

In certain instances a sonant r arose in Latin

sonant r developed differently from the Indo-European r above

described, regularly

becoming

er.

Thus

in the

Nominative Sin;

gular of r^-stems, ager, for example, was originally *agros

by

Syncope (see

92) *agros became *agrs, whence by assimilation

er,

*agr(r), and by development of r to

in -ris

ager.

Similarly, stems

developed an er in the Nominative Singular.

*acrs, then *acr,

Thus

acris

gave

first

whence

acer.

Other instances of the

same change are

libertas for *libr-tas (root libro-), acerbus for

*dcr-bus; incertus for *incrtus (from *incritos,

cri-)
;

secerno for *secrnd (from *secrino,

i.e.

75. 2,

root
;

75. 2,

root m-)
Latin

agellus,

*ager-lus for *dgr-lus, from agro-.

Similarly sonant / sometimes arose

secondarily in

and

developed as
lus ( 92).

el, e.g.

catel-lus, for *catl-lus,

by syncope

for *catlo-

The

Nasals, m, n.

The Nasals as Consonants.

101.
1.

As consonants the Latin nasals exhibit few

peculiarities.

Before j,
2)
;

became

n, e.g.

venid for %gemjd (with labio-velar

g;

94.
2.

qnonia?n for *quoi7ijam.

On

the tendency of
20.

to disappear before labials,

and n

before dentals, see

2-4.

The Nasals as Sonants.

102.

In the Indo-European

parent-speech,

whenever roots
en, on,

which, in their strong grade, contained em,

om ;

became
vocalic

reduced

to the

weak grade (see

e or 0)

64. b),

the
i.e.

or n (by the dis-

appearance of the

became

sonant,
n.

endowed with

character, usually indicated

by m,

English has these sounds

in butn (written buttoii), rhythin, etc.

1.

These Indo-European sonant nasals developed

in

Latin
(for

regularly as

em and

en,

e.g.

septem

(for

*septni)

decern

112
*decm)
ped-em,

THE LATIN SOUNDS.

;

niilitem, etc.,

for

*pedm,

niilitm, etc.
;

memento

for *tne-mn-tod; tentus for *tn-tos (root ten-)

and

in the suffix

-men
2.

for

-mn,

e.g.

riomen.

Like the liquid sonants (see

ioo.

i,

2)

the nasal sonant

as na, e.g. antae

is

sometimes long, and then regularly develops

;

for *ntae; gna-tus {{or *gn-tos

root gen-)

gna-rus (for *gn-ros).

An

instance of
'

is

perhaps seen in (d)materies (for *dmteries,

'

from root dem-,

build

cf.

dom-ns)

The Semi-vowels
103.
1.

j, v.
e.g. tres

Primitive intervocalic

regularly disappeared,
for *monejo.

for *tre-es,
2.

from

trejes ;

moneo,

etc.,

When
;

following a consonant, primitive

/ became

i,e.g.

venio

for *venjo
3.

capio for *capjo; medius for *medjos.

Intervocalic v also often disappears,

accompanied by concontio for co{v)en;

traction of the vowels

tid;

which

it

separated,

e.g.

latrina for *la(v)atrina ; nolo {or *no{v)olo

;

jucundus

for

ju(v)icundus
all

junior

for *juvenior.
v.

Yet

this

law does not affect

instances of intervocalic
4.

av and ov in unaccented
;

syllables regularly for

became

u, e.g.

abluo for *ablavd

denuo

for

de novo ; impluo

Hmplbvo ;

in-

duo

for

*indovo

suns and tuus for earlier sovos and tovos, owing

to their frequent enclitic (unaccented) use.

5.

Before

0,
;

v regularly disappeared,

e.g.

sudor

for *soidos

from *svoidos

coenom

for

*quoinom

secundus for *sequondus

73. 4)
;

socrus for *svocrus (from *svecrus ;

somnus

for *sop-

nus from *svop-nos (earlier *svep-nos)

soror for *sosor ( 98. 1),

from *svos-or (earlier *svesor).

CONSONANT CHANGES.

113

CONSONANT CHANGES}
Initial Combinations.

104.

1.

Initial

consonant combinations often drop the

first

consonant.
a)

Thus
lost

Mute
1)

in

tilia

for

*ptilia

(Gr.

Trrekza)

sternuo

for

*psternuo.
2)

/in

latus lor *tlatus (root

(cf.

tel-)

^/

in Ju-piter for

* Djeupater

Gr.
(cf.

Zeu's

for *Aieus).
tz<z/z/.t

3)

in

/<2<r

for *gtact
;

yaAaKTo?), also in
;

for

gnattcs

notus for gnotus

yet the

appears in the
e.g.

archaic language and in compounds,

(for

ignotus

*ingnotus)

cognatus (for *con-griatus).

By

analogy cognomen takes a

b) s lost
1) before
cf.

(for

*comnomeii).

mutes
ster-no,
;

torus
stor-ea,

for *storus
'

(root .r&r-, stor-;

for *slego
'

mat

')

tego

(cf.

crreyco)
-

further, in

/zj*,

locus, latus,

broad,'

for

J-//7.S

stlocus, stlatus.

Early Latin

still

has stlocus

{e.g.

CIL.

v.

7381) and

stlatus, while stlis is regu-

larly

used in the phrase Xviri stlitibus judicandis.

i.

Cf
2)

also Quintilian,

4. 6.
:

Before liquids and nasals

ninguit,

in lubricus for *s lubricus ;

nix

(for

*sninguit,

*snix)

mirus

for

*smirus ; nubo for *snubo.

c)
2.
dfr-

v lost in /### for *vlana

radix for *vradix.

;

becomes
*j

in bellum

(and derivatives)
cf.

in
;

bonus and

(earlier

^>
'

Gr.

St? for

*S/:is)

bimus

for

*dvi-him-us,

of two winters.'

The

early forms dvel-

lum, dvoiiorutn are preserved in inscriptions, and as

1

See especially

Stolz, Lateinische

Grammatiffi,

62-69

Lateinische

Lautlehre, pp. 295-334.

114

THE LATIN SOUNDS.

archaisms in the poets.
bd-,

from the foregoing

e.g.

By the side of forms roots, we find also forms

bes), diennium.

with
with

dimus, des

(=

These are
is

archaic or dialectal, but dims, a classical word,

apparently for *dvei-ros.

Consonant Changes
105.
Simplification
of

in

the Interior of Words.

Compound Consonant Groups.

i.

In

the case

of groups of three or

more consonants, one or more

suscipio for *subs;

were regularly dropped in the formative period of the language

to
facilitate
;

pronunciation.
;
;

Examples are

cipio

asporto for *abs-porto

{cf.

ostendo for *obs-tendo

disco
;

misceo for
for

*mig-sceo
Hllucstris ;

Gr. ixiy-wfxi)

for

*di-dc-sco

illustris
;

suesco for *suedsco

;

ecferri for *ecs(ex)ferrz

;

pastus

for *pasctus

mulsi for *mulg-si

;

ultus for *ulctns

; ;

qiiintus for
;

*quinctus

arsi for *ardsi

;

tortus for *torctus

ursus for *urcsus

spars! for *spargsi

cere
;

biniestris for *bimens-tris

{cf.

poscere for spores;

Tuscus

for

*Turscus
for

Umbrian Turskwri)
urna
for

alnus for
{cf.

*a/snus ; fulmentum

fulc-mentum ;

*urc-na

urc-eus), qtiernus for *querc-nus.

Here

also

belong such compound forms as ignosco for *ingridsco ;

;

cognosco for *co?igndsco

2.

agnosco for adgnosco.

is

Often such simplification

merely preliminary to further

si,

changes,

regularly so when the groups

;
'

sm, sn
e.g.

arise.

Com'

pensatory lengthening (89) then takes place,

tar,' for

pihwi,

mor-

*pinslum, *pislum
by- form
*es/i'go ;

ala for *acsla, *as/a.

e.g.

The
eligo,

preposition

e as a

'

of ex arose in this way,

enormis for
estab-

*ecs/igo,

*ecsnormis, *esnormis
it

after e

became
Transduco

lished

in

compounds,

came

to

be used separately.

So
is

also

tra- arose, e.g. traduco for transduco, *trasduco.

result of
3.
'

the

Re-composition

'

( 87. 3).

Where two
owing

of three consonants in a group are a mute and a

liquid,

to facility of pronunciation, simplification does not

CONSONANT CHANGES.
take place,
tion are
e.g.

I15

astrum, antrum.

Other groups easy of pronunciasculps'i,

sometimes preserved,
to analogy.

e.g.

serpsl, planxi,

though

these may be due

ports, being

Compounds

like transcribo, trans-

much

later than the formative period of the language,

are not to be regarded as exceptions.

Assimilation.

106.
first

1.

Assimilation
is

is

designated as

'

regressive'
'

when

the
'

of two consonants
the second
is

assimilated to the second,

first.

progressive

when
2.

assimilated to the

By
be

regressive assimilation the following changes take place

to
ee, e.g.

occurro.

bg to bf to

gg, e.g. suggerd.

ff, e.g. suffero.
e.g.

bp to pp,
dc to

supportd.

cc, e.g.
e.g.

accurrd.
aggero.
(*sed-la)
;

dg to gg,
dl to
//,

e.g. sella

lapillus (*lapid-lus).

dn
dp
tc
ts

to nn,

e.g.

mercennarius for *merced-narius.

juss us for *jud-sus (root judh-)

ds to

ss, e.g.

to pp, e.g. apporfo.

to to

cc, e.g. ss, e.g.

siccus for *sit-cus (cf. sit-is)

quassl for *quatsi.

for

pm
nm
11s

to

mm,
mm,
e.g.

e.g.

summits

*supmus.
i.e.

pf'to ff,
to

e.g. officina
e.g.

for *opficina,

*opificina; see 92.

i.e.

gemma

for *gen-ma,
i.e.

'sprout' (root <?#-).

92.
s, e.g.

nlX.o

II,

ullus for *unlus,

to ss,

*unulus ; see

sometimes

which was

later simplified to

in adjec-

tives in -osus.

The

earlier

form was formonsus,

etc.,

whence

formossus
rl to
//,

{cf.

98. 2),formosus.
*sler-la
;

e.g. stella for

agellus for *ager-lus (see 100. 3)

{cf.

paullus (classical paulus) for *paur-lus

Gr. -navpoi).

Il6
3.

THE LATIN SOUNDS.

By
//,

progressive assimilation the following changes occur

e.g.

Id to

percelld for *perceldo.

Assimilation affects only a

;

primitive Id; in valde

(
;

valid e

92), for example, the Id

remains unchanged.
In to
//,

e.g.

pellis for *pelnis

;

In resulting

from Syncope ( 92),

is

as in ulna for *ulena

this
Is

volnus for *vol-inus,

not affected by

change.
e.g. velle

to

//,

for *velse ;

facillumus for *facilsumus.

rs to rr, e.g. ferre for *fer-se ; torr~ere for *tors-ere.

rs,

Secondary
1)

for

rtt,

as in versus for *verttos (see

108.

generally

remained unchanged, but

sometimes became
proversus)
4.
tial.

in the colloquial language such

an

rs

ss or s, e.g. prossus,

prdsus for prdrsus

(i.e.

Partial Assimilation.

Sometimes

assimilation

is

only par-

Thus
a)

labial nasal

may become
for

dental, or a dental nasal

may

become
mute,

labial,

owing to the influence of the following

e.g.

centum
;

*cemtum

ventum

for

*vemtum
whence

(root guem-)

con-tend o for *com-tendo,

etc.,

arose con- as a separate form of the preposition com-.

b)

voiced mute

may become

voiceless before a following

;

voiceless sound, e.g. ac-tum (for *ag-tum)

*scribsi.
c)

scrip-si for

The

labial

mutes p and b are changed

e.g.

to the correspond-

ing nasals before n,

somnus
;

for

*sop-nus (earlier
for
lit.

*svep-nos
{cf.

;
;

104.

2.

b)
for

Samnium

Sab-nium
'opposite

SabinT)

antemnae
(

*ant-ap-nae ;

fastenings,'

hence

yards.'

Metathesis.
107.
in

Metathesis or transposition
for

is

perhaps to be recognized
for *pat-rid ;

/undo

*fud-no

unda

for

*ud-na; pando

and

tendd for *te-tn-o (reduplicated present).

CONSONANT CHANGES.
Other Consonant Changes.
108.
i.

11/

An

original dt or tt

became

ss, e.g.

sessus for *sed-tus

passus

for *pat-tus.
s in

After a long vowel or diphthong such an ss

became
(

the Augustan era, though retained in Cicero's time

usus, earlier
i

98.

2),

e.g.

ussus,

for

*uttus

diznsus,

earlier

divissus, for *dii idtus.

In such forms as lapsus, pulsus, nexus

(=

nec-sus), fixus, s has not

developed phonetically, but has

etc.

simply been borrowed from words like sessus, fisus,

followed by r an original dt or
tt

When

became

st (instead of ss), e.g.

claustrum for * claud-trum

pedestris for *pedettris.

tt,

In syncopated
for *cedate

forms and compounds, dt simply became

(cf.

e.g.

cette

cedd), attendd ;

i.e.

these forms belong to a period in which

the change of
2.

dt, tt to ss

was no longer operative.

a parasitic

Between

m
;

and

/,

developed,

e.g.

exemplum

for

*exemlom

templum

for *tem-lom.

Such

developed also

between
~emptus

and

s in siimps'i, contempsl,

and between

and

/ in

and contemptus.

This phenomenon apparently was conis

fined to accented syllables, though hiemps

in

attested occasionally

good Mss.
3.

An
:

original -srf>r {fi

became

br.
t/i),

The

steps in this change were

//-,

first

from sr to

Eng.

then to

whence

br.

Examfor

ples are

sobrinus for *sosr-~inus {*sosr-, from % sosor, earlier form

;

of soror ; see 104. 2 b)

tenebrae for *tenesrae

(cf.
;

;
;

membrum

*memsrom ; funebris
*muliesris
(cf.

for *funesris

funes-tus)
98. 1)
;

muliebris for
;

mulier-is for *mulies-is

fibra for *fs-ra

cf.filum for *fis-lom.

4.

For

the

disappearance

of

before

/,

m,

n,

r,

b,

d, g,

combined with lengthening of

a preceding short vowel, see 89.

Il8

THE LATIN SOUNDS.

Consonant Changes at the End of Words.

109.

i.

Single consonants are usually retained.

is

Final s does

not become r phonetically, but

changed

after the
;

analogy of the
Final

r arising by rhotacism in the oblique cases n in the Nominative Singular of

;z-steiris

see 98. 4.

disappeared prior to
e.g.

the existence of Latin as a separate language,

in

homo

for

*hom-o{n)\ *caro(/l),
final

etc.

After a long vowel

or a diphthong,

d is found

in early inscriptions, but disappeared toward the

close of the archaic period.

Examples are
e.g.

Ablatives Singular of
for

the
for

first

and second declension,

also certain

praeda

praedad; Gnaivo
e.g.

Gnaivo d ;
etc. ;

Adverbs and Prepositions,

extra,

supra,

pro- for prod-, which latter appears in prodesse.

So

also se- for sed-,

2.

which

latter

appears in

seditio.

Geminated consonants
(cf. as-sis)
;

are not written at the

end of a word

thus as for *ass

so/el for
;

*fell,
it is

i.e.

*fels ( 106. 3)

far

for *farr, i.e. *fars ( 106. 3)

yet

probable that gemi(for *hodc),

nated consonants were spoken in these words, e.g.hocc

not hoc
;

so fai'r, fell, ass,

ess,

thou art

;
'

this last is the regular

form in Plautus.
3.

Groups of two consonants

plified,

at

the end of a

word

are sim-

a)

By dropping the second,

e.g.

mel

for

*melt; lac

for *lact;
is

os for *ost ; cor for *cord.

Iwfert,

volt, est,

the final consonant

retained after the analogy of agit,

etc.

regular exception to the

e.g. ops,

general principle
see 27)
b)
/

is

seen in

final

ps and x,

urbs (bs

= ps ;

7-ex, lex.

By dropping

the

first, e.g.

miles for *mzlets

pes for *peds ;

;

and

in final syllables in -ns,

as

agros

for

*agrons

turris

for

*turrins.
4.

Final

-nts,

-nds,

-rts,

-rds,

-Its

lost

the

/,

e.g.

mon{f)s,

fron(d)s, concor(d)s, ar(t)s, pul(t)s.

mitted,
e.g.

Final -nx,

-Ix,

-rx are per-

lanx, falx, merx.

CONSONANT CHANGES.
Disappearance of Syllables by Dissimilation.
110.

1 1

By

a natural tendency,
first

when two

syllables

began with the

e.g.
;

same consonant, the

for *denfitio

syllable

was often dropped,

d~ebi-

lifare for * debilitatare ;

;

calamifosus for * calamifatosus

for *fiortitoriu7n
; ;

dentio

portorium

veiieficus

for *veneni-

ficus ; voluntarius for *voluntatarius

semodius for *semi-modius.

CHAPTER

VII.

INFLECTIONS.

DECLENSION OF NOUNS AND ADJECTIVES}

A- Stems.
111.
(
(

In the Indo-European parent-speech there was Ablaut

62) in the suffix of ^-sterns.

The weak grade

of a,

viz.

66), occurred in the Vocative Singular.

a.

Elsewhere the

suffix

remained
112.
gular

Nominative Singular.
-a, e.g.

1.

The

original

Nominative Sin-

had

* porta.

But -a was shortened to -a before the

Possibly this shortening was
Singular,
2).

beginning of the historical period.

owing

to the influence of the Accusative

where *-am
relation of

regularly

became shortened

to

-am

88.

The

the Nominative to the Accusative in <?-stems, z/-stems, and /-stems

might

easily

have led to such shortening.

Cf.

the following pro-

portional representations
servos
friictus
:

servom
\

fructum
Tgnim

porta

portam.

ignis

Possibly the law of Breves Breviantes ( 88. 3), by which */uga,

*fera, *rotd,
etc.,

regularly

became fuga, /era,

Nominatives.
to

rota, etc., led to the

extension of -a for -a to
these influences
final a.

all

Either one or both of

may have operated

produce the shortening of

See, in general:
v.

Brugmann, Grundriss,
and
vi.
;

ii.

184-404;

Lindsay, Latin
75-88;

Language, chaps,

Stolz, Lateinische

Grammatik

%
,

Som-

mer, Handbtuh der Lateinischen Laut-

und

Formenlehre, 179-265.

120

AS TENS.
2. e.g.

121
of Masculine 5-stems,
'

The Latin
'
'

has developed a
'

number
optio,

agricola,

farmer

(probably originally

farming
'

')

cf. optio,

m.,

centurion's

assistant/

from

Other

languages exhibit

this

same

f, phenomenon,

choice,

selection.'
e.g.

Greek.

Thus

vcavids, 'a youth,'

-s

probably goes back to a lost

*veavCa, 'youth'
sig-

(abstract), the
nification
;

being appended to indicate the Masculine

many Greek Masculines in -ds, -775. The mediaeval Latin word bursa,/., meant company of students,' but subsequently became individualized to mean 'a student' (German
so further
'

Bursche)

so camerata,/., 'roomful of comrades,' later

'

comrade

'

(German Kamerad).
justice
('

Cf. also English justice (the quality)

and

magistrate'); Spanish justicia, by change of gender, also

covers these two senses.

113.
in

Genitive Singular.

The ending of the

Genitive Singular

Indo-European was
( 64. a), as

-s, -es, -os,

the different forms representing

Ablaut

the result, probably, of varying accentual con-

ditions of the parent-speech.

In the case of ^-sterns, the casea- of the

ending was

-s,

which united with the

stem and gave

It is

-as.

This termination appears in but a few Latin words.

served in familias,
familias,
etc.,

pre-

in

the

combinations pater familias, mater

is

but elsewhere

archaic, e.g. vias (Enn.

Ann. 421

Vahl.) , forfunas (Naevius)

114.

The
This

Genitive Singular in -ae goes back to an earlier -at

(dissyllabic),

which

is

found in the poets as

late as the

Augustan

Age.

termination -at apparently arose by appending the

-1

Genitive termination
porta-i.

of the ^-sterns directly to the stem,

ai,

e.g.

Whether

at

became

aeby

regular phonetic processes,

ae, is

or under the influence of the Dative

and Locative ending

uncertain.

115.

Dative Singular.
-ai.

The Indo-European case-ending of the

But
this

Dative Singular was

had already

in the

Indo-Euro-

122

INFLE C TIONS.
final -a
;

pean parent-speech contracted with the

ducing
116.
*-ai,

of the stem, pro80. 1).

whence successively

-at,

-ae ( 86

Accusative Singular.

The case-ending was

e.g.

-m

in Indo-

European.

This in combination with -a of the stem must have

given a primitive Latin *-am,

final

*portam; but the vowel

in all

syllables in

had probably become shortened before the

2).

beginning of the historical period ( ZZ.

117.

Vocative Singular.

There

was no case-ending

in

the

Vocative Singular of ^-sterns in the Indo-European parent-speech.

The Vocative simply had

Whether the Vocative
tion or
is

the

weak form a of

the suffix a (

in).

in actual use represents this original forma-

merely the Nominative employed in Vocative function

cannot be determined.
118.

Ablative Singular.

The

Indo-European case-ending

of

the Ablative Singular seems to have been

with some preceding

this

vowel,

i.e.

-ad, -ed, or -od.

In the noun-declension,

case-ending

belonged
In Latin

in
it

Indo-European exclusively
was transferred

to the ^-sterns (see 130).

to 5-stems also,
-ad,

combining with the

preserved in early
i.

final -a of the stem to

produce

which
;

is

in-

scriptions, e.g. praidad,

17.
is

CIL.

i.

63, 64

sententiad, CIL.

196. 8,

These

inscriptions belong to the period of Plautus

but

it

generally thought that such Ablatives were probably archaistic

at that time.

Before an

initial

consonant,

final

d when

following

a long vowel regularly disappeared. a while two forms must have existed,

Theoretically, therefore, for

an ante-consonantal
etc.

form,

praida,

etc.,

and an ante-vocalic form, praidad,

form
early

But the

ante-consonantal
before 200 b.c.

became predominant,

probably
Locative

119.

Locative Singular.

The
-1.

case-ending of the

Singular in Indo-European was

In -^-sterns

this

combined

A- STEMS.
with -a of the stem to produce
-at,

123
a long diphthong ( 86),
as in the case

which then became shortened to

of the Dative ( 115). 120.

-at, later -ae, just

Nominative and Vocative Plural.

The
must

original

case-

ending of the Nominative Plural in Indo-European was

nouns.

-es for all

In the case of ^-sterns,

this

-es

early have conis

tracted with final -a of the stem to *-as.

This *-as

the regular

termination of the Nominative Plural of 5-stems in the other

Italic

dialects,

Oscan,
1

Umbrian,
-as,

etc. ;

but has entirely disapthe termination

-at,
is

peared in Latin.

Instead of
to

we have
*portai.

which goes back

analogical, after the

an

original

This

formation

Nominative Plural of ^-sterns

of ^-sterns
is

in -oi ( 131).

The Vocative
ployed
121.
in a

Plural

simply the Nominative

em-

Vocative function.
Plural.

Genitive

The

case-ending

of

the

Genitive
this

Plural in

Indo-European was -dm.

With the -a of the stem

case-ending must have early contracted to *-am, a termination

which has entirely disappeared

from

all

the

Italic

dialects.

Instead of *-am the Latin has -arum, a termination borrowed

from the Genitive Plural of the Pronominal Declension.

This
earlier

-arum
*-asom

is ;

developed by Rhotacism

98.
in

1)

from an
e.g.

compare Homeric Greek forms

-aw,

Oeawv for

Oed(a-)(j)v.

the poets,

The forms ending in -um, which sometimes occur in e.g. caelicolum, Dardanidum, are new formations, posAeneadum (from Aeneades).

sibly in imitation of the <?-stems ( 132), possibly after the analogy

of such Genitives as
122.

Dative and Ablative Plural.

The Indo-European parentthe

speech

had no special form

all

for

Ablative

in

the

Plural.
is

The

Ablative Plural, in

languages in which that case occurs,

identical in form with the Dative.

l

The genuine Dative and Ab-

few possible vestiges occur in the early language.

24

INFLE C TIONS.
5-stems in -abus (on -bus, see
is

lative Plural of

144) appears only

e.g.

in a few

words where distinction of sex

etc.

important,

equabus,
-is,

filiabus, libertabus,

Elsewhere we have the termination

which

is

historically

an instrumental formation borrowed from the

of the

tf-stems.

The
was
-ois

termination
(see

Instrumental
the
-Is

Plural

of the

tf-stems

the termination -ais,

By analogy which regularly became

133).
t

^-sterns created

(see 80. 2).

Nouns

in

-ia

sometimes contract the

e.g.

with

the

-is

of the
taeniis.

termination to

-is,

Virgil,

Aen.

v.

269,

taenis

for

Words
(for

in ia, e.g.
;

Maia

(the adjective) have

1.

-is, e.g.

Kalendis Mais

Mails)

see 80.

123.

The Accusative

Plural.

The case-ending of the Accusa-ns.

tive Plural in

Indo-European was
i.e.

The n disappeared

accord-

ing to 109. 3,

portas for *portans.

0-Stems.
A. Masculines and Feminines.
124.

In the Indo-European parent-speech there was Ablaut

( 70) in the suffix of ^-sterns.

Both forms of the strong grade

in the Vocative
;

occur, e
tive

and

o.

The former appears

and
partially in

and Locathe

Singular,

the Ablative

the latter in

remaining cases.
125.

Nominative Singular.

This

is

formed by appending

-s

to the stem, e.g. horto-s, later hortus

( 76. 1).

On

ager, see

100. 3.

126.

Locative and Genitive Singular.

In

the

Locative Sin-

gular the suffix took the form e ( 124), which, with the Locative

case-ending

1,

gave by contraction
is still

-ei,

whence

regularly

-1.

The

Locative function
in

apparent in humi,
;

belli,

domi, heri, also

etc.,

town names,

e.g.

Corinthi

and

in quarti, quinti,

in such

expressions as quarti die, quinti die ( 173).

0- STEMS.
It

125

was formerly thought that the Latin Genitive Singular of o-

stems was a Locative that had taken on a Genitive function.

this position is

But

no longer tenable.
B.C.,

For

in the

Senatus Consultum
ei
I,

de Bacchanalibus of 186

where Indo-European

is

still

regularly written El, the Genitive ending appears as

that
this,

showing
Besides
different

we have a

different formation

from the Locative.

{e.g.

the Genitive Singular of z'<?-stems

imperium)

is

from the Locative.

For while throughout the Republican period

-i,

the Genitive of /^-sterns ends in

the Locative of such stems

is

ends in

-it, e.g.

Brundisti.

The

Genitive, therefore,

probably

distinct in origin

from the Locative, but what the origin of the

Genitive
in

-1 is, is

not clear.

Words

in -eius
3).

formed the Genitive

-el, e.g.

Pompei from Pompeius (82.

127.
early
-oi.

Dative Singular.

The

Indo-European case-ending
final o

-at

combined by contraction with

Perhaps we have
this

of the stem, producing

(shortened to -oi ;

86) in Numasioi

in our earliest Latin inscription,

torical

CIL.
(

xiv.

4123.

In the his-

period -oi has

become

86).
regular

128.

The Accusative Singular.

to

The

ending -m

is

appended
( 76. 5)-

the

stem in

0,

e.g.

horto-m,

classical

hortum

129.

Vocative Singular.
e.g.

The stem with the

is

^-suffix serves as

a Vocative,

hort-e

there

no case-ending.

Not only proper

-1

nouns in
for -te)

-ius but all

nouns

in -ins regularly

had

(by contraction
fill, Vocatives

in the Vocative Singular.

But barring

from other than proper names are rare.

cally

Forms

in -ie are practi-

unknown, except
Ablative
in

as cited

by the grammarians.
<9-stems
originally

130.

Singular.

were the only

class

of

nouns

Indo-European that
;

had a special Ablative

case-ending

other nouns, so far as they exhibit a special ending

126
for this case,

INFLECTIONS.
have borrowed
it

from

^-sterns.
e,

The form
or
o, i.e.

of this

case-ending
or
-od.

is

with a preceding vowel, a,

-ad, -ed,

As

the

case-ending

appears only in contraction, the

vowel cannot be determined.

The stem appears

and
to

in

two forms,

one in

o-

and one

in e- ( 124), e.g. rtcto-

recfe-.

With the

former of these the case-ending combined

produce *rectdd,
in early Latin,

and with the

e.g.

latter *rected.

Forms with d appear

.

poplicod, facilumed
B.C. in

(= facillime)

Later (probably shortly

before 200

the ordinary speech) the

in -e

disappeared

see

118.

The forms
etc.

became appropriated

as Adverbs,

recte,

facillime,

131.

Nominative and Vocative Plural

The Nominative
-os.

Plural

of <?-stems in Indo-European was originally formed by appending

the case-ending -es to the stem, giving Indo-European

This

termination appears in the other

brian,
etc. ;

Italic dialects,

Oscan,

Um-

but in Latin the osteins have borrowed the termina-oi.

tion of the Pronominal Declension, viz.

tradition of this

appears in pilumnoe, poploe, cited by Festus (p. 205, ed. Miiller).

But

final

oi regularly
is

became

i,

the classical

termination,

e.g.

horfi ; di

common

as the

Nominative-Vocative Plural of deus.

132.

Genitive Plural.

The
e.g.

original termination

was -dm, the

result of contraction of final o of the

stem and the case-ending

(

-dm

121).

This termination, shortened to -dm

42.

1),

appears in early Latin,

Romanom, and
words in the
25. 6.
is

in

the form

-um

( 76. 5) is also regular in certain

classical period, e.g.

ta/entum, modium, deum,

etc.

(Gr.

a).

The

usual end-

ing -drum

is

of secondary origin, and

formed

after the

analogy

of the Genitive Plural of ^-sterns ( 121).

133.

Dative and Ablative Plural.

is

The

so-called Dative

and

Ablative Plural

in reality

an Instrumental.

The Indo-European

6-STEMS.
form of the termination was
( 86),
122.
-dis.

127
first -dis

This in Latin became

and then
In
;

-eis, -is (

81. 4), the classical termination.

Cf.

-/^-sterns -its often

;

contracts to
is

-is,

e.g.

conubls for

conubiis

so fills, auspicis

dis

common

as the Dative-Ablative

Plural of deus.

134.

Accusative

Plural.

The

Indo-European

case-ending

was

-11s.

Latin *horto-ns would represent

the primitive forma-

tion; this

became

horfos ; 109. 3.

b.

B.

Neuters.
present

135.

In

the

singular

these

no special

peculiarity.
as

The Nominative,
ending, which
is

Accusative,

and Vocative have -m

case-

Indo-European.

136.

The Nominative, Accusative, and Vocative Plural have

is

-a.

This ending

in all probability identical with that of the

;

NomiThus
'

native Singular of -^-sterns

i.e.

certain Feminine collective nouns

so used syntactically.
'

came
an

to be felt as Plurals

and were
be

original *juga (Latin

juga) meaning
to
felt

collection of yokes

(cf.

German das

Gefoche)

came

as a Plural

and was con-

strued accordingly.

The
still

use of the Singular in Greek with a

Neuter Plural subject, apparently dates from the time when the

Neuter Plural was

a Feminine Singular.

In Latin

this -a

of the Nominative, Accusative, and Vocative Plural of ^-sterns

was transferred also

%maria,

to consonant,

t-,

and
'

?7-stems (e.g. *nomina,

'

*cornua),

and when (by the

etc.,

Breves Breviantes

law

88. 3) the -a of jnga,

was shortened
stems,

to -a, this shortening

was

extended
etc.

also

to

other

giving

nomina,

maria,

comua,

28

INFLE C TIONS.
Consonant Stems.
A.

Masculines and Feminines.

The original Mute stems.

137.
s,

case-endings are seen to best advantage in the

Nominative and Vocative Singular.

final
e.g.

The
;

case-ending

is

which combines with the

Gr. 32, 33,

consonant in the ways enuprincep-s

niiles ;

merated in

dux.

The

Nominative serves also as Vocative.

138.
Genitive Singular.

Of
-s,

the
-es,

three
-os,

forms of the Indosecond,

-es,
is

European case-ending,
one which
regularly
-is

viz.

the
to
a),

the

appears

appended

consonant
e.g.

stems.
milit-zs.

This becomes

according to
-os
5),

73.

2.

ped-h,

Traces of the ending

for -os

are seen in early Latin nomin-vs {-us

ace.

to

76.

Castor- vs, honor-vs,

'

etc.,

perhaps

also in opus in the phrase opus est,

it is

necessary.'

Cf. 341. 2.

139.

Dative Singular.

The

Indo-European case-ending was

-i,

probably -ai which regularly became

milifi for *nulitai.

e.g.

ped-i

for

*ped-ai

140.

Accusative Singular.
after

The

Indo-European case- ending

necessarily

was -m, which,

( 102. 1)

consonant,
as -em, e.g.

became sonant
;

and developed

pedem

for

*pedm

principem

for *principm.

141.

Ablative Singular.
stated,

In the Indo-European parent-speech,

Ordinarily the Genitive served also

as already

there was no separate form for the Ablative

Singular except in ^-sterns.

as Ablative.

In

Latin consonant

stems
-i

the

ending

-e

is

the the

Indo-European Locative ending

ends in

75.

3).

But

after

analogy of ?-stems, the Ablative of consonant stems sometimes

id,
-1, e.g.

airid

= aere)

conventioni

153.

In the

CONSONANT STEMS.

29

Ablative Singular of adjectives with consonant stems, the termination

-i

has

become
So also

practically universal
in

(e.g.

prudent!, aitdaci,
e.g.

felici, etc.).

town-names, to denote place where,

etc.

Tiburi, Carthagirfi, Lacedaemoni,

the
-e

To

denote place whence,

forms are used,

e.g.

Carthagine.

142.

Nominative and Vocative Plural.

the

The
-es,

Indo-European
seen
in

case-ending of
-es (e.g.

Nominative Plural was

is

Greek
which

cf>v\aK-es),

but

not preserved in Latin.

88. 3.

Plautine canes,
-es

pedes, turbines,

etc.,

come under
all

The ending

appears regularly in

nouns of the so-called Third Declension

;

has been borrowed from the z-stems

fact that

see

154.

Owing

to the

the

Nominative and Accusative Plural were regularly

(e.g.

alike

in

consonant stems
fact
Plural,

indites,

milites)

and owing
-is

to

the

further

that

many consonant stems took

the -z-stems
this
-is

in

the

Accusative
that

after

(159.
was

1),

it

happened
to

by proportional analogy
Plural.

transferred

the

Nominative
sively
(

to

early

The phenomenon is confined almost excluLatin, where we find such forms as iovdicis,
etc.

judicis) , homiriis,

143.

Genitive Plural.

The

regular ending

-um

is

for earlier

-om, from -om ; see

121.

144.

Dative and Ablative Plural.

The

Indo-European end-

ing was -bhos, which

This appears (97. 1. b). once or twice in early Latin, but soon became -bus ( 76. 5).
-bos
i

became

The
is

of -ibus, the regular

termination of
156.

all

consonant

steins,

borrowed from the /-stems;

145.

Accusative

Plural.

The

Indo-European

ending

-ns

became
*-ens,

-ns ( 102. 1) after a consonant.

-es ;

This regularly became

whence

109. 3. b.

3O

INFLE C TIONS.
B.

Neuters.
are

146.

The Nominative and Accusative Singular

formed with-

out case- ending.

Plural, see, 136.

For the -a of the Nominative and Accusative

Stem-Formation of Consonant Stems.

147.
70).
1.

Several formative suffixes originally

showed Ablaut

( 62,

Thus:
S-Stems.

Stems

formed with

the
;

suffix

-os

(-tis)

e.g.

gen-us,

had

in certain cases the suffix -es-

thus originally

Nom.

*gen-os, Gen. *gen-es-es, Dat. *gen-es-ai, later gen-us,

gen-er-'i ( 98. 1).

gen-er-is,

In some words the

e.g.

-os- suffix

of the

Nomina-

tive

invaded the oblique cases,

.

temp-us, Gen. temp-or-is (for

*temp-os-es)
temp-er-e.
-es-

Yet the

-es- suffix

appears in the adverbs temp-er-i,

Cf. also temp-es-tas, temp-es-tivus,

/.

where the

original
is

has

been protected by the following

Pignus, which
-es-

ordinarily declined pignus, pignoris,

Latin,
2.
e.g.

had the

suffix

in early

pigneri (Plautus).

Nasal Stems.

The

suffixes of

many

nasal stems originally

had Ablaut
a)

( 62, 70).
suffix -on-

Thus

The
lost

(lengthened from
-en-,

-o?i-,

strong grade;
-11-.

62)

had another strong form,

have

and a weak one,

Most words
e.g.

the -n- grade, and show only -on- or -en-,

(

umb-o

for

umb-o{ii)
(for

109.

1),

Gen. umb-on-is,
;

etc.;

ord-d(n),

ord-in-is

*ord-en-is,

73. 2)

turb-d(ii),

turb-in-is.

Car-o{ii),
suffix.

Gen.

car-n-is,

shows a trace of the weak grade of the

b)
viz.

The
-ien-,

suffix -io(ii)- had another form of the strong grade,

and a weak grade

Italic

-in-.

The weak grade appears

Umbrian,
etc.,

in

the

other

languages,

Oscan,

but not in

Latin, where as a rule

we have only
-ien-

-ion, e.g. actio{ii), acti-on-is,

though Anio(n)
Anienis,
etc.

shows

('protracted

form'; 62. 3)

in

I-STEMS.
c)

The

suffix
.

-mo(n) had
-?nen-

also

the

grades -men- and -mn-.

e.g.

Sometimes the

grade appears in the oblique cases,

/io-mo(n), ho-min-is, etc.; sometimes the -nio{n) of the

tive

NominaIn the
for

appears throughout,

e.g.

sermo, sermonis.
suffix.
e.g.

d)

Neuters in -men show two forms of the

stands
for

Nominative -men
*no-mn.

-mn

( 102. 1),
is

no-men

In the oblique cases min-

for men-, e.g. rio-min-is for

*no-men-es ( 73. 2).

3.

^-Sterns.

Some

of these

originally

had Ablaut

in

the

suffix.

Thus

a)

Nouns of
originally

relationship

in

-ter,

e.g.

pater,

mater, frater.

These
-tr-

had three forms of the

62).

suffix, viz. -ter-, fer,

and

(weak form;

The Greek

has clung quite closely to

the original declension

-tr-

7ra-Trjp, ira-rp-os, Tra-rip-a.

In Latin the

form of the

suffix

has gained the supremacy in the oblique

cases; in the Nominative, -ter represents earlier *-ter (88. 2).

b)

Nouns of agency

in -tor originally

had three forms of the

all

suffix, viz.

-tor-, -tor-, -tr-.

In Latin these have

-tor

practically

been reduced
88.
2).

to one, -tor

(Nominative
-tr-,

being for earlier *-tor;

The weak grade

however, appears in the corree.g. vic-tr-ix, gene-tr-ix, etc.

sponding feminine nouns of agency,

/-Stems.
A.

Masculine and Feminine

originally

z-Stems.
in

148.

These

had Ablaut

( 62, 70)
-ei-,

the

suffix.
-1-.

The

strong forms of the suffix were

-oi-; the

weak form

Many
cn-acri?
*(3aTL<s)

original /-stems have passed over in Latin into the -io{n)

class ( 1470 2. b).

for *oTa-Tis)
;

Examples are
;

statio

(earlier *statis
;

cf.

Gr.

-ventio

(earlier

*ventis

cf.

Gr. /ftuns for

-tentio (earlier -tentis ; cf.

Gr. racus for *rarts).

is

149.

Nominative Singular.
-s,

This

regularly

formed by aplost the

i

pending

e.g.

Ignis,

turri-s.

Several nouns

have

32

INFLE C TIONS.
s

before

by Syncope
;

( 92), e.g.

pars

for *part-(i)s (cf.

partim)

gens for *gent-(i)s

metis for *ment-(i)s.

Gr.

38. 3.

150.

Genitive

Singular.

The

Indo-European

termination

seems to have been

grade of Genitive
-eis,

-eis, i.e. ei

(strong form of suffix)

138).

-s,

weak

case-ending (

But

this

termination

while preserved in Oscan and Umbrian, has disappeared in

Latin.

The termination
Dative Singular.

-is is

borrowed from consonant stems.

151.

The

Indo-European case-ending
suffix
-ei,

-ai

was appended to the stem with the

thus giving, for

example, *turrei-ai, whence by contraction *turrei, turri.

152.

Accusative Singular.
e.g.

The regular ending -m

-etn

is

appended

to the stem,

turri-m.

The termination

(borrowed from

the consonant stems) has, however, largely

-int.

displaced primitive

See Gr.

37.

153.

Ablative Singular.

There
-d,

was no special form

for the

Ablative Singular of z-stems in Indo-European.

ever,

The

Latin, how-

formed an Ablative in
: :

e.g.
:

turrid, after the analogy of

:

0-stems (hortos
-^/-forms,

hortom hortod:

turn's

turrim

turrid).
;

These
the

however, are attested by only scanty examples

-i.

d
in

early disappeared ( 109. 1), leaving the termination

But

most nouns the ending

replaced this
-i.

-e,

borrowed from consonant stems, has

-J.

Adjectives, however, always have

154.

Nominative Plural.
-ei-

The
The

suffix of the

Nominative Plural

took the form

148).

Thus the primitive formation would

1

be represented by
j,

*turr-ei-es.

between vowels

first

became
then
for

and then

regularly disappeared.

The

resulting *turrees
7roA.s

became

turres

by contraction.

Cf. in

Greek

= e)

*7roAei-es.

1STEMS.
155.

133
V

Genitive Plural.

The

ending -um
turri-um.

is

appended

to

the

stem, ending in the

/-suffix, e.g.

156.

Dative and Ablative Plural.

is

The
1.

Indo-European end/-suffix, e.g. tur-

ing -bhos
ri-bus.

appended

to the stem,

ending in the
b
;

On

-bus for -*bhos, see 97.

76. 5.

157.

Accusative Plural.

The
(

termination was
109. 3. b).
-is, is

-ns

hence

originally turrins, whence turns

-Is,

The

termination

which

is

often used instead of

borrowed from the con-

sonant stems.
B.

Neuter

/-Stems.
-/

158.
( 75).

1.

These changed the

final

to -e

by a regular law

Stems of more than two

thus

syllables

then usually dropped

it,

the

-e

developed, while
;

dissyllabic
rete.

stems retained

e.g.

calcarie), animalie)
2.

but mare,

The

case- endings of

Neuter /-stems are

in general the

same

as for Masculines

and Feminines.

In the Ablative Singular the

{i.e. -i-d)

termination

-i is

regular.

On

the -a

of the Nominative

and Accusative

Plural, see 136.

Consonant Stems that have partially adapted themselves to the

Inflection of /-Stems.

159.
tically

As stated in the Grammar,

confined to the Plural,
-is

40, the adaptation

is

prac-

viz.

the Genitive and Accusative,

where -ium and

take the place of the normal

:

-um and

-es.

Several distinct groups of words belong here

1.
e.g.

One

of the most important classes consists of nouns in

etc.

-es,

aedes, nubes,

What

has led to the adaptation of these

words to the
Plural
tically
is

inflection of /-stems in the Genitive

;

and Accusative
in the

not certain

but the fact that stems of this class pracin

never show

-/"/;/

the Accusative

Singular or
-is

-/

Ablative Singular, whereas regular /-stems in

frequently

show

34

INFLE C TIONS.
makes
it

these endings,
-is,

impossible to regard nouns in

-es,

Gen.

as actual z-stems.

2.
i.e.

Nouns

in -ids,

Gen.

-tatis,

may

possibly represent z-stems,

-I-

civitat-i- ;

yet the absence of -im and

is

forms in the Accusai

tive

and Ablative Singular

against this.

Cf.

above.

(/-Stems.

A.

Masculine and Feminine

zz-Stems.
suffix
first

160.

Like the /-stems, the ^-sterns had a

which appeared

in three forms, viz. -eu-, -ou-,

and
;

-u-.

The

two were strong

the last weak.

See 64.

70.

161.
-s,

Nominative Singular.

The Nominative Singular appends

e.g.fructu-s.

162.

Genitive Singular.

The Genitive Singular had the strong

or
-ou-.

form of the
tive

suffix, either -euits

To

this

was added the Genithus giving

case-ending in

weakest form,

viz. -s ( 138),

*fruct-eu-s, or *fruct-ou-s,

whence

regularly fructus ( 85).

viz.

Early

Latin also shows two other formations,

seriatu-is

in -uis

and

-uos, e.g.

and

senatu-os.
-is

These represent the other forms of the

being for earlier
-es (

Genitive case-ending,

138).

The
traction

termination -us cannot be explained as the result of con-

from either

-uis or -uos.

Neither ui nor uo contracts to

u.

In Plautus and Terence ^-sterns largely follow the analogy of

0-stems and form the Genitive Singular in
-1,

e.g. senat'i.

163.

Dative Singular.
to the

The

Indo-European case-ending

-ai

appended

stem with

suffix -eu-

gave *fruct-eu-ai, whence

not formed from that in
u.

regularly fructui.
-ui
in

The Dative
;

in -u

is

by contraction

for -ui

does not contract to

friictu being

The forms

u are probably Locatives,

for earlier *fructeu,

a peculiar terminationless formation, found also in Sanskrit.

U-STEMS.
164.

135
ending -m
is

Accusative Singular.
e.g.

The

regular

ap-

pended,
165.
e.g.

fructu-m.

Ablative Singular.

The

earliest

Latin formation had

-d,

fructud.

This, however, was not inherited from the

Indo-

European, but was a new formation, specifically Latin.

See

153.

The -d was soon dropped,

Nominative Plural.
i.e.

giving fructu.

166.

The
;

original formation

would have
-f-

been

in *-eu-es,

the strong form of the sufhx ( 160)

-es

the

Nominative case-ending
*-u-is,

*-eu-es

would regularly have become

which would have remained uncontracted.

in -us must, therefore,

The

regular

Nominative Plural
origin
;

be referred

to another

it

is

probably an Accusative that has taken on a NominaCf. early Latin

tive function.

Nominatives in
in

-Is

from z-stems,
( 142).

which are likewise Accusatives

167. Genitive Plural.

Nominative function

Fructu-u?n,

etc.,

are for earlier rue tu-om.

e.g.

On

-om, see

121.

Genitive in -utn also occurs,

currum, in

place of curruum.

Inasmuch

as this formation appears in Plautus

(long before the change of -uom to

-uum;
is

57. 1. c),

currum

cannot be explained as from curruum, but

after Genitives in

an analogical formation

-um from

<?-stems ( 132)

and consonant stems.

168.

Dative and Ablative Plural.

The regular Indo-European

( 97.
1.

case- ending *-bhos

larly

became Latin -bus

stem in

b),

and was reguLater, either

appended

to the

u-, e.g. fructu-bus.

owing

to the influence of

consonant and 2-stems, or to the ten6.

dency of u
-ib us.

to

become

before labials (

2), -ubus often

became

169.

Accusative Plural.

The

primitive

formation would be

represented by *fructu-ns (case-ending -ns),

fructus
;

whence

regularly

109. 3. b.

36
B.

INFLE C TIONS.
Neuter
u- Stems.
peculiarities.

170.

These are not numerous and present few

The long u

of genu and comic has been explained as being possi-

bly an original dual formation,

'

two knees/

etc.

1 and U- Stems
171.
1.

The

only z-stem in Latin

;

is vis.
i

The

terminations of

the Singular follow those of z-stems

has probably been short-

ened

in

the

Genitive,

though the actual quantity cannot be

for

proved.

The Accusative vim

etc.,

*vim

is

regular; 88.

2.

In

the Plural, vires, virium,

result

from the conception of the

98.
1.

stem as
2.

vis-,

whence

*vis-es, vires, etc.;

easterns are represented

by sus and grus, both of which take

the endings of consonant stems, shortening u regularly to u before

vowels.

Subus

is

not a contraction of suibus, but represents

;

the original formation

subus and suibus are the result of analogy.

Ie- Stems.

172. /e-stems are represented by nouns in

facies, species, etc.

-ies, e.g. rabies, acies,

The

suffix

-u- originally had Ablaut ( 70) in

-i-

Indo-European, appearing in the forms

lost all traces

and -u-

but Latin has

of the

z-suffix

and has

-ie

throughout.

Two

original

.f-stems

(spes

and

fides)

have also adapted themselves to the

same declension

as the -z^-stems, along with res

;

and

dies,

which

were originally diphthong stems

see 180.

172 a
where.
173.

Nominative Singular.

The

case-ending

is

-s as

else-

Genitive

Singular.
-s,

The
The
i

primitive
Lucretius,

Genitive of the
iv.

-ie

stems

ended

in

e.g.
is

rabies,

1083.

But the

regular termination

-a.

of this

is

probably borrowed

ESTEMS.
from
^-sterns, precisely as in case

137

of the ending -ai of 5-stems

-i,

subsequently 2 was regularly shortened before

when

a consonant
in early

preceded the termination,

e.g.fidei, spei, rei, plebei,

though
-i

Latin, forms like fidei, rei are found.

Genitive in

also arises

by the contraction of
Aen.
i.

ei to -ei,

whence

-i, e.g.

pernicu, dii (Virgil,

636).

The ending
see

-e, e.g.

acie, die (in

such expressions as

quinfi die, posiridie, pridie,

etc.), is

not Genitive, but a Dative-

Locative
*diei.
-e (cf.

formation;

174.

The
this

original

formation was
-ei

But under certain conditions

86)
;

diphthong

became

hence die

for *diei.

174.

Dative Singular.

In

the Indo-European parent-speech

to have
;

the Dative

and Locative seem

-ei

become merged
-e (
1

in a single

formation in

(long diphthong)

whence

73).

But

for

the most part this original Dative in -e has been supplanted by

the Dative in

-ei,

new formation modelled on

the Datives of

consonant stems.

174 a

Accusative Singular.

This

is

formed regularly by ape

is

pending the case-ending -m, before which

( 88. 2),
e.g.

regularly shortened

aciem for earlier *aciem.

175.

Ablative

Singular.

No

traces

of forms with

-d are

found, though
etc.

it is

likely that acie, etc., are for

an

earlier *acied,

This formation would be secondary, after the analogy of the

Ablative Singular of ^-sterns.

176.

Nominative

Plural.

The

Nominative

case-ending
e.g.

-es

(see 142)
for *acie-es.

combines

by contraction with the stem,

acies

177.

Genitive Plural.

The

termination

-erum

is

after the

analogy of -arum of the ^-sterns and -drum of the <?-stems.

38

IXFLE C TIOXS.
Dative and Ablative Plural.
( 97. 1. b), is

178.

The ending

-bus, for Indo-

European -bhos
179.

appended

directly to the stem.

Accusative Plural.

The primitive Latin formation would

(

be represented by *aciens, whence acies

109.

3. b).

Stems ending
180.
1.

in a Diphthong.

Res,

originally a

diphthongal stem,

viz.

*reis,

had

Decome
2.

res in the

Indo-European period.
Singular of navis was originally
;

The Nominative
Dative nam.
is

*naus.

This form disappeared

tive navis,
3.

navis

is

new formation

after the

Geni-

Bos

probably not a genuine Latin word, but

is

borrowed

from one of the

Italic dialects (XDscan?); o represents earlier ou.

The

oblique cases are formed from the stem bou-, u becoming v

between vowels.
bo bus
4.
is

The Dative

Plural bubus
bos.

is

regular (for *boubus)

modelled on the Nominative

of Ju(filter)
was,
104.

The stem
and

in
1.

Indo-European, *Djev-.

Initial

^'regularly

became j

a); hence *JDjevthis

became

*Jev-,

further Jov- ( 73. 3J.

From

stem are formed

the oblique cases Jov-is, Jov-i, Jov-em.

of the simple stem, namely *Jev,

It
is

The Vocative consisted which became *Jeu, Ju- ( 85).

(i.e.

this

last

which, combined with -piter

pater,

73.

2),

gives Jupiter (Juppiter,

a

88.

1). really a Vocative, but

used as

Nominative

as well.

The
1

original

Nominative was *Djeus, with a

f

'

by-form

'

*Difeus,
the

god of the
three

sky,'

god of
ago

day.'

From
But

the latter

came
'now

com-

mon noun
day,'
'

dius, 'day,' preserved in nudiustertius,

the third

days

'

86).

after the analogy of the

Accusative diem, there arose also the Nominative Dies seen in

the archaic Duspiter, which
to Jupiter.
is

the real Nominative corresponding

This same

dies, as a

common

noun,

day,' passed over

into the inflection of the z'^-stems.

COMPARATIVE AND SUPERLATIVE.

Formation of the Comparative and Superlative. 1
181.

39

The Comparative.

1.

The

regular Comparative Suffix in

Latin was -jos-

('protracted form' -jos-;

-is-

62. 3), with -jes- as

another form of the strong grade, and

as

weak grade

( 62).

But
line

-jos-, -jos-

alone

survived in Latin.

In the Nominative Mascu-

and Feminine the

original formation
i

was

-jos.

Following a

consonant,
cases s

regularly

became

103. 2),

and

in the oblique
;

became r

( 98. 1), e.g. melioris for *meliosis

the r was

subsequently

transferred

by analogy to the Nominative.

The

Neuter took the

e.g.

suffix -idsis

and kept

melius.

Minus

not for

u ( 76. 5), *min-ios (which would be imposs,

changing

to

sible in

Latin), but was

probably originally a Noun, *minvost

whence *minos

( 103. 5), minus,

Gen. *mineris.

This became
analogv

an Adjective and developed a Masculine minor,

of other Comparatives.

after the

in

The Indo-European parent- speech had another suffix, which some languages developed Comparative force, viz. -tero-, -tera-,
2.

e.g.

Greek

^axco-repos.
'

But in Latin
to,'
i

this suffix

retained

its

primi-

tive force of
lit.

having a relation

connected
'

with,' e.g. ex-terus,

* interns,

'having a relation to the outside, outer

posterns,
suffix

citer'etc.
-tor- to
3.

These were

felt

as Positives

and took the regular

denote Comparative relation.

is

Plus

for *plo-is,

from the root pie-, plo-,

'

fill,'

'full' (

62).

This *plois became *plois ( 86), whence plus ( 81. 1).

archaic

In the
find

hymn known

as the

Song of the Arval Brothers we

the form pleores from pie-, the other phase of the root.

182.

The

Superlative.

We

have three Superlative

suffixes in

Latin

See Lindsay, Latin Language,

;

92

p. 404; Stolz, Lateinische Grammati&, Sommer, Handbuch der Lateinischen Laut- itnd Formenlehre, 302 ff.

- -is

appears in plus (see

3,

below) and in the Adverbs mag-is, nim-is.

140
i.

INFLECTIONS.
-mo-, -into- seen in
'winter,'
lit.

sum-mus
l

mus ; bruma
prox-imus
2.

*sub-mus ( 106. 2) ; /rzshortest day,' for *breu-ma {brev-is) ;

for

pessimus ; also in extre-mus, postre-mus ; supre-mus ; plur-imus,

(for *proqu(f)simus)
6. 2),
;

max-imus

(for

*mag(i)s-imus).

-tumus, -timus (

seen in ci-timus, ex-timus, in-timus,

(

pos-tumus, ul-timus, op-timus, for op(i) timus, from ops

earlier citumus, etc.

92)

This

suffix

originally

had much the same

retains
etc.
its

meaning

as tero-, tera- (see 181),

e.g.

and
;

still

primitive

force in several words,

3.

llgi-timus

fhii-timus,

The

suffix -issimus is

of uncertain origin.
-isto-

It

can hardly be

for -istimus, a

mingling of

(seen in the Greek Superlative

ending

-io-to?)

and -mus ;

for -istimus could not

become
been

-issimus.
offered.

No

plausible explanation of the suffix has as yet

is

Acerrimus

probably for an original *acr-is-imos, whence by

92)
acrsimos,

Syncope
(

*acersimos
is

100.

3),

acerrimus

106. 4).

Similarly facillimus
-is-,

for *fdcil-is-imos, *facilsimos,

facillimus ( 106. 3);

sents the

in the forms

assumed

as original, repre(

weak form of the Comparative

ii.

suffix

181).

Cf.

Brugmann, Grundriss,
4.

p.

158.
43.

On

the quantity of i in -issimus, see

Numerals. 1
Cardinals.

183.
spot
'

1.

Unus

is

for earlier oinos

81. 1. (cf.

Gr.

oLvt),

the 'one-

on dice).

German

ein

and English one are the same word

Greek
2.

els for *o-e^-s is

is

not related to unus, but to semel, singu/i.

Duo

for earlier

*duo according to

88. 3

cf.

Gr.

Sixo.

The formation was Dual.

1

See Brugmann, Grundriss,

ff.;

ii.

164-181;
",

Lindsay, Latin Language,

pp. 408

Stolz, Lateinische

Grammatik 9
306

91; Sommer,

Handbuch der

Latdnische Laut- und Formenlehre,

ff.

1 ; '

CARDINALS.

14

The stem shows Ablaut weak grade tri-. The former stem
3.

Tres.

( 64. c), strong grade trei- t

originally

appeared

in

the

Nominative,
tri-, viz.

*trej-es,

whence

*tre-es, tres.

The
;

other cases have

109. 3).

tri-um, tri-bus, tri-a, tris (for *tri-ns

4.
is

Quattuor.

The Indo-European form from which guattuor

early
after

descended was probably *quetvores ; but the Latin form

its

lost

inflection,

the

analogy of the other indeclinable

88. 3. /

numerals; -irregularly became or;

primitive e to a,

The change
is

of the

and the doubling of the

cannot be referred to
perfectly natural

any recognized
cf.

law.

The change

of v to u

16. 1./.

5.

Quinque.

The Indo-European form was *penque ;

Initial

cf.

Skrt.

panea, Gr.
tion of the

ttvt..

qu- in Latin

is

the result of assimilabi-bo for

i is

first

syllable to the

second

cf.

Indo-European
in

*pi-bd (Skrt. piba??ii).

The change
i is

of e to

accordance

with

73. 2. b.

The long

probably borrowed from quintus,

for quinctus.
6.

Sex.

This comes from

Indo-European
(for 07: e).

*sex, a by-form of

*svex, seen in
7.

Greek

e|,

Doric p

Septem.

The Indo-European form was

*septm, which regu-

larly developed in Latin as * septem ( 102. 1).

8.

Octo
a

is

descended from an Indo-European

('

*octo.
oc-,

The form
'

was

Dual
'?).

two

sets

of fingers

;
'

root ac-,

sharp,

'pointed
9.

Novem. The Indo-European form was *nevn, which

102.
1
;

in
cf.

Latin would regularly have appeared as *noven (

Eng. ni-ne ; German

neu-ri)

-em for -en

is

probably due to

association with the following dec-em.

10.

Decern

is

for

Indo-European *dekm ;

102. 1.

n. 'Eleven'

to 'Nineteen.'

These are regularly formed by

etc.

composition, undecim,
73.

2.

tredecim,

On

-im

for

-em,

see

For tredecim we should expect * tredecim according

{cf.

to

89

sedecwi for *sec(s)decim

105).

The

remains

42
'

INFLE C TIONS.
Eighteen
'

unexplained.

and

'

Nineteen

'

were usually expressed

by duodeviginti, undevtginfi.
12.
zn,
'

Vlginti.
is

The Indo-European form was

*vi-kmfi, in which

two,'

for *dvt,

an original Neuter Dual, from the root *du-

*kmti, whence in Latin * -gen ft, -gin ft ( 102. 1) was also Dual, in
the sense of
'

tens.'
cf.

The change
'

of k to

is

peculiar,
die'

though not
point
').

unexampled
13.
'

dig-itus for *dic-itus

(from root
all

Thirty' to

Ninety.'

These

end

in -ginfd,

which in
etc.),

Indo-European was *-konta

a Neuter Plural

(cf.
'

Gr. rpiaKovTa, recrcrapdKovTa,

;

meaning

'

tens

*-konta shows the strong grade

lies at

of the root whose

weak grade *kntis

the basis of vlginti (see

above)
-a
c,

-ginfd for *gonta

due

to the influence of vtginfi.

136.

The

is

a vestige of the original ending mentioned in

Tri- in triginta
is

On g for
Its

see above.

probably a Nominative Plural

is

Neuter.

The
it

-a in quadra-, quinqua-, sexais

secondary.
it is

precise origin

uncertain.

As regards quadra-,
It is

best to dis-

connect
word.
14.

entirely from quattuor.

probably an independent

Centum

is

for

an Indo-European *cntom, whence the Latin

;

form by regular phonetic process

red is the same word.
15.

102.

1.

Eng. hunde-,

in
'

hundone.'
is

Gr.

-kclt6v

has prefixed
difficulties.

for -h,

The Hundreds present few

by assimilation
( 90).

Tre-cenfi

for
(for

*tri-eenti

Quadrmgenti, oetingenti
-ing-

quattuor-, octo-) have

borrowed the

from quingenfi

(for

*quinq-genti ;
-ing-

105. 1)

and septingenti
Sescenti
is

(for *septem-gent~i),

where

developed regularly.

for sex-centi,
is

according to
'

105. 1.

Sexcenti, which also occurs,

;

the result of

Re-com-

position
16.

'

87. 3.

On^-

for c in -gen ft see above, 12.

this

Mille.

The most probable etymology of

it

word

is

that

which connects
'

thousand.'

Doric xv^ ia (f r *X* a ^ La)> The Indo-European form of this was *gheslia, which
with Greek
-xiXia,

in Latin

would regularly develop

as *h~elia

( 89; 97. 3. A.),

;;/

and, by assimilation ( 90), *hilia.

The

initial

would repre-

DISTRIBUTIVES.
sent

43

sm-,

weak form of

the root sem-,

Cf. also
'

'

one/ seen

in

sem-per

sem-el, sim-plex, sin-gull.

Greek

fx-ta

for *(or)/xta.

Hence

originally in Latin *sm(h)ilia,

one thousand.'

On m

for initial

sm-, see 104.

1.

b).

Ordinals.

184.

1.

Primus

for *pris-mos
;

{cf.

pris-cus, pris-tinus)

is

Superlative formation
2.

89.
;

Secundus
:

(for *sequondos

103. 5)

is

from sequor; hence

originally
3.

'

the following.'

Tertius

may be

for *tri-tios,

whence by Syncope

( 92)

*trtios,

then tertius (100. 3).

are

4.

Quartus, Qulntus, Sextus

formed from the respective

cardinals by adding -tus.

The

route followed in the development

of quartus
5.

is

too devious to be here described.

Septimus, Decimus are probably for an original *septm-tnos,

Before m,
u, z;

*decm-mos.

developed into the sound variously

represented by
6.
7.

6. 2.

Octavus

is

for

an

earlier *octovus.
cf.

Nonus

is

for

*noven-os ;

183. 9.
suffix
1.

8.

Vicesimus and the other tens are formed with the

-tinio-, i.e.

vicesimus, earlier vicensimus, for *vice?it-timos ; 108.

9.

Centesimus and the Hundreds.

Inasmuch
it

as the
felt as

element

-esimus was

common

to

all

the tens,

came

to

be

an inde-

pendent ordinal

suffix,

and was appended

etc.

to the

stems of the

hundreds, centum, ducenfi,

The

suffix

-timo- would have

given *centum-timus, or else *cesimus for *cent-timus.

10.

Millesimus follows the analogy of the hundreds.

Distributives.

185.

1.

Singuli shows the

'

weak form of the root sem-,

'one,'

seen in sem-el,
suffix -guli is

once,' sim-plex, sem-per, etc.

The

origin of the

not clear.

144
2.

INFLECTIONS.

The

other Distributives are formed with the suffix -no-,

;

e.g.

bint for *bis-nt

trim for

*tris-nt.

Beginning with septent, the

-~~en~i,

Distributives are formed by the

suffix

which

is

borrowed

from sent

(for *secs-itt ; 105.

1; 89).

The

cardinal form to

which

this suffix is

added, usually loses

e.g. sept{eni)~en~i,

its final

syllable,

sometimes
viceni.

the last two syllables,

nov(em)eni ; dent,

Multiplicatives.

186.
2.

1.
is

Semel, 'once,'
for dvis,

is

from the root sent-;

185.

1.
;

Bis

preserved in the Glosses of Festus

104.

2. c).

Cf.

Greek

8ts.

For Latin

dis,

see 104.

2.

3.

Ter

is

for *tris {cf. Gr. rpt's) in

unaccented position.
*trr, ter;

The
106.

sequence of development would be

3; 100.
4.
5.

*tris, *trs,

3.
is

Quater

possibly for *quatrus, *qualrs, quater(r)

100. 3.

The

other Multiplicatives are formed by the suffix -tens,

-ies (see 20. 2),

which

is

variously explained.

Some
l

see in
'

it

the Participle of
six.'

eo,

so that sex-tens would

it

mean

literally

going

Others identify

with the Sanskrit suffix -yant,

great.'

PRONOUNS}
Personal Pronouns.
187.
First Person.

1.

The Nominative Singular,

an Indo-European *ego.
is

ego,

for

earlier ego ( 88. 3), represents

2.

The Genitive Singular,

met,

simply the Genitive Singular

Neuter of the Possessive

of met

metis,

used substantively.
Latin
the

By

the side

we have

also

in

early

Genitive

mts.

This

probably goes back to an Indo-European Genitive- Dative-Loca1

See Brugmann, Grundriss,

vii
;

ii.

407-459; Lindsay, Latin Language,

chap,

Stolz, Lateinische Grai?i?natik, z 89,

90

Sommer, Handbuch der

Lateinischen Laut-

und Formenlehre,

266

ff.

PRONOUNS.
tive

I45

form *mei or

*?noi,
-s.

whence *mi.

To

this

was added the

Genitive termination
3.

The Dative Singular, mihi,

is

probably descended from an

This would regu-

Indo-European *megh-oi or *megh-ei, Locative.

larly

appear in Latin as *niehi ( 97.

1

3.

A;
in
1,

81. 2).

The change

of e to

took place

first

when

*?7iehi

was

unaccented position
see 88. 3.

73. 2.

On

the shortening of the final

Mi may
fxoi (also

be a contraction of mihi or may be identical with Greek

Locative).
4.

The Accusative and Ablative Singular,

Before
ttied

nie,

was

tried in early

Latin.

Originally tried was Ablative only, with the case-ending

discussed in

130.

an

initial

consonant

tried

would
Accu-

become

nie,

remaining

before vowels.

The
nie

original

sative Singular

was me, but the existence of

and med side by

the side of

side in the Ablative naturally led to the rise of

med by

me in the Accusative. 5. Nominative and Accusative Plural, nos, is apparently an inherited Indo-European formation. The form was originally
the already existing

Accusative and was thence transferred to the Nominative also.

6.

Genitive Plural.

Nostri, nostrum are the

Genitive Singu-

lar

and Genitive Plural of the Possessive Pronoun used with subIn early Latin we find also nostrorum and (as

stantive force.

Feminine) even nostrarum.

7.

Dative and Ablative Plural.

its

Nobis has

apparently bor-

rowed

termination

-bis

from vobis ; see below.

188.

Second Person.
tu-.

The Indo-European stem was

form
tu
te-

tve-,

with

weak grade
1.

collateral

also appears.

Nominative Singular.
rv- in

corresponds

to

German
is

du,

Greek
2.

Homeric

rvv-q.

Genitive Singular.

Tui

like

mei (187. 2)

the Genitive

of the Possessive Pronoun used substantively.

also a Genitive tzs to

Early Latin has

187. 2).

be explained

like

mis (see

146
3.

INFLECTIONS.
Dative Singular.
for
-i,

Tibi

is

for

an

earlier *tebhei, *tebi, tibi

on

e,

see under mihi, 187. 3.

3.

On

the shortening of the

final
4.

see 88.

Accusative

The origin of the termination -bhei is uncertain. In both Accusative and and Ablative Singular.

Ablative

we have

te,

with fed as an alternative form in early Latin.

On

the origin and relation of the two formations, see

187. 4.

5.

Nominative and Accusative

Plural.

Vos
it

represents

an

Indo-European formation.
Accusative only.
6.

Like nos ( 187. 5),

was originally

Genitive Plural.
;

Vestrum,

vestfi are of the

same formation
vestrum,

as nostrum, nostri
vestri, result
7.

see

187. 6.

Vostrum,

vostri, for

from association with nostrum,

nostri.

Dative and Ablative Plural.

Vobis

is

formed with the

suffix

-bliis,

the relation of -bis in vo-bis to

;

-bi in ti-bi

being perhaps de-

termined by that of Hits to Hit

istis to isti, etc.

The Reflexive Pronoun,

189.

The stem

of the Reflexive

is

*sev-, with the collateral

forms
1

*se-, sv-.

Genitive.

Sui, like mei

and

tui, is

the Genitive Singular of

the Possessive used substantively.

2.

Dative.

Sibi, earlier sibi, is for *sebkei, *sebi.

See under

mihi, 187. 3.
3.

On

the shortening of the final

Ablative.

Accusative
se,

and

In both Accusative
on
trie,

see 88. 3.

and Ablative

we have
origin

with sed as an alternative form in early Latin.

On the

and

relation of the two formations, see

187. 4.

The
190.

Possessive Pronouns.
-os {-us) to the

These are formed by appending

formed by adding the

stems or

other form of the Personal Pronouns.

1.

Me-us

is

suffix -os to *mei, the

Indo-

European Genitive form mentioned

in 187. 2.

This *mei-os

PRONOUNS.
regularly

147
Singular

became mens.
is

The Vocative
for *mei-e,

mi

is

either the

old

Genitive mi, or
mi.

which by

loss of its -e

might

become *mei,
2.

Tu-us

is

from the stem

tev-,

whence

originally *tev-os, later

tovos ( 73. 3), preserved in early Latin.

In

enclitic position ov

became
cf.

u,

whence

tuos,

tuus

see 103. 4.

With Latin

*tev-os,

Homeric Greek
3.

Te(/r)oV
sev-,

Su-us

is

from the stem

whence

originally *sevos, later

e.g.

sovos (73. 3), preserved in early Latin.

In enclitic position,
suos, suus
;

pdtrem sovom, ov became u y whence

see

103. 4.

With primitive Latin *sev-os

cf.

Homeric Greek
sv-.

e/ro?

for *cre/:os.

The weak form

in

of the root sev was

It is this

which appears
e.g.

Greek

6's

for 07:0s,
sis

and

traces are present also in Latin,

in

such forms as
early Latin.
4.

(Dat.-Abl. Plu.), for *svis (root svo-), found in

Noster and vester are formed by adding the

suffix -tero- to

nosis

and
l

vos-,

with Syncope of the e

cf.

Gr. ^/xe-Tepo?.

The
early

suffix

the

same

as that already considered 181. 2,

with,'
'

and had the mean-

ing

connected

having a relation

to.'

The

form

voster

became

vester according to 76. 3.

The Demonstrative Pronouns.

Hie.

191.

1.

The stem of

hie

was

ho-,

ha-.

To

the regular case-

forms of this stem was often added the

to -c
2.
;

suffix -ee, frequently

'

reduced

-ce itself

represents a pronominal stem meaning

here.'

Nominative Singular.
Masculine.
(i.e.

a)

Hie

is

now explained
this

as for *ho-ce.

In unac*hice,

cented

enclitic)

use,

would regularly become

hois

whence

hie ( 76. 4).

The element

thought originally to

have been a Nominative form of the same type as Indo-European

*so (Skrt. sa, Gr. 6 (for *cro), Gothic sa),
i.e.

a Nominative consisting

148

INFLECTIONS.
The
i

of the stem alone without case-ending.

of hie was thus short

by

origin,

and the word always has

1

in Plautus.

Where we
Kicc;

find
i.e.

an apparent
the syllable

in later times,

we should probably read

is

is

long, but the vowel

ix.

short.

An

instance of hicc

occurs in CIL.
below,
b)
c).

60, hicc est.

This cc comes from hocc.

See

Feminine.

Haec
*porta).

for *ha-i-c(e),

adds

(a formative element to

recognized elsewhere in the inflection

original
c)

of this pronoun)

an

*ha

{cf.

Neuter.

Hoc
was

is

for

*hocce,
to

earlier

*hod-c(e),

in

which

-d

is

a case-ending

peculiar
short.

the

Pronominal

Declension.
a

The

o of hoc

Whenever the word makes

initial

long

syllable in verse before

an

vowel

it

is

probable that the

Romans pronounced
664.
3.

hocc, e.g. hocc erat in Mss. of Virgil,

e.g.

Aen.

ii.

Before consonants they pronounced hoc,

Genitive

hoc templum.
Genitive

Singular.
*hoi)'-os,

The

earliest

form

of the

Singular was

whence hoi(j)us, preserved

is
still

in early Latin,
clear.

The

exact nature of this formation

far

from

The

classical

form hujus seems most

likely to
1.

have developed from

hoi-(j)us in accordance with

4.

81.

Dative Singular.
*hoijei,

The

original

form of the Dative Singular

was probably
function.
( 103. 1)

a Locative formation that took on Dative

From
i.

this,

by disappearance of the intervocalic j

earliest

and contraction, arose the

197. 26.

Latin form,
is

viz.

hoice, CIL.
certain.
5.

The

exact

way

in

which huic arose

un-

Accusative Singular.

Hunc,

hanc are simply

for

earlier

*ho-7n-ce, *ha-m-ce, with obvious phonetic changes.

Ablative Singular.

Hoc, hac
all

for earlier *hod-c (e) ,

*had-c (e)

represent the same

Ablative formation as regularly seen in a-

and

0-stems
7.

118, 130.

Plural Forms.
^-sterns,

These

follow the regular termination of

a-

and

except the Nominative and Accusative Plural Neu-

PRONOUNS.
ter,

149

haec,

where

-ai, -ae

(instead of -a) exhibit the

same

as noted

above in connection with the Nominative Singular Feminine.

Is.

192.

1.

The

root of this

pronoun

is

ei-,

weak form

1-

( 62).
eja-

By appending
2.

the suffixes -o-

and

-a

we

get the stems

eo-, ea-.

ejo-,

or (by disappearance of the intervocalic j)

Nominative Singular.
Masculine.
-s.

a)

Is shows the root in the

weak form with the

case-ending
b)
c)

Feminine.
Neuter.

Ea

is

for *ej-a

see above,

1.

Id shows the weak form of the

-d.

root with the Pro-

nominal case-ending
3.

Genitive Singular.
*ei/os,

The
eius,

original formation

is

thought to
;

have been
82. 3.

whence

the correct classical form

cf.

4.

Dative Singular.

El
stem

for *e-ei, earlier *ej-ei

ejo- (see 1).

was

in forma-

tion a Locative from the

In the Pronouns the

Locative served also as Dative.

5.

Accusative

Singular.
1).

Eum,
and

earn

represent

an

earlier

*ejom, *ejam, (see

6.

Ablative

Singular.

Ed

ea,

earlier

eod,
is

ead,

were

formed from the stems

as that of a7.

*ejo-, eja-.

The case-ending
all

the same

and

<?-stems.

Plural Cases.
*ejo-, eja-.

These

are

formed regularly from the

ei (for *ej-oi)

stems

In the Nominative Plural,

it is

repre-

sents the original formation;

z

for ei

by assimilation

90);

is

from

it

by contraction.
els, its, is.

Cf. also the corresponding Dative-

Ablative forms,
8.

Idem

is

simply

is

with the suffix -dem.

1.

For the Compen-

satory Lengthening, see 89.

5O

INFLE C TIONS.
Iste,
Ille,

Ipse.

193. These three pronouns presumably contain in their second

syllable the
'

Indo-European pronoun

*so,

'

he

'

*sa,

'

she

'

*tod,

that.

'

But by association and analogy the second element has

become much modified.

194.
Iste.

The

first

syllable of iste

is

of uncertain origin.

It

was apparently an unchangeable element.

*so, *sa, *tod,

would

arise *isso, *issa,

By the addition of *istod. The regular AccuGreek

tov,

sative of

*so was *tom, *tam, *tod

{cf.

rdv,

ro(S)),

whence
/

*istom, *istam, *istod.

eventually caused *issa to

The preponderance of forms with become ista and *isso to become

ille,

*isto, later iste ( 76. 6),

influenced by

ipse.

195.

Ille.

If olle to i

was the original of

ille,

as

is

usually held, the

change from

can be accounted

for only

on the ground of

adaptation to such forms as

*ol-so,
*ol-se,

iste, ipse, is.

Olle, itself,

whence

olle

( 76. 6;
*ol-sa, olla.

106.

3).

may be for The Feminine

would

similarly have

been

been

*ol-tod,
//

and the Accusative *oltom,

The Neuter would have *oltam, *oltod. Then the

forms with
those with

might naturally have gained the supremacy over

It.

196.

Ipse.

I- here

seems the root of

is [cf.

early Latin eapse,

is

eu7npse, eopse, etc.), while the origin of the sufnx -pse

obscure.

The Neuter, ipsum Noun Declension.

197.

(instead of *ipsud), shows transition to

the

Declension of Iste,

Ille,

Ipse.

With

the

exception of

the forms istud,

Mud

already mentioned, and the Genitive and

all

Dative Singular, these

Declension.

show the usual terminations of the Noun

illius,

The

Genitives istius,

ipsius are

formed by
Locatives

appending the Genitive ending

-os {-us) to isfi,

Ml,

ipsi,

PRONOUNS.
from the stems
is to-,
i//o-,

151

ipso-.

These Locative formations

served originally as both Dative and Genitive in the Pronouns.

Later the Genitive was differentiated from the Dative.

The
198.
1.

Relative, Interrogative, and Indefinite Pronouns.

These are

all

formed from the same

root,

which ap-

pears as qui-, quo-, qua-.

2.

Nominative Singular.
-s.

Quis shows
quo
-f- i,

the stem qui- with the

case-ending

Qui

is

for

a formative element which

hie, u,

appears elsewhere in the Pronominal Declension (see under

191. 2. b)
;

7)

oi in

accented syllables regularly becomes

but

for oi in qui

acter of the word.

may perhaps be explained by the enclitic charQuae is the regular Feminine of the Relative.
the

The formation
which appears
clension.

is

same

as seen in hae-c ( 191. 2. b).

Qua,

in

the Indefinite Pronoun, follows the

Noun De-

Quo-d and qui-d append the

Singular.

regular pronominal ter-

mination to their respective stems.

3.

Genitive

Cujus,
hujus ;
era

for earlier quoi(j)us, *quoijos,

191. 3.

seems best explained

4.

like

Dative Singular.

Cui seems

to have

developed

in the first
14.

century of the

Christian

from the

earlier

quoi ; see

Quoi was probably a Locative formation.

5.

Accusative Singular.

Quern

for

*qui-m has followed the

e.g.

analogy of z-stems having -em for -im,

152.
6.

turrem, ovem,

etc.

Ablative Singular.

Besides
we

the

regular

quo,

qua,
all

quo,

which present no

peculiarities,

find qui used for

genders
a

and

(in early Latin) for

both numbers.

This

may have been

genuine Ablative form (qui for *quld), or an Instrumental.

7.

Plural Forms.

Quae

is is

analogous to hae-c;

191.

7.
it

The Dative and

qui-.

Ablative quis

from the stem quois

133);

has no formal connection with qui-bus, which

from the stem

52

INFLE C TIONS.
Pronominal Adjectives.

199.
also the

Several Adjectives of pronominal meaning have adopted

Pronominal Declension in the Genitive and Dative Singualter ;

liter,

lar, viz. alius,

neuter ; ullus, nullus ; solus, totus, unus.

Alius takes also the pronominal -d in the Neuter Singular.

CONJUGATION}
Introductory.
200.

As compared with Greek and

Sanskrit, the Latin in

its

verb-system exhibits extensive deviations from the original conjugational system of the

Indo-European parent-speech.
:

The

fol-

lowing are the most important points of difference

i.

The

Latin has lost the augment,

i.e.

an

initial e-,

prefixed to

the secondary tenses of the Indicative as a symbol of past time.

2.

The
The

strong

{i.e.

unsigmatic) Aorist has disappeared almost

entirely.
3.

original Perfect Indicative has

become merged with

the sigmatic Aorist.

The

result

is

a tense whose inflections are

derived from both sources, and whose meanings are Aoristic as

well as Perfect.
4.

The
The

original

Middle Voice has disappeared, being super-

seded by a new inflection peculiar to Latin and Keltic.

5.

Subjunctive and Optative do not appear as separate

moods, but have become fused into one, designated Subjunctive.

6.

In the Imperfect and Future Indicative of the a- and

~e-

conjugations

we meet new

formations in -bam and

-bo,

which, like

the r- Passive, are peculiar to Latin

and

Keltic.

See in general: Brugmann, Grundriss,

viii.;

Language, chap,

Stolz, Lateinische

Lindsay, Latin ii. 4601086; Grammatik^,^% 96-118; Sommer,

Handbuch der Lateinischen Laut- und Formenlehre, 317-391.

CONJUGA TION.
7.

53

In the Personal

Endings the distinction between primary

effaced.

and secondary endings has become

8.

Several

new

tense- formations
e.g.

have developed which

are

peculiar to Latin,

the Perfect Indicative in -vi and -ui, the

etc.

Pluperfect Subjunctive in -issem,

Formation of the Present Stem.

201.

Thematic and Unthematic Formation.

The
The

Latin inone, char(e,

herited two distinct types of Present formation.

acterized by the presence of the variable or thematic vowel

o)
is

before the Personal Endings,

illustrated

is

called Thematic.
;

This type

by dicu-nt

(for *dico-nt)

dici-tis (for *dice-tis).

The

other type of Present formation has no thematic vowel, and hence

is

called Unthematic.

Unthematic presents

originally

had Ablaut

( 62).

The

strong form of the root appeared in the Singular,

the reduced form in the Plural.

primitive accentual conditions.

This change was connected with

Presumably the accent

originally

rested on the root syllable in the Singular,

Plural.

on the endings

in the

In Greek, the Unthematic Conjugation

-/u

is

represented by the
e.g.

verbs

(jL-dr)-[Ai,

rt-dc-fxev)

while -w verbs are thematic,

\ey-o-/Av, Aey-e-Te.

Classification of Present Formations.

A. Unthematic Presents.

202.
Latin
;

Unthematic
for the

Presents

are

but scantily represented in

most part they have passed over into the thematic

following verbs are the chief representatives of

inflection.

The

the class
1.

Do, das, ddt (for earlier dat)

1

Plural da-mns, da-tis, dant.

For the personal endings

in this

and the other verbs, see 235

ff.

: :

54
2.

INFLE C TIONS.
Eo.

The two forms of the

The
*ej-o 1
*ei-s
*ei-t

root were

ei-

(strong),

and

-i

(weak).

primitive inflection for Latin, therefore,

:

would have

been theoretically somewhat as follows

*i-mos
*i-tis

*i-nt

In the First Singular *ejv regularly became eo ( 103.

1)

*eis

became
strong
3.

is

( 82)

and

*eit,

it,

later

It.

The

Plural

seems to

have abandoned early the weak form of the root in favor of the
;

imus,

itis,

eunt, therefore, represent *ei-mos,

ei-tis, ej-ont.

Sum.

The

strong form of the root

is

es-,

the

weak

s-.

The

original conjugation for

Latin, therefore,
:

would have been

theoretically

somewhat

as follows

*es-nfi
*es-s
es-t

*s-mos
*s-tis

*s-nt
this.

The

historical forms

show considerable deviation from

es as

Traces of *ess are seen in the regular use of

Latin verse.

long in early

The presumption is that ess represents Plautus's pronunciation. The First Singular sum, along with su-mus (for *so-mos), and sunt (earlier sont) may represent a special thematic The Second Plural es-tis is formed from the strong formation. root, like the Second Singular. Enclitic forms V and 'st sometimes occur for the Second and Third Singular.
1

These are often

The Indo-European

inflection
ef/xi)

was presumably
*i-mos
(cf.

*ei-mi (Gr.
*ez-si
*ei-ti (Gr.
2

Gr.

X-fiev)

*i-te (cf. Gr. i-re)

elcri

for *e?Ti)

*i-enti

The Indo-European

inflection

was presumably
*smos
*ste

*esmi
*essi
*esti

*senti

'

CONJUGA TION.
joined in writing with a previous word,
e.g.

55

bonust

bonum

'st

morast
4.

mora

st.

The usage
in the

is

poetic and colloquial.

Edo.

Unthematic

forms occur only in the Second and

Third Singular, and

Second

Plural.

The

root shows Ablaut,

appearing in some forms as

ed-, in others

as ed-, altered to Is1).

by

euphonic change,
5.

e.g. est (ox

Fero.

Fers, fert,fertis

*edt; estis for *edtis ( 108.

show apparent unthematic forms,

probable that the above Latin

but in view of the fact that this verb follows the thematic conjugation in Sanskrit

and Greek,
( 92).

it is

forms arose by Syncope

6.

Void.

The only forms which are

volt,
volti's).

certainly unthematic are

vult, vultis (earlier

The

root in the

Singular was

normally *velvolt,

(cf.

vel-im, etc.), but *veld

and

*vel-t

became
is
;
'

void,

according to

73.

5.

The Second
vet-,

Singular vis

not for
cf.

*vel-s,

but comes from the root

also

meaning

'

wish

in-

vitus.

Volumus,
e

volunt

have followed the thematic inflection

73. 5.

with o for

according to

Vultis (earlier

volti's) is

most

naturally explained as for *vl-tis,

is

whence

voltis ( 100. 1).

Nolo

for * nevoid^ *novolo ( 83. 3) and

maid

for *mag(e)vold.

B.

Thematic Presents.
:

203.
I.

Of these
Class.
-f-

there are the following classes

Root

The Present stem

vowel
e

consists of the root in

.

its

More exactly the root / appeared in that phase of the strong grade which gave its name
strong form
the thematic
to the different Ablaut Series ( 62).

Thus

roots of the ^-Series

etc.

had
is

e,

ei(i),

eu{u)

those of the ^-Series had d,

The

^-Series

most

fully

represented.
leg- e -/ _,

Examples are
leg- ; teg- e -

e-Series

root

/
_,

root
;

teg- ;

veh- e '
e.,

.,

root

veh-;

deic- e
;

_,

root deic- (later die-; 82)

-J
.,

feid-

root feid-

(later/J^-)

deuc- e

root deuc- (later due-).

'

156
d-Series
a-Sei'ies
:

INFLECTIONS.
.,

ag- e-/

-,

root ag- ; caed- e -/

.,

68).

vadced-

e
-

root vad-.
root ced-.

ISeries
o-Series

e -/ e -/

.,

rod-

.,

root rod-.

II.

Reduplicating Class.
root
-f-

The

Present

Stem
e

is

formed by

prefixing to the
syllable,

the

thematic vowel
initial

a reduplicating

which consists of the

its

consonant of the root

+
6 -/

i.

The
sed;

root appears in
{cf.

weak form
;

( 62).
.

Examples:
e -/
.

gi-gn-

.,

root gen-

Gr. yi-yv-o-fxai)

si-d- e -/

for *sz-sd

( 89), root

also apparently originally *di-do {cf.

;

reddo

for *re-d(i)-do

by

Syncope
*se-,

89).

Sis to, root sta,

and se-ro

for *si-so ( 98. 1), root

do not

strictly
{cf.

belong here.

They were
{a)l-{(r)r}-jxi),
;

originally

unthematic

formations

Gr. (o-)i-o-t^/u,

but have passed in

not properly a redu.

Latin into the thematic conjugation

plicated
i7ri(38a for

bibo

is

formation.

The root was pib- {cf Skr. pibami; Gr. The Latin word results from assimilation *e7ri-7ri7?-Sa).

ofp

to b.

III.

T-Class.

This
t
e

class, like

the preceding,
its

is

but sparingly
form, to
'

represented in Latin.

The
..

root appears in
:

strong
e

which

is

appended
.,

Examples are

nec-t

.,

plec-t e ~/

pec-t*l
IV.

flec-t

e -j

.,

N-Class.

The Present
_.

Stem

is

formed with a nasal

;

infix

before the final consonant of the root

to this

is

appended the
form.
e~

thematic vowel '-/

ples
:

The
;

root appears in the

weak
;

Exam/
_,

find

'/

.,

root fid-

rump- /
6'

.,

root rup-

jung

root

jug-.

Originally the infix was confined to the Present system, but in

it

souie words, as jungo,

appears throughout the entire verb,

e.g.

jungo, junxi, junctus.

Perfect

In other verbs the nasal appears in the

Indicative, though

not in the

Perfect

Participle,

e.g.

fingOi finxi,

fictus ;
..

stringo, strinxi, strictus.

V.
suffix

NO-Class.

To

the root in

its

weak form

is

added the
unthematic.

e'

Originally

verbs

of this class were

CONJUGA riON.

The primitive suffix was nu- in the Singular, and nu- in Plural. The Personal endings were appended directly to these suffixes, so
that a verb like stemd, for example, was once inflected
*ster-nii-o
*ster-nif-s
:

*ster-nu-mos
*ster-nu-tis

*ster-nu-t

*ster-nu-nt

But % ster-nu-mos regularly developed

to sternimus.

Thus two
the

forms of the Plural {sternimus, sternuni) were identical with the

thematic inflection and hence led to stemd, slernis,
Singular,
diets,
tolld,

stei-nit in

after

the

analogy of dicimus,

dicitis,

dlcunt to died,
li-nd,
si-rib,

dieit.

Other examples are sper-nd, tem-nd,

for * tl-nd ( ioo. i).

VI.
se e '/
.,

SCO-Class.

The

Present stem

is

formed by appending
(g) nd-scd,

to the root,

e.g. hl-scd, gii-scd, cre-scd,

posed for

*porc-scd, suesed for *sued-sed.

Many

secondary formations also occur, as

gem~i-scb, trenie-scd ;

especially derivatives from contract verbs, 2&fld7'~escd, from flored ;

/abased from labd; and even from nouns and adjectives, as lapidesco, roresco, duresed.

The

inceptive or inchoative

meaning of numerous

scd- verbs is
is

not an inheritance from the Indo-European parent speech, but

a special development of the Latin
formation,
e.g.
itself.

Many
etc.,

verbs of this

naseor, disco, posed, hiscd,

show no trace of

the inceptive force.

VII. jO-Class.
suffix f' / o- to

The Present Stem

is

formed by appending the

a root or stem.

Several different formations belong

:

under

this

head, the chief of which are the following

tf^yV-Presents from roots ending

in a consonant.

Here/
all

becomes

i,

e.g.

jac-id for *jac-jd; capid for *cap-jd,

and

the

so-called verbs in -id of the

originally of this formation

Third Conjugation.

Some

verbs

have passed over into the inflection of

e.g.

contract verbs in

-id, -ire

(see b below),

venid, venire.

158
b)
/,

INFLECTIONS.
/^.-Presents from roots and stems ending
in a vowel.

The

here

becoming

intervocalic,

disappears
e-

and the concurrent

1- verbs)

vowels (except in the First Singular of

contract.
1)

and

regularly

Examples

Monosyllabic roots

i?np?e-mus for * imple-jo-mos, root pie-;

intramus
2)

for *intrajomos, root tra-.

:

Dissyllabic verb-stems

domamus

for *

do-majo-mos, stem

doma-.
3)

Noun and

Adjective stems in

-a, e, 1:

curamus, stem cura-;

rubemus, stem rube-; firiimus, stem fini-.

These ^-contracts form the so-called

First Conjugation, the e-

and ^-contracts the Second Conjugation, and the

Fourth Conjugation.
c)

z-contracts the

Causatives in eje '/

.,

e.g.

mon-eo, doc-eo, torr-eo.

These

all

take the <?-phase of the strong form of the root

regularly suffer contraction
gation.

( 64).

They

and

form a part of the

Second Conju-

d) Verbs in -ojo probably once existed in Latin, but have dis-

appeared.
*arojo)
to
;

Thus

a?'d,

arare was probably originally *aroo (for

cf.

Gr. dpoo).

The

adjective aegrotus

is

likewise possibly

be referred to an orignal *aegro.

Tense Formation

in

the Indicative.

The Imperfect.
204.

The termination -bam

explained
as

in

the

Imperfect

Indicative

is

plausibly

representing an

Indo-European

Aorist,

*bhvam, from the root bhu-.

to

This seems to have been appended

the stem of the verb.

some oblique case of a noun derived from

The

primitive formation would be represented

etc.

by *monebhvam,

*/egebkvam,

This theory of the origin of the Latin Imperfect

finds confirmation in Slavonic,

where the Imperfect consists of a

the past tense of the verb 'to be.'

case-form of a verbal noun

CONJUGA TION.

59

Early Latin has both -ibam and -iebam in verbs of the Fourth
Conjugation.

The ending

-iebam, however,

is

later in origin than

-ibam, and was borrowed from /^-verbs of the Third Conjugation,

e.g.

capiebam.
has been suggested that the element preceding the -bam in
Infinitive.

It

the Imperfect was an old

Cf.

such compounds as

are -facid, 'to

make

dry.'

Erani

for earlier

*es-am

( 98. 1) exhibits the

same
etc.

praeterite

formation as that assumed for *b/w-am in

amabam,

The Future.
205.
1.

The Future

in -bo.
,

The

Future in -bo

is

analogous

to the Imperfect in

-bam

-bo

is

probably the Present of the root

bhu-, so that

amabo

(for

*ama-b/wo;

204)
ich

literally

means

'I

become

loving.'

Cf

the analogous

German

werde

lieben.

On
is e.g.

ama-, monefound also

scibo,
2.

in this formation, see 204.

The Future

in -bo

in

verbs of the Fourth Conjugation in early Latin,

audibo.

The Future in -am.

This formation,
to
is

regular in the Third

and Fourth Conjugations, .is

Subjunctives, that have

in reality a Subjunctive, or rather

two
1st

come

be ranked as Indicatives.

The

Singular in

-am

(for

*-am)

an ^-Subjunctive

the remaining

forms are ^-Subjunctives.

3.

See 221; 222.

in such archaic

The future

in -so.

This formation appears

for *es-o

forms as dixd,faxo, which are in reality Aorist Subjunctives that

have come to be ranked as Indicatives.

is

The Future

of sum, ero,
;

similarly a Present Subjunctive,

(<r)a>,

( 98. 1)

cf.

Ho-

meric Greek

Attic

w (by contraction). The

Perfect.

The
206.
1.

Reduplication.

In Verbs beginning with a Consonant.

The Redupli-f-

cation in such verbs regularly consisted of the initial consonant

e.

l6o

INFLECTIONS.
the root began with
se, sp,

Where
e.g.

or

st,

the

sc, sp,

or st appeared

in the reduplicating syllable, but the s

sci-ci-di
e,

was

lost in the root syllable,

(early

Latin) spopondi,

vowel,

was assimilated to the root

The vowel when the

ste-fi.
e.g.

reduplicating
latter

was the
sci-cid-l,

same

in the Perfect as in the Present,

mo-mord-i,

pu-pug-i, di-dic-J, spo-pond-i ; but the original forms with <?are often

found in early Latin,

e.g.

memordi, pepugi, spepondi, fhefhaked

CIL.

xiv.

4123.
has disappeared very largely in Latin, yet
e.g.

The Reduplication
traces of
its

earlier

presence are sometimes distinguishable,

( 92)
;

in

rettuli for *re-(te)tu/i

reppuli for *re-pepuli

repperi for
fidi, scidi

re-i^pe^peri

reccidi for *re-{ce)cidi.

(ef.

In the same way

represent an earlier *fepdi, *scecidi

2.

early Latin scicidi).

In Verbs beginning with a Vowel.

e.

The Reduplication here

it,

consisted in prefixing
egi for *e-agi
epi,
;

Only a few verbs have preserved

e.g.

edi

for *e-edi ;

-epl{iox *e-apT) in coepi, for *co-

root ap-

~emi for *e-enu.

in Latin

Some

scholars refuse to recognize a

Reduplication

verbs beginning with a vowel, and explain

the long vowel in the foregoing Perfects in other ways.

Stem Formation of the Perfect.

A. The Primitive Perfect.
207.

In the Indo-European parent-speech the accent rested

syllable in the Singular of the Perfect,
It

on the root

but on the

Personal Ending in the Plural.

was probably owing to these

root

primitive accentual conditions that the strong form of the

appeared in the Singular, the reduced form

in the Plural.

The

special phase of the strong form appearing in the Singular was

that containing o or 3 (see the various Ablaut Series,

62

ff.).

Several of the Indo-European languages, as Sanskrit, Greek, and

the Teutonic, have preserved with

more

or less fulness the original

CONJUGA TION.
Ablaut of the root in the Perfect
;

but in Latin there has been a

uniform

'

levelling'

either the strong

form has invaded the Plural

Singular.

(the usual sequel), or the

weak form has invaded the

Examples of the former process may be seen

in totondimits, spo-

pondimus ; of the
mentioned.

latter in ce-cid-i, tu-tud-i.

In most Latin verbs,

however, other formations have largely displaced both of those just

This has

come

about, partly as the result of phonetic

changes, partly from the workings of analogy.

is

The whole

subject

too intricate for detailed consideration here.

p.

See Lindsay, Latin

Language,

494

f.

B.
208.

The Perfect
-si,

in

-si.

The

Perfect in

which appears chiefly in roots ending

is

in labial, dental,

and guttural mutes,

by origin an Aorist which

has passed over to the Perfect inflection.

Cf

Latin dix-i with

Greek, e-Se^-a.
fect
(in

Some

verbs have preserved both the true Per-

and

this Aorist Perfect, e.g. peperci

and parsi

pupugi and

compounds) -punxi ; pepigi and

C.

(in

compounds)
-vi.

-panxi.

The Perfect
is

in

209.

The

Perfect in -vi

new formation which has

itself.

devel-

oped

in the separate history of Latin

The
-vi is

origin of this

suffix is

not clear

according to one theory,

borrowed from
where

such Perfects asfavi, iavi,fovi, movi, vovi,juvi, v really belongs to the stem.
1

solvi, volvi,

Cf., for

example, Greek
oi8-a
ola-da.
td-fjLv

ta-re
ta-aai,

ol5-e

or Gothic

vait
vaist

vit-um
vit-u\>

vait

vit-un

62

INFLE C TIONS.
D. The
Perfect in
-ui.

210.
is

The

Perfect in -ui

is

a development of that in -vi

-vi

thought to have been added to extended forms of the roots

*gen-e-vi (root gen-), *dom-a-vi

e.g.

(root dom-),

whence genui,

domui;
extend

103.4.

From forms

like these the

category might easily

itself.

Its diffusion

was probably assisted by the existence

of such Perfects as fui, plui

rut, indui, exui, iinbui, etc.

for

early

fuvi (Ennius), pluvi

The
211.

Inflection of the Perfect.

In

its

inflection the Latin Perfect presents a mingling of

Perfect and Aorist forms.

The

exact determination of the details

of this fusion furnishes one of the most difficult problems of historical Latin

grammar

the following explanations can claim only

a certain degree of probability.

212.

The type

of Perfect inflection existing in Latin prior to

the fusion of Perfect

as follows
:

and Aorist may be

partially reconstructed

Singular
1.

Plural
1

vidl
?

vid-i-mus
?

2.

3.

*vide

*vid-ent (for *vid-nt)

Of

these forms vidt in the First Singular represents an Indo*void-ai.

European middle,

The Second

Singular and Second

Plural cannot be conjectured with any degree of satisfaction.

213.
viz.

With

this true Perfect

were fused certain sigmatic Aorists,

an

.y-Aorist
i.e.

and an

-is- Aorist.

These were
to the

originally unthe-

matic,

the endings were

appended
to

stem without the

No

attempt

is

here

made

take account of the Ablaut.

CONJUGA TION.
help of connecting vowels
these
-is-

63

201).

The

inflection of

one of
:

Aorists

may be
;

hypothetically reconstructed as follows

Singular
1.

Plural
75. I; 98. I;

*vid-er-em (for *vid-is-m

s

102. \)*vid-is-mos
*vzd-is-tis

2. 3.

*vid-is (for *vid-is-s )

*vid-is-t

*vid-er-ent (tot*vid-is-nf)

214.

Just

what furnished the starting-point

is

for

the

formal

fusion of the two tenses

is

not clear

vidistis in the

Second Plural

the Aorist form

to
e~

so

is

viderunt in the Third Plural, with % -ent

the analogy of other tenses,
is

changed

-u?it after

e.g.

regunt,
it

amab-unt;

(for e) in -~ernnt

of uncertain origin.

Probably

was borrowed from the Perfect Third Plural in

-~ere,

which

is

certainly a different formation, though not at present well under-

stood.

The

scansion -erunt,

frequent in poetry, preserves the

earlier quantity.

In the Singular, vidi has already been explained

as originally a

Middle which has assumed the function of the and the

First Plural, vidi-mus, are
is difficult

Active.

First Singular, vidi,

Perfect forms ( 212).

explanation.
Perfect

The Second
*visti.

Singular vidisfi

of

Possibly the primitive form of the Second Singular

If so, vidisfi

may have been

may be

a contami-

nation of *visti (Perfect) and *vidis

influence

(Aorist), helped

on by the

of the Second

Plural vidistis.

Perfect Hnsti, however, involves

The assumption of a difficulties. The Personal Ends

ing of the Second Singular Perfect was -tha in Indo-European.

Cf.
-ta.

Greek

oto-0a for */roiS-0a.

In Latin -tha after

should become

Influence

of the

Second Singular Middle ending *-sai

;

(= Latin -si) has been suggested

The Third
ing,
/,

also of the First Singular ending -i.

Singular *vide early

assumed the regular Personal Endvidit.

of the other tenses.

This gave *videt,

Some have Some

thought that in the true Perfect in Latin the primitive Third

Singular was *vidi (a Middle form, like the First Singular).

evidence in favor of
of
-it in early

this

view

is

found in the regularly long quantity

Latin poetry.

64

INFLECTIONS.
The
Pluperfect.

215.

The

Pluperfect Indicative in -eram seems to have devel:

oped by proportional analogy

videram

videro

eram

ero.

The Future

Perfect.
is

216. The Future Perfect Indicative

an Aorist Subjunctive.
1
;

Thus videro
-is- is

is

for a primitive *veid-is-o ( 75.

suffix as already

98. 1), in

which

the

same Aorist

mentioned

in 213, 215.
-it,

The

inflection follows that of Presents in

-o, -is,

except in

the 3d Plural, which has -int instead of -wit, probably owing to

the influence of the Perfect Subjunctive ( 219), which
larly
it

regu-

resembles in the other persons and numbers.

-is,

In strictness
-inius, -itis.

the terminations of the Perfect Subjunctive had

Hence, by confusion of the two formations, the

appears in the
dederis.

-i-

sometimes
iv.

Future Perfect,

e.g.

Horace,

Odes,

7.

20,

The
thematic and an unthematic.

Optative.

217. There were two Optative formations in Indo-European, a

Greek

\v-o-i-fxi

represents the for-

mer,
type
tive

(Tra-L-rj-v

the latter.

In Latin probably only the unthematic

is

to be recognized.
(

Owing
353),
all

to the thorough fusion of Opta-

and Subjunctive

Optative forms are traditionally

known
218.

as Subjunctives.

Present Optative.

Only a few forms occur.

-ie-

The

special
-i-

suffix of the

unthematic Optative was

in the Singular,

in

the Plural.

Thus
root
es-,

the primitive inflection of the Present Optative of the

'

to be,'

was

Singular
1.

Plural
;

*s-ie-m (siem
s-ie-s

88. 3)

s-i-?nus
s-i-tis

2.
3.

s-ie-t

*s-t-nt (s-i-nf)

CONJUGATION.
Siem,
sies, siet

65

are

common
sim,

in early
is

Latin.

The
after

classical in-

flection of the Singular,

sis, sit,

formed

the analogy

of the Plural.
sie?it after

Similarly in early Latin also

etc.

we

find siemus, sietis,

the analogy of siem,

The weak form

Other

of the root, as
illustrations

above, regularly appeared in the Plural.

this

of

Optative

are
l

velim (for *vel-i~e-m, after vel-i-mus), riolim,

eat
'),

malim, edim (edd,

dii-U?i,

possim.

219.
-erim
is

Aorist Optative.

The

so-called

Perfect Subjunctive

in

by origin an Aorist Optative.

suffix -is-

The

tense

is

formed by

means of the Aorist

to

already mentioned in 213, 215,

suffix ie-,
:

which

is

further

appended the Optative

was

i-

218).

Thus the

original inflection of ihderim

*veid-is-ie-m
*veid-is-ie-s

*veid-is-i-mus
*veid- is-i-tis
*veid-is-i-nt

*veid-is-ie-t

By change
the regular
*videriem,

of ei to

( 82),
i

by rhotacism
before r
(
ie

( 98. 1),

and by
gave

development of
Plural
to
1

to e

75. 1), this

etc.,

viderlmns.

But the

of the

Singular
giving

was early changed

after the analogy of the

Plural,

*v~iderim, videris, *viderit.

The long vowel was

in the

regularly shortened

in the 1st

and 3d Singular and

is

3d

Plural, but

was retained in

the 1st and 2d Plural, and the correct inflection

videris.
is
:

common
-itis, -is,

in the

2d Singular.

Hence

viderimus, videritis, and probably also

The forms

in -imus,

where they occur, are to be

with
the

explained as the

result

of confusion

Future Perfect
is

(216).

trace of the long vowel in the

3d Singular

found in

Plautus, Mercator, 924, adduxerit.

Another Aorist formation was by means of the

of
-is-.

suffix -s- in place

This

is

seen in dixim,faxim, ausim for earlier *dic-s-te-m,

etc.

66

INFLE C TIONS.

The
220.
to

Subjunctive,

Two

formations, both descended from Indo-European, are

be recognized.

One

of these

is
;

characterized by the suffix a

the other
is

and belongs
the
suffix e,

to the Present tense

characterized by

and appears not only

Both these

in the Present, but in the

other tenses as well.

suffixes take the place of the

thematic vowel of the corresponding Indicative formations.

221.

A- Subjunctives.

Examples

are

moneam

(for

*mone2.

jam)
the

reg-a-m, audiam, earlier *regam f

Singular,

* audiam;

88.

In

3d

and 3d

Plural also, the a has

become

regularly

shortened, but traces of the original quantity are

early Latin,
e.g.

preserved in

Plautus, Famulus, 489, faciat.

222. E-Subjunctives.
1.

Anient (for *ama-fe-ni) evidently has preferred this type, to

avoid the identity of Indicative and Subjunctive, which would have

resulted from the ^-formation here
;

*ama-ja-m,

etc.,

would have
~e

given

*amam,
2.

* amas, * amat.

For the shortening of

cf.

in *anient,

see 88.

For the

e in

amet, ament,

221.

Traces of the

original quantity are preserved in Plautus, Curculio, 208, amet.

2.

The

so-called Future Indicative of the Third

is

and Fourth

Conjugations

(outside the First Singular, which

is

an ^-Subjunc-

tive) a Present

Subjunctive of the ^-formation which has

e.g.fer-e-s, audi-e-s, etc.

come

to rank as
3.

an Indicative,

The Imperfect

Subjunctive also belongs here.

-s-

There are

two formations, both

a)

Aorists in origin

Without connecting vowel.

Examples are

ess-em, ferrem,
;

for *fer-s-~em ( 106. 3), vellem for *vel-s-em ( 106. 3)

ama-r-e7n

for

*ama-s-em

98.

1)

ntone-r-em for

*mone-s-em, audi-r-em

for * audi-s-em.

b)

With connecting vowel,

e.g.

reg-e-rem for reg-e-s-em ( 98. 1).

CONJUGA TION.
4.

6/

The
:

Pluperfect Subjunctive
vidissem
:

may be
:

the result of proportional

analogy

vidisse

essem

esse.

The Iterative.
A. Active.
223.
that

Present, Second Singular.

this

The

most probable view

the

is

which regards

form
will

as consisting of

simple stem.
to

The Imperative,
which
it

then,

be analogous to the Vocative,

bears in general meaning a strong resemblance.

:

Exam-

ples are

i,

es,

ieg-e, cura (for * cura-je),

mone

(for *mone-je),

audi (for *audi-je).

Verbs
e.g.

in id of the

Third Conjugation follow

diic,

the root class ( 203. 1)

for dice, duce, face, fere

cape.

Die,

fac, fer are probably

e.

by dropping

off the final short

224.

Present, Second Plural.

This
;

is

formed by adding

-te
e.g.

(Indo-European ending of the secondary tenses) to the stem,

i-te,

fer-te,

es-te,

legite

(for * lege-te

73.

2),

amate,

monete,

audite.

225.
is

Future, Second and

earlier -tod,
etc.

Third

Singular.

The
Stem,

termination
ito,

-to,

appended

to the Present

e.g.

ferto,

esto,

legifo,

Originally this formation

Strictly, too,
it

had

Plural as well as
;

Singular force.

was a Present, not a Future

the

Future force
-tod
is

is

a special development of the Latin.

e.g. licetod,

The ending

preserved in early Latin,

datod, violatod.

226.

Future, Second and Third Plural.

is

The
a

termination

of

the Second Plural -tote

-to.

simply a pluralization of the Singular

is

The Third

Plural termination -nto

new formation
i.e.

(cf.

225) after the analogy of the relation existing between the

Third Singular and Third Plural of the Present Indicative,

sunto
:

esto

sunt

'

est

regunto

regito

regtmt
affiant

regit

amanto

am a to

*amat

68

INFLECTIONS,
B. Passive.

227.

The Present.

The

Second Singular ending

-re repre-

sents an original -so, so that Latin seque-re (for * seque-so ; 76. 6)

corresponds exactly to Greek

in -mini
is

e7re(o-)o,

hrov.

The Second

Plural

probably an old Infinitive which has taken on the funcCf. the

tion of the Imperative.

Homeric use of

the Infinitive as

an Imperative.
Xeytfxevai,

According to

this

view Latin legi-mini

= Greek

both forms being originally the Dative of a verbal noun

Cf. ger-meji, Dat. ger-mini.

with the suffix -men.

228.

The Future forms are the

result of

appending the Passive

-r ( 235) to the corresponding Active forms.

The Personal

Endings. 1

A. Active.
229.
1st

Singular.

In

the Indo-European parent-speech -o

was the termination of the primary tenses of the Thematic Conjugation, while -mi was the termination of the
gation.

Unthematic Conju-

Secondary tenses had -m only.

Latin shows no traces

of -mi (on sum, see

202. 3)

-o

appears in the Present, Future,

and Future Perfect

Indicative.

Elsewhere in the Indicative and

everywhere in the Subjunctive (including some original Optatives)

-m appears,
230.

e.g.

amabam, amaveram,

sim, essem,

etc.

2d Singular.

The
for

Indo-European
Latin
-s

endings

were

-si

(primary)

and

-s

(secondary).

may

represent the
final short

secondary ending, or original *-si

vowel, so that
* leg-e-si.
legis,

may have lost its example, may be either for

*kg-e-s or

The endings

of the Perfect Indicative and of the Imperative have already

in 211
ff.,

been considered

223

ff.

CONJUGA TION.
231.

69
-ti

3d Singular.
-t

The
.

Indo-European
Apparently

endings

were

(primary) and

(secondary)
Cf.

in the earliest Latin, -t

e.g.

had become
feced,

-d.

early
-ti,

inscriptional

forms,

fhefhaked,
-t

fecld,

sied

on

the

other

hand,

became

and

very early

supplanted

the

-d of the

secondary tenses.

The
and
-/

closely related

Oscan

dialect exhibits this distinction of -d

assumed
232.
1

for early Latin.

st Plural.

The only ending appearing

The Latin ending
i

in Latin

is

-mus,
to

earlier *-mos,

which seems to stand in Ablaut relation ( 62)

Greek
233.
(the

-pes (dialectal).

2d Plural.

-it's

probably stands for

-f

-te

Indo-European ending of the secondary tenses)

either from the 2d Singular or the 1st Plural. '&

s bor-

rowed
234.

3d Plural.
-lit

The Indo-European
-ns.

endings were

-nit

(pri-

mary) and
-nt,

(secondary).

In the Italic languages

-nit

became
this dis-

while -nt

became

Oscan and Umbrian preserve

tinction,

but in Latin *-ns has disappeared, being everywhere

supplanted by -nt (for -nti).

B. Passive.
235.

The

distinguishing characteristic of the Latin Passive

is

the presence
tion,
is

of final

r.

This formation, in
Italic
is

its

wide

applica-

found only in the

family.

and Keltic groups of the Indo-

European

Its origin

not yet sufficiently clear to warrant

an attempted explanation here.

the Sanskrit ending -re of the
fectly certain
:

Some have connected it Perfect Middle. One thing is

is

with
per-

Latin r does not arise from the reflexive se as was

formerly held.

In general the Latin Passive

an outgrowth of
1st Singular

an earlier Middle.
1st Plural,

With the exception of the

to

and

Middle forms are seen

have been at the basis o f

the developed inflection.

70
236.
1
is

INFLECTIONS.
st Singular.
-or, e.g.

Where

the Active

form ends in

-o,

the

Passive

regor (earlier -or;

-in,

88. 2),

amabor.
-in, e.g.

Where
amer,

the Active ends in

the Passive has r instead of

amabar.

The

originally long

vowel before

-r

sometimes appears

in Plautus, e.g.

Asinaria, 62,fateor; Amphitruo, 559, loquar.

237.

2d Singular.

This

is

in origin

a Middle, formed with

the Indo-European ending *-so, the

tenses in the Middle.
Cf.

termination of secondary
is

Thus sequere

for *seque-so (

98.

1).

Greek

e7re-(cr)o, Ittov.

The ending
-s,

-ris arises

secondarily from

-re

by further appending

the ending of the 2d Singular Active.

(

Thus

sequeris for *sequere-s

73. 2).

This was possibly the

result of

an

effort to distinguish the Indicative

2d Singular from

the Imperative.

238.

3d Singular.

The

origin of the

3d Singular

in -tur

is

too obscure to be considered here.

239.

1 st

Plural.

In place of
regimu-r.

-s

of the Active ending -mus

we

have the Passive

240.
tion
;

-r, e.g.

2d Plural.

We probably have here a periphrastic formapresumably stand

for legiminl estis, in

legimini,
is

etc.,

which

legimini
fxevot.

a Middle Participle of the same type as Greek Xeyo-

This formation must have originated in the Present Indic;

ative

legebamifft, legemini,
after the

legamim, legerlmirii are

all

secondary,

formed
241.

analogy of legimini.

3d Plural.

The

origin of the

3d

Plural in -n tur

is

too

obscure to be considered here.

The
242.

Infinitive.
Infini-

In Latin, as in other Indo-European languages, the

tives are oblique cases

of verbal nouns which have become stereo-

CONJUGATION.
typed by usage.

171
cases have contributed

The Dative and Locative

most largely

to this category.

A. Active.
243.
Present.

This
-es-,

was apparently
suffix.

in origin the Locative of

?-eg-er-e

a noun with an
*reg-es-i (

-os-

Thus

for a

primitive

141), as

though from a
-si), e.g.

Nom.

*reg-os.

Unthematic

verbs appended -se (for

for *ve/-se.

esse, fer-re, for */er-se; vel-le

244.
-is-

Perfect.

The
In

Locative

-se

(for

si)

is

appended

to the

Aorist stem ( 213, 215),

e.g. vid-is-se.

245.

Future.

such forms as dicfurum

esse, it is

probable

that originally dicfurum was

not a Participle, but an Infinitive.

The form

has been plausibly explained as being contracted from

dictu *erom,

where dictu

is

Supine, and *erom (for *es-om;

98. 1)
is

the old Infinitive of the root es- (-esse).

This Infinitive

pre-

served in Oscan and Umbrian, though lost in Latin.

force of dictu

The
'

original

*erom would be

'

to

be

for saying/

i.e.

to

be about

to say' (on dictu see 252. 2).

The

foregoing explanation ac-

cords excellently with the use of dicfurum and similar forms without esse and (in early Latin) with a Plural subject,
cos
i.e.

e.g.

credo inimi-

meos hoc dicfurum,

'will say this'

'I believe

my

enemies are

for saying this,'

i.

(C. Gracchus, cited by Gellius,

esse

7).

After

the

analogy of periphrastic forms, dicturum

into

subsequently

came

vogue (though the form with

rise

esse

never came to be

predominant) and thus gave

in -urus, -a, -um.

to the Future Active Participle

B. Passive.
246.

Present.

Such

forms as

reg-'i,

dic-i are

Dative forms;
e.g.

139.

Other verbs append the Dative ending to -^-sterns,

etc.; so

curart, ?noneri, audiri, for *cura-es-i,

ferri for

*fer-s-~i.

172
Cf.

INFLECTIONS.
243.

No

Passive signification originally attached

;

itself to

these Dative Infinitives

essentially

at the outset they could not have differed

Infinitives of the Active.

from the Locative

The
is

dif-

ferentiation into Active

and Passive meanings was purely arbitrary.

of unthink that -er represents

The

Passive Infinitive in -ier (archaic and poetical)

certain origin.

Some
-ere.
e.g.

the
fairly

apocopated
frequent in
Agier,

Active ending

This seems to have been

biber for bib ere
;

colloquial Latin,
therefore,

tanger for tangere.

and

similar forms

might represent Passive

Infinitives

with an added Active termination.

247.
e.g.

Perfect and Future.

iri.

Periphrastic forms
its

are used here,

dictus esse, dictum

The

latter consists of the

Supine com-

bined with the Passive of eo in

impersonal use.

The
248.
*-s-nt-s

Participles.
suffix

Present Active.
(

The

here
;

is

-nt-,

e.g.

-sens for

102. 1)

in ab-sens, prae-sens

regens for *rege-nt-s.

*ej-o-, e.g.

The

oblique cases of tens are formed from the stem

eufitis for *ej-o-ntis.

249.

Future Active.
Perfect

250.

See Passive. The

245.

suffix

was

-tus,

earlier -tos, ap-

pended
tus,
t,

originally to the

weak form of the

root, e.g. dic-tus, duc-

tentus for *tri-tos ( 102. 1).

Where

the root
e.g.

ended

in

d ox
-sus
e.g.

ss or s arose

phonetically ( 108. 1),

sessus for *sed-tos

By an extension this spurious ending, became appended also to some guttural and liquid stems,
usus for *ut-tos.
tap- sus,

fix us, pulsus.

251.

The Gerundive.
is

The

origin of the termination -endus,

-undus

not yet determined.

CONJUGA TION.
Gerund and
252.
1.

J3

Supine.
is

The Gerund.

The

Gerund

probably a developas virtus colenda est

ment of the Gerundive.

might easily give
larly
2.

Such expressions
colendum

rise to a

est (impersonal), while simi-

patriae defendendae causa might generate a d~efendendi causa.

The Supine.

The

Supine in -um
suffix -tu- ;

is

an Accusative of a
is

Verbal noun formed with the

the Supine in -u

Locative formation from the same stem

(cf.

163).

CHAPTER

VIII.

ADVERBS AND PREPOSITIONS.

AD VERBS.
253.

Adverbs

are, in the

main, case-forms which have become

stereotyped as the result of highly specialized usage.

The

cases

most frequently thus employed are the Accusative, Ablative, Locative,

and Instrumental.
Accusatives.

254.

These

result

from various syntactical usages.

Thus
i.

Accusative of Result Produced (Gr.

176. 2

3)

e.g.

/nul-

lum, pl'erumque, plurimum,

comparatives,
2. etc. e.g.

aliquid, facile, fortius,

and other

Appositives,

vicem, partim, etc.; 310.

e.g..

3.

Limit of motion,
Ablatives.
in

foras.

255.
1.

Here
-~e

belong

Adverbs
;

(for -Id;

130)

from

<?-stems,

e.g.

pulchre,

sane
'

certissinie.

Bene and male

-0 (-for

result

from the operation of the

Breves Breviantes' law ( %%. 3).

2.

Adverbs

in

-od ;

130)

from

^-sterns, e.g.

cerfd,

continud.

Cf. early Latin meritod.

Cito

and modd

result

from

the operation of the 'Breves Breviantes' law ( 88. 3).

3.

Adverbs

in -a

(for

-ad ;

118) from ^-sterns,

e.g.

extra,

supra, 'infra, contra, supi'a, ultra, citra, juxfa.

exstrad,

Cf. early Latin

suprad.
1

Many

words, clearly Ablative in form, apparix.

See especially Lindsay, Latin Language, chap.

174

PREPOSITIONS.
ently

75

became Adverbs through the medium of Instrumental conCf. structions, e.g. una, recta, qua, ed, eadem {sc. via), etc.

34i- 5-

256. Locatives.
i.

Here belong
e.g. heri,

True Locatives,

vesperi, hunii, belli, nfilitiae, donii,

;

postndie ( 126;

173),

meridie

die

crastim

;
;

nocfu ; temere
also the Pro-

(originally, 'in the dark,'

and so

'blindly,' 'rashly')

nominal Adverbs
2.

Ju-c, illi-c, isti-c ( 197).

e.g.

Ablative in Locative function,

/oris.

257.

Instrumental.

Here

belong

sponte,

forte,

repente,

numerd, 'promptly' (originally a musical term,

'with the
e.g.

music,'

'with the beat'), saepe (originally, 'with frequency').

258.

Even

a few Nominatives have

become Adverbs,

adver-

sus ; rursus for reve?'sus ; prorsus for proversus.

259.

Many

adverbs
;

were originally

phrases,
;

e.g.

denuo

for

ae novo ( 103. 4)

ilico for in

*stloco ( 89)
-iter also

admodum.
e.g.

Some
breviter

have thought that Adverbs in

for breve iter, etc.

belong here,

Cf.

German kurzweg.

PREPOSITIONS. 1
260.
Prepositions are in the

main Adverbs which have come

to

have special uses in connection with certain cases.

Historically

they belong to a relatively late period in the development of language.

Originally the cases alone sufficed for denoting relations,

but as greater precision became necessary, the requisite definiteness of meaning

came

to

be expressed by various Adverbs, which

;

ultimately crystallized as Prepositions

bial

yet an independent adver-

usage often remained.

1

See especially Lindsay, latin Language, chap.

ix.

76
In the
earlier

ADVERBS AND PREPOSITIONS.

period of their employment, Prepositions enjoyed
later,

considerably more latitude of usage than

being freely com-

bined with almost any oblique case

ultimately, however,

most of

them became
This
is

restricted

to

combination with particular cases.

truer of Latin, for example, than of Greek,

is

where the older

also

freedom

quite apparent.

The Oscan and Umbrian

show

greater latitude than Latin.

261. A, ab, abs, au-.

1.

A, ab, abs go back

of the
final o, this

to

an Indo-European

* apo,

Greek

a-rro.

By

loss

became
and

in Latin ap-, seen probably in ap-

erio.

But

in composition
b,
e.g.

in phrases before voiced

for

consonants

p became

abdo

for

*ap-do; ab gene re

*ap genere, and

ultimately the form with b supplanted that with p.

Abs

is

formed
its

from ab by appending

-s,

probably the Genitive ending in

weak form
subs

( 138),

an element frequently employed

Cf.

in amplify-

ing prepositional and adverbial formations.

ic-;

ex

= ec-s)

from

(in suscipio for * sub-s-cipio ; 105. 1)

e

from sub; obsAttic cU,

from ob ; also Greek

by the side of Ik; by the side of

e.g.

ev<?,

whence

by the side of

a/x<f>is

d/x<t.

seems to have
(for

developed from abs

in

compounds,

avello

from *asvelio

*abs-vello; 105. 2),

and then

to have detached itself as a 'by-

form' of ab, abs.

2.

Au-,

Sanskrit

ava,

goes

back

to

an Indo-European ave.

It

appears in Latin only in aufugio, and auferd for *ave-fugw,

*ave-fero

by Syncope (92).

Cf.

auspex for *av(i)spex; augu-

rium,
3.

etc.

form of *apo with aphseresis of the

initial
cf.

vowel

is

po-, seen in

pond

for *po-s-{i)no

92; 89);
cf.

po-situs.
off,

Po-

also possibly appears in po-lio (root //-;

4.
is

It-no),

'rub

polish.'
later,

form af found in early inscriptions and occasionally

It is

of uncertain origin.

probably merely a dialectal variation

of ab.

PREPOSITIONS.
262.
tions

JJ

Ad

is

cognate with English

at.

In early Latin inscrip-

we

find a

form

ar-,

used before

and v

in composition,

e.g.

arfueruntj arversus ; also ar-biter, arcesso in classical Latin.

is

Ar-

probably of dialectal origin.

263. 264.
265.

Ambi-, Greek

a/x^l, is

probably an old Locative.

avrt, is

Ante for

*a?iti,

Greek

probably an old Locative.

261. 2) with

Apud seems an appended d.

266.
circus,
'

to

be Indo-European *apo

Circum,
ring,

circa,

circiter are
;
'

all

connected with the noun

is

circle,

circus

circum

the Accusative Singular,

;

used
late

first

as Adverb, later as Preposition

circa

is

probably a
(

formation after the

analogy of extra, supra

255.

3).

Circiter probably contains the

Cf. inter, propter,

Comparative

suffix -ter (

181).

sub ter.

267.
cis,

Cis, citra are

from the root

ci-,

'this.'

On

the final

-s

of

see

261.

2.

Citra has the comparative suffix ( 181).

On

the formation, see 255. 3.

268.

Clam evidently contains the root of

is

celo,

'

conceal.'

The

formation
269.
is

uncertain.

Com- (cum).

See 58.^).

The

relation of co- to co?n-

not clear.

270.

Contra.

See

255. 3.

271.

De

is

obscure in

its

formation and

its

relationship.

272.

Erga,

ergo

are

obscure

in

etymology and
(f)epyov,

formation.

They can have no connection with Greek

work.

1/8
273.

ADVERBS AND PREPOSITIONS.

Ex,
ec-,
ef-,
e.

See

105.

2.

On

the final s of ex

(=ec-s), see

261.

2.

274.

Extra

is

formed from ex by means of the Comparative

suffix tero- (

181).

On

the case-formation, see

225.

3.

275.

In

is

the unaccented form of Indo-European *en, Greek

h.

The

original

form of the Preposition

Ivho-Ot, 'iv-Sov.

is

seen in early Latin

is

en-do.
(indi-)

Cf.

Greek

Another form of endo

indu-

seen in indi-genus, ind-oles, and in several early Latin

e.g.

words,

indu-gredi.

276.

Infra.

Cf. inferns,

and see

255.

3.

277.

Inter, intra are

;

formed from
181; 255.
3.

in

by means of the Com-

parative suffix -tero-

278.

Intus contains the same suffix as seen in dnnnitus, f7in-

diftis, etc.

279. Jiixta
'

is

from the stem juxta-, a Superlative of jugis,

connected,'

'

continuous.'

For the case-form, see

255. 3.

280.

Ob

is

from an Indo-European
k-rr-i,

*ofi-i,

a Locative formation

kindred with Greek

(

to

which

it

stands in Ablaut relation

62).

The form
(

ob

has developed

from *op, exactly as ab

from *ap
is

261. 2); yet op- probably appears in op-erio, and

preserved in Oscan.

281.
irepi.

Per

is

for

an Indo-European *peri (Locative).

Cf.

Greek

282.
*posti.

Post, early Latin poste, apparently goes

back to a Locative

PREPOSITIONS.
283.
Prae, praeter.

179

Prae

is

very likely a Dative from pra-,

an extension of pr- (weak form of per-).

Cf pro(d) from

pro-.

Praeter bears the same relation to prae as inter to in; subter

to sub.

284.
forms.'

Pro, prS-, por-.

Pro

and pro were Indo-European

only
in

'by-

In

Latin, pro-

appears

composition,

chiefly

before
e.g.

(e.g.

profugio, profteor, proficiscor), but also elsewhere

.

protego, pronepos)
is

The d of prod-, seen

is

in prodesse,

pro dire,

etc.,

not original, but

Por-,
e.g.

probably borrowed from retro(d) or

red-.

in

por-tendo, porrigo, polliceor (for *por-liceor)

may

represent pr-,
all

weak form of

the root per- (

100.

2),

with

which

the above words are ultimately connected.

285.

Prope,

propter.

Prope
is

is

for

pro-\-pe.

Cf. quip-pe.
in, etc.

Propter bears the same relation to prope as inter to

286.

Re-, red-.

Reis

the earlier form; the

of red-

is

of

uncertain origin.

287.
(sequor).

Secundum

an Accusative from secundus,

lit.

'following'

288.

Se-, early

Latin
;

sed-,
so-,

preserved in

seditio,

an Ablative formation
sent the Ablaut of
se-.

seen in so-cors,

may have been so-brius, may repre-

289.

Sub, subter.
v-n-o

The

Indo-European form
rough breathing).

is

*upo.
initial
'ks,

Cf.
s
is

Greek

(with irregular

The
viz.

explained as containing a reduced form of ex,

so that

*{k)sup would represent the primitive formation.

For the change

of/

to b, see 261.

2.

On

subter,

cf. inter.

180
290.
*uper.

ADVERBS AND PREPOSITIONS.

Super,
Cf.

supra.

Super goes
Supra

back

to

an Indo-European
breathing).

Greek

v-n-ip

(with irregular rough

sustains the

For

the

initial s,

see 289.

same

relation to super

as intra to inter.

291.
-eris,

Tenus

is

probably the Accusative of an obsolete tenus,

lit. 'a.

stretch,' root ten-.

292.

Trans

is

probably the Present Participle of *trare seen in

; i.e.

intrare, penetrare

originally trans

flumen

niilites

duxit meant

he led his troops, crossing the river. 293.

On

tra- } see 105. 2.

Uls, ultra

from root

ol-,

'that'

{cf.

olle;

195), are the

pendants to

cis, citra.

294. Versus, versum,

etc.

See

258.

CHAPTER
SYNTAX.

IX.
1

THE
Names
295.

CASES.

of the Cases.

The English word

case

comes from the Latin

-m-wo-is
'

casus,

which

was a translation of the Greek word 7nwts.

(from trWcd,
'

fall), as a grammatical term, primarily denoted a

viation,'

change' or

de-

and was accordingly

'

first

employed

to

denote the oblique

cases, as being

deviations

'

(7n-wcreis)

from the Nominative.

-tttuo-ls,

The

Nominative
it

itself,

therefore,

was not

at the outset a

though

early

came

to bear this

name.
of the cases were
:

296.

The Greek names

ovo/iao-TiKi] (sc. TrTwo-t?),

yevLKrj, SoTi/07,

Nominative.

Genitive.

Dative.

aiTLaTiK-r},

Accusative.

kXtjtlkt],

Vocative.

See especially Brugmann und Delbruck, Grundriss der Vergleichenden

vols,
iii-v

Grammatik,
1

{Vergleichende

Syntax, by Delbruck), Strassburg,

Gram?7iatik.

893-1 900.

Landgraf, Historische

Lateinische
et

Riemann

et

Goelzer, Grammaire Comparee du Grec

du

Latin, vol.
Sprache,

ii.

Paris, 1899.

Drager, Historische Syntax der Lateinischen

Leipzig,

2 vols. 2d edition.

1878,

1 ii.

88 1.

Kiihner,
1

Ausfuhrliche
878.

Grammatik der Lateinischen

3d
edition.

Sprache, vol.

Hannover,

Schmalz, in Muller's Handbuch der

vol.
ii.

Klassischen

A/tertumswissenschaft,

Munich, 1900.

Riemann, La Syntaxe Latine.

mar,
vol.
ii.

4th edition.

Paris, 1900.

Roby, Latin Gram-

5th edition.

London, 1888.
181

82

SYNTAX.
so called because
it

The Nominative was

for

it

was the case employed

naming a substantive when

meant
yevtKo's

was simply cited as a word.

is

The

significance of the term ya/1/07

it

in dispute.

Some have

thought

'the case of source or origin.'

is

But the usual

meaning of

against this view.

It

probably meant 'the

case of the genus,' or 'the generic case.'

This view accords with

the regular use of the Genitive to restrict the

meaning of another
it

word by denoting the

class or yeVo? to

which

applies,

e.g.

love

ofpare?its, 'fskers of'men ,' tons of earth.

The Dative was

is

called Sotlk^, 'the case of giving,' though this

simply one prominent function of the case.

In calling the Accusative

cutultikt},

the

Greeks intended

to

designate this case as the 'case of

(clItul).

effect,' i.e.

of the thing caused

Here again the name designated but imperfectly the

For the Accusative indicates
also the per-

functions of the case.

son or thing affected, to say nothing of other uses.

~K\-qTiKr)

means

'calling case' or 'case of address.'

297.

The Romans
(sc.

in devising grammatical terms for their

'Ovo/xao-ri/oj

own

language simply translated these Greek names.

became

No7ninativus

casus)

In translating

yevi/oj

by Genetwus the
of source,

Roman grammarians

falsely interpreted the case as that

or origin, misled doubtless by the frequent use

of the

Greek
was

Genitive in that function.

falsely

Aon/07 became Dativus.

airiariKTJ

AiTiaTt/07

rendered Accusafivus, as though

accuse.
KXrjTLKrj

were derived

from

atTiao/x.ai,

became
the

Vocafivus.

The Greek
therefore
'

had no Ablative, and

obliged to coin a

for this case

;

new term
viz.

they

Romans were named it Ablatwus,

fairly

the case

of taking away.'
uses of the case,

This designation was

accurate for certain

;

those of the true Ablative

but

it

ignored
It
is

the Instrumental and Locative uses of the case ( 331).

uncertain just

duced.
tilian's

when and by whom these Latin names were introThey had become established as current terms by Quina.d.).

time (90

REVIEW OF CASE-THEORIES.
Review
298.
of Case-Theories.

83

Since the beginning of the last century, there has been

much

discussion concerning the original force of the cases both

individually

and

collectively.

299.

The

Localistic Theory.

The chief representative of

und Bedeutung.

this

was Hartung, who

set forth his views in

1831 in a work entitled

Ueber die Casus, ihre Bildung

Hartung started
from the

with the assumption (largely a correct one, according to the views of most investigators) that in language the development
is

concrete to the abstract,

nite sense concepts,

that words
later

at the outset indicated defi-

which

came

to

be used

in transferred

meanings.
in

Applying

this principle to the cases,

he assumed that

Greek and Latin there had been

(in addition to the

Nomi-

native and Vocative) three cases, one to designate each of the

three definite local relations, from, in, and

ciple
first

to.

Applying

this prin-

to Greek, he explained the Genitive as

the from-ca.se,
fo-ca.se.

the Dative as the in-ca.se, the Accusative as the

Latin, substantially the

For
that

same explanation was given, except

the Dative of the Greek has in Latin, according to Hartung, been

differentiated into
latter

two cases, Dative and Ablative, of which the

has entirely absorbed the ///-function, while the Dative has

developed new meanings.

Hartung's theory has been styled 'through-going' Localism.
It

asserted that the original Indo-European case-system (apart from

Nominative and Vocative) had

cases,
in.

originally

been limited to three

:

which expressed the three natural space relations

in the individual languages

to,

from,

Wherever

more

cases appeared

(as in Latin or Sanskrit), these

('

were held to be differentiations

Zersplitterungen

'

of the original three.

Whatever may be true

of the meaning of individual cases, comparative

sively proves that

grammar concluit

Localism in the form in which Hartung held

84
absolutely untenable.

SYNTAX.

is

case-system

of at least

six

clearly

distinguished oblique cases must have existed in the Indo-Euro-

pean parent-speech.

300.

The Logical Theory.

Michelsen,

in his

Casuslehre der

lateinischen

Sprache vom causal-localen Standpuncte aus, pub-

lished in

1843,

endeavored to apply logical categories to the

explanation of the cases.

According to him two principles are

fundamental
ity.

1) Causality (including cause

and

effect)

2) Final-

Hence

in every sentence,

he holds, we must have a cause, an

as the case

effect,

and a purpose.

The Nominative he regarded The

expressing the cause, the i^ccusative the case of the effect, the

Dative as the case of finality or purpose.

Genitive
these

and
cases

Ablative were also given special treatment, though

were regarded as not essential

Michelsen's theory
is false

to

logical

completeness.
is

But

in principle.

Language

not founded

on

logic,

and any attempt

to explain forms of speech as primarily

identical with logical categories will probably always be fruitless.

301.

The Grammatical Theory.

In

1845 appeared Rumpel's

Casuslehre in besonderer Beziehung auf die griechische Sprache.

This book was a protest against the Localism of Hartung on the

one hand and the logical theory of Michelsen on the other.

Rumpel asserted the purely grammatical character of The Nominative he defined as the case of the Subject,
sative as the case

the cases.
the Accu-

used to complete the meaning of the verb, the

Genitive as

the

adnominal case

or

case

used

to

complete

the meaning of a noun, while the Dative was used to modify the

meaning of the sentence

a verb,
to
it

as a whole.

Where

the Genitive limited

as

was explained as denoting an internal relation

relation,

opposed

an external

such as that denoted by the Accusative.

As Rumpel concerned himself only with Greek, he propounded no

theory of the Ablative.

THE ACCUSATIVE.
302.

l8$
better

Subsequent Views.

Rumpel's theory shows much

is

method than

either Hartung's or Michelsen's.

Yet the gramDiscussion

cases are

matical theory of the cases

since Rumpel's day has

not universally true.

that while

shown

some of the
were

undoubtedly grammatical
tainly local.

in their origin, others

just as cer-

To

the Grammatical cases belong with certainty the

Nominative and the Genitive, the former as the case of the subject, the latter as

the adnominal case.

To

the local cases belong

with certainty the Ablative, as the from-ca.se, the Locative, as the

in-ca.se,

and the Instrumental,

Diversity of opinion

as the case denoting

exists as to the

association

with.

still

Dative and to

some
take

slight extent as regards the Accusative.

it is

If

we regard
;

the

Dative as originally the case of direction,

it

a local case

if

we
a

as originally used to

modify the sentence as a whole,

is

it is

grammatical case.

The Accusative
but there
is

usually regarded as simply

is

completing the meaning of the verb, and

a grammatical case
;

therefore classified as
for considering
it it

some warrant

as originally denoting the goal of motion, in

local.

which case

would be

See

311.

The
303.

Accusative. 1

The

distinction

between the Accusative of the Person or

Thing Affected (Gr.

175) on the one hand and the Accusative

is

of the Result Produced (Gr. 176) on the other,

one of funda-

mental importance.

Other designations are often employed to

distinguish the two types.

Thus the Accusative of the Person or

But these designations are

Thing Affected

is

called External Object, the Accusative of Result

likely

Produced the Internal Object.

to prove too philosophical for elementary pupils.

German
'

scholars
'

employ

also the designations

1

'

Akkusativ des Affekts

and

Akkuthe

sativ des Effekts,

terms which might be advantageously imitated

in English,
x

if

our language only had the noun Affect.

311.

When

For the original force of the Accusative, see

86

SYNTAX.
to the Accusative,

Greek philosophers gave the name atnariK^

they had in mind only the second of the two uses of the Accusative

now under
or, as

consideration,

viz.
it,

the Accusative of the Result

Produced

they designated
').

of the Thing Caused ('Internal

in transferring the
it

Object,' 'Effect

The Romans,

Greek name

of the case to Latin, should have rendered

as Caitsativus (a designation actually

by some such word

employed by Priscian) or
would have been accu-

Effectwus.

Either of these would, like the Greek original, have

{cf.

been a defective name

rate as far as
it

296), but

it

went.

304.

The Accusative with Passives used as Middles.

The
2.

treatment of the Accusative after Passive Verbs in Gr.

is

175.

a)

based on the elaborate discussions of Schroder,

Der

Accusativ

nach Passiven Verben in der Lateinischen Dichtersprache, Grossglogau,

1870

Engelhardt,

Passive Verba

mit dem Accusativ,

his Ausfilhr-

Bromberg, 1879; and the treatment of Kiihner in

liche Lateinische

Grammatik,

ii.

71. b).

The

explanation of the
is

Accusative as Synecdochical
given for

{cf.

Gr.

180), which
It

sometimes

this construction, is

not adequate.
is

might explain

such phrases as ductus tempora hedera, but

irrational for

galeam

induitur, riodd sinus collecta, laevo siLspensi loculds lacei'td,

and

many

others.

On

the other hand, the interpretation of the Pas-

sive in

such instances as a Middle, and the Accusative as the

all

Direct Object, furnishes a satisfactory explanation of of this type.

phrases

Sometimes by an extension of usage the Middle

indicate that the subject lets
himself, or has
it

is

employed

to

some
Aen.

action be

consummated upon
his hair cut.

done.

Cf. English he
ii.

had

An

illustration of this is Virgil,

273,

per pede~s

t?-ajectus lora,

'having had thongs drawn through his

in

feet.'

For a few instances

which a Synecdochical Accusative occurs with Passive verbs,

see 307.


THE ACCUSATIVE.
305.
tions
1

87

Accusative of Result Produced.

The

different construc-

grouped together under Gr. Cognate Accusative as the

176. 1-5, are often referred

to the
all

original

from which they have

is

developed.
its

The Cognate
it

Accusative, however,
to regard
it

so restricted

in

scope that

seems better

as a subdivision of a
Cf.

larger category rather than as the basis of such a category.

Brugmann,
Tv-rrTeiv e'A/os

Griechische
{strike

Gram/natik
i.e.

3
,

439.

a wound,

produce a

who wound by
2,

classifies

striking)

and

vLKav vlktjv,

win a

victory, as parallel subdivisions of the gen-

eral category of the

Accusative with Verbs of producing.

306.

Accusative of Person Affected and of Result Produced

Dependent upon the Same Verb {Gr.

acter of this construction
sative of Result
is
is

178).

The
e.g.

true

char-

best seen in phrases where the Accu-

a Neuter

Pronoun or Adjective,

fe

haec rogo,

id nie doces, the essential point being that the Latin was able not
only to say id doces (Ace. of Result)

and ml doces (Ace. of Per-

son Affected), but to combine the two constructions in a single

phrase.
in
It is a

misconception to regard the Accusative of Result

less the

such sentences as any

Direct Object than the Accusathe two Accusatives

is
is

tive

of the Person Affected.

Each of

Direct Object equally with the other.

There

no

essential differ-

ence between the construction of haec in haec me rogas and the

construction of haec in haec rogas.

In

many

instances the Accuetc., is

sative of Result with verbs of asking, teaching,

clearly of
fe

secondary origin,
celavi

e.g.

t~e

sententiam rogo, after

tl

hoc rogo;

sermonem

after fe

id celavi.

307.

The Synecdochical
little

or Greek Accusative

{Gr.

180).
Cf.

There can be
Quintilian,

doubt that

this construction is a
it

Grecism.

ix. 3. 17.
its

Some have claimed

who
imitate

as a genuine Latin

idiom, but

almost total restriction to the poets of the imperial

age and to the prose writers

them

is

against any such

88

SYNTAX.
The names
'

theory.
tive of

Accusative of Specification' and 'Accusathis construction.

Respect' are sometimes used to designate

it

With Passive verbs

is

better in most cases not to recognize a

Synecdochical Accusative.
usually be classed under

Apparent cases of the construction can

Gr.

175.

2.

d), but in some twenty inthat

stances in the Augustan poets

in

and in about twice

number

Lucan,

Silius, Statius,

and Valerius Flaccus, we must recognize

the Synecdochical Accusative with Passive verbs.

308.

Accusative in Exclamations.

This construction
is

is

appar-

ently the result of ellipsis.

Just what verb

is

to be supplied in
is it

thought in particular instances,

rial

not always clear, nor

mate-

that

it

should be determined.
of the Infinitive.
is

309.

The Accusative as Subject

The Accueum

sative as Subject of the Infinitive

an outgrowth of the use of

history of the construction
like jussi

the Accusative as Direct Object.

The

may be illustrated as follows abire, eum was originally the

was a noun
being:
in the Locative
to

In an expression

object of jussi, while the Infinitive

( 243), the force of the entire

phrase

a going ( 351). But in course of time the eum abire came to be felt as a whole and as sustaining an

I ordered him

object relation to the verb, a conception which led to such expressions as jussit pueros necari,

where pueros could never have been

once the construction of the Accusaestablished,
its

the object ofjussit

tive

When

with

the

Infinite

became

extension

was

rapid.

Expressions like jussit pueros necari easily led to dixi

esse,

pueros necatos

whence pueri necati

esse

dicebantur and other

types of Infinitive usage.

310.

Id genus, muliebre secus,

etc.

1.

Id genus
it

is

clearly

appositional in origin, as indicated by the fact that

regularly oci.e.

curs only in combination with a Nominative or Accusative,

not

virorum id genus, but usually

viri id genus, viros

id genus,

etc.

THE ACCUSATIVE.
2.

89

Muliebre secus, virile secus, while doubtless of the same

origin as id genus, have nevertheless

in actual use.

advanced a stage beyond

it

We
lit.

find not only liberi muliebre secus, 'children of

the female sex/

'children, the female sex' (of children), but

also liberdi'iim (liberis) muliebre secus.

3.

Meam

vicem,

tuam vicem,

etc.

The appositional or prediho die sese excruciari meai?i

cate origin of this phrase seems to be indicated by such early Latin

usages as Plautus, Mostellaria, 355,

qici

vicem possit pati, 'who can

tute;' Captivi, 697, ut
release
4.

let

himself be tortured, as

my substivicem, 'to
for us two.'

eum

remittal nostrum
lit.
'

amborum

him

in return for us two,'

as

an exchange

Magnam
ix.

partem,

maximam
is less

partem.

The

appositional

origin of these phrases

v.

certain, yet expressions like Livy,

14 and

37.9,

maximam partem ad arma

trepidanfes caedes

opp7-essit,

seem

to point in that direction.

311.

Original Force of the Accusative Case.

(</

Rumpel

in his

Casuslehre, published in 1845

3 OI )> contended

that

the

Accusative served simply as the complement of the verb, and that

all

the varieties of meaning, such as limit of motion, duration of

etc.,

time, direct object,

are but varieties of this primary function.

Rumpel
case,

accordingly regarded the

this
It is

Accusative as a grammatical

and

view has

been maintained by most subsequent

all

scholars.
e.g.

advocated to-day by

the leading authorities,

Delbruck, Brugmann, Hubschmann,

others.

Holzweissig,
is

Gadicke,

and

This theory,

it

must be admitted,

both simple and

rational.

Yet there have always been some scholars who have

recognized the goal-notion as representing the original force of

the Accusative.

While

it

is

impossible to prove the truth of this

its

latter theory, yet the

arguments in
:

favor deserve consideration.

They
1.

are the following

The antecedent

probability of the existence of a case denotis

ing

to

a place, person, or thing,

very great.

It is

admitted that

I90
the parent-speech

SYNTAX.
had an m-ca.se (the Locative) and afrom-ca.se

(the Ablative), so that a to-case might naturally be expected as

the complement of these.

2.

There are advantages

for the Accusative.

in starting with a concrete, tangible

meaning

Language undeniably develops from

the concrete to the abstract.

3.

The

goal-notion

is

shown by the testimony of those Indoliterature

European languages whose

reaches furthest back, to

this

have been an extremely primitive force of

Sanskrit

case.

Thus

and Homeric Greek exhibit the goal-meaning of the

it

Accusative, while the vestiges of

historic times
it

in Latin indicate that in pre-

had been more frequent.

rus, to

Thus the use of town

as exsequias Ire, mfttias

names and of domum, domos,

pessum
dare,

denote the goal of motion,

and the occurrence of such expressions

Ire,

venum

dare, point to a freer use of the

in

same

kind in early times.

force.
It is

The Supine
become

-um

also

shows

this primitive this goal-use

noteworthy that in post-Homeric Greek

obsolete.

of the Accusative had

stands

Post-Homeric Greek
In both of

upon the same ground

as Latin in this respect.

these languages the practical disappearance of the goal-notion in

historical

times would seem to indicate that

as other

uses de-

veloped the original function gradually passed away.

4.

The

other uses of the Accusative

may

all

be

satisfactorily
first

derived from the goal-use as the original one.

As the

and

most obvious developments must be considered the Accusative of

Extent of Space and of Duration of Time.
processit

Thus

znginfi

rriilia

would

originally

have meant

'

he advanced to the

limit of

twenty miles,' whence arose secondarily the notion of extent.

Similarly viginfi annos vixit
to the limit of

would have meant

originally 'he lived

twenty years,' whence secondarily 'he lived throughIn the case of the Direct Object, the Accusa-

out twenty years.'

tive

may

also have orginally designated the limit of the action of

the verb.

Thus aedes struxit would

originally have

meant 'he per-

THE DATIVE.
formed an act of building, the goal of which was a house.'
larly video

191
Simi-

hominem,

'I

perform an act of seeing, the goal of

which

is

a man.'

Cf. the similar

idiom prevalent in certain Ro'

mance languages, e.g. Spanish yo veo al hombre, lit. I see, to the man' = 'I see the man.' The so-called Accusative of Specification,

which, so far as
),

it

appears in Latin,
least
like

is

apparently a Grecism

307

would be the
Yet expressions

obvious development of the goal-

notion.
as

umeros

similis deo,

lit.

'like a

god

to the shoulders,'

may

be explained as

originally

meaning
i.e.

'looking to the shoulders,' 'as regards the shoulders;'

the

shoulders are conceived as the thought limit to which the state-

ment

is

referred.

The
312.

Dative.

The Dative probably

of.

originally designated motion towards,

It

motion in the direction

was accordingly a
it

localistic

case.

Some, however,

as Delbriick, regard
it

as a

grammatical case, and

think that originally

like the so-called
ficult to

was a mere sentence modifier, very much

Dative of Reference.

But

it is

much more

dif-

develop the notion of direction from the force of the

It

Dative as a sentence modifier than vice versa.

simpler to assume this concreter meaning as

therefore seems

the original one.

In that case the poetical construction of the Dative to denote

direction of motion

(Gr.

193) would represent the

original

meaning of the

case.

313.
is

Dative of Indirect Object.

The Dative of Indirect Object

case.

a very obvious development of the notion of direction, just asas the original

sumed

meaning of the Dative

would
originally have
'

Thus
'I
;
'

tibi

hoc
in

died, 'I tell

you
'

this,'
;

meant

tell this

your direction

so tibi ignosco,

pardon you

nana

nobis

inipendet, 'ruin threatens us.'

192
314.
It
is

SYNTAX.
Indirect Object with Verbs signifying 'Favor/ 'Help,' etc.

common
many
is

conception that the Latin

is

peculiar in con-

struing

verbs of these meanings with the Dative; but this

erroneous, and largely due to the loss of inflections
original distinction

impression
in English,

whereby the

between the Anglo-

Saxon Dative and Accusative has become obliterated, so that the

English
'

Objective

'

is

As a matter of

fact

commonly many verbs

felt

as

an Accusative.

of the category under consider-

ation were intransitive in Anglo-Saxon

and

in

Teutonic generally,

and accordingly governed the Dative

gives clear illustration of this.

case.

Modern German

Cf. e.g. ich glaube Ihnen, ich verzeihe

Ihnen, ich traue Ihnen, ich helfe Ihnen.

Latin, therefore, does

not differ from English and the other Teutonic languages in taking the Dative with these verbs
;

on the other hand there

is

a strik-

ing agreement,

when we come

to examine the matter from the

historical point of view.

315.

The

Indirect Object with

Compound Verbs.

It

is

a mis-

conception to suppose that

the mere fact of composition with

certain prepositions was the occasion of the

employment of the
Some-

Dative ease.
essentially

Prepositions

when

prefixed to neuter verbs often

modify the previous character of the verb.

times they
tive)

make
2.

the verb transitive

{i.e.

the verb becomes transi(e.g.

and

it

then governs the Accusative

a).

mire magistratutn.

Cf Gr.
meaning

175.

More

frequently a neuter Verb,

when come.g.

pounded with
as to

a preposition,

becomes only

so far modified in

admit an indirect object, not a direct one,

perlculis incurrit.

Sometimes

also

composition changes the char-

acter of a transitive verb,

making the compound incapable of

e.g.

governing a direct object, though admitting a Dative,

obsequor.

But
to

in all these
fact
all

the use of the Dative should be

to the

referred not

the

of composition, but

meaning of the verb.

Least of

should the Dative be regarded as depending upon

THE DATIVE.
the preposition,

93

an
of

error often

propagated in the minds of

elementary pupils.
316.

The Dative

Reference

is

an outgrowth of the original

It is

notion of direction belonging to the Dative.

a somewhat less

obvious development than the Dative of Indirect Object, representing as

it

does a somewhat weaker relation.

Thus

in a sentence

like nobis hosfes in

conspectum venerant, the Dative represents the

direction of the thought as a whole rather than of the action in-

dicated by the verb.

The name
is

'Dative of Interest' sometimes

in

applied to this construction

'

somewhat narrower
is

scope than

Dative of Reference,' and hence

'

less satisfactory.
'

The
'

sub-

division of the construction into

Dative of Advantage and Dative

of Disadvantage

'

is

also quite useless.

These designations tend

to obscure the real character of the construction, calling attention,

as they do, to

what

is

merely accidental.
'

division of the

Ac-

cusative of Direct Object into

Accusative of Advantage' and

justified.

'Accusative of Disadvantage

'

would be equally

317.

The Ethical Dative.

This
is

is

simply a special phase of

the Dative of Reference, and

rate category only because
it

entitled to recognition as a sepaits

represents the Dative in

most
con-

attenuated force,

often, in fact, quite untranslatable.

It is

fined to the Personal Pronouns.

318.

Dative of Agency;

Dative of Possession.

These
it

are

both developments of the Dative of Reference.

Thus haec mihi

is

agenda sunt originally meant

reference to

'

this is to

be done and

with

me

that this

is

true,' i.e.
'

'I must do this.'

Similarly
it

nobis sunt agri originally us that this

is true,' i.e.

meant

there are lands,

and

is

of

'we have lands.'

319.
Object,

Dative of Purpose.
is

This,

like

the Dative of Indirect

a perfectly obvious development of the original notion

'

194

SYNTAX.
Thus
receptifi canere, 'to
'

of direction belonging to the Dative.

sound

the signal for a retreat/

;
'

was originally
rei

to

sound the signal

in the direction of a retreat

publicae cladi sunt similarly

meant

'

they are in the direction of

damage

to the state.

The
320.
case,

Genitive.

The

Genitive

is

best regarded as primarily an adnominal

i.e.

as originally
It
is

used with nouns to define their meaning

more

closely.

therefore a grammatical, as opposed to a

local, case.

The
and

use of the Genitive with verbs must be regarded

as

as secondary,

developed from

its

use with nouns by some

association or analogy.

321.

Genitive with Nouns.

The

special kind of closer deter-

mination expressed by the Genitive, depends upon the context.

There was no one type from which the others developed, but

all

of the varieties enumerated, in Gr. 195 (excepting the Genitive

of Quality)
are

equally

primitive.

Most of these
Genitive
is

call

for

no

special comment, but the Objective

noteworthy as
first

exhibiting at times a wider extension of application than at

belonged to

it.

Theoretically the Objective Genitive

is

used only

with verbal nouns whose corresponding verb governs the Accusative.

Thus amor patris corresponds

to metuere deos, etc.

to

amare patrem, metus

dedrum

But by an extension of usage we

frequently find the Genitive used with nouns derived from verbs

which govern other

case construction

cases,

and even from verbs which admit no

Typical examples are
excessus
vitae,
:

whatever.

consifetudo

hominum, 'intercourse with men';

from
life
;
'

'departure

~ira

praedae amissae,
argenfi oratio,
are
usually

'

anger on account of the loss

of the booty';
relations,

'talk

about the

money.'

These
by

however,

more accurately expressed

means of

prepositions.

THE GENITIVE.
322.
Genitive of Quality.

95

This seems

to have

been of second-

ary origin and to have developed from the Subjective Genitive.

Thus
man.'

homo magnae

virtuiis

was probably
the

originally

'Virtue's

In conformity with this origin,

Genitive of Quality-

more commonly denotes a permanent quality, as opposed to the Ablative of Quality, which was primarily employed to designate
qualities

which were more or

less transitory.

For a completer

statement of the difference between the Genitive of Quality and

the Ablative of Quality, see

345.

323.

Genitive with
equally

Adjectives.

This
is

construction

must be

regarded as
nouns.

primitive

with

that of the

Genitive with

Cupidus

la?/dis, for

example,

just as original a construc-

tion as cupiditas laudis.

As regards the construction with

similis with the Genitive

similis,

many

fine-spun theories

have been propounded to account

for

the difference between

and

similis with the Dative.

The

dif-

ference, however,

is

probably merely one of chronology and not

of meaning.

In the earliest Latin This

is

we

find similis construed only

with the Genitive.

probably Plautus's unvarying usage.

in,

Later the use of the Dative begins to creep

doubtless after the

analogy of

par and

similar

words construed with the Dative, and

as time goes
until

on the Dative gains the supremacy more and more,

Latin
the Genitive
is

in

Silver

comparatively rare.

See

Jones,

Thomas M.,

Case- Constructions of Similis

and

its

Com-

pounds, Baltimore, 1903.

324. Genitive with Verbs.

If the Genitive

was primarily an

adnominal case,

its

use with verbs must be of secondary origin,

and

is

due either

to

some analogy whereby the verb adopts the

else to the
ellipsis

construction of a

noun of kindred meaning, or

of a governing word.

I96
325.
Genitive

SYNTAX.
with Memini, Reminiscor, Obliviscor.

With
felt as

verbs of remembering the use of the Genitive apparently comes

from associating the verb with mentor.

Thus memini was

its

memor sum.
memini.

Obliviscor

followed

the analogy of

opposite

Cf

English differ with after the analogy of agree with.

L.,

See Babcock, C.

Study in Case-Rivalry, being an Inves-

tigation regarding the

Use of the Genitive

and
(

Accusative with
Studies
in

Verbs

of _Re?ne?nbering

and

Forgetting.

Cornell

Classical Philology, Vol. XIV.)

New
etc.

York.

Macmillan, 1901.

326.

Genitive with Admoneo,

felt

Here the verb of remindmemorem

reddere,
this principle.

ing was probably

as equivalent to aliquem

and was construed with the Genitive on

327.

With Verbs

of Judicial

Action the Genitive

ellipsis

is

plausibly

explained as

resulting

from an

of the governing word,

is

crimine, judicio, nomine.

Thus Verrem avaritiae coarguit

for

to

be regarded as standing

Verrem avaritiae crimine coarguit;

Occasionally
cri-

'he convicts Verres on the charge of avarice.'

mine was expressed,

rationis crimine ;
iii.

e.g.

Tacitus, Annals,

vi.

14. 2

cecidere conju-

44. 8 maiestalis crimine reum.

328.
is

Genitive with Pudet, Paenitet,

to

etc.

The

Genitive

here

held

depend upon the noun notion implied

in the verb.

Thus pudet suggests pudor;

co7'dia, etc.

paenitet, paenitentia ; miseret, miseri-

329.

Interest and Refert.

The Genitive here

i.e.

is

probably the
interest

Subjective Genitive used

predicatively,

patris

rem

familiarem curare
curare.

is

quite analogous to patris est

rem fa?niliarem

For the Ablative Singular Feminine of the Possessive

with refert and interest, see

349. 3.

THE ABLATIVE.
330.
Genitive with Other Verbs.
e.g.

97

With

verbs of plenty

and
is

want,

compleo, impleo, indiged, the Genitive, where used,

employed

after the analogy of its use with adjectives of plenty

after
this

and wa?it; thus compleo

But with most verbs of
construction.

plemcs ; indiged after egenus,

is

etc.

category the Ablative

the regular

Potior

when construed with

'

the Genitive follows

the analogy of pofens,

master

of.'

The
331.

Ablative.
i.e.

The

Ablative

is

a so-called syncretistic case,

a case

resulting from the fusion of

more than one

original case.

The
and

Ablative represents three original Indo-European cases, viz. the

true Ablative or fr07n-ca.se, the Instrumental or with-ca.se,

the Locative or in-ca.se.

Evidences of the fusion referred to are

found both in the forms and in the functions of the so-called

Ablative.
a)

Forms : Only

a portion of the forms designated as Ablative

are historically such.

Thus

in ^-sterns the Ablative Singular

is

true Ablative (e.g. porta, for

portad';
is

118).

In the Plural of

^-sterns the so-called Ablative

probably an Instrumental.

The
141),

same

is

true of ^-sterns as of ^-sterns.

-e {e.g. niilite) is

In Consonant stems the

(

Ablative Singular in

probably a Locative

while the Plural forms ending in -ibus are true Ablatives.

-i-,

In the

-u-,

and

-e-

stems both the Ablative Singular and the Ablative

Plural are true Ablatives.

b)

Functions

The

triple function

of the so-called Ablative

also points clearly to a triple origin of the case.

Thus we

find

from-uses, with-uses, and zVz-uses (the

others) side by side.

last

much

rarer than the

Notions so radically distinct could hardly

have developed from a single original case.

By the Romans, of course, the They were totally ignorant of its

recognized
its

Ablative was

felt

as a single case.

syncretistic origin, although they

great diversity of function.

198
332.

SYNTAX.
Causes of Syncretism in the Latin Ablative.

Despite
wash with

their radical differences of

meaning, the Locative, Ablative, and

Instrumental cases naturally possessed certain points of contact.

Thus aqua lavare might have meant

water
'

originally either

'

to

or

'

to

wash

in water,'

i.e.

might be expressed either by

Similiarly
'

the Instrumental or

the

Locative.
'

equo vehi might

'

mean

'

to

be borne on a horse on

or

by a horse

onus umerd

sustinet, 'he bears the load

his shoulder' or 'with his shoulder';

etc.

carris ve?iiu?it,

'

they

come

with carts 'or 'on carts,'

These

examples

all

show points of contact between the Locative and

Instrumental.

The
or
'

Ablative and Instrumental also have certain

points of contact.

Thus Ira
or

ard'ere might
'

mean

either 'to burn

with anger

'

from anger
'

lacte

vivunt might

mean

either

'they live from milk

'

by

milk,' etc.

Points of contact between

less

Locative and Ablative are naturally

English expressions as
'

much
and

frequent, yet such

to receive at the

hands
'

of and
is

'

from the
from the

hands of;'
west,'

'

the wind

is

in the west'

the wind

show

that even here contact was possible.

Ablative,

Instrumental, and Locative, therefore, to a certain

extent occupied

common ground

in the field of thought,

and

this

circumstance ultimately led in Latin to a complete fusion of the

three

and

to the establishment of a single syncretistic case,

the

Ablative.

Genuine Ablative Uses.

333.

The

true Ablative designated dissociation or the point of

departure.

When

the dissociation
;

is

external,

we

call
is

the coninternal,
is

struction Ablative of Separation

when

the dissociation

we

call

it

Ablative of Source, a construction which in prose

con-

fined to narrow limits.

The

Ablative of

Agency

is

also a develop-

ment of the

true Ablative, the agent being conceived as the source

;

from which the action emanates

e.g.

in a Caesare accusatus est

the action was primarily conceived as emanating from Caesar as

its

source.

THE ABLATIVE.
334.
Ablative of Comparison.

99

This

construction also reveals

the original conception of point of departure.

Thus

melle dulcior

primarily

meant

'

sweeter, reckoning from honey as the standard,'

and so

in similar expressions.
this

An examination of Cicero's
Comparison
is

orations

shows that in

writer the Ablative of

mainly

restricted to negative sentences, to interrogative sentences imply-

ing a negative, and to a few stock phrases such as luce clarius,

latins opinione, etc.

When plus,
lents of plus

minus, longius, and amplius are used as the equiva-

quam, minus qua?n,

appositional.
l

etc.,

the plus, minus,

etc.,

were
urbes

probably

originally

Thus amplius
cities,

viginti

incenduntur originally meant

fired.'

twenty

(aye)

more were

This explanation, of course, involves the assumption that

originally a different order of the

this type,
e.g.

words existed

in sentences of

viginti urbes, amplius, incenduntur,

and

this

assump-

tion

is

borne out by the repeated occurrence of

xii.

this order, e.g.

Tac. Ann.
for fifteen

43 quindecim durum alimenta, non amplius, 'food days, not more'; Livy xxix. 32. 5 cum quinquaginta,
equitibus,
'

haud amplius,

with

fifty

horsemen, no more.'

For a

detailed discussion of the Ablative of Comparison, see Neville,

K. P. R., The Case- Construction after the Comparative in Latin.

{Cornell Studies in Classical Philology, Vol. XV.)

New

York,

Macmillan, 1901.

Instrumental Uses
335.

of the Ablative.

The Instrumental was

primarily the case of association or

with-case.

336.
first

Ablative of Accompaniment.

This

is

logically

one of the

and most obvious developments of the

is

sociative idea.

The

construction

not frequent, however, being confined mainly to

military expressions.

Gr. 222.

1.

200
337.

SYNTAX.
Ablative of Association.
strictly applies

Besides the idea of accompaniin Latin, yet

;

ment (which

only to persons in connection with a

verb of motion) the Ablative also sometimes denotes association. This construction was never

common
,

it

should be

recognized in a limited set of expressions

Jungere, miscere, mutare, permutare

thus with jungere, cone.g.

assuetus,

libido
'

scelere

jimcta,

'

lust

joined with
'

crime

;
'

mella vino miscere,

to

mix

honey with wine

war
*

bellu?n agricultura

permutant,
'

'

they exchange
to
toil
'

for

farming

;
'

assuetus
').

labore,
all

accustomed

(lit.

familiarized with toil

In

of these expressions and in

it

some

others of less frequent occurrence,

seems better to recognize

the primitive sociative force of the Instrumental, rather than the

Ablative of Means, as
is

done

in

Gr. 218.

7.

For a

fuller

discussion of this Ablative of Association, see Bennett in Transactions of the

American Philological Association, Vol.

XXXVI

(1906),

pp. 64

rT.

338.

Ablative of Attendant Circumstance (Delbriick's

'

Instru-

mental

der Begleitenden

Umstande
is

;
'

Vergleichende

Syntax,

195).

This
'

construction also
in

a direct

outgrowth of the

sociative

idea inherent
*

the Instrumental.

Thus dat sonifu

no one
cele'

magno stragem means occasions

loud crashing
brate
dies
;

destruction in connection with a

fletu faxit,
;
'

nemo mea funera

'

let

my

obsequies with weeping

exstinguitur ingenti
etc.

lilctu,

he

under circumstances of great sorrow,'

339.

The
with

Ablative of Manner

is

another obvious development

gravitate, loquitur,

of the sociative idea.

Thus

in

magna

'he

speaks

great

impressiveness,'

the

'impressiveness'

was

primarily conceived as an accompanying feature of the speaking.

is

'Manner'

differs

from 'Attendant Circumstance' in that

e.g.

it

regularly restricted to abstract words,

etc.

celeritate,

virtute,

dignitate,

THE ABLATIVE.
340. Ablative of Accordance.

201

The construction treated under

mea senconnected both with Manner on
viz. sifis

Ablative of
tentia, etc.,

Manner
seems

in

Gr.

220. 3,

moribus,

to be closely

the one

hand and Attendant Circumstance on the

and pronounced
that
it

other.

The

type

is

so definite

deserves clear recogni-

tion in our Latin teaching.

Another excellent example of the

3,

construction
bio
cii77i

is

seen in Cic. de Sen.

pares autem vetere prover-

paribus facillivie congregantur, 'according to the old

'

proverb, "birds of a feather flock together."

341.

Ablative of Means.

The

notion of

Means

is

an outis

growth of the idea of Association.

primarily 'he,
this sociative

Thus, hostem telo percussit

along with a spear, smote his enemy.'

Out of

idea the notion of

means or instrument developed

secondarily.
in

Yet there are few instances of the Abative of Means

in

which traces of the sociative notion are not apparent, and

this

some cases
With
is

idea

is

very prominent,

e.g.

deos precibus adorare,

'to worship the gods with prayers.'

1.

iitor,

fruor,

fungor,

potior,

vescor,
i.e.

the Ablative

of

Means
2.

a natural result of the Middle,

self,'

reflexive, use of these

verbs, 'benefit one's

'enjoy one's
is

self,' etc.

With opus

est the Ablative

a secondary construction after

the anology of usus est with the Ablative.

'there
is

In usus

est aliqua re,

need of something,' the Ablative was

'there
is

originally

one of

Means,

lit.

service

by means of something.'
as

notion of use the. notion of need arose secondarily.

ich brauche etwas,
earlier
'

From the Cf. German

need something,'

an outgrowth of the

meaning,

'

use something.'

Besides the use of usus est

e.g.

with the Ablative,

est, 'this
is

we

find usus

used predicatively,

hoc usus

necessary.'

Now

in the case of opus, the predicate

;

construction was probably the earlier

opus

is

best taken as the

Genitive of ops, 'help, service.'

relic of

The formation would then be

etc.

Genitives of the type of nominus, necessus,

138).

202

SYNTAX.
est

At the outset hoc opus

'this is necessary.'

meant

'this is of service,' secondarily

Early Latin exhibits

its

many

instances of this

predicative use of opus in

the

original meaning, 'of service,'

{e.g.

and
ii.

same

force

is

noticeable at times in Cicero

19. 4),

de

Or.

296), Livy

{e.g. xliii.

and

later writers.

The

construction

opus est aliqua re seems to be historically later than the predicate

construction,

and

to have

developed
this

after the

analogy of usus est

aliqua

re.

It is in
it

view of

theory of the origin of the con-

struction that

has been classed in the Gr. as a subdivision of

the Ablative of Means.

3.
of,

With

contineri, consistere. constare, consist of, be

composed

the Ablative was probably originally one of Means.

Such

is

the view of Ebrard, de Ablativi, Locativi, I/istru men talis usu, p.

645.

Ktihner and

Roby

also give this explanation for the Ablative

with constare and consistere ; the use with contineri they explain as
Locative.
'

But

all

three words originally

it

had the same meaning,

seems unnecessary
to

hold together, be held together,' and

adopt different explanations for the separate verbs.

regard the Ablative with
all

Some

scholars

three verbs as a true Ablative usage.

This view

is

based upon the occurrence of ex with the Ablative

with constare.

But prepositions are a very uncertain guide in

Often more than one case relation
;

such matters.
the same verb

is

possible with

and often

a verb in

its

developed meaning takes

it

a different construction from that which

Delbriick, Vergleichende Syntax, L, p. 230.
4.

originally had.

See

Quid hoc nomine facias;

quid

me

fiet?

Delbriick
in

in

his

Ablativus, Localis,

Instru men talis, p.

17

(published

1867),

explained the case in expressions of this type as a true Ablative.

Ebrard's collections for early Latin, however, showed that the construction was
(

rather Instrumental in origin, and Delbriick

I.,

now

Vergleichende Syntax,
5.

p.

248) adopts this view.

Ablative of the

Way

by which.

This

construction seems

to

be one of considerable antiquity, and deserves recognition as

THE ABLATIVE.
an independent type of the Instrumental.
Latin, but in
It

203
appears not only in
Illustra-

several other
:

Indo-European languages.

tions for the Latin are

utjugis Octogesa?n pervefiiret, 'that he might

;

reach Octogesa by way of the mountains'

porfis erumpunt; fru-

mentum quod flumine Arari subvexerat.

reisen,

Cf.

German mit der Bah?i

com-

where the

traveller

is

evidently conceived as keeping

pany with the road.

342.

Ablative of Cause.
its

Cause

is

sometimes referred to the

this

true Ablative for

ard'ere

origin.
'

In accordance with

theory ira

meant

originally

to

burn from anger.'

way.

The

Sanskrit often

employs the Ablative

mental origin
sions as
is

in this

On

the other hand an InstruCf.

equally conceivable.

such English expres-

burn with anger, howl with pain, leap with joy, green
Instrumental as well as

with envy; the Sanskrit employs the

the Ablative to denote this relation.

Other Indo-European

lanit

guages also use the Instrumental to denote Cause.

While

is

impossible to prove that Cause has developed exclusively from

the Instrumental conception, yet
least
is
it

is

likely that this case has at

;

had the greater share

in

propagating the construction

such
126).

now

the opinion of Delbriick (Ve7gleichende Syntax, L,

ii.

Cf. also Kiihner, Ausfuhrliche Gra?n?natik,

p. 291.

343.

Ablative of Degree of Difference.

;

This

seems an out-

growth of the Ablative of Means

i.e.

uno dil

lo?igiore?n ??ie?isem

faciunt meant primarily 'they make the month longer by means

of one day,' and so on.

344.

Ablative of Price.
the

Price
At the

was

in

its

origin a develop-

ment of

Means

notion.

outset, the construction

e.g.

must

have been confined to verbs of buying,

emit, 'he

puellam vlginti minis

bought the

girl

by means of twenty minae.'

With verbs
selling
;

of selling the price was not strictly the

means of

but

204
after the analogy of verbs

SYNTAX.
of buying, such verbs early

came

to

take the Ablative construction.

construction
is

still

further extension of the

seen in

its

application to verbs of costing, being

;

worth,

etc.,

and
e.g.

also to the adjectives vilis, 'cheap'

carus, 'dear,'

'too dear,'

HS

sex milibus constat,

'it

costs

6000 sesterces';

asse carum, 'dear at a farthing.'

The use
and
(Gr.

of tanti, quanfi, pluris, minoris with verbs of buying

selling is the result of a transference of the Genitive of

Value

203.3) from verbs of valuing, estimating,

selling.

etc.,

to verbs of
Cf-

buying and

Such a transition

is

psychologically easy.

our English I wouldn't give a penny for that (a phrase of buying)

in the sense of

/ don't value

that at a penny.

345.

The Ablative

of Quality

is

an obvious outgrowth of the

sociative force of the Instrumental case.

Thus

in a sentence like
'

serpens
glides

immani

corpore labitur, the original idea was

the serpent
distinct

on with

its

huge body,' as though the body were a

serpent.

accompaniment of the
tive in
this

But

in course of time the Abla-

such cases came to be

felt as

a modifier of the noun.

In

way such expressions as acerba tuens immani corpore serpens became possible. Here the phrase immani corpoi'e can be conceived only as an Ablative of Quality, limiting serpens
; it

cannot

be associated with the verb

In conformity with
denotes more or

as in the first example.

its origin,

the Ablative of Quality primarily

Qualities

less transitory qualities.

which are the

naturally

mere outward
permanent.
denotes

accompaniment of an action are

observation

not

The
is

sometimes made

that the Genitive

ititej-nal qualities,

whereas the Ablative primarily denotes

In the phrase hortatur

external ones,

not sufficiently exact.

ut bono animo
quality
is

sint,
;

'he urges them to be of good courage,' the

yet the Genitive could not here be used
internal,
'

internal

for

while the quality

'a

is

it is

transitory.

On

the other hand,

man

of high purpose

is

in Latin vir

magm

animi, since a per-

THE ABLATIVE.
manent and not a passing
of usage the Ablative
is

205

quality

is

intended.

By an

extension

sometimes employed, where ambiguity

would not
Genitive
is

result,

to indicate

permanent

characteristics

but the

not used to denote temporary qualities.

Physical and
Ablative.

bodily characteristics are regularly designated by the

For an excellent discussion

of

the

Ablative

of Quality,

see

Edwards, Geo. V., The Ablative of Quality and the Genitive of

Quality.

New

York, 1900.

346.

Ablative of Specification.

This
'

seems

to

be a develop-

ment of

the sociative force of the Instrumental.

Thus Helve Hi

virtute praecedunt

meant

originally

the Helvetii with their valor

are superior'

so pede claudus, 'lame with his foot.'

also

The Means

conception

may

have assisted in the propagation of the

construction.

347. Ablative Absolute.

is

The Ablative Absolute

trie

construction

an outgrowth of the sociative force of the Instrumental.

Thus
helping

in Plaut. Trin. Prol. 13,

rem paternam

adjutrice perdidit, the

sense is: 'he lost his property (in connection) with

me
te

him'

so frequently

me

judice, 'with

me

as judge

'

praesente

'with you present.'

clothes torn,

Cf

further scissa veste, passis eapillis, 'with

and

hair dishevelled.'

At

first

the Ablative in such

phrases modified

the verb of the

sentence, but ultimately the

of,

original construction was lost sight

and the phrase

as a

whole

came

to be felt as a kind of loose modifier of the rest of the

sentence (Ablative Absolute).

See Brugmann, Die lateinischen

ff.

to-ParHcipia, Indogermanische Forschungen, Vol. V., p. 142

Others

have regarded

the Ablative Absolute

as

a Locative

development.
Locative
is

This theory was suggested by the fact that the

the case absolute in Sanskrit.

little

That
it

fact,

however,

would be of

significance for Latin unless

can be shown that

the Locative was the case absolute in the

Indo-European parent-

speech.

But there

is

nothing to show that such was the case.

206
In fact each
absolute.
tive

SYNTAX.
language seems to have developed
its

own

case

In Sanskrit we have the Locative, in Greek the Geni;

and Accusative

in

Gothic there are traces of the Dative

the
Accusative.

modern German employs

therefore, there
lar case.
is

As regards Latin,
any particuevi-

no anterior probability
is

in favor of

The

question

simply one of evidence, and the

dence points to an Instrumental rather than to a Locative

origin.

Those who advocate a Locative

nings of the construction
tive, e.g.

origin

would find the beginLoca'

in
l

the

temporal force of the

;

Servio regnante,
'

in the time of Servius reigning

bello

confeclo,
this

at the

time of the war having been finished,'

etc.

But

explanation seems

Another theory,
wald,

much less natural than the former. that of Bombe (De Ablativo Absolute,
explains bello confecto,
etc.,

Greifs-

1877), refers the Ablative Absolute to the true Ablative

for its origin.

Bombe

as

'

after the

war having been

tive to

finished.'

But no such use of the true Ablais

denote time after which

true,

known

for Latin.

Moreover,

if

Bombe's theory were

predominance

we should expect

a predominance of

time-words in the early history of the construction; but no such

is

found to

exist.

Locative Uses of the Ablative.

348.

The Locative seems

at,

to

have designated originally the

is

space in or within which something

the

done.

From

this

meaning
Verglei-

notions

on subsequently developed (Delbriick,

p.

chende Syntax,
rally fall into

I.,

183)
:

The

Locative uses of the Ablative natu-

two classes

Place Relations and

Time

Relations.

349.
1
.

Place Relations.
its literal'

These may be
umeiro.

either literal ox figurative.

:

In
'

force the Locative

may mean

a)
b)
c)

in,'

as premit altu?n corde dolorem.

'

on,' as pharetrarti fert

'by,'

'near,'

as

litore

curvo exstruimus

toros.

This

last

appears to be rare.

THE ABLATIVE.
The
preposition, however,
is

207

usually necessary to express these

relations,

except in poetry and late prose, and in the classes of

words specified in Gr.

228.

1.

Some
recipere,

recognize a Locative use in tenere se castris, aliquem fecfo

pugna

vincere; but

all

of these easily admit interpretain the phrase

tion as Instrumental usages,

it

and

conquer in

battle

is

significant

that the

Sanskrit regularly employs the Instru-

mental case.
2.

In figurative uses the Locative function of the Ablative

is

restricted to very

narrow

limits.
lit.

Here belong, however, a few

'

phrases such as aniniis pendent,

they are in suspense in their

;

minds'

{cf.

the Singular aninii in aninii pe?idere)

'

stare pro missis,

'to stand

by one's promises;

stare conventis ;

manere promissis.
P- 39> Delbriick

In

his Ablativus, Instncmentalis, Localis

(1867),

formerly
in

pronounced

in favor

of

recognizing a Locative usage

connection with glorior, detector.

I.,

But now

in his Vergleichende

Syntax,

p.

253, this scholar regards the construction as Instru-

mental in origin.

The same

explanation

is

also to

be preferred
Instru-

for laetor, gaudeo, etc.

Similarly with

fido and confido an

as

mental origin

is

the

more probable, inasmuch

we

find this case

used in Slavic with verbs of trusting.

3.

Refert and Interest.

The Ablative
'

Singular Feminine of

the Possessive with refert originally limited the re (Ablative of res,

1

thing

')

of refert.

If the construction

was Locative

in origin,
'

mea

refert

may have
;

originally
i.e.

meant

it

bears towards

my affair
is

(Goal

Locative

351),

'it

concerns me.'

The use

of the Ablative

Singular Feminine of the Possessive with interest

origin,

of secondary
conseas

being modelled on the construction with

similarity of

refert, in

quence of

meaning.

Some regard mea

re has also

r~efert

equivalent to ex

mea

re fe?'t ;
;

mea

been explained as

a stereotyped Dative ( 86. b

i.e.

174), and even as a Nominative,

for

mea

re (s) fert, with retention of the original long

in

a of the

Nominative

mea;

112. 1.

208
350.

SYNTAX.
Time
Relations.

The transference of the Locative from

is

space relations to relations of time

easy and natural.

In

this

way arose the notions of time at which and within which. some
B. G.
frequency in the best prose of
all

The

use of the Ablative to denote duration of tune, which occurs with

little
i.

periods, e.g. Caesar,

is

26. 5,

eaque tota node continenter lerunt,

probably not
rather to be

a development of the time within which, but

referred to an Instrumental origin.
to

is

This use of the Instrumental

denote duration of time would correspond to the use of the

Instrumental to denote the

way

by which ( 341.

5).

351.

Locative of the Goal.

Sanskrit and
This
is

Greek both exhibit

a goal use of the Locative.

the result of extending to

verbs of motion a conception primarily belonging only to verbs of

rest

Cf. in English he

went among

the Indians, after he

is

among

the Indians.

Examples

in Latin are confined chiefly to the archaic

;

period.

Thus, ford ponit (Ennius)

loco

collocare

(Lucilius)

certa parte reponunt (Lucretius).

Genuine Locative formations,

e.g.

humi,

dorni, etc., also

occur in this sense,

domi adveniens.

Surviving Locative Forms.

352.

The

chief genuine Locative formations in

common

use

are enumerated in Gr. 232.

Beside these we should probably

recognize the Locative of an

z/t-stem in noctu,

and (by association

with noctu) in diu.

On
-Is

die, as

the Locative of dies in such ex1.

pressions as quarfi die, postridie (for posteri die), see 256.

Plural formations
in

from a- and <?-stems are more safely

regarded as Instrumentals

which have taken on

all

the functions

of the Ablative, Locative included.

Plurals in -ibus of the Third

Declension are certainly Ablative in form.

the Third Declension,
e.g.

Formations in

-e

of

Sulmdne, are original Locatives;

141.

THE SUBJUNCTIVE.

209

THE MOODS.
The Greek name
i.e.
'

Latin Names of the Moods.

353.
1.

for

mood was
as

ey/cAto-is, literally

'

in-

clination' or 'turn/

turn of thought.'

The Romans

transferred
is

this designation to their

universal designation for

own language mood among

modus, which

the

the Latin grammarians.

still

Yet traces of the influence of the Greek designation are

seen in the definitions given by the grammarians.

to

be

Thus

Priscian,

probably following the tradition, defines modi as diversae inclinationes animi, varios eius affectiojies dei?ionstrantes (Keil,

matici Latini,V o\.

Vol.
I.,

11., p.

421. 17).
:

Diomedes

(Keil,

GramGram. Lat

p.

338) gives the heading

De

modis sive inclinationibus

verborum, indicating that hiclinatio was sometimes used as an

alternative designation.
2.

The Greeks recognized

TrpoaTaKTLKrj

five ey/eAto-eis, viz. o/awrn/oj (IndicaevKTLKrj

tive),

(Imperative),

(Optative),

VTroraKTLKrj

(Subjunctive), d7rape/x<^aTos (Infinitive).

3.

'OpLa-TLKrj

was variously rendered by the Latin grammarians

or indicativus.
lyKAto-is

as

modus finitus, pronuntiativus,

precise,

Neither of these
6/310-1-1*77

designations was
'

however, as

meant

mood

of definite statement' (from 6oiw, 'bound,'

'limit,' 'define,'

'state definitely').

Hence

definiUvus would have been a better

name.
v

4.

Ey/oWi?

7rpo(TTaKTLKrj

meant mood of command, and was

as

literally translated
5.

by the Romans

modus imperativus.
of the Greek Optative; but

"EyKAicns evKTiKrj was the

name

the designation was good for only a small portion of the uses

of the Greek Optative,

viz. its

employment

in wishes.

It

did not

apply with accuracy to the Potential uses of the mood.

1

The

On

the

names of the Moods,

see especially Jeep,

Zur

Geschichte der Lehre

j

der Redeteile bei den lateinischen Grammatikern, Leipzig, 1893

PP 216-236.

210
Romans, having no
times used
(Keil,

SYNTAX.
special verbal forms recognized as Optative,
optativus.
as

had no need of the designation modus

it,

Yet they somePriscian remarks

ad imitationem Graecorum,
Vol.
II., p.

Gram. Lat.

407).

But

it

should be noted that

the

Romans never used

the

name
it

optativus to designate a group

of inflected forms.

.With them
viz.

designated merely a syntactical

use of the Subjunctive,

the Subjunctive in wishes. the Greek

evKTiKrj,

They

thus

made
6.

the

name narrower than

whose

syntacti-

cal province
J/

extended beyond what

its title

designated.

EyKXicrts v-rroTaKriKrj

meant 'mood of subordination' and was

the Greek designation for what

we

ordinarily call the Subjunctive.

it

But the name was a poor one, since

applied only to the uses

of the Subjunctive in subordinate clauses, and implied that these represented the original function of the mood.
It

ignored the

independent Volitive uses (Hortatory, Jussive, Deliberative, Prohibitive), also the so-called Anticipatory uses.

The Romans
names
quite
as

translated woTa/m/o^ usually by subjunctivus, less

(cf.

frequently by conjunctivus

Jeep, Redeteile, p. 224, footnote

course,
as

3),

misleading, of

the

Greek

original

from which they were taken.

7.

'Axape/x^aTos
infijiitivus

was rendered

by the

Roman

grammarians

modus

or infinitus.

The
354.
is

Subjunctive.

1.

Origin of Subjunctive Forms.

The Latin Subjunctive

of the Indo-European

the result of a fusion of two original

moods

parent-speech, the Subjunctive and the

Sanskrit kept
early

Optative.

Greek and
in Latin they

them

distinct
in

from each other, but

a
single

became merged

characteristic

meaning of each.

mood endowed The following table

with

the

indicates

the origin of the different formations appearing in the so-called

Subjunctive

THE SUBJUNCTIVE.
Subjunctive Forms.
i.

211

Optative Forms.

All regular Presents,

e.g.

amem,

I.

Presents in -im,
ndlitti,

e.g.

sim, possim,

2.

moneam, regam, audiam; 221 f. All Imperfects, eg. essem, amdrem,

monerem,
etc. ;

mdlim, velim, edim, duim;

218.
2.

222. 3.
e.g.

All Perfects,

e.g.

viderim, amdve-

3.

All Pluperfects,

amdvissem,

dl-

rim,

etc.;

219.

xissem,
2.

etc. ;

222. 4.

Original Force of the

Subjunctive.

The

Indo-European

Subjunctive exhibits two meanings which seem to have been the

source of
a)
all

others

= 'I
tain

The Subjunctive expresses the will of the speaker, e.g. surgat This use implies a cerwill him to rise/ i.e. Met him rise.'
power or authority on the part of the speaker,
i.e.

he

is

repre-

sented as willing something over which he has control or volition

hence the name 'Volitive' has been given to characterize

use of the mood.
F)

this

Alongside of

this Volitive notion, the

Indo-European Subthat of

junctive also possessed a second force,

(precisely like a Future Indicative)
.

pure futurity

The Greek,

particularly of

the

Homeric
;

dialect, frequently exhibits this

Future force of the

it

Subjunctive
in Latin.

but

it

is

uncertain whether

we should recognize

In Latin the Subjunctive has a Pure Future force only

in subordinate clauses,

and

this

may be

traced to a different origin.

Yet

it

should be borne in mind that the so-called Future ero was

;

in reality a Present Subjunctive ( 205. 3)

also

audiam,

rega??i, etc. ;

while the so-called Future Perfect

All of these

is

an Aorist Subjunctive

( 216).

formations bear witness to a Pure Future force as

having once existed in the Latin Subjunctive.

The connection
than might at
first

of meaning between the Future force and the

Volitive force of the

Indo-European Subjunctive

is

much

closer

appear.

Thus the English

he's to go clearly

stands on the border line between the two meanings, and

may

be

interpreted either as Volitive,

= let

him

go, or as Future,

= he

will go.

212
It is

SYNTAX.
probably impossible to explain satisfactorily the relation-

ship to each other of these two uses of the Indo-European Subjunctive.

Some have regarded

the Volitive notion as the original

one and the Future notion

started with the Pure

as derived

from

that.

Others have

Future notion as fundamental and have

Others have regarded the
as representing

deduced the

Volitive uses from this. 2

two functions as equally primitive and

merely two

phases (the Subjective and Objective) of the same thought. 3

No
not

attempt to solve
acceptance, nor

this
is

problem, however, has

it

commanded
its

extensive
is

likely to.

Fortunately

solution

necessary to our purpose.

The two meanings of

safely accepted,

the Indo-Euro-

pean Subjunctive may be

even though we are

unable to determine their mutual relations.

For the views of those who deny

that the

Indo-European Sub'

junctive possessed any definite fundamental force (or

griff'), see below, 356.

Grundbe-

355.

Original Force of the Optative.

Here

we note two

dif-

ferent, but closely related meanings, as in the case of the

Sub-

junctive.
a)

Thus

The Optative
e.g.

is
'

used to express an act as wished for by the

speaker,
authority,

veniat,

may he come
is

'

The element

of power,

and

volition

which characterizes the corresponding use

lacking here.
in

of the Indo-European Subjunctive

b)

Alongside of the notion of wishing, we find both

in Latin

Greek

and

another notion,

viz.
e.g.
;

that of a contingent futurity

aliquis dicat,
'

(Delbriicks Bedingte Zukunft),

say'
;

some one may

'

crediderim,
is

'

should believe'

quis ftutet,

'who would think?

This
1

obviously a weaker type of Future than that belonging to

the view of Delbriick in his Conjunctly
1 1
ff.

This

is

und
p.

Optativ

im Sanskrit

und
2
3

Griechischen, p.

Notably Goodwin in Greek Moods and Tenses,

371

ff.

The view advocated

in the earlier edition of this book.

THE SUBJUNCTIVE.
the Subjunctive (in Greek), just as in
its

213

meaning of wishing the

Optative expresses a weaker phase of thought than the Subjunctive.

The problem
ings of the

of the mutual relationship of the different meanis

Indo-European Optative

even more

difficult

than for

the meanings of the Subjunctive.

Delbriick in his Konjunktiv

und

Optativ started with the wish meaning as fundamental, and derived

the Potential uses from that.
p.

Subsequently (Altindische Syntax,

302) he has expressed the conviction that the wish meanings and

Potential meanings are distinct in their origin. 1

Goodwin (Greek

Moods and
original.

Tenses, p.

384

ff.)

starts

with the Potential force as

But scholars are

far

from agreed as to accepting any of

It is safer, at

these theories of relationships.

present at

least, to

content ourselves with recognizing the existence of the various Optative functions, even though

we cannot determine

their origin

and mutual

relationships.

For the views of those who deny that the Indo-European

Optative possessed any precise fundamental
'Grundbegriff'), see the following section.
force

whatever

(a

356.

Some eminent

syntactical investigators have contested the

propriety of attributing to the Indo-European

Subjunctive
'

and

Optative any precise narrow fundamental value (a

Grundbegriff').
.
. .

Thus Abel Bergaigne (De Conjunctivi

quissima.
junctive
Paris, 1877,

et Optativi

vi anti-

pp.

41-50; 57-73) urged

that the Sub-

and Optative

alike originally

covered the entire range of

modal conception outside

embraced by the
Indicative,

that of positive categorical assertion

and

that the specific Subjunctive

and

Optative uses found in the various Indo-European languages are

the result of selection in this wide
attitude
field.

Closely related to this

of Bergaigne

is

that of Morris

{American Journal of

But in his Vergleichende Syntax, IV.

2. p.

373, he apparently returned to

his earlier view.

214
Philology, Vol.

SYNTAX.
XVIII.
p.

392

ff
.

also

On
iv.).

Principles

and Methods

in Syntax, especially chapters

iii.
'

and

Morris recognizes in

Subjunctive and Optative no

actual functions of these

Grundbegriff,' but urges that the

as

moods have developed

etc.

a result of

context, gesture, intonation,

Yet to most investigators the phenomena of

linguistic

growth

seem

to point to the early existence of a fairly definite value for

every inflected form.

The

existence, also, in

Old Indian, Iranian,

Greek, Latin, Gothic, and Slavic of a number of substantially the

same

specific Subjunctive

and Optative modal uses seems impos*

sible to

account for except upon the basis that the value

in

of these

moods
cf.

Indo-European was a

fairly precise

and

definite

one

Delbriick,

Die Grundiagen der Griechischen Syntax,

so-called

p. 116.

357.

The

Latin

Subjunctive,

as

an amalgamation

of the original Indo-European Subjunctive and Optative, might

naturally be expected to exhibit
tions, viz.:
Volitive
>

all

four of the original significa-

Indo-European Subjunctive.

Pure Future "

\

Indo-European Optative.

Contingent Future

As
them,

a matter of fact
viz.

it

represents with certainty only three of

;

the Volitive, Optative, and Contingent Future

all

and

from these three primary uses are to be derived

existing Sub-

junctive constructions in Latin, not only in principal, but also in

subordinate, clauses.

The absence of the Pure Future use of the Subjunctive in Latin may be accounted for by the fact that the Subjunctive in that use early came to be felt as Indicative, and as a result various
Subjunctive formations actually

became

Indicatives, ero,

audiam,

It is

not necessary that this assumed value was absolutely primitive in

It

Indo-European speech.

may have been

the result of development.

THE SUBJUNCTIVE.
viderd,
etc.

21

( 205.

2,

216).

This transition to the Indicative

of those Subjunctive forms which possessed the Pure Future force

naturally resulted in the restriction of the remaining forms to the

Volitive use.

CLASSIFICATION OF SUBJUNCTIVE USES.

Subjunctive in Principal Clauses.

A. Original Uses.
358.
Volitive Subjunctive.

a) Jussive,

expressing a

command.

This use

is

found most

commonly
1)

In

the

Third Singular and Third Plural of the Present

toquatur, Met

tense,

e.g.

him speak'; /oquantur,

\et

them

speak.'
2)

In the Second Singular and Plural Present.

The Second
Met

Singular often has indefinite force, but not necessarily so.

An example is utare itiribus, 'use your a man use his strength (indefinite).
'

strength,'

i.e.

The

Perfect tense

is

sometimes employed in the Jussive.

It

calls attention rather to the

summary performance

of the act, while

is

the Present represents the act as in progress.

This

in

accord

with the origin of the two tenses, for the Perfect was by origin an
Aorist (219).
Jussives
Cf.

under

d,

and

360,

a.
e.g.

accompanied by

ut, ufi

occur in early Latin,

Plaut.

Capi. 115, ufi adserventur, 'just

739, ut caveas
1. 4, is

let

them be watched

!'

Bacch.

Ter. Ad. 280, ut

siet.

omne reddat;
in these

Cato, de Agr.
like expressions

ut bene aedificatum

Ut

and

an adverb,

probably

originally indefinite, corresponding to

the indefinite qui, 'somehow,' 'only,' 'just.'

of the adverb qui are well substantiated,
1.

The
:

three meanings

viz.

Relative, 'in which way,' 'as.'

'

2l6
2.

SYNTAX.
Interrogative,
Indefinite,
'

how
'

?'
cf.

3.

'somehow';
only.'

modo, originally

'

in

away,' 'in

some

way,'

'

somehow,'

In case of the corresponding adverb ut we have

1.

Relative

ut,

'in which way,' 'as.'

'

2.

Interrogative ut,
If

how?

'

3.

we recognize

the Indefinite ut,

we

get for ut the third of

the three meanings which are assumed for qui.

The

value here

suggested for ut seems to occur also in uti-nam, and to be sup-

ported by the use of qui and ut interchangeably with independent

Optatives
b)
;

see 359.
resolution.
e.g.

Of determined
hoc quod coepi

This rare usage

'

is

confined to

the Present First Singular,

Terence,
"enarrem,

H autontimorumenos
'

273
first

mane :
c)

primum
began.
is

wait

I'm bound

to finish telling

what

Hortatory.
is

This

confined to the Present First Plural,

e.g.

and

a mingling of a) and b),

to speak,

loquamur,

'let

us speak,'

i.e.

Tin bound
d)
lar

and do you speak.'

Prohibitive.

This occurs in the 2d and 3d Persons Singu-

and Plural of the Present and Perfect Tenses.

The

earlier

theory as to the Prohibitive was that the Second Singular Perfect

was employed of a
lar

definite

Second Person, while the Second SinguThis view has been

Present had a general (or indefinite) force.

to be false

shown

by the exhaustive examination of the subject by

3.

Elmer, American Journal of Philology, 1894, No.

In the

Grammar
citement.

and the Appendix

had given

my

adhesion to Elmer's

view that the Perfect Prohibitive expressed special emotion or ex-

Renewed examination

of the question, however, has

compelled

me

to

abandon

that attitude

and to accept the conbetween the

clusions of Delbnick,

who holds

that the difference

Present and Perfect tenses was one of the kind of action desig-

nated by the verb, the Present indicating an act (or state) going
on, the Perfect an act (or state) conceived of without reference

THE SUBJUNCTIVE.
to continuance.
for the Perfect
e)

21

This accords with the origin of the two tenses,

was an Aorist

219).

Deliberative.

This occurs in affirmative questions inquiring

after

the will or

command
:

of the person addressed,

e.g.
is

quid
your

faciam, in the sense

will that

what do you bid

Plaut.

me

do

?
'

'

what

do?'

Cf.
:
:

Trin.

59, sequere :: quo

?

sequar?
: :

Aul. 651, redde hue

quid reddam

Capt. 839, gaude

quid

gaudea?n

The usage, accordingly, consists simply in the inquiry An English analogy may perhaps be recognized after a command. in 'what let's do?' i.e. 'what do you say (direct) that we do?'
?

The name

'

Deliberative

'

is

by no means an accurate designa-

tion of the usage

here under consideration.

There

is

nothing

deliberative in an inquiry after orders.

We shall
is

come

later,

under

the head of

'

Derived Uses,' to a usage which

also to a
'

truly deliberative.

We

shall

bear the

come name

number

of other uses which traditionally

Deliberative,' though

no deliberative character

inheres in them.

See

363.

359.

Optative

Subjunctive.

The
e.g.

Optative

Subjunctive

ex-

presses a desire or

hope

for the fulfillment of a wish.

Both the

Present and Perfect tenses occur,

tibi ;

Plaut. Pseud. 714, dene sit

Verg. Aen.
is

i.

603, di tibi praemia digna fera7it.

The

Per-

fect

less

frequent than the Present.

When
to
its

used,

it

ordinarily

differs

but slightly in value from the Present, denoting the sumact, as

xii.

mary performance of an

opposed

continuance (see
averterint.

358, a, d),

e.g.

Cic. Phil.

14,

quod di omen

But

occasionally the Perfect Optative has true Present Perfect force,

e.g.

Cic. de Rep.

iv.
i.e.

8,
'

cui quidem
I

v'ere

auguraverim,

lit.

'

may

have prophesied,'

hope

I
is

have prophesied.'
not infrequently accompanied by

The Optative

Subjunctive
e.g.

strengthening particles,

Plaut. Trin. 923, qui istum

di per-

da?it; Aul. 785, ut ilium di perdant.

Cf. the use of utw'ith. the

Jussive

(358,

a, 2).

Utinam

is

also frequent.

2l8
360.
to the
(

SYNTAX.
Subjunctive of Contingent Futurity.

This corresponds

second of the two meanings of the Indo-European Optative

355. b).

From
:

this

general notion have developed the follow-

ing special uses

a)

Subjunctive

of Pure
say.'

Possibility, e.g. aliquis dicat, aliquis

dixerit,

'some one may

This

is

the most obvious develop-

ment of the notion of contingent

As regards the use of
lays stress

futurity,

but

it is

rare,

being con-

fined chiefly to phrases of the type cited in the

above examples.
;

tenses, the Perfect (originally Aorist

act,

219)

upon the accomplishment of the

while the Present

calls attention to its progress.

Cf. 358, a, d.
is

b)

Where some condition

i.e.

implied or expressed,

e.g.

velim,

'I should wish,'

'if

were to have

my way';

dicas, 'you

would

say,' i.e.

'

if

you should have occasion to express an opinion.'

This use occurs also particularly in the First Singular of the Perfect (Aorist, 219), e.g. dixerim, 'I should say';

crediderim, 'I

should

believe.'

Where

the

condition

is
(

expressed,

we

get

Conditional Sentence of the Second

ris, si

Type

Gr.

303),

e.g. laefe-

veniat,

'you would rejoice,

Potential
b)
;

if

he should come.'

The name
under a) and

is

usually given to the Subjunctives cited

but this

name

is

somewhat inexact; see

365.

B.

Derived Uses.

361.

The

uses here enumerated are secondary developments

in

from those cited above

358

ff.

362.

Extensions of the Jussive and Prohibitive.

a) Corresponding to the Jussive loquatur there developed an

Imperfect use,
'

e.g.

loqueretur, in the sense,

'

he was

to speak,'

i.e.

he should have spoken.'

This use

is

manifestly a derived one,

since one cannot

now

will a

person to have done in the past what

do.

he obviously has

failed to

An

expression

like

loqueretur,

THE SUBJUNCTIVE.
therefore,

219
analogy of loquatur.
e.g.

must have been formed

after the

The

Pluperfect Subjunctive also occurs in this sense,

'

eui7i

imifdtus esses,

you ought to have imitated him.'

is

The

Volitive

character of these expressions

tive is regularly
rie, e.g.

shown by the

fact that the

tale facer'es,

nega-

Plaut. Pseud, 437, tu

;

rie

'you
rie.

ought not to have done any such thing'

poposcisses,
F)
'

Cic.

ad

Att.

ii.

1, 3,

you ought not

to have asked.'

An example of this is Cic. de Sen. 58, originally this sibi kabeant anna, they may have their weapons
The
'

Permissive.
'

'

meant
so,
'

let

them have
have.'

! '

i.e.

'

let

them have,

for aught I care,'

and

they
to

may
i.

In

this

way a recognized permissive value

Other examples are

came

attach itself to the Subjunctive.

1,

Tibullus,

58,

tecum dummodo sim, segnis vocer ; Accius, Fr.

'

dderint, duin metuant,

they

Permissive Subjunctives,

may hate, provided they fear.' These when negative, imply that one does not
e.g.

need
ea7n

to

perform the act involved,

Plaut.

Capi. 947, at ob

rem mihi Ubellam argenfi

etc.

rie dicis,

'you don't need to give

me,'
c)

The

Concessive.

This

is

found in the Present, Perfect, and

Pluperfect tenses.

The

Perfect in this use refers to the past.

sit

Examples:
Ennius
is

Cic. Brut. 76,

Ennius

perfectior,
ii.

'I grant that

more

finished';

Academica,

75, at dissolvit Idem.

'

Mihi quidem ndn videtur ; sed

refuted
;
'

dissolvent,
'

but I grant that he

it
'

Verg. Aen.

iv.

603, fuisset,

grant that

had been
grant that

done
there

'

Cic. de Sen. 34, ne sint vv'es in senectute,

is

not strength in old age

;
'

Or. 101, nemo

is,

inquies,

um-

quamfuit.

Nefuerit,

'I grant that there wasn't.'

d) Subjunctive of Acquiescence.
the familiar/^/ of comedy, 'so be
e)

Here belong expressions

it,'

like

'very
is

well.'
is

Subjunctive of Supposition.

This

infrequent, but
iii.

exem-

plified in

such expressions as Cic. de

et

Off.

54, v'endat aedes vir

.

bonus ; pestilenfes sint

etc.,
'

habeantur salubres ;

quae?-o

num,
let

let

us suppose the case of a

good

man

selling a

house

us

'

' '

220
suppose the house
ask whether,'
etc.
is

SYNTAX.
unwholesome, but
is

considered

safe,

...

363.

Extensions

of the Deliberative.

These
e.

are

all

outgrowths
:

of the original use mentioned in

358.

We

distinguish

a) Questions of purely rhetorical character, implying that the

thing mentioned
fect tenses

is

impossible.

The
'

Present, Imperfect,
!
'

and PerI to

occur
:

in this use, e.g.

quid faciam

what

am

do ?

in the sense

'

there's nothing I

can do.'

The Imperfect

repre-

sents this present use projected into the past, e.g. quid facerem,

'what was

I to
is

do?' implying the impossibility of doing anything.

rare, but
is

The

Perfect
'

found in Plaut. Amph. 748, ubi ego

have heard
it ?
'

audwerim,
heard
it?

where

am

I to

i.e.

'

how can

have

b) Questions implying the idea of duty, obligation, or propriety.

The
'

tenses used are the Present

is

and Imperfect.
'

characteristic
I to

example

quid faciam in the sense of what ought

I

do ?

what should

do ?

'

This

is

a perfectly natural and legitimate

in

outgrowth of the original idea contained

1

quid faciam
question
is

( 358. e),

what do you bid

me do ?

'

Whenever
is

this
'

addressed

to a person

whose authority
to

respected,

what do you bid

I

me

do?' becomes tantamount

'what ought
'

to

do?'
it

So

in the

Imperfect, quid facerem often means

what was

my

duty to do ?

This

is

simply quid faciam projected into the past.

is

The

negative
'isn't

of this usage
it

non,

e.g.

nan haec faciam, non haec facerem,

?
'

(was n't

it)

my

duty to do these things

So also in Cic. pro

I

Arch. 18, hunc ego non di/igam, non admirer? 'ought

love,

not to

ought

not to admire this

man?'
?

So also in expressions
e.g.
'

introduced by cur, quare, and rarely qum,

Caes. B. G.

i.

40,

cur despe7-arent,

why should

they despair

The

negative non (instead of ne)

is

to be regarded as a perfectly natural

consequence of the derived nature of the usage.

THE SUBJUNCTIVE.
c)

221
'

Real Deliberative questions.

'

Although the name

all

Delibera-

tive

is

used as a designation of

the related idioms here conis

sidered, yet the only real

Deliberative Subjunctive

found in
is

expressions like quid agam, quid facia?n, where the speaker

actually pondering to pursue.

what decision

to take or

what course of action

d) Repudiating Questions, in which the speaker repudiates with

scorn some

command

or imputation, or expresses his disdain at

some proposal of another person.

Questions

The

origin of the Repudiating

may be
is

seen in passages like Plaut. Mil.

: :

Glo. 496,

zncine ausculta, quaeso

ego auscultem

tibi.

that the inquiry

listen

uttered with contempt.

The context shows Hence the idea is, I

'

to

you!'

Sometimes we have the Indefinite ut

tibi ego

358.

a. 2), e.g.

Ter. And. 618,

ut cr'edam ?

In these cases we see

that the usage originated in an inquiry after a

command, but

that

the indignant attitude of the speaker developed a repudiating force.

As a

result a

new category was formed, and we

e.g.
is

find Repudiating

Questions, where no vestige of an inquiry after a

cernible, or even imaginable,
rids

Plaut. Capt.

command is dis207, fingitis fugam

: :

fugiamus !

The idiom

even transferred to the past,

tibi

e.g.

Plaut.
ris

Men. 678, pallam quam

dedi mihi redde

!

mihi tu dedei.

pallam, 'you gave

me

a cloak
:

'

Cic.

ad

Quint,

3.

1,

ego te

vid~ere ridluerim.

Pluperfect

Cic.

pro Sulla, 45, mihi cujusquam

?

salus tanti fuisset, ut

meam neglegerem

364.

Extensions

of the Optative.

The
For
it

use of the Imperfect

and Pluperfect Subjunctive

utinam adfuisses,
is

in expressions like
if

utinam tu valeres,

also secondary.

the primary force of

for-

the Optative was to denote a wish,

must have looked

ward

to the future

hence

its

employment with reference

to the

present and the past must be a derived usage, after the analogy

of sintfelic'es,

etc.

The Imperfect and

Pluperfect Subjunctive, in expressions like

222

SYNTAX.

those cited above, do not strictly express a wish, but rather a

regret at the present non-existence or the previous non-occurrence

of something. In these derived uses utinam

is

almost invariably used.

very

few exceptions occur

in poetry.

365.

Extensions of the Subjunctive of Contingent Futurity.

There are three derived uses

a)
e.g.

The Present 2d
can

Singular in the sense 'you can, one can,'

videas, 'you

see.'

In

its

origin, the Subjunctive of the

possibility, e.g. dicas

Contingent Future denoted mere objective

= 'there's

a possibility, you will

say,'

'you

may

say.'

In

the

derived usage this objective possibility

becomes

subjective,

'you may' becomes 'you can.'

of these
trol,
is

Strictly speaking, only the

second

Potential.

For potentiality involves capacity and con-

which mere

possibility does not.

F)
limits,

The 2d

Singular Imperfect.

This

is

restricted to

narrow
'one

being found chiefly in such expressions as

vider'es,

could see'; cemeres, 'one could observe';

believe.'

cr~ede?-es,

'one could
like that, is

The usage

is

an extension of a) above, and,

Potential in the strict sense of that term.

c)

The Imperfect and

e.g.

Pluperfect in the conclusion of contrarysi adfuisses, vidisses.

is

to-fact conditions,

si adesses, vid'er'es ;

The

exact

way

in

which

this use has

grown up
But
if,

one of the most

difficult

problems of Latin syntax.

as

seems probable,

the protasis in such conditional sentences was in origin partly

Optative, partly Jussive
(e.g. adesses,
'),

'would that you were

if this

here,'

vid'er'es,

'

then you would see

may be assumed,

then

the use of the Imperfect

lation,

and Pluperfect would be

a kind of assimi-

induced by the regular correspondence of tense and

mood

in other conditional sentences.

THE SUBJUNCTIVE.
Subjunctive in Dependent Clauses.
366.

223

Parataxis and Hypotaxis.

In

the earlier stages of lan-

guage there were no subordinate clauses.

Sentences were joined

by co-ordination.
tive

For example, an independent use of the Indicamoneo, desi-

was followed by an independent use of the Subjunctive, or by

e.g.

another Indicative without any conjunction,

nant,
lit.
'

eos

warn them,

let

them

cease.'

In course of time in such

felt

combinations the one clause came to be

to be introduced

as subordinate,

and

by various connecting

particles
is

('subordinate

conjunctions').

The

stage of co-ordination

called Parataxis;

that of subordination, Hypotaxis.

In Latin the paratactic form

of expression often survives, even

when
is
is
;

the

hypotactic

relation

has

become

clearly developed.

This

especially noticeable in

the early and colloquial language, but

found also in the best

see, for

prose in certain categories of expression

example,

381

f.

For further discussion of Parataxis, see Bennett,

in

in Cornell Studies

Classical Philology, Vol. IX. p. 66

i?i

ff. ;

Morris, Principles

and

Methods

Syntax, p. 113

ff.

Subjunctive of Purpose.
367.
ne, quo,
1.

The
tibi

Subjunctive clause of Purpose

It

is

introduced by

ut,

qit'i,

and Relative Adverbs.

was probably Jussive in

'

origin, e.g.

do pecmiiam ut panem emas originally meant

;

give you

money

just purchase bread.'

For

this force

of

ut,

see

358. a. 2; 359.

In course of time the ^/-clause came to be

to the other,

felt

as subordinate

and ut from being an adverb came

In
this

to

be

felt as

a subtit.

ordinate conjunction.
2.

way arose

the purpose clause with

Negative clauses of purpose introduced by ne were quite

ut.

analogous in origin to those introduced by

intres probably
in
!'

Thus
;

tibi obsto

ne

meant

originally

stand in your way

don't

come

Ultimately this Parataxis developed into Hypotaxis.

'

224
3.

SYNTAX.
Quo
as

an Ablative of Degree of Difference

is

regularly con-

fined to use in connection with comparatives.

The

Subjunctive

with quo arises in the same way as with other relatives.

4.

See 4.

Qui, quae,

etc.,

in relative clauses of
e.g. tibi

purpose had practically

legas,
;
'

a demonstrative force,

librum do quern
'

give
it
!

you a

book
5.

to read,' originally

meant

I give

you a book

read

Relative Clauses with dignus, indignus,

classified in Gr. 282. 3

and idoneus have

been

under Relative Clauses of Purpose.

This has been done partly on account of the meaning of such

clauses, partly in

view of the other constructions found with

dig-

nus, idoneus,

etc.

As regards the meaning of the

it

relative clause

with digitus, indignus, idoneus,

seems impossible

to separate a

sentence like dat mihi surculos quos seram, 'he gives

to plant,'

me

shoots

from dat mihi surculos dignos quos seram,

to plant,' originally
elegit
fit

'

he gives

me

shoots

fit

'he gives

So homilies dignos
originally
:

quos

me fit mitteret may

(in order)
fit

shoots
well

to plant.'

have meant

'he selected
'

men,

to

send them,' and

In each case the
is

then, secondarily,

he selected
is fairly

men

to send.'

Subjunctive clause

one of Purpose.

This view

further

confirmed by the other constructions found with dignus, idoneus.

Thus we repeatedly
e.g.

find an Infinitive

employed with these words,

'worthy to be

Verg. Eel.
'

5. 53, et

puer
7.

ipse cantari dignus,

praised

Pliny, Paneg.

4,

dignus

eligi,

'

worthy to be chosen.'
i.

The Gerund with ad also occurs, e.g. ad imitandum ; and sometimes even an
ut haberes
///-clause

Cic. Rep.

18. 30, dignus

///-clause, e.g. eras

dignus

(cited

by Quintilian from an

early

author).
this

The
similar

cannot be regarded as one of Result in

is

and

cases, as

done by Kiihner, Ausf. Gr.

ii.

p.

858 d), since the

action

is

viewed purely as one contemplated, not as one accom-

plished.

Some

regard the relative clause with dignus,

It
is

etc.,

as a Clause of
fol-

Characteristic.

of course quite true that dignus, with a

lowing relative clause, does express a characteristic in a general

THE SUBJUNCTIVE.
way
;

225^ not a Clause of

but the relative clause

itself is certainly

Characteristic in the technical sense of that term.

See

371.

368.
tive

It is

obvious that only those purpose clauses are of primi-

origin in

which the main clause and the subordinate clause

refer to different persons.

Thus

in a

sentence of the type pecureferred directly

is

niam mutuor

tit

libros

emam, emam cannot be

to a Volitive origin, since the Volitive Subjunctive

not naturally

used to represent a person as exercising his authority and volition

over himself.

Sentences like the

last,

therefore, are

more probcited

ably of later origin and formed

in 367.

upon the analogy of those

Clauses of Characteristic.

369.

The Clause
its

of Characteristic

is

a relative clause devel(

oped from the Subjunctive of Contingent Futurity

probable that in
origin
it

360).

It is

was confined to a limited number

of words such as passim, velim, no fan ,

ma fan, audeam,
Thus a nemo

credam,
est qui

putem,
possit,

etc.,
lit.
'

following negative expressions.

there
is

is

no one who would be

is

able,' is so nearly equivit

alent to 'there
force.

no one who

able,' that

early took

on

this

Similarly in such expressions as

nemo

est qui velit, nolit,

malit, audeat, credat, putet.

In

all

these cases the notion of con-

tingency

is

so slight as easily to disappear, leaving

the

relative

clause essentially

one denoting a

fact.

370.

Clauses of Characteristic as Distinguished from Relative

Clauses of Purpose.

Difficulty

is

often experienced

in

distin-

guishing Clauses of Characteristic from Relative Clauses of Purpose.

This

difficulty results chiefly

from the fact that a Relative

Clause of Purpose

may denote

a characteristic of an antecedent
characteristic.

in the general sense of the

word

Thus

in Cicero,

Brutus,

56 scribebat 07-ationes

quas

alii

dicerent,

'he

wrote

'

226
speeches
dicerent
for
is

SYNTAX.
other

persons

to

deliver/

the
;

clause

quas
the

alii

Relative
in a

Clause

of Purpose

but
'

at

same
'

time
its

it

does

certain

sense

indicate a

characteristic

of

antecedent.

One

essential difference

between the Clause of

consists in the

Characteristic
fact that the

and the Relative Clause of Purpose

former denotes an action or state contemporary with

or anterior to that of the

main

clause, while the Relative Clause

is

of Purpose denotes an action which

the

future relatively to that of

this principle

main

clause.

In accordance with

expressions
i.

like nihil

habeo quod again, 'I have nothing to do' (Hor. Sat.

9.

19);

nil scio

quod gaudeam,

'I don't

know anything
I

to rejoice

about' (Plaut. Capt. 842) are Relative Clauses of Purpose.

these

Did
I

sentences
'

doing

and

'

mean respectively don't know anything

find sentences

'

have nothing that

am

that I

am

rejoicing about

(contemporary action), they would be Clauses of Characteristic.

At times we

which are ambiguous.

The

syn-

tactical nature of the relative clause will

then depend upon the

interpretation.

good example
oblectet,

is

Ter.

quae tuam senectutem

either

Phormio 433 hafrebis 'you will have some one

to cheer
'

who cheers'
371.

(Characteristic) or

'some one

(Purpose).

Clauses of Characteristic Denoting Cause or Opposition.

like o fortunate

In sentences

adulescens qui tuae virtu lis

is

Home-

ruin praeconem inveneris there

an apparent violation of the

to
'

principle that the Clause of Characteristic refers

an ante-

cedent not otherwise defined' (Gr.

as this

283. 1)

but in such cases

we may

explain the relative as referring to an indefinite

antecedent to be supplied.
force

According to

this

view the original

'
:

of the

above sentence would have been

etc.

fortunate

man, (one) who has found,'

ut qui, utpote qui,
etc.,

The

frequent

employment of
this view.

'as being

one who,' supports

The

use of the Second Singular in the subordinate clause would

then be a species of attraction.

THE SUBJUNCTIVE.
372.
Clauses
of

227

Characteristic

Introduced by

treatment in Gr.

283. 4 follows that of

ff.

nische Forschungen, Vol. IV. p. 226

Quin. The Brugmann in IndogermaBrugmann sees in the first

qui,

element of

this qiiin

an indeclinable Relative
for

which he thinks
Plural.

was capable of standing According to

this view,

any case either Singular or

quin might be equivalent to qui non, quae

non, quod non, etc.; the quin mentioned in 383, 391 must

then be regarded as a separate word.

Clauses of Result.

373.

Clauses of Result, introduced by

ut,

ut non, quin, qui, are

viz.

a development of the Subjunctive of Contingent Future,

its

from
b).

second phase, where there

is

a condition implied ( 360.

tale est
is

Thus

in the sentence hoc flagitium

ut quivis oderit, the

original

meaning was
'

'
:

this outrage
(i.e.

of such a nature as anyone

it).

you please would hate'

if

he should see

From

this to
it,'

the meaning of such a nature

is

that

anybody you please hates

it is

an

easy transition.

At the outset
malit,

probable that such Sub-

junctives as possit,

velit, nolit,

audeat figured largely in the

establishment of this category, since in these verbs the transition

from the idea of contingency to that of actuality

easy;
cf.

is

particularly

370.

374.

Relative Clauses of Result are simply a development of

the Clause of Characteristic.

At times

it

is

not easy to decide

indi-

whether the clause

is

one of Characteristic or of Result, and

vidual interpretations of the

differ.
eii7n

same sentence would doubtless often

For example,

in the sentence given in Gr. 284. 2 habltis

consulem qui parere vestris decretis non dubitet, the clause

. .
.

qui

Characteristic,

dubitet might be
'

felt

by some simply
'

as

a Clause of

a consul of the sort that

but the clause also

admits the interpretation 'a consul such that he does not hesitate
'

and

in that sense

it is

a clause of Result.

V
:;
c

575.

Qssses
:

f Sesalt
:

w. II

.*-_li

:::
"

:-

zvf-si

mi
:
r

iize: r:
'
.

J.
:

::
.
..

J"

v.: ;.
f
f
I

in

: :

rr
-

in
l_llSf

."

-..-fs
-j

::

J
-

iini.fi
-._-.

si 7:

W
:

mini

--- ---" ;

Hill

IS

~~l"f-T~T

J _ 5 L.

2 .1 _ E =

."":

."__;.
:
:

J.:_
r

Iitr.iilr:
:

:t
::

"mi
i::

~~ii

rziiirre.

is

::. :.

:z

3'""
*r_zz

:.LSl

rii-Sr=
:s

Izzz

i-:-i

:v
s:

_-

if

5_:_r:-;_;
'

-n_si_

if f . i i

nfii

ne

ifiii::

Tie lemrril

ismi

eis:_~ iissfs ~~~~ lie

susil

in iT

~~ ~ uies.

sac

be

m
>m_
::
:

-_r

;":
:

Tie

'7:~f"

lie

:_:s
his

lie

:::r::z: f":sztiora of Hale in 7T. F%hk CmsriLr*Ki, CwrmeB Sim&es '--. lie /i ^-ilmse :s simT i ::rn :: sii":
_

nine n:

Cttm-

:it

_.;

if

riiriiierii

T
::
7

eirie:

is

i 1:111 ::

inni
i

:._

~_

use
_*
"

Iiiiiiierisii
.

is

~i~

mj
t

nier Telii

f
i

~is~
::

is

::

LTiiiiieiisiii
: :

sumr
_

.*"-
.-*

i ee rs-ii ir
=
.

~"~~~
^

s:

:::::

in:

:ier-

sni- sni"
~

z._"

zzTzzz
-i _i use

ne
"_

sirimii fiisii_

_ ~_z

in

i
:

me
7.

ii er
z

in z
leierIT

ibf

~~

""~re
7

ii:-iliise

~i:

:n:ii
11

-~~~~

:.

::

1.11

Ifllf ISei

Iflllf

Th

SUBJUNCTIVE

: :

;_

379

"--.-.

e=e sxe

f:

H=>

~.f

'

ClussiC'dl

Hufait&v Vol. L. z
t

rzi
r zz :z

La

::

Izi.-Z
"

::

:zfzz VZ:

zzzz
~_

Zzrr Zzzzrz ~z:

zrZzrzL.
j

.-

1 "5c
t

~~~
:

'--.''
iz-r
:
:

r^l IZrZZ

rtfr:

::i

;zzzz
I-

::

-~^ ~ez : 7 I:
_

:ir
[

::zi

::

Zc

zzez

::"

t
:

i-Z-:: r:z
.

'izzz

:>:

::

7 -t

-i -=

i-

r:

Z
-

z
j_--_

~-z -_

2*1
: :

Zzz
:
-

::

ztZt ire :zez rzzzzi>i

"

lz z
f
:

_.
.

- :

-..::.:
7 it:
:

zi z z
:.

-Z_

..

"

_z

=c

Z: .it:
7
t

:r

:z:t z
zz
*

=z:z
~~~
"
'..

zezzf:

zlt5t

z z:i:z
-

r:

:t":t;t1":
1Z

mt z
zie

zzz
: I
'

znzzzzz
"

z
I
."

'i:
-

ZLfZZZZcl

iZ

izZzfez
:::
:

z
'

::I
"*

Lt't.iTZLez
r

zifz
t
.

ZZ ZZ

7zr Szi

~li

Z-

zz:
382.

:;

------

As

:.

zzz
15
:

Ziizzzzrz
ize
:

::

zzze:

if

zzzzr
z.

zz:Zi
:
'

_?:

nkf
:

z zz?
:

vz

i;

7~

n:

Zz

:."

7-Z

i.

'zi z :z

Tzrzrsz

; ;

230

SYNTAX.
is

In either case the dependent clause

of Jussive origin

ut

is

simply the adverbial particle which we have already met

in inde-

pendent sentences (see 358. a; 359).

ence between
tibi

The
tibi

original differ-

impero hoc mihi des and

impero ut hoc

mihi

des,

could hardly have been more than that between 'I

you, give

command me this.'

me

this,'

and

'

command

you, just give

Probably even

this distinction

soon passed away, and

the two forms of expression

lent in force.

came

to be felt as practically equiva-

383.
type,
rise

Taking now our

tibi

impero {ut) hoc mihi des as the

it

let

us consider a variety of Extensions to which

gave

a)

'

Extensions within the Present.'

it

After the analogy of

tibi

impero (ut) hoc mihi des,

like:

became

natural to form sentences

mihi imperat (ut) hoc

tibi
ill!
illi

sibi

dem

imperat (ut) hoc

sibi
sibi tibi

des
det
det

imperat (at) hoc

imperas (ut) hoc

mihi imperas (ut) hoc

tibi

dem.
mihi
des,

Our
to

original typical sentence, tibi impero (ut) hoc

:

was undoubtedly once paratactic

'

command you
'

give this

me

'

But the developments

just

enumerated could obviously

Extensions

never have stood in Parataxis ; they are analogical

within the Present.'
b)
tibi
'

Future Extensions.'
(tit)

An

illustration

of these would

be,

imperabo

hoc mihi des.

sentence like this could

It is

obviously never have stood in Parataxis.

simply

tibi

impero

(ut) hoc mihi des projected into the Future.

c)
(tit) (tit)
'

Past Extensions.'
da7'~es.

An
Here

illustration

would be,

tibi

imperain
i?npe?'d

hoc mihi

similarly

we have our

tibi

hoc mihi des projected into the past.

THE SUBJUNCTIVE.
d)
'

23
tibi
'

Negative Extensions.'
(tit)

These are exemplified by

These
'

non

impero
also
e.g.

hoc mihi des.

Negative Extensions

may
non
(tit)

be combined with Extensions of the kinds already noted,

tibi

non imperdbo
(tit)

(tit)

hoc mihi des (Future)

;

tibi

imperdvi

hoc mihi dares (Past)

mihi non ifnperat

hoc sibi dent (Within the Present).

e)
tibi
'

Interrogative Extensions.'

These are exemplified by quare These 'Interrogative Extensions'

impero (ut) hoc mihi des?

at the

may

same time
e.g.

also

be Future, Past, Within the Present,

or Negative,

quis tibi imperat (ut) hoc mihi des ? quis tibi

imperavit (ut) hoc mihi dares ? cur mihi non imperavisfi (ut)

hoc

tibi
'

darem

?
e.g.

f)
dem.

Conditional Extensions,'

si

tibi

impero (ut) hoc

tibi

These may
Negative
e.g.
;

similarly

be also Future, Past, Within the Pres-

ent, or

or they

may

contain a combination of these

tibi

Extensions,

si

mihi non imperavisfi (ut) hoc

darem,

a Conditional Negative Past Extension.

g)
te
'

'

Extension by Analogy of the Meaning of the Verb.'

Thus
:

oro (ut) abeas undoubtedly represents an original Parataxis

(Just)

go away

beg you,'

'

beg you
te

to

go away.'
'

Now
induce

after the

analogy of this we get

'

exoro (ut) abeas,

you

to

go away,'

succeed in

my
'

request that you go away.'

'

Similarly after tibi suadeo (ut) abeas,

I advise

you

to

go away,'

we

get tibi persuaded (ut) abeas,

succeed in

my

advice that
te

you go away,' 'I persuade you

(ut) abeas nor
tibi

to

go away.'

Neither

exoro

persuaded (ut) abeas could have stood in

an original Parataxis.

Such combinations would have

failed

to

make

sense.
'

recognition of the foregoing varieties of

Extensions

'

is

of

great importance for an understanding

of Substantive

Clauses

Developed from the

eties of subordinate

Volitive,

and

in fact for

many
the

other vari-

clauses of Subjunctive origin,

e.g.

Purpose
Optative,

Clauses,

Substantive

Clauses

Developed

from

232
Clauses
of Characteristic,

SYNTAX.
Result
all,

Clauses,

etc.

No

theory of

origin can possibly explain

or even any proportionally large

part of the

phenomena

ordinarily classified

under any one of

these syntactical usages.

large part of the instances belonging

under any

single syntactical category (Purpose, Result, Volitive

etc.)

Substantive Clause,
sort or another.

represent analogical Extensions of one

Classification of Substantive Clauses Developed

from the Volitive.

Developed from
384.

the Jussive

and

Prohibitive.

With Verbs

of

Ordering or Commanding.

Without

ut.

Original Uses: Plautus, Poen. 1155 died mi'hi ft Ham despondeas,

'

bid you to betroth your daughter to me.'

:

Extensions

Plautus,

Stichus,

624

dixi, in

carcerem

ires,

'

ordered you to go to prison' (a Past Extension).

With
Original Uses habeatis cu?-ae,
'
:

ut.

Plautus,
I

Men. 990 died ut imperium meum

bid you heed

my

orders.'
tibi

Extensions

Plautus,

Men. 784

edixi

ut caveres (Past

Extension)

With ne and ut
Examples
*
:

ne.
lie

Plautus,

Merc. 465 ad portum

;

bitds
s1

died

tibi,

I tell

you not to go
'

to the harbor'

Mil. Glo. 185

hoc

ei dicitd

ut ne digrediatur,

tell

her not to depart.'

385.

With Verbs

of

Begging and Requesting.

Without
ut.
s1

Original Uses

Plautus, Merc. 992

pacem

facialis oro,

'

beg

you

to

make

peace.'

'

THE SUBJUNCTIVE.
Extensions:
Plautus,

233
ignoscamics peccatum
'

Amph. 257 orant

suom,

'

they entreat us to forgive their fault

(Extension within

the Present).

With
Original Uses
:

ut.

Plautus, Cure. 629 quaeso ut mihi dicas,

'

beg

you

to tell me.'

Extensions: Plautus, Cas. 532 orabat ut properarem, 'he entreated

me

to

make

haste

'

(Past Extension)
'

Men. 1048 possum,

exorare ut pallam reddat,

succeed

in

my

request that she return

the cloak,' an Extension after Analogy of

Meaning of the Verb

3%3-

Similarly impetro ut,

'

succeed in

my

request that.'

With
Original Uses
1
:

fie

and ut

tie.

Plautus, Bacch.

1013

quaeso ne

me
te

deseras,

1 I

beg you not

to desert

me

Rud. 627 quaeso ut

ne pigeat

'

beg that you be not

Extensions
:

loth.'

Plautus,

Cist.

302 earn exores ne

tibi

suseenseat,

'induce her not to be vexed with you'; Bacch. 533 impetravi ut ne quid e~i suseenseat, I succeeded in my request that he
'

cherish no anger toward him.'

Some

regard the clause with exoro

it

and impetro

as

one of

Result, but

is

abnormal to have ne or ut ne with a Clause

of Result.

We

have no sure instance of any such Result Clause

Furthermore, affirmative clauses depenut,

in the entire Latinity.

dent on exoro and impetro often lack

in

which

is

never lacking

Result

Clauses.

It

is

therefore

much

simpler and

more

natural to explain such usages as analogical Extensions.

386.

"With Verbs of Advising.

The

origin of the Subjuncis

tive Substantive

Clauses after verbs of advising

:

indicated by
!

Plautus,

Men. 569 male habeas

sic censed,

'worry him

That's

my

advice.'

'

234

SYNTAX.
Without

ut.

Original Uses
furere,
frenzy
!

Cic.

in
this

Cat.

ii.

9 eos hoc moneo, desinant

:

'

give

them

warning

let

them cease
tibi dein

their

Extensions: Plautus, Merc. 1015 meant sordrem

des,

sha-

'You advise me

to give

you

my
ut.

sister.'

With
Original

Use:

Plautus,

Trin.

674 moneo hoc ut reputes, 'I

advise you to consider

this.'

Extensions
'

Plaut. Persa
to

842 hortantur tuo ut imperio paream,

they exhort

me

obey your bidding.'

With
Original Uses
te,
'
:

ne,

ut

tie.

Plautus, Persa

680 ne permittas do mum, moneo,

urge you not to hie yourself home.'

:

Extensions
not to go.'

Plautus, Stick.

608 suades ne

bitat,

'

you urge him

Under
ling,

this

head belong clauses with verbs of inducing, impele.g.

persuading,

Plautus, Epid. 8 7 perpuli

'

ut

ce~7is~ei-et,

'

induced him to believe

'he persuaded her to

Bacch. 964

persuasit, se ut amitteret,
in

let

him go'; Mil. Glo. 1269 induxi

animum
all

tie

oderim,

'

I've persuaded myself not to hate her.'

'

In

these cases the usage represents an

Extension after the Anal-

ogy of the Meaning of the Verb'

( 383. g).

Some

regard the

Subjunctive Clause after verbs of inducing, persuading, impelling,

as a Clause of Result.

But the same arguments are

to

be urged

against this

view as previously in the discussion of the nature

of the

clause

used with exoro and impetro,

verbs

viz.

the fact that

(instead of
lacking.

negative clauses with these

///

have

ne,

ut
is

lie

non), while in affirmative

clauses

the ut

often

See

385.

'.

'

'

' ;

THE SUBJUNCTIVE.
387.

235
'

With

facio, particularly

with

fac, facite, facito,

see to

it

Without
Original Uses
1
:

ut.

Plautus, Poen.
!

1035 linguam compescas, face,

'

see that you hold your tongue

it
!

originally

'

hold your tongue

see to

Extensions: Plautus, Men. 890 fac scia/n,

'

see to

it

that I

know/ 'make me know' (Extension within the Present).

With
Original Uses
that he receives
:

ut.
it

Plautus, Persa 526 ut accipiat, face, 'see to

it!'

Rud. 12 18 fac ut
faciam ut

exores, 'see that

you

persuade him
Extensions
:

Plautus, As. 28
;

scias,

Til

see that

you

know' (Future Extension)

reperiret,
'

Aid. 26 feci tliensaurum

ut hie
'

saw

to

it

that he

discovered the treasure

(Past

Extension)

With
Original Uses
:

lie,

ut

ne.

No

suitable

examples are at hand.

1145 fac ne metuam, 'see to
fear
! '

Extensions:
that
I

Plautus, Most.

it

have

no

occasion

for

(Extension within the

Present)
Especially interesting are the clauses with
verbs.
efficid

and related
n~e

Here belong
lacessas,
'

Virgil, Eel. 3. 51 efficiam

it

posthac

quem-

quam
sectifi

I'll
;

bring
Cic.

about that you do not

i.

challenge

anybody hereafter
ut
n~e

ad Fam.

2.
'

4 hoc videmur esse con-

quid cum populo agi

possit,

we seem
lie

to

have accom'

plished

this, viz.
iv.

that no business can be

done with the people

Q. Curtius,
'brought
it

14. 4

Macedonas

assecufos

quis tuto locus esset,

about that no place was safe'; Cic. pro Mifone, 13.

34 adepti estis ne quem civem metueretis, 'you have achieved your end, of standing in fear of no one.' All of these clauses
are probably to be regarded
as

Extensions after the Analogy

'

236

SYNTAX.
The
origin

of the Meaning of the Verb ( 383. g).

of the
'

usage probably goes back to clauses with fac, facite,

to
it

factio,

see

(that).'

From

fac, the

first

Extension seems to have been

to the other forms of facto ; but in our earliest Latin there are

many more

instances oi fac, facite,

all

factio,

followed by Substantive

Clauses than of

the other forms of facio combined.

c

From
succeed

facto the next Extension seems to have been to

in one's effort to see to
it

efficio,

(that)

'

and from

efficio

the construc-

tion was

extended to other verbs of closely equivalent meaning,

such as assequor, consequor, adipiscor.

Many

regard the dependent clause with these verbs

{tie,

as

one

of Result, but the employment of negatives

ut ne) and the

fact that the affirmative clause often lacks ut, point to a Volitive

origin.

Beginning with Cicero we find ut non in negative clauses

after facto, efficio,

which seems

to

show

that the Clause of Result

also

is

used with these verbs.

to it

Other verbs of seeing

388.

are curd, video.

With

cave, cave ne.

Expressions of the cave abeas Type.

The most
is

plausible theory as to the origin of these expressions

is

that cave abeas

formed on the analogy of fac abeas.

Cave
tie.

Original Uses: Plautus, Most. 324 cave ne cadas,

'

take care

you don't

fall
:

! '

Originally

'

Don't

fall
tie

Take care

Extensions
389.

Plautus, Pseud.
of

478

quid noceat cavero.

With Verbs

Permitting, Granting, Allowing.

Without
Original Uses
:

ut.
licet,

Plautus, Trin. 1179 vide as

'

originally 'see
l

you may

'

then,

You may

see

'

Amph. 806

sine dicat,

permit

him

to speak.'

'

THE SUBJUNCTIVE.
Extensions:
Plautus,
;

237
'let
'

Cist.

454 sine dicam,

sivt

me
I
let

speak'

(Present Extension)
live
'

Mil. Glo. 54

viverent,

them

(Past Extension)

With
Original Uses
:

ut.

Plautus, As. 43

dono ut expers

sis,

'

permit

you to be exempt.'
Extensions
:

Plautus, As. 847 potestatem dedi ut esses, 'I gave

be.'

you the opportunity to

Here we have a noun taking

the

place of the verb in a Past Extension.

390.

With Verbs

of

Deciding, Resolving,

etc.

Without
Original Uses
videant,
'
:

ut.

An

original use

would be

decernimits consules

we decree,
:

let

the consuls see to

it.

Extensions

The

foregoing was evidently the starting-point for

Cat. 29. 2 senatus decrevit darent operain

expressions like
consules,
'

Sail.

the Senate decreed that the consuls should give heed.'

With
Original Uses
:

ut.

No

suitable

examples are

at

hand.

Extensions: Plautus, Pseud. 549 rus ut irem constitueram, 'I

had resolved

to go to the farm'

(Past Extension).

391.

With opus

est, iisus est,

necesse est, oportet.

ut.

Without
Original Uses:
Lucretius,
iii.

593 fateare, necesse est, 'you must admit'; originally 'admit! you must'; Cic. de Fin. ii. 26
m~e

ipsum

curies, oportet,

'you ought to love

me myself

originally

'love

me my

myself! that's your duty.'

:

Extensions
'

Plautus, Poen. 1244 mihi patronus sim necesse

est,

must be

my own

defender.'

238

SYNTAX.
With
ut.

Original Uses
t~e

Plautus, True.

500 nunc

tibi

opust aegram ut
ill';

adsimuTes, 'now you must pretend that you're

Mil. Glo.

132 nunc

ad

vie ut veniat usust.

Several scholars regard the Substantive Clause after necesse est

as

one of Result.

But

if

the clause were one of Result,

it

would

be impossible to account for the practically invariable absence of

ut in
this idiom.

Moreover, we find that the clause with opus est

vii. 6.

takes ne as a negative in Pliny, Epp.

videretur.

3 opus esset ne reus

392.

With
to,'

sequitur,

reliquum

est,

restat,

in

the sense 'it

remains

'the next thing is

to.'

Without
Original Uses
:

ut.

These seem

lacking.
xv. 21. 6 reliquum est tua77i profec-

Extensions: Cic.

ad Fam.

tionem

amore prosequar,

'it

remains for

me

to

attend

your

departure with affectionate wishes.'

These expressions

also are followed

by Substantive Clauses of
viz.
'

Result, but they then have another meaning,

that.'

the fact remains

393.
after

Substantive Clauses Introduced by Quominus and Quin

Verbs of hindering.

As
is

explained in

Gr.

295.

3.0,

Substantive Clauses introduced by quominus are probably developed

from Purpose Clauses.

in the Volitive, since
( 368. 1).

However, they have

their ultimate origin

Purpose

a development from the Volitive

The

original character of Subjunctive Clauses of this

kind

may be

seen in an expression like fornfido viros impedit

:

quominus

velint, originally
willing,' i.e.

'fear hinders

men,

in order that they

willing.

may not be
minus
lit.

prevents

them from being

by which
not,'

Quo'in

means 'by which the

less,

and hence

order that not.'

THE SUBJUNCTIVE.
The
origin

239

of Substantive Clauses introduced by quin with

is

expressions of hindering
serve consideration
:

not altogether clear.

Two

views de-

(1)

Quin
In

in

such clauses

may be
and

relative

adverb,
lit.
'

com-

pounded of qui
not.'

(old Instrumental),

ne, 'not';

by which

this sense,

quin would be the exact equivalent oi quominus,

after expressions of hindering

and the Substantive Clause with quin

would have the same origin as that with quominus.

(2)

Qian

in such clauses

may be

the interrogative quin,

is

'

why

not?'

In that case the Substantive Clause

developed from the

Deliberative.

Cf
i.e.
'

Plautus, Amphitruo,
lit.

560 quin loquar, numspeak?

quam

potes deterrere ;
;

'why am

I not to

You cannot
;
'

prevent

it

'

You cannot prevent me from speaking

nequeo quin dicam.

Tri-

nummus, 641
dering are

retiiieri

Clauses introduced by quin after negative expressions of hin-

sometimes
its

classified

as

Result

Clauses.

It

is

of

course true that in

negative
dicate

developed meaning the quin-c\dMSQ

at

after
in-

expressions
(negative)

of hindering does
result,
e.g.

times

seem

to

nee

impediti sunt
l

quin face-

rent

may be conceived

as literally
do.'

meaning
But
this

nor were they preis

vented so that they didn't

conception

just as

possible in case of quominus-clauses after negative expressions of

hindering,

and even more so

in case of quominus-clauses after

affirmative expressions of hindering.

Thus,

t~e

impedio quominus
'I

haec facias might theoretically be conceived as meaning

hinder

you so that you do not do

this.

'

But quo minus

is

clearly a pur-

pose particle, so that the original purpose character of the quomi-

^^-clause seems beyond question.

Any

consistent treatment of

Substantive Clauses must have regard to their origin, not merely

to the English rendering.

Thus, in a sentence like

eis

persuasit

ut exirent, 'he persuaded them to go out,' the z//-clause might

seem

at first sight to indicate a Result,

but an examination of such

clauses clearly shows that they are developed from the Jussive.

'

240
Clauses introduced by
sarily
lie

SYNTAX.
after verbs of hindering are not neces-

developed from the Jussive,

is

as suggested in
it is

Gr.

295.

3.

This

the

more probable view

<^z;z-clauses,

but

also possible that, like

quominus and
Purpose Clauses.

they

have

been developed

from

394. Substantive Clauses in Sentences of the

est cur, nulla causa est quin, etc.

Type

nulla causa

{Gr.

295. 7).

These
This
is

have
is

been explained
view,

as

developed from the Deliberative.

the

among

others, of

Schmalz {Lat. Synt?

350), and

supii.

ported by the history of these clauses.

17.
1

Cf. e.g. Cic.

'

ad Fam.

quin decedam nulla causa

!

est,

originally
'

why

shouldn't I
I

go away

There's no reason'
Cf. Ter.
'

later

there's

no reason why
est,

shouldn't go away.'

Andria 600 quid causae

is

quin

in pistrinum proficiscar,
set out for the

what reason
'

there

why
is

I shouldn't
?

mill
'

'

originally

what reason

there

Why

shouldn't I set out?

Substantive Clauses Developed from the Optative.

395.
After Verbs of

wishing and de si ring {Gr.

is

296.

1).

The
evident.

Optative origin of these Substantive Clauses

It

sufficiently

should

be noted, however, that in colloquial lan{cf.

guage void sometimes has the force of commanding

English authoritative

the

/ want,

e.g.

in

/ want you
'

to

understand).

In such cases the Substantive Clause with void must be referred to

a Volitive origin,
e.g.

void earn ducds,

want you to marry

her.'

396.

After Verbs of fearing {Gr.

296. 2).

Instructive for

the history of the construction are such early Latin uses as Ter.

Andr. 277
ferre,
'

Hand verear si

in te sit sold situm

sed ut vim queas

;

should not

fear, if it

were to depend on you alone

;

but

may you be

able to withstand compulsion

705 dies hie mi ut

sufficient

satis sit vereor

ad agendum, 'may

this

day be

(I'm

afraid though).'

THE SUBJUNCTIVE.
Substantive Clauses of Result.

24

397.

Expressions
ill,'

like accidit

ut aegrotaret,

'it

so

happened

that he was

show

clearly the origin of the Substantive Clause

of

Result.

But the Result notion early became weakened in

these clauses,

and the substantive notion became so prominent

by ut occur where not only no
never could have existed,
'

that Substantive Clauses introduced

notion of Result exists, but where

verisimile

it

e.g.

non

est

ut

ille

anteponeret,

it's

not likely that he preis

ferred'

accedit ut doleam, 'another fact

est

that I

am

suffering';

praeclarum
them'
;

ut eos amemus,

'it's

a noble thing that

we

love

reliquum est ut virtus

is

sit frugalitas , 'the fact

remains that

economy

a virtue.'

Substantive Clauses Introduced by Quin after

Won Dubito and

Kindred Expressions.
398.

In the expressions non dubito quin, quis dubitat quin, non

est dubiicm quin,

hand dubium

est quin, the qztin-clause

is

probquis

ably developed from the Deliberative

Subjunctive.

Thus

dubitat quin in virtute divitiae sint originally meant 'why shouldn't

there be riches in virtue
find
!

who doubts

it?'

It

seems

difficult to

any ground in the history or signification of these clauses

for

regarding them as Clauses of Result, a view advocated by some.

Indirect Questions.

399.

The

origin of the Subjunctive

in

Indirect Questions

is

not yet clear.

The

construction

is

manifestly a relatively late

one

in

the development of Latin syntax.

Plautus and Terence

frequently

employ the Indicative

in

such sentences.

Conditional Sentences.

400.

The treatment

in the

Grammar

follows the

traditional
is

classification,

which has regard exclusively to what

each instance.

implied

in the Protasis in

242
401.

SYNTAX.
Conditional sentences are the development of an earlier

Parataxis ( 367).

Thus we may assume

was bene
est,

that the earliest type

it

of

si

valet,

bene

est,

valet,

'

is

well

he

is well.'

The

conditional force was purely the result of the context, which

indicated that valet was something assumed.

As language devel-

oped, the fact that one clause was related to the other as an

assumption or condition was brought out more definitely by the

use of st; yet conditional sentences without
less
si

occur with more or

frequency in

all

stages of the Latin language (Gr. 305. 2).

They

are simply a relic

of the earlier paratactic stage.

si.

The

ori-

gin of the conjunctional use of

was as follows

Si was originally

an adverb meaning
sentence with
well so,
est,

so.

The most
is

primitive type of a conditional

in bene est
si,

si

would be seen
he
well.'

valet, i.e.

'it

is

{viz.

that)
is

In

this expression si limits

bene
si.

and valet
use of

really

an appositive of the adverbial idea in

is

The

si as

a conjunction
si cf.

secondary and the result of

its

association.

With

English so in such expressions as so you

pay me, I shall

be satisfied.

402. Conditional Sentences of the

Second Type.

Here

the

Subjunctive in the Protasis was originally Jussive in

character.

Thus a sentence

like si videat, credat

lit.
'

would, in
(i.e.

its earliest

form,

have been videat, credat,

see),

let

him

see

assuming he should
the Subjunctive of

he would then believe.'

The Apodosis
'

is

Contingent Futurity, conventionally called

Potential.'

403.

Conditional Sentences
is

of

the

Third Type.

The

origin

of this type
Optative,

obscure.

Perhaps the Protasis was originally an

lit.
'

i.e. si

adesset, bene esset,

that he were here

it

would be

well.'

The employment
tive

of oportuit, decuil, deb'ebam, and of the Indica-

of the Periphrastic Conjugations in Apodoses of Conditional

is

Sentences of this type

frequently the result of

ellipsis.

Thus

in

THE SUBJUNCTIVE.
si

243
ituri,

Pompeins

occisus esset, fuistisne

ad arma
in

the thought

is

'were you about to proceed to arms (and would you have done
so
? )

had Pompey been

slain ?

'

So

eum

patris loco colere debe:

bas, si Tdla in te pietas esset the full sense

is

'it

was your duty to

revere him (and you would

now be doing

it),

had you any sense

of devotion.'

Subordinate Adversative Clauses with Quamvis.

404.

Here

the Subjunctive was originally a paratactic Jussive.

originally

Quamvis was
in the

quam

vis,
est,

'as
id,

much

as

you

wish.'

Thus
tamen
is

sentence,

quod turpe

quamvis

occultetur,

hones turn fieri no n potest, the original meaning was: 'what

let it

base,

be concealed as much as you wish, cannot become honorIn

this

able.'

way quamvis ultimately developed

'

into a Conjunc-

tion with the force of

although.'

Clauses of Proviso with

Dum, Modo, Dummodo.

Thus
in

405.
senibus,

These were

all

originally Jussive.

manent ingenia

modo permaneat studium

only
interest

et industria, the original sense

!

was:

'let

and vigor remain

(then) old men's

faculties remain.'

Dum

was originally an oblique case of a noun

in oderint,
!

meaning

'while.'

Hence

dum

metuant,

the original

sense was 'let

them

fear the while

(then) they

may hate.'

Some

regard the clause of Proviso with

('while').
junctive,

dum

as

originally

temporal

But that view fails to account for the use of the Subalso ignores the fact that the negative with the
is

and

dum-

clause of Proviso

always ne.

INDEX.

[The References are to Sections and Paragraphs.]

A.
a, a,

abs, 261.
ac, 93. 1.

pronunciation,

4.

changes, 71.

Accent, 54

f.

a-Series, 66.
a, 72.

accestis, 47. 2.

Accusative, syntax, 303

original force, 311.

ff.

a-Series, 67.
a,
'

from,' 261.

a-

Stems,
;

in

f.
;

of person or thing affected, 303. of result produced, 303; 305; 306.

261.

ab, 93. 2

96. 1

abjetis, 15. c.

with passive used as middle, 304. synecdochical, 307.

Ablative, 331 f. absolute, 347.

of

Greek, 307.
in exclamations, 308.

accompaniment, 336.

as subject of

inf.,

309.

of accordance, 340. of agent, 333.

Accusdfivus, 297.
deer, 92; 100. 3. acerb us, 100. 3.

of association, 337. of attendant circumstance, 338. of cause, 342. of comparison, 334. of degree of difference, 343. of duration of time, 350. of manner, 339. of means, 341. of price, 344. of quality, 345. of separation, 333. of source, 333.

acerrimus, 182. 3. Acquiescence, Subjunctive

ad, 262.
in composition, 58. a.

of,

362. d.

Adjectives, 181

ff.

admodum,

259.

admoneo, with genitive, 326. Adverbs, 253 f.

in -e, 130.

in

e,

257.

in -0, 130.
in -0, 255. 2.

of specification, 346. of time at which, 350.

of time within which, 350. of

Adversative clauses, 404.

in indicative, 400. 3. adversus, 258. ae, pronunciation, 10.

way by which,

341, 5.

Abldtlvus, 297. Ablaut, 62 f.

in cas'_-endings, 70.

aedes, 97. 2. b. aegrotus, 203. VII. d.

in suffixes, 70.

Aesculapius, 91.
a/, 261. 5.

Ablaut-Series, 62
abluo, 103. 4.

f.

agceps, 20.
1.

1.
;

-abrum, -acrum, -atrum, 51.

1

agellus, 100. 3

106. 2.

For words containing hidden quantities and spelling, see the special lists, p. 52 and p. 79. 244

for

words of doubtful or varied

INDEX.
ager, 92 100. 3. aggulus, 20. 1.
;

245

ar versus, 262.
Aspirates, 31 asporto, 105.
;

97.

agricola, 112. 2.
ai, 86. ai,

1.

changes, 80.
earlier

form of

ae, 10. 1.

aj (j), 109. 2. Assimilation of consonants, 106. of vowels, 90.

attingo, 71. 5.

aid, 80. 1.

airid, 141.
-al, 88. 2.

au, 86.
azz,

pronunciation, 12.

a/ for
_
*

/,

100. 1.
;

ala, 89

105. 2.

au, changes, 84. au-, 261.

auceps, 92.

ali quid, 254. 1.

allium, 88.
alnus, 105.

1.

azidiam, 221.
audies, 222. 2.

1.

Alphabet, 1. a mare m, 222,

ambi-, 263.

audirem, 222.
3.

3.

aufero, 261. 3. aufugio, 261. 3.

ambo,

97. 1.

<5.

Augment,

200. 1.

amem,
tf?z

222.
?7,

aurora, 86.
atirufex, 76. 4.

for

102. 2.

anceps, 92.

a^,

97. 3.

A.
3.

ausim, 219. auspex, 92.

#/, 93.

animal, 93. miser, 23 97.

;

A.
B.
, ,

antae, 102. 2.
#;zzV,

264.

96.

1.

antemnae, 106. 4. . antequam, with subjunctive, 379. Anticipatory subjunctive, 379.

anticus, 57.
2.

pronunciation, 27.
1.

bacca, 88.
256. 255.

basiuvi, 98. 3.
&?//z,
<fe;z,

N.

1. 1.

Antisigma, 1. 5. Aorist Optative, 219.

sigmatic, 200. 3. strong, 200. 2.
aperio, 96. 1
a^?.r, 36. 3.
;

(Indo-Eur.), 97.
;

1.

to<?, 96. 1

203.
1.

2.

blmestris, 105.
tozz, 185. 2.

261. 2.

to, 186. 2.
bobus, 180. 3.
&, 180. 3.

Apocope,
apud, 265.
-ar, 88. 2.
ar-, 262.

93.

Bosphorus, 31.
-br-iox bracca, 88.
1.

3.

-sr-, 108. 3.

ar, for r, 100. 2.

arbiter, 262.

Breves Breviantes, 88.

I.

3.

arbosem, 98.
ardor, 92.

breviter, 259.

bruma, 182.

I.

are/acid, 204.
arfu'erunt, 262.
-ar/- for -a/?-, 99.

bubus, 180. 3. bucca, 88. 1.

armiger, 76. 4. armus, 100. 2. aro, 203. VII. a.

<z/\57,

C.
.94,

105.

1.

C.

pronunciation, 25. Gaius, 1.3.

246
C.

INDEX.
1. 4.

= centum,
1.

-clo- for -tlo-, 95. 1.

caecus, 11.

Clodi us, 84.

<5.

1.

caedo, 104.
caelebs, 11.

cludo, 87. 2.

Cn.

Gnaeus,

1. 3.

caelum, 1 1. caerimonia,

co- in
11.

co- in

compounds, compounds,

58. b. 6. 58. b. 6.

caeruleus, 99.
caesaries, 98. 3.

COCUS, 57. 2. N.

calamitosus,

no.

coemeterium, coenum, 81. 2

<rc/>?,

n.
;

103. 5.

calcar, 93. <ra/7<?, 103. 2; 203. VII.

81. 2;

206. 2.
1.

coeravere, 81.
coetus, 81. 2.

Cardinal numerals, 183.

ari>, 100. 2.

Carthagine, 141. Carthagini, 141.

Cases, 295
,

cog?zomen, 104. 1. a. cognosco, 105. 1.

collis, 76. 1.

ff.

wz-,
296.
tffc,

<%>#,

in composition, 58. .

names,

Case-endings, see c-stems, 0-stems, Case-theories, 298 ff.

Caslorus, 138.
catellus, 100. 3.

Comparative Degree, 181. Comparison, 181 f. Compensatory Lengthening,

compleo, with genitive, 330. Concessive Subjunctive, 362. con clausus 87.
,

89.

catus, 69.

I.

Cauneas = cav(e) n(e) eas, 16. Causal clauses, 376. introduced by ;, 377.
causa, 98.
2.

1. /.

conditio, 25. 3.

Conditional Sentences, 400

conditus, 65.

ff.

confido with ablative, 349.

2.

mfo,

88. 3.

t^wa, 11.

centesimus, 184. 9.

Conjugation, 200 ff. consistere with ablative, 341. Consonant changes, 104 ff.

3.

centum, 106. 4; 183. 14.

cerno, 75. 2.
ceteri,

Consonant stems, 137 ff. that have partially adapted themselves to 2-stems, 159.

n.
1.

Cethegus, 31. 3.
*#*, 108.
>&,

Consonants, 15 ff. 94 ff. Consonants doubled, 34.

;

31. 2.

final, 109.

cineris, 75. I.

constare,
' ,

'

consist

of,'

341. 3.

circa, 266.

cost,' 344.
2. 2. 3.

cir titer, 266.

contempsl, 108.

circum, 266.
cw, 267.
^/Y^r, 181. 2.

contemptus, 108.

contmere, with ablative, 341. Contingent Future, 355. b.

con tid, 103.
3.

citimus, 182. 2.

Y5, 88. 3 257. 1. '^a, 255. 3; 267. clam, 268.

;

1.

contra, 255. 3. contyb erndlis 6. 2.

,

conubis, 133.
;

Claudius, as grammarian, 1. 5 16. 5. Clauses of Characteristic, 369 ff. distinguished from relative clauses of purpose, 370. denoting cause or opposition, 371. introduced by quin, 372.
,

conubo, 89.

1.

convention}., 141.

convicium, 25.
coquo, 96.
>r,

3.

1.

109. 3.

cordis, 100. 2.

INDEX.
cornu, 100.
,

247

2.

die, 223.
dfortf,

corpulentus 108. 4,
<?j-,

82.

67.

didici, 206. I.
dfo?,

coventio, 103. 3.

173.
86.

Crassupes, 76. 4.
cratis, 100. 2.

<&'<?;#,

Diespiter, 180. 4.
<*4?* J 3 8
. , ;

cresco, 203.

VI.
2. .

73- ^

94- 35.

cribrum, 97.
->-<?-,

for

-^/tf-,

95. 1

99.

with ^zzz-clause, 367. dingua, 95. 2.

', 14;

198. 4.

cujus,

198. 3.

Diphthongs, 10 ff. 80 ff". Diphthongal stems, 180.

;

culleus, 88. 1.

dirimo, 98.
^'j-

1.

culpa, 76. 1.
<r//z-Clauses,

dlrus, 104. 2.

377

f.

not afo- in compounds, 48.

;

cuppa, 88.

1.
/J.

disco,

curamus, 203. VII.

curtus, 100. 2.

64 105. 1. Dissimilation of syllables, no.

Distributive numerals, 185.

dins, 86.
dlvissio, 98. 2.

curvus, 100.

2.

D.

dlvus, 82.
<^J.rz,

4
a?,

95-

208.

pronunciation, 28.
500,
1. 4.

dixim, 219.
6frr<3,

D=

205, 3.

dacruma,

95. 2.
1.

dlxti, 47. 2.
<$J,

dajjtma, 88.

202.

1. .
/>.

Dative, 312 f. of agency, 318.

of indirect object, 313. of possession, 318.

do ceo, 203. VII.

domamus,
dovii, 256,

203. VII.
1.

domu'i, 103. 4.

of purpose, 319. of reference, 316.

donee, with Subjunctive, 380.

with compounds, 315. with verbs signifying favor,'

'

'

help,

Double consonants, 32 f. Doubled consonants, 34. drachuma, 91.

a^r, 223.
ducentl, 183. 15.
?<?,

etc.,

314.

Dafivus, 297.
de, 271.

64.

debllitare,

no.
;

duim, 218.
183. 10.
tf'awz,

decent, 102. 1

introducing a Proviso, 405.

380.
2.

decimus, 184. 5. Declension, in

dum, temporal,
ff.

duo, 183.

De-composition,
deeram, 88.
4.

87. 2.

to,

186. 2.

deesse, 88. 4.

E.
e,

de lector, with the ablative, 349. 2. Deliberative Subjunctive, 358. e; 363.

pronunciation,

5.

,
?, ?,

Demonstrative Pronouns, 191

deni, 185. 2.
dentio,

ff.

from from

a, 71. 1.
1,

75.

1.

no.
;

changes, 73. ^-Series, 64 f.

2.
e,

denuo, 103. 4 259. dk (Indo-Eur.), 97.

-dhlo- 97. 2.
y

^-74'from,' origin, 105. 2.

c.

^-Series, 65.

248
^-Sterns, 172
ea, 192. 2. b. ea, 192. 6.
f.

INDEX.
<?;, 192. 5.

eumpse, 196.
euntis, 76. 5.
,

ea (adverb) 255. 3. eadem (adverb) 255.

,

3.

261. 2. :, 273 ex, in Composition, 58. c,

;

earn, 192. 5.

exaequo, 87.

1.

eapse, 196.
ec-,

273.
1. d.

exemplar is, 99. exemplum, 108.

existumb, 80.
2.
1.

2.

ecferri, 105. I.

ecus, 57.

exquaero, 87.

edi, 206. 2.

ex s ilium, 71.

7.
3.

edim, 218.
/<?,

exstra (d), 255.

exsulto, 71. 3.

202. 4. 273187.

*/-.

egl, 206. 2.
<?^-0,

exsulo, 71. 7. ex terus, 181.

I.

extimus, 182.

2.

ei,

82.

, 86.
ei,

extra, 255. 3. extremus, 182.

1.

192. 4.
;

^z/j, ejus, 82. 3

192. 3.

F.
21. /j pronunciation,

Eleven,
z for

etc.,

183. 11.

-ellus, 51. 2.
;;?,

/
2.
1.

origin of letter,

1. 3.

102. 1.
2.

y#, 223.
facile, 254.
,

emi, 206.
for

1.

emptus, 108.
<?;z

7?,

102.

facilUmus 182. fagus, 97. 1. fl.

_/#/.r,

3.

o
orj^.

endo,

100. 1.

eo, 192. 6. eo,

'

familias, 113.

go,' 202. 2.

famulus, 91.

eopse, 196.

fanum,
_/ar (r)
,

65.

-^z, 206.

2.

/tfrz, 97. 1. a.

equabtis, 122.

109. 2. 219.

equus, 57.
-ir, 88. 2.

2.

faxim,
y/
(/)
,

/a.r5, 205. 3.

^r for

r, 100. 3.

109. 2.

eram, 204.
-er cuius, 51. 5-

/*/, 86.

femina, 11;
feres, 222. 2.
/*?ri?,

97. 2. #.

^r^a, ergo,

ttjq..

-emus, 51.
ero, 205. 3.

3.

erus, 23.
y, est (edo),
^j-(j-),

97. 1. a; 202. 5. 243. ferre, 106. 3 errem, 222. 3.

;

f
50. 2; 202. 4.

festus,
yz"ay,
yfrfz,

65.

109. 2.

yra, 108. 3.

<?.<?, 243. essem, 222. 3.

172.

206. 1.
2.

-estus, 51. 4.
et,

93-

/#?, 64; 82. with ablative, 349.

filiabus, 122.

Ethical Dative, 317.

eu, 85.

z, pronunciation, 13.
eu, 86.
I

films, 90. Final Consonants, 109. yfzziJ, 203. IV.

INDEX.
Jingo, 97. 3. A.

249

Jtnimus, 203. VII. flxus, 108. 1.

fiecto, 203. III.

b.

gn, pronunciation, 20. 4. gn, quantity of vowel before, 38.

-gn- for -en-, 94.
3.

foe dus, 64; 81. 2. fons,fontis, 41. fords, 254. 3.

forts, 256. 2.

Gnaivod, 109. 1. gndrus, 102. 2. gnatus, 102. 2 104.

;

1.

gnosco, 203. VI. gnotus, 104. 1.

for inus, 97.

3.

B.

/brte, 257. 3.

Gracchus, 31. 3. gradlor, 97. 3. B.

fortunas, 113.

gram en,
301.
1.

97. 3.

forum, 97.

2. a.

Grammatical

theory

of

the

cases,

frons, fr otitis, 41. fruor, with ablative, 341. y5, 97. 1. a 210.
;

^r^,
30-

171. 2.

Guttural,

distinguished

from

Palatal

fulmentum, 105.
fiimus, 97. 2. a. /undo, 97. 3. A.
;

1.

Gutturals, 94
107.

f.

funebris, 108. 3. fungor, with ablative, 341. furuos (= furvos) 98. 2.

,

H.
1.

h,

pronunciation, 23.

hac, 191. 6.

futtilis, 88. 1.

Future Indicative, 200. 6; 205. Future Perfect Indicative, 216.

/at/*, 210.

haec (Fem.), 191. 2. b. haec (Neut.), 191. 7.

hallucinari, 88. I.

hanc, 191. 5. harena, 23.

G.
g, 94-

Hartung's theory of the cases, 299. haruspex, 23.

hellud, 88. 1.
/k?rJ,
>&J<;,

^ pronunciation, 29.
g, earliest form, 1.3.

256

1.

gaudeo, 86. a. gaudeo, with ablative, 349.

191.
1.
f.

2.

hlc (Adverb), 256.

gemma,

106. 2.
,

Hidden Quantity, 36
hiemps, 108. 2. hiems, 97. 3. A.
hieto, 71. 6.

Genetivus 297.
Genitive, original force, 320. with adjectives, 323.

with nouns, 321.

of quality, 322.

hisco, 203.
/$<?,

VI.
c.

191. 2.

with verbs, 324 genus, 70.

gero, 98.
I. 1.

ff.

hoc (Ablative), 191. hoc (c), 109. 2.

;

6.

Gerund, 252.

gg, for ng, 20. 1. g/i (Indo-Eur.), 97. gignb, 203. II.
gllsco, 203.
gluttire, 88.

hohis, 23 97. 3. A. honorus, 138. Hortatory Subjunctive, 358.

3.

hortei, hortl, 81. 4.

horteis, horfis, 81. 4.

VI.
1.
1.

hostis, 97. 3. B.
/zzV,

glorior, with the ablative, 349. 2.

ghtttus, 88.
,?,

191. 4.

hujus, 191. 3. hum'i, 256. 1.

94. 3.

quantity of vowel before, 39.

-<;;-,

-gm- for

hunc, 191. 5. Hypotaxis, 367.

2 SO

INDEX.
in
-r<?,

243. 246.

i,

pronunciation,
ISat, 80. 2.
ei,
<?z,
?2,

6.

in -isse, 244.
in -ri,
-z,

*. i,
i

from 1 from J from i from from zfrom

z
z z

in -ier, 246.

82.
81. 3.

Inflections,

in

ff.

infra, 255. 3.

78.

inquirb, 80. 2

a,
?,

71. 2; 5.
73.
a.

Instrumental case, 331 332; 335. uses of the Ablative, 335 ff.
;

for/, 103. 2.

intellego, 87. 1.
1.

consonans, 15. longa, 36. 3.

ff.

intelligo, 87. 1.

inter, 277.
interest,

z-Stems, 171.

z-Stems, 148
id, 192. 2. c.

with the ablative, 349. wiih the genitive, 329.

3.

inter ieisti, 50. 3.

idem, 192. 8. id genus, 310.

z^zls-,

Interrogative Pronoun, 198.

1.

intimus, 182. 2.

idoneus, followed
68.
-tens, -ies, 186. 5.

by

^zzz-clause, 368. 5.

intramus 203. VII.

,

b.

intus, 278.
ipse,
is,

196

197. 182. 3.

ignosco, 105.

1.

192.
;

zyfor/, 15.
ilicet,
z7?V<?,

3.

-issimus, 43
z-yzV,

204.

r-4; 197.
256. 1.

76. 4; 259. illacrymant, 6. 2.

z'/zV,

-ister, 51. 4.
z-fz*z,

195

197.
1.

illic,

256.

J.

-i litis, 51. 2.

/ 103.
1.

illTisfris,

105.

/,

Imperative, formation, 223 f. Imperfect, Indicative, 200. 6; 204.

Subjunctive, 222.
3.

/,

pronunciation, 15. defence of the character,

in

2.

j acid, 203. VII.

compounds,

60.

impi emus

203. VII. b.

y<?-class of verbs, 203.

VII.

impleo, with genitive, 330.

impltio, 103. 4.
z;z,

jticundus, 103. 3. jungo, 203. IV.

275.
in

junior, 103.
Jupiter, 104.
Juppiter, 88.
jussus, 106.
jtisti, 47. 2.

3.

z>z,

composition, 58. d.

180. 4.

incertus, 100. 3.

1.

Inchoatives, 49. Indefinite Pronoun, 198.

Jussive Subjunctive, 358. a; 362.

2.

indigena, 76. 4. indigenus, 275. indigeo, with genitive, 330. by followed indigmis,
368. 5-

juxta,2SS- 3; 279-

^J-clause
, 25. 4.

K.

Indirect Questions, 399.

indoles, 275.

indu-, 275.

/,
/,

indugredi, 275. induo, 103. 4. Infinitive, formation, 243

pronunciation, 17. earliest form, 1. 1.

50, 1. 4.

L=
fff

/ for r, 99.

INDEX.
o_
/,

251

100.
1.

la for / 100.

malim, 218. maid, 202. 6.

3.

Labio- Velars, 94; 97.

lac, 104. 1
;

mancipiiim, 71.
"*. 75-

4.

109. 3.

Lacedae?noni, 141. lacruma, 6. 2; 95. 2. laetor with the ablative, 349.

/a/ztf,

3[ 93Masculine a-stems,
<S.

112. 2.

materies, 102. 2.
2.

Matiita, 86.

100.

1.

maxumus
#Zi?,

6. 2.

lapillus, 106. 2.

187. 4.
187. 4.
2.
/)
;

lapsus, 108.

1.

#z</,

lama,

16. 1. .

medius, 97.
wz^z, 187. 2.
z^/, 109. 3.
/).

103. 2.

larva, 98. 1. lalrina, 103. 3.

latus, 100. 1
;

104. 1.
1.

membrum,

108. 3.

lavacrum, 95.

legimini (Imperative), 227. Lengthening of vowels, 89.

cevi, 82. 2.

memini, with genitive, 325. memordi, 206. 1.

Menerua,

98. 1.

viercennarius, 106. 2.
me?'idie, 256. 1.

levir, 95. 2.
/<?7.y,

82. 2.

Messalia,
mens, 190.
z;zz

88.

1.

libertabus, 122.
liber fas, 100. 3. //$!?/ {lubel), 78.
/zV<?/\

Metathesis, 107.
1.

with the subjunctive, 389.

?;zz

(Dative), 187. 3. (Vocative), 190.

1.

lingua, 95. 2. /z';z<?, 203. V.

Michelsen's theory of the cases, 300.

Middle
;

voice, 200. 4.
;

Liquids, 17 f. 99 f. as sonants, 100.

lis,

?7zz7zz,

88. 3 187. 3. militiae, 256. 1.

104.

1. b.

mille, 183. 16.

I'Mera, 88.

1.

Localistic theory of the cases, 299. Locative uses of the Ablative, 348 f.

millesimus, 184. 10. miluos, 16. 1. .

mina, 91.
minoris, 344.

Locative of the goal, 351.

locus, 104. 1. b.

minus, 181.
zzzzj,

Logical theory of the cases, 300. Long diphthongs, 86.

lubet, 78.
Iubidd, 6. 2.

187. 2.
I.

mis ceo, 105.

miser, 98. 3.

miseret, with genitive, 328.

77ZZJ7,

lucrum, 99.
Indus, 81.
i.

98. 2.
;

modo, 88. 3 237. 1. modo, introducing a proviso, 405.

M.
m, pronunciation, 19;
20. 3.

moenia, 81.
mollis, 76. 1

2.
;

106. 3.

M=
'

m, 102. o
1000,
1. 4.

motnordi, 90

206. 1.

moneam,

2.2.x.

viaestus, 11.

moneo, 203. VII.

1.

c.

magis, 181.

monerem, 222.

3.

magislres, 131.

mons, montis, 41.

maior, 80. Mains, 80. male, 255.

1. 1.
1.

Moods, names
morbus, 97. muccus, 88.
1. 1.

of,
/).

353.

252
muliebre sccus, 310.
muliebris, 108. 3.
2.

INDEX.
Nouns, declension,
novem, 183.
noveni, 185.
novos, 73. 3.
7,
t?^,

in

ff.

9.
2.

mulsi, 105. 1. multa, 76. 1.

Multiplicatives, 186.

novitas, 76. 4.

multum, 254. munia, 81. 2. Mutes, 24 ff.

muttlre, 88.

94

ff.

722",

pronunciation, 20. 2. quantity of vowel before, 37. quantity of vowel before, 40.
etc.,

1.

nudiustertius, 86.
?z//a

N.
n, pronunciation, 20.

n
7z

for

171,

101.

1.

n, 102.

causa quin, Numasioi, 86. <5. Numerals, 183 f. mimero, 257. numerus, 76. 2.

394.

adulterinum, 20.
, 102. 2.
;

1.

7z-class of verbs, 203.

IV.
<7,

O
pronunciation,
7.
5,
<?

afor

101 f. Nasals, 19 f. as sonants, 102.

changes, 76.

from

2,

73. 3.

Nasal Stems,
natus, 104.
1.

147. 2.

^-Series, 68.
lost, 93. 2.

?iavis, 180. 2.

^-sterns, 124.

f.

nd, quantity of vowel before, 40.

ne,
'

0.773

verily,' II.
1.

from

azz,

84.

1.

nee, 93.

0-Series, 69.

necesse

est,

with the subjunctive, 391.

I.

ob,

nectd, 203. III.

280. 93; 96. 1 ob in composition, 58.

;

e.

neglego, 87.

obliviscor, with genitive, 325.

negotium, 25. 3. nexus, 108. 1. 77/j pronunciation, 20. 2. 7?/ quantity of vowel before, 37. -nguont, -nguontur, 57. 4.
nihil, 90.

occupo, 71. 4.
occultus, 100.
1.

octavus, 184. 6.
octingenti, 183. 15.
0C#?, 183. 8.
<?,

pronunciation, 92
;

11.

nimis, 181.

1.

offievna,
o/',

106. 2.

ninguit, 97. 3. B.

86.

7zm, 90.
7Z7w>, 97. 3. B.
Tzo-class of verbs, 203.

oi,

changes, 81.

oinos, 11.

V.

oitilis, 11.

nobis, 187. 7.

0/
1.

from

/,

101.

noctu, 86; 256.

nolini, 218.
7z<?/<?,

oliva, 73. 5.
olle, 195.

103. 3

202. 6.

ote,

81. 4.

Nominativus, 297. ndminus, 138.

nonus, 184. 7.
/zoy,

onustus, 76. 5.
operio, 96. 1
;

280.

oportet, with the subjunctive, 391.

187. 5.

Optative, 217.
original force, 355.

noster, 190. 4.
nostri, 187. 6.

Subjunctive, 359
6.

364.

nostrum, 187.
notus, 104. I.

optimus, 182.

2.

0^/0, m., 112, 2.

INDEX.
optumiis, 6. 2. opus, 138.
ph,_ 31. 2; 5.

253

piaclum, 95.
2.

I.

opus

est,

with ablative, 341. with subjunctive, 391.

pietas, 76. 4.

pignosa, 98.

1.

or, 88. 2.

pilleus, 88. 1.

or bus, 97. I. (J. Ordinals, 184.

piliem, 105. 2.

pilumnoe, 131.
f.

Orthography, 56
os, 109. 3.

plectd, 203. III.

pledres, 181. 3.
1.

ostendd, 105.
ou, 85.
<??^,

plerumque, 254.
plied, 87. 2.

1.

for , 85.
86.

plddo, 84.
//, 210.

1.

die,

Pluperfect Indicative, 215. Subjunctive, 222. 4.

A
^

96. 1.

plurimum,
96.
1.

254. 1.
1.

/, pronunciation, 26.
for
b,

plurimus, 182.
pluris, 344. ///7j, 181. 3.
30.

paenitet, with genitive, 328.

Palatal distinguished
Palatal Mutes, 94 palea, 100. 1.
f.

from Guttural,

pluvi, 210.

poena, 81.

2.

polio, 261. 4.

pandd, 107.
-panx'i, 208.

polliceor, 284.

Parasitic Vowels, 91.

Pompei, 126. Pompeius, 82.

3.

Parataxis, 366.
parjetis, 16.
1. c.

pond, 261. 4. pons, p on tis, 41.

pontufex,
6. 2.

pa?-ricida, 88. I.

parsi,

208.
4.

poploe, 81. 4; 131.

partem, 310.

populus, 91.
^?r-, 284.

Partial Assimilation, 106. 4.

parti ceps, 71. 1. Participles, formation, 248 pastus, 105. 1.

pecto, 203. III.

porrigo, 284.
f.

porta, 100.

2.

portendo, 284. portdrium, no.

poscere, 105.
1.

pedestris, 108.

I.

peior, 82. 3.

posed, 64.
349. 2.

203. VI.

pellis, 106. 3. pendere animi, pepugi, 206. 1.

/<?r,
/<??-

Possessive Pronouns, 190.

ariiniis,

possini, 218.
^atf, 282.

281.
in composition, 58./".
;

posterns, 181.

postrhnus, 182.

1.

Perfect Indicative, 200. 3 inflection, 212.

,

206

ff.

postridie, 173; 256. I.

in

-si,

208.

postmnus, 182. 2. Potential Subjunctive, 365.

potin, 108. 4. potior with ablative. 341.
1.

in -ui, 210.
in -vi, 209.

Permissive Subjunctive, 362. b. Personal Endings, 207 229 ff. Pronouns, 187 ff. pessimus, 182. 1.
;

with genitive, 330.

^r<2, 283.

praedad, 109.

1.

praestigiae, 99.

254
praeter, 283.

INDEX.
0MW, 198. ^zm, 198.
2.

prehendo, 97.

3.

B.
f.

7.

Prepositions, 260

0<?, 198. 6.

in composition, 38.

quoad with Subjunctive, 380.

quod, 198.

Present Optative, 218. stem, 201 f.

pr'idie, 173.

quoniam, 101.

1.

primus, 182. 1; 184. priusquam, 379.

pro-, 284.

I.

-quos, -quont, etc., orthography, 57. 1. 0Z"^, 9308MZTC, 57. 3.

<

pro-, 284.

R.
r, 1.
r,

procus, 64. b. prodesse, 109.

pronunciation, 18.
100.
/,

Progressive Assimilation, 106. 1; 3. Prohibitive Subjunctive, 358. d; 362.

Pronominal Adjectives,
Pronouns, 187
prope, 285.
propter, 285.
ff. ff.

199.

r from r from

99. 98. 1.

j,

-r for -J in

nominative, 98.
3.

4.

Pronunciation, 3

r-Stems, 147.

ra from
r<?-,

?,

100. 2.
1. .

radix, 104.
286.
reccidl, 206.

prorsus, 238.
prot'mus, 73. 2. a. Provisos, 405.

1.

Re-composition,
recta, 255. 3.
r<?<^-,

87. 1.

proximus,

182.

1.

pudet, with genitive, 328.

pulcer, 31. 3
;

286.
II.

76.

1.

puleher, 25.
-punxi, 208.
/#><*, 88.
1.

1. ^".
1.

Reduplicating class of verbs, 203. Reduplication, 206.

refert, 349. 3.
refert,

pulsus, 101. 1; 108.

with genitive, 329.

2.2.1.

Reflexive Pronouns, 189.

1.

pupvgl, 90; 206. purpura, 90.

regain,

regerem, 222. 3. Regressive Assimilation, 106.

Relative pronouns, 198.
relincmit, 57. 2.

2.

0,25.4; 94.
$rff,

198. 6; 255. 3.

quadra, 183. 13. quadraginta, 183.

13.

quadringenfi, 183. 15. quae, 198. 2.

reliquum est with the Subjunctive, 392. rem, 86. remltuscor with genitive, 325. remus, 89.
repente, 257. 3.

quant l, 344.
quartus, 184. 4. quater, 186. 4.
quattuor, 183. 4. quern, 198. 5.
querela, 89.

repperl, 206.

1.

Repudiating Questions, 363.

res, 180. 1.

d.

Result Clauses, 373

rettull, 206. 1.

f.

397.

Rhotacism,
1.

98.
5.
I.

quernus, 105.
0J, 198. 2.
0///V/,

198. 2.

languages, 36. Root class of verbs, 203. rubemus, 203. VII. b.

rtibro-, 97. 2. c.

Romance

qulnqiie, 73. 2. 96. 1; 183.5. qulntus, 105. 1; 184. 4.

Rumpel's theory of the cases,

301.

INDEX.
rumpo, 203. IV.
rursus, 259.
sextus, 184. 4.

255

rutundus, 90.

jz,

Shortening of Vowels, 88. origin as a conjunction, 401.

88. 3
;

.rafo',

189. 2.

siccus, 106. 2.
jJ^o, 64;
s,

89; 203. II.

pronunciation, 22. between vowels, 98.

1.

j/<f2,

2l8.

2.

siemus, 218.
silua, 16. 1. .
jz'wz,

j-Stems, 147.
-s

-s

from from

ns, 109. 3. b.
ts,

218.

109. 3. b.

similis, construction, 323.

sacerdos, 65.

simplex, 73.
JM, I90.
.H.tf<?,

2. ^.

saeclum, 95. 1. saeculum, 91.

salignus, 94. 3. salvus, 100. 1.

si?iguli, 185. 1.
3.

203. II.

-T0-,
<;.

288.

Samnium,
scala, 89.

106. 4.

soboles, 90.

satin, 108. 4.

sobrimis, 108. 3. sobrius, 288.

societas, 76. 4.

scicidi, 206. 1.
scidi, 206. I.
scilicet,

204.

socius, 94. 2. socors, 288.

-sw-class of verbs, 203.

j^, 189. 3.
.$-,

VI.

socrus, 73. 4; 103. 5. solium, 64; 95. 2.

288.

sommis, 106.

4. .

secerno, 100. 3.

-w,
;

88

103. 5.

semndum,

287.
184. 2.

ww, 73.

4;
f.

103. 5.

secundus, 103. 5 secwitur, 57. 2.

Sounds, 62
sparsi, 105.

sovos, 190. 3.
1.

j^-, 189.

3.

sedes, 62. 3.
seditio, 109. I.

spepondi, 206.
speres, 171. 1.

1.

sedulo, 76. 4.

sperno, 203. V.
jr^?.r,

segmentum,

94. 3.

172.
f.
;

sella, 106. 2.

Spirants, 21

98

f.

sernel, 186. 1.

#0\
j\r

Semivowels, 103.
semodius, no. sempitermis, 99.
j#, 185. 2.

257. 3. spopondi, 206. 1.

.W

from from

<#, 108. 1.

#, 108. 1.

septem, 102.

183. 7.

stabulum, 91; gy. 2. c. stare with ablative, 349.

stella, 106. 2.
stellid, 88. 1.

2,

septenl., 185. 2.

Septimus, 184. 5.

sepulcrum, 31. 3. sequere (Imperative), 76. 6 227. sequitur, it remains,' with subjunctive,
;
'

V. stemud, 104. 1.
sterno, 203.
jfe/J,

206.

1.
<5.

392.
j^rc?,

stlatus, 104. I.
;

75. 1

203. II.

i'/ZJj,

104. 1. .

sescenti, 105. 1.

stlocus, 104. I. $.

.s^, 183. 6.
sexcenti, 87. 3.

stratus, 100.

1.

strenna, 88.

I.

256
Strong grades of roots, 64.
stuppa, 88.
1.
;

INDEX.
ion-, -in-, 147. 2. b.

sub, 93. 2; 96. 1

289.

mon-, -men-, -mn-, 147. men-, -mn-, 147. 2. d.

mo-, -ma-, 182.
1.

2. c.

sub in composition, 58. g. Subjunctive, 200. 5.

uses, 358
in
ff.

on-, -en-, -n-, 147. 2. a.

os-, -es-,

147.

1.

of Contingent Futurity, 360; 365.

ter-, -tr-, 147. 3. a.

Dependent Clauses, 366

ff.

ff.

tero-, -ter a.-, 181.

tor-, -tor-, -tr-, 147. 3. b.

formation, 220

original force, 354. in principal clauses, 358

sui, 189. 1.
ff.

sulcus, 76. 1.

of purpose, 367.

syntax, 353 ff. Substantive Clauses, 381 ff. after verbs of ordering and com-

sum, 202. 3. sumus, 106. 2; 182.

sumpsi, 108.
super, 290.
2.

2.

manding, 384.
after verbs of

Superlative degree, 182.

begging and request-

ing, 385.
after

Supine, 252. 2. Supposition, Subjunctive

supra, 255. 3. suprad, 255. 3.

of,

362.

e.

verbs of advising, 386.

after /acid, 387.

after cave, 388.

after verbs of permitting,

surrexe, 47.

2.

grant/ fig,

sus, 171. 2.

allowing, 389.
after verbs of deciding and resolving,

suscipio, 105. 1.

390-

with opus
oportet, 391.

est,

necesse

est,

usus

est,

suspicio, 25. 3 90. suus, 103. 4; 190. 3. Syllables dissimilated,

;

no.

division
est, etc.,

of, 35.

with sequitur, reliquum

392.

Syncope, 92.
Syncretism, in ablative, 331; 332. Syntax, 295 ff.

after verbs of hindering, 393.

developed from Deliberative, 394. after verbs of wishing and desiring,

395after verbs of fearing, 396.

T.
*.
t,

95-

of Result, 397.

pronunciation, 24.

with quin, 393

f.

/-class of verbs, 203. III.

tanti, 344.
fe,

developed from Volitive, 384 ff. introduced by quo-minus, 393. developed from Optative, 395 ff. after non dubito, etc., 398.
subter, 289.
subfilis, 90.

188. 4.

fed, 188. 4.

tego, 104. 1. b.

tegula, 62. 3.

temere, 256.
temo, 89.

1.

subus, 171. 2.

temno, 203. V.

succus, 88.

1.

sudor, 103.
Suffixes,

5.

templum, 108.
203. VI.
tendo, 107.

2.

su'escb, 105. 1;

tenebrae, 108. 3.

of a-stems,

in.

of z-stems, 148.
of c-stems, 124.

tenere castris, 349. tentus, 102. 1.

tenus, 291.

I.

of -stems, 160.
ios-. -ies-.

tenuia, 16.

1. a.

181.

-r, 88. 2.

INDEX.
ter, 75.

257

2; 186. 3.

z?-stems, 160.
(5^r, 97. 2. .

ter tins, 184. 3.

th, 31. 2;

4.
;

z, 14;

83.
/. o'

Thematic Conjugation, 201

tibi, 88.

203.

ul from
lilluS,

100.

I.

188. 3.

I06. 2.

tilia, 104. 1.

tinguo, 73.
&r, 188.
-tlo-,

2. .

uls, 293. ultimus, 182.

/zV-,
zit///^,

2.
;

2.
1.
;

95.

255. 3 105. 1.

293.

zW/<?, 100. 1

203. V.

-um

in genitive plural of a-

and

c-s terns,

-tor, 88. 2.

42.

torreo, 203.

VII.
1.
1. .

umbilicus, 76.

2.

torrere, 106. 3.

tortus 105.
torus, 104.

umbo, 76. 2. umerus, 23.

iimor, 23.

A*, 93totondl, 90.

/cc/<?j-,

, 255.

3.

-undo,

51. 5.

190. 2.

trans, 292.

-unculus, 51.5. uncus, 76. 1.

trans, in composition, 58. .

unda, 107.
unguis, 76.
1.
;

transduco, 87. traxe, 47. 2.

3.

Unthematic Conjugation, 201

ZWZ2W, 183. I.
-?z<?.y,

202.

trecenti, 183. 15.

tredecim, 183.
#-r, 183. 3.

n.

-uom, -uont,
105.
I.

etc., 57. 1. .

zzrzza,

tribubus, 168.
triginta, 183. 13.
zVzzzz,

-urn us, 51.

-zzj-

3.

in genitive singular, 138.

185. 2.
1. .

-usculus, 51. 5.
-ustus, 51. 4.

zWo, 104.

triu?npus, 31. 3. /, 188. 1.

utor with ablative, 341.

1.

tugiirium, 90.
taz, 188. 2.

V.
v, v,

-tumus, -timus, 182. 2. Tuscus, 105.

tazzj,
1.

pronunciation, 16. changes, 103. 3; 4.

valde, 92.

103. 4;

190. 2.

variego, 71. 6.
97. 3. A. Velar gutturals, 94.
t/<?/$<?,

U.
2/,

1.

pronunciation,

8.

velim, 218.
velle, 106. 3
;

K, 78.

243.

from a^, 103. 4.

vellem, 222. 3.
veneficus,

from u from m from

, 79.

<?<,

103. 4.

no.
103. 2; 203. VII. a.

a, 71.
<?,

3; 4.
;

venio, 101. 1;

76. 1

2.

venire, 94. 2. ^. ventum, 106. 4.

from au, 84. 2. a from eu, 85. z7 from <?z, 81. 1. u from ozz, 85. u cbnsonans, 16. 1.
K-stems, 171.

Verbs of judicial
versum,
-us, 76. 3

action, 327.
;

a.

294. vertex, 76. 3. vescor, with ablative, 341. vesperi, 256. 1.

vester, 76. 3
;

I.

190. 4.

258
vestri,

INDEX.
vestrum, 188.
1.

6.

void, 202. 6.

veto, 76. 3.

voluntarius,
volvo, 73. 5.

no.
2.
<5.

vhevhaked, 206,
vias, 113.

vorarc, 94.
3.

vicem, 310.

ww,

188. 5.
1. .

vichii, 185. 2.

vlcesimus, 184. 8.
vicus, 81. 3.
videlicet, 204.

-zw, -vom, -vont, 57. vostri, vostrum, 188. Vowel gradation, 62

7.
f.

viderimus, 219.
z//J,

Vowels assimilated, 90. Vowels shortened, 88.

212.
4.
.

vidimus, 108.
viduus, 97.
viginti, 73. 2.

W. Weak
grade of roots, 64 f.

2. ^.
<5;

183. 12.
1.

vincere pugna, 349. vinclum, 95. 1.

X.
x, pronunciation, 32. x, origin of the letter,
1. 2.

vinum,
z^j,
'

81. 3.

virile secus, 310. 2.

thou

wilt,'

202. 6.
y, pronunciation,

Y.
9.

vitulus, 91.

vivus, 94. 2. ^. z-^jj, 188. 7.

y, origin of the letter, 1. 5.

Vocativus, 297.
z/<?/-,

orthography of words

beginning
z,

with, 57. 1. a. Volitive, 354 358.

;

pronunciation, 33.
origin of the letter,
1. 5.

2,

Date Loaned

\;

uu

BOSTON UNIVERSITY

1719 02815 8204