The Rose

of the



and development

of the Compass-Card


Silvanus P, Thompson, D.Sc., F.R.S.


at the International Historical Congress, April, 1913,
the Proceedings


of the British Academy, Vol.

Published for the British Academy

By Humphrey

Milford, Oxford University Press


Corner, E.G.
Shillings net




D.Sc., F.R.S.

at the International Historical Congress, April 5, 1913.

This inquiry embraces three chief points : I. The origin of the Names of the Winds.


The The

origin of the arrangement of the Rose of Thirty-two Points.

origin and significance of the distinctive Cards.

marks used on Compass


this inquiry,

the origin ; The mariners' compass, as we know it to-day, consists of a light circular disk or card, beneath which is attached a magnetic needle or

the construction of the compass lies outside the scope of some preliminary considerations are necessary concerning and these must be stated briefly. of the compass itself

system of magnetic needles. The card is provided at its centre with a small cap, by which it is poised movably upon a pin. The whole is enclosed in a hollow box or bowl covered with a flat lid of glass

and the compass box or bowl
of brass to enable

suspended within two hinged rings to conserve its proper horizontal position in spite

of any tilting movement to which it is subjected by the rolling or pitching of the ship on which it is carried.
' divided out into thirty-two ( points or c rhumbs of equal angular breadth forming a rose or star, and to these are affixed the initials of the names of the thirty-two points. The magnet





needle, or system of needles, is affixed to the card parallel to the direction marked on the card, so that when left to itself, the card, obeying the directive force which acts on the needle, sets itself


in a direction pointing (magnetically)

North and South


the several

points (or, strictly, pointers) marked on the card then indicating the several directions that the mariner may know in which way to
steer in order to follow his desired course.


The card


also usually


at its

North point with a

fleur-de-lis or other distinctive sign.

The magnetic needle
of the earth's globe.

controlled in

pointing by the magnetism

2 N-l

power of the lodestone was known to the ancients. (see Fig. 3. certainly which have yet been identified as of earlier date than compass-cards 1500 bear a fleur-de-lis and those of them which bear initials carry only the initials of some of the eight principal winds. The fact that this attractive power could be communicated to iron or steel by the lodestone was known to Lucretius and to the Chinese. . 7. causing it to point in times. : The attractive as appears 2. and whether in them the North point bore a only six of their points fleur-de-lis. Plate IV). as above described. The windroses . As will be were so marked and almost seen. Neither is there here due to the any question considered as to the errors or perturbations iron on the ship. or worse etymology. The compass. and it was independently known to the Chinese. in la Cosa. and the invention of the compass has been claimed for many different It may therefore be worth nations on the most absurd grounds. on the pilot for Mappamundi drawn in 1493. as to the means for compensating those perturbations or correcting those errors. in all essential particulars by Columbus in his voyages of discovery in the West Indies certainly that it is doubtful possessed every one of the features enumerated. the truth of errors arising established 1. who claimed for that knowledge . has remained practically unchanged The compasses carried for four centuries. probably For none of the there was no fleur-de-lis on any of them. The directive action of the lodestone and of the piece of iron or steel which had been touched by the lodestone.2 It is PROCEEDINGS OF THE BRITISH ACADEMY no part of the present paper to enter upon the reasons for the irregularities which are found to exist in the directive force of ' ' the earth. from the writings of Plato and of Euripides . nor. again. was unknown in Europe in classical was known to the Chinese. Flagrant from bad copying. save whether all or any of their thirty-two points were marked with initials. is more perplexing to the student of the history of the than the unhistorical way in which the chroniclers have compass transcribed and transmitted erroneous statements and unfounded Nothing which should have been tested. and which produce those local variations of the compass which the mariner encounters everywhere. while to set down a few of the more important points as now assertions. nor as to of presence near the compass of pieces those due to the incidental presence of magnetism in the iron of the ship's hull or fittings . have been sedulously propagated in the interests of mistaken patriotism . but the north and south direction. who was Columbus have no 1500 by Juan de fleur-de-lis and no initials.

after 1500. by the narrative of Cardinal de Vitry (1219) . quoted by Bongars . a triangle. The first Levant and throughout the Mediterranean. D. 1269. Practically without exception. weather to the e Stella maris '. 16. . In Europe. by Brunetto Latini and by a reference (1263) in Las Siete Partidas of Alfonso X. primitive form of mariners' compass was a floating the needle being supported on a reed or a cross of reeds apparatus. 139. with a glass lid. is that of Pierre de Maricourt (Petrus Peregrinus).C. the magnet was certainly used in navigation in the latter half of the needle at sea before twelfth century. by the maritime laws of Wisbuy by the letter to Pope Honorius (1218). states that sailors resort in cloudy 7. and the various stages in the . provided with a divided scale. and with ( sights ' for taking bearings. a star. The . the compass-cards or windroses used in mariners 5 compasses were from the beginning divided out into 8. (1260) . as attested by Alexander Neckham (1157-1217). 8. The is application of the directive action of the needle to navigation of later date. as also in the Baltic. by by a fleur-de-lis. 121. growth of that myth can be traced. In this form it was used in the 6. though Klaproth regards that date as mythical. is A. Roger Bacon in his Opus Minus (second half of the thirteenth century) . 1297 . and in 1680 were still using a floating compass. The east point was often marked with a cross. That it was in use in the twelfth and thirteenth centuries is attested by the ancient poem of Guyot de Provins (1190). upon the surface of water in a basin. and in all probability at the port of The earliest description of a pivoted Amalfi but the attribution (by Mazzella in 1586 two hundred and eighty years after the event) of this step to a mythical personage called Flavio di Gioia is unhistorical. a group of stars. or. 9. The first authentic mention of the directive properties of the lodestone and of the needle 4.D. In or about the year 1302 the improvement was made of affixing to the pivoted needle a light card whereon was painted the rose of the winds. They had no pivoted compasses before A. the North point being distinguished a dart.. who describes the needle as mounted on a dart.THE ROSE OF THE WINDS 3 the date 1100 B. it The was introduced by Western navigators in the sixteenth century.D. by Bai'lak of Kibdjak (1242) by Hugo de Bercy (1248) . It is doubtful whether the Chinese used the magnetic 5. Such evidence as exists points to this improvement having been made in Southern Italy. and 32 equal divisions . application of the directive action of the magnet to travelling on land was known to the Chinese from A. magnetic needle placed in a box. a trident.

Trucraf 8' aeXXa? iravroioiv dvcfjuov. Eupoy. It is for in 8 Od. as in miners* dials. four separate ' winds are mentioned. and South. Daniel viii. and Rev. vii. in Acts xxvii. re. Scholars have written much on the winds recognized in antiquity. xxxvii. known prior to NNE. North. 9 . and xi. and opticians* compasses. and on the names assigned to them. Paul. West. : . 8. of question of the number by name: Boperjs. Mark xiii. The rhumbs of the eight principal winds were usually painted black. xxiv.. the majority advised to depart thence. &c. while the rhumbs of the sixteen quarter-winds were commonly painted red. Long before the invention of the mariners* compass the ancients had adopted the arrangement of the names of the winds symmetrically around a circle representing the horizon . xlix. 31 . Apparently in primitive civilization the peoples reckoned but four winds. V. No examples of this use are.. Homer recognizes only are the of names the winds. The passage may be thus translated And because the haven was not commodious to winter in. surveying instruments. seems to have arisen with Flemish navigators. ORIGIN OF THE NAMES OF THE WINDS. 12-14. Ezek. 7 .. sometimes they were painted alternately blue and red. the rhumbs of the eight half-winds being usually painted green 10. In the windroses of thirty-two points the rhumbs of the eight principal winds were usually black . The same number is mentioned in the four 1 New 2 Testament. . ZiQvpos. such an arrangement conThe so-called ' points * of the compass stituting a Rosa ventorum. namely those that blew from the four quarters of the horizon East. (tyvpos vorof re KO\ ftopfrjs aldpriyevcTrjs /xeya Ku/ia Kv\iv8otv. 295. 2w> 8' ftpos T' en-ctrf. and IVoroy. 1536. occasionally blue or gold .4 PROCEEDINGS OF THE BRITISH ACADEMY six other of the eight principal points were almost invariably marked with the initials of the Italian names of the winds. 4. appears to have originated with Stevinus of Bruges about the year 1595. The placing of the compass-card in a fixed position beneath the needle. about which there has been considerable confusion. see Jer. 36 . but was early adopted by the Portuguese and Spanish. 1. on compass-cards. Four winds only are recognized in the Old Testament . however. In the account of the voyage of St. in Matt. 11. The use of the initials of the Frankish names of the winds N. NE. remarked by Aulus Gellius that Homer recognized no more than four the description of the storm he says that all the winds conspired : .

being some forty degrees northward of East at midsummer. it being a haven of Crete lying towards Libs and Choros. But such names as Euros. Auster. Some of the names are clearly directional .' In the Vulgate the names are given as Africus. and Kaikias are proper names denoting personified winds which seemed to have special properties. or Thrakias). thus. one malevolent and the other beneficent. obviously refers to the direction in which appeared the seven circumpolar stars . loosing thence they sailed by Crete. as well as the Italian Levante and Ponente. local The last-mentioned is a purely name for a North-East wind. though used for the North wind (in place of Boreas. and the name Meridies (in place of Notos or Auster) sun. Zephuros. to lie up . But the same term became used for the North-East wind in Spain. The Euraquilo of Crete. example is afforded by the mediaeval use in the Italian Mediterranean Some name Greco for the North-East wind which blew from Greece. supposing they had gained their purpose. of the where certainly it did not blow from Greece. . 4) says that Boreas and Zephuros blew from Thrace . and whether his Euros was opposite to his Zephuros. in different localities there were found prevalent Moreover. In many cases there seems to have been no clear distinction between the name of the wind and the name of the geographical direction from which it blew. and Euraquilo. Homeric scholars have. For seeing that a cold East wind in spring and a mild East wind in autumn were looked upon as two different personages. and these frequently received specific names. Auster. favourable if by any means they might reach Phoinike. Aquilo. and it appears doubtful whether his Boreas was directly opposite to his Notos. the name Septentrio. and the Skiron of Athens are examples of the names of purely local winds.THE ROSE OF THE WINDS 5 Late Greek writers mention other winds having many different names. found some difficulty when comparing For instance. But not long after there arose against it a typhonic wind called Euraquilo. Arctos. however. and were personified after the manner of Greek thought. and similarly with the shifting position of . And when Notos blew softly. are obviously directional. two different names would be allotted to them. The East and West points seem not to have been regarded as fixed in the same way as the North and South For the position of the points. Homer in these with the usage of Homer and Hesiod. of the local or topical names given to winds found wider and spread into regions far from the locality of origin. The is essentially the name of the direction of the midday names Oriens and Occidens. sunrise varied with the season of the year. and some forty degrees southward of East at midwinter sunset. An usage. the Iliad (ix. winds which did not seem to blow from any one of the four cardinal points of the horizon . Corns.

and does not imply or qualify the direction of the wind. while that from the place of winter f Exortus et called 'ATrrjXitoTTjs. Thrakias and Meses. with their appropriate geographical topical or unfavourable. Probably the most consistent account of the views of antiquity is He was confronted with the fact that Notos that given by Strabo. and the names directions. (or his Editors) course of development. The equally divided rose of eight winds is a later conception. Hippocrates. also that 8vcrarj$ (impetuous) is not a separate westerly wind. and Theophrastus. three winds from the East and three from the West. remarking occasus mobilia et varia sunt. We find other opinions in other writers. that Euros and Apeliotes were used for winds from the East. have not always been successful . meridies septentrionesque statu perpetuo stant et manent/ Of the four cardinal winds he regarded two as sunrise is : fixed and two as movable.6 PROCEEDINGS OF THE BRITISH ACADEMY Attempts to identify the names of special winds. minds. making a northern triplet. To the two fixed winds. of winds. out that dpyeo-Trjs (whitish) is merely an epithet of NOTOS . That the wind which blows from the place of summer sunrise is called Eu/>oy. he placed two winds. one (subsequently known as Libonotos) unnamed. Anaximander. there were now apparently added two sets of triplets. and the developments were conflicting and even contradictory. North and South. Hippocrates adopted six as the number of the winds. Varro. at the side of Boreas. but some confusion exists as to their directions. but they were not equally spaced out in their directions. In Aristotle we find the arrangement of twelve in the North. and room had to be found for more in the disposition of them around the circle of the horizon. making eight winds in all. but afterwards he named Phoinikias as being placed opposite to Thrakias. and Argestes were used as names of winds blowing from the North . and Here he points Zephuros and Dusaes for winds from the West. On still leaving four triads so evolved are At first these had no names. but is an attribute of Ztyvpos. The : Thrakias (Kaikias Northern Aparktias Boreas Eastern \ Apeliotes (Euros (Libs Southern [Euronotos] Phoinikias Notos [Libonotos] Western \ Zephuros [Argestes . He also framed a southern triplet by placing two additional winds at the sides of Notos. But eight winds did not exhaust the number which had received definite names. and the imperfect state of knowledge of geography prior to the fifteenth century added to the To fit such winds into an ordered system occupied many confusion. he accepts.

C. Whereas the Aristotelians had viewed the winds as issuing from different points. who adopted it in his Portulan . Lips. Africus. the Northwest Corus. named Aparctias and Boreas 3 by the Greekes/ : principal For a later example of the unequally-divided rose of twelve winds. and Libs. riseth the South wind [w4wster] : winter the Southwest. were of equal width . but this rose never found acceptance either on land or at sea. Therefore every quarter of the heaven hath two winds apeece. 1 . From the equinoctiall from the rising sunne-rising bloweth the East wind Subsolanus both : : thereof in Mid-winter. To this Winds at Athens. In the meantime Eratosthenes had modified the scheme in St. 8 The Historic of the World. Euros. Pliny. refers to the schemes of four and twelve winds. p. and adds that ( the Moderne sailers of late daies. four and eight collateral. Suetonius. 1 attributed to Timosthenes. Niirnberg. and Skiron. Philemon Holland's Translation (London. each adorned with an emblematic figure in relief. the Southeast Vulturnus. however. Apeliotes. erected about system the fundamental witness is the Tower of the 100 B. foure winds and no more. equally divided into twelve. 22. founde out a meane betweene and they put unto that short number of the first. is known as the Rose of Timofind references to it in the writings of Varro. From the North-waine or pole Arctike. and he had simplified the arrangement by fixing the number of spaces as 2 eight. but from that in Summer season. and others. Zephuros. Eratosthenes assigned to them definite spaces or regions. and the arrangement of the circle of the winds. though with latinized forms of the names in Western Europe. see Schoner's Opusculum Geographicum. sthenes. most frequently other of his immediate disciples. by Andronicus Cyrrhestes. in his Natural History. is the Northeast wind Aquilo. and by the same Greekes they are tearmed Zephyr us and Argestes. bloweth the North wind Septentrio betweene which and the Sunne-rising in Summer. Notos. is set forth in the writings of Alexander Aphrodisiensis and It is. the Greekes call Apeliotes. the admiral of Ptolemy Philadelphus. down to the end of the Middle Ages. according to Vitruvius. as follows Boreas. We Vincent of Beauvais. which gives the names in Latin and Greek. Kaikias. which. 1533. : It has eight flanks facing the eight winds. 2 Vitruvius also gives a diagram of twenty-four named winds . Isidor. the West wind Favonius commeth. which they tooke out of the later.THE ROSE OF THE WINDS / 7 and This arrangement of twelve unquestionably goes back to Aristotle. two ways. 1601). and the later Eurus. From the equinoctiall The former of these From the mid-day and from the sunne-setting in MidThey also name these two Notus going downe of the Sunne.

34. . of Orontius Finaeus. narrates how the beginning of the ninth century. and continued through 1 Eginhard. however. De Mundi Cosmographia. and is found in Seneca and Margarita writers. in a magnificent Manuscript which unfortunately perished in the fire of the Library of Strassburg. Vita et Qesta Karoli Magni. such a multiplicity of names. The northern nations had never possessed. down to the Philosophica of Reisch in the sixteenth century. as used throughout all Europe to-day. and the tradition is recurrent that the actual denominations of the thirtytwo points. The same scheme appears again sphere in a Latin Manuscript in the fourteenth century on a planiin the Library of Arras. at combined them two-and-two winds. was repeated in the Hortus Deliciarum of the Abbess Herrade de Londsberg. that monarch. 1555. \Eginhard. in the eleventh century. The Flemish mariners.8 PROCEEDINGS OF THE BRITISH ACADEMY The rose of twelve winds continued to later be used. The Latin rose of twelve winds had found acceptance throughout the Roman Empire from Egypt to Spain. and particularly those of Bruges. but with further variations. using the Latin names. in the siege of 1870. The Sphera sive stance that the making-up of the names of the subsidiary points by combining the names of the four cardinal winds imposed the adoption of monosyllables for those four. gives two windroses of sixteen points with the Frankish names . finding in the Teutonic tongues the names of the four cardinal points only. were framed by the pilots of Bruges. the form of which suggests that by that time another Frankish adaptation to sixteen winds had already become Paris edition. as follows : in such a way as to denote the twelve \ (Circius Septentrio Nordwestren Nordren Nordostren (Aquilo Subsolanus Eurus (Vulturnus Auster Austroafricus (Euroauster Ostnorden Ostren Ostsundren Sundostren Sundren Sundwestren (Africus I Zephyrus (Chorus Westsundren Westren Westnordren ^ t r This notation. with some slight changes. 1 in his life of Charles the Great. p. and the author comments on the circumcurrent. early adopted these Frankish terms .

Nevertheless. BRUNETTO LATINI Tramontainne Grec Levant Siloc AZAPH Tramontane* Graecus Oriens Sillocus (or Imber) Meridies Midi Aufrique (or Garb. a transition from the rose of twelve to the rose of eight winds. It reappears in the famous Mappamundi in the Cathedral Library of Hereford. using Levante for Oriens. and that of Sillocus for the South East wind.THE ROSE OF THE WINDS 9 the Middle Ages. namely. but with the accompanying gain of a more minute secondary and tertiary subdivision. and in the work of the same name by the Venerable Bede. or Lebech) Couchant Maistre Garbinus (or Lebex) Occidens Magister Here we meet with the arrangement destined classical satisfied twelve-point nautical needs. for eight. and the names adopted by these two writers may be compared. In this list of the winds of Azaph we meet for the first time with two new names. The Cosmographie of Azaph. eight that of Magister for the North West wind. Le Bex) continued also as a synonym. The occurrence of these two names fixes the or even . the Italian form of the classical Auster. exhibits this transition in process. while for Garbin or Lebex (which appears to be of Arabic origin) they more commonly used Africa. that to supersede the because it better which. won its way amongst the seamen of the Mediterranean. with its subdivisions into sixteen and thirty-two. Isidor. about the end of the eleventh century. It appears in the De Natura Rerum of St. which Eratosthenes had effected in the system of was accompanied by a change too in the current names. and in the De Proprietatibus Rerum of his contemporary Bartholomeus de Glanvilla. It is repeated by Vincent de Beauvais and by Albertus Magnus. in the Polychronicon of Higden (1357). Most of the fifteenth and sixteenth-century editions of the Geographia of Claudius Ptolemaeus give the rose of twelve winds. The transition by the interposition of was in progress. They modified the names by substitution of vernacular equivalents. cited above for an example of the rose of twelve. It occurs in the Cosmographie of Azaph. it gives also a planisphere with In the somewhat later Livres dou Tresor of Ser Brunetto Latini the' same simplification appears. For Meridies they adopted Ostro. half-winds and of simplification Aristotle was to be repeated. rose. The same kind quarter-winds. and Ponente for Occidens. though Libeccio (or Labetes.

Dechales^ Riccioli. Africa. 1588). 250) points out at some length the differences in usage between Southern Europe and Northern. 1545) bear the names or initials of the Flemish winds. Flemish designations appear also in the windrose eight. with its Italian names. Flemish names are exclusively used in the Trattado da Agulha de marear written by Joao de Lisboa in 1514 (printed 1903 at Lisbon).10 PROCEEDINGS OF THE BRITISH ACADEMY in the Italian Mediterranean. In the Mappamundi of Marino Sanudo. and Tramondoes tana. which reappears in Silocco. twelve. there are twelve several stanzas in ottava in winds. the Flemish names of the winds are given. 1674. 1556) and Pedro de Medina (Valladolid. on the Pisan charts . In Italy. A rose of thirty-two points . the eight the date of this transition from the twelve-point rose to the eightafforded by the is circumstance point or sixteen-point rose while in the poem (middle fifteenth century) entitled that A La Sfera. though the lastnamed also gives one with Italian initials. in the Lexique Roman of Raynouard . in his del Arte de navegar (Seville. of Joannes Maria Tolosani. in 1514 at Florence. Dechales (Cursus Mathematicus. In the Arte de navegar. and Fournier give elaborate tables of the names of the winds in various languages. T. in the chart of Ruiz de Estrada e Pefiato (1525). Martinus Cortes (Seville. in the Ars Magna of Lully . or subsequent windroses and charts under many forms Sirocco. Xaloque and Saluq in certain Arabic sailing-charts. of Gregorio rima recounting by name the kindred poem called La Nuova Dati. Such then was the order of developments which issued in the What the influence eight-point windrose. of 1321 (now in comment on names are given in Italian. 1545). just as surely as origin of the system the occurrence of the other terms Greco. Siloc. The drawings of compasses given by Nunez (1536). of the Crusades was in bringing about its adoption is a matter for conjecture only. and thirty-two points. the name Magister for the favourable North West is preserved to us in the familiar form of Maestro. p. in Brunetto Latini . Roderigo Camorana. : for the South East wind. and are displayed variously in roses of four. Issalot. ii. is the familiar Scirocco. Brussels). Sillocus. of Pedro de Medina (Valladolid. four stanzas similarly reciting the published names and virtues of the sixteen winds. Exaloch. De Lanis. uses always the Compendia Flemish names. The Jesuit writers of the seventeenth century Kircher. It remains to be remarked that the Portuguese mariners at an early date seem to have adopted the Flemish designations of the winds in preference to the Italian. on the Catalan Atlas . and throughout wind Southern Europe. Laxaloch. we find the Sfera.

27. 1585) gives one planisphere with Antwerp names in edition of 1533. large and small. the four principal ones being within the periphery.THE ROSE OF THE WINDS drawn by Willebrord Snellin gives the full names in Dutch his : 11 Tiphys Batavus (Lugd. Mappamundi in the Cologne edition of Macrobius. The Arte de navegar of Pedro de Medina (1545) gives a planisphere surrounded by eight cherubs' heads. A depicts sixteen cherubs. usually depicted around the planisphere map by cherubs' heads blowing. Noort-ten-oosten. Noort-oost. The Leuanrc Poneotc CECCO D'ASCOLJ. and the twelve auxiliary ones grouped in triads outside. shows twelve cherubs' heads with four cherubs. Batav. Noort-oost-ten-noorden. The Cosmographia of Barocio (Venice. another with thirty-two. of 1521. Noort-noortoost. &c. The various editions of the Cosmographia of Pietro d' Abano (Bienewicz) are rich in such.. another with twelve. fol. . three languages. In the works of the contemporary cosmographers we meet with or many diagrams of the winds. 1624) Noort. 1521.

Nufiez. names ORIGIN OF THE ARRANGEMENT OF THE ROSE OF THIRTY-TWO POINTS. (See page 11. in Bibl. 2. since the time of Carolus Magnus. 1651. * soon as the compass was used for sailing. De Ventis. impossible to sever the develop- 2 Chaucer. Between the eight rays are eight cherubs' heads blowing. 2 3 4 The last-named says Schoner. and Riccioli. in the Prior to the introduction of the primitive compass. 3 4 Nunez. Division into thirty-two by the intercalation. xvi. and. On the other hand the division of each quadrant into ninety degrees. of eight half-winds and sixteen quarter-winds. Norimb. between the eight principal winds. when out of sight of land. As authorities for this statement may be cited Chaucer. with far greater precision than before. The Astrolabe. seamen Levant and in the western part of the Mediterranean could not direct their courses. the introduction of the compass the sailing of direct courses became comparatively safe and habitual . (2) that their primitive maps were exceedingly crude. nautical chartographers from the thirteenth century onwards preferred the division into eight.' the delineation of their land-maps. the division of the horizon into eight or twelve winds was found altogether It is certain that so insufficient for the purposes of navigation. 31. of the same date. . Schoner. This circumstance undoubtedly accounts for two facts : (1) that their navigation was at any rate in doubtful weather largely confined to With coasting routes . sixteen.12 PROCEEDINGS OF THE BRITISH ACADEMY edition of the The Milan Acerba of Cecco d> Ascoli (F. Bononiae. 1390. Pt. 1533. as proposed by Pierre de Maricourt (Petrus Peregrinus) in 1269. contains a remarkable cut the earliest known professed of the Italian eight-point picture of a compass in any printed book the having the usual initials and cross marked upon its rays. that the division into thirty-two winds has been exclusively expressly employed. Opusculum Geographicum. . of sixteenth century Riccioli. MS. by successive bisection of the angles. was too fine to be practical. and of the winds printed opposite each in the margin. 1 It is. II. indeed. cap. they were able to delineate. and thirty-two. de Stabili). with any precision.) rose. afforded the most natural way of attaining the Whatever the plan retained by geographers for requisite precision. with the knowledge of the direction in which their courses were set. Nat. Lib. by chartographers and 1 seamen. the outlines of the coasts in the sailing-charts which almost suddenly superseded the older forms of map.. Almagestum novum. Paris.

now in Madrid. and go back to the Byzantine Stadiasmos of late example. and other primitive forms. They are all executed on parchment. distinguishing them from all previous kinds of map. straight courses can. or English. the scale at the margins being from that at the centre. naturally. most of them executed by Italian chartographers . In mere coasting. starting usually from some central point. and they present a number of features in common which distinguish them from all other maps. The more closely one regards the question. so small a part of the whole globe as the Mediterranean region is all that is represented. laborious compilations of distances and landmarks along the routes. and in this case the distances are distorted. of real sailing-charts. which becomes covered as with a web of lines in a circle fine lines. presenting a system of rays. in which the directions of the coast lines are often ludicrously wrong. planispheres. is the presence of the so-called loxodromic lines. in the fourth or fifth century. Great Rutter. the important thing to know is the distance from Angular port to port. and distances are of secondary importance. hence the term loxodromic. direction is everything. They all have the loxodromic lines. only be true upon the assumption that the projection of the map is a true projection on indicate a plane (as in the method of projection different known as Mercator's). mappaemundi. These lines are almost without exception thirty- . But in sea-sailing the distances are of relatively little moment. and uniting a subsidiary series of centres arranged the map. characteristic thing about them. were in abundant use before the advent carefully stated. the more insistently is one convinced of the immense changes brought about by the introduction of the compass. When. A Totally distinct from all the earlier maps. while the distances from port to port are The The Portulani. the scale of distances will not vary greatly between different parts of the same map. around These are supposed to indicate direct routes along the recognized That they directions of the winds. or haven -finding guides. and almost invariably on single skins. of Robert Norman. printed in London in 1590. and the loxodromic lines will practically represent The vast majority of the sailing-charts in question straight courses. are the sailing-charts which made The most their appearance at the outset of the fourteenth century. of this kind of treatise is The Safeguard of Saylers. are of Italian origin. to direct the mariner from port to port. character of such charts and maps as have come down to us from pre-compass times may be judged by the typical examples of the maps accredited to Ptolemy.THE ROSE OF THE WINDS ment 13 of the compass from that of the sailing-chart. however. and the aspects of the various headlands.

Petrus Roselli (1466. one near Bilbao. near Carcasonne .. lions. the network of lines. Mappamundi of the Medicean Atlas (1351). at isthmus of Suez Giovanni da Carignano (early fourteenth). the sixteen quarter-winds red. Santarem. nor of Rome. follows : Pisan Chart (end of thirteenth century). at least in the charts after : 1375. to the . Colombaria. 1 some the cities are indicated by convenWild animals. Florence). gaily coloured . or by figures of eastern potentates. one in Spain. and give of inland places Nor is there any fixed size of radius for Most of these charts depict the seaboard almost the names of places on the coast. Bibl. have thirty-two lines and these unquestionably were prior to 1331. p. nor of Naples. later and reindeer adorn the outlying regions of the maps. and at one or more subsidiary points serving as centres for other systems of loxodromic lines other roses or rosettes are frequently On comparing chart with chart it is seen that these loxopainted. the other south of Rome. in Sardinia . and in tional castles or churches. Jachobus de Zeroldis (1446. vague mark political divisions by heraldic banners or devices. between Palermo and Gaeta . Naz. elephants. Corsini. but rarely the names nor do they give many inland features beyond . in the Catalan Atlas (Paris) 1375. exclusively. Pizigani (1367). Archivio di Stato. one in the Aegean. near Toulouse . Giacomo Bertran (1482). Florence). in the sea . to north of Sicily . Diego Homen (1558). east of Cyprus . one near Naples.14 PROCEEDINGS OF THE BRITISH ACADEMY two 1 in number. 275 of his Essai sur I'histoire de la cosmographie et de la cartographic pendant le moyen age. south of Rome. Soc. Florence). 3 no meridian of Greenwich. the eight halfwinds green (rarely blue) . Paris). the other near Sardinia Petrus Veschonte (1311). Catalan Chart (early fifteenth. bears. . another south of Naples . Chart of Badia di Cava (circa 1492). 2 These are the colours used in the Perrinus Vescoute Chart (Florence) 1327. south of Marseilles . Perrinus Vesconte (1327). The later charts indications of mountain ranges or forests. 2 At the centre of the network of loxodromic lines. Florence). not even of Jerusalem. there is usually depicted a decorative rosa ventorum. Guillelmo Soleri (1385. another to the north of Spain . off Cape Finisterre . Florence). The sailing-charts of the priest Giovanni da Carignano and of Petrus Vesconte. . very rarely eight or sixteen. The lines are invariably drawn in colour the eight principal winds black . Bibl. dromic centres are differently chosen in different maps. camels. Nat. there being no fixed point. in the Levant. both of Genoa. in the sea Baldasare di Majolo Visconte (1583. erroneously states that there are no charts with thirty-two rhumbs upon them before the Pizigani chart of 1367 . i. vol. to east of Sardinia . the same author (1570) in central Italy. Gratiosus Benin- casa (1467). between Sardinia and Sicily . . . Bibl. and in practically every one of the hundreds of charts of the period down to 1600 which have survived. two centres. Catalan Atlas (1375. between heel of Italy and Greece . and he argues fallaciously that Italian and Catalan sailors had never undertaken navigation on the high seas before the Portuguese expeditions to the Canaries from 1331 to 1344. Pinelli (1384). ' The principal loxodromic centres in some of the best known charts are as .

that ship on a chart. the only exception appears to be in certain charts of Grazioso Benincasa. like a series of little bays. their individuality. Louis against Tunis in 1270.' The use of such loxodromic charts is recorded at various dates. Thus the chronicler Nargis. that is at the West. have the North at the top. in some such form as this : Guillmo Solerij ciujs Maioricaru me fecit Ano Ant. The question has been debated whether in these loxodromic charts the lines were laid down first. and the decorative features received remarkable development. and the coast lines then drawn in over them . the lines being drawn. Theobald Fischer that the chart was drawn over the lines. The coast lines in the earlier maps. preserved at Palma (Majorca). of 1439. but no map. concave toward the sea. Even for a century after the invention of printing all the sailing-charts were drawn by hand . he wrote on it with his own hand : large geographical skin Amerigo Vespucci has paid 130 gold ducats. They were highly prized by navigators. who sailed in the Crusade of St. The chartographer's signature is usually given at the left. was shown by the captain the position of the Vasco da Gama mentions. formerly in the e For this possession of Vespucci. or this Jachobus Vesconte de Maiollo composuit hanc cartam in Janua anno : domini MDLXV. being anxious after a storm to know where they were. tells how the king. and are most frequently drawn in a peculiar manner by joining together little arcs of circles of different sizes. covered with loxo- . The hand of the type current inscriptions are mostly written in a Gothic in the fourteenth century. No such charts* appeared before the latter end of the thirteenth century. Mccclxxxv . of 1548. who seems to have worked by hand upon a printed or The chartographers appear to have jealously guarded etched outline. having fifteen charts. Nordenskjold maintains that the lines were drawn over the chart . 1 Blundevile. as is shown by the circumstance that on the chart of Gabriel de unaltered the lines Vallsecha. gives a blank chart. though the draughtsmanship of the coast shows great improvement. the pilot whom he engaged at Malinda to take him to India. Dnj. or whether the map was first drawn. had a map of India e without compass lines '. howunfinished. Most of these charts. in 1498. now in Florence. with some surprise. and from that epoch to the end of the sixteenth century they preserved almost same features. 1 ever. Kretschmer cites the case of the Atlas of Vesconte de Maiollo. though not all. in his Exercises (1594).THE ROSE OF THE WINDS 15 Islands are often strongly coloured to distinguish them the more are depicted with a precision never attempted readily. be concluded that for all charts the lines were drawn first. and the loxodromic lines added afterwards. of which one is It must not.

until about the year 1655. and found a discrepancy between them. The differences observable. working out their maps from observations made at sea. The fact of a declination gradually became known . In the fourteenth century at that the declination of the needle was unknown . contrasting notably with the crude inaccuracies and variations of all the maps of the antecedent period. but that in chart. But no such supposition is admissible . That discovery. and was eastward him both Genoese and Flemish compasses. * Columbus in his great voyage of 1496 took with . while the Flemish compass showed none. Paris. The Flemish pilots probably had adopted the practice. and.. usually attributed to Columbus. many cases the sixteen red lines of the quarter-winds earliest were added after the outlines of the coasts had been drawn on the On comparing together some of the is Mediterranean struck with the extraordinary accuracy of the coastal features which they depict. Gilbert in 1600. which he calls the shape and figure of the first liniaments of the Mariner's carde. probably being independently observed by pilots and attributed to local error or faulty compasses. and proceeds to instruct the reader ' how to set down the places of the land or sea therein '.16 PROCEEDINGS OF THE BRITISH ACADEMY the that The writer after examining many of the original charts in museums of London. &c. 2 seems to have been more or less known to others. of setting the needle slightly obliquely under the card to compensate for the declination occurring in that region. and usually also the eight green lines of the half-winds. 1 In the middle-Mediterranean in the thirteenth century the declination was westward. for example. It would charts one almost seem as though there had been an accepted standard map of the Mediterranean which successive chartographers had adopted and reproduced. afterwards general in certain ports. unknowingly. dromic * So long ago as 1624 lines. But the fact that the declination varied from place to place was quite unknown. Milan. It probably constituted the ' s secret which Sebastian Cabot when dying declared to have been a divine revelation to him. drawn after the old manner'. At a point three degrees west of Flores the Genoese compass showed a westerly deviation. the chartographers. is convinced in almost all cases the black lines of the eight principal winds were drawn first. Florence. and the striking similarity which they present one to another. and reproved by Dr. for there are differences which exclude the possibility. and epoch * it was relatively small in the middle Mediterranean. remain to attest the slow change in the variation of the compass. set their coast-lines across the loxodromic lines at angles which differ slightly between one lustrum and another. in the slope of the Adriatic coasts. It became zero about the middle of the fourteenth. between the first quarter and the close of the fourteenth century. Still less known was the fact that the declination suffered a saecular change .

1534. 73). now in or in some of the maps in the Venetian Atlas (circa 1489) Venice. 2803) of the British Museum. These initials. Louis in the Sixth Crusade.. (No. Another feature of these sailing-charts is the method of indicating have seen the principal winds by the initials of their names. This practice was carried on into the fashioning of the sailing-charts. by a face appearing half-way across a circle. Venice. Venice (?). or framed in the margin of the Mappamundi. around the next change tana rosette at the centre of the loxodromic system of lines. set in eight of the sixteen subsidiary points of the network. 1.. in the Hamy collection . and in a sixteenth-century map of North Peru. Plate I) of 1436. of 1485. No. 2712). We how in the early windroses of the prior age the eight. were usually inscribed in medallions or circles. Plate II). by Bartolomeo Olives of Majorca.point star Levant . were prepared by compass. and the Levante almost invariably with a cross.THE ROSE OF THE WINDS 17 Willebrord Snell pointed out that the internal evidence of the charts proves that the compass was used in their construction as early as the thirteenth century . and Signor Enrico Rostagno (of the Biblioteca Laurentiana at Florence) is of opinion that the maps of St. or later with the is sometimes The fleur-de-lis. Another example i& in the British Museum (Egerton MS. Labetes. Where initials are used they are invariably those of the Italian names of the winds . of about 1508 . and how they were represented by heads of cherubs blowing. in the chart of Petrus Roselli of 1465 (Fig. Other examples occur in the Majorcan named and marked by an eight. or symbols. wherein the Tramuntana . and other islands in The the Isolario of Bartolommeo da li Sonetti. inscribed in Gothic lettering inside coloured circles. 4. or twelve. or twenty-four winds were named. Ponent. was to place the eight initials. but the Tramonmarked with a star or a dart. by Gabriel de Vallsecha. in the Atlas of early American charts in the Egerton MS. marked Migiorno the other points having the names Greek. as may be Andrea Bianco chart (Fig. 1 vi 2 N-2 . chart of 1447. many of which bore brightly coloured margins marked with the names of the winds. 1244-7. Axaloch. On the title-page of that work are two excellent pictures of early Italian compass-roses with the initials of the winds. An example is found in the Majorcan chart of 1385 by Guillelmo Soleri. labelled now at Florence. * afforded by the printed maps of Crete. afford another excellent example. No. Patmos. or bearing their initials. the South is indicated . seen in the The maps in the Ixolario of Bordone. or symbols representing them. or else placed where the loxodromic lines reached the margin. now in the Vatican. and Mestra. in the British Museum (Egerton MS. set around the rose. [sic] is in red the East by a red cross.

or of the Holy Family. and with smaller rhombuses or triangles between them for the half-winds and quarterwinds. drawn as a star with eight rhomboidal rays for the eight principal winds. in the later maps.18 PROCEEDINGS OF THE BRITISH ACADEMY third stage was to construct at the centre of the loxodromic net a rose of eight. or thirty-two points. these central windroses has already been noted. With the progress of time the embellishments became more elaborate. for example. adorned with the device of a ship. The art of the miniaturist lent to them its fascinating aid. are Virgin. sixteen. From the Isolario of Bartolommeo da li Sonetti. find to We the centres of the roses. or of an armillary sphere. The decorative treatment given ISLAND OF PATHOS. and to place on the eight principal rays the initials or symbols formerly inscribed in the margins. and even with miniatures of the Thus ornamented. the roses of the American chart of 1500 of Juan de la Cosa. repro- .

appartenente al Capitano di Vasselli quadri (1612). Again. Manwaring in his Seaman's Dictionary (1644). but it is significant to note that in the earlier charts. in the celebrated Atlas of John Rotz (1542) in the Additional MSS. hence charts with loxodromic lines to indicate the courses were the natural result of experience in sailing by the compass.winds are red. however. Pantero : L'Armata Navale (1614) . with the introduction of the compass into sea-sailing. green for the half.winds. The colour treatment of the rhumbs varies. but to justify certain conclusions that may be drawn from the comby parison. or with the map? A map with sixteen loxodromic lines would be of little service with a compass having only the eight principal winds marked upon it . even in the early examples. not rigidly adhered to. for the most part. Did then the division into sixteen or thirtytwo points arise first with the compass itself. These particulars have been dwelt upon not merely to show the intimate connexion between the compass-card and the sailing-chart. and that finer subdivisions were needful .THE ROSE OF THE WINDS 19 duced by Lelewel. : the colours given to the rhumbs of the thirty-two point star correspond with those given to the corresponding loxodromic lines. It has W. why should the chartographers of the beginning of the fourteenth century adopt green lines for the halfwinds. while the eight triangles of the half-winds are painted green. been pointed out that. But. No. Falconi's Breve Instruzione . and the sixteen triangles of the quarter. and by Sir H. 1 . viz. In any case it would be very simple for the maker of either the chart or of the compass to add to steer the extra divisions. Ruscelli^s Espositioni sopra la Geografia di Crescentio^s Nautica Mediterranea (1602) . For example. black for the principal winds. four of them usually in black. when considering the various types of roses. except in the charts originating at Mallorca or those with inscriptions in the Catalan dialect^ the roses bear on the eight rays of the principal winds the respective initials in Gothic capitals. referred to They are also Bourne in his Regiment of the Sea (1574). and indeed commonly. as we shall see. and red for the 1 quarter-winds. 11548. if the compass by which This convention is. the old system of eight or twelve winds became inadequate. the four cardinal winds are painted in gold. and red lines for the quarter-winds. Fabritius Paduanus's De Ventis (1601) Fournier's Hydrographia (1671) . Specific instructions for the use of these colours exist in several treatises Tolomeo (1561) Pantera^s . and the other four of the eight principal points are blue. while a compass with thirty-two winds would naturally be associated with a chart marked to correspond. of the British Museum. and four in red. as well as being the needful complement in the equipment of the navigator who intended by the compass.

originated 1 * with the Amalphitanians. If we may infer that step from the appearance on the sailing-charts of the thirty-two loxolines. p. as developed pari passu. did not possess the same colorations on their corresponding rhumbs is ? The correspondence it in the choice of colours most significant. But the wellattested tradition embodied in the line. 157. cit. where the South West wind blows from Africa. the year of may be earlier. dromic Petrus Veschonte's chart. remains undisturbed. seems therefore to be well founded so 3 ' far as regards two of its principal features: (1) that the invention mounting of the rosa ventorum upon the needle . ' invention 5 of the mariners' compass by the unknown navigator of Amalfi. century. For instance. bringing up the number of points to thirty-two.20 PROCEEDINGS OF THE BRITISH ACADEMY these observations had been made.. that the civic arms of the city of Amalfi bear emblazoned the figure of a compass. is then the date cannot be later than 1311. and the North wind across the mountains (the Apennines). Certain Op. least likely to occur so long as the compass remained a mere floating needle of the primitive Levant type. ' Ultimamente in Amalfi hanno collocate tra quest! otto . or even so long as the needle a fixed card. Another allegation. the year 1302 for the commonly supposed to date from the end of the twelfth The tradition. as in Peregrinus's compass of 1269. it has been urged that geographically the only place where the North East wind blows from Greece. now century. or at the of the fourteenth. consisted in the that this ' invention : Prima dedit nautis It usum magnetis Amalphis. Africa.. giving rise to the respective names of Greco. On this it suffices to observe that those denominations were in general use as the Italian names of the winds at least one or two centuries earlier. like an astrolabe. and those with which they were destined to be used. seems to have escaped the notice of recent writers that. The necessary step in evolution must have been that of the suggests that both compass and chart But that development would not be in the was pivoted to move over fastening of the movable card with its coloured rose upon the pivoted needle . That it was made at very beginning Amalfi must remain a matter of tradition for the matters which have been urged as evidence in support are all fallacious. if the Pisan chart. the addition of the eight halfwinds and the sixteen quarter-winds. or within a box having fixed graduations on its edge. and Tramontane!. must have been Amalfi. and that step was subsequent to 1269. (2) * took place at the end of the thirteenth. has long ago been shown to be wholly devoid of foundation in fact. according to Crescentio. attributed to Panormitan. and in Paris. which is strong in fixing really so old.

The West point. Hyperboreus. and that the characteristic features. was rarely indicated by the symbol of the setting sun. We are so accustomed to see the North point marked on the compasscard with a fleur-de-lis that it is somewhat surprising to be told that the compass-card existed two hundred years without any fleur-de-lis. The most Scirocco. Tramontana). O for Ostro.winds and quarter-winds. M M that symbol being. such as Septentrio (from the seven stars) or Arctos (the Bear). in the usage. The name Tramontana became applied not only to the wind that blew across the mountains. were developed in connexion with the loxodromic in sailing-charts of the Italian navigators. per piu commodita de* Naviganti. The South West point is usually marked with A for Africo. marked with the winds in a rose of eight principal points in the Italian manner. 255 he adds : essere poi che in Amalfi. The three in variables are the North East with G for Greco. and with subdivisions corresponding to half. or were the names of winds such as Boreas. in Gothic capitals. the names for the North. and. the initial T is often used to denote the North point. but never are the initials for the venti principal! altri otto.' ( . and hence to the geographical North. instead of being inscribed with L for Levante was more commonly marked with a cross. It has been mentioned above that the majority of the early Italian compass-roses bore. originated Southern Italy at the end of the thirteenth century . and but rarely with for Mezzod\. the initials of the Italian names of the eight principal winds. however. and the North West with variable of all in the manner of its indication is the North point.' Puo On p. and very The South point is usually marked with rarely with G for Garbino. There were variants. of the compass-card. In windroses. and in early compass-roses.THE ROSE OF THE WINDS 21 it is that the compass with the pivoted card. though almost always marked with a P for Ponente. by metonymy. The East point. but also to the North Pole Star (Stella itself. None of these various names North lent its initials to mark the North point except Tramontana. fusse 1' una [la carta] e 1' altra [la bussola] in 32 punti o venti scompartita. less often with L for Libeccio. a late addition. or Tramontana. the South East with S for for Maestro. Aparctias. detti da loro meziventi. in fact. as inscribed on In the pre-compass era maps were either astro- nomical. ORIGIN OF THE DISTINCTIVE MARKS USED ON COMPASSCARDS. even to the magnetic needle because it pointed northwards. including the system of colouring adopted. Aquilo.

che in mar XK vera norma. by Marco Boschini (Venice. and may therefore be regarded as distinctively denoting a magnetic needle. . is sometimes found. Plate IV). Egerton MSS. of 1545. Brussels). representing a female figure. 3 as in the rose of the Catalan Atlas (1375). 1545). A narrow. 3 The close association of the symbols of the star and of the dart with the North is attested by a remarkable etching in a book on Painting. of the other names Boreas. No. 1660). however. so used. extending from the middle of the rose to the North. This mark is not found in the pre-magnetic windroses. Tramontana. Plate II). and in the Mappamundi Marino Sanudo. Septentrio. like an arrow-head. 2803. 7. in bela forma. A triangle for the North is one of the four symbols found in the MS. the still earlier rose of painted black. and sometimes in association with it. che stabile. twelfth century). the North in the compass-roses of the Catalan or Majorcan type. A star. &c. as in the rose of the Pinelli chart of 1384 Nicolao di Combitis in the Marciana library at Venice. marks the North in the Planisphere of the Cosmographie d'Azaph (MS. also in the rose of the Agnese chart now in Stockholm. A group of in the chapter seven stars appears in the compass-card (Fig. A star. Paris. marks (Fig.22 PROCEEDINGS OF THE BRITISH ACADEMY Other symbols. and in the similar chart at Upsala . are found quite as frequently as the letter T. Opposite it is the following quatrain 2 : 1 Voi che La Come rapresenta. Atlas World in the British Museum. or a group of seven stars. of for instance. Another occurs in the chart of Juan de la Cosa (1500) A broad triangle. 2 the oldest atlas of the New Occasionally the triangular or arrow-head mark is associated with the star-symbol. Plate III). 4 & 5. 1321 (MS. p. A commoner form is that in which the North triangle or rhumb of for the mark the rose is (Fig. stela. 1 in the same library. al bon Peota. while above is an eight-point star. also in Venice. e immota 7 me Sia scorta al Marinier. and in In this instance the coloured rose appears to be of later date than the chart itself. dart-like triangle. the compass on the chart of Jachobus de Giraldis. dated 1508. called La Carta del navegar pitoresco. also in Fontano's book just mentioned.. A A North may be described as a triangle representing an arrow-head or a dart. emblematic of the North wind. 2. and on that of Battista Agnese. 11. Plate VI) printed ' Dei Venti e della Bussola da navicare * by Giovanni It Quintino Sofo in the Delia Guerra di Rhodi of Jacopo Fontano star or a group of seven stars appears in the (Venice. of 1426. is found in the sailing-chart (late fifteenth century) in the Archives of Lucerne (Figs. more frequent compass-cards of the Catalan type mentioned below. 616. with a javelin .

in his Tratado da Agulha de marear (1514). ' Nigredo ilia Septentrionem '. as in the rose of Andrea Bianco (1436). 204) that the earliest known to is on a Majorcan chart in the Library of Parma (No. 1620). fallacious inferences having been it is important to Similarly. Plate V) Museo Civico at Venice. or as the sailors often call it. Many drawn from the occurrence 1 of the fleur-de-lis.. Ventis et Navigatione. said to now also in the Atlas of found. the South point pee (foot). it is not known on any compass or chart before the sixteenth century. which is believed to have belonged to Cardinal Richelieu. p. the Prince of Wales's feathers. No. 1613. 3. 4. No one has hitherto been able to account for the introduction of this to exist 2 symbol . in the interior of A the compass printed in the De Angelus as the head Blondus. in the British Museum (Egerton MS. head also appears. in Paris. surmounted by a dart. while the East and West are respectively inscribed braqo de leste and braco daloeste [sic]. 1546. also in that of Giorgio Callapoda (1552) (Fig. but its form suggests a trident rather than a lily. 12. as in the strange Plate VI) of Pierre Garcie. the device used to indicate the North is the fleur-de-lis. A trident occaper nigrum radium intelligendo Tramontanam instead of a dart. the North point. in that of Giacomo Castaldi (1564). in the Isolario of Benedetto Bordone (Venetia.THE ROSE OF THE WINDS 23 that of the chart of Petrus Roselli (Fig. marked with a fleur-de-lis. the pilot of Rouen. 'laltra che e tutta negra e tramontana'. conjectured to be of date 1493 . Plate II) in the British Museum. 819). in the use of a human head. to mark the North. . Joannes de Oliva. In all modern compasses. 10. practically without exception. The use of a black in the In the triangle for the North is recognized also by various writers. Father Bertelli says (Studi storiti. is inscribed cabeqa (head). but this does not seem to have been reproduced in any of the published 3 him collections of charts. Plate II). Also. at Venice. De Ventis of Fabritius Paduanus (Bononiae. of Michael That mariners regarded the North is emphasized by the fact that in a compass-rose of Joao de Lisboa. 1601) occur l the phrases : Tramontanam denotat '. of late fifteenth century . ' on the sailing-chart of the Monastery of Badia di Cava Tirrena (circa 1492) (Fig. as in the rose sionally appears Nigredo ilia ' '. and in the chart engraved on copper by Nicolaus di Nicolay (1560) at Venice. compass-card (Fig. 1894. 1584. the supposed earliest one being that on the chart in the Bibliotheque Nationale in Paris. occasionally. the triangle is associated with the initial T. 1534). and in the rose of the late fifteenth-century have been in the possession of Richelieu . Yet one other symbol for the North is chart. occasionally.

et en particulier sur la fieur de Us. the heralds adopted as a symbol of sovereignty the emblem which had struck them as having this remarked that on several of the decorative charts of the fifteenth century for example. But the banners.24 PROCEEDINGS OF THE BRITISH ACADMEY The fleur-de-lis has been for scrutinize closely the circumstances. Whence then comes the appearance of the fleur-de-lis as the chief device on the modern compass-card ? The answer which may hesitatingly be given. and scutcheons used in that cereearlier than this. and has derived it from the lotos ornament of Egypt. It seems to have been appropriated by at Rheims in 1179. 105. Recherches sur Forigine du blason. lent itself to this development. of 1339. Only after the year 1500 does the fully-formed fleur-de-lis appear. which may be traced through various incomplete stages. hangings. the fleur-de-lis is found marked on the banners that indicate the political standing of some of the cities. and in others of the fifteenth century which bear compass-roses devoid of fleurs-de-lis. in his Dissertation sur Saint Louis. the East before the time of the Crusades. orfleur de roi . after examination of the various treatments of the marking of the North point. of Genoa on which the loxodromic rosettes bear no roses. 3 Another strange development of the ornament at the North point is the bunch of grapes. From Louis VII to Henry IV the emblem preserved its primitive form in the royal blazonry of France. quoting a dissertation by Bullet. In particular it was adopted by the kings of France. on a field of practice began of strewing with golden fleurs-de-lis azure all mony. the Philippe Auguste. It was in use for regal decorations in centuries a heraldic emblem. the word K in Celtic signify- Ducange. with the motto Florida florenti floreat Florentia flore. is that it is a purely decorative 3 development of the dart. that the original form w&sfleur de li. of 1791. that after the first Crusade. political influences 1 would tend to promote its employment. de Beaumont states 1863. or serifs. at whose coronation. confirms this by saying ing king. of 1525. The T of Tramontana. and that of Florence bears argent a single lily gules of the peculiar Florentine form with two projecting stamens. the star. 2 It may be . On p. and the T of earlier use. with its drooping appendages. and the still earlier map of Giovanni da Carignano. Doubtless when the usage of a fleur-de-lis had once found acceptance. by the dukes of Anjou. Louis IX ordained that the fleur-de-lis should be borne by all the princes of royal blood. reproduced in Nordenskjold's Periplu*. Adalbert de Beaumont 1 has traced back its use as a decorative emblem to Byzantine times. not so much as being a lily but rather as emblematic of royalty. Paris. Carolus Magnus had given the fleur2 de-lis to the city of Florence. The banner of Paris bears azure five fleurs-de-lis or. significance. which appears on the rose in the chart of Ruiz de Estrada e Penato. Dulcert's. and then most frequently in the Portuguese compass-roses. and by the Bourbon family. when Christian heraldry began.

of an actual compass and the designs drawn in the roses of the sailing-charts. while in the case of the latter there is nothing to hinder the draughtsman from depicting salient ornaments that project forth beyond the circumference of the rose. Similarly on a rose of a sailing-chart of Africa. such as abound on the roses of the florid charts of the sixteenth century would be impossible on a real compass-card. at Weimar. is doubtless printed from the very wood-block which was used to print the compass-cards in that port. though the roses of the charts may be in the main pictures of the com pass.THE ROSE OF THE WINDS ' 25 It must be borne in mind. Projecting spear-heads. of 1584. bespeaks the courtier in the artist . as further discoveries extended the knowledge of the coast-lines . illustrates ' the manner for to knowe the Wyndes of all the points of the Sey compass* with a gorgeously-coloured rose having a golden lily and a golden cross standing outside the periphery. of 1527. in the British Museum. pilotage et encrage de la Mer. The rude compass (Fig.cards of the day. crosses. Internal evidence shows that in many cases the sea-charts were revised or improved by later hands. they must be received with caution. however. Plate VI) of Pierre Garcie of Rouen in his treatise Le Grand Routier. as appears to be the case with the fine compass-rose on the early fifteenth-century chart of Jachobus de Giraldis. reproduced in Jomard's Monuments de la Giographie. . 12. The celebrated Atlas of John Rotz (1542). So. and in many cases the decorative embellishments are of subsequent date. and lilies. pyramids. and a large P at the West. The former must of necessity present a circular periphery in order to turn freely within the compass box. while in another chart (sixteenth century) in the Biblioteca Marciana a triangle extends at the North. as to their respective title to discovered lands equally proclaims that it is really no inherent part of the compass-rose which it adorns. a Latin cross at the East. as Tudor writers called it. a lily and a cross stand out as excrescences outside the circle . The gorgeous fleur-de-lis on the map painted on vellum by the order of Henri II. while the splendid blue lily depicted in 1529 by Diego Ribera in his map the famous map on which Pope Julius II drew the line to decide the dispute between Spain and Portugal. the letters T and L are tacked on to the apices of the North and East points respectively. It is possibly the oldest compass-card in existence. that there is a distinction to be drawn between the designs of the roses actually used on the flie 3 . On the rose of the Andrea Bianco chart in Venice (1436). So far as the author is aware there is not preserved in any museum in the world any actual compass or compass-card older than the late sixteenth century.

Distillation. one may are. often as a cross pattee. Gunpowder. In the rose of Cecco d' Ascoli (see p. Angleterre. and hence the reason why the fleurde-lis is marked on the card This erroneous view appears to have ! originated with Gassendi. In a compass preserved in the Galilei Museum in Florence.26 PROCEEDINGS OF THE BRITISH ACADEMY As to the that a fleur-de-lis arguments which have been founded on the circumstance is so generally found on compass-cards. that in 1302. etched by P. In these None have etchings appear three figures of the mariners' compass. Florentino. Outre qu'il est constant que les mots de Nord. appears usually as a Greek or a Maltese cross. Italic bref tous les Europeans auroient-ils mis en ce lieu plustost les armes de France que de leur Nation. Sud. Lapis Polaris (the Lodestone). sepulchres. was governed by French Venanson. ont plus de force que ceux que nous pourrions auoir de quelque auteur passionne pour son pays Je dis que la Fleur de lis. Galle from the drawings of Joannes Stradanus. marbres. argues ' e compass was invented by the mythical Flavio di Gioia. 1808. How little the fleur-de-lis was considered as pertaining to the compass may be judged from the remarkable group of etchings published about the year 1600 under the title of Nova Reperta. sometimes as a Latin cross. It continued to be supposed to be about 1650. Danemarc. the traditional date when the Naples. p. et choses semblables. and dedicated to Aloysio Alamanno. or into a pair of parallel cross-lines. They relate to the recent discoveries of America. rarely as a cross fitchee. Silk' growing. comme pierres. even on English . argues thus Puis que toutefois les arguments qui se tirent des monuments. de la pointe de 1'aiguille dont toutes les Nations : de 1'Europe se seruent pour monstrer le Nord. the kingdom of Naples. : now see of what value they Father Fournier. it appears marked. fleur-de-lis. The It cross at the East point underwent many variations of form. et que c'est d'eux qu'ils out appris 1'vsage : car a quel propos les Matelots de Noruegue. 399. It sometimes degenerated into an arrangement of intersecting circles. in his Hydrographie (1671). temoigne assez que ce sont les Francois qui 1'ont bastie de la sorte. of date as a pair of dolphins. Guaiacum. the Printing Press. and Harness. 11) it resembles a cross-handled dagger. Est. prisco operta cuncta saeculo '. dont on se sert sur 1'Ocean pour monstrer les Runs des vents sont mots Fra^ais. : la boussole nautique. Andrew's cross attached outside the periphery of the rose. The compass on the title-page is a simple eight-point star with an arrow-head at the North point and three stars at the South. as in the Weimar map of Africa. Clocks. et Oiiest. traditions immemoriales. Occasionally it takes the form of a St. desquels on se servoit du temps de Charlemagne. in his work De Pinvention de princes of the house of Anjou. in which Amalfi is situated.

Another strange compass-rose is that printed in the De It has twenty-six Ventis et Navigatione of Blondus (Venice. engraved by John Sellers about 1660. distinctive As marks for the West or sometimes a plain circle. while the North point bears a fleur-de-lis. on p. details. Having now reviewed the development of the compass-card and its we may attempt some classification of the forms which appear . It bears in the centre a crude human figure with the head to the North. the cross has degenerated into a couple of S-shaped marks above and below the East point.THE ROSE OF THE WINDS 27 compasses. points marked with the Latin names of twenty-six winds. the South an arrow-head. Delia Fabrica Uso di diversi Stromenti di Astronomia (1598). Mer (1661) states that the Cleirac in his Us et Coustumes de la West point is marked with an imperial et double-headed eagle. This he regards as a new compass. M(eridies). the T of the Tramontana being replaced by a javelin-head . In a fine card in the author's collection. point one occasionally meets with a P. the four cardinal rays of which are marked with the Latin initials S(eptentrio). has the usual thirty-two English names in initials surrounding a thirtytwo point rose. 286 of the Paris edition of 1801. and speaks of those who formerly sailed ( cum observatione antiqui pyxidis et cultu Ursae minoris '. a form which survived even to the nineteenth century. of date about 1640. or a double P (from which possibly the heart-shaped ornament is derived). picture A of a compass so designed is given in Pigafetta's Premier voyage autour du monde. Of singularities in the design of compass-cards a quaint instance is afforded by the woodcut print of Pierre Garcie (Fig. erected over the centre of the needle. and O(ccidens) . 12. and the West a heart-shaped ornament. described as ' He alleges reasons for abandoning the old compass with eight or twelve winds. the East a cross. Riccioli in his Almagestum (1651). down to the eighteenth century. 1546). In the middle is an inner circle in which are inscribed the usual eight symbols of the Italian rose. or sometimes the figure of the setting sun with rays. 1519-22. instrumentum et dux navigantium '. Another English card in the author's collection. and in the centre of this there is drawn a human face with the top of the head towards the North. and is Pixis vel Buxolus. Plate VI) already mentioned. speak of compasses which possess. and arms spread out from the elbows to the East and West. an upright stylus to carry a little flag to indicate the direction of the wind. O(riens). and for preferring this new pyxis with its twenty-six winds. The East and West names appear to have been reversed in printing. and Gallucci.

the eight half-winds green. appendages of an arrow. adopts the three ranges of but uses a different colouring. the rose of the anonymous chart. but with the late addition of a lily in gold. No.28 PROCEEDINGS OF THE BRITISH ACADEMY Amid the vast variety of treatment we may admit prior to 1600. and without initials. The eight-point rose of Andrea Benincasa. 1). xx. of 1514. 1560. with eight black. An 2. the North ray being black. The is It is the basis of many more developed forms. Italian type with eight winds. 2). and sixteen red. The Atlas of John Rotz (1542) has a rose with four of the principal winds gold. 8813) has eight winds blue. 2712). and the . Typical examples are the Andrea Bianco rose of 1436. Plate II). fine sixteenth-century triangles. eight green. eight green. collection of M. of the Bibliotheque de 1' Arsenal at Paris A (No.winds. of early sixteenth century. and having a broad arrow-head at the North. roses of To the same type. and . with four black. The rose of Cecco d'Ascoli of 1521 (page 11) is another example. Italian here is The characteristic feature type with thirty-two winds. of Ancona. eight in green. This is characterized by having the four cardinal winds in blue and the other four winds in red or gold. eight gold. represented. F. 1476. belong the Matheus Prunes. of the same epoch. Hamy. in the British 4. of 1465. has a dart added externally at the North. without any triangles for the half-winds or quarter. and the sixteen quarter-winds red. 3. which has four rays blue and four red. and sixteen red. a cross. 260. decorated with a lily. This type bears no initials. with eight winds in with the gold. Joao de Lisboa's rose. Plate III). Typical specimens are that (Fig.. save that the North is black . three leading types In this type a star of eight rays 1. The rose of Nicolaus de Nicolay. the other seven of gold. and the rose of the chart of Badia di Cava Tirrena (Fig. rose in the Spanish MSS. affords a Portuguese example of this type. infra No. the other four blue. has eight half-winds in addition. four gold. is in the eight-point rose in the chart of date 1501-2. (Egerton MS. marked initials or symbols of the eight principal winds. Catalan or Majorcan type. A design by Cabot (see XXX Jomard. of the second half of the fifteenth century. with the same colouring. in the Biblioteca Ambrosiana rhumb (No. with internal circle for the setting sun. Plate II) Museum on the chart of Petrus Roselli. reproduced by Kretschmer (Plate of his collection). 5. sixteen red. in the possession of Prince Corsini and the undated rose in the Museo Civico of Venice. but with some modifications. 1561. and a another. 3. : Lucerne chart of late fifteenth century gives an example (Fig. the arrangement of three series of coloured triangles. sixteen in red .

or sometimes thirty-one. No. 73. are all close to the type. This type occurs chiefly between 1500 and 1600. 43. of 1466. Plate * IV) of the exquisite little anonymous atlas. Other examples are found in the roses (Figs. it reappears surmounted by a fine fleur-de-lis Rome in 1562 (Additional in his chart of 1529.. Another is the moderno 9 depicted in Bordone's Isolario (Venice. black triangle extending from the centre of the rose to the North point the other winds being represented by fifteen. and that in the Gregorio Dati manuscript of 1420 in the Bibliotheque Nationale. in 1500 and 1520. No. No. Archivio di Stato.THE ROSE OF THE WINDS 29 almost identical one. Closely akin are the roses of examples. one of them the oldest known compassand the rose of the Berlin sea-chart (circa 1450). Catalan chart of 1429 (Florence. is a cluster of seven stars. of 1482 (Florence. in the collection of M. is exemplified in a form used by Juan de la Cosa (1500). at the end of Egerton MS. of 1561 and 1564. the chart of M. in which the subsidiary winds are added.. of date about 1540. 1607 (Florence.. No. the chart of G. at Mantua. 4. those of this type are characterized by the presence of a large. 16) . having the eight principal winds in blue and gold. the charts of Giacomo Castaldi.. Sideri. Black dart type. di Majorca. of 1562 (Egerton MS. the map of Diego Homen No. usually of equal size. (Fig. by the same hand. made in MSS. It is of some interest to note that while this type occurs along with the arrow-head (Catalan) type in the charts of Juan de la Cosa. 2856) . . Hamy . and the quarter-winds in white. the eight half-winds in red. No. ranged round the periphery. 42). Plate III) 1534). of 1487 (ibid. So are the roses of two fifteenththe Catalan Atlas in Paris rose century manuscripts in Paris that in the Arsenale (Italian MSS. smaller triangles. 1916) . No. of the Lucerne chart (late sixteenth century) . of 1556. described and Both these have crosses at the East. narrow. 5 & 6.. A modification of this type. Bibl. Archivio di Stato. the chart of Jachobus Vesconte de Maiollo.. 19). 8.. in the possession of Above the arrow-head for the North. The figured by Kretschmer. the Minorcan chart of Giacomo Bertran. 2858) . No. in the British Museum (Egerton MS. ' the chart of Agnese Battista at Stockholm . Although in some of the roses of the other types a black triangle indicates the North. 8) . Naz.. 7) . also in the sketch of a compass in the (Egerton MS. fol. that of Vincenzo de Demetrio Volzio di Ragusa. the chart at Upsala . Italian manuscript. cited by Bertelli. in these Prince Corsini. of uncertain date. No.. No. A typical example is presented by the roses . the e Richard King bossolo da nauigar chart of 1501-2. 2854). the chart of Angelus Eufredutius.

to have gloried in inventing new designs from page to page of their atlases. of 1567 and of 1578. did the them seem Indeed. and P. all in Around the outside are sixteen small triangles. and M. Another is that in the anonymous An sixteenth-century chart in Florence (Archivio di Stato. To this class belong the roses (Fig. 3489. No. but with other features like those of the Italian roses of The roses of the charts preserved in the various thirty-two winds. In the middle are the eight symbols of the O. usually in red. An actual compass-card by John Sellers (circa 1660). between them. both in colouring and detail of form. of 1600. in red. of early seventeenth century. though in many cases the identity of the artist is unmistakable. at Paris (circa 1520). and those of Giacomo Russo. S. A typical example is the rose of Demetrius Voltius. Since 1650 the disappeared. and which occurs also in English compass-cards. William colour. in the writer's collection. but is devoid of It is singular to find in the De Magnete of Dr. Gilbert. or any other. sometimes white. 1557. old compass. Marseilles. or Thus in a Portuguese atlas in the Bibliotheque de F Arsenal. others appear. in the Galilei Museum in Florence. 185. This form. 1681) an almost identical navegar rose is engraved. whether Naples. of Messina. reproduced in Nordenskj old's Periplus. and before the seventeenth century was over all coloration and all use of the Italian initials compass-cards of all of any kind. Portuguese type. Hamy. is of the same type. is practically the Catalan type. PROCEEDINGS OF THE BRITISH ACADEMY Messina type. a card of the needle. : Italian cross. museums of Venice present great variety. having four winds In the Arte pratica de (including the lily) in blue and four in red. is also closely akin. p. 8. and G. in the collection of 6. winds dart. nations have presented little variation European . we find a rose with a black dart. but with a fleur-de-lis replacing the arrow-head. for the quarter. of Luis Sarrao Pimentel (Lisboa. probably English. sometimes in alternating blue and gold. of 1593. in black . No. M. same pattern used in a dipping- Besides these types many with mixed characteristics. It is not possible to assign definitely any special typical design to any particular port. the latter being Harleian MS.30 5. Genoa. some of designers of these roses always adhere to one type.. A. of subsequent date to the preceding.winds. with sixteen smaller rhumbs. 17). Nor. some showing Arabic or Moorish influence. Plate V) in the charts of Joannes Martines. of intermediate forms. After 1600 nearly every compass bore a fleurde-lis.

Havre. bearing at its North point the device Amongst here copied. tention will be much more than a plausible suggestion. others are said to exist in Probably others remain Possibly in unknown in Munich and in Madrid. compass. presumably English. illustrating with a wealth of detail the construction of the compass and the design of the compass-card from the earliest time to the present date. his con- S. or calamita. which are strictly not fleurs-de-lis in their intention. Marseilles. At the date when this the privately-printed Schiick of Hamburg. museums or private collections. Bruges. to have been a customary proved hereafter be rainette. origin The study of the Note. the introduction of which on the compass-rose at the end of the fifteenth century is discussed in the authors paper. the compass-roses described by Herr Schiick is one in an early sixteenth-century Portuguese atlas in the Riccardi Library in Florence. with a coloured card. Naples. The dart is in black. But Herr Schiick puts forth a new and ingenious conjecture as to the origin of the fleur-de-lis mark. issued in 1911 by Herr A. . elaborate repertory runs counter to the main conclusions expressed in the author's paper. Genoa. added February 1914. Hamburg. exists in the Galilei Museum in A Florence . the two side-appendages are gold-coloured. Numerous similar forms. P. Rouen. and that it depicts a lancet-shaped magnetic needle supported between its two wooden floats. Nothing in this remarkable and paper was written the author had not seen work Der Compass. Venice or other old maritime centre. could be cited from other If this should early examples. Herr Schiick suggests that this mark is essentially a representation of the form of the primitive floating compass. there may be found sailing-compasses with roses coloured in the ancient manner. in the Mediterranean. T.THE ROSE OF THE WINDS It is to 31 be hoped that further research will be made as to the existence of actual compass-cards of early times. Lisbon. he agrees that Herr Schiick at least has made out a case worthy of consideration. which consists of a series of forty-six magnificent plates in colours. While the author does not admit that this hypothesis has been proven by Herr Schiick. together with a Catalogue describing the hundreds of figures in those plates. of any such will further elucidate the problems and development here considered. form of the which preceded the pivoted compass.


Fig. from sailing-chart of Andrea Bianco. known .Plate I Fig. as the Portolano Pinelli. of 1436. 1 Loxodromic rosette with initials of Italian names of the winds. now in Paris. 2 Windrose from the anonymous sailing-chart of 1384.


) Rose from the No.) .Plate II Rose from the sailing-chart di Cava Tirrena. (circa 1492) in the Archivio della Badia (Italian thirty-two point type. 2712 in sailing-chart of Petrus Roselli. (Catalan or Majorcan type. the British Museum. Egerton MS. of 1466.


(Eight-point type. with black dart. 6 Rose from sailing-chart in the archives of Lucerne.) . circa 1500. circa 1500.) Fig.Plate III Rose from sailing-chart in the archives of Lucerne. (Black-dart type.


the British Museum. 8 Egerton MS. No.Plate IV Fig. Fig. 7 Rose from sailing-chart of Juan de la Cosa. (Black-dart type. . 2854. 1540.) Rose from in sailing-chart. of 1500.


1552. 1567.Plate V Fig. (Messina type. 9 Rose from sailing-chart of Joannes Martines.) Fig. . 10 Rose from sailing-chart of Giorgio Calapoda.


Plate VI Fig. Fig. . Pilotage et Encrage de la Mer. 1545. from the Dei uenti e delta bus8ola da nauicare of Giovanni Quintino. Rouen. 1584. of Pierre Garcie. Venice. 11 Figura della Bussola. 12 Compass-card printed on title-page of Le Grand Routier.



Oxford Printed by Horace Hart. at the University Press .