Oarn bridge ; ^^^.i^:igi||i|D^.::;]NFI.UENCE VR7i8.9 R433 LIBRARY OF LEONARD PEARSON VETERINARIAN CAMBRIDGE BIOLOGICAL SERIES. General Editor : — Arthur E. Shipley, M.A., F.R.S. FELLOW AND TUTOR OF CHRISt's COLLEGE, CAMBRIDGE. THE ORIGIN AND INFLUENCE OF THE THOROUGHBRED HORSE By the same Author. THE ORIGIN OF METALLIC CURRENCY AND WEIGHT STANDARDS. New Edition in preparation. "Die epochemachende Untersuchungen von William Eidgeway, ' The Origin of Metallic Currency and Weight Standards.' " — Devtsches Rundsc}iaii, June 5, 1897. "It is the induction which is the real strength of the present work. The collection of sure facts is so large, and the facts themselves hang so well together, that we cannot help accepting what they point to — at least until we see whether an adversary can make an equally good collection on the other side. But we do not expect to find this done." — Economic Journal, vol. ii, p. 704. "One of the most important and brilliantly original works on any archaeological subject which has appeared for many years past." — - Saturday Review. "A book of profound erudition, and of the first value to everyone interested in the early history of civilization." — Scotsman. THE EARLY AGE OF GREECE, Vol. i (Vol. II in the press). "No more lucid piece of argument has been produced for many years. Mr Ridgeway takes no step which is not sure. He trusts neither to prejudice nor to speculation. He admits nothing save facts, and being an eminent anthropologist he does not reason as though Greece were a province set in a vacuum far apart from the civilization of the world." — Spectator. "With the main arguments we are fully in accord. The finds both in Greece and elsewhere, on which it is largely based, appear so far as we can test the matter to be accurately stated, and no material evidence seems to have been ignored. In this part we can hardly believe that his position will be seriously questioned.... We think the legends are most likely to be the battle-ground with his opponents. But they have never been critically examined before in the light of archaeological discoveries, nor has anyone so successfully illustrated them by the way in which historical heroes, like Alexander or Charlemagne, are treated in saga and song ; and when so examined their consistency with themselves and with the finds is indeed remarkable. It is no small confirmation of their value that by following them Mr Ridgeway has been enabled to explain better than any of his predecessors the origin of the Homeric j^oems." — Athenaeum. " Der vorliegende erste Band des auf zwei Bande berechneten Werkes verdient wegen des Inhaltes und wegen der Art der Stoffbehandlung aufmerksame Beachtung An dieser Stelle muss das Hervorgehobene geniigen und wird wenigstens das eine gezeight haben, dass der vor- liegende Band der interessanten Schift, auch schon wegen des reichen Materials in archjiologischer und priihistorischer Beziehung, ein sorg- faltiges und eingehendes Studium verdient." — Neue Philologische Rund- schau, 1902, pp. 132-5. " .Jetzt beginnt er in einem grossen Werke, dessen erster Band vorliegt, die Frage nach Griechenlands Jugendzeit in peinlich sorgfaltiger und ausfiihrlicher Weise zu behandeln, und das Buch diinkt uns so interessant, dass wir in einer kurzen Uebersicht seiuen Gedankengang hier wiedergeben woUen." — Allgemcine Zeitung, 1901, p. 260. THE ORIGIN AND INFLUENCE OF THE THOROUGHBRED HORSE BY WILLIAM RIDGEWAY, M.A., F.B.A., Hon. D.Litt. DISNEY PROFESSOR OF ARCHAEOLOGY IN THE UNIVERSITY OF CAMBRIDGE, FELLOW OF GONVILLE AND CAIUS COLLEGE, CAMBRIDGE, SOMETIME PROFESSOK OF GREEK IN QUEEN'S COLLEGE, CORK, HON. MEMBER OF THE ARCHAEOLOGICAL SOCIETY OF ATHENS. A most absolute and excellent horse." Shakespeare. IVITH NUMEROUS ILLUSTRATIONS Cambridge at the University Press 1905 CAMBRIDGE UNIVERSITY PRESS WAREHOUSE, C. F. CLAY, Manager. aonUon : FETTER LANE, E.G. (Slasgoto: 50, WELLINOTON STREET. ALSO ILonBon: H. K. LEWIS, 136, GOWER STREET, "W-.C. ILeMJjig: F. A. BROCKHAUS. iftfto lork: THE MACMILLAN COMPANY. Bombag mt> Calcutta: MACMILLAN AND CO., Ltd. K H33 [All Ilights reserved.] to TO JAMES COSSAR EWART, M.D., F.R.S., REGIUS PROFESSOR OF NATURAL HISTORY IN THE UNIVERSITY OF EDINBURGH. Digitized by the Internet Arciiive in 2009 witii funding from Lyrasis IVIembers and Sloan Foundation http://www.archive.org/details/origininfluencOOridg PREFACE. THE present work is an attempt to solve one of the chief problems in the history of the horse, the most important (the ox not excepted) of all the animals domesticated by man. The nature of the inquiry rendered it necessary to treat not only of all the chief breeds of domestic horses known in historical times, but also to take a survey of all the other living Equidae, as well as of the ancestors of the genus. The question of coloration naturally holds a prominent place in the investi- gation, and I have tried to trace historically the origin of the various colours found in domestic horses. At the same time I have endeavoured to indicate, though it must be con- fessed rather cursorily, the influence exercised on the history of the chief nations of the ancient, medieval, and modern world by the possession of horses, and especially by the acquisition of what I venture to term the Libyan horse. I have also tried to point out the lessons of supreme importance to the breeder which can be learned from the contemplation of the injury wrought to breeds of great value by the ill-judged and un- scientific introduction of alien blood, a practice in no small degree due to a lack of historical knowledge, and to a general belief that all our domestic races of horses, like those of tame pigeons, have been obtained solely by artificial breeding from a single wild species. I must add a few words respecting the method of writing zoological names and the transliteration of foreign words. In zoological terminology I have, of course, conformed to the rule Vlll PREFACE of the editor of the Cambridge Series in which this work appears, but my classical friends will understand ray qualms when I have to describe, for instance, Burchell's zebra as Equus burchelli instead of using the usual Latin form, Equus Bur- chelli. I have used the form ' Prej valsky's Horse' as the nearest English equivalent for Equus przewalski/i since the letter-combination Pt^ze has no phonetic significance for the English reader. In Arabic words consistency was impossible, for although I myself use only the vowels a, i and u in trans- literation, e.g., Muhammad instead of Mahomet or Mahomed, in extracts from other writers I had, of course, to preserve faith- fully the forms which each employed according to his fancy. It only remains for me to express my gratitude to the many kind friends who have aided me in various ways : Dr James Cossar Ewart, F.R.S., Regius Professor of Natural History in the University of Edinburgh, has given me much of his time and has read through all my proofs, and twice through those of the first two chapters ; whilst Mr R. I. Pocock, F.Z.S., the Superintendent of the Zoological Society's Garden, Regent's Park, has given me valuable aid by reading the proofs of the chapter on the Living Equidae. Mr A. E. Shipley, F.R.S., Fellow of Christ's College, and University Lecturer in In- vertebrate Zoology, and editor of the Cambridge Biological Series, has also read the proofs and has aided me with various suggestions. Had it not been for their criticism and advice the shortcomings of this book, of which no one can be more sensible than the writer, would have been still more numerous ; but for the many that remain I alone am responsible. To Dr W. L. H. Duckworth, Fellow of Jesus College, University Lecturer in Physical Anthropology, I am indebted for various important references, and above all for having called my attention to a hitherto unpublished head and neck of a quagga in the Elgin Museum, which I figure and describe (pp. 438-9) ; Mr A. W. Howitt, Hon. D.Sc, the well-known Australian ethno- logist, of Metung, Victoria, has supplied me with the valuable account of the feral horses of Eastern Victoria, which I have embodied ; Dr R. S. Conway, Professor of Latin in Manchester University, and late Fellow of Gonville and Caius College, PREFACE IX made for me in Florence the very careful measurements of the chariot from Thebes ; whilst Sir Ernest Clarke, M.A., St John's College, Secretary to the Royal Agricultural Society, has supplied me with several valuable notes which I have embodied, and has also aided me in obtaining the use of blocks and photographs. Dr M. R. James, the Provost of King's College, Cambridge, and Director of the Fitzwilliam Museum, pointed out to me the drawings of Stradanus, reproduced in this work. I had the advantage of talking over many points in my book with the late Captain Maurice H. Hayes, the well-known author of The Points of the Horse, etc. Mr G. P. Bidder, M.A., Trinity College, long resident at Naples, and through him Mr E. Neville Rolfe, M.A., H.B.M. Consul for southern Italy, have given me much valuable information respecting the horses of Naples and southern Italy. I am also indebted for very important references or for help in various other ways to the following: Mr A. A. Bevan, M.A., Lord Almoner's Professor of Arabic and Fellow of Trinity College; Mr H. M. Chadwick, M.A., Fellow of Clare College; Mr J. G. Frazer, M.A., F.B.A., Hon. D.C.L., Hon. LL.D., Hon. D.Litt., Fellow of Trinity College ; Dr C. S. Myers, Gonville and Caius College ; Rev. Canon Greenwell, F.R.S., D.C.L. ; Rev. T. T. Gray, M.A., Senior Fellow of Trinity College, Dublin ; Mr J. Bass Mullinger, M.A., Librarian of St John's College; Mr R. C. Bosanquet, M.A., Trinity College, Director of the British School at Athens ; Mr George Coffey, M.A., Keeper of the Museum of the Royal Irish Academy; Mr A. W. Mair, M.A., Professor of Greek in the University of Edinburgh, and Fellow of Gonville and Caius College ; Mr Carl Hagenbeck, of Hamburg (for photographs) ; Mr C. W. Hawes, M.A., Trinity College (for the use of photographs) ; Dr J. Venn, F.R.S., President of Gonville and Caius College ; Mr J. A. Venn, B.A., Trinity College (for photographs) ; Dr M. M. Hartog, Professor of Natural History in Queen's College, Cork ; Mr Cecil Bendall, M.A., Professor of Sanskrit in the University of Cambridge, late Fellow of Gonville and Caius College; Mr Harold Littledale, M.A., Trinity College, Dublin, Professor of X PREFACE English Literature in University College, Cardiff; Mr Clement Gutch, M.A., Lecturer of St John's College ; Mr E. H. Minns, M.A., Fellow and Librarian of Pembroke College ; Dr A. C. Haddon, F.R.S., Fellow of Christ's College, University Lecturer in Ethnology, and late Professor of Zoology in the Royal College of Science, Dublin; Mr F. W. Green, M.A., Jesus College; Dr Scharfif, the Head of the Department of Natural History in the National Museum of Ireland, and Director of the Zoological Gardens, Dublin ; Dr Cecil H. Smith, Keeper of the Greek and Roman Antiquities, British Museum ; Mr A. H. Smith, M.A., Trinity College, the Assistant-Keeper of the same Department; Dr Budge, Christ's College, Keeper of the Department of Egyptian and Assyrian Antiquities, British Museum ; Dr Oscar Montelius, Director of the National Museum of Sweden ; Dr Brunchorst, Director of the Bergen Museum ; the Rev. J. Roscoe, C.M.S., Uganda, and Mrs Roscoe; Mr P. W. Sclater, F.R.S., late Superintendent of the Zoological Garden, Regent's Park ; Mr H. Platnauer, Curator of the York Museum ; Mr Gordon Taylor, Curator of the Elgin Museum ; Don Angel Cabrera, Madrid ; Mr W. W. Skeat, M.A., Christ's College; Mr J. Stanley Gardiner, M.A., Fellow and Dean of Gonville and Caius College ; Mr E. C. Quiggin, M.A., Fellow and Lecturer of Gonville and Caius College ; Mr R. C. Punnett, M.A., Fellow of Gonville and Caius College ; Mr Alfred Newton, M.A., F.R.S., Professor of Zoology in the University of Cambridge, and Fellow of Magdalene College; Mr William Bateson, M.A., F.R.S., Fellow and Lecturer of St John's College, Deputy for the Professor of Zoology; Mr J. Graham Kerr, M.A., Professor of Natural History in the University of Glasgow, late Fellow of Christ's College; Mr T. McKenny Hughes, M.A., F.R.S., Professor of Geology in the University of Cambridge, and Fellow of Clare College ; Dr Walter G. Headlam, Fellow and Lecturer of King's College ; Rev. T. J. Pulvertaft, M.A. ; Mr F. A. H. Marshall, M.A., Christ's College; Mr Theodore M. Davies (through Mr Percy E. Newberry); my brother-in-law, Mr Arthur Warren Samuels, LL.D., K.C. ; my niece, Miss K. F. Samuels (for photographs); Lt.-Col. Herbert Irwin, PREFACE XI Warvvickshire Regiment ; Mr G. E. Low, Foster Place, Dublin (for photographs) ; Mr Alfred J. Smith, Rendlesham, Suffolk ; Mr W. B. Redfern; and Prof. H. F. Osborn, Hon. D.Sc, Columbia University, U.S.A. For the loan of blocks I have to thank the Department of Agriculture in Ireland (through Mr T. P. Gill), the Council of the Royal Irish Academy, the Council of the Cambridge Philosophical Society, and the Council of the Cambridge Antiquarian Society ; and for permission to copy illustrations or to obtain electrotypes I am indebted to the Trustees of the British Museum, the Council of the Royal Zoological Society, the Council of the Royal Agricultural Society, the Council of the Shire Horse Society, the Council of the Hellenic Society, Messrs A. J. Holman and Co., and Messrs Longman and Co. WILLIAM RIDGEWAY. Fen Ditton, Cambridge, Lammas Day, 1905. CONTENTS. CHAPTER I. PAGE Introduction : The Ancestors of the Equidae . . 1 CHAPTER II. The Existing Equidae 12 CHAPTER III. The Horses of Prehistoric and Historic Times . . 82 CHAPTER IV. The Origin of the Libyan Horse 425 CHAPTER V. (Supplementary.) The Development of Equitation 478 Addenda 506 Index 513 LIST OF ILLUSTRATIONS. FIG. PAGE 1. Hind-leg of Arab of coarse type, with callosity like that of Prejvalsky's Horse ......... 13 2. Chestnut on right fore-leg of a Prejvalsky Horse .... 14 3. Hock Chestnut (right) of a Prejvalsky Horse .... 14 4. Hock Chestnut (right) of Iceland Pony of Cart-horse type . . 14 5. Ergot (Fetlock Wart) on left fore-leg of a Common Donkey . . 15 6. Ergot (Fetlock Wart) on hind-leg of a Common Donkey . . 1.5 7. Ergot on right fore-leg of a Chapman Zebra ..... 1.5 8. Ergot on left fore-leg of a Connemara Pony ..... 15 9. Ergot on right fore-leg of an Arab ...... 15 10. Ergot on hind-leg of an Arab ........ 15 11. Typical Yellow-dun ' Celtic ' Pony : North Iceland ... 17 12. Hind-leg (left) of ' Celtic ' Pony showing no hock callosity . . 19 13. Typical ' Celtic ' Pony : North of Iceland 20 14. Flat-nosed variety of 'Celtic' Pony: Hebrides .... 21 15. Black Hebridean Pony without hock callosities .... 22 16. A Faroe Filly . '. 23 17. Faroe Pony 25 18. Prejvalsky's Horse 27 19. Young Prejvalsky Horses and their Mongolian foster-mothers . 29 20. The Kiang 44 21. The Kiang 45 22. The Onager 47 23. Assyrians lassoing a Wild Ass ....... 49 24. The Nubian Wild Ass 50 25. The Nubian Wild Ass 51 26. The Somali Wild Ass 52 27. The Somali Wild Ass 53 28. The Somali Zebra 59 29. The Mountain Zebra 62 30. Head of Grant's Zebra 63 31. Skin of an unborn Foal of Grant's Zebra : Uganda ... 64 32. Skin of a young Grant's Zebra 65 33. Skin of a full-grown Grant's Zebra : Uganda ..... 67 84. Burchell's Zebra (Grant's variety) 68 35. Burchell's Zebra (Grant's variety) from Kilima Ndjaro ... 69 36. 'MatoiJO,' Prof. Ewart's Chapman Zebra 70 37. Typical Burchell's Zebra 71 38. The typical Quagga 73 39. Lord Morton's Quagga 77 LIST OF ILLUSTRATIONS XV FIG. PAGE 40. The Daniell Quagga 79 41. A Prehistoric Horse 84 42. Head of Prehistoric Horse : Gourdan . . . . . .85 43. Miniature Axe : Hallstatt 91 44. Eemaius of a Chariot found at Driffield, Yorkshire ... 96 45. Bronze Bits: Ireland ......... 98 46. Thracian Coin showing Ox-cart ....... 106 47. Grave-stone : Mycenae ......... 107 48. Norse Pony from the Isle of Eodo ....... 119 49. Norwegian Ponies near Mundal Glacier ...... 120 50. The last of the old Lofoden Ponies 121 51. An old Icelandic Horse-light ........ 123 52. Ancient Scythians taming Horses ....... 131 53. A Buriat Horseman .......... 135 54. A Group of Buriats ready for the Kace ...... 137 55. Buriat Women setting forth to Hill Shrine on a Feast-day . . 139 56. The Tangum of Tibet 155 57. Black Arabian of the Imam of Muscat ...... 173 58. A bay Arabian 181 59. The Horse of Anatolia 189 60. The Turk 191 61. Impression from the Signet of Darius Hystaspes .... 193 62. Tiglath Pileser III in his Chariot 195 63. Head of a Horse from the Chariot of Assur-bani-pal . . . 196 64. Assyrian Lion-hunt . . . . . . . . . . 197 65. Coin of Aemilius Scaurus ........ 202 66. Coin of Aulus Plotius 203 67. ' Hittite ' Bas-relief showing Lion-hunt 215 68. Seti I in Battle 217 69. Chariot found in a tomb at Thebes ...... 225 70. The Barbary Horse 241 71. The Moorish Horse 242 72. A Libyan Woman on Horseback 243 73. The Dongola Horse 249 74. Carthaginian Coins .......... 255 75. The Jennet of Spain 259 76. The Sardinian Horse .......... 275 77. Archaic Metope showing a Quadriga : Selinus .... 277 78. Fragment of sculpture from Tarentum 279 79. Fresco showing a Samnite Warrior : Paestum .... 281 80. The Neapolitan Courser 285 81. Vase from Enkomi, Cyprus ........ 288 82. Vase fragment from Enkomi, Cyprus ...... 289 83. Greek Horsemen from the Parthenon ...... 297 84. Head of the Horse of Selene : Parthenon ..... 299 85. Coin of Potidaea 301 86. Coin of Philip II of Macedon showing a Jockey on a Eace-horse . 302 87. Coin of Philip II of Macedon showing a Horse-soldier . . 303 88. Head of one of the Horses of the Quadriga from the Mausoleum . 305 89. Eomau Denarius .......... 308 90. The Lombard Horse 315 91. The Hungarian Horse ......... 319 92. Eoman Contorniates ......... 329 93. The German Horse 336 XVI LIST OF ILLUSTRATIONS FIG. PAGE 94. The Flanders Horse 337 95. The Horse of Cleves 338 96. The Horse of -Tuliers 339 97. The Friesland Horse 340 98. The Danish Horse 341 99. The East Prussian Horse 343 100. Sleii^nir, Odin's Horse 347 101. The English Horse 361 102. King WiUiam HI on a Great Horse 367 103. Shire Stallion, ' Stroxton Tom ' 369 104. Typical Suffolk Punch, 'Saturn' 375 105. Flying Childers 383 106. The Godolphin Barb 384 107. The Yorkshire Coach-horse 385 108. A Hackney I\Iare 387 109. Plinth of the Cross at Kells, Co. Meath 389 110. Irish Horseman : Book of Kells 391 111. Sword of La T^ue type in its sheath : Connantre, Marne . . 394 112. Iron Sword, La Tene type, in Bronze Scabbard : Hallstatt . . 394 113. Bronze Shield: Bingen 394 114. Bronze Fibula: Marne 395 115. Alderwood Shield : Ireland ........ 395 116. Bronze Fibula : Ireland ......... 396 117. Bronze Fibula: Ireland 396 118. Bronze Shield : Co. Limerick ........ 397 119. The O'Donovan Shield : Skibbereen 399 120. Yellow dun Connemara Pony ........ 403 121. A rich Yellow-dun Connemara Mare and Foal .... 404 122. Light-grey Connemara Pony ........ 405 123. New Forest Ponies 406 124. Light-grey Connemara Filly 407 125. Connemara Pony : Clifden district ....... 408 126. Connemara Gelding of larger type ....... 409 127. Connemara Pony used as a hunter ...... 410 128. Irish Draught Stallion 411 129. Irish Thoroughbred Stallion 415 130. Irish Hunter Gelding 417 131-3. The Elgin Quagga 438-9 134. The Muscovy Drake ' Hans ' and his hybrid offspring by a white Aylesbury duck .......... 461 135. Chapman's variety of Burchell's Zebra ..... 463 136. A Zebra-pony Hybrid 477 137. Coin of Messana showing Mule-car 489 138. Ancient Irish Rein-ring with ' late Celtic ' ornament . . . 493 139. Ancient Irish Rein-ring ......... 495 140. Medieval Persian Stirrup 499 141. Prick-spurs 500 142. Rowel-spurs 501 143. Old English Horse-shoes 503 CHAPTER I. INTRODUCTION: THE ANCESTORS OF THE EQUIDAE. Multaque turn interiisse auimantum saecla necessest nee potuisse propagando procudere prolem. nam quaecunque vides vesci vitalibus auris, aut dolus, aut virtus, aut denique mobilitas est ex ineunte aevo genus id tutata reservans. Lucretius v. 855-9. Next to the history of the various branches of the human race there is no more interesting and important subject for man's study than the origin and development of the breeds of domestic horses, the noblest of all the creatures that man has subdued to his will, and the acquisition of which has been, as will be shown presently, one of the chief factors in the rise and supremacy of the great nations of the ancient, medieval, and modern world. It has long been a matter of dispute among naturalists whether all our domestic horses have had a multiple or a single origin. Colonel Hamilton Smith ^ held that they are descended from five primitive and differently coloured stirpes — the bay (represented by the tarpan), the white, the black, the dun with a striped back (represented by the horses of the Ukraine), and finally the piebald stock of Tibet. M. Sanson^ went further and divided the Equides caballines of our actual epoch into eight species (especes) which have severally their 1 "The Horse," Naturalist's Library, Vol. xii. pp. 160 sqq. - M. Sanson first published his subdivisions in his " Nouvelle Determination des Especes Chevalines du genre Equtis" (6 Dec. 1869), Comptes-Rendus, lxix. pp. 1204-7; then in Migrations des Animaux Domestiques, p. 9, and in his Traite de Zootechnie (ed. 2), Vol. iii. pp. 9 — 105. K. H. 1 2 THE ANCESTORS [CH. own osteological types. The district where each is principally found is denoted by its Latin praenomen : — (1) Equus caballus asiaticus, (2) E. c. africanus, (3) E. c. germaniciis, (4) E. c. frisius, (5) E. c. helgicus, (6) E. c. britcmnicus, (7) E. c. hibernicus, (8) E. c. sequanius. Sanson divided all the horses hitherto known as ' Oriental ' or 'Arab' between his two first species — Asiaticus and Afri- canus, as he conceived that they had two separate places of origin denoted by the names which he assigned to them. The Asiaticus he conceived to have originated and been domesticated in central Asia, whilst from the existence of a peculiar breed of black horses, commonly with white feet, known as Bongolawi, from the fact that they are found round Dongola in Nubia, he was led to maintain that this breed had " originated in north- east Africa, probably in Nubia." He declares that there are distinct osteological differences between Asiaticus and Africanus, holding that the former has a flat forehead, and a straight chaffron, which gives its head a rectangular profile, that it has prominent orbits projecting beyond the plane of the forehead, a long head, a large chest, a round barrel, a large rounded croup, and a tail borne far from the body, whilst Africanus has a forehead rounded like the segment of a globe, and the lower part of the chaffron slightly convex, features which give its head a bousque or moutonne look ; the orbits are not salient, the ears are longer and are less divided apart at the base, the body not so capacious, the chest not so large, the sides less curved, the croup more like that of a mule, the tail carried near the body, the thighs always slender, and the legs longer than in Asiaticus, and it differs from the latter in the number of its lumbar vertebrae, and by the absence of hock callosities \ Sanson derives ultimately his remaining six classes from his asiaticus; several of them are known by other names, germanicus as Danish, his frisius as Flemish, whilst his hritannicus comprises the Norfolk or Black Horse, the Suffolk Punch, and in France the Boulonnais and Cauchois (Caux), his hibernicus includes all the ponies of the United Kiugd(jm and 1 Op. cit. (ed. 4), Vol. III. p. .52: "Les membres posterieurs sont dcjJourvus dc chataignes." l] OF THE EQUIDAE 3 the Breton in France, whilst finally his sequanius is identical Avith the Percheron so highly esteemed in France. M. Pietrement^ adopted Sanson's principle of an eightfold subdivision, but carefully restricted the term species to Equus caballus, describing the eight classes as races. But not be- lieving that the argument based on geographical distribution was of itself sufficient he rejected Sanson's africanus, and on the grounds that Sanson admitted that his asiaticus originated in central Asia, he assumes that it was primarily domesticated by the Aryans, and accordingly terms it aryanus to distinguish it from what he holds to be the other Oriental race (wrongly regarded as African by Sanson). As Pietrement considers that the latter was domesticated in central Asia by the Mongolian or Tartar-Finnish peoples, and that it was brought thence by the Hyksos into Egypt, he gives it the name of mongolicus. The aryanus and the mongolicus of Pietrement thus corre- spond respectively to the asiaticus and the africanus of Sanson. It will however be observed that Pietrement's nomen- clature is based on several unproved assumptions ; first, that the original home of the Aryans was in central Asia, secondly, that the Dongola horse was brought into Africa from Asia some two thousand years before Christ, thirdly, that it was the Hyksos who brought it there, and fourthly, that the Hyksos were Mongols. We shall presently see grave reasons for doubt- ing the validity of the grounds on which M. Pietrement has based his terminology. Darwin'^ rejected not only Hamilton Smith's five stirpes, but also Sanson's E. c. africanus on the ground that the latter involved the assumption " that osteological characters are subject to very little variation, which is certainly a mistake," and he was thus inclined to follow those naturalists who " from the fertility of the most distinct breeds when crossed, look at all the breeds as having descended from a single species," and he held^ "that it is not probable that each larger breed, which in the course of time has supplanted a previous and smaller 1 Les Chevaux prehistoriques et historiques, pp. 13 sqq. 2 Variation of Animals and Plants under Domestication, Vol. i. pp. 53-J: (ed. 2), 3 Ibid. Vol. II. p. 423. 1—2 4 THE ANCESTORS [CH. breed, was the descendant of a distinct and larger species : it is far more probable that the domestic races of our various animals were gradually improved in different parts of the great European-Asiatic continent, and thence spread to other countries." He thus left Africa out of account as a possible source for a race of horses. It will be observed that those who hold a single oiigin for all domestic horses base their belief on " the fertility of the most distinct breeds when crossed." Yet this cannot be regarded as a true criterion, for animals which are admittedly distinct species, such as the dog, the wolf, and the jackal among carnivores, and the common ox, the zebu (Bos gaurus), and the yak (Bos grunniens) among herbivores, freely interbreed and produce fertile offspring. But though Darwin leaned to the belief that all our horses come from a single stock, he carefully pointed out that, " as several species and varieties of the horse existed during the later Tertiary period, and as Rutimeyer found differences in the size and form of the skull in the earliest known domesticated horses, w^e ought not to feel sure that all our breeds are descended from a single species\" He elsewhere^ points out that " as the savages of North and South America easily reclaimed the feral horses, so there is no improbability in savages in various quarters of the world having domesticated more than one native species or natural race." Since Darwin wrote it has been generally held that all our domestic horses have had but a single source, whether they be the fine horses of slender build and great speed, of which the Arab is the type, or the heavy cart-horses, whose origin is commonly found in the coarse, thickset horses of Europe and upper Asia, of which the unimproved Mongolian pony is the repre- sentative, or hunters, roadsters, carriage-horses and trappers, which are as everyone knows, the result of a judicious blending of the two first-mentioned classes. Thus M. Sanson^ now holds that all our domestic breeds had a single origin, and divides recent Horses into two groups — long-headed and short-headed — 1 Variation of Animals and Plants under Domestication, Vol. i. p. 53. 2 Ibid. Vol. I. p. 54. ^ Traite de Zootechnie (ed. 4, 1901), pp. 2, 3. IJ OF THE EQUIDAE 5 each of which consists of several races, while Capt. M. H. Hayes ^ maintains that " no breed of horses possesses any distinctive characteristic which serves to distinguish it from other breeds." But on Dec. 2nd, 1902, Prof J. Cossar Ewart, F.R.S., read before the Royal Society of Edinburgh a paper in which he pointed out the existence in the Western Islands of a variety of horse hitherto unnoticed. A week earlier the present writer had laid before the Cambridge Philosophical Society a summary of the evidence which led him to conclude that the hitherto generally received notion that the Arab horse was the ultimate source of our thoroughbred and half-bred horses had no historical foundation, that the Arabs had only got their fine breed of horses from North Africa at a period later than the Christian era, and that on the other hand there was the clearest evidence of the existence in Libya of a fine breed of horses for a thousand years before the Arabs ever bred a horse, and finally it was maintained that from this North African stock all the best horses of the world have spnmg, and that it is a variety entirely distinct from the clumsy, thickset, slow horses of Europe and Asia. The object of the following pages is to set out at length the evidence for the conclusion just stated, and to trace the important part played by this Libyan horse and its descendants in the history of the world. It was only at a comparatively late epoch in the history of mammals that the ancestors of the horse made their first appearance, for it is not until the Tertiary period that hoofed animals begin to occur. It is among two extinct families of the Perissodactyles — the Lophiodontidae and the Palaeotheriidae — that we meet what appear to be the earliest ancestral forms of the horses and the tapirs of to-day, though it cannot be affirmed that an unbroken line of descent from any forms yet known can be made out for the existing Equidae. Yet we can at least point to a series of forms, the salient osteological features of which have led to a belief in the relationship of our horses to these primeval Perissodactyles ^ We may start with certain 1 Points of the Horse (Sril ed., 1904), p. 422. 2 Beddard, Mammalia (1902), pp. 247-8; Flower and Lydekker, Mammals (1891), p. 380. 6 THE ANCESTORS [CH. forms in the Eocene of both Europe and America. In the Phenacodiis of the American Lower Eocene the feet still retain the primal five digits, whilst in the Eocene of both Europe and America occur the diminutive Hyrcwotherium, which had its fore-limbs four-toed, but its hind three-toed (as is the case with the tapir, which is in many respects the most ancient of existing forms referable to the Perissodactyle order), and also Eohippus belonging to the same sub-family, but which is slightly more primitive, as its hind feet have a vestige of the first digit. Pachynolophus (or Orohippus), found in both Europe and America, shows molars somewhat more advanced towards the equine type. From this last form the Anchitherium found in the Upper Miocene is not far removed in structure ; but, though it is a little nearer to the horse in several respects, it is not now considered to be in the direct line of descent, as it is considerably larger than some succeeding forms. Since both Hyracotheriimi and Pachynolophus occur in both the Old and New Worlds, from them may have sprung the true horses of both hemispheres. But from this point there is now a bifur- cation, for Mesohippus, the next step towards Equus, is as yet only known in America, as is also the case with its successors Miohippus, Desmatippus, and Protohippus. The last-named (found in the Lower Pliocene), and which was about the size of a modern donkey, had three toes on each foot. As Mesohiptpus has not been discovered in the Old World we are left only with Anchitherium (already described) and Hipparion (which had come from America) in that area. The latter was very widely distributed, occurring not only in North America, but also in Asia, Europe and Africa. Its remains have been recently found in considerable numbers at Pikermi near Athens (a fine specimen from which place is now in the National Museum of Natural History), in the isle of Samos, and in Egypt. In the typical North American and European forms there were three digits, but in the Indian Hipparion antelopinum the lateral digits seem to have disappeared. We have already given reasons for not placing Anchitherium in the direct line of ancestry of the horse, and zoologists now hold that Hippai'ion must likewise be excluded. It became extinct probably owing to excessive specialization. l] OF THE EQUIDAE * 7 Three years ago the American Museum set on foot under the direction of Prof. Henry Fairfield Osborn a special exploration into the fossil history of the horse. " The object was to connect all the links between the Lower Eocene five-toed, and Lower Pleistocene one-toed horses and to ascertain the relations of the latter to the horses, asses, and zebras of Eurasia and Africa. The first great result obtained is the proof of the multiple nature of horse evolution during the American Oligocene and Miocene. Instead of a single series as formerly supposed there are five, one leading to Neohipparion the most specialized antelope- like horse which has ever been found, a second of inter- mediate form probably leading through Protohippus to Equus as Leidy and Marsh supposed, a third leading to the Upper Miocene Hypohippus, a persistently primitive probably forest or swamp-living horse with short crown teeth adapted to browsing rather than grazing, and three spreading toes ; this horse has recently been found in China also. A fourth and fifth line of Oligocene-Miocene horses became early extinct. This poly- phyletic or multiple law," says Prof. Osborn S "is quite in harmony with the multiple origin of the historic and recent races of horses as recently established by Ridgeway and Ewart. The Pliocene horse of America still requires further exploration before we can positively affirm either that all the links to Equus are complete, or that America is indubitably the source of this genus. The Lower Pleistocene of America exhibits a great variety of races ranging in size from horses far mote diminutive than the smallest Shetland to those exceeding the very largest modern draught breeds. Yet all these races became extinct, not surviving into the human period, as was the case in South America. The relation of these North American races to those of South America and of Asia and Africa is again a subject requiring further investigation in which it is necessary to exercise the most extreme accuracy." In the recent Equidae each foot consists of a single complex digit, but digits li. and IV. are complete in the embryo and also survive though degraded in the adult, and there is a callosity 1 "Evolution of the Horse," a paper read before the British Association (Section D), Tuesday, Aug. 23, 1904 (Nature, 22 Sep. 1904, p. 520). 8 THE ANCESTORS [CH. (termed castor or chestnut) on the inner side of the fore-leg above the carpus ; the tail is furnished with long hair, either at the end or throughout its whole length. The lateral digits sometimes survive to a considerable extent, as was apparently the case with Julius Caesar's favourite charger^ Fossil remains of horses are found abundantly in deposits of the most recent geological age in almost every part of America, from Escholtz Bay to Patagonia. According to Sir C Lyell- remains of no less than twelve species referred to seven genera have been discovered in the Pliocene and Pleistocene formations of that country. Recent investigations show that North America in pre-glacial times possessed at least nine perfectly distinct wild species of Equidae. These varied much in size ; thus Equus complicatus of the Southern and middle Western States and E. occidentalis of California were as large as small cart- horses, E. tau of Mexico was extremely small, whilst others, such as E. fraternis of the South-eastern States, were inter- mediate. " Some of the American pre-glacial Equidae were characterised by very large heads and short, strong limbs, some by small heads and slender limbs ; and though the majority conformed to the true horse type, two or three were constructed on the lines of asses and zebras^" Yet no horses, either wild or domesticated, existed in any part of America at the time of the Spanish conquest, which is all the more astonishing having regard to the very favourable conditions of soil and climate as demonstrated by the thousands of horses now ranging the Pampas of South America, all descended from seven stallions and five mares introduced by the Spaniards*, whilst the mustangs of Texas, sprung from a like small beginning, prove that North America was no less suited to be the nurse of horses. Dr Munro^ has ingeniously suggested that a satisfactory 1 Pliny, N. H. viii. 42. 64. - Principles of Geology, Vol. ii. p. 340 (11th ed.). * J. C. Ewart, "The Multiple Origin of Horses and Ponies," TransactioiiD of Highland Society, 1904, p. 2. ■* Azara, Natural Hintory of the Quadrupeds of Paraguay and the River La Plata (Eng. trans.), pp. 4 — 5. ^ " On the Prehistoric Horses of Europe and their supposed domestication in Palaeolithic Times," Archaeological Journal, Vol. lix. pp. 112-3. l] OF THE EQUIDAE 9 explanation of the course which led to the extinction of the American horses will be found in the fact that after coming to the end of their evolutionary tether in the attainment of speed — the sole means by which they could escape from their enemies — they fell an easy prey to one or more of these animals, who meanwhile had succeeded in improving their methods of warfare in the struggle of life. But it is obvious that if the great carnivorae had exter- minated the horses, and thereby brought about their own destruction, they would certainly have eaten up the bisons and tapirs before they themselves had perished of hunger, for it cannot be supposed that these animals escaped because they were fleeter of foot than the Equidae. It may well be that the destruction of these American horses was due not to the continual ravages of mighty carni- vores, but to the insidious inroads of far meaner foes, for we must not forget that there are no feral horses in Paraguay, because an Hippohosca or an Oestrus attacks the umbilical region of young foals, and produces ulcers, which invariably cause death unless human aid is interposed ^ I do not for a moment suggest that the extinction of all 1 Col. Hamilton Smith, "The Horse," Naturalist's Library, Vol. xii. p. 248, Edinburgh, 1841. Though Azara does not mention this in his Spanish version (from which the English trans, was made), yet (English trans., p. 66) speaking of the wild cattle of Paraguay he says that "from August to January, which is the calving time, the cows are driven in Paraguay twice a week to the rodeo, in order to free them from a certain worm which infests them, more especially the calves, at the umbilicus, and to such an extent that, without this assistance, they would inevitably perish. The same malady occurs in Corrientes and the Pueblos of the Missions : but in Monte Video and Buenos Ayres it is so little known that it demands no particular attention ; nor are the herds during the above-mentioned months collected so frequently as usual, for the pregnant cows might be injured thereby, and many of the young calves would be lost." My friend Prof. Graham Kerr writes as follows: "The fly appears to be the ordinary blow-fly, which lays its eggs in the drying-up end of the severed umbilical cord and on the blood round it. The larvae hatch out in a few hours and cause ulceration and the death of the calf. Estancieros regularly round up the cattle every few days, and dress the calves affected with medicated glycerine. I have not personally seen the maggots on calves, but I have seen them on adult cows." Darwin cites Azara for the statement about the horses, but he used Azara's French edition, and probably Col. Smith did the same, as we shall see later on. 10 THE ANCESTORS [CH. the Equidae on the American continent was due solely to the insect scourge of modern Paraguay. But as it is clear that though lions abounded in South Africa and preyed largely on zebras, they never threatened extermination to the horse family in Africa, whilst on the other hand the tsetse-fly and horse sickness are as deadly to Equus caballus in certain areas of that region as is the insect pest of Paraguay, it seems far more probable that the extinction of the horses of North and South America was due to the inroads of mean and obscure forms of life rather than to the onslaughts of the great flesh-eating monsters of the young world's prime. It is generally admitted that the ancestors of the living Equidae passed from America into the Old World, for before the Ice Age it was perfectly possible for American horses to cross into Asia by land bridges in the vicinity of Behring's Straits ; thence they extended into Europe, and finally reached Africa either from Asia or by the land bridges which then linked Europe to North Africa. " One of the earlier immi- grants, Equus stenonis, has left its remains in Pliocene deposits of Britain, France, Switzerland, Italy, and the north of Africa. While E. stenonis was extending its range into Europe and Africa, two others, E. sivalensis and E. namadicus, were finding their way into India, and yet other species were doubtless settling in Eastern Europe and Central Asia\" Thus, as Africa now contains several species of zebras, so Europe at the Pleistocene period was inhabited by several species of horses. Some palaeontologists believe that the Indian species E. siva- lensis and E. namadicus became extinct, and that E. stenonis gave inse through one variety (E. 7'ohustus) to the modern domestic breeds, and by another {E. ligeris) to the Burchell group of zebras. Hipparion and certain prehistoric South American species were characterised by a fossa or depression in front of the orbit for a facial gland (probably similar to the scent gland in the stag), found also in E. stenonis- and its later ally E. quaggoides and in E. sivalensis (cf p. 1 50). In some modern horses, which have so-called Eastern blood 1 J. C. Ewart, op. cit. p. 3. 2 K. W. Lyclekker, Proc. Zool. Soc. (1904), Vol. i. p. 427. l] OF THE EQUIDAE 11 in their veins (for instance the race-horse Bend Or) there seems to be a vestige of the pre-orbital depression'. Again, Equus sivalensis was usually characterised by large first pre-molar (wolf) teeth in the upper jaw, whilst large functional first pre- molars are found in some horses of South-eastern Asia {e.g. in Java and Siilu ponies) and in some zebras, as for instance Grevy's zebra and in a zebra of the Burchell type found near Lake Baringo. It is hence held by some that lineal, hut somewhat modified descendants of E. sivalensis of the Indian Pliocene period still survive, and that E. sivalensis was a lineal de- scendant of Hippainon. But it will be presently found that the horses of Java and Sulu have no pretensions to be regarded as aboriginal. Osseous remains show that horses were widely distributed over Europe in the Pleistocene period, but it has not 3^et been determined how many species of horses inhabited Europe during and immediately after the Glacial period, nor yet from which of the pre-glacial species prehistoric horses were descended. Bones and teeth from deposits in the south of England seem to indi- cate that during the Pleistocene period several species of horses ranged over Western Europe. The Pleistocene beds of Essex yield bones and teeth of a large-headed, heavily-built horse, which probably sometimes measured over fourteen hands high. From the ' Elephant bed ' at Brighton, portions of a slender- limbed horse have been obtained : and Kent's Cave, near Torquay, has yielded numerous fragments of two varieties or species, which differed somewhat from the Brighton and Essex species. One of these in its build approached the Essex horse, the other the slender-limbed species of the Brighton ' Elephant bed.' Although the latter has hitherto been described as very small, according to Prof Ewart, if we are to judge from the bones in the British Museum, "it may very well have reached a height of 12"2 or 13 hands." If there were two or more species in the south of England, which then formed part of the Continent, " it is probable that yet other species inhabited South and Middle Europe and the North of Africa-." 1 E. W. Lydekker, loc. cit. ^ J. C. Ewart, op. cit. p. 4. CHAPTER II. THE EXISTING EQUIDAE. Hark ! I hear horses. Macbeth, iii. 8. The early ancestors of the horse which first crossed from the western hemisphere into Asia have had a much more successful career than their American cousins, for in spite of various gi'eat flesh-eating animals which once preyed upon their ancestors in Asia and Europe, and still continue to do so down to our own day in Africa, the genus Equus at this present moment (without including the probably extinct quagga) comprises at least some fifteen species or subspecies: — (1) Equus caballus (the horse) ; (2) E. caballus celticus (the Celtic pony discovered recently by Prof Ewart); (3) E. przeiualskii (Prejvalsky's horse); (4) E. kiang (the kiang); (5) E. onager (Indo-Persian wild ass) ; (6) E. hemippus (Syrian wild ass); (7) E. asinus (the African wild ass) ; (8) E. somalicus (the Somali wild ass) ; (9) E. grevyi (Grevy's Imperial or Somali zebra); (10) E. zebra (Mountain zebra); (11) ^. cratvshayi; (12) E.foai; (13) ^. granti (Grant's zebra); (14) E. chapmani (Chapman's zebra); (15) E. burchelli (Burchell's zebra); (16) E. quagga (the quagga). HORSES. We have seen that in the true Equidae each foot consists of a single complex digit, but with vestiges of the second and fourth digits, and there is a callosity on the inner side of the fore-leg above the carpus. Besides this there is in the common horse {E. caballus) also a callosity on the back and lower surface of each fetlock joint in the centre of the tuft of hair which covers that part, and a callosity on each hind-leg a CH. Il] THE EXISTING EQUIDAE 13 little below the true hock joint and immediately over the cuneiform bone (Figs. 1 — 4). The callosities near the knee and hock are termed chestnuts or castors, and those under the fetlock ergots. The chestnuts on the fore-legs of all Equidae are more or less oval in form, those of ordinary domestic horses being usually about 2 inches long. The hind chestnuts (which Equus caballus almost alone of the family possesses) are somewhat similar in shape to the fore ones, though a little smaller and narrower (Fig. 1). The ergots in all members of the genus are more or less round (Figs. 5 — 10), and in ordinary domestic horses are less than a quarter of the size of the chestnuts. As Prof. Ewart has shown, the front chestnuts cor- respond to the wrist pads, the ergots to the middle portion of the trilobed sole pad in the dog and cat. Other zoologists^ hold that the chestnuts are the re- mains of scent glands, similar to those found in some species of deer and other animals. According to M. Sanson the absence of the hind chest- nuts is of frequent occurrence among the horses and ponies of North Africa, although they are almost invariably present in ordinary breeds : in very rare instances there are no chest- nuts on the fore-legs of domestic horses. Though ergots are generally present, Captain Hayes has " noticed their frequent Fig. 1. Hind leg of Arab of coarse type, with callosity like that of Prejvalsky's horse. 1 Sir W. H. Flower, The Horse, p. 170 (but he held that the ergot on the hinder aspect of the horse's pastern ajjpears to represent one of the pads, which are still functional in the foot of the tapir). Mr E. W. Lydekker, F.R. S. (Proc. Zool. Soc, 1903, pp. 199 — 203) has attempted not "so much to show what the equine callosities represent," but rather "from palaeontological considerations the imjDrobability of their being vestigial foot-pads." 14 THE EXISTING EQUIDAE [CH. absence in pure-bred Arab horses and in thoroughbreds." " The nearer a horse approaches the heavy draught type, the thicker is the growth of the callosities on his legs\" We shall presently find (p. 19) that in the typical 'Celtic' pony the hock callosities are wanting, as is not unfrequently the case in North African horses, the front chestnuts are small, whilst /T-^ "'•""n-' :"!*'■' Fig. 2. Chestnut on right Fig. 3. Hock Chestnut Fig. 4. Hock Chestnut fore-leg of a Prejvalsky horse (right) of a Prejvalsky (right) of Iceland ponj {natural size). horse (natural size). of cart-horse type (na- tural size). the ergots have entirely vanished, as in the pure-bred Arabs and thoroughbreds noticed by Capt. Hayes. Through the kindness of Prof. Ewart, I am enabled to figure (full-size), from drawings made from his own animals, examples of chestnuts and ergots in various Equidae (Figs, 1 — 10). 1 Hayes, Poiiits of the Horse (ed. 3), p. 319. THE EXISTING EQUIDAE '1- 15 Fig. 5. Ergot (fetlock wart) on left fore- Fig. 6. Ergot (fetlock wart) on hind-leg leg of common donkey {natural size). of common donkey [natural size). .7, "v'-i^. Fig. 7. Ergot (fetlock wart) on right Fig. 8. Ergot (fetlock wart) on left fore-leg of a Chapman zebra (natural size). fore-leg of a Connemara pony (natural size). .rfm.': '#H, mil Fig. 9. Ergot (fetlock wart) on right Fig. 10. Ergot (fetlock wart) on hind- fore-leg of an Arab {natural size). leg of an Arab {natural size). 16 THE EXISTING EQUIDAE [CH. (1) Equus caballus has hitherto been distinguished from the rest of the family by the tail being covered with long hairs from its base to its end, and by having chestnuts on the inner sides of its hind-legs, as well as on its fore-legs. But recent investigations have rendered this statement no longer accurate, as will be made clear when we come to speak of Ewart's Celtic pony, and Prejvalsky's horse. The Equus caballus has a forelock, a longer mane, and shorter ears than its relatives the asses and zebras, whilst in proportion to its size its limbs are longer, its head smaller, and its hoofs broader. A further distinction between the horse and the other Equidae, first pointed out by Tegetmeier and Sutherland*, is the length of the period of gestation, which in the horses is eleven months, whilst in the asses and zebras it exceeds twelve months. It is a matter of dispute whether the true Equus caballus survived in a wild state in Europe down into the historical period, for although Pliny^ declares that the north of Europe produces troops of wild horses, just as Asia and Africa produce wild asses, and Strabo'' states that wild horses were found in the Alps, and also enumerates them among the wild animals of Spain, it has been maintained that these were not indigenous, but merely the descendants of domesticated horses which had run wild. There are abundant records of the existence of wild horses in upper Europe in not only the early but late Middle Ages^. Thus St Boniface was rebuked by Pope Gregory III. (a.D. 732) for permitting his German converts to eat the flesh of wild horses as well as of tame, and wild horses were ap- parently eaten by the monks of St Gallon about a.d. 1000. In a Westphalian document of 1316, the fishing, game, and wild horses of a certain forest are assigned to one Herman, and there seem to have been wild horses in the Vosges in Merovingian times and even at the end of the 16th century. There were wild horses in Pomerania in the 12th century and there were at the same period wild horses in Silesia, from which Duke ^ Horses, Asses, Zebras, Mules, and Mule Breeding, p. 2, 2 N. H. VIII. 16. ^ 207, 163. * Hehn, Cultivated Plants and Domestic Animals, pp. 37-8. II] THE EXISTING EQUIDAE 17 Sobeslaus carried away herds of wild mares not a few iu A.D. 1132, whilst there is good proof that the woods of Prussia contained wild horses down to the time of the Reformation, Fig. 11. Tyjjical yellow-dun 'Celtic' pouy; North Iceland. The Teutonic Knights hunted wild horses and other game chiefly for their skins, and Duke Albert in 1543 sent an order to the commander at Lyck bidding him to take measures for R. H. 2 18 THE EXISTING EQUIDAE [CH. the preservation of the wild horses, whilst far into the 17th century the horse was hunted in Poland and Lithuania. It is almost certain that many of these horses were merely feral, but it is quite possible that some of the troops contained a genuine wild element, though greatly modified by being crossed with escaped domestic animals. But, as it will soon be seen that one, if not two genuine wild varieties have survived in eastern Europe and eastern Asia down to our own times, it is probable that troops of genuine wild horses may have lingered in parts of Europe down to a comparatively recent date (cf. p. 348). Naturalists not unreasonably view with suspicion the sup- posed primitive wildness of the tarpan and Prejvalsky's horse, inasmuch as the haguals of South America, the mustangs of North America, the hrumhies of Australia, the kumrahs of Nigeria, and the muzins of Tartary are admittedly feral, whilst it is not unlikely that the same holds true of the wild horses of Northern Tibet. (2) E. caballus celticus. Prof. Ewart\ in the paper already mentioned, called attention to the existence of a distinct variety of horse, to which he has provisionally given the name of Equus caballus celticus. It is a true pony, and not a dwarf horse ; it has a small head, with prominent eyes, small ears, a heavy mane, slender limbs, small joints, and well formed small hoofs. It has similar characteristics to those Arabs which have no ergots, and at the most only minute hock callosities, but with the essential difference that instead of having long hair up to near the root of the tail, the hair on the upper part of the tail forms a fringe or taillock (Fig. 11). It has been found in Connemara and the north of Ireland, in Barra and other islands of the Outer Hebrides, and seems to have been common at one time in the island of Tiree, where ponies are now extinct, whilst the same characteristics are observed in many of the ponies imported into this country from Iceland and the Faroe Isles, a fact of considerable importance when we come to discuss the history of the horses of the North at a 1 Times, Tuesday, 2 Dec, 1902, p. 10; for a fuller abstract see Nature, Vol. Lxvii. (1903), p. 239 ; "The Multiple Origin of Horses and Ponies " (Tram, of Highland Society, 1SJ04), pp. 19 sqq. "] THE EXISTING EQUIDAE 19 later stage in this work\ Prof. Ewart thinks that there is some evidence of its occurrence in the New Forest, and he holds that it is " conceivable that the Celtic pony in its present form never existed in the East, but that it is a modified descendant of a small horse, which left the ancestral home in Central Asia and reached Europe long before the arrival of neolithic man." He pointed out that the drawings in the Dordogne caves suggest the existence of a small horse that might very well correspond to the Celtic pony, and further, that in Pleistocene deposits bones had been found of two kinds of horses, one a horse with small head, slender limbs, and small teeth, which again suggested the Celtic pony. In the Celtic pony not only are the hock cal- losities wanting (Fig. 12), but the front chestnuts are small, and, still more remarkable, the fetlock callosities (ergots) have entirely vanished : in asses and zebras the ergots are always present, and in some cases still play the part of pads. The Celtic pony is hence not only more specialized — further re- moved from the primitive type — in its mane and tail, but also in having got rid of the fetlock pads (ergots) and the hock (heel) callosities. Capt. Hayes has frequently noticed the same absence of erg-ots in North African and Arab horses. " Except in size I have been unable to discover any difference between the skeleton and teeth of the Celtic pony and the small horse of the ' Elephant bed ' of the Brighton Pleistocene. 1 Mr F. H. A. Marshall, B.A., Christ's College, Cambridge, has recently noticed a Welsh pony without hock callosities {Nature, 13 Aug. 1904). •2 9 Fig. 12. Hind-leg (left) of 'Celtic' pony showing no hock callosity. 20 THE EXISTING EQUIDAE [CH. In the northern part of Iceland, where the few pure specimens of the Celtic pony survive, only a height of twelve hands (48 inches) is reached — under more favourable conditions the height would probably be 50 to 52 inches, the size of some of the ' Elephant bed ' horses and the smaller variety of the desert- bred Arab to which the small slender-limbed occidental pony closely approximates." The Celtic pony learns rapidly what the trainer wishes, and responds with alacrity. " In a few days its education is completed" Fig. 13. Typical 'Celtic' pony: North Iceland. Ewart regards the pony selected as a type (Fig. 13) as an almost pure representative of a once widely-distributed species. The pony in question proved sterile with stallions belonging to five different breeds, as well as with a Burchell's zebra and akiang; but she at once bred when mated with a yellow-dun Connemara-Welsh pony, which closely approximates to the Celtic type, and she has this year been successfully mated with a Hebridean black Celtic pon}^ (Fig- 15). 1 J. C. Ewart, op. cit. p. 25. 11] THE EXISTING EQUIDAE 21 In a recent paper, based partly on their own observations in Iceland and the Faroe Islands, Mr F. H, A. Marshall and Mr N. Annandale' support Prof. Ewart's doctrine and give some useful details as regards the form, size, and colour of the ponies in those two regions, as well as some interesting historical details, to which we shall refer at a later page (p. 416). Though I'iG. 14. Flat-uosed variety of 'Celtic' pony: Hebrides. in recent times some efforts have been made to improve the Iceland ponies by the introduction of Norwegian stallions, it is probable that the majority used for stud purposes are still of 1 "The Horse in Iceland and the Faroes," Proc. Camb. Phil. Soc, Vol. xii. (1903), pp. 300-1. 22 THE EXISTING EQUIDAE [CH. pure Icelandic bloods " In the case of the Faroes the Norwegian stallions have only been introduced, apparently, within the last ten or twelve years, and in spite of this admixture, of which it does not appear that Prof. Ewart was aware, the general characters of the majority of the Faroe and Fig. 15. Black Hebridean pony without hock callosities. Icelandic ponies are those of the ' Celtic type ' (Fig. 13). This was made evident by an examination of a number of Faroe ponies 1 " The Horse in Iceland and the Faroes," Proc. Camb. Phil. Soc, Vol. xii. (1903), piD. 298-9. Prof. Ewart doubts " if Norwegian ponies have recently been introduced into Iceland." "] THE EXISTING EQUIDAE 23 in Thoi'shavn this summer (1903), for out of eleven individuals only one had large hock callosities, and this animal, though of small size, differed from the others in being exceptionally clumsy in build, with a large head, strongly suggestive of a cart-horse. The remaining ten ponies either had the hock callosities much reduced in size, or, as in the case of two indi- viduals, had no hock callosities. The height of these ponies Fig. 16. A Faroe filly ^ varied from about eleven to thirteen hands. A number of Icelandic ponies, averaging about thirteen hands high, were seen on board ship on their way from Reykjavik to Denmark, and of these six were examined and found to have no hock callosities ; while another, which had been imported into the Faroes, had the same peculiarity." 1 From a photograph kindly given me by Prof. Ewart. 24 THE EXISTING EQUIDAE [CH. " About a dozen Faroe ponies have very recently been imported into this country, and of these fully one-third have no hock callosities, while the others have them very small. All these ponies were characterised by having short hairs in the upper part of the tail. Of two Icelandic ponies, also recently arrived in Scotland, the hock callosities are absent in one and reduced in the other, while the tail characters are similar to those of the Faroe ponies." The authors give an illustration, here reproduced (Fig. 17), of a Faroe pony of a better type, which was stated by a native to " closely resemble the animals Avhich existed in the Faroes before the recent introduction of Norwegian blood." In this animal the ' Celtic ' characters strongly predominate, the shorter hairs in the upper part of the tail being especially noteworthy, as this character is occasionally absent in the Norwegian cross-breeds \ " So far," write Messrs Marshall and Annandale^ " as we have been able to discover, the chief, if not the only difference between the Icelandic and Faroe breeds, while they remained jjure, was that of colour, for while the former was, and still is, typically either light dun, with a dark line down the centre of the back and often with dark transverse stripes on the legs ; the Faroe ponies, according to Landt, a most trustworthy observer, were, at the beginning of last century, generally red, and occasionally black, the skewbalds sometimes seen among them at the present day being possibly descended from Icelandic ancestors." Mr Daniel Brunn^ in a very valuable little work on the ponies of Iceland, the Faroes and Greenland, gives numerous illustrations showing the various types and colours of these animals, and according to his statements the Icelandic ponies can hardly now be described as " typically light dun, since there are many skewbalds, chestnuts and bays." These ' Celtic ' ponies of Iceland and the Faroes are, as we shall see below, very different in form from the now extinct ponies of the 1 " The Horse in Iceland," loc. cit. p. 301. Messrs Marshall and Annandale, and the Council of the Cambridge Philosophical Society, have most kindly allowed me to use their block. ^ loc. cit. ^ Hesten i Nordboernes Tjencste paa Island, Faer^erne og Grfinland, Saertryk af " Didsskrift for Landokouomi " (Kjobenhavn, 1902). "] THE EXISTING EQUIDAE 25 i f* it' 26 THE EXISTING EQUIDAE [CH. Lofoden Isles (p. 121, Fig. 50), and from those of the isle of Rodo off the coast of Norway (p. 119, Fig. 48). (3) Prejvalsky's Horse {E. przewalskii Poliakoff) is so called from the fact that a skin was presented to that traveller by the chief magistrate of Zaizan ; the latter had obtained it from Kirghis who hunt wild camels in the deserts of Central Asia. It was first described by Poliakoff in 1881. It has callosities on its hind-legs and its hoofs are like those of Eqiiiis caballus^, but it differs from the latter in having a short, erect mane, no forelock, and by the tail, on which the long hairs only begin at the lower third of the dock instead of at the root. The ears are of a moderate size, and Poliakoff maintained that it was a distinct species. Ten years later the brothers Grijimailo found many of these animals in the desert of Dzungaria, and shot three stallions and one mare. Recently the energy of Mr Carl Hagenbeck of Hamburg (commissioned by the Duke of Bedford), has enabled naturalists to study living specimens for themselves. His agents, who employed nearly 2,000 Kirghis for the purpose, captured thirty- two foals (17 stallions, 15 fillies), which were fostered by common Mongolian mares (cf. Fig. 19). Mr Hagenbeck's account of these young animals, accompanied by an illustration, the first taken from a living specimen, was published by Mr Tegetmeier^ " The young wild horses were obtained from three districts, and, according to their descent, certain variations in colour are to be distinguished. The districts where they were caught are south of the Mongolian town Kobdo, long. 93" E. (Greenwich). To the west the territory is a large plain, of which the great Altai mountains are the eastern frontiers. The northern and southern frontiers are formed by two rivers, both of Avhich flow from the Altai, the Kui-kuius in the north, the TJrungu in the south. The plain is bounded 200 miles from Kobdo by the Tusgul Sea, into which both rivers discharge. The foals in this territory are of the following colours : directly after their birth, the head, the ears, the neck, the 1 The hoofs are long as in 'Arabs,' not broad as in cart-horses. 2 Field, 31 Ang. 1901, p. 391. Mr Hagenbeck also supplied a map showing the localities mentioned. n] THE EXISTING EQUIDAE 27 shoulders, the collar, the back, and the croup are light red, passing into whitish colour. The nose, the throat, the chest, the belly, and the legs are of whitish colour. The white colour blends with the upper colour on the middle of the body. The mane consists of light red-brown hair, the eel-back is marked pink-coloured ; it ends in the tuft of the tail, the curled hairs of which are light red-brown, white, and black. The upper short-haired portion of the tail shows a whitish colour. The Fig. 18. Prejvalsky's horse. lower jaw beard consists of reddish hair, which is about six inches long. The coat is smooth, except the croup, which bears curled hair. There are slightly-marked cross stripes on the withers. The eye has a whitish iris. " The second and middle territory is situated about 200 miles south from Kobdo. The Altai mountains surround this plain. The foals caught here bear light ashy-brown hair on the upper part of the body, marked similar to the former foals. The nose 28 THE EXISTING EQUIDAE [CH. is white as well as the under side, just the same as the foals from the west, only the outside of the legs is slightly tinted, and the fetlocks are black. The mane and the spine are a deep brownish colour and the beard also. They also have the cross stripes, and stripes are to be seen in parallel lines at the shoulders. The skin is smooth, except the light curled mane. The eye has a darkish iris. " The third variety of the horses come from the territory of the Zagan-norr Lake only, a small plateau on the southern branch of the Altai mountains, about 100 miles in a south-east direc- tion from Kobdo. The coat of these foals has a pale, full yellowish-brown colour, only interrupted by the white belly hair, and the distinct black bands at the outside of the legs from the black hair of the fetlocks to above the hocks. The nose is whitish. The mane and the curled hair of the tail are black. The spine is an intense red-brown colour. They have also cross stripes and shoulder stripes of a blackish colour. The lower jaw beard is of a reddish colour. All these foals bear a more or less curled coat, which is also to be seen on the legs. The eye is blackish." At the close of 1901 several of the animals secured by Mr Hagenbeck reached this country, and specimens are now in the Zoological Gardens, and in the possession of the Duke of Bedford, the Hon. Walter Rothschild and Professor J. C. Ewart\ Since then Mr Hagenbeck has imported a second batch of young Prejvalsky horses, some of which with their Mon- golian foster-mothers are here reproduced (Fig. 19) from a photograph kindly sent me by him. Thus the habitat of this animal, as at present known, is a tolerably confined region, being a quadrangular area bounded on the north by lat. 48°, on the south by lat. 46°, on the west by long. 84°, and on the east by long. 90-1''. Mr Hagenbeck informs me that wild horses of another variety are said to exist 600 miles south of Kobdo, that is, somewhere in the great Gobi desert. 1 Tegetmeier, Field, 11 Jan. 1902, p. 68 (with illustration of those in the collection of the Duke of Bedford) ; 8 Mar. 1903, p. 362 (notice of specimens in Eegent's Park, and in Mr Walter Rothschild's collection). "] THE EXISTING EQUIDAE 29 30 THE EXISTING EQUIDAE [CH. At the time when Poliakoff's paper appeared zoologists had settled down to a firm belief that no true wild horses existed, or indeed had existed for a very long time, since Sanson and Pietrement had concluded that all primitive wild horses had disappeared in prehistoric times. True it was that Pallas had declared that he had seen wild horses with suberect manes in Tartary, and Moorcroft and the brothers Gerrard, when they penetrated into Independent Tartary and within the borders of China, met with numerous herds of wild horses, scouring along the table-lands some 16,000 feet above the sea, but it had become a matter of ftxith with many naturalists that all the wild horses of Asia were sprung from the common Russian country horses turned loose for want of fodder during the siege of Azov in 1697. But in both the eighteenth and nineteenth centuries there were not wanting those who neither believed that all the known wild horses were genuine nor yet committed themselves to the belief that none but feral horses still survived. Thus Pallas, who had himself travelled in Asiatic Russia, was inclined to the same belief as his predecessor Forster, who was disposed to think that all the wild horses in Asia from the Ukraine to Chinese Tartary were descended from strayed domestic animals ; Pallas * himself thought that all the wild horses from the Volga to the Ural were the progeny of domestic animals, and that all those from the Jaik, Don, and Bokhara were of the Kalmuck and Kirghis breed, remarking that they are mostly fulvous, rufous and Isabella, whilst he noticed that those on the Volga were usually brown, dark-brown, and silver grey, some having white legs and other signs of intermixture. Linnaeus'"' held that, though the wild horses of the Don were sprung from the horses that had escaped at the siege of Azov, true wild horses survived in Bessarabia and Tartary, whilst Col. Hamilton Smith came to similar conclusions from the information which he himself obtained from Russian officers of experience whom he met in Paris at the time of its occupation by the Allies in 1814. His statements are so important in reference to Prejvalsky's 1 Travels in Russia and Northern Asia, Vol. t. pp. 37C-8 (French trans.) ; Vol. VII. pp. 89-92 ; PI. I. (in atlas) shows a tarpan of the feral kind. ■^ Systema Naturae, p. 432 (Kerr's trans.). Il] THE EXISTING EQUIDAE 31 discovery, or rediscovery, as it may prove to be, that I shall give them in his own words^ : " Whatever may be the lucu- brations of naturalists in their cabinets it does not appear that the Tahtar or even the Cossack nations have any doubt upon the subject, for they assert that they can distinguish a feral breed from the wild by many tokens ; and naming the former Takja and Muzin, denominate the real wild horse Tarpan and Tarpani. We have had some opportunity of making personal inquiries on wild horses among a considerable number of Cossacks of different parts of Russia, and among Bashkirs, Kirghis, and Kalmucks, and with a sufficient recollection of the statements of Pallas, and Buffon's information obtained from M. Sanchez, to direct the questions to most of the points at issue. From the answers of Russian officers of this irregular cavalry, who spoke French or German, we drew the general conclusion of their decided belief in a true wild and untameable species of horse-, and in herds that were of mixed origin. Those most acquainted with the nomad life, and in particular an orderly Cossack attached to a Tahtar chief as Russian interpreter, furnished us with the substance of the following notice. — The Tarpani form herds of several hundred, subdivided into smaller troops, each headed by a stallion ; they are not found unmixed, excepting towards the borders of China ; they prefer wide, open, elevated steppes, and always proceed in lines or files, usually with the head to windward, moving slowly forward while grazing — the stallions leading and occasionally going round their own troops ; young stallions are often at some distance, and single, because they are expelled by the older until they can form a troop of young mares of their own ; their heads are seldom observed to be down for any length of time : they utter now and then a kind of snort, with a low neigh, somewhat like a horse expecting its oats, but yet distinguishable by the voice from any domestic species, ex- cepting the woolly Kalmuck breed. These animals are found in the greatest purity on the lake Karakoum, south of the lake of Aral, and the Syrdaria, near Kusneh, and on the banks 1 "The Horse" {Naturalises Library, Vol. xii.), pp. 100-5. - Cf. Pallas, Travels (French trans.). Vol. v. p. 378. 32 THE EXISTING EQUIDAE [CH. of the river Tom, in the territory of the Kalkas, the Mongolian deserts, and the solitudes of the Gobi : within the Russian frontier there are, however, some adulterated herds in the vicinit}' of the fixed settlements, distinguishable by the variety of their colours and the selection of residence less remote from human habitation. Tarpans are not larger than ordinary mules, their colour invariably tan, Isabella, or mouse, being all shades of the same livery, and only varying in depth by the growth or decrease of a whitish surcoat, longer than the hair, increasing from midsummer and shedding in May ; during the cold season it is long, heavy, and soft, lying so close as to feel like a bear's fur, and then is entirely grizzled ; in summer much falls away, leaving only a certain quantity on the back and loins ; the head is small, the forehead greatly arched, the ears far back, either long or short, the eyes small and malig- nant, the chin and muzzle beset with bristles, the neck rather thin, crested with a thick, rugged mane, which like the tail is black, as also the pasterns, which are long ; the hoofs are narrow, high, and rather pointed ; the tail, descending only to the hocks, is furnished with coarse and rather curly or wavy hairs close up to the crupper; the croup is as high as the withers ; the voice of the Tarpan is loud, and shriller than that of the domestic horse ; and their action, standing, and general appearance, resemble somewhat that of vicious mules. " The feral horses, we were told, form likewise in herds, but have no regular order of proceeding ; they take to flight more indiscriminately, and were simply called Muzin. They may be known by their disorderly mode of feeding, their desire to entice domestic mares to join them, by their colours being browner, sometimes having white legs, and being often silvery grey ; their heads are larger and their necks shorter ; but their winter coat is nearly as heavy as that of the wild, and there is always a certain number of expelled Tarpan stallions among them ; but they are more in search of cover and watery places, the wild herds being less in want of drink and more unwilling to encounter water, being even said not to be able to swim ; while the Muzin will cross considerable rivers." Il] THE EXISTING EQUIDAE 33 I have given this long extract because the account which it contains of the habitat, the colour, the appearance and habits of the true wild Tarpan, and the respects in which it differed from the feral or intermixed herds has a most important bearing on our present discussion. It is to be carefully noted that the Kirghis and Cossacks from whom Hamilton Smith obtained his information respecting the true wild Tarpau, maintained that the pure wild horses were only to be found in that very region where Prejvalsky obtained the skin of one killed by the Kirghis camel-hunters, where the brothers Grijimailo shot their specimens, and where the Kirghis have captured the numerous foals imported by Mr Hagenbeck. It is further to be remarked that the young wild horses obtained by Mr Hagenbeck differ in colour ac- cording to the three different localities from whence they were procured, and that the foals from the western district have their heads, necks, ears, shoulders, back and croup a light red, passing into whitish colour, the nose, the chest, the belly and the legs being of a whitish colour, whilst the mane is light red- brown, the eel-back is pink, ending in the tuft of the tail, the curled hairs of which are light red-brown, white and black, whilst the upper short-haired portion of the tail shows a whitish colour. The foals further east have light ashy-brown hair on the upper part of the body, the nose and under side are white, just the same as the foals from the west, only the outside of the legs being slightly tinted, whilst the fetlocks are black. The mane and spine are a deep brownish colour, and the beard also. The western foals have a whitish iris, the more eastern have a darkish iris. The foals from the most easterly district (Zagan-norr Lake) have a coat of a full yellowish-brown colour, only interrupted by the white belly hair and the distinct black bands at the outside of the legs from the black hair of the fetlocks to above the hock. The nose is white, the mane and the curled hair of the tail are black, and the spine is an intense red-brown colour, the lower jaw beard is of a reddish colour. All these foals bear a more or less curled coat, which is also to be seen on the legs ; the eye is blackish. Let us now compare these descriptions with R. H. 3 34 THE EXISTING EQUIDAE [CH. that given by Smith of the true wild Tarpans which nearly a century ago were declared " not to be found unmixed except towards the borders of China," and which were then found in their greatest purity on the lake Karakoum, south of the Sea of Aral and the Syrdaria near Kusneh and in the Gobi desert, whilst those within the Russian frontiers of that time were adulterated and distinguished by the variety of their colour from the pure herds further east. The true Tarpans " are not larger than ordinary mules, their colour invariably tan, Isabella, or mouse, being all shades of the same livery." Now this would describe very well Mr Hagenbeck's foals from the two most easterly districts, but does not agree with the red heads, necks, backs, and croups of those from the western area. Again, the true Tarpan had a small head, the forehead greatly arched (which we shall soon see to be a characteristic of at least some of the Prejvalsky horses), " the neck crested with a thick, rugged mane, which like the tail is black, as also the pasterns." Now this description does not at all agree with the Prejvalsky foals from the tvest, for the mane is a light red- brown, and the curled hairs of the tail are light red-brown, white and black, whilst the legs are white, but it tallies quite well with the foals from the second district, which are light ashy-brown coloured instead of red on the head and back, have black fetlocks and the outside of the legs slightly tinted, and have the mane and spine of a deep brownish colour, and the beard also, thus coming much closer to the description of the pure Tarpan, with its black fetlocks and legs; while the foals from the eastern district, i.e., closest to China, which have a coat of a full yellowish-brown colour, and have not only black fetlocks, but also distinct black bands on the outside of the legs to above the hock, exactly correspond with the picture given us of the unadulterated Tarpan. From these considerations it would appear that (1) Prej- valsky's horse is nothing more than the Tarpan of the older writers ; (2) that if pure Tarpans still survive they are those of the Zagan-norr Lake, and (3) that the divergence in colour of these animals which characterizes those found in the middle district, and in a still greater degree those of the most westerly Il] THE EXISTING EQUIDAE 35 area (especially in the white colour of their legs), is to be ascribed to their being mixed with feral horses rather than to any variation due to environment or any other natural cause. Sanson and Pietrement viewed with suspicion Prejvalsky's discovery, and Pietrement placed the animal under the same sub- species of Equus wdth Equus caballus. In this country Dr Sclater took the same view as Poliakoff, whilst Sir W. Flower thought that it might be an accidental hybrid between a Kiang and a Mongolian or some other kind of horse. Flower's caution was quite justifiable at a time when only a single skin was known, although it seems not very likely that such accidental unions as he postulated would occur between different species of Equidae in a state of nature, in view of the well-known objection of the herds of half- wild horses in the Caucasus to intermix in any way. Yet, though many specimens both living and dead, which have since come to hand, render it very im- probable that Prejvalsky's horse is a mule, the theory has retained its hold upon some naturalists down to the present time, who, however, have made no attempt to test the theory by experiment. It is to the indefatigable energy and enthusiasm of Pro- fessor Cossar Ewart, who has done more than any living man to advance our knowledge of the Equidae, that we owe the experiments which seem likely to settle the question finally. It is best to let him speak for himself^ : " With the help of Lord Arthur Cecil I succeeded early in 1902 in securing a male wild Asiatic ass- and a couple of Mongolian pony mares — one a yellow dun, the other a chestnut. 'Jacob,' the wild ass, was mated with the dun Mongol mare, with a brownish- yellow Exraoor pony, and with a bay Shetland-Welsh pony. The chestnut Mongol pony was put to a light grey Connemara stallion. Of the four mares referred to three have already (June) foaled, namely the Exmoor and the two Mongolian ponies. The Exmoor having foaled first, her hybrid may be first considered. It may be mentioned that the Exmoor pony 1 "The Wild Horse" [Equus przeivalskii Poliakoff), Proc. Royal Soc. of Edinburgh, 1903, pp. 460-8. - This animal, now in the Zoological Garden, Eegent's Park, is ,an onager indicus ( — hemionus indicus, cf. p. 43). ^ 3—2 36 THE EXISTING EQUIDAE [CH. had in 1900 and again in 1901 a zebra hybrid, the sire being the Burchell zebra ' Matopo ' (Fig. 36) used in my telegony experiments. In the case of her Kiang hybrid the period of gestation was 335 days (one day short of what is regarded as the normal time), but she carried her 1900 zebra hybrid 357 days, three weeks beyond the normal time. The Exmoor zebra hybrids are as nearly as possible intermediate between a zebra and a pony ; the Kiang hybrid, on the other hand, might almost pass for a pure-bred wild ass. In Mendelian terms the Exmoor pony proved recessive, the wild ass dominant. In zebra hybrids the ground colour has invariably been darker than in the zebra parent ; but the Kiang hybrid is decidedly lighter in colour than her wild sire, while in make she strongly suggests an Onager — the wild ass so often associated with the Runn of Cutch. Alike in make and colour the Kiang hybrid differs from a young Prejvalsky foal." This comparison Pro- fessor Ewart was enabled to make by means of his hybrid foal with the skin of a very young Prejvalsky foal (for which he was indebted to Mr Carl Hagenbeck). " I have never seen a new-born wild horse ; but if one may judge from the conformation of the hocks, from the coarse legs, big joints, and large head of the yearlings — to their close resemblance to dwarf cart-horse foals — it may be assumed they are neither characterized by unusual agility nor fleetness. The Kiang hybrid, on the other hand, looks as if built for speed, and almost from the moment of its birth has by its energy and vivacity been a source of considerable anxiety to its by no means placid Exmoor dam. When four days old it walked over twenty miles ; on the fifth day instead of resting it was unusually active, as if anxious to make up for the enforced idleness of the previous evening. In the hybrid the joints are small, and the legs long and slender, and covered with short, close-lying hair. In the wild horse the joints are large, and the ' bone ' is round as in heavy horses. " As to its colour it may especially be mentioned that the hybrid has more white around the eyes than the wdld horse, but is of a darker tint along the back and sides and over the hind-quarters. Too much importance, however, should not be Il] THE EXISTING EQUIDAE 37 attached to differences in colour ; for though the two hybrid foals, which have already arrived, closely agree in their colora- tion, subsequent foals may differ considerably, and it is well known that young wild horses from the western portion of the Great Altai mountains differ in tint from those found further east. " Of more importance than the coat-colour is the nature of the hair. A Prejvalsky foal has a woolly coat not unlike that of an Iceland foal. In the hybrid, the hair is short and fine and only slightly wavy over the hind-quarters. It thus differs but little from a thoroughbred or Arab foal. " The mane and the tail of the hybrid are exactly what one would expect in a mule ; the dorsal band, 75 mm. wide over the croup in the sire, has in the hybrid a nearly uniform width of 12 mm. from its origin at the withers until it loses itself halfway down the tail. The tail, which differs but little from that of a pony foal, is of a lighter colour than the short, upright mane, while the dorsal band is of a reddish-brown hue. In the wild horse the dorsal band is sometimes very narrow (under 5 mm.) and indistinct. In the Kiang sire there are pale, but quite distinct stripes above and below the hocks, and small faint spots over the hind -quarters — vestiges apparently of ancestral markings : but in the hybrid there are neither in- dications of stripes across the hocks or withers, nor spots on the quarters. In having no indications of bars on the legs, or faint stripes across the shoulders, the hybrid differs from Prejvalsky colts ; it also differs in having a longer flank feather and in the facial whorl being well below the level of the eyes. As in the Kiang and some of the wild horses, the under surface of the body and the inner aspect of the limbs are nearly white. "In the hybrid the front chestnuts (wrist callosities) are smooth and just above the level of the skin; but instead of being roughly pear-shaped, as in the Kiang, they are somewhat shield-shaped, as in the Onager. In the wild horse the front chestnuts are elongated. In the Exmoor dam the hind chest- nuts, (hock callosities) are 27 mm. in length and 10 mm. wide. In the sire there is a minute callosity inside the right hock. In the hybrid the hind chestnuts are completely absent. In 38 THE EXISTING EQUIDAE [CH. the absence of hock callosities the hybrid differs from the wild horse, in which they are relatively longer than Clydesdales, Shires, and other heavy breeds of horses. In the hybrid, as in the sire and dam, there are smooth, rounded fetlock callosities (ergots) on both fore and hind limbs. " In the wild horses the hoof is highly specialized, the 'heels' being bent inwards (contracted) to take a vice-like grip of the frog. In the hybrid the hoof closely resembles that of the pony dam ; it is shorter than in the Kiang, and less contracted at the ' heels ' than in the wild horse. The Kiang hybrid further differs from a young wild horse in the lips and muzzle, the nostrils and ears, and in the form of the head and hind- quarters. The wild horse has a coarse, heavy head, with the lower lip (as is often the case in large-headed horses and in Arabs with large hock callosities) projecting beyond the upper. The nostrils in their outline resemble those of the domestic horse, while the long, pointed ears generally project obliquely outwards, as in many heavy horses and in the Melbourne strain of thoroughbreds. Further, in the wild horse the forehead is convex from above downwards, as well from side to side, hence Prejvalsky's horse is sometimes said to be ram-headed. In the hybrid the muzzle is fine as in Arabs, the lower lip is decidedly shorter than the prominent upper lip, the nostrils are narrow as in the Kiang : and even at birth the forehead was less rounded than is commonly the case in ordinary foals. The ears of the hybrid, though relatively shorter and narrower than in the Kiang, have, as in the Kiang, incurved dark-tinted tips, and they are usually carried erect or slightly inclined towards the middle line. In the wild horse the croup is nearly straight and the tail is set on high up as in many desert Arabs. In the hybrid the croup slopes as in the Kiang and in many ponies, with the result that the root of the tail is on a decidedly lower level than the highest part of the hind-quarters. Further, in the young wild horses I have seen the heels (points of the hocks) almost touch each other, as in many Clydesdales, and the hocks are distinctly bent. In the hybrid the hocks are as straight as in well-bred foals, and the heels are kept well apart in walking. Another Tl] THE EXISTING EQUIDAE 39 difference of considerable importance is, that while the wild horse neighs, the hybrid makes a peculiar barking sound remotely suggestive of the rasping call of the Kiang. " The dun Mongol pony's hybrid arrived five weeks before its time, and, though perfect in every way, was short-lived. Only in one respect did this hybrid differ from the one already described. In the Exmoor hybrid the hock callosities are entirely absent; in the Mongol hj'brid the right hock callosity is completely wanting, but the left one is represented by a small, slightly hardened patch of skin, sparsely covered with short white hair. In zebra hybrids out of cross-bred mares the hock callosities are usually fairly large, while in hybrids out of well-bred pony mares the hock callosities are invariably absent. The Exmoor^ pony, though not as pure as the Hebri- dean and other ponies without callosities, has undoubtedly a strong dash of true pony blood; the Mongol pony is as certainly saturated with what, for want of a better term, may be called cart-horse blood." Prof Ewart thus sums up the results of his experiment: " From what has been said, it follows that a Kiang-Mongol pony hybrid differs from Prejvalsky's horse (1) in having the merest vestiges of hock callosities ; (2) in not neighing like a horse ; (8) in having finer limbs and joints and less specialized hoofs ; (4) in the form of the head, in the lips, muzzle, and ears; (5) in the dorsal band ; and (6) in the absence even at birth of any suggestion of shoulder stripes and of bars on the legs." After this experiment it does not seem likely that zoologists will continue to hold that Prejvalsky horses are the offspring" of Kiangs and feral Mongolian ponies-. But as some naturalists had maintained that Prejvalsky horses in nowise differed essentially from an ordinary horse and held that the colts brought from Central Asia were the progeny of escaped feral Mongol ponies, and as others again asserted that they failed to discover any difference between the 1 The Exmoor ponies are said to have derived some good blood from a famous stallion Katerfelto. 2 As these pages are passing through the press the Prejvalsky horses belong- ing to the Duke of Bedford have themselves triumphantly refuted the charge of their being merely mules by having this year (1904) produced offspring. 40 THE EXISTING EQUIDAE [CH, young wild horses in the London Zoological Gardens and Iceland ponies of a like age, Prof. Ewart again resorted to the experimental method. To test the first of these assertions he mated his chestnut Mongol pony with a young Connemara stallion ; to test the second he purchased an Iceland mare in foal to an Iceland stallion. " The chestnut Mongol mare produced a foal the image of herself. The foal, it is hardly necessary to say, decidedly differs from the Prejvalsky colts recently imported from Central Asia by Mr Hagenbeck, and it decidedly differs from the wild ass hybrids described above. The Iceland foal, notwithstanding the upright mane and the woolly coat, for a time of a nearly uniform white colour, could never be mis- taken for a wild horse, and the older it gets the difference will become accentuated." " If the Prejvalsky horse is neither a wild ass-pony mule nor a feral Mongolian pony, and if moreover it is fertile (and its fertility can hardly be questioned), I fail to see how we can escape from the coDclusion that it is as deserving as, say, the Kiang to be regarded as a distinct species \" It will be obvious that in view of the facts that the Prejvalsky horses from the two western districts agree in the colour of their legs with the adulterated herds of Tarpans described by Hamilton Smith, while they differ essentially in colour from that of the true Tarpan, and that on the other hand the Prejvalsky horses from the easternmost district correspond accurately to the description of the genuine Tarpan, it would be unwise to maintain that all the Prejvalsky horses imported by Mr Hagenbeck are genuine wild animals unmixed with feral blood, though in view of the evidence which I have set forth one is justified in holding that the Prejvalsky horses from the Zagan-norr Lake are possibly perfectly genuine, and if not absolutely pure from all admixture, at least so little tainted that they practically give us a true picture of the primitive wild stock. Indeed, if they are impregnated with the blood of feral horses, their resemblance to the ass in the absence of the forelock, the upright mane, 1 op. cit., pp. 467-8. Il] THE EXISTING EQUIDAE 41 and the character of the tail render it all the more certain that there was a primitive variety of horse which had these characteristics so strongly marked that they cannot be easily blurred by crossing with horses of the ordinary domestic type. Quite recently Dr Salensky^ has urged strongly that Equus przewalskii is a true variety of Equus. He gives the charac- teristics of the Prejvalsky horse as the considerable size of the head, the want of a forelock, the upright mane, the back and shoulder stripes, the characteristic form of the tail, which in some particulars resembles that of the koulan, the size of the ears, which are smaller than in the ass or koulan, and the coloration of the rump, the lower parts of the body, and the striping on the legs, and he holds that the examination of the skull and skeleton leads us, as do the external marks, to the conclusion that Prejvalsky's horse represents a special type, which forms a peculiar race of the sub-species of Equus standing next to Equus cahallus. In comparing Equus pi'ze- walskii to other horses he considers that the Tarpan comes first in importance, a view obviously correct in face of the considerations which have been urged above. He starts by citing Gmelin's notice of the Tarpan, the earliest in modern times at least (in his Reise durch Russland). That traveller had the opportunity of seeing them at Bobrowsk (gov. Woronesh), and he describes them as mouse-coloured " with short and crisp mane " {mit kurzer und kraushaariger Mdhne) and says that their legs were black from the knee to the hoof, the head disproportionately thick. The ears sometimes long as in the ass, and hanging, the tail always shorter than in domestic horses, being sometimes well furnished, sometimes sparsely. But Salensky relies chiefly on the official description of a Tarpan captured in 1866 in the Zagradoff steppe on the property of Prince Kotschubei (gov. Cherson) and which was still in the Zoological Gardens at Moscow in 1884. This animal had a forelock but had no callosities on its hind-legs. It was a dark mouse-colour, the legs from hocks and knees down to the pasterns being very black, whilst it had a mane 1 Equus przewalskii (Comptes-Kendus of Imperial Russian Academy, 1902), from which my illustration of Prejvalsky's horse (Fig. 18) is taken. 42 THE EXISTING EQUIDAE [CH. 48 cm. long hanging down on the left side of the neck. Un- fortunately no minute study was made of its tail, but, as far as can be seen from a photograph, the tail resembled that of Prejvalsky's horse. There are in existence two Tarpan skeletons, one at St Petersburg, the other at Moscow. On the ground of the skull measurements Czerski came to the con- clusion that the Tarpan has all the marks of the group of Oriental horses, being connected on the one side with the Arabian, on the other with the Scottish race to which the ponies belong. The skull comes very near to Equus przewal- skii, although it does not agree with any fully developed skulls of this kind. The number of lumbar vertebrae agree in both Tarpan and Prejvalsky horse, as both have five, but this does not amount to much, as the same occurs in other horses, whilst there are asses with six such vertebrae. The most genuine re- semblance between the Tarpan and the Prejvalsky horse is the black colour of the legs below the knees, a feature very persistent (says Salensky) in the Prejvalsky horse \ and which separates it from hybrid asses, in which the legs are always half or wholly white. But Salensky points out that there are some essential differences between the Tarpan and Prejvalsky horse; these are the presence of a forelock in the Tarpan, a longer mane falling down at the side, and a tail more like that of a horse. " Ail these marks indicate that the Tarpan is a type more specialized towards the horse side than is Equus przeival- skii. Too much stress cannot be laid on the absence of the hock callosities in the only known Tarpan, for such a feature is well known among true horses. The Prejvalsky horse represents a more universal form between the horses and the asses, and this leads to the assumption that more than any other kind of the genus Equus it comes nearest to the common stem-form of horses, asses, and half-asses." When the reader bears in mind the evidence obtained by Colonel Hamilton Smith in 1814 — that there were no pure tarpans within the Russian frontiers — he will at once see that the tarpans described by Gmelin as having sometimes long, some- times short ears, and that the Moscow tarpan with its long hanging mane (in which it differed from the tarpans observed by 1 But this is disproved by the facts cited on pp. 27 and 32. Il] THE EXISTING EQUIDAE 43 Gmelin and Pallas), with its forelock and its horse-like tail, were not genuine wild tarpans, but belonged to the mixed herds of eastern Russia. We must therefore reject Salensky's view that the Moscow tarpan represents a variety of Equus distinct from Prejvalsky's horse, by being more specialized towards the side of the horse. The hypothesis that it was a cross between the true tarpan and a feral horse will likewise account for Czerski's conclusions that it approaches the Arab on the one side and the British ponies on the other, for, as has been already pointed out, Arab horses sometimes lack the hock callosities. We shall presently find that mouse-colour — the hue of the Moscow tarpan — when found in horses is an indication that crossing has taken place. Later on in this investigfation it will be shown that mouse-colour and dark mouse-colour in horses are a sure indication of an intermixture of breeds. We may therefore conclude that whilst the tarpan of eastern Asia and the Prejvalsky horse with black legs from Zagan-norr Lake are identical, we must hold that the tarpans of eastern Europe and western Asia have probably been largely crossed with escaped domestic animals for at least two centuries, and probably much longer. To the three kinds of horses which have been just set forth above I shall venture to add a fourth — Equus cahallus libycus. ASSES. Side by side with Prejvalsky horses the brothers Grijimailo found two varieties of wild asses in the desert of Dzungaria. The wild asses of Asia fall into a group distinct from those of Africa; the older zoologists divided them into E. kiang, E. onager, and E. hemippus, which were regarded by some as distinct species, but by others as merely races of the same species, the Equus hemionus of Pallas. The best modern authorities now make at least five subdivisions S—^. hemionus, E. hemionus kiang Moorcroft, E. onager, E. onager indicus, and E. onager hemippus, whilst of course there may be other races as yet unidentified. All have a dorsal band, but no shoulder 1 Dr W. T. Blanford, F.K.S. (Indian MammaUa, p. 470, 1891) holds that all are simply local races of the same species [E. hemionus). 44 THE EXISTING EQUIDAE [CH. stripe, while their ears are a little shorter than those of the Abyssinian ass. (4) The Dzeggetai {E. hemionus). Mr Lydekker has described a wild ass^ obtained in Kobdo north-west of the Gobi Desert, now in the possession of the Duke of Bedford. In its make and action " as well as in the general type of coloration, this wild ass agrees essentially with the wild ass of Ladak and Tibet. Both in the winter and summer coats it lacks, however, i'Ki. 2(». The KiaiiL'-. the distinctly rufous-chestnut tint so characteristic of tlie latter, while it is further characterized by the much less marked contrast between the light and dark areas of the coat." The light areas on the muzzle, buttocks, legs, and under parts being ' Isabella '-coloured* instead of pure white, and thus much less sharply differentiated from the fawn of the rest of the body. 1 P. Z. S., 1904, p. 431 (with Plate); cf. Pallas (vii. 92) for Mongol di>hi(i(jnetei. - From a photograph by the Duchess of Bedford. * i.e. the colour of the soiled linen of Isabella of Castile. "] THE EXISTING EQUIDAE 45 " The general colour is pale sandy fawn, with the tips of the ears, mane, and dorsal stripe (which is continued down the tail), brown, and there seems to be but little difference in this respect between the summer and winter coats. The dorsal stripe is narrow as in the kiang, and is thus distinct from that of the ghor-kar and onager, which is broader and bordered with white." Mr Lydekker regards this animal " as the true Eqaus hemionus of Pallas, which came from Mongolia, and is known Fig. 21. The Kiang i. to the natives as chigetai (dzeggetai). It is certainly entitled to be regarded as subspecifically distinct from the kiang of Tibet and Ladak, and the latter should be known as Equus hemionus kiang (Moorcroft)." (5) The Kiang {E. hemionus kiang) lives in the upper Indus valley, Tibet, and Mongolia, seldom at a lower altitude than 10,000 feet. It (Figs. 20, 21) differs from the onager 1 This illustration is from a photograph (copyright) of a kiang formerly in the Zoological Garden, Regent's Park, by Mr L. Medland, F.Z.S., well known for his photographs of living animals. 46 THE EXISTING EQUIDAE [CH. in being larger, exceeding 13 hands, and by the narrowness of its dorsal stripe compared with that of the latter animal. Its hind-quarters are much more developed in length and strength. In colour it is rufous-bay, whilst its voice is said to differ from the strident bray of the onager. (6, 7, 8) The Onager {E. onager), the Onager In- dicus, and the Hemippus {E. onager hemippus) differ so slightly in habit ^ that they may be described together. These animals are found on all the great plains of Asia, Chinese Tartary, Tibet, the Panjab, Afghanistan, Western India, Balu- chistan, Persia, and Syria, It is called koulan by the Kirghis, ghuran or ghur by the Baluchis, ghor-khur in Hindi, ghour or kherdecht in Persian, in all of which a common element may be recognized. Zoologists now discriminate between E. onager indicus (which is found in North-western India and Baluchistan), E. onager (found in Persia and Turkestan), and E. onager he7nippus (found in Syria), whilst it seems probable that the onager of Turkestan- differs in some respects from that of Persia. The onager indicus is not so dark in colour as the kiang, whilst the typical onager (Fig. 22) " is very white, and in fact might be described as a white animal with a yellowish blotch on the side, another on the neck, and some yellow on the head'V They are usually found in herds of from four to forty, and in spring the mares and foals sometimes congregate in still larger numbers. The ears (Fig. 22) are large, the hair of the tail is short at the base, but grows gradually longer towards the end, which is of a black colour, whilst the mane is erect. The dorsal stripe is dark brown, sometimes with a white edging, and varying in breadth, but normally broader than that of the kiang. Some specimens show a cross stripe on the shoulder, and sometimes the legs show feint rufous bars. It varies in height from 11 to 12 hands. It has been supposed to outstrip in speed the fleetest horses, a notion 1 Blandford, Indian Mammalia, p. 470. " According to Pallas (vii. 92), the koulan of Upper Asia is brownish-yellow with brown dorsal stripe and two bars on legs. ■■' I am indebted for this accurate information to Mr E. I. Pocock, F.Z.S., Superintendent of the Zoological Garden, Eegent's Park. "] THE EXISTING EQUIDAE 47 apparently as unfounded as the belief that it is incapable of being domesticated. The groundlessness of the former has been demonstrated by the capture of these animals in Cutch by sportsmen mounted on Arabs, Walers (horses from New South Wales), and country breds\ whilst the latter is shown to be erroneous by the fact that some of the Indians in the army of Xerxes drove chariots drawn by 'wild asses-.' Fig. 22. The Onager^. From this it is clear that the peoples of western Hindustan, who did not possess horses, had made the wild ass obedient to the yoke. In Carmania (included in modern Persia), a region bounded by the Indian Ocean and Persian Gulf on the south, and by ^ Tegetmeier, Horses, Asses, etc., pp. 23-5. " Herod, vii. 86, ijXavvov 5^ KiXyiras Kal dp/xara' virb 5^ rolai. cipfxaai virrjaav iTTTTot Kal ovoi dypioi. ^ The illustration is from a photograph (copyright) from the specimen in the Zoological Garden, Regent's Park, by Mr W. P. Dando, F.Z.S., official photographer to the Zoological Society. 48 THE EXISTING EQUIDAE [CH. Persia on the west, down to the time of Strabo^ "asses on account of the scarcity of horses" were "generally made use of in war. They sacrifice an ass to Ares, who is the only god worshipped by them, for they are a warlike people." The Carmanians closely resembled the Persians and Medes in their customs. That the onager was regularly captured and domesticated in Assyria in ancient times is clearly established by one (Fig. 23) of the bas-reliefs discovered by Sir A. H. Layard at Kouyunjik (Nineveh). The relief, which is one of a series of slabs recording scenes in the life and hunting expeditions of Assur-Bani-Pal (B.C. 668 — 626), represents two of the king's attendants lassoing a wild ass. The other asses are seen running away-. In ancient times they were well known in Syria, as is clear from the frequent allusions to the wild ass in the Old Testament. Aristotle^ speaks of "those animals called mules Qiemionoi) in Syria which are so termed because of their similarity to mules {i.e. domestic mules), though they are not really of the same kind, for they breed freely," and else- where he states that "in Syria there are animals termed mules (hemionoi), which, though they are quite a different species from the domestic mules and resemble the wild asses (ol ajpioi ovof^), get their name from a certain resemblance to tame mules. Like the wild asses and the domestic mules they surpass in fleetness. These mules breed freely, as is proved by the fact that some were brought into Phrygia in the time of Pharnaces, the father of Pharnabazus, and still continue there. There are now but three, but formerly there were nine." Apparently there were two kinds of wild asses in Syria, differing but slightly from each other, the one known as the ' wild mule ' (hemionos), the other as the ' wild ass ' (onos), the former being probably darker in colour than the latter. Thus one corresponded to the variety termed hemippus or onager indicus, the other to the typical onager. 1 726. - C. Keller (Abstimrn. alt. Haustiere) takes them for Prejvalsky horses, but cf. Horses on Figs. 62-4. 3 Hist. Animals, i. 6. •* Hist. Animals, vi. 36. "] THE EXISTING EQUIDAE 49 K. H. 50 THE EXISTING EQUIDAE [CH. According to the Book of Wonderful Stories' attributed to Aristotle, one of the wild asses led the herd, and whenever any of the younger stallions attempted to serve a mare, the master stallion became furious, pursued the colt until he caught him, then stooping down behind him he emasculated him with his teeth. A like belief is still current in India respecting the wild asses of Cutch. In early days the wild ass was well known in Paphlagonia, for Homer ^, when speaking of the Eneti who came from thence Fig. 24. The Nubian Wild Ass^. to aid Piiam and the Trojans, describes their land as "the home of wild mules." There can be little doubt that the wild mule 1 10 (831 a 22) : ^acrli' ev ^vpia twu aypLwv ovwv eva dcpriyeiadai ttjs dy4\r]s, iireidav 5^ tis vewTcpos 03v twv wuiKwv eirl tlvo. difKeiav ava^rj, tov a.s etrLXijiplovt. - W. Kidgeway, "Why was the Horse driven before he was ridden?" Academy, Jan. 3rd, 1890, p. 14. Ill] AND HISTORIC TIMES 95 Celts and other inhabitants of Gaul, whose chief strength lay in their horsemen {equites). It will presently be shown that this change in the method of using the horse on the Continent was due not solely to the development in size of the indigenous animal, but to the fact that by the middle of the second century B.C. the Gauls had procured from southern Europe horses of a size far superior to their own and better adapted for riding. Dio Cassius, when speaking of the Caledonians and Maeatae, two chief tribes of northern Britain, says that they "went to war on chariots, as their horses were small and fleets" Since the countr}^ which these tribes inhabited would have been much more easily traversed by men on horseback than by wheeled vehicles, it is clear that they used chariots because their ponies, which, in part at least, may be represented by the 'Celtic' ponies of to-day, were too small to carry a full-grown man for any considerable time or distance. The statement of Dio Cassius concerning the practice of the tribes of northern Britain is completely confirmed by the discovery of the remains of a considerable number of chariots in Yorkshire barrows. In one (Fig. 44) of the sixteen tumuli known as Danes' Graves, situated in the parish of Driffield, Mr J. R. Mortimer- and Canon Greenwell, in 1897, discovered the remains of two adult bodies (Fig. 44), the iron tires of two wheels and other pieces of iron belonging to a chariot, two iron snaffle-bits, and several rings and ornaments of bronze belonging to the horse- trappings, though not a single bone of a horse was found. The wheels had apparently been taken from the axle. The tires measure respectively 2 ft. 6f in. and 2 ft. oj in. in diameter, both being If in. broad and -^ in. in thickness. The iron hoops for the naves likewise survived, being 5 in. in diameter (inside), fin. wide, and nearly a quarter of an inch thicks ^ Dio Cassius, lxxvi. 12 (ex XijihiUni epit.), ^TparevovTai. de eTri re apfiaroiv, 'iinrovs '^xofTes iMKpovs koL raxcis, koX ire'^ol 5i dm kt\. - Ann. Report of the Yorkshire Philosophical Soc. for 1897, pp. 3-4. 3 Ibid., p. 10. Tlie Yorkshire Philosophical Society have most kindly lent me the block for Fig. 41. '-i*. c. +^ D Fig. 44. Eemains of a Chariot found at Driffield, Yorkshire. CH. Ill] PREHISTORK^ AND HISTORIC HORSES 97 In a group of over 200 small barrows closely resembling the Danes' Graves, which once existed at Arras near Market Weighton, the remains of three chariots were founds The fragments of another chariot discovered in one of the barrows at Hessleskew, were presented to the Museum of the Yorkshire Philosophical Society (which possesses also the relics from Arras and Danes' Graves), by the Rev. E. W. Stillingfleet in 1865^ The evidence derived from the finds at Silchester and along the Roman wall confirms the statement of Dio Cassius respect- ing the small size of the British horse. There can therefore be little doubt that the cause which led the Sigynnae to drive in chariots had induced Britons to follow the same custom, even in a country beset with forests and morasses, and where it is obvious that riding on horseback, as in medieval times, would have been much more convenient than driving in chariots, had horses of sufficient size been available. The Belgic tribes of Britain at the time of Caesar's invasion used both horsemen and chariots, for on learning of Caesar's^ intended landing, "they sent forward cavalry and charioteers, which formed their chief arm in warfare." Caesar describes elsewhere the value of the war-chariots, and their method of handling them. " At the first onset they drove the cars in all directions, hurled their javelins, and by the din and clatter of horses and wheels commonly threw the ranks of the enemy into disorder, and making their way amongst the squadrons of the enemy's cavalry they leaped down from their chariots and fought on foot. The charioteers then little by little withdrew out of the fight and placed their chariots in such a way that if they were hard pressed by the enemy they could readily retreat to their own side. Thus in battle they afforded the mobility of cavalry, and the steadiness of infantry. 1 Oliver, Hixtory of Beverley, p. 4 (footnote), cited by Mr Mortimer, loc. cit. " For information about all these finds I am indebted to my friend Canon Greenwell, and also to Mr Platnauer, the curator of the Museum of the Yorkshire Philosophical Soc, York, who (through my friend and former pupil, Mr C. Gutch) most kindly supplied me with photographs of the Arras and Hessleskew remains. An elaborate monograph by Canon Greenwell on all the chariot-burials found in England will appear very shortly in Archaeologia. ^ B. G. IV. 24. R. H. 7 98 THE HORSES OF PREHISTORIC [CH. Daily practice enabled them to pull up their horses when in full speed on a slope or steep declivity, to check or to turn them in a narrow space, to run out on the pole and stand on the yoke, and to get nimbly back again into the chariots" This statement, confirmed by other ancient writers^ puts it beyond doubt that it was not lack of intrepidity or agility that induced the Britons to drive their horses instead of mounting on their backs, and also shows that their cars were not scythed. The evidence just offered for the diminutive size of the Irish horse combined with the fact that in the oldest Irish epics the horse is not ridden, for Cuchulainn and Queen Medhbh are always represented as fighting in chariots, renders it highly Fig. 45. Bronze Bits: Ireland^. probable that here also the use of the chariot in a country singularly difficult for vehicles was due at least in part to the smallness of the steeds. It is not unlikely, that as the domesticated horse was intro- duced into Britain, so also was he brought into Ireland at no very remote date, for all the bits and trappings hitherto known belong to the Iron age in that country, where as I have elsewhere argued iron found its way at a comparatively late 1 B. G. IV. 33. 2 Juvenal, iv. 126. •* The larger and more richly decorated bit is one of a pair found along witli a pair of the well-known spur-shaped objects, on the hard turf bottom of a bog at Atymon, co. Mayo, in 1891. Ill] AND HISTORIC TIMES 99 epoch. The bronze bits here shown (Fig. 45), each being one of pairs in the Museum of the Royal Irish Academy, belong beyond doubt to the late Celtic period, as is demonstrated by the beautiful ornament on one of the bits and its fellow, as well as on the pair of mysterious pieces of trapping, found at the same time, one of which is figured on a later page. It has been urged that it was easier to learn to drive than to ride. But is this true ? Under modern conditions it may be so, when a person's first essay in driving is made on some old and steady animal tightly embraced in harness and shafts. But when primitive man first subdued the little wild horse, was it easier for him to learn to drive two of these animals, when simply attached by means of a yoke and pole, with free play for their heels, their first instinct being to kick to pieces the rattling, creaking wheels and axle, which formed the primi- tive car, or to learn to sit firmly on his backM The South American Indians found no difficulty in acquiring the latter art when they obtained the horse from the Spaniards. As the Britons were famous for their intrepidity in running out and standing on the chariot pole, and as we shall presently see, Odysseus and Diomedes had no hesitation in getting on the backs of Thracian steeds, it is clear that it was not from fear that either Achean or Briton drove habitually in a chariot instead of riding on horseback. But though mounted men formed the chief weapon of the Gauls in their death-struggle against the Romans, it is clear from both literary and monumental evidence that at no long time previously had the chariot been in universal use among all the Celts of Gaul and north Italy. Thus Diodorus^ makes it plain that down to a late date they, like the Homeric Acheans, had regularly gone to war in two-horse chariots, containing each a warrior and a charioteer: the former first hurled spears called saunia at the foe, and then dismounted to finish the combat at close quarters with the sword, the latter being doubtless of that type known as La Teue (Fig. 111). The opening of many tumuli in Champagne has brought to light 1 W. Kidgeway, Academy, 1890, p. 91. - v. 29. 7—2 100 THE HORSES OF PREHISTORIC [CH. the remains of Gaulish chieftains, who were interred, seated on their chariots, the horses and trappings being buried along with them. These interments, as is proved by the swords and fibulae of the La Tene ^ type, cannot be earlier than 400 B.C. and are probably to be set at least a century or two later". This evidence is completely corroborated by that of Livy^, who narrates that in the great battle fought at Sentinum in Etruria (292 B.C.), in the third Samnite War, when the Romans under Fabius Maximus and Decius Mus overthrew the com- bined Samnites and Gauls, the latter had a thousand chariots (esseda) and cars (carri), the charge of which completely routed the Roman cavalry, and would have decided the battle in favour of the allies, had not Decius Mus, following the example of his father at the battle of Vesuvius in the Latin War, dedi- cated himself and the enemy's host to the infernal gods, and by this act of devotion gave fresh courage to his legionaries to make a stand which led to ultimate victory. It is therefore clear that when the Gauls entered Italy at the end of the fifth or the beginning of the fourth century B.C., they like the Sig3'nnae on the north side of the Danube were drivers of chariots and not yet riders of horses. How then did it come to pass that though the Gauls of north Italy are still using chariots in 292 B.C., yet by Caesar's day the peoples of Gaul had universally discarded the war- chariot and were employing cavalry alone ? Fortunately suf- ficient evidence has survived from antiquity to enable us to trace the way in which this important change was effected. I am now going to show that the Gauls of north Italy had taken to horseback by the latter part of the third century B.C., and probably much earlier, that the Transalpine Gauls had fully adopted the same practice by the iniddle of the second century B.C., whilst even the Belgic tribes of the Continent had ^ Ridgeway, Early Age of Greece, Vol. i. p. 410. 2 Morel, La Champagne Sonterraine, p. 23, Pll. i. x. etc. =* X. 28-30, " Essedis carrisque superstans armatus hostis ingenti sonitu equorum rotarumque advenit, et insolitos eius tumultus Eomanorum conteiruit equos." Ill] AND HISTORIC TIMES 101 discarded the chariot by Caesar's time, though their kinsfolk who had crossed into south-eastern Britain still retained it in use side by side with cavalry. Later on it will be shown that the Celts of Noricum and the Danube had begun to ride on horseback in the early Iron age, though retaining the use of the chariot, and that by the beginning of the third century B.C. the Celtic tribes of this region had developed a highly organised cavalry system. Furthermore, it will be shown that this change from chariot- eering to riding went pari passu with the importation of superior horses from the Mediterranean area into the Upper Balkan and into the countries beyond the Alps. When Hannibal arrived in north Italy (B.C. 218) he first came into contact with the Romans in the cavalry engagement on the Ticinus (Ticino). Here his Numidian horsemen, who rode without either bridle or saddle, and his Spanish cavalry who used bridles, at once proved their superiority not only to the Gallic horsemen, whom Scipio' had placed with his javelin- throwers in his front line, but also to the cavalry of the Romans and the best of their Italian allies which were superior to that of his Gallic auxiliaries. In the year 170 B.C. envoys arrived in Rome from Cinci- bulus, a king of the Gauls. One of the king's brothers addressed the Senate and complained that C. Cassius, one of the consuls of the previous year, had ravaged the lands of the Alpine peoples, who were in alliance with Rome, and had carried thousands of persons into slavery. At the same time envoys came also from the Carni, the Istri, and lapodes with similar complaints. As Cassius was absent in command of an army in Macedonia, the Senate could not take any immediate action, but wishing to appease the anger of the injured tribes they not only sent commissioners to examine on the spot into the charges brought against Cassius, but also loaded the Gaulish envoys with presents, especially the two brothers of Cincibulus. It was decreed that they should be given two torques made out of five pounds of gold, five vases made out of 20 pounds of 1 Livy, XXI. 46, " Scipio iaculatores et Gallos equites in fronte locat ; Komanos sociorumque quod roboris fuit, in subsidiis." 102 THE HORSES OF PREHISTORIC [CH. silver, two caparisoned horses {equi phalerati) with their grooms, cavahy armature, and cloaks, and gifts of apparel for the members of their retinue, bond as well as free. These were the voluntary gifts of the Senate. But at their own request each of the two brothers was granted the right of buying ten horses, and permission to export them out of Italy ^ Twenty stallions brought back by the Gallic chieftains to their home beyond the Alps would in a very few years produce a great effect on the quality of the little indigenous horses, even if no fresh blood was imported. But Caesar himself, in a passage shortly to be fully cited, points out that the Gauls were always importing foreign horses and paying very long prices for them. But as Caesar contrasts the excellence of the Gallic horses with those of the German, it follows that the horses imported by the Gauls must have been brought from the countries lying south of the Alps and Pyrenees. It is very significant that on the series of silver Gaulish coins, the earliest of which may be dated from about 150 B.C., and which from the first commonly display native types and not imitations of Greek or Roman issues, a horseman is one of the most favourite types, whilst practically the chariot nowhere appears, although it forms the regular type on the reverse of the gold coins imitated from the gold stater of Philip II. of Macedon, which bore on one side the head of Apollo, on the other a two-horse chariot I It would appear that by Caesars time the Belgic tribes, who occupied all the region bounded by the Marne, the Seine and the Rhine, had given up the use of chariots, although their brethren who had crossed into south-eastern Britain continued to employ them in warfare, for Caesar does not refer to the use of war-chariots by the former in any of his campaigns against them. Probably by his time they had obtained horses of a kind fully suited for cavalry, and had therefore given up the use of the chariot in war; their relations in Britain, though able to put a considerable number of mounted men into the field, 1 Livy, xLiii. 5, " Ilia petentibus data, ut denorum equorum eis commercium esset, educendique ex Italia potestas fieret." 2 Bidgeway, Origin of Metallic Currency, p. 'JO. Ill] AND HISTORIC TIMES 103 still retained the war-chariot, whilst with the original tribes of the interior, who had none but very small horses, the chariot apparently still reigned supreme. The monuments of northern Italy demonstrate likewise that though the horse was in use in Italy from the beginning of the Iron age, as is proved by the discovery of bronze horse- bits associated with remains of that period both at Este and Bologna, the chariot was still employed by the Umbrians, though the art of riding on horseback was becoming known. At Sesto Calende, near the point where the Ticino issues from the southern extremity of Lake Maggiore, was found a tomb of the Iron age. It contained a helmet made of plates of bronze rivetted together, two bronze greaves, a very short sword, a lance-head, arrow-heads, two horse-bits, two iron circles (the tires of the chariot-wheels), two large hollow objects, and other pieces of iron belonging to the chariot. The horse-bits are bronze mounted in iron. There was also a bronze bucket ornamented with horsemen, footmen, stags, birds, and dotted circles, and dotted lines. This bucket is one of a class well known in the region lying on both sides of the head of the Adriatic, and may be assigned to the sixth or seventh century B.C. ; from its evidence we may infer that the peoples of those regions had learned to ride the horse at a period much earlier than the tribes beyond the Alps, a conclusion in complete harmony with the evidence of Herodotus respecting the Sigynnae. To the question of the development of riding in Italy we shall presently return. The evidence here stated makes it clear that the Umbrian tribes who had passed down into Italy in the Bronze age and had subdued or driven back into the mountains the aboriginal Ligurians, employed the horse and the chariot from the early Iron age onwards, and that when the Celts, the close kinsmen of the Umbrians, crossed the Alps at a later date, they too came as a chariot-driving and not as a horse-riding folk. We have seen, that though by Caesar's time the Gauls were well- supplied with cavalry, and the war-chariot was virtually extinct, yet they never abandoned its use until they had obtained a superior breed of horses from the southern side of the Alps. 104 THE HORSES OF PREHISTORIC [CH. The Veneti, who lived at the head of the Adriatic, and who have left their name in Venice, had a peculiarly good breed of horses at an early date, or to speak more accurately one family among them owned this valuable possession. These horses were branded with a representation of a wolf. "They were remarkable," says Strabo\ "more for speed than for beauty." According to the story the breed had its origin thus : A man notorious for the readiness with which he became surety for others happened to fall in with some hunters who had a wolf in their nets. They asked him in jest if he would go bail for the wolf on condition that he would become responsible for all the damage the beast had done. He agreed, and the wolf being set free at once went and drove a herd of unbranded horses to the steading of his benefactor; he accepted the gift and branded them with a representation of a wolf His descendants kept both breed and brand, and in order to retain the pure strain in their own hands made it a rule never to part with a mare. The statement that these horses were more remarkable for speed than for beauty would of itself suggest that they were only an improved breed of the little horses of central Europe. This is actually confirmed by ancient testimony, for Aelian^ when describing under the name of Lycospades the horses called Lycophori by Strabo, speaks of them as in appearance thickset and short, and also with flat noses. Dionysius, the despot of Syracuse (B.C. 405-367), got some of the Venetian breed for his stud, in consequence of which Venetian colts became known in Greece and the breed long enjoyed a high reputed But by Strabo's time it had died out and the Veneti had given up altogether the breeding of horses *. There was a shrine in their land said to be dedicated to Diomedes, in which white horses were sacrificed to the herol White horses, such as those bred by the Veneti, were held in great esteem in Sicily, and it is highly probable that the famous four-horse chariot drawn by white horses in which Dionysius regularly rode was horsed by the imported Venetian 1 215. ^ H. A. XVI. 24, TTjv oij/LV txovcTL ffwearpaixtilvav kol ^paxiiav, in Si aifjirfv. 3 Strabo, 215. •» Id. 211. « Id. 215. Ill] AND HISTORIC TIMES 105 steeds and their progeny. This practice, which was not fol- lowed by either Hiero or his son Gelon, was revived by the foolish Hieronymus, who clad in purple and wearing a diadem used to drive forth from the palace in a quadriga drawn by white horses, like a second Dionysius^ White horses were apparently in favour with the Sicilian aristocrats at the close of the fifth century B.C. According to Diodorus Siculus^ Exaenetus of Agrigentura, on returning home after his victory with his chariot at the Olympic games in 412 B.C., was brought into the town escorted by 300 bigae drawn by white horses. The statement that white horses were sacrificed by the Illyrian Veneti in a shrine called after Diomedes by the Greeks has every stamp of truth, for we know that it was a general practice amongst the Illyrians to sacrifice horses to a deity identified with Cronus by the Greeks. Moreover, we shall presently see that white horses were held in special esteem by the tribes of Germany, and we shall find that the sacrifice of horses was a characteristic of the religion of the Teutonic and Scandinavian peoples. The value set on white horses by the Sicilian Greeks and by various other peoples both ancient and modern was due not to any superiority in speed or other qualities, but rather to the sanctity attached to animals of a white colour, as for instance to white elephants in Further India, and to white asses in Persia. At the dawn of history all the peoples of the Balkan penin- sula like those of the Italian seem to have kept horses, but they all appear to have used the chariot and never mounted the steed. The Upper Balkan was occupied almost wholly by the closely related Illyrian and Thracian tribes on whom the fair-haired people known as Celts to the Greek writers of the classical period, were constantly pressing down. These Celts were distinguished from the indigenous tribes, not only by their xanthochrous complexion, but by the fact that whilst all the Illyrian and Thracian tribes tattooed, the Celts never followed this custom. The Thracians tattooed themselves with figures of animals such as deer, which were probably their 1 Livy, XXIV. 5. - xiri. 82. 106 THE HORSES OF PREHISTORIC [CH. tribal badges, or even totems. The wolf-brand used to mark the horses of the Veneti was probably the badge or perhaps totem of the clan which owned them. Though the Thracians were using oxen for draught (Fig. 46) in the sixth century B.C., and though by that time the riding of horses must have been very familiar to them from their Greek neighbours on the south and the Scythians on the east, yet it seems certain that two-horse chariots continued to be used by certain peoples in Thrace down to late times. On the hills Fig. 46. Thracian coin showing Qj^^^^j. which surround the valley oi the Kritchma, the last affluent of the Maritza (ancient Hebrus) before the latter reaches Philippopolis, there are many large tumuli, which have been partially explored during the last fifty years'. In a pit close to the most remark- able of these, called Doukhova Moghila (" The Barrow of the Spirit"), in 1851 the brothers Shkorpil found the remains of a chariot and a pair of horses. Ten years later a peasant found the remains of another chariot and pair of horses close to the same spot. MM. GuerofiP and Berti commenced working at this spot, and at a point nearer to the tumulus they found a body in an upright position, the skull broken, and with an arrow-head still sticking in one of the ribs. Horses and chariots had been placed in trenches running east and west and had then been covered with earth. They found various objects in iron, bronze ornaments for the bridles, iron bits with bronze attachment, and bronze statuettes for ornamenting the body of the car, consisting of horses, bears sitting-up, a Poseidon, and two plaques bearing in low-relief horses' heads, incrusted with silver. lu 1877-8 the Russians quartered at Philippopolis excavated at the same spot and were said to have found a chariot as well as a silver disc, a silver cup, and a three-legged table. In 1888 sixteen pits were opened and according to MM. Shkorpil each grave ^ Georges Seure, "Voyage en Thrace," Bull, de Corrcspondance HelUnique, Vol. XXV. (1901), pp. 156 sqq.. Figs. 11—23. in] AND HISTORIC TIMES 107 contained the bones of two or four horses of a small breed (Equas caballus minor, Linn.), but it does not appear that more than four or eight chariots were discovered. In 1898 M. Dobrusky found near the old workings the fittings of iron and bronze belonging to a chariot and various human skeletons. In 1899 Fig. 47. Grave-stone, Mycenae. 1900 M, Seure made further explorations aud discovered a chariot, which he has described with admirable minuteness, and the remains of horses of a small size. M. Seure would refer this interment to the fourth century A.D., aud would assign it to a settlement of Scythians in Thrace, on the ground that the 108 THE HORSES OF PREHISTORIC [CH. Thracians did not bury slaves and horses with their chiefs, whilst the Sc^'thians did both. But there is nothing in the ancient statements^ respecting the Thracian funeral customs to hinder us from believing that the Thracians might occasionally so honour a great chief inasmuch as concubines were regularly put to death and all kinds of victims were offered. On the other hand, as the Scythians did not use chariots in the time of Herodotus, it is most unlikely that they would have resumed their use in the centuries after Christ, when all other peoples had taken to horseback. The monuments of the Bronze age of Greece, commonly termed the Mycenean period, furnish the earliest evidence of the use of chariots and horses in that country. These monuments are the relics of the Pelasgians, who were the indigenous people of Greece, and also the close congeners, of the Illyrians and Thracians; and the grave-stones of the acropolis of Mycenae, on three of which are sculptured in low relief a man driving a two-horse chariot with four-spoked wheels (Fig. 47), may be placed in the fourteenth century B.C. The Homeric poems furnish us with very copious evidence from at least 1000 B.C. respecting the method of employing horses, their breeding, their management, and their colours, not only for Greece itself, but also for Thrace and Asia Minor. We shall first examine the evidence for Thrace. That the Thracians used chariots in war is shown by the episode of the slaying of Rhesus the Thracian king and twelve of his best men. Dolon, the Trojan spy, when captured by Odysseus and Dioraede, said that if they desire "to steal into the throng of the Trojans, lo, there be those Thracians, new-comers, at the furthest point apart from the rest, and among them their king Rhesus, son of Eioneus. His be the fairest horses that ever I beheld, and the greatest, whiter than snow, and for speed like the winds. And his chariot is fashioned well with gold and silver, and golden is his armour that he brought with him, marvellous, a wonder to beholds" 1 Herod, v. 4 ; Pomponius Mela, ii. 2, 18 ; Solinus, x. 28. 2 II. X. 433 sqq. (Lang, Leaf, Myer.s). Ill] AND HISTORIC TIMES 109 "Then the twain went forth through the arms, and the black blood, and quickly they came to the company of Thracian men. Now they were slumbering, foredone with toil, but their goodly weapons lay by them on the ground, all orderly, in three rows, and by each man his pair of steeds. And Rhesus slept in the midst, and beside him his swift horses were bound with thongs to the topmost rim of the chariots" Then Diomede fell to slaying the sleeping Thracians^ " but whom- soever he drew near and smote with the sword, him did Odysseus of the many counsels seize by the foot from behind and drag him out of the way, with this design in his heart, that the fair-maned horses might lightly issue forth, and not tremble in spirit, when they trod over the dead ; for they were not yet used to dead men." Then when Diomede is slaying Rhesus himself, " meanwhile the hardy Odysseus loosed the whole-hoofed horses, and bound them together with thongs, and drave them out of the press, smiting them with his bow, since he had not taken thought to lift the shining whip with his hands from the well-dight chariot. Diomede pondered, whether he should take the chariot where lay the fair-dight armour, and drag it out by the pole, or lift it upon high, and so bear it forth," but yielding to the monition of Athena he " swiftly sprang upon the steeds, and Odysseus smote them with his bow, and they sped to the swift ships of the Acheans." When they had come thither Nestor was the first to hear them, and when they had leaped down to earth the old man asked them whether they had won the horses by stealing into the press of Trojans, or had some god given them to them. " Wondrous like," said he, "are they to the rays of the sun. But never yet saw I such horses, nor deemed of such ^" Plainly then the white horses of the Veneti to which we have just referred were no new feature in the countries south of the Danube, but it is clear from the words put in the mouths of both Nestor and the Trojan Dolon that white horses were unknown both in Greece and on the Asiatic side of the Aegean. 1 II. X. 469 sqq. ^ U. x. 480 sqq. ^ H. x. 543 sqq. 110 THE HORSES OF PREHISTORIC [CH. This inference is fully borne out by the evidence touching the colour of horses supplied by the poems themselves. The Iliad and the Odyssey, whicli present us with immortal pictures of the fair-haired Acheans, who in the early Iron age became the masters of the Pelasgians of upper Greece and the Peloponnesus, represent those heroes as breeders and drivers of horses. The warrior goes to battle in a two-horse chariot witii his charioteer beside him, as was the practice of the Celts of Gaul down to the century before Christ. We have also clear evidence respecting the type and colour of their horses. The evidence of the Iliad amply suffices to show that the horses bred and used by the Acheans were almost uniformly dun-coloured, for the epithet xanthos, commonly applied to them, was used by the Greeks to describe the colour of gold and golden-coloured hair. In two passages at least this epithet is applied generically to Achean horses. Achilles when he rejects in scorn the gifts proffered by Agamemnon, exclaims, " Kine and goodly flocks are to be had for the lifting, and tripods and yellow-dun {xanthos) horses can be bought ; but to bring back man's life neither harrying nor earning availeth when once it hath passed the barrier of his lips^" Again, Nestor relates how once he headed a foray into the land of the Eleans — the land in which Pelops and the Acheans had especially established themselves — " and a prey exceeding abun- dant did we drive together out of the plain, fifty herds of kine, and as many flocks of sheep, and as many droves of swine, and as many wide flocks of goats, and yellow-dun {xanthos) horses a hundred and fifty, all mares, and many with their foals at their feet"''," The horses of Achilles which had been given to his father Peleus by Poseidon himself were named Xanthos {Dun) and Balios {Dapple), " swift horses that flew as swift as the winds, the horses that the harpy Podarge bare to the West W^ind, as she grazed on the meadow by the stream of Oceanus^" Here we have the earliest reference to the belief so common in classical times that the fleetest horse came from the West 1 II. IX. 407 sqq. - II. xi. 080 sqq. ^ II. xvi. 149 sqq. in] AND HISTORIC TIMES 111 and that in that region the mares became impregnated by the west wind, an aetiological myth to explain the swiftness of steeds, who rivalled in speed the strong west wind from the Atlantic. As halios is regularly applied to deer and to the lynx^ there can be no doubt that when it is used of horses it means dappled, and accordingly Achilles' steed was a dappled-dun. These famous steeds had very heavy manes. After the slaying of Patroclus they kept " apart from the battle weeping, since first they were aware that their charioteer was fallen in the dust beneath the hands of man-slaying Hector. They stood unmoved abasing their heads unto the earth. Hot tears flowed from their eyes to the ground as they mourned in sorrow for their charioteer, and their rich manes were soiled as they drooped from beneath the yoke-cushion on both sides beneath the yoke ^" Again, when at the funeral games of Patroclus Achilles lays down prizes for the chariot-race, he says that if in some other's honour the Acheans were holding games he himself would compete : " but verily I will abide, I and my whole-hoofed horses, so glorious a charioteer have they lost, and one so kind, who on their manes full often poured smooth oil, when he had washed them in clear water. For him they stand and mourn, and their manes are trailing on the ground, and there stand they with sorrow at their hearts^" In the horses of Achilles, with their long, heavy manes, one dun-coloured, the other dappled-dun, we can recognize the same breed of horses as those used by the Sigynnae of central Europe in the fifth century B.C., and this identification will gain further confirmation from colour and other arguments as our investi- gation proceeds. As the poet seems carefully to note any peculiarity of colouring in horses, such as those of Rhesus already described, and those of Aeneas to be discussed later on, 1 Eur. Hec. 90, Ale. 579. " II. XVII. 437 sqq. : oy'Set ivuTKLtixpavTe Kaprjara' dcLKpva 34 aos, the other e(TKLa.Tpo7]Kvia. As the men seen on vases or other Daphnae fragments are painted black we may be certain that the white nude figures are women and not men or boys. ■* Petrie, op. cit. PI. xxxi. figs. 13, 14. ' The Greeks held that there were Amazons in Libya as well as in the region round the Black Sea. As the Sarmatian women who hunted on horse- back with their husbands gave rise to the myth of the eastern Amazons, so the fable of the Amazons of Libya arose from the horse-riding Libyan women (cf. Bidgeway, Early Age of Greece, Vol. i. pp. 651-2). Ill] AND HISTORIC TIMES 245 beautiful dark horses can be no other than Libyans, who at the time when these vases were painted were a standing danger to the Egyptians, and must have been perfectly familiar to the Greeks who were employed to secure the Egyptian frontiers. The Scythian women, however, even down to the second century a.d. had not learned to ride, as do the Tartar women (cf. p. 139) of to-day, but they still travelled from place to place in their waggons as in the time of Herodotus. We have seen that the setting on of the tail is a character- istic of the pure-bred Arab and his well-bred derivatives, a point to which great attention is paid by horse-fanciers in Western Asia and India. Though unfortunately in none of the Daphnae pictures of horses ridden by women has the tail survived, yet on two other fragments^ we can clearly see that the horses' tails are set on high like those of well-bred Arabs and the horses seen under the chariot of Seti I (p. 217). Again, the Daphnae pottery- yields the earliest known repre- sentation in painting of the winged horse Pegasus, whose birth- place according to the legend was the Libyan desert. The winged steed was naturally modelled after the fleetest courser known to the artist, who has simply added wings to indicate his supernatural swiftness. In the Pegasus of the Daphnae vase the tail is clearly set on high. It is now plain that as in the Egyptian paintings of the New Empire so in the sixth century B.C. the high set of the tail, as well as the dark colour — two of the features of the Kohl breed of Arabia — characterised the horses of North Africa. In the small lightly built and docile horses of the Libyans we recognise the light built Barb and 'Arab' of to-day, a horse of matchless swiftness, when carrying a light rider, such as the typical Numidian horsemen described by Livy (c£ p. 241), but not well adapted for a heavy weight or for draught purposes. The horse ridden by the Libyan women on the Daphnae frag- ments is not a thoroughbred, but a fine cross-bred horse. We have seen proof that the great desire of the ancients was to 1 Brit. Mus., B 129, 10 (two horsemen) ; B 125, 8 (a horse with the upper part of tail well defined, though a portion is lost). 2 Brit. Mus., P 105. 246 THE HORSES OF PREHISTORIC [CII. obtain horses of large size, and we shall presently find that the same principle has been always at work in medieval and modern times. Thus in the Iliads the horses of Rhesus are commended by Dolon as 'the largest' that he had ever seen, and the Nisaean horses of the Persian kings were ' the largest ' in Asia. Again, we have seen that in modern times the Arab tribes of the Persian Gulf have fatally injured their fine breed of horses by crossing it with large Persian strains in order to produce for the Indian market animals capable of carrying greater weight, and we have also noticed that Niebuhr in the eighteenth century preferred the half-bred horses of Syria to the Kochlani (p. 167). It is more than probable that the same principle was at work in Egypt from a comparatively early date, and also in those parts of North Africa occupied by Greek and Phoenician colonies. It will be presently shown that long before the Christian era horses were imported into Libya from Europe, and that in Roman times the North African horses crossed with Spanish blood were especially esteemed. It is therefore probable that in the horses of the Daphnae fragments the Greek vase-painter has pourtrayed animals produced by crossing the Libyan and the Asiatic horse, and we shall furnish evidence of similar half-bred horses at Carthage, in Sicily, and in Greece by the fifth century B.C. In fact the horses of Daphnae, Carthage, Sicily, and Greece stood to the small slender uncrossed horses of the Numidians in much the same relation as do the coarse Arabian horses found to-day in Syria and Irak to the pure-bred Keheilans, It is now clear that for many centuries before the Arabs ever owned a horse, all the Libyan tribes possessed a most notable breed, which in size, shape, speed, colour, and docility, very closely resembled the Kohl breed of Arabia. As it has been shown that Egypt was exporting horses into Asia Minor in the time of Solomon, and that Arab tradition points to Egypt as the region from whence the best horses were obtained in the time of Muhammad, and as Egypt derived her horses in great part from Libya, we are justified in concluding that the ancestors of the Kohl breed of Arabia came from North Africa. This conclusion is strongly corroborated by a fact already Ill] AND HISTORIC TIMES 247 mentioned. It will be remembered that Al Khamseh is un- known in North Africa. Yet if the Kohl breed had been developed in Arabia, and had been brought by the Arabs into North Africa at the time of their conquests or at a later period, the fivefold division of their famous breed would have been most religiously preserved by the Arabs in their new homes. On the other hand, if the Kohl breed really originated in North Africa, being merely the ordinary horse of that region, there would have been no reason for dividing it into special families, for it is only when a strain of special quality is intro- duced from elsewhere that people begin to pay attention to pedigrees, as in the case of our own thoroughbreds. If the Arabs derived their Kohl breed from North Africa, there would have been every reason for paying great attention to purity of race and carefully discriminating between different sub-families sprung from a common stock, in some instances crossed with Asiatic horses. This is in perfect accord with the statement of Abd-el-Kader that the five families of Al Khamseh are but ramifications of the ancient Ahwaj race, which, though not to be found any- where in Arabia, is said to still exist in the Sahara. The historical evidence therefore is unanimous in pointing to North Africa as the source of the Arabian Kohl breed. The Libyans had domesticated horses at a very early period and had learned not only to drive them in pairs under very light chariots, but had also inveuted the four-horse chariot. Moreover, they soon learned to ride these horses, for if their women habitually did so in the seventh century B.C., we may infer, not unreasonably, that the men had learned to ride at a still earlier period. The Libyans who furnished horses and chariots to Xerxes in B.C. 480 must have come from the country bordering on Egypt and were therefore probably the Marmidae. By the time of Christ the chariot had been discarded by all the Libyans with the exception of the Pharusii and the Nigretes who lived south of the Atlas, and who probably only retained them (like the Persian and Seleucid kings) because the chariot when fitted with scythes was still thought to be a valuable engine in war. 248 THE HORSES OF PREHISTORIC [CH. Strabo, unfortunately, does not give us any information respecting the colour of the Libyan horses, but we can infer from the pottery pictures of Daphnae that they were of a dark colour, which harmonizes well with the fact that the horses on Egyptian monuments are usually painted brown. As it will be remembered that dark colour is the characteristic of the pure-blooded Arabian, and that the grey and white Arabian horses come from regions where Turcoman horses have been in universal use, the agreement in colour between the Arab horses of the Kohl race, and those of ancient Libya, and the Egyptian monuments indicates clearly that the Arabs have not derived their famous breed from the dun or white horses of Persia and Upper Asia, but from a dark-bay stock of Libya. Turning to the North African horses of to-day, we find that in the Barbary States the prevailing colours are dark-bay, brown, chestnut, black and grey. In Morocco the horses are said to be of every colour, but black and chestnut are considered the best. The black colour as well as grey is probably due to importations from Europe and Asia dating from a long time back down to recent times. In the last century the Sultan of Ducaila imported a black English thoroughbred stallion and thereby obtained a splendid breeds On the other hand we find the Roman Senate sending to Masinissa, the Numidian king, two militar}^ cloaks fitted with two golden brooches each, two horses fully equipped, and two sets of cavalry accoutrements, including breastplates^ Plainly, then, from the second century B.C. the Barbary horses could have been crossed with the heavier and stronger strain of Italy. In later centuries the Barb has been largely crossed with imported Syrian Arabs, which of itself is sufficient to account for the occurrence of grey and black amongst them. The Barbs of Algeria have in modern times been much crossed with French and English blood, and have consequently lost a good deal of their original type. It is only in Morocco that it has been kept at all pure, for few horses are imported into that country. The Morocco Barbs are excep- ' Hamilton Smith, op. cit., p. 224. - Livy, xxx. 17. Ill] AND HISTORIC TIMES 249 tionally hardy, enduring and useful animals, and this breed is officially preserved by the reigning sultan ^ The Barb ranges from 14 to 15 hands high, but is said to be a little smaller than the Arab, with flat shoulders, round chest, ioints inclined to be long, and the head particularly beautiful. In the sandy plains south of the Atlas (the ancient home of iiG. 73. Dongola Horse. the Pharusians and Nigretes), the Mograbins of the West rear horses, which are brown or grey, known as Shrubat-ur-Reech, " Drinkers of the Wind ; they are rather low, shaped like greyhounds, and destitute of flesh-." The horses found in the region of Dongola are bay, black, and white, but not grey, and never dappled. The blacks are 1 Hayes, Points of the Horse, pp. 627-8. - Hamilton Smith, p.' 227. 250 THE HORSES OF PREHISTORIC [CH. the finest; they have all white legs (Fig. 73), sometimes the white extending over the thighs and occasionally over the belly^ We have seen at an earlier stage that the presence or absence of hock callosities has been generally taken as one of the chief means of differentiating Equus caballus from the asses and zebras (p. 16). It is therefore to be carefully noted that, whilst the hock callosities are present in Prejvalsky's horse, and are of specially large size in domestic horses of heavy breeds, accord- ing to Sanson^ they are frequently absent in North African horses (as is the case with Ewart's 'Celtic' pony). The presence therefore of such hock callosities in Arab horses of a coarser type may be due to their having in their veins a considerable ad- mixture of Asiatic blood, as is certainly the case with many 'Arabs' from Turkish Arabia, and the tribes of South Arabia, which border on the Persian Gulf (pp. 183-6). It has already been pointed out that white feet and a star or blaze on the forehead are characteristic of the pure Anazah bay horses, and it has also been shown that the fine black horses with white feet — the best bred by the Turcomans — are the result of crossing the horses of Upper Asia with so-called Arab blood (p. 133), that black horses are never found in Al Khamseh, and that the best black horses found in Syria belong regularly to the Jelfon breed ; it has also been shown that the 'Arab ' horses sent in large numbers into Egypt are exported from Syria, where a large proportion of the horses are grey, such horses being the result of crossing common Turcoman mares with pure Arab horses. It is therefore more than probable that all the black, grey, and white horses of North Africa have been derived either from Syrian and other horses of impure breed, or else have resulted from the blending in Africa itself of Asiatic and European horses with a native dark breed. As we have already seen that there were no horses in all North-east Africa or in Nubia down to the time of Strabo, 1 Hamilton Smith, o}). cit. pp. 229-30, PI. 10*, from which my illustration is copied. 2 Zootechnie, Vol. iii. p. 53 (ed. 4). Under the heading of "caract^res zoo- techniques generaux" he says that "les membres posterieurs sout depourvus de chataignes" in the " Race africaine" {E. c. africanus). in] AND HISTORIC TIMES 251 and as we have now shown that grey horses and black horses with white feet result from crossing the Upper Asiatic and the so-called Arab, it follows that the breed of Dongola is not a primeval stock {E. c. africanus) as was held by Sanson, but only a blend of comparatively modern origin. It is now clear that Egypt could have obtained from the Libyans the horses which she exported into anterior Asia in the tenth century B.C., and it is likewise certain that when the Greeks planted Cyrene in B.C. 632, they found the Libyans not only employing the two-horse and four-horse chariot, but also generally riding on horseback. Herodotus explicitly tells us that "the Greeks learned from the Libyans to yoke four horses to a chariots" It is therefore not without significance that the four-horse chariot and the ridden horse (keles) were only given places in the Olympic contests (the former in B.C. 680, the latter in B.C. 648) in the same century that saw the founding of Cyrene. The four-horse chariot does not seem to have been ever employed by any of the peoples of Upper Europe, by Vedic Aryans, Persians, Assyrians, Canaanites or Egyptians or by the Homeric Acheans. For although in two passages of Homer mention is made^ of "four male horses yoked together," these only refer to the occasional practice of attaching one trace-horse or two to the regular pair, under special circum- stances, either in war or for racing, an idea which may very well have been borrowed from Libya long before the foundation of Cyrene, for it is plain from Homer that the Homeric Greeks were well acquainted with that country. Moreover, it is very significant that one of the passages from Homer where mention is made of a four-horse chariot is a simile, whilst the other does not refer to a contemporary event, but is in a tale of a bygone age when king Augeas reigned in Elis before Pelops and his Acheans came and conquered the old Pelasgian inhabitants, Nestor relates how Neleus, the Pelasgian king of Pylus, had sent a four-horse chariot to Elis to race for a tripod and how Augeas had kept the horses and sent back the charioteer without them. But as the legends of the Bronze Age of Greece have frequent 1 IV. 189. - II. XI. 699-702 ; Odtjss. xiii. 81. 252 THE HORSES OF PREHISTORIC [CH. references to intercourse with Libya, we need not be surprised if the practice of yoking four horses abreast had been learned from that land. On the other hand it seems clear that the Achean conquerors clave fast to the practice of the region from whence they had come and continued to use only their national two-horse cars. In the chariot-race at the funeral games of Patroclus there is not even a hint of the use of the four-horse chariot or even of a trace-horse, a fact which contrasts strongly with the constant appeai'ance of the quadriga on the black- figured vases dating from the seventh century B.C., when the four-horse chariot had been reintroduced. Not only is it probable that the Greeks first learned the use of the four-horse chariot from Libya, but it is also not unlikely that it was from the same region that they first learned to mount on the back of the horse. Although the Homeric Acheans never normally rode their dun-coloured horses any more than they drove them in teams of four wdien they raced or went to battle, yet there is at least one simile in the Homeric poems which shows that riding on horseback was by no means unknown to the poet. Thus Odysseus, when his raft was shat- tered, " bestrode a single beam, as one rideth on a courser (Jceles) and stripped him of the garments which fair Calypso gave him^;" and in another simile, in the Iliad, if we do not hear of riding on horseback we have the earliest picture of a circus-rider, — " a man right well skilled in horsemanship, that couples four horses out of many, and hurrying them from the plain towards a great city drives them along the public way, many men and women marvelling at him, and unerringly ever he leaps and changes his stand from horse to horse, while they fly along-." Now when we bear in mind that the first horse ever ridden according to Greek legend was Pegasus, that this famous steed was born in Libya, and that he was obtained there by Perseus, the renowned king of Mycenae, in the Bronze Age, before the coming of the large-limbed, fair-haired Acheans from central Europe, it is not strange that the Greeks of the Homeric (Iron) Age knew that the horse could be ridden as well as driven, although the new 1 Od. V. 371. - n. XV. 679 sqq. Ill] AND HISTORIC TIMES 253 lords of Greece did not ride their dun-coloured horses, but only drove them in pairs under chariots. From the great docility of the Libyan horse and his descendants it is highly probable that the Libyans from its first domestication could mount it without difficulty, and as they themselves (cf. p. 241) were small, light- built men, their horses could carry them with ease at a date when the European horses, which were not so tall as the Libyan, could not carry for any great distance their large-limbed masters, such as the Sigynnae in the fifth century B.C. or the Acheans some seven centuries earlier. It is thus not at all unlikely that it was from Libya that the practice of riding on horseback first became known to the inhabitants of Greece. The islanders from Thera, who planted Gyrene, were not slow in fulfilling the prophecy of Medea^ that " instead of short- finned dolphins they should take to themselves fleet mares, and reins instead of oars should they ply, and speed the whirlwind- footed car," for Gyrene soon became famous as "the city of fair steeds and goodly chariots-." Pindar glorified her king, Arcesilas, for his victories in the chariot-race, and later her native poet, Gallimachus, sang of his " home famed for her steeds^" When Alexander had conquered Egypt and conceived the idea of visiting the shrine of Amnion, the Cyreneans sent envoys to make submission to the world-conqueror, bringing a crown and rich gifts amongst which three hundred war-horses figured prominently ^ Still earlier than the Greeks the Phoenicians had begun to plant colonies along the coast of North Africa, and they therefore soon became possessed of the noble horses of Libya. When Garthage, in the fourth century B.C., first began to coin money it is significant that she placed the horse and palm-tree on her coins (Fig. 74), whilst the horse alone is seen on the issues of Panormus, her most important settlement in Sicily. It will be noticed that the horses seen on these coins are not pure-bred, but rather fine cross-bred horses, like the horse ridden by the Libyan woman (p. 244). In reference to that 1 Find. Pyth. iv. 17. - lb. iv. 1. 3 Strabo, 837. * Diod. Sic. xlix. 2. 254 THE HORSES OF PREHISTORIC [CH. horse and the others pourtrayed on the pottery from Daphnae I have pointed out that, as in modern and medieval times people have bred horses for practical purposes rather than for speed alone, so the ancients were actuated by similar motives. No doubt the same reason influenced the Cartha- ginians, both in Africa and Sicily, and we shall see that the Sicilian and other Greeks of the fifth century B.C. and the Romans of the Republic and Empire were affected by the same considerations, though for purely racing purposes the Romans gave preeminence to the pure African horse. We have seen that the high set of the tail is one of the characteristics of the Kohl breed (Fig. 58). The tail of the well-bred horse on the Carthaginian coin of Panormus (Fig. 74) indicates that this was already a feature of Libyan horses in the fifth century B.C. At how early a date the Libyan horse made its way into Spain it is impossible to say, although from certain Greek legends, to one of which we have already referred, the Iberians may have known of the famous fleet steeds of the Libyan shore for many centuries before the date of the Carthaginian settle- ments in the peninsula in the third century B.C. There is, however, the clearest evidence from the Roman historians that the Libyan horses had been brought over to Spain in large numbers by 219 B.C. When Hannibal in that year prepared to march into Italy with 90,000 foot and 12,000 cavalry, a very considerable pro- portion of the latter consisted of Numidians, who rode their own native horses, without bridles, whilst the remainder was composed of Spanish horsemen who rode with bridles. The Romans first became acquainted with the Numidian horses and horsemen, whom afterwards, to their cost, they were to know too well, at the moment when Hannibal after seizing the passage of the Rhone was slowly ferrying his thirty-seven elephants across that river. On learning that Scipio had disembarked his army at the Massaliot mouth of the Rhone, Hannibal despatched a body of five hundred Numidian horse to reconnoitre. Scipio had also sent his cavalry forward with a like object. As soon as the Numidians met the enemy they Ill] AND HISTORIC TIMES 255 charged with their wonted impetuosity, and a desperate en- counter took place, but after some two hundred had fallen on either side, the Romans eventually drove their assailants back to Hannibal's camp\ As soon as he crossed the Alps and descended into Italy the cavalry skirmish at the Ticinus {Ticino) demonstrated that Fig. 74. Carthaginian Coins («. and c Carthage, h Panormus). the Carthaginian cavalry was far superior to that of the Romans and the Gallic cavalry serving with them, the latter in turn being inferior to that of the Romans. Though Livy gives no description of the horses ridden by the Numidians in Hannibal's army there can be no doubt that they brought their own horses with them, as was the case ^ Livy, XXI. 29. 256 THE HORSES OF PREHISTORIC [CH. with a body of eight hundred of the same bold horsemen who, after the Roman conquest of North Africa, served with the Romans in their campaign against the Ligurians, and saved the Roman consul and his legions from a complete defeat. Livy's graphic description of these Numidians and their small slightly built horses has already been given (p. 241). Hannibal had committed the care of Spain to his brother Hasdrubal, and had left him, in addition to a large body of African infantry, a force of cavalry composed of three hundred Libyphoenicians and eighteen hundred Numidians and Mauri- tanians from the region bordering on the Atlantic \ Over two thousand Libyan horses were thus at this time alone sent into Spain and kept there permanently. As probably most of them were stallions, since the Numidians and Moors did not use geldings and kept their mares for breeding, the influence which these two thousand horses exercised on the native breed within a short time must have been very great. For it is more than likely that the Spaniards would have sought eagerly to obtain the services of superior sires for their mares. We need not then be surprised that at the time when Posidonius, the Stoic philosopher who travelled in western Europe about 90 B.C., visited Spain, the Iberians and Celti- berians possessed horses of fine quality more or less impregnated with Libyan blood. He tells us that the Iberians used " cavalry interspersed among their footmen, that their horses were trained to traverse the mountains, and to sink down on their knees at a word from the rider, in case of necessity. They had also a practice not confined to them — two men mounted one horse so that in the event of an engagement one might be at hand to fight on foot." He does not mention the colour of the ordinary Iberian horses, but he gives us the very important information that the horses of the Celtiberians, who occupied northern Spain, were " rather starling-coloured " (i.e. dark grey flecked with white), but that they lost that colour when transported into southern Spain, and he compared them to the Parthian horses, " for indeed they are superior to all other breeds in fleetness and endurance." 1 Livy, XXI. 22. Ill] AND HISTORIC TIMES 257 We can at least gather from the statement of Posidonius that the Celtiberian horses, that is, those of northern Spain, differed essentially in colour from those of the Iberians, that is from the horses of Andalusia and other parts of southern Spain. But as the former seem to have been grey, it is clear that the latter were not that colour, but were probably mono- chrome. Unfortunately Strabo gives us no information respecting either the colour or form of the wild or feral horses which he mentions incidentally. We have therefore no means of judging whether they were more akin to the horses of the South or to those of the North, but, as in modern times dun-coloured horses (not unfrequently with a dorsal stripe) are found in the sierras of Spain, it is not unfair to infer that the old dun-coloured horses of upper Europe and Asia formed the substrate of the grey Celtiberian horses, just as the upper Asiatic dun horse is a main element in the grey horses of western Asia at the present day. Whether these dun horses were indigenous in Spain or brought in by the Celts in their invasion in the sixth century B.C., or whether they were of the heavy type, or a light type such as the ' Celtic ' pony, it is of course impos.sible to say. A well-known story told by Pliny^ indicates that horses of extraordinary fleetness were bred in Lusitania, for the tale went that in the region of the town of Olisipo and the Tagus the mares were impregnated by the West Wind and brought forth an offspring of surpassing fleetness, which however lived only for three years. From this we may fairly infer that a very swift breed of horses existed in that part of Portugal, and that they were not the ordinary slow upper European horse ; but whether they were a slight built indigenous race connected with the ' Celtic ' pony, or whether they were the descendants of horses introduced from Libya, it is of course impossible to say. On the other hand, we also learn from Pliny^ that north- 1 H. N. VIII. 166 : constat in Lusitania circa Olisiponem oppidum et Tagum amnem equas Favonio flante obversas animalem concipere spiritum, idque partum fieri et gigni pernicissimum ita, sed triennium vitae non exccdere. Of. Tasso, Gerusalemme Liherata, canto vii. 2 loc. cit.: in eadem Hispania Gallaica gens et Asturica equini generis,— hi R. H. 17 258 THE HOKSES OF PREHISTORIC [CH. west Spain produced a breed of horses, those of Gallaecia (Galicia) and Asturia (Asturias), the former known to the Romans as celdones (Galicians), the latter (which were of smaller size) as asturcones (Asturians). The pace of these horses was an easy amble, but we are not told their colour. It has been universally held up to the present^ that "when the Saracens conquered Spain in the eighth century they brought from Africa many saddle-horses of the Eastern type, which, later on, were crossed with the heavier native Spanish horses." But the evidence just cited renders it certain that the fine docile breed of Libyan horses had been planted in Spain some ten centuries before the Saracen conquest, and some three or four before the Arabs ever owned a horse. The Barbary blood in the horses of Andalusia and Grenada was largely reinforced by the Saracen conquest, for the Moors, like their ancestors who served in Hannibal's army, brought over with them their own native horses. Let us now examine the colours of the modern Spanish breeds of horses. It is of course in southern Spain that the Barbary blood especially prevails, and through the Middle Ages down to our own times the horses of Andalusia, Grenada, and Estremadura have been especially esteemed. The pre- dominant colour is bay, next to which come black and grey. According to a Spanish proverb, " a mulberry-black horse is what everyone should wish for, though few may possess." Black horses without a white mark or with only a star in the forehead are especially valued. Early writers, such as Absyrtus, note that Libyan horses " be of like goodness, and of like shape to the Spanish, save that the Libyans be stronger made, longer bodied, thicker ribbed, and broader breasted." This similarity was doubtless due not only to the fact that Libyan horses had been imported into Spain, but that later on, as certain writers tell us, Spanish horses were imported into North Africa and crossed with the suntquos celdones vocamus, minore forma appellatos asturcones, — gignunt quibus non vulgaris in cursu gradus, sed mollis alterno crurum explicatu glomeratio, unde equis tolutim capere incursum traditur arte. 1 Hayes, Points of the Horse, p. 508 (ed. 3). in] AND HISTORIC TIMES 259 native horses there. It was the fine-bred Andalusian horse which was termed the Spanish jennet by our older English writers, as is clear from Blundeville's description of that animal : " The Jennet of Spaine is finelie made, both head, bodie, and legs, and very seemlie to the eie, saving that his buttocks be somewhat slender, and for his fine making, lightnesse, and swiftnesse withall, he is verie much esteemed, and especiallie of noble men, as Camerarius saith, which Oppianus also affirmeth, saieng : that the Jennet in swiftness passeth the Parthians and Fig. 75. The Jennet of Spain. all other horses whatsoever they be, even so far as the Egle exceedeth all the birds in the aire, and as the Dolphin passeth all the fishes in the sea, but therewith he saith that they be but small of stature, of small strength, and of small courage, all which things seeme to agree verie well with those Jennets that be brought hither into England, unlesse it be the last point. For I have heard some of the Spaniards to set such praise on their Jennets' courage, as they have not letted to report, that they have carried their riders out of the field, I cannot tell how manie miles, after the Jennets themselves have been shot cleane through the bodies with Harquebushes. 17—2 260 THE HORSES OF PREHISTORIC [CH. Which report I have heard to be true by divers of our owne souldiers, which if it be true in deede, it doth better countevaile their small stature and little strength, which is manifest to all men's eies that do behold them." This description is admirably illustrated by Stradanus' drawing of the ' Hispanus ' here reproduced (Fig. 75). The horses of northern Spain are grey and rufous-grey, and are of a smaller size than the Andalusian^ In them we have no difficulty in recognizing the Celtiberian, Galician and Asturian breeds of Strabo and Pliny. In the dun-coloured" animals (often striped) of the sierras we have the less modified descendants of the old European horses, but in view of the continual inter- crossing for many centuries of the Libyan and European blood, and in face of the facts relating to the dun- coloured striped horses of Kattywar (p. 157), it would be rash to assume that these striped dun horses of Spain are really pure representatives of the old dun horses of Europe. The curious statement that the Celtiberian horses lost their distinctive speckled coats when removed into southern Spain is readily explained. The actual horses exported did not change their colour, but as they became mixed with the dark Libyan stock in that region, the latter blood predominated and the grey horses merged into the dark bay and black. The horses of the Asturias and other mountainous areas of Spain are probably descended from the old European large- headed horse, which may have continued in a wild state in Spain down to the Christian era, since Posidonius mentions horses among the wild animals of Spain. Of course these horses may have been simply feral horses, but on the other hand there is no reason why genuine wild Equidae should not have still survived in wild and mountainous districts. We have seen that Strabo compared the Celtiberian horses, that is, the horses of northern Spain, to the Parthian horses, which were the descendants of the Nisaean horses, the best in Asia in the fifth century B.C. ; it has also been shown that ' Hamilton Smitb, op. cit. p. 247. - Isidore of Seville {Origij. xii. 1) says that duu (cinereus) horses are the worst (but of. p. 348). Ill] AND HISTORIC TIMES 261 the Libyan horses were being imported into Asia Minor through Egypt at least as early as 900 B.C., and additional evidence will soon be offered to show that the Libyan dark bay horses with a star in the forehead were known in north-west Asia Minor by at least 1000 B.C. It has been demonstrated that in western Asia bay pre- dominates largely among the pure North African breed of the Anazah tribes, and that all the breeds of western Asia which have originated from the light-coloured upper Asiatic horse by the admixture of Arab, i.e. Libyan, blood are generally grey, rufous-grey, iron-grey, and black, as are to-day the horses of northern Spain which are derived from the old European light- coloured horse improved by Libyan blood. As then the crossing of the bay Libyan blood with the dun- coloured horse of Asia and Europe has produced the same results in Spain and western Asia, we are justified in concluding that grey horses are not an original stock as has been held by some, but are the result of the crossing of Libyan and Asiatic blood. The same holds true of black, for it is found in the areas where the two primitive stocks overlap, whilst it seems unknown among true Asiatic-European horses on the one hand, and pure Libyan horses on the other. We shall soon see that among the feral Pampas horses, which are descended from Andalusian ancestors, barely one in two thousand is black. It seems certain that the grey and black element in the horses of Andalusia and possibly even in those of Morocco, is derived from the grey half-bred horses of the Asturias and Murcia. Again, as it has been shown (p. 157) that the dun-striped horses of Kattywar are the result of crossing the upper Asiatic dun horses with Libyan blood, so too the striped dun horses of the sierras of Spain are probably due to crossing the old dun European horse with the same Libyan stock. South America indeed not only had fossil horses, such as the small Hippidium of the Pampas of Argentina and various long-limbed species, but (unlike North America) had horses at the first advent of man, as is proved by the existence of horse remains associated with worked stone implements, pottery 262 THE HORSES OF PREHISTORIC [CH. and fire refuse in the lake formations of the Upper Plei- stocene \ But nevertheless there seems no doubt that it was extinct before the coming of the Spaniards, and that the Pampas horses of South America are all descended from Andalusian horses introduced by the Spaniards in 1535, when the city of Buenos Ayres was founded by Don Pedro de Mendoza. The place, however, was almost immediately deserted, the inhabitants passing over to Paraguay by water in such haste, and with such lack of means of transport, that they were unable to carry along with them all the horses brought from Andalusia ; five mares and seven horses were left behind on the plain. The city was refounded in 1580 by Don Juan de Garay, accompanied by sixty colonists from Paraguay. " These in- dividuals found that a considerable breed had already sprung from the above-mentioned mares, and set about domesticating those which they were able to catch ^." The ministers of State opposed this, asserting that they belonged to the king. After protracted litigation it was decided in 1596 that the wild horses should be the property of whosoever should take the trouble to capture them. This is the origin of the innumerable herds of wild horses which are met with to the south of the La Plata as far as Rio Negro, and even throughout Patagonia. These horses were at first, as now, called alzada and cimarrona, but the Querandese (commonly known as Pampas Indians) having given them the name of bagualada, the Spaniards adopted this name, and these horses are generally known as baguales. In Azara's time there were also baguales to the north of the river La Plata. These horses seem to have descended from some mares abandoned by the Spanish settlers of San Juan Bautista, a town founded by John Romero in 1553, right opposite Buenos Ayres, where the San Juan debouches into the La Plata. It was soon attacked by Indians, and the inhabitants crossed the river into Paraguay and were obliged, ^ H. F. Osborn, The Century Magazine, Nov. 1904, p. 13; Sir H. H. Johnston, British Mammals, p. 274. 2 Azara, Natural History of the Quadrupeds of Paraguay and the River La Plata (translated by W. P. Hunter), pp. 4-5. Ill] AND HISTORIC TIMES 263 in their hasty flight and from want of vessels, to leave their mares behind them. The baguales form immense herds, sometimes numbering 12,000 according to Azara. They frequently entice away domestic horses, who remain ever after with their wild comrades. Travellers often used to find themselves unable to continue their journey, their relays of fresh horses, which were always driven loose before them, being enticed away by the wild horses. The baguales' mode of attack is not in line of battle, but some precede the others, forming a vanguard, and the rest follow in a column, which is never broken or interrupted, and at most only changes its direction if they are frightened \ Azara estimates the proportion of bays amongst these horses • to be about ninety to ten zains, that is, entirely dark-coloured without any white ; and there is not one black in two thousand ; pied and greys occur sometimes, but are then invariably individuals escaped from domestic conditions^ As the grey and rufous-grey horses of the Asturias stock are very common in the hills and northern states of South America, the grey horses which occasionally occur on the Pampas are probably of this breed. Now as the Pampas horses have been living under natural conditions for the last three centuries, and do not show any tendency to grey, black, or pied, whilst bay forms their universal colour, it seems clear that the latter is the inherent colour of the Libyan horse, and that it does not tend to black, grey, or pied, except when crossed with the European-Asiatic stock. The history of the domestic horse in North America is in many respects closely parallel to its story in the southern part of the same continent. Although North America, as we have seen (pp. 6, 7), played a leading part in the evolution of the Equidae, and although in the early Pleistocene period owing to the favourable conditions of environment there were great numbers of and several kinds of horses, such as E. cumplicatus 1 Azara, op, cit. pp. 5-6. ^ Hamilton Smith, op. ctt. p. 175. 264 THE HORSES OF PREHISTORIC [CH. (about the size of a small Western broncho originally found near Natchez and now traced all over the Southern States from the isles of the Gulf of Mexico to South Carolina), E. pectinatus (a large horse with very elaborate grinders found in the North-Eastern and Middle States), E. pacificus (found on the coast of California and Oregon, perhaps the closest of all fossil horses to E. caballus), and E. excelsus (Nebraska), there is as yet no evidence that the indigenous horses had survived down to the first arrival of man in that continent. But it would be rash to dogmatise on this point since in a part of a quarry in Nebraska where Prof. Osborn and his associates on the Whitney expedition obtained the remains of hundreds of E. excelsus, "all the large limb bones were found broken in two." 'This,' says Prof. Osborn ^ "suggested to me the possi- bility that these large bones, the only ones known to have contained marrow, had been broken by man, who was primi- tively a great marrow-eater, but we searched in vain for any collateral evidence of this hypothesis. To my knowledge, no human remains have been found associated with those of the fossil horse in North America ; but I confidently expect that such association will be discovered, as it has been in South America." It is therefore absolutely certain that there were no horses in North America at the time of its discovery by Columbus. Indeed it seems likely that the Indians would have domesticated the indigenous horses, if any such still survived, since they were so quick to tame and utilise at a later date the feral horses of the western prairies. It is with the latter horses and their origin that we are now concerned. It seems certain that the thousands of wild horses that down to sixty years ago roamed the western prairies, were all descended from the horses introduced by the Spaniards, and that accordingly their history is very much the same as that of the haguals of the Pampas. But, whilst we know definitely that the Pampas horses are all sprung from a dozen Andalusian horses, we have no such explicit statements respecting the ancestry of the wild ^ Tlie Century Magazine, November, 1904, p. 12. Ill] AND HISTORIC TIMES 265 horses of North America. Again, whilst the Pampas horses show a remarkable uniformity of colour, it was very different with their kindred in the North, for they appear to have worn coats of many colours in the vast area over which they ranged from Mexico to the Red River (since, according to Dr Richardson \ they never seem to have advanced beyond the 52nd or 53rd degree of latitude). Thus, according to Catlin^, these wild horses were of all colours — black, grey, roan, and roan pied with sorrel — and F. Micheaux^ describes two wild horses from Mexico as roan, whilst Darwin* on the authority of Dr Canfield says that in certain parts they are mostly duns and striped, and Osborn states that " in Mexico and various other parts of America, the descendants of the horses in- troduced by the Spaniards are frequently of a dun colour with distinct dorsal, shoulder, and leg stripes^" Since Darwin adduced these dun-coloured horses of Mexico as cases of reversion to the dun colour of a primitive ancestor, and as the same horses are still being constantly cited in support of the same doctrine, it is most important that we should ascertain as far as possible the history of the horses of Spanish descent in Mexico and other parts of North America. This is all the more desirable in view of the facts which we have already elicited concerning the striped horses of Java and the dun- coloured and striped horses of Kattywar also cited by Darwin, and generally accepted without question as instances of the survival of or reversion to the primeval livery of the horse. It will be noticed that the contention of Darwin and others — that in the dun-coloured horses of Mexico and Texas we have instances of reversion — assumes that no horses of a dun-colour and having stripes were introduced into North America by the Spaniards. In fact Darwin tacitly assumed that as the Pampas horses are all sprung from Andalusian horses which are normally of a dark colour, the Spaniards brought none others than dark ^ Fauna Boreali-Americana (1829), p. 231. 2 Indian Tribes, Vol. ii. p. 57 (cited by Darwin). 3 Travels in North America (Eng. trans.), p. 235 (cited by Darwin). * Variation, Vol. i. p. 64. 5 The Century Magazine, November, 1904, p. 14. 266 THE HORSES OF PREHISTORIC [CH. Audalusian horses to Mexico and Texas. But it would be just as absurd to assume that because amongst the hills and northern states of South America grey and rufous-grey of the Asturian breed are very common, we have here instances of reversion. We have just seen that in addition to the fine dark horses of Andalusia (which have shown no tendency to revert to dun on the Pampas) Spain also possesses horses of all sorts of colours — dun, white, dun with stripes, various shades of grey, rufous-grey or roan, brown and black — and we were enabled to conclude that all these colours except dun and white and possibly striped dun were due to the intermix- ture of the Libyan and European-Asiatic horses. We also saw that when pied and grey horses were met on the Pampas, they were not reversions or ' sports,' but were invariably animals which had escaped from domestic conditions and were almost certainly of the Asturian and Murcian stocks. Bearing therefore in mind that the horses introduced by the Spaniards into the northern parts of South America were of light and mixed colours, we must be prepared to find that horses of a similar Asturian or Murcian origin as well as Andalusians were brought by the first Spanish settlers to San Domingo, and later to Cuba (settled in loll), by Cortes to Mexico in 1519, and still later to Florida and to the western bank of the Mississippi. We need not then be surprised, if the descendants of these Spanish horses wear liveries of all colours — black, grey, roan, and roan pied with dun (sorrel), dun, and striped dun — just like the horses of the northern states of South America descended from the breeds of northern Spain. We must not then hastily assume that the dun horses often with stripes found in Mexico and the Western States are instances of a reversion to a primal colour, when it is far more probable that they have simply retained the livery which dun- coloured striped ancestors brought with them from the sierras of Spain. If it can be shown that the Spanish Conquistadores carried into Mexico and Texas not only Andalusian horses, but also those of inferior breeds, we shall have an easy explanation of the many colours (including striped dun) found to-day in their posterity. Fortunately we have the clearest evidence Ill] AND HISTORIC TIMES 267 respecting not only the first introduction of horses into Mexico, but even minute descriptions of the animals themselves, whilst there is also reliable evidence for the first carrying of horses to the western bank of the Mississippi, though unfortunately we have no details respecting the colour or colours of these animals. This however is of no great matter since it can be shown that they were the same kind of animals as those brought to Mexico. After Columbus had discovered the New World, the Spaniards first settled in Hispaniola (San Domingo), and hither horses were introduced very soon, for beyond doubt they were already in that island when Diego Velasquez crossed over to Cuba in 1511 and conquered that island with little opposition. As the islands formed the base for all subsequent expeditions to the northern parts of South America, to Darien, to Yucatan, to Mexico, and to Florida and the lands beyond the Mississippi, we need not be surprised to find horses of the same kinds in modern times in the northern parts of South America, in Mexico, and the Western States. The difficulty of transporting horses from Spain to Hispaniola and Cuba in the small ships of the time naturally rendered these animals very scarce and very dear for some time in the islands, as is put beyond all doubt by the fact that when Hernando Cortes set forth (1519) from Cuba to conquer the empire of Montezuma he was only able to take sixteen horses with him. From the statements contained in the depositions at Villa Segura, it appears that the cost of the horses for the expedition was from four to five hundred pesos de oro each^ Bernal Diaz, Cortes's comrade, who wrote the immortal account of the conquest of Mexico^, was a lover of horses and he has given us in his work a minute account of each of the sixteen horses brought to Mexico. There were eleven horses and five mares and these were of many different colours. There were only two jennets, that is, line-bred horses, the rest being cross-bred animals. Cortes himself had a chestnut ' without any white ' (zaino), there were two termed simply 'chestnut' (castano), three 1 W. H. Prescott, The Conquest of Mexico, Vol. i. pp. 260-1, note. 2 Bernal Diaz, La Conquista de Nueva-Espaiia, cap. xxiii. 268 THE HORSES OF PREHISTORIC [CH. ' dark chestnuts ' (castano escuro), two bright ' chestnuts ' (cas- tano claro), and one 'perfect chestnut' (perfecto castano); there were three 'greys' (rucia), all being mares, one of which is described as very heavy (muy poderosa) ; and there was one ' all dark sorrel/ mane and tail included (alazan tostado), and there was an 'egg-coloured' (overo), t.e. light yellow-dun, and a ' light yellow-dun ' somewhat blackish-red above (overo algo sobre morcillo), and one ' dark '-coloured (escuro). As this passage has such an important bearing on the history of the horses not only of Mexico but also of the northern parts of South America, and the Western States of North America, and as it has hitherto escaped the notice of those who have written about American horses, I subjoin the original in a footnoted 1 El capitan Cortes, un caballo castano zaino, que luego se le murio en San Juan de Uliia. Pedro de Albarado y Hernando Lopez de Avila, una yegua castana muy buena, de juego y de carrera; y de que llegamos a la Nueva-Espana el Pedro de Albarado le compro la mitad de la yegua, 6 se la tomo por fuerza. Alonzo Hernandez Puertocarrero, una yegua rucia de buena carrera, que le compro Cortes por las lazadas de oro. Juan Velazquez de Leon, otra vegua rucia muy poderosa, que Uamabamos la Eabona, muy revuelta y de buena carrera. Cristobal de Oil, un caballo castano escuro, harto bueno. Francisco de Montejo y Alonso de Avila, un caballo alazan tostado : no fue para cosa de guerra. Francisco de Morla, un caballo castano escuro, gran corredor y revuelto. Juan de Escalante, un caballo castano claro, tresalvo : no fue bueno. Diego de Ordas, una yegua rucia, macborra, pasadera aunque corria poco. Gonzalo Dominguez, un muy extremado jiuete, un caballo castaiao escuro muy bueno y grande corredor. Pedro Gonzalez de Trujillo, un buen caballa castaiio, perfecto castano, que corria muy bien. Moron, vecino del Vaimo, un caballo overo, labrado de las manos, y era bien revuelto. Vaeno vicino de la Trinidad, un caballo overo algo sobre morcillo: no sali6 bueno. Lares, el muy buen jinete, un caballo muy bueno, de color castano algo claro y buen corredor. Ortiz el miisico, y un Bartolom6 Garcia, que solia tener minas de oro, un muy buen caballo escuro que decian el Arriero : este fue uno de los buenos caballos que pasamos en la armada. Juan Sedefio, vecino de la Habana, una yegua castaiia, yesta yegua pario en el navio. Este Juan Sedeno paso el mas rico soldado, que bubo en toda la Ill] AND HISTORIC TIMES 269 Cortes set sail from Cuba for Yucatan in February, 1519, with 663 men, 200 Indians, and sixteen horses. In his first battle two horses were killed and in the second another, and all the survivors were more or less severely wounded. Cortes later on was joined by Alvarado at Vera Cruz with twenty horses and one hundred and fifty men. Cortes had mortally offended the governor of Cuba by reporting direct to Spain, and the latter sent out a force under Narvaez, who was to supersede Cortes and send him back in chains to Cuba. Narvaez had eighteen vessels, which carried nine hundred men, of whom eighty were cavalry. Cortes by this time had only five mounted men, but by a successful night attack he captured Narvaez and his whole army. The common soldiers were only too ready to transfer their allegiance to so vigorous a captain as Cortes, and the latter had now eighty-five horsemen. The conquest of Mexico was accomplished in 1521, and adventurers from Spain soon poured in, bringing other horses from the Antilles from time to time. We need not hesitate to believe that the horses brought by Alvarado and Narvaez were of the same kinds and colours as those of Cortes, the colours of which we have just enumerated. As Cuba had been settled from Hispaniola and by 1538 had already great numbers of horses, we may safely assume that the horses brought to Darien, if not by Balboa (1510) most certainly by Pedrarias, who was established there before Cortes sailed for Mexico, and the horses brought into Peru by Pizarro in 1526 were of the same kinds and colours as those described by Bernal Diaz, and in some of which we have recognized the descendants of the ancient grey horses of the northern parts of Spain and of the dun-coloured horses of the same country. We need not then be surprised to find horses of many colours including (striped dun) in Mexico and in the Western States, and it is now certain that the dun horses with black stripes which Darwin and so many others have supposed armada, porque trujo uu navio suyo, y la yegua y un negro, e cazabe 6 tocinos; porque en aquella sazon no se podia hallar caballos ni negros sine era a peso de oro, y a esta causa no pasaron mas caballos, porque na los babia. For fuller explanation of the horses' colours see Addenda. 270 THE HORSES OF PREHISTORIC [CH. to be instances of reversion to the aucestral type are nothing of the kind, but only the continuance of characteristics brought by the ancestors of such horses from Spain to Hispaniola, and Cuba, and thence into Mexico. It is absolutely certain that the horses of the northern parts of South America, to which we have already referred (p. 266), are descended from animals first brought to those regions by the Spanish colonists. That such was the case not only with the Isthmus of Darien, and the adjacent regions, but also with Peru, is beyond doubt. Francisco Pizarro, the conqueror of the kingdom of the Incas, was already in Hispaniola in 1510, for in that year he took part in an unfortunate attempt to plant a colony at Uraba on Terra Firma. In the following year he was at Darien with the famous Vasco Nunez de Balboa, and accom- panied that bold cavalier when he scaled the mountain ridge of the Isthmus and was the first European to gaze on the limitless expanse of the Pacific. After the gallant Balboa had been put to death by his rival Pedrarias, Pizarro attached himself to the latter, who with a view to pushing forward the discovery of new lands for plunder on the w^estern coast, transferred the capital of the colony from Darien to Panama in 1519, at the very time that Cortes had set forth to subdue the Aztec empire. Pizarro started on his first expedition southwards from Panama in 1524 with about one hundred men in two small vessels, but he does not appear to have had any horses. However, in his second attempt to reach the realms of gold, in addition to one hundred and sixty men he had been able to procure a few horses at Panama ^ There can be no reasonable doubt that these horses, like those brought by Cortes to Mexico, had come from Hispaniola or Cuba, and were of the same kind. The important aid rendered to Pizarro by these horses in his wars with the Indians is probably familiar to most readers, for the natives at first supposed that the Spanish cavaliers were a kind of centaurs, and Pizarro was able to extricate his forces from one dangerous position by the consternation that seized the Indians when a cavalier fell from his horse, as they were ^ W. H. Prescott, Conquest of Peru, Vol. i. p. 240. Ill] AND HISTORIC TIMES 271 not prepared for the division of what seemed a single being into two^ We thus obtain an easy explanation of the reason why the horses of the northern parts of South America resemble so closely in their colours those of Mexico. There is another source other than the horses brought into Mexico from which may have sprung wholly or in part the wild horses which formerly roamed Texas and the other regions west of the Mississippi. Ferdinando de Soto had gone to the Spanish Indies when Pedrarias was governor " and there he was without anything of his own save his sword and target." Pedrarias made him captain of a troop of horsemen (which clearly shows that there were already horses at Darien), and by his commandment he went with Pizarro to the conquest of Peru. He was at the capture of the Inca Atahualpa and at the taking of Cuzco and got a goodly share of the booty. He then married the daughter of Pedrarias, and the emperor Charles V made him governor of the island of Cuba and Adelantado or President of Florida ^ In April, 1538, he sailed from San Lucar in Spain to Cuba and Florida with an armament of 600 men. He arrived at Santiago in Cuba on Whitsunday, and " as soon as they came thither a gentleman of the city sent to the seaside a very fair roan horse, and well furnished for the governor, and a mule for Donna Isabella, and all the horsemen and footmen that were in the town came to receive him at the seaside I" The chronicler then states that " in this country (Cuba) there are many good horses, and there is green grass all the year. There be many wild oxen and hogs whereby the people of the island are well furnished with fleshy" Having sent his ships to Havannah " the governor and those who stayed with him bought horses and proceeded on their journey. The governor's company 1 Prescott, op. cit. Vol. i. p. 252. - The Discovery and Conquest of Terra Florida by Don Ferdinando de Soto and six hundred Spaniards, his foUmvers, loritten by a Gentleman of Elvas, employed in all the action, and translated out of Portuguese by Richard Hakluyt (reprinted from the ed. of 1611 by W. B. Eye : Hakluyt Society), p. 4. 3 Ibid. p. 17. * Ibid. p. 19. 272 THE HORSES OF PREHISTORIC [CH. which went by land were one hundred and fifty horsemen^" De Soto and his company reached Havannah in March, 1539, and on May 30th they landed in Florida. " They set on land two hundred and thirteen horses, which they brought with them, to unburden the ships that they might draw less water^." The horses were weak with travelling upon the sea. It was not the first time that horses had been in Florida, for in 1525 Allyon had sailed thither with six ships, carrying five hundred men and between eighty and ninety horses, but the expedition met with nothing but disaster and the remnant got back to San Domingo. In 1527 preparations were made in Spain for a fresh expedition to Florida. Pamphilo de Narvaez set sail from San Lucar in 1527, and after wintering at Cuba he landed in 1528 at Santa Cruz in Florida. He and his comrades were reduced to great straits and had to kill a horse for food every third day, while they were constructing at Baya de Caballos frail vessels in which they hoped to reach the Spanish settle- ments, but most of them were lost at sea. De Soto's men discovered the site of Narvaez' camp at Alpaca and found there skulls of horses^ The horses brought by Narvaez cannot then have contributed any element towards the equine population of America. After three years' wanderings, in which he committed the most wanton cruelties on the Indians, de Soto reached the Mississippi, which he crossed about lat. 35° N. in 1541. " He desired to send news of himself to Cuba that some supply of men and horses might be sent unto him." By this time he had lost two hundred and fifty men and one hundred and fifty horses^ He wintered in Autiamque and in the next year (1542) he departed from it to see Nilco, which the Indians said was near the great river, "with determination to come to the sea and procure some succour of men and horses, for he had now but three hundred men-of-war and forty horses, some of them lame, which did nothing but help to make up the number : and for want of iron they had gone above a year unshod, and because they were used to it in the plain country, it did them no great harm^" 1 Hakluyt, op. cit. p. 21. 2 j^,-,; p_ 25. » Ibid. p. 43. 4 Ibid. pp. 111-12. 5 Ibid. p. 115. Ill] AND HISTORIC TIMES 273 De Soto died by the Mississippi, and to conceal his death from the Indians, whom he had persuaded that he was immortal, his followers by night cast his body into the Father of Waters. His goods were sold by auction : they consisted of two men slaves, two women slaves, and three horses and seven hundred hogs, the posterity of the sows which he had brought from Cuba to Florida. " For every slave or horse they gave two or three thousand ducats : two hundred ducats for a hog^" Luys de Moscoso now took the command and determined to return to Minoya and to make ships there. " They shipped twenty- two of the best horses that were in the camp : the rest they made dried flesh of-." They made their way down the river with great difficulty, fighting against the river-side natives. Finally, as the canoes in which they were conveying the horses made such slow progress, Luys de Moscoso determined to go on shore and kill them. " As soon as they saw a place con- venient for it, they went thither and killed the horses and brought the flesh of them to dry it aboard. Four or five of them remained on shore alive : the Indians went unto them after the Spaniards were embarked — the horses were not acquainted with them and began to neigh and run up and down in such sort that the Indians, for fear of them, leaped into the water, and getting into their canoes went after the brigandines, shooting cruelly at them^" According to another account of the expedition, written by Hernandez de Biedma^ when de Soto landed in Florida he had six hundred and twenty men and two hundred and twenty- three horses^ He says that when they came to the great river " we resolved to make four large pirogues, each capable of containing sixty or seventy men and five or six horses, and we spent twenty-seven or twenty-eight days in constructing them." The river was crossed where it was about a league broad and the depth from nineteen to twenty fathoms. In their subsequent wanderings they reached a place where the 1 Hakluyt, op. cit. pp. 126-7. - Ibid. p. 148. ^ Ibid. pp. 153-4. •* This account is given by Eye in an appendix to his reprint of Hakliiyt's translation of The Discovery, pp. 173 sqq. 5 Ibid. p. 178. R. H. 18 274 THE HORSES OF PREHISTORIC [CH. Indians made abundance of salt (salt springs ?), and they spent six months in the construction of the brigandines : " we took with us some canoes, into which we put twenty-six horses \" There is no reason why the half-dozen horses thus abandoned on the bank of the Mississippi should not have become the pro- srenitors of all the wild horses of Western America. It is clear that the Indians did not kill them at once, and the horses would thus have had time to escape into the open country, where it would be difficult for the natives to capture or slay them. There was no greater likelihood of their being killed by the Indians than there was in the case of the oft-cited animals abandoned at Buenos Ayres and San Juan Bautista, when these two infant settlements were attacked by Indians. Only one doubt remains. On which bank of the Mississippi were the horses deserted ? The chronicler gives no direct statement, but as the brigandines were built on the western bank, and as the writer does not state that they had crossed to the other side when Luys de Moscoso determined to kill the remnant of the horses, it is probable that the horses were abandoned on the western bank of the great river. As all de Soto's horses like those of Cortes were brought from Cuba, the colours of these animals would have been the same as those of the horses which formed the parent stock of Mexico. Accord- ingly, even if the horses of the Western States are sprung from de Soto's animals, and not from feral horses from Mexico, the occurrence of dun with stripes in such animals is not to be regarded as a reversion to a primitive ancestor. The horses of Sardinia and Corsica, as might naturally have been expected, show much North African blood. The Sardinian seldom exceed 13"2 hands (185 m.), and are in colour black, chestnut, and bay, but rarely grey^ whilst the Corsican range from 11"2 hands to 13*2 and are black, chestnut, sometimes bay, but rarely grey^ They are steady, active and courageous, and capable of enduring cold and hunger. The horses of Sardinia and Corsica had well-known char- acteristics in medieval times, for Stradanus gives a picture of 1 Hakluyt, oj). cit. p. 199. ^ Cuyer and Alix, Le Cheval, p. 632. 3 Ibid. p. 613. Ill] AND HISTORIC TIMES 275 the ' Sardonicus ' (here reproduced, Fig. 76) ; whilst Blundeville says that " the horses that come out of the Isles of Sardinia and Corsica, as Volaterranus saith, have short bodies and be very bolde and couragious, and unquiet in their pace, for they be of so fierce and hote cholerike complexion, and there- with so much used to running in their countrie, as they will stand still on no ground, and therefore this kind of Horse requireth a discreete and patiente I'ider, who must not be overhastie in correcting him, for feare of marring him altogether." Fig. 7C. The Sardinian Horse. The Carthaginians had occupied Sardinia from an early time and it is probable that Libyan horses were introduced into that island by them. Whether this be so or not it seems probable that Libyan blood had got into both islands from a comparatively early period, and that partly from it originated the modern breeds. The prevalence of black, chestnut, and bay, and the rarity of grey shows that although the Libyan blood had been crossed with European, the former was really the chief element in the blend. 18—2 276 THE HORSES OF PREHISTORIC [CH. Corsica was noted in the sixth century A.D. for a very diminutive breed of ponies, for Procopius^ states that in that island "just as human beings become dwarfs, so are there herds of horses which are but little taller than sheep." The horses of Sicily to-day resemble closely those of Sardinia both in colour, conformation, and character, being black, chestnut, and bay, but rarely grey. As the Carthaginians from an early period had factories in Sicily, and as there was a vast trade between the Cyrenaica and the Greek cities of that island, it seems highly probable that the famous horses of Sicily, which so often carried off the prizes at the great games of Greece, were largely of the Libyan stock. It is very noteworthy that of the fourteen victories with horses and mules celebrated by Pindar, twelve were carried off by animals from Sicily and Cyrene, two only falling to Greeks of Greece proper — Megacles of Athens and Herodotus of Thebes. Syracuse, Agrigentum, and Camarina were specially distinguished for their horses, and Pindar praises Psaumis of Camarina " as a man most zealous in the breeding of horses," and it would appear that he had actually entered at the same time at Olympia a chariot, a mule-car, and a race-horse. Agrigentum continued to be devoted to horses until its destruction by the Carthaginians in 405 B.C., for when a citizen named Exaenetus won the chariot-race at Olympia in 412 B.C., he was met on his return and escorted into the city by three hundred chariots each drawn b}^ two white horses. The city was likewise famous for the splendour of the monuments reared over successful race-horsesl The coinage of the Sicilian cities amply proves their pride in their horses. Victories in the great contests are frequently commemorated in their types. Thus some Agrigentine coins have a quadriga driven by winged Nike, suggested by such a victory as that of Exaenetus, whilst in the earliest issue of Syracuse (before 500 B.C.) the tetradrachms bear a four-horse chariot, the didrachms a man riding one and leading a second 1 De Bell. Gothico, ii. 4. 24: evravda uia-irep dvdpojiroL vdvvoi -ylvovrai, ovtoo drj tIvup Xirwwv ayiXai elcrl twv TrpojSdTUiv dXlyij} fxei'^bvwv. - Diod. Sic. XIII. 82. 6. Ill] AND HISTOEIC TIMES 277 horse, and the drachms a horseman. Gela not only prided herself on her racing successes, but on what was far more important, her fine cavalry, for her coins show a four-horse chariot with Nike floating above, and very constantly an armed horseman, spearing a prostrate foe, or else striking downwards Fig. 77. Archaic Metope, showing a Quadriga; Selinus. with his spear. These coins give us some of the earliest numismatic representations of riding. But by far the earliest representation of Sicilian horses now extant is the quadriga on one of the metopes of the archaic temple of Selinus (founded B.C. 628), which date from the latter half of the seventh cen- tury B.C. The metopes, which are in high relief and extremely vigorous in execution, have a special interest as they are the 278 THE HORSES OF PREHISTORIC [CH. most ancient extant specimens of Greek sculpture of the historical period. The quadriga relief which represents a chariot and horses in elevation (Fig. 77) is the most remark- able. The Sicilian horses, to judge by the monuments just cited, as well as the Carthaginian issues of Panormus (Fig. 74 b, p. 255), were cross-bred animals, for doubtless the desire to obtain strength made the Sicilians prefer half-bred to pure Libyan horses. It seems almost certain that the white horses which the Sicilian Greeks, like so many other peoples, esteemed for their colour far beyond their true merits, were derived from central Europe, for we have already seen that Dionysius, the despot of Syracuse, purchased some of their best strain from the Veneti, who certainly possessed white horses. In the fourth century A.D, the Sicilian horses were nearly as much esteemed for the circus as the Cappadocian and Spanish ^ South Italy shared with Sicily the fame of her riding horses, and from at least the sixth century B.C. the great cities of Magna Graecia, such as Sybaris, Croton and Tarentum, were renowned for their cavalry, which formed their chief arm in war. Before the fall of Sybaris (510 B.C.) five thousand of her citizens used to ride on horseback in procession on high festivals. They taught their horses to dance to music at their banquets, an accomplishment which brought about the de- struction of Sybaris, as the Crotonians before the great battle in which the Sybarites were destroyed caused flute-players to play one of the tunes to which the Sybarite horses were accustomed to dance. As soon as they heard the tune they stood on their hind legs, unseating their masters and rendering them an easy prey to their foes^ We have noticed as we advanced that extraordinary docility characterised the Libyan horse and its derivatives, such as the Arab, the Turk of Western Asia, and the little ponies of Java, and we are told by Ibn Batuta^ that he saw horses (which were certainly Arab) dancing before the Arab sultan of Sumatra, and that he had already seen ^ Veg., Ars Vet. iv. 6, 2 : Cappadocum gloriosa nobilitas, Hispanorum par vel proxima in circo creditur palma, nee inferiores prope Sicilia exhibet circo. 2 Athenaeus, xii. 520; Pliny, H. N. viii. 157. ■^ Voyages, Vol. vi. pp. 236-7. Ill] AND HISTORIC TIMES 279 a similar performance taking place before a king of Southern India, where, as we have seen, all the horses were imported from Arabia ; again, the Iberian horses, i.e. the horses of Southern Spain, derived directly from Libya, were noted for the same docility, and their descendants, the Pampas horses of South America, retain that quality, whilst the dancing and Fig. 78. Fragment of Sculpture from Tarentum (4th cent. b.c). performing horses in modern hippodromes seem always to be Arabs. The extreme readiness of the Sybarite horses to learn dancing itself points to their having in their veins a considerable infusion of Libyan blood. Thurii, which was founded on the ruins of Sybaris in 443 B.C., revived the horse-breeding tradition of the older town, and it is most important to note that according to Tacitus^ ^ Ann. XIV. 21. 280 THE HORSES OF PREHISTORIC [CH. the Romans first learned horse-racing {certamina equoruni) from that city. The Tarentines especially prided themselves on their horsemen, which formed their chief arm in war and one of which was the type on their coins from the end of the fifth century B.C. down to the Roman conquest in B.C. 272. That the best horses of Tarentum were well-bred with a large admixture of Libyan blood is put beyond doubt by a marble fragment procured near Tarentum^ and now in the British Museum (Fig. 78). The length of the horse's head from the end of the mane over the forehead to the lip is 0'46 m., the height from the bottom of the cheek-bone to the top of the head is 0"34 m. ; the lower lip is gone, and the ears are broken. The bridle, as usual in such cases, was probably added in bronze, as is shown by rivet-holes. From a comparison with Tarentine coins and other considerations the fragment may be placed in the latter half of the fourth century B.C. The head differs from that of the horse of Selene from the Parthenon (p. 299) in the greater length of its fore-part, from which Prof Michaelis was disposed to infer that the Tarentine horse was of more slender proportions, and furthermore it shows more faithful representation to nature than is seen " in that wonderful com- bination of idealism and realism which is so conspicuous in the head of Selene's horse." The bony ridge below the eye, to which are attached the masseter and zygomatic muscles, is more strongly marked in the Tarentine than in the Parthenon horse ; the nose is slightly curved, the eye is large though not so prominent as in the latter, whilst on the other hand the eyes of the famous bronze horses from St Mark's at Venice lie deep in their sockets, and are overshadowed by rather strongly marked brows. The mane in the Tarentine fragment is cut short, but is not so stiff as is usual with Attic horses, and it falls more freely, hanging in a double forelock over the forehead, as is also the case in several slabs of the Parthenon. If we compare the Tarentine head with that of one of the colossal horses from the Mausoleum (p. 305) and with that of a horse from the Amazon frieze of the same monument, we at once see that the ideal Tarentine horse in the second half of the fourth century B.C. 1 A. D. Michaelis, Jour. Hell. Stud., Vol. iii. (1882), pp. 234-9, PI. xxiv. Ill] AND HISTORIC TIMES 281 was a far better bred animal than the ideal Asiatic horse of the same period. The Tarentine fragment is held to have been part of a chariot group, but there is no reason why it may not have been part of a monument erected over a favourite horse, as Fig. 79. Fresco showing a Samnite warrior: Paestum. was the fashion at Agrigentum (p. 276), or it may have formed part of a dedication, such as a group of horses and captive women — the work of Agelaidas of Argos — dedicated at a much earlier period by the Tarentines to celebrate their victory over their inveterate foes the Messapians. 282 THE HORSES OF PREHISTORIC [CH. We are fortunately not without good evidence for the colour of the best horses of Southern Italy in the fifth cen- tury B.C., and by implication also of those of Sicily. A fresco from Paestum now preserved in the Naples Museum ^ shows a Samnite warrior on horseback (Fig. 79). The horse is bay with white stockings. As the painter would take for his model the most typical war-horse of his own time and country, it is clear that the best horses of Southern Italy at that date were good cross-bred horses full of Libyan blood, as we had already inferred from the marble horse's head from Tarentum. We shall soon find that the typical South Italian horses of to-day are also bay in colour. It is most interesting to find that Virgil represents as mounted on such a horse as that ridden by the Samnite warrior in the fresco not merely the Roman youth who took part in the Trojan game, but also Turnus, king of the Rutuli, a fierce warlike tribe of the same Umbro-Sabellian stock from which the Samnites were sprung^ In Virgil's day this colour was closely associated with the best horses of Thrace, where, as we shall soon find (p. 304), much fine blood had been introduced from the south by the fourth century B.C. The Roman youth described by Virgil " rode a Thracian steed of two colours, as he had white markings — a white forefoot and a white mark on his forehead^" whilst Turnus' charger is also described as a Thracian steed with similar white markings'*. But, as we have seen that bay with white feet and a star in the forehead is a regular concomitant of the North African horse 1 E. Neville-Rolfe, Handbook to the Naples Museum, p. 19. I am indebted to my friend Mr G. P. Bidder, M.A., Trin. Coll., Camb., for calling my attention to this fresco. - Ridgeway, Early Age of Greece, Vol. i. p. 257. ' Virg. Aen. v. 565-7: quem Thracius albis Portat equus bicolor maculis ; vestigia primi Alba pedis, frontemque ostentans arduus albam. ^ Virg. Aen. ix. 48 : Improvisus adest, maculis quem Thracius albis Portat equus. In Georg. in. 82 he puts bay first, grey second, and white and yellow-dun last : Honesti spadices glaucique ; color deterrimus albis et gilvo. Cf, Varro aj). Non. 2, 87 : equi colore dispares item nati, hie badius, iste gilvus, ille murinus. Ill] AND HISTORIC TIMES 283 and his derivatives, it is not surprising that Tarentum, Croton, Sybaris, and later Thurii were famous for their cavalry. It is very significant that the best Italian horses in Roman times were those bred in Apulia, where, as Varro^ tells us, herds of brood mares were pastured. When speaking of noble breeds of horses he only mentioned the names of two Italian kinds- — the Apulian and Roseau. The Apulian thus stands first of those of Italy, the Roseau horses bred in the fertile district near Reate being placed second, whilst he does not even allude to any breed in Northern Italy. The same practically holds true at the present hour, for Calabria still breeds excellent bay horses, which are of a finer build than those of Central Italy, as for example the horses of the Roman Campagna-*. The superiority of the horses of Southern Italy continued right down through the Middle Ages, the stock being reinvigorated from time to time by fresh importations from North Africa, especially during the period of Saracen domination, and also from Spain at the time when Naples was closely connected with that country. The fame of the horses of Naples was spread all over Europe, and not only does Stradanus include the Neapolitanus in the 'stable ' (Fig. 80), but Blundeville in 1580 exhorts horse- breeders, if possible, to procure a Neapolitan stallion, and elsewhere he thus praises this horse: "The Napolitan, which we commonlie call a courser of Naples, is a trim Horse, being both comelie and stronglie made, and of so much goodnesse, of so gentle a nature, and of so high a coureage as anie Horse is, of what countrie soever he be. He is easilie knowne from all other Horses, by his no lesse cleane, and strong making, his limmes are so well proportioned in everie point, and partlie by his portlinesse in his gate, but chieflie by his long slender head, the nether part whereof, that is to say, from the eies downward, 1 Re Rust. II. 7 : Horum equorum, et equarum greges qui habere voluerint, ut habent aliqni in Peloponneso et in Apulia, primum spectare oportet aetatem, quam praecipiunt videndum ne sint minores trimae, maiores decern annorum. 2 Id. II. 7, 6: de stirpe magni interest qua sint, quod genera sunt multa. Itaque ad hoc nobiles a regionibus dicuntur, in Graecia Thessalici, a terra (Apulia) Apuli, ab Rosea Eoseani. 3 This fact I owe (through my friend Mr G. P. Bidder, M.A., Trin. Coll., Camb.) to Mr E. Neville-Rolfe, H. B. M. Consul for Southern Italy. 284 THE HORSES OF PREHISTORIC [CH. for the most part is also somewhat bending like a Hawke's beak, which maketh him to rein with the better grace, and yet the Italians do both write and saie, that these coursers be nothing so strong now, as they have been in times past, partlie perhaps for that like industi'ie of late daies hath not beene used in breeding them, as in times past, and partlie for that nature doth decaie everie day more and more, as well in man as in beast. But howsoever they be, in mine opinion, their gentle nature and docilitie, their comelie shape, their strength, their courage, their sure footmauship, their well reining, their loftie pace, their cleane toothing, their strong galloppiug, and their swift running well considered (as which things they have in maner by nature) they excell numbers of other races, even so farre as the faire greihounds the fowle Mastiffe curres." In modern times the harses of Naples are partly imported, partly native-bred ^ The little carrozzella horses are not Italian, as the best come from Sardinia {supra, p. 274), and the others from Tunis; the latter being known as 'Turkish horses ' {cavalli Turchi). The carrozzella ponies seldom exceed 14*2 hands ; they cost from £6 to £16 at three years old when they are taken into work. They are regularly driven with a nose-band, and not with a bit, though some of the nose-bands furnished with teeth {morgi dentati) are as severe as any bit^ The native horses of Naples (cavalli nostrali) are bred in Salerno and Calabria and run to fifteen hands high, the average being about 14'2 to 14'3. The outside price is about £40. The prin- cipal horse-fair is at Foggia. The Calabrian horse is usually a dull bay with a black stripe down the back, black legs, and a tan muzzle. As the records of Roman history belong to a period con- siderably later than that from which we have very full and comparatively complete documents for the history of Sicily, the ' For this information respecting the NeapoHtan horses of to-day I am indebted to Mr E. Neville-Rolfe, B.A., H. B. M. Consul for South Italy, through my friend Mr G. P. Bidder, M.A., Trin. Coll., Camb. 2 For this information I am indebted to Mr G. P. Bidder, who has taken much trouble to get me information about the structure and use of the nose- band. He tells me that the riding-horses are controlled by the bit, and that carriage-horses are driven with bits and English harness. m] AND HISTORIC TIMES 285 Greek colonies of Southern Italy, and for Greece itself, we shall now resume the history of the horse in Greece proper, and then take up the story of that animal in Central and Upper Italy. We saw that the horses of the Acheans bred in Thessaly and Elis were dun coloured, and that they were not ridden, but driven in pairs under chariots. But the horse and the chariot had already been known in the Bronze Age (Mycenean period), before ever the sons of the Acheans had come. This is proved by the grave-stones found on the acropolis of Mycenae over the Fig. The Neapolitan Courser. famous shaft-graves which contained the rich treasures buried with the royal Perseid house, which had reigned in the Bronze Age in Argolis. Several of their stelae (Fig. 47, p. 107) show in low relief a two-horse chariot in which is seated some personage, not unlikely the ancient chief whose mouldering bones were uncovered by Schliemann. Moreover the Homeric poems have at least one reference to the horses kept by the men of the pre-Achean time. The Iliad^ speaks thus of one steed of that bygone age: "Not even if he drove at thy back divine Arion, ' XXIII. 346. 286 THE HORSES OF PREHISTORIC [CH. swift steed of Adrastus that sprang from the gods." A line in the lost epic called Thehais told how "Adrastus fled from Thebes wearing sorry garments, and with him dark-maned Arion." According to Antimachus Adrastus "was the first of the Danaans that drove two high-praised steeds, fleet Caerus and Thelpusian Arion, whom, near the Oncean grove of Apollo, Earth herself brought forth a wonder for mortals to see^" According to another statement Poseidon himself was the father of Arion, which is not without significance, as we shall soon find the same god as the father of Pegasus by the Gorgon Medusa. According to these stories it was only in the days of Adrastus — about 1350 B.C., if we follow the traditional chrono- logy—that the horse and chariot first got into Peloponnesus. The fact that a divine origin is ascribed to Arion seems to indicate that the horse had hitherto been as unfamiliar to the people of Argolis as it was to the Aztecs when Cortes landed with his Spaniards and his horses in Mexico. Arion is described as ' dark-maned,' which indicates that his body colour was lighter than his mane and tail, a feature common both to the dun-coloured horses of Europe and Asia and to the bay horses of North Africa. It is worthy of notice that the wheels of the chariots on the Mycenean grave-stones (which may be assigned to the four- teenth century B.C.) have only four spokes like the chariot found in a tomb at Egyptian Thebes (p. 225) said to be about the same period, and in this respect they stand in contrast to the Homeric chariot with its eight-spoked wheels. Centuries before the planting of Cyrene the Greeks had a firm belief that by the side of the Atlantic were bred steeds of surpassing swiftness. Already we have read in Homer of swift horses begotten by the west wind and foaled by a harpy beside the stream of Ocean. But still more significant is one of the most familiar of Greek legends, the myth of Pegasus, the first of all horses that bore a rider on his back. The renowned winged steed, begotten by Poseidon himself on the Gorgon 1 Paus. VIII. 25. 7. Ill] AND HISTORIC TIMES 287 Medusa, was born in Libya, where he had sprung from the body of his mother, when that monster fell to the sword of Perseus^ Thus, then, this most famous of stories shows clearly that the early Greeks held that not only the best of all horses were bred in North Africa, but also that the Greeks knew not how to ride on horseback until they had borrowed that practice along with the Libyan horse, which was a little taller and infinitely superior in speed to the little dun horses of Europe. So great a feat was the mounting on horseback considered that Athena herself, according to the myth, had to bridle and subdue Pegasus for her favourite Bellerophon of Corinth^. In each case Poseidon — who was the chief deity of Libya, as well as of the indigenous population of Greece — is repre- sented as the sire of these, the most famous horses of early Greece. The evidence of the myths, as far as it goes, taken along with that of the chariot-wheels, points to North Africa as the region from which the Greeks of Peloponnesus first heard of the horse and the chariot, and later on learned the art of mounting the horse itself. This view is confirmed by certain pieces of evidence derived from recent discoveries in Crete and Cyprus. Even if the Libyan horse had never been seen on the mainland of Greece by Homeric times it is not at all improbable that it was known in Crete and Cyprus, which were not only in close communica- tion with Egypt for many centuries before the Achean conquest of Greece, but had also constant intercourse for trading purposes in Homeric days with the coast of Libya, as is proved by more than one passage in the Odyssey^. We have seen that the Mycenean chariot (Fig. 47), like the Libyan and the earliest Egyptian chariot (pp. 224-5), had only four spokes. It is therefore interesting to find that not only are the chariots in the pictographs found by Mr A. J. Evans in the palace of Cnossus furnished with four-spoked wheels, but that similar chariot-wheels appear on the vases of the late Mycenean period found at Enkomi (Figs. 81-2) and Curium in Cyprus^ We 1 Apollod. II. 3. 2. - Paus. ii. 4. 1. ^ IV. 85, XIV. 295, XVII. 442 sqq. * Excavations in Cyprus, by A. S. Murray, A. H. Smith, and H. B. Walters 288 THE HORSES OF PREHISTORIC [CH. further saw that the earliest representations of Pegasus are found on the pottery from Daphnae (p. 245), and that the winged steed carries his tail like the pure-bred 'Arab,' from which we were led to conclude that the swift Libyan horse had been taken as a model for the winged steed. A vase from Enkomi and another from Curium show horses which carry their tails in the true North African fashion. This clearly Fig. 81. Vase from Eukomi, Cyprus. demonstrates that about B.C. 1000 the Libyan horse was familiar to the Cypriote vase-painter, although a fragment of another vase (Fig. 82) indicates that he was also acquainted with horses of the ordinary European-Asiatic type, as is shown by the way (1900), p. 49, nos. 981, 1113 (Enkomi), p. 73, no. 136 (Curium). The chariot shown on the carved ivory combined draught-board and box (op. cit. p. 12) found at Enkomi, has wheels of six spokes Hke the common Egyptian form of chariot evolved from the earlier four-spoked wheel (supra, p. 221). Ill] AND HISTOEIC TIMES ' 289 in which the tail is set on and carried. As the myth of Pegasus does not appear in Homer, and as the Cyprus paintings come at the close of, or after, the Homeric period, the appearance of the Libyan horse on such vases is quite in accord with the literary data. Meanwhile the Acheans had come down from central Europe with the dun-coloured horses of that region yoked to chariots with wheels of eight spokes. In Homer only dun-coloured horses are known in Greece itself; white horses, though known in Thrace, apparently not yet having been imported into Greece, while the one bay horse mentioned, of which we shall speak at greater length, was bred in Asia. Fig, 82. Vase fragment from Enkomi, Cyprus. We have seen that by the tenth century B.C. Libyan horses were being imported from Egypt into anterior Asia for the kings of Syria and the Hittites, and we have found that bay or chestnut colour, frequently accompanied by a white mark on the forehead and white 'stockings,' is the characteristic colouring of the Libyan horse and his purest derivatives down to the present day, being thus clearly distinguished from the lighter colours — dun, rufous-grey, grey, and white — which form the liveries of horses either of pure Asiatic and European ancestry, or with but a small infusion of Libyan blood. Now the Iliad yields at least one piece of evidence that the Libyan horse had made its way not only into Syria and the land of the Hittites, but even as far as the north-western comer of Asia Minor before 1000 B.C., R. H. 19 290 THE HORSES OF PREHISTORIC [CH. though it was then extremely rare and therefore very much coveted by horse-fanciers. Like Pegasus and Arion, and, indeed, like most things strange and rare, brought from distant lands, this strain was held by the Homeric Greeks to be of divine origin. The Trojans are regularly spoken of as "tamers of horses" {iTTiroSa/jiOL) and we are told that King Erichthonius, "who became richest of mortal men," had "three thousand mares that pastured along the marsh meadow, rejoicing in their tender foals. Of them was Boreas enamoured as they grazed, and in semblance of a black -maned horse he covered them. Then they, having conceived, bare twelve fillies. These, when they bounded over Earth, the grain-giver, would run upon the topmost ripened ears of corn and break them not; and when they bounded over the broad backs of the sea, they would run upon the crests of the breakers of the hoary brine. Erichthonius begat Tros, the father of Ganymede, who was caught up by the gods to become the cup-bearer of Zeus^" There can be little doubt that these horses were of the European- Asiatic stock. Since horses from the Ocean and from Libya are described as the offspring of Zephyrus, the West Wind, and of Poseidon respectively, so the Boreas-sprung horses of the rich plains of the Troad, peopled by Dardanians and Mysians, who had crossed over from Thrace — the home^ of Boreas, the North Wind — were plainly regarded as of Thracian origin. Though Tros had inherited so excellent a breed of horses from his father Erichthonius, he was able to introduce a far better strain into his stud. In his combat with Aeneas Diomede especially wished to capture the horses of the Trojan hero, for, said he to his charioteer, "they are of that breed whereof far-seeing Zeus gave to Tros as recompense for Ganymede, his child, because they were the best of all horses beneath the daylight and the sun. That blood Anchises king of men stole of Laomedon, privily putting mares to them, whereof a stock was born to him in his palace, even six; four kept he himself and reared them at the stall, and the other twain gave he to Aeneas, deviser of ^ Iliad, XX. 219 sqq. - Iliad, xiv. 395, xxii. G92. Ill] AND HISTORIC TIMES 291 rout. Them could we seize we should win us great renown \" Diomede accomplished his desire, and the horses of Aeneas became his. One of them is described "as bay all the rest of him, but in the forehead marked with a white star round like the moon I" In the chariot-race at the funeral games held in honour of Patroclus Diomede drove these horses and they easily carried off the first prize ^ from the other Achean chiefs, whose horses had all been bred in Greece'* and, as we have seen, were all dun-coloured {xanthos). As this is the only bay horse mentioned in Homer, and as it stands in strong contrast to the dun-coloured horses of Achilles and the other Acheans, and to the white horses of Rhesus from Thrace, we may reasonably infer that in this rare strain declared to be sprung from the gods we have the earliest mention of the bay horses of North Africa. But as Pegasus, the winged horse of Libya, was the offspring of Poseidon himself, the poet's reason for ascribing a divine origin to the dark-bay steeds is now obvious. In the determination of Anchises to obtain by fair means or foul the services of a first-rate sire for his mares we see the same anxiety as is now evinced by Arabs of other tribes to obtain the use of Anazah stallions, and by Turcomans and Kurds to secure the use of Arab sires for their mares. We have thus important indications in the Homeric poems that the best horses known to the inhabitants of Greece at the end of the second millennium B.C. were those of Northern Africa; we have also adduced evidence to show that a like belief was held at a later period, and coupled with a further belief that the first horse that was ridden came from the same region. This belief, taken together with the fact that the race with ridden horses was only added to the list of contests at Olympia at a time when not only had the Greeks become well acquainted with North Africa but were already establishing themselves in the Cyrenaica, leads us to conclude that the Greeks not only 1 Iliad, V. 265 sqq. 2 Iliad, XXIII. 454-5, (poivL^ = 'date-palm,' hence 'date-coloured' like Arab. ku-mait (p. 177), and Lat. spadix (borrowed from Sicily) = ' date-palm ' and Lat. hadius (whence It. haio, Fr. haie, Eng. hay) from Gr. /3dis (from Coptic hai) = ' palm-branch,' and must therefore mean either bay or chestnut (cf. xvppos). 3 Iliad, XXIII. 512 sqq. * Iliad, xxiii. 287 sqq. 19—2 292 THE HORSES OF PREHISTORIC [CH. learned from Libya to yoke four horses to a chariot, as stated by Herodotus, but also learned from thence to ride on horse- back. Nor need we be surprised if the Libyan blood had made its way into Asia Minor at a date anterior to its first advent into Greece, for, as we have seen, the Libyan horses could pass by land into Asia Minor, whilst the broad expanse of sea between Africa and Greece would interpose a formidable barrier to the transport of horses, at least in any quantity, from Libya to Greece, in the days when only ships of small tonnage sailed the seas. It may naturally be said that, if at a comparatively early period the Libyan strain had made its way into Greece, where dun-coloured horses were in general use, we ought to find indications of cross-breeding in the colours, such as have been furnished by the existence of black and grey horses at different points in our wanderings through Asia, Africa, and Spain. First of all let us deal with the question of white horses, which cannot be regarded as tests of cross-breeding, but which must have come down into Greece from Illyria or Thrace, where they were already known in Homeric days ; and then we shall proceed to the evidence for black and grey horses in post-Homeric and classical times. The names of heroes, which occur in early legends, prove that very soon after Homeric days both white horses and black horses, as well as dun-coloured, had become known in Greece. Thus a famous hero, worshipped at Daulis in Phocis, was called Leucippus (" He of the White Horses "), whilst there was a Theban worthy by name Melanippus (" He of the Black Horses "). Now since the myth of Pegasus and the type of Mycenean chariot-wheel indicate that the Greeks early knew the horse of North Africa, and the Homeric poems make it certain that the Acheans had plenty of dun-coloured animals, and as the black horses of Greece must have been a cross similar to those already enumerated, the evidence of Homer and the later myths seems perfectly correct. At what date cavalry began in Greece to replace the chariot in war it is difficult to say with precision, but various considera- Ill] AND HISTORIC TIMES 293 tions render it certain that it was a fully recognized military arm in the seventh century B.C. It will be remembered that the race with the ridden horse was instituted at Olympia in B.C. 648, a fact in itself sufficient to prove that the riding of horses had become a matter of great importance by that time. Moreover when Solon constituted his four new classes at Athens in the beginning of the sixth century, the second, termed the Knights (^iTrnrels:), was composed of those who had suf- ficient property to keep a horse, and serve the State on horse- back in time of war. The Knights naturally were a very aristocratic body, and in the subsequent political struggles they are always found on the conservative side. Just as the medieval gentleman was known by his horse, his hound, and his hawk, so the keeping of horses was the mark of an aristo- crat at Athens and in other parts of Greece ; and as the effigy of a medieval knight is often distinguished by his hound at his feet, or sometimes by his falcon on his wrist, so a horse's head often occurs on Athenian tombstones, indicating that the dead was of a knightly family. Nor was this exceptional. Pausanias^ the traveller "saw not far from the river Crathis in Achaia a tomb on the right of the road with a faded painting of a man standing beside his horse." The hound too occa- sionally got his place on his master's tomb. At Tritia in Achaia the same writer saw a remarkable monument of white marble, adorned with paintings by the eminent artist Nicias : " An ivory chair is seen with a comely young woman seated on it : at her side stands a maid-servant with a parasol. A young and beardless man stands erect, wearing a tunic with a purple robe over it : beside him is his servant holding his darts, and with some hounds in leash. I could not learn their names, but anyone could guess that a husband and wife are here buried together^" The archaic black-figured vases' and those of succeeding 1 VII. 25. 13. 2 vii_ 22. 6. * British Museum Cat. of Greek Vases, Vol. ii. nos. 130, 132 (both show bigae), no. 133 (two youths riding a horse-race), nos. 135, 374, 375 (fight be- tween two warriors on horseback), no. 581 (third horse white), nos. 606, 545, 546, 547 (third horse white). 294 THE HORSES OF PREHISTORIC [CH. styles furnish some evidence respecting chariots and horses from the seventh century onwards. As the four-horse chariot and the ridden horse had been added to the Olympic contests in the seventh century, it is but natural that four-horse chariots and horsemen as well as two-horse chariots often occur on the black-figured and later vases. The horses on the archaic vases are always rendered in black like the chariots and men, but as time goes on we meet representations of chariots in which one of the four horses is white. On a vase representing the death of Hippolytus^ the quadriga is drawn by two white and two yellow horses. On another vase the quadriga of Helios is drawn by two black and two white horses. On the earlier vases the chariot wheels are either four-spoked or of an ancient form which has no spokes, but has a diametrical bar crossed by two others at right angles, whilst on the later vases eight-spoked wheels make their appearance. When we remember that white horses were held sacred among the Germans, Illyrians, and medieval Tartars, and that the sacred chariot of the Persian Zeus was drawn by sacred white horses, we are not surprised to find that in a representa- tion- of a Gigantomachia Zeus is seen in a chariot drawn by four horses of that colour. Moreover, just as the Illyrians and Persians sacrificed white horses, so did the Greeks of the fifth century B.C., in the ratification of solemn oaths ^ As the vase-painters had no hesitation in representing Achilles in a four-horse chariot*, although in Homer that hero like all other Acheans drives only a pair, and as also the painters pourtrayed ancient worthies in contemporary armour rather than in the equipment of the Mycenean and Homeric periods to which they were supposed to belong, so they took for their models the chariots and horses of their own day. The same, as we shall see, holds equally true of the great dramatists, who had as little scruple as Shakespeare and his 1 British Museum Cat., Vol. iv. nos. 279, 305, 258, 487. - British Museum Cat., Vol, iv. no. 237. ^ Ar. Lys. 191-2: ris hv ovv yivoLr'' hv 6pKos ; KAA. el XevKbv wbdev I'ttttoj' Xa^oucrai. rbfiiov ivTe/j.ol/ji.eda. ■* British Museum Cat., Vol. ii. no. 289. Ill] AND HISTORIC TIMES 295 contemporaries in committing anachronisms. As then both dark, yellow, and white horses appear on the vases, we may infer that not only the old dun, but also white and dark- coloured horses of various hues were very common in Greece by the sixth century B.C. This inference is fully borne out by the literary remains. Thus not only are Castor and Pollux described as riders of white horses by Ibycus^ and Pindar^, but the same epithet is applied to the goddess Persephone by Pindar'' and to the Day by Aeschylus*. As I have already pointed out, the sanctity of white probably gave to white horses a fictitious reputation, and accordingly the epithet ' white-horsed ' is applied to the Thessalians® and Thebans'' by Pindar. The vase showing the death of Hippolytus makes it plain that dun horses continued to be used as in Homeric days, an inference fully borne out by the fact that Xanthippus (" He of the Dun Horses ") was a regular name in the great Attic family of the Alcmaeonidae, the father of Pericles the famous statesman being so named. That horses of good blood had raced and won at Olympia before the close of the sixth century may be inferred with certaint}' from a group of sculpture — the work of the famous artist Ageladas — dedicated at Olympia by Cleosthenes of Epidamnus in Epirus, who was victorious with the four-horse chariot in B.C. 516. He set up statues of his chariot and horses, of himself and his charioteer. " The names of the horses also are inscribed : Phoenix {Bay) and Corax {Raven) and on either side of them the side-horses Cnacias {Dun) on the right and Samus on the left." This Cleosthenes was the first horse-breeder in Greece who dedicated his statue at Olympiad The names of the two yoke-horses — the most important and therefore the best — demonstrate that they had a good infusion of Libyan blood. The description of the chariot- race at Delphi given by Sophocles in his Electra^ affords some information about the colours of horses in the fifth century, for, as the poet introduces Ismene^ the daughter of Oedipus, 1 Fr., 16. 2 pyf/j. i_ 66. 3 01. vi. 95. * Persae, 386 ; cf. Soph. Ajax, 673. ^ Pyth. iv. 117. 6 Pyth. IX. 86. " Paus. vi. 10. 6—8. » 701 sqq. ^ Oed, Col. 312 : Airuaias eVt iruXov ^e^uaav ; cf. Aristoph. Pax, 73. 296 THE HORSES OF PREHISTORIC [CH. riding on an Aetnaean steed, a breed which only became famous after Hiero of Syracuse (b.c. 478 — 467) had founded the city of Aetna, we may confidently assume that in his account of the chariot-race and horses he is simply describing those of his own time. There were ten entries for the race, one Achean, one Spartan, two Libyans, Orestes with a team of Thessalian mares, an Aetolian with dun-coloured mares, a Thessalian from Magnesia, an Aenian (also from Thessaly) who had a team of white horses, an Athenian, and the tenth was a Boeotian. It is unfortunate that, though he introduces two Libyan charioteers, he does not tell us the colour of their horses. Yet, since he specially mentions both the white team from Thessaly and the dun-coloured from Aetolia, it is probable that the Libyan horses were neither of these colours, but were bay or some other dark hue. In his comedy of the Clouds (B.C. 423) Aristophanes intro- duces a young spendthrift, who by his passion for horse-racing has overwhelmed his father in debt. The latter is sued by Pasion the inoney-lender for a sum of twelve minae, which he had borrowed to purchase for his son a ' starling-coloured ' {yfrap6<;) horsed We have found a similar epithet {viro-^apo'^) applied by Strabo to the horses of the Parthians and to those of northern Spain. The excessive price paid for the horse (which was four times the amount of the normal ransom of the heavy-armed soldier during the Peloponnesian War) shows that the animal was plainly first-rate according to the Greek standard of that day ; but as ' starling-coloured ' means that the animal was a bluish-black with light-coloured speckles all over, in other words, iron-grey, and as we have found this and other shades of grey to be the regular outcome of crossing the Asiatic-European horse with Libyan blood, we may safely con- clude that the best horses of the day were the result of crossing the old dun horses of Greece with the blood imported from Libya. The literary evidence is amply corroborated by the 1 Nuh. 1225; Aristotle {H. A. viii. 18) says 8 5e \pa.ph% iarl ttoikIXos. In the Nubes we read of other horses, such as Ko-iriraTlas (branded with a 9), but it is possible that it might refer to a blaze down a horse's face ; also of a (ra/j.(p6pa$ (branded with the letter San), but this also might mean some kind of blaze. Ill] AND HISTORIC TIMES 297 monuments. We naturally turn to the horses sculptured by Pheidias to adorn the Parthenon, — to the immortal steeds of Selene and Helios on the east pediment, and to the mortal- ridden horses of the frieze (Fig. 83). Pheidias naturally modelled his immortal steeds after the best living horses which he had seen, and these, as we have just shown, must have been good cross-bred horses. A glance at the head of Selene's horse (Fig. 84) shows that it is not the embodiment of the ISea 'iinrov Fig. 83. Greek Horsemen from the Parthenon. (the Urpferd) as Goethe held, but is rather modelled after a good half-bred horse, probably such an animal as that for which Pheidippides paid twelve minae. The head is long, the cheek and jaw refined, the eye large and prominent, and the nostril well shaped ^ We shall presently find that great length of 1 There are earlier remains of horse-sculptures from the Acropolis, and Olympia has also yielded similar remains. Strongylon (who made in bronze a representation of the Wooden Horse of Troy, the base of which has been found on the Acropolis of Athens, apparently dating from the last quarter of the fifth century b.c.) was famous for his sculptures of horses and bulls. 298 THE HORSES OF PREHISTORIC [CH. head is a characteristic of the progeny of thoroughbred and ordinary horses in modern times. If further proof is required that by 400 B.C. the Greeks had plenty of well-bred horses, Xenophon's^ description of the points of a good war-horse will suffice. The first points to be looked to in the colt are his hoofs, which were naturally of especial importance, as the Greeks never shod their horses, in conse- quence of which the Athenian cavalry horses suffered greatly in the latter part of the Peloponnesian War, after the Spartans had occupied Decelea. As the Athenian cavalry had to be constantly scouting, their horses' hoofs got worn down on the hard ground, and their animals fell lame^ "Thick hoofs are to be preferred to thin ones, for they give a firmer tread, and they ought to be high both before and behind, for high hoofs raise the frog far above the ground ; but low ones tread equally on the strongest and softest part of the foot like in-kneed men." An older writer, Simo, cited by Xenophon, declared that horses with good feet may be known by the sound, and Xenophon commends this as a just observation, for, says he, "a hollow hoof rings against the ground like a cymbal^ The colt's pasterns must neither be too upright like those of a goat, which renders him uncomfortable for riding, nor too sloping, for then he will scratch and gall his fetlocks when ridden among clods or over stones ; he must have plenty of bone in the leg, supple knee- joints, strong shoulders, a broad chest, which is a mark both of beauty and strength, as it keeps the legs wide apart ; the neck as it rises from the chest should not fall forward like that of a boar, but it should grow upwards like that of a cock, and should have an easy motion at the parts about the arch ; the head should be bony, the cheek small, for then he can see things immediately before his feet, his eye should be prominent, 1 Be Re Equestri, 1. - Thuc. vii. 27. •' This phrase puts it beyond all doubt that the words x'^^f'^'^/'o^o' iTriroi, " hoises that stamp with hoofs of bronze " (Ar. Eq. 552), often quoted to prove that the Greeks shod their horses, simply refer to the ringing sound like that of a bronze cymbal produced by the hollow hoof. As Pindar uses the same epithet x^-^K^f^poTos of Demeter (Isth. vii. 3), in allusion to the use of cymbals in her worship, it is clear that Aristophanes, like Xenophon, means that the horse-hoofs ring like cymbals. in] AND HISTORIC TIMES 299 and his nostrils wide. When the upper part of the head is somewhat large and the ears rather small a horse's head is more like what it ought to be." These words of Xenophon would serve well for a description of the head of Selene's steed. " A high shoulder makes the rider's seat more secure and makes the shoulder appear more firmly attached to the body; a double back, that is, when the flesh rises on both sides of the spine, is much softer to sit upon and more pleasing to the eye than Fig. 84. Head of the Horse of Selene : Parthenon. a single one." As the Greeks did not use saddles, they had every reason to dislike a horse ' as lean as any rake.' " The colt should have a good barrel, both for appearance and for easiness of riding ; the loins should be short and broad, for then the horse more easily raises his fore-parts and brings forward his hinder ones ; his haunches should be broad, and well covered with flesh; and if these parts are compact, they will be the lighter for running and make the horse much swifter; the thighs should be broad and straight, for in that case he will set 300 THE HORSES OF PREHISTORIC [CH. his hiud legs well apart and will thus have a quicker and firmer step, a better seat for a rider, and will be better in every respect." The description of the head shows that Xenophon knew what a well-bred horse was like, and indeed by his day it is highly probable that the best horses in Greece were more and more saturated with Libyan blood. We saw that various shades of grey, then black, and finally chestnut and bay are the sure tokens of the gradual increase in the amount of Libyan blood in the horses of Asia, North Africa, and Spain. We have had already clear evidence of grey and black horses in Greece, but it is not until the fourth century B.C. that we have undoubted mention of chestnut. That the Greeks of the fourth century B.C. had chestnut horses with yellow manes is rendered certain by a famous story told by Plutarch \ how on the eve of the battle of Leuctra Pelopidas dreamed that he saw the daughters of Scedaus lamenting round their graves, and that their father urged him, if he wished for victory on the morrow, to sacrifice a fair-haired (^avdi]) virgin to his daughters. Pelo- pidas told his dream to the seers and chiefs, and they long disputed, one party urging a human sacrifice, the other main- taining that such acts found no favour in the sight of the Father of all. At that moment a filly with bright yellow mane came galloping up and stopped short before them. Theocritus the seer cried out to Pelopidas, " Here is the victim you want. Let us wait for no other virgin; take thou and use God's gift." Again, Aristotle'"' when describing the colour of the bisons (bonasi), which still survived in Paeonia down to his day, says that the colour of their hair was a mixture of ash-colour and red, not such as that of the horses termed paroai, that is, of the colour of the parous, a reddish-brown snake, sacred to Aesculapius. As red {pui^ros) is the term used for a bay horse in Revelation, it is probable that the Greeks used pa7'oas to distinguish chestnut from bay. As in Homeric days Thessaly had excellent dun and dappled-dun horses, so her horses and cavalry enjoyed great 1 Pelop. 20-2. ^ H. A, VIII. 32 : to 8^ xP^t^'^ '''°^ Tpix^/J-aros ^x^' " fJ-^cov recppoO sai irvppov. Ill] AND HISTORIC TIMES 301 reputation in Greece in classical times. Not only do Thessalian horses figure in the lists of winners at Olympia, but the Thessalian cavalry were by far the best in Greece. We have just seen that Pindar speaks of the Thessalians as riders of ' white horses,' and that Sophocles represents a chariot from Thessaly as drawn by white horses. Probably soon after Homeric days, the white horses of Thrace had made their way into Thessaly, and they doubtless were improved by superior blood from the south in later times, for not only was Thessaly noted for white horses, but also for grey, a sure indi- cation of crossing with Libyan blood, since Statius describes the Thessalian mares of Admetus as white spotted with black, i.e. 'flea-bitten grey^' Moreover, the studs of Thessaly probably exercised a most important influence on the horses of Macedonia and the rest of the upper Balkan. No better illustration of the horses of that region in the fifth century B.C. can be given than the coins of Potidaea (B.C. 490 — 432) which show Poseidon Hippios (of the Horse) mounted on horseback with his trident levelled against some foe (Fig. 85). The type- was doubtless suggested by the image of Poseidon which stood in front of the city, and we may reasonably believe that Fig. 85. Coin of in the horse on which the god is seated the artist has pourtrayed the best bred horse with which he was acquainted. In Strabo's time^ (a.D. 1) large numbers of good horses were bred in Greece, especially in Peloponnesus. Arcadia was much noted for its horses. It was then a mere solitude, as many of its ancient cities had been destroyed by constant wars, and the inhabitants of many others had been deported to Megalopolis when that city was founded by Epaminondas to be a bulwark against Spartan aggression. There were therefore wide expanses of pastures given up to herds of mares. On ^ Thebais, vi. 336 : noctemque diemque assimulant, maculis internigrantibus albae. - Head, Hist. Numorum, p. 188; Herod, viii. 129. 3 378. 302 THE HORSES OF PREHISTORIC [CH. this account, says Strabo, the Arcadian breed of horses is most excellent, as is likewise that of Argolis and that of Epidaurus. Doubtless it is one of those famous Arcadian horses that we see on a medal dedicated " to the Arcadians " by one Veturius, sometime after the establishment at Mantinea of the cult of Antinous, in the reign of Hadrian \ That bay was the colour of the best Argive and other horses of the Roman world is rendered certain by the description of the steed of Gnaeus Seius, whose owners from Seius down to Antony came to such miserable ends, that he gave rise to the proverb used of an unlucky man, He has Seius horse in his stable^. This horse bred in Argos was of first-rate strain, of " unusual size, carried Fig. 86. Coin of Philip II of Macedon, showing a jockey on a race-horse. his neck well, was of a bay colour, had a flowing mane, and far excelled in every other good point that a horse can have." As in Homeric days, so in later times, mule-breeding was carried on in all parts of Peloponnesus, except in Elis^ and flourished espe- cially in Arcadia^ The large uninhabited areas in Aetolia and Acarnania in Strabo's time rendered this district no less suited for horse-breeding than the plains of Thessalyl Thrace had been famous for its horses and chariots in Homeric days. At what time the horseman finally displaced the chariot we have no means of judging, but it is probable that from at least the sixth century the war-chariot had been superseded by mounted men. In the ^ Head, Hixtoria Numorum, p. 373. - Aulus Gellius, Noct. Att. iii. 9. 3: maguitudine inusitata, cervice ardua, colore poeniceo, etc. 3 Paus. V. 9. 2 ; cf. v. 5. 2. •* Strabo, loc. cit. = Ibid. m] AND HISTORIC TIMES 303 times anterior to Philip II (B.C. 359 — 336) the military force of Macedonia seems to have consisted, like that of Thessaly, in a well-armed and well-mounted cavalry, formed from the sub- stantial proprietors of the country, and in vast numbers of targeteers^. But until Philip formed his famous phalanx after the model of the Theban system established by Epaminondas, the Macedonian infantry was little more than a rabble of shepherds and cultivators. Philip did not merely discipline this raw material into the best infantry that the world had yet seen, armed with the sarissa, a pike 21 feet in length, but he paid equal attention to his cavalry, and it was to this arm that he owed largely his superiority over the autonomous Fig. 87. Coin of Philip II of Macedon, showing a horse-soldier. States of Greece and the Illyrians and Thracians. When Philip came to the throne he bestowed great care upon all that appertained to horses and horse-racing. He sent both chariots and ridden horses to compete at Olympia, and, according to Plutarch'^, he celebrated his victories in the chariot-race on his coins (probably in the biga on the reverse of his gold staters), and he commemorated his victory in the horse-race^ by placing on his silver tetradrachms a representation of the winner ridden by a naked boy — doubtless the jockey — bearing a palm or crowning his horse (Fig. 86). The news of this victory reached 1 Demosth. Philippic in. p. 123. " Alex. 4: ras ev ' OXv/jLwiq. viKas tGiv dpfMaruv e7xapdrrwc roh vofxiir^iacnv ; Alex. 3: i] 8' 'OXv/Mirlaa-w 'iTnrui Ki\r]TL veviK-qKivai. Head, Hist. Numonim, p. 197. 3 Ibid, 304 THE HORSES OF PREHISTORIC [CH. him at the same time as that of Alexander's birth. The jockey was evidently a boy* selected for his light weight, for he is much smaller in proportion to his horse than is the Macedonian trooper (Fig. 87). This difference is possibly due not simply to the fact that one is a full-grown man, and the other a boy, but also to the superior size of the race-horse, which is shown by the set-on of its tail and other points to be a far better bred animal than the ordinary cavalry mount. Others of his coins represent one of his famous Macedonian horse-soldiers, bearded, wearing a hat and cloak (Fig. 87). His cavalry was of two kinds — the Sarissophori or Lancers, apparently light-horse, armed with the sarissa, but probably in this case not more than 14 feet long, and the heavy cavalry who carried, not a javelin for throwing, but the xyston, or pike, suited for thrust- ing in hand-to-hand fight. But apparently the best Macedonian horses were imported from lands further south, if we may judge by the fact that Bucephalus, the famous charger of Alexander, was bred in Thessaly by Philonicus of Pharsalus. Alexander when a boy saw him and purchased him from his breeder for sixteen talents. The horse derived his name either from his appearance, or from a brand on his shoulder in the form of a bull's head^ When Alexander succeeded his father all he had to do was to perfect the military organization which he had inherited. In his vast conquests it would appear that his cavalry was even superior in efficiency to his infantry. The heavy cavalry, which had from of old formed the chief arm of Macedon, and which had been supplemented by the Sarissophori, were known as the Companions. According to Arrian^ at the battle of Arbela there were eight distinct squadrons of the Companions, most of which, if not all, were named after different districts or towns. There was a distinguished royal squadron called the Agema, or leading body, at the head of which Alexander generally charged. Alexander divided these squadrons into half-squadrons in 330 B.C. Not only was the cavalry furnished by Macedonia proper, 1 Cf. Waldstein, Art of Pheidias, p. 415. - riioy, H. N. Mil. 42. 64. •* Anal. i. 2. 89. AND HISTORIC TIMES 305 III] but also by the region known as Upper Macedonia, a fact in complete harmony with the evidence of Herodotus, that in the fifth century B.C. all the peoples south of the Danube had discarded the war-chariot, which was only retained owing to Fig. 88. Head of one of the Horses of the Quadriga from the Mausoleum at Halicarnassus. the small size of their horses by the people north of that river. The conquests of Philip and his great son must have done much to improve the native Macedonian horses. The former became master of Thessaly and the Thessalian horses, whilst R. H. 20 306 THE HORSES OF PREHISTORIC [CH. the latter secured possession not only of the best breeds of Asia, but was able to add to his stud the best blood of Libya, for instance when he marched into Egypt the Cyrenians sent him a present of three hundred horses, doubtless the best that their land could furnish. The effect of this may be traced in the fact that Thrace and Macedonia continued to be famous for their horses under the Roman domination, and, as we have already seen (p. 282), the Thracian horses were noted for their dark colour, with white feet and a white mark on the forehead*, whilst in Byzantine times the emperors of the East kept large studs of brood mares in that region^ In Alexander's time the best horses of Asia were strongly- built animals, showing not much breeding, if we may judge from a fragment of one of the horses (Fig. 88) from the famous chariot- group that once surmounted the Mausoleum (built B.C. 351 — 341) at Halicarnassus. This with the other surviving fragments of the sculptures is now in the British Museum. Thus the suc- cessors of Alexander, whether in Egypt, Asia, or Europe, were equipped with the best war-horses that the world had yet seen, and the Seleucid kings of Syria, who had become the masters of the famous Nisaean race (pp. 192-3), paid great attentiou to the breeding of horses, a circumstance which probably led them to place a mare suckling her foal as a favourite type on their coins. In the struggle of Rome against the Macedonian and Asiatic monarchs the superiority of the latter in cavalry was one of the chief difficulties with which she had to contend. Let us now return to Italy and trace as far as we can from the scanty records the history of the horse at Rome and in Upper Italy. I have already shown that from the early Iron Age the Umbrians had been using chariots. Accordingly we would be almost justified in assuming that the Romans had done the ' Virgil, Aen. v. 565 : quern Thracius albis portat equus bicolor maculis, vestigia primi alba pedis, frontemque ostentans arduus albam. 2 Procopius, cle Bello Vandalico, i. 12: tTrel jSaaiKfi/s 'itnrois 6ti ndXicTTa TrXeiOTOts tov (rrpaTTjybv evravda ibwpetro iK rwv ^aaiKiKCsv 'nnro(pop^iwi', a ol viixovTM is TO, iirl Qpij.Kr]s x^/ata. See also Bell. Goth, iv. 27. Ill] AND HISTORIC TIMES 307 same, for the Patricians were Sabines in origin, the latter being an Umbro-Sabellian tribe. But the terminology of the chief offices of State under the Republic renders it probable that in early days the leading men of the State rode in chariots and not on horseback. The Consuls, Praetors, and Aediles were termed Curule (from currus, a chariot), and they had chairs called sellae curules (' chariot chairs '), and horses termed Equi curules^ (' chariot horses '), were supplied to them at the public expense, even at a time when consuls and everyone else rode on horseback. This seems to point to a time when the chiefs went to war in chariots as did the northern Gauls down to the third century B.C. In the fact that the victorious general down to late times rode in a chariot to the Capitol we may recognise a survival from the time when the chief who had gone forth to do battle in his war car on returning victorious drove his chariot in triumph. through the city^ All this indirect evidence is strongly corroborated by the fact that the Romans in their campaigns against Pyrrhus in Lucania(278 B.C.) used two-horse chariots of peculiar equipment against that king's elephants ^ We have seen that not only the Illyrians, but also all the Celto-Teutonic peoples of Upper Europe regularly sacrificed horses. It is therefore but natural to find that the Romans, whose upper classes at all events were closely related to the Celto-Teutonic peoples, should have had the same practice, for Pliny* especially points out that when a horse was sacrificed on public solemnities the flamen was forbidden to touch it. Whether the Romans preferred white horses for sacrifice like the Illyrians, Greeks and Persians, we cannot say, but it is absolutely certain that they used them to draw chariots on solemn occasions. Thus the Senate decreed that Julius Caesar after his return from Africa should ride in a quadriga drawn by 1 Livy, XXIV. 18. " The dismounted chariot seat would not unnaturally be used as a seat of dignity. ^ Vegetius, iii. 24. 12. ■* H. N. xxviii. 9 : damnat equinum tantum inter venena, ideo flamini sacrorum equum tangere non licet, cum Romae publicis sacris equus immo- letur. 20—2 308 THE HORSES OF PREHISTORIC [CH. white horses^ The Romans so firmly believed in the superior fleetness of horses of that colour that " to outstrip with white horses" became a proverbial expression for an easy victory^, and VirgiP represents Turnus, king of the Rutuli, as drawn by horses " which surpassed in whiteness the snow, in fleetness the wind." At what time the Romans began to ride on horseback it is impossible to say, but apparently they had begun to do so early in the regal period. The original legion was said to have been supplemented by three centuries of horsemen — the Ramnenses, the Titienses, and the Lucres each supplying one hundred men. In the constitution of Servius Tullius eighteen centuries of horsemen were included amongst the ninety-eight centuries of the First Class. The Romans owed little to their cavalry in their conquests in Central Italy and at no time in their history did their strength lie in their horsemen. The conquest of Tarentum and the rest of Southern Italy furnished the Romans with horses of a better quality for cavalry purposes, and as I have already mentioned, it is very significant that the best horses of Italy in Varro's day were the horses of Apulia, no doubt de- scended from the famous breeds developed by the Greeks of Southern Italy from constant importation of Sicilian and Libyan horses. It is not without significance that when the Romans for the first time issued silver money in 268 B.C. — four years after the conquest of Southern Italy — they placed on their coins Castor and Pollux (Fig. 89), Fig. 89. Roman • r rp , Denarius ^ ^JV^ borrowed irom coins oi iarentum or Bruttii. It was but natural that the Romans should keep bringing up the good horses of Apulia to improve ^ Dio Cassius, xliii. 14: iiri re XevKwv 'iinrwv kclI fiera pa^Soi^xw kt\. '^ Hor. Sat. i. 7. 7-8: Sisennas, Barros, ut equis praecurreret albis; cf. Plant. Asin. ii. 2. 12: nam si se huic occasioni tempus subterduxerit, nunquam edepol quadrigis albis indipiscet postea. ^ Aen. XII. 84: qui candore nives anteirent, cursibus auras. These white horses are said to have been the gift of Orithyia, the wife of Boreas, the North Wind. But Virgil as the practical farmer (p. 282) knows that white and dun horses are inferior to bay. Ill] AND HISTOBIC TIMES 309 the native breeds of Central and Upper Italy, and I have shown on an earlier page that in 218 B.C. the horses of the Roman cavalry were superior to those of the Gauls of Northern Italy. We have already seen that from the outset of the Second Punic War the Roman cavalry was inferior to the Numidian and Spanish horsemen in Hannibal's army, and it is highly suggestive that as soon as Scipio conceived the idea of carrying war into Spain, his hrst care was to secure the alliance of Syphax, the Numidian king, and to enter into relations with the Masaesylian Massinissa. Indeed it was in no small degree to the aid of the latter, who joined Scipio with a few followers when that general landed in Africa in 204 B.C., that the Romans were ultimately successful in the closing scene of the drama which culminated in the battle of Zama (B.C. 202), " The Romans made a province of that part of the country which had been subject to Carthage, and made over the rest to the rule of Masanasses (Massinissa) and his descendants, beginning with Micipsa. For the Romans paid particular attention to Masanasses on account of his great abilities and friendship for them, for he it was that formed the nomads to civil life and directed their attention to husbandry, and he taught them to be soldiers instead of robbers'." During the years that intervened between Zama and the final destruction of the hapless city Massinissa and his Numidians effectually kept Carthage from regaining anything of her ancient power. Before Massinissa died (B.C. 141) he had so extended his kingdom that it completely enveloped the Roman province, since it reached even as far as the western Syrtis, and exceeded both in extent and population the territory ruled by Carthage, even in the zenith of her power. Carthage once destroyed, the Romans began to look with alarm upon the kingdom of Numidia, with its vast hordes of swift horsemen. On Massinissa's death Scipio Aemilianus had constrained his three sons to share the kingdom, but as two of them soon died, the whole had lapsed to the survivor Micipsa. This able man improved his father's capital Cirta {Constantineh), and established there a colony of Greeks, and he raised it to such importance that it could put in the 1 Strabo, 832. 310 THE HORSES OF PREHISTORIC [CH. field 10,000 horse and twice that number of infantry. Micipsa had two legitimate sons, Adherbal and Hiempsal, and a bastard, Jugurtha, who far exceeded the others in energy and ability. Micipsa at his death divided the kingdom between all three, but quarrels soon broke out, and Jugurtha in no long time became master of the whole. Then commenced the long struggle between Rome and Jugurtha in which the latter routed the Romans more than once, and even made a consular army pass under the yoke. It was not until Rome had found a military genius in Gains Marius, a rude soldier of fortune, and had been induced by him to remodel her whole military system, that Jugurtha was at last vanquished and led in triumph to the Capitol. I have already shown that Micipsa and his successors paid great attention to the breeding of horses and that no less than 100,000 colts were bred annually in their dominions. It was by means of his overwhelming superiority in cavalry that the Numidian king was able for so long to bid defiance to the legions of Rome. In B.C. 125 the Romans had for the first time permanently established themselves in Gaul, after the overthrow of the powerful Ligurian tribe of Saluvii\ and we have seen how in one of their campaigns against the Ligurians a Roman army had been saved from destruction by a body of Numidian cavalry in the Roman service. When Julius Caesar commenced his war of conquest (B.C. 60 — 56), he found that the chief strength of the Gauls lay in their cavalry, which was composed of the ruling class who had crossed the Rhine and become the overlords of the indi- genous population, the latter forming their serfs and dependents and following them to war. The Gauls, as has been repeatedly shown, possessed horses of fine quality, derived from southern lands at great cost. The conquest of Spain had supplied the Roman army with some good horses and cavalry, and now the 1 W. H. Hall {The Romans on the Riviera and the Rhone, pp. 49, 53, etc.) gives the best account of the Ligurians and the Roman conquest. As these pages are passing through the press, the news of the sudden death of my gifted friend has reached me. The memory of his rare qualities of heart and brain will be long cherished by his many friends. in] AND HISTORIC TIMES 311 subjugation of Gaul gave her henceforth, until the barbarians burst through her frontiers, a practically unlimited supply of serviceable horses. In the time of Augustus the ordinary carriage horse and hackney used at Rome and in Central Italy was a pony' termed mannus'^. As the word is Celtic, it would indicate of itself that the Romans obtained these animals either from Cisalpine or Transalpine Gaul or from both, even if they were not distinctly alluded to as Gallic in the literature of the time^ In their campaigns against the Germans in the first century of our era the Romans seem to have relied entirely on the Gallic provinces for their supply of war-horses^, whilst St Jerome^ writing in the fourth century after Christ, mentions the high value set by worldly men upon Gallic geldings. We shall soon adduce evidence to show that the Roman manni were bred by the Ligurians in what is now north-west Italy and Provence, of which region they were one of the chief productions for export in the first century B.C. (p. 321). The Romans had used geldings for pack-horses from at least the firsf century B.C. and we know not how much earlier. Indeed Cantherius'', 'a gelding,' means properly a ' pack-animal ' and got its secondary meaning from the circumstance that such animals were usually unsexed (cf. p. 167). ' Isidore, Orig. xii. 1, defines mannus as equus brevior, quern vulgo brunitum, vel brunitium vocant. Brunitum vel brunitium are emended to burrichum vel burrichium, as there is a gloss ^ovppixots, but this may be quite unnecessary. ' Lucr. III. 1076: currit agens maunos ad villam praecipitanter ; Prop. iv. 8. 15: hue mea detonsis avecta est Cynthia mannis (where detonsis means "with hogged manes"); Hor. Od. iii. 27. 6: rumpat et serpens iter institutum, si per obliquum similis sagittae terruit mannos; Id. Epod. 4. 14: et Appiam mannis terit; Id. Ep. i. 7. 76: impositus mannis; Ovid, Am. ii. 16 ^Ji.: rapientibus esseda mannis ipsa per admissas concute lora jubas ; Sen. Ep. 87: ita non omnibus obesis mannis et asturconibus et tolutariis praeferres unicum ilium equum a Catone deseriptum ? 2 Hor. Od. I. 8. 6: cur neque militaris inter aequales equitat, Gallica nee lupatis temperat ora frenis. * Tac. Ann. ii. 5 : fessas Gallias ministrandis equis. 5 ad Zech. ix. 9. ^ Varro, R. R. ii. 7. 15, but the cantherii mentioned by Plautus (died 187 b.c.) may be simply pack-animals, not unsexed. 7 From Ok. /cai/^^Xtos := pack-ass (from KavdriXia, 'panniers'); cf. Cic. N. D. m. 5. 11. 312 THE HORSES OF PREHISTORIC [CH. As by the end of the Republic and the first century of the Empire Rome was mistress of a large part of the known world, and was in a position to obtain not only the best horses within her own wide bounds, but also to acquire their best steeds from those tribes who did not own her sway, and, as in the first century of our era, chariot-racing had become a furious passion at Rome, and immense sums were spent on it b}^ the four great factions of the Blue (Veneta), the Green (Prasina), the White (Albata), and the Red (Russata), so named from their distinc- tive colours, it is obvious that if we could ascertain what was the best breed of racers at this time, we would be justified in concluding that this was the best in the known world'. In 1903 several fragments of a long Latin- inscription were found at Rome, built into a wall to the north of the castle of St Angelo. They turned out to be part of a document of which other fragments had already been published^ The inscription had been set up in honour of Avilius Teres, a renowned charioteer in the second half of the first century A.D., and it contains a recital not only of his racing career and how he first drove for the Blue (Veneta) and then changed to the Green (Prasina), but what is more to our purpose it gives a list not only of the horses' names which he steered to victory, but also mentions the breeds to which each belonged. Although the inscription is very incomplete, yet in forty-two cases adjectives giving the horse's nationality can be read. Thirty-seven horses are described as Afer, i.e. from that part of Libya comprised in the Roman province of Africa and in the modern Barbary States ; one is styled Maurus, Mauritanian, one Hispanus, one Gallus, and two Lacones, i.e. Lacedaemonian. Consequently thirty-eight out of forty-two are actual North African horses, whilst from what we have seen of the history of the horses of Spain we know that the South Spanish horses at this time were almost purely Libyan, and Caesar's evidence respecting the Gallic horses and the constant importation of first-rate horses from the south by the Gauls in the two centuries before Christ 1 Suet. Calig. 55 ; Vit. 7. 14 ; Dovi. 7. ^ Revile Archeologique, 1903, Juillet et Aout, no. 160, p. 160. 3 Corp. Inscript. Lat., vi. 10053, 10054, 33943. Ill] AND HISTORIC TIMES 313 make it equally plain that the excellence of the Gallic horse was due to his Libyan ancestry. Finally, the history of the horses of Greece renders it certain that the best breeds of that country were saturated with the Libyan blood. The absence of all mention of any Asiatic horses — Parthian, Armenian, Cappadocian, or Arab — is the clearest proof that the racing men of the time did not look to Arabia or any other district of Asia for horses of preeminent speed, and this com- pletely corroborates the evidence of Strabo in the first part of the same century that the Arabs neither bred nor kept any horses at all. It is now beyond all doubt that from the dawn of history down to the early centuries of our era the Libyan horse surpassed all others in swiftness, and that no horse was able to compete with him save those of Spain, Gaul, and Greece, which were themselves wholly or in great part sprung from the same blood. Of course a very diiferent class of horse from the prize- winners of the circus was required for war, hunting, and other practical purposes, and for such the horses of the Parthians, so highly commended by Strabo, and other good breeds of Asia Minor, Greece, and Italy, which we have shown to have been the result of crossing the Asiatic-European horse with the Libyan, were admirably adapted. For the chief breeds of horses in the second century after Christ we have the evidence of Oppian, who flourished about 180 A.D. In his treatise on Hunting (Gynegetica^), he says that though each country has its own breed of horses, he will only mention the most important, and then enumerates the Etruscan, Sicilian, Cretan, Mazicean, Achean, Cappadocian, Mauritanian, Scythian, Magnesian, Epeian, Ionian, Armenian, Libyan, Thra- cian, and Erembian. His enumeration is not according to order of merit or geographical position, but to meet the exigencies of the hexameter metre. The horses of Libya (by which he pro- bably means the Cyrenaica), the Mazicean- (Numidian), and 1 I. 166—200. 2 The Mazices of Oppian are the same Numidian tribe as the Mazaces and Mazices of Caesar and Suetonius, and are to be identified with the Libyan tribe of Maxyes, who, according to Herodotus (iv. 191), lived west of the river Triton. 314 THE HORSES OF PREHISTORIC [CH. Maiiritanian horses are mentioned separately, in addition to the well-known old breeds of Greece and Asia Minor. It is very significant that in his wide survey of the best breeds of horses Oppian is silent respecting the Arabs, and from this we are justified in inferring that down to the second century A.D., even if the Arabs had by that date begun to breed horses, their steeds were not yet recognized as of any special merit. It will be noticed that the only Italian breed mentioned is the Etruscan, from which we may infer that for practical purposes it had by that date overshadowed the Apulian breeds. We have seen that in Spain, Africa, and in Western and Central Asia black horses are a regular result from the blending of the African with the European-Asiatic horse, and it is probable that the same holds true for Italy. Down to modern times Tuscany, Ancona, and the region of Bologna have been noted for fine breeds of black horses, all of which have been much influenced by African blood derived from Lower Italy. But as the horses of Tuscany were the best Italian breed known to Oppian, it is not unreasonable to conclude that from these horses are descended, in part at least, some of the fine black horses of modern Tuscany. Horses of similar colour but heavier build are found in Lombardy, but these are probably in good part descended from heavy horses brought by the Teutonic invaders of that region, of whose horses we shall soon speak. These horses of Lombardy have been much influenced in later centuries indirectly by Libyan blood through Turkish and Hungarian horses and also by the admixture of heavier horses from Upper Europe (p. 362). Stradanus pourtrays in his Insuher (Fig. 90) a typical example of the Lombard horse of the 16th century. Let us now return to Central and Upper Europe. We saw that the horses on the north side of the Danube were remark- able for their small size in the 5th century B.C., and that down to the time of Caesar the Germans still possessed only their primitive, unimproved, large-headed horses. But as the Gauls beyond the Alps had shown the keenest desire to improve their native breed by importing horses of superior blood at great cost HI] AND HISTORIC TIMES 315 from southern lands, we might assume even without evidence that the Gauls of the lower Danube had for several centuries before Christ been importing horses of improved kinds from Macedonia, Epirus, and Thessaly. But there is not wanting evidence that for a good many centuries before our era the Celts, who dwelt in what is now Styria^ had begun to ride on horseback. At Strettweg, near Judenburg, in that province, a cremation grave of the early Iron Age — formed of large round stones — contained a remarkable series of objects, the most Fig. 90. The Lombard Horse. interesting of which was a small bronze waggon. The vehicle is a simple platform on four wheels, each of which has eight spokes. At each end are the heads of two animals : on the middle of the car stands a woman, nude save for a girdle j there are four figures of men on horseback, each of whom carries a round shield with a central boss and wears a conical cap. There are altogether thirteen figures on the waggon. In this vehicle we have probably a model of the waggons on which the Celtic tribes conveyed their women and children as they 1 Bidgeway, Early Age of Greece, Vol. i, p. 428. 316 THE HORSES OF PREHISTORIC [CH. wandered into southern lands. The occurrence of horsemen dressed and armed in the fashion of the early Iron Age proves that the Gauls on the north-east of the Adriatic had learned by that time freely to ride on horseback. But as the tribes on the north side of the Danube continued to fight in chariots because their horses were so small, we may infer that the Celts of Styria had been able to obtain horses of a better kind from the lower Balkan peninsula. It was almost certainly owing to this advantage that the Gauls of that region had been able to develope a very re- markable cavalry organization, which formed the chief element of success in their invasions of Macedonia and Greece. The description of the admirable cavalry system of those who invaded Greece and got as far as Delphi in 279 B.C., de- monstrates that they had long before procured good strains of horses from Macedonia and northern Greece, and had learned to utilise them, just as quickly as at a later time did their kinsfolk in Gaul, as soon as they had obtained horses of superior quality from Italy or Spain : " When Brennus persuaded his people to invade Greece the assembled army numbered one hundred and fifty-two thousand foot, and twenty thousand four hundred horse. But though that was the number of cavalry always on service, the real number was sixty-one thousand two hundred ; for every trooper was attended by two serfs, who were themselves good riders and were provided with horses. When the cavalry was engaged, the serfs kept in the rear, and made themselves useful thus : If a trooper (ol LTTTreuovTe'i) had a horse killed, the serfs brought him a fresh mount : if the trooper himself was slain, the serf mounted his master's horse ; but if both horse and man were killed, the serf was ready mounted to take their place. If the master was wounded, one of the serfs brought the wounded man off the field to the camp, while the other took his place in the ranks. These tactics, it seems to me, were coj)ied by the Gauls from the Persian corps of the Ten Thousand, known as the Immortals. The difference was that in the Persian corps the places of the dead were filled up by enlistment after the action, while with the Gauls the squadron Ill] AND HISTORIC TIMES 317 was brought to its full strength on the field of battle. This organization they called trimarcisia ('three-horse' system) in their own tongue ; for you must know that the Celtic for a horse is marca^." This valuable passage gives us a picture of a society and a military organization closely resembling the feudal system that sprang up in all the countries conquered by the Teutonic tribes after the downfall of the Roman Empire. We have here the medieval knight attended by his squires, though in the present case the latter are not freemen, but belong to the conquered people, and have to follow their lord to war. But the institution here set forth did not belong merely to the Gauls of the Danube, for it would appear that wherever the Celto-Teutonic tribes from Central Europe pushed their con- quests, they established a like system. The squires, here termed serfs (douloi'^) or bondsmen, are identical with the am- bacti^ of Gaul in Caesar's time, where there was a ruling class of knights (eqitites) who had passed over the Rhine and con- quered the old melanochrous population of France, the latter becoming the vassals and dependents of their Celtic conquerors. These Celtic lords spent all their time in war, and the greater each was in birth and power, the more ainbacti or clientes had he around him. That the institution of ambacti was no new feature among the Celto-Teutonic tribes is shown by the fact that Ennius*, the father of Roman epic poetry (239 — 169 B.C.), knew of it as a Gallic term. I have elsewhere^ shown that the fair-haired Acheans of Homeric Greece had come down from Central Europe into Greece, and conquered the old dark Pelasgic inhabitants, making them into their vassals and dependents, and compelling them to follow them to war. We have seen, on an earlier 1 Paus. X. 19. 4 sqq. (Timaeus, of Locri, was probably Pausanias' authority). 2 For this use of SoDXos (and dovXeia) for a serf or vassal population cf. Aristotle, Pol. ii. 5. 22; Thue. v. 23. * Caesar, B. G. vi. 15: plurimos circum se ambactos clientesque habent. With (unbactus cf. Gothic andbahti = service, andbahts — servant. ■* Paul, ex Fest. p. 4 (Miiller) : ambactus apud Ennium lingua Gallica servus appellatur. 5 Eidgeway, Early Age of Greece, Vol. i. pp. 337 sqq. 318 THE HORSES OF PREHISTORIC [CH. page, how one of these dependents of Agamemnon had to purchase exemption from following his lord to Troy by pre- senting him with a famous mare. It is clear then that long before the Christian era the plains of the Danube were producing horses of excellent quality, and this they continued to do under Roman rule\ The Huns of course brought their own hardy horses from the steppes, but it is almost certain that these were soon improved by crossing with the already improved breeds developed by the old Celtic occupants of that region. Certainly by the fourth century A.D., when Vegetius^ wrote, the Huns were famous for their horsemanship, and he praises the horses of the Huns and other northern peoples for their hardiness and freedom from disease, though left out on pasture through the winter frosts, and never stabled. He remarks that from the example of the Huns the Romans of his own time, who wished to save expense in matters of careful grooming and horse-doctoring, pretended to follow the example of the Huns, whose horses, though left uncared, had such excellent constitutions ^ But Vegetius points out that the northern horses were naturally of a hardier stock, whilst the Roman horses were not only of a more delicate constitution, but were reared more tenderly, being housed from the time they were foals. Accordingly when he describes the chief breeds of horses fitted for the war, the race-course, and for the road, he puts the Hunnish horses at the head of the war-horses, next in order being the Thuringian and the Burgundian, and thirdly the Frisian^ He^ gives a very full account of the Hunnish horse, which I give here in Thomas Blundeville's admirable version. 1 E.g. Dalmatian and Epirote horses (Veg. Ars Vet. iv. 6). 2 Ee Mil. III. 26. 15. * Ars Veterinaria, iii., prol., sect. 1. * Veg. Ars Veterinaria, iv. 6. 3. The mss. read Frigiscos. The Frigisci cannot mean anything else than Frisian, for they must be a northern breed owing to their association with Toringos. 5 Ars Vet. iv. 6. 5: Hunniscis grande et adimcum caput, extantes oculi, angustae nares, latae maxillae, robusta cervix et rigida, iubae ultra genua pendentes, maiores costae, incurva spina, cauda silvosa, validissimae tibiae, parvae bases, plenae ac diffusae ungulae, ilia cavata, totumque corpus angu- losum, nulla in clunibus arvina, nulli in musculis tori, in longitudine magis HI] AND HISTORIC TIMES 319 " The Hungarian hath a great and hooked head, and his eyes stand ahnost without his head, his nostrils are narrow, and his jaws broad, his neck is long and rough, with a mane hanging down nearly to his knees, he hath a large bulk, a right back, a long bush tail, his legs be strong, his pasterns small, and his hoofs full and broad, his guts are hollow, and all his body is full of empty corners, his buttocks are not filled with fat, neither do the brawns of his muscles appear, of stature he is more in length than height, and therewith somewhat Fig, 91. The Hungarian Horse. side-bellied, his bones are also great, he is rather lean than fat, which leanness is so answerable to the other parts of his body, as the due proportion observed in his deformity, maketh the same to be a beauty. And as touching his inward dis- position, he is, as Vegetius saith, both temperate and wise, and able to abide great labour, cold and hunger, and very meet for the war." " Camerarius also saith that they be very swift, and quam in altitudine statura, propensior venter, exhaustus, ossa grandia, macies grata, et quibus pulcritudiues praestat ipsa deformitas: animus moderatus et prudens, et vulnerum patiens. 320 THE HORSES OF PREHISTORIC [CH. if they be provoked by some injury, they will both bite and strike, otherwise not. Their pace is a trot." The Hungarian horses have been continually improved by the introduction of Libyan blood, derived largely in later centuries through Turkish channels. Accordingly it is not surprising that the Hungarian horse, drawn by Stradanus (Fig. 91), in the " Stable of Don John of Austria," shows little resemblance to the animals described by Vegetius except as regards the copiousness of the mane and tail, which were probably inherited from the ancient horses of the Danubian region. The old Hungarian horse was usually of a bay colour and without any white on the legs, but grey, dun, and chestnut were likewise often found. Since the early part of the last century this type has been entirely changed owing to the constant importation of English thoroughbreds, when the Government began to breed for military purposes and en- couraged the fai-mers to do likewise. " In almost all cases the Government stallions were half-bred English, and these were placed at breeding depots all over the country^" As is well known, Hungary at the present time supplies some of the best cavalry horses in the world. I have already pointed out that the black horses of Western Asia, Spain, and Italy all result from a mixture of the African bay with the indigenous horses of Asia and Europe. If this principle is sound, the same colour ought to characterize strongly the horses of the Upper Balkan and Danubian regions. But large black horses are so distinctive a feature not only of this area, but also of those lying to the east, that the cavalry of Austria and Russia has been regularly mounted on horses of this colour. Our evidence now makes it clear that black is not an original colour of the horse either in Europe, Asia, or Africa, but that it is an artificial product arising from the mixing of the African stock with the Asiatic-European indigenous horses in the three southern peninsulas of Europe, in Syria, Anatolia, and other parts of Western Asia. Let us next pass into Northern Italy and France. Now 1 Hayes, Points of the Horse (ed. 3), pp. 531-2, embodying the notes of Mr Keynolds (M.K.C.V.S.). Ill] AND HISTORIC TIMES 321 since the horses of the Celtiberians, who occupied all Northern Spain, were iron-grey, and as the horses of Southern Spain were bay, and occasionally black, and as the Gauls were import- ing horses of superior blood from the south from 172 B.C. and probably much earlier, and in Caesar's time were paying large prices for horses from southern lands, it is but reasonable to expect that some of the oldest breeds in France should show characteristics similar to those of the cross-bred horses of Spain. All French authorities are agreed that the fine breeds {traces legeres), of which there were several of great antiquity and excellence in France, are derived from the ' Oriental ' or ' Arab,' or in other words from the Libyan horse. In Strabo's time^ the Ligurians were noted for a particular breed of horses called ginni, i.e. jennets, which the geographer (probably following Posidonius) mentions amongst the chief products and exports of that region. That these jennets were not mules but ponies is made absolutely certain by the statement of Aristotle^ that " the animals called ginni are stunted horses and bear the same relation to horses that dwarfs do to well- grown men." It is therefore certain that the Ligurians had an excellent breed of ponies before the Christian era, and as these ponies were sent down into Italy, we can have little doubt that they were the manni of the Roman writers of that period. The Ligurians, who lived on the Italian side of the Alps, in their struggle against the Romans seem to have had no cavalry, but when in 125 B.C. the Romans undertook for the first time to carve out a province on the Gallic side of the Alps, they came into contact with the powerful tribe of Saluvii, whose capital was Arelate (Aries), and who were well mounted on the horses which they bred in the plains east of the Rhone. To-day in the same district we meet the horse known as the Camargue. It is reared in a half-wild state, and its origin is ascribed "to the introduction of Arab or Numi- dian blood in the neighbourhood of Aries in 125 B.C., when Fulvius Flaccus, the Roman general, occupied the country." 1 202. 2 Hisi_ Animal, vi. 24. R. H. 21 322 THE HORSES OF PREHISTORIC [CH. The original Libyan blood thus obtained is supposed by French writers to have been augmented by the establishment of the colony of Julia (circ. 24 B.C.), and later at the time of the Saracen occupation of Provence (730 A.D.), and later still at the time of the Crusades. The Camargue is a small horse (1'32 — 34 m. = 13'1 hands). His head is a little big, but well set on. His feet are large and often flat ; his coat is always grey. His head and feet point clearly to his European ancestry, whilst the modification of his colour points equally clearly as in the horses of Northern Spain, Syria, Mesopotamia, and Asia Minor, to his Libyan blood. As we have shown that not only were Gallic chieftains importing horses across the Alps long before the Roman occupation of Provence, but that the Celtiberians by the beginning of the first century B.C., and we know not how much earlier, possessed a breed of grey horses, it is more than probable that Libyan blood had been introduced into the region of the Rhone at a time long anterior to 125 B.C. Again, as we have seen that the osseous remains of the horses used by the Helvetians in the first century B.C. (p. 93) are declared by Dr Marek to agree in their fundamental characters, size excepted, with the so-called Oriental races of horses, whose typical representative is the 'Arab,' and as this Helveto-Gallic horse was 1'35 to 1'41 m. in height, almost the same as that of the modern Camargue, we are led to conclude that the Ligurian ginni of Strabo's day were not only the ancestors of the modern Camargue but were the same breed as the horses whose bones have been found in the settlements of the La Tene period. The crossing of the Camargue with the Arab in modern times has given excellent results ^ The Basses Pyrenees and the Hautes Pyrenees are the seat of an ancient breed known variously as that of Navarre, Tarbes, or Bigourdan^. It was derived from Andalusia according to some at a period later than the Arab conquest of Spain, accord- ing to others at the same time, but from what w^e have already seen it is probable that it had imbibed much Libyan blood 1 Guyer and Alix, Le Cheval, p. 013. - Cuyer and Alix, oii. cit. p. 609. Ill] AND HISTORIC TIMES 323 at a far earlier date. In form it is said to fall about midway between Arab and Andalusian. The breed was decadent by the close of the eighteenth century, and has since then been greatly modified by being crossed with Arab and English thoroughbred stallions, the former being used exclusively under the Empire and the Restoration, but in 1833 the English horses were introduced, and since then the horse of Bigourdan has gTadually supplanted that of Tarbes, being taller than the latter. The crossing has increased its height, which is now 1'60 m. (16 hands), and its head is longer than that of the thoroughbred. In the Eastern Pyrenees we meet the horses of Ariege, which are reared on the plateau of Laderg, at a very consider- able height above sea-level. They have all the characteristics of a mountain type — being ugly and angular, but very hardy and useful. Their relatively large size (1"45 — 50 m. = 15 hands at most) is due to the excellence of the mountain pastures. " Everything points to the belief that they are Spanish in origin," for they preserve in great part the character of the Andalusian. The coat is generally blacks The Limousin horse was the glory of old France, for it was esteemed above all others for the saddle, and the royal stables were filled with animals of this breed. The majority of hippo legists are agreed in dating the origin of the Limousin stock to the conquest of Spain and Southern France by the Saracens in the beginning of the eighth century A.D. It has the distinctive characteristics of the Barb. The chief centre for its breeding was in Haute Vienne. It was in full decadence in 1770, and in modern times it has been much crossed with Arab, Spanish, and English thoroughbred blood, and has consequently lost its ancient shape and qualities. Since 1830 English thorough- breds have been exclusively used and have given the race greater size, but not so good a barrel. The Auvergne horse is also ' Oriental,' i.e. Libyan, in origin, being absolutely identical with the horse of Ariege, its colour being generally black. They are not so fine as the Limousin, 1 Cuyer and Alix, op. cit. p. (309. 21 2 324 THE HORSES OF PREHISTORIC [CH. for the head is larger in proportion to their size, but in modern times they have been modified by English thoroughbred blood \ The Landes furnish a breed of horses without doubt partly Libyan in origin. They are reared in a half-wild state, de- pendent entirely on the scanty pasture of their native wastes. They are of small size (I'lO — 30 m. = 11 — 13 hands) with small square heads. They are hardy and untiring. The results from crossing them with big breeds, especially the English thorough- bred, have always been bad, but when these little mares are mated with an Arab, whose height and feeding is much more on a par with their own, the results are excellent. The horses bred in Bas-Medoc are the result of crossing the indigenous mares with English thoroughbred or Anglo-Norman stallions. Their height should make them suited for cavalry of the line, but in every other respect they are ill-suited for this purpose, being ugly, awkward, nervous, and bad-temperedl The horses of Morvan (whose seat is in Saone-et-Loire and Nievre) are ' absolutely identical ' with the horses of Auvergne and those of Ariege (Eastern Pyrenees). This breed had a great reputation under Louis XV, and is still valued, though degenerate^. Though the departments of Western France supply but few fine horses, yet Brittany from time immemorial has had an excellent breed, 'absolutely identical' with that of Morvan, Auvergne, and Ariege. The best French authorities maintain that the Breton horses are 'Asiatic,' i.e. Libyan, in origin. According to M. Sanson the introduction of this type must go back to the Celtic epoch, and he places it under his E. c. hibernicus (p. 2). These horses, known as bidets, are reared everywhere in the mountains and plains of Brittany, especially near Guingamp, Cartaix, Loudeac, Brest, Morlaix, and Redon. They have short, square heads, and they do not exceed 1-50 m. (15 hands) in height. They have been spoiled to a certain extent by the introduction of English thoroughbred stallions^ This breed, which thus extends right across France, from Ariege through Auvergne, and down the Loire into Brittany, ^ Cuyer and Alix, op. cit. pp. 605-6. - Cuyer and Alix, op. cit. p. 609. =» Cuyer and Alix, op. cit. p. 617. ^ Cuyer and Alix, op. cit. p. 614. Ill] AND HISTORIC TIMES 325 we shall find to be of great importance when we come to inquire into the origin of the Irish Hobby and Ewart's Celtic pony (p. 18). The little horses of the Mouse, Moselle, and Meurthe (the ancient pi'ovince of Lorraine), are held by Sanson to be ' Arab ' in origin. They have great toughness and endurance, though ugly in shape. They are now only to be found amongst the poorest people. Their decadence commenced with the intro- duction of Ukraine blood in 1757, and has been aggravated by the introduction since 1807 of Belgian, Percheron, and Anglo- Norman blood. Alsace formerly possessed a breed of small horses of 'Asiatic' type, but at the present day the Alsatian horses are of little value*. The breeds of which we have just been speaking are all of a dark colour, like the horses of North Africa and Andalusia, from which they are sprung, and indeed they are commonly black. But, as in Spain we found not only the pure or almost pure Libyan horse of a dark colour, but also a grey breed, partly Libyan and partly derived from the old European stock, which still exists in Northern Spain, the land once occupied by the Celtiberians, and as we have already found such a grey breed in Provence, it is but natural that breeds of a similar origin and colour may be found in Central France also. The most famous of all the French half-bred or intermediary horses is the Percheron, who is as much renowned in his class as the English race-horse is in his. The centre of production of the Percheron is what was formerly the little province of Le Perche, distributed now between the departments of Orne, Sarthe, Eure-et-Loire, and Loir-et-Cher, the actual geographi- cal area of the breed only covering a portion of each of these departments '^ The principal breeding centres are Mortagne, Bellesme, Saint-Calais, Mont Doubleau, and Courtomer. There are two kinds of Percheron, the little and the big. It is the small or Percheron postier, that was so universally used for posting and for coaching. The head is a little large, souvent 1 Cuyer and Alix, op. cit. p. 617. . ^ Cuyerand Alix, oj). cit. p. 640. 326 THE HORSES OF PREHISTORIC [CH. camuse, the forehead large, the eye small but quick and in- telligent, the neck of moderate length with a long fine mane, the back short, the croup round and muscular, and well-rounded sides, the tail set a little low, strong legs with large joints, short pasterns. It is generally grey. The little Percherons horse the Paris omnibuses and the French artillery. Various theories have been advanced touching the origin of the Per- cheron, some holding that he is an Arab become heavy under a particular kind of work and feeding in the course of some centuries, others consider him the outcome of the blending of the Breton with the Boulonnais, whilst M. Sanson ^ makes it into a separate species, E. c. sequanius, and holds that it de- veloped in the Parisian basin of the Seine (Sequana). MM. Cuyer and Alix ^ accept this view, believing it to be confirmed by the discovery at Grenelle of a skull of Equus caballus, the only quaternary skull of Equidae known up to the time when they were writing (1884), of which the typical characteristics are those of the Percheron breed. Though it is possible that this skull may be that of a cross between the ' Celtic ' pony and the heavy built horse of the Solutre type, yet the grey colour of the Percheron taken in conjunction with that of the Camargue, and the same colour in the horses of Northern Spain, about whose ancestry we are fairly certain, render it far more likely that the Percheron is the outcome of blending the old heavy European horse with Libyan blood derived through Spain and Italy. There are also horses known as the large Percheron, but they must not be confounded with the small or true Percheron, for in the plain of Chartres there are horses of various other breeds, some of them very large and heavy — Breton, Boulon- nais, Flemish, Picard, Norman — but as the mode of rearing tends to assimilate all these horses to the older breed, they are commonly called Percherons and sold as such. It will be remembered that by 100 a.d. the German tribe of Tencteri, who had settled on the left bank of the Rhine, were distinguished from all other German tribes by their love of ^ Zootechnie, Vol» iii. p. 105, . - Op. cit. p. 641. Ill] AND HISTORIC TIMES 327 horses and their finely organised cavah-y (p. 115). It was pointed out that their superiority in horses over their kindred was due to the fact that they had been able to obtain a better class of horses from their Gallic neighbours, who had been importing at great cost fine horses from southern lands long before the time of Caesar. By the third century the Tencteri, like the Ubii and other tribes, who dwelt on the left bank of the Rhine, had lost their identity under the common term of Franks, which had gradually supplanted the older name of Germans ^ But their kindred on the other side of the Rhine from the Main down to the sea maintained their autonomy in their ancient seats from which they were one day destined to sally forth to conquests pregnant with empire. Next to the Franks on the east lay the Thuringians, whilst on the south from the Main as far as Basel came the Burgundians. As the Tencteri had been able to obtain superior horses and to organise a fine cavalry in the first century A.D., it was but natural that some of the other tribes should soon follow their example. It is not then surprising to find that Vegetius'"* {circ. 380 a.d.), in his list of breeds suited for war, places the Thuringian and the Burgundian next after the Hunnish, and gives the third place to the Frisian. Unfortunately, Vegetius does not mention the characteristic colour or colours of these different breeds of war-horses, but it is not improbable that many of them were already of a dark colour. Certainly by the beginning of the sixth century a dark colour with blaze on the face characterised the best Roman war-horse of the day. This is rendered clear by the story of the famous fight which took place near the Tiber, when Belisarius, Justinian's great general, with a thousand of his cavalry came suddenly upon a party of the Goths, who were bent on the capture of Rome and had already crossed the river. Belisarius himself fought like a 1 Procopius, de hello Gothico, i. 12: 'Prjvos 8e is tov ihKeavbv ras €K(3o\as iroutTai, \Lfwai re ivravda, ov 5rj Tepixavol to wciKaibv t^KrjvTO, ^dpjSapoi' idvo%, ov TToXXoO \6yov to /car' apxas d^iov, ot vvv ^pdyyoi KoKoui'Tat.. 2 Ars Veterinaria, iv. 6. 2 : ad bellum HunDiscorum longe primo docetur utilitas patientia laboris, frigoris, famis. Toringos deinde et Burgundiones iniuriae tolerantes. Tertio loco Frisiscos non minus velocitate qnarn continua- tione cursus invictos. Postea Epirotas, Samaricos, ac Dalmatas, licet contu- maces ad frena, habiles armis [ac bellis] asseverant. 328 THE HORSES OF PREHISTORIC [CH. common soldier. "He happened to be riding at that moment a horse well tried in war and who knew well how to carry his rider through in safety. All his body was dark-coloured, but his face from the top of his head to the nose was pure white. Such a horse the Greeks called phalios ('bald') and the barbarians balas ('bald'). This horse was recognised by deserters from Beli- sarius, who had joined the Goths, and they immediately shouted out, "Strike the bald-faced horse." Nothing but the devotion of his body-guard saved Belisarius and his noble charger^ The gallant war-horse here described must have differed essentially from the ordinary post-horses of the day, which were kept at the public expense along the great roads of the Empire, and on which Belisarius himself once made a memorable journey, when Justinian, on hearing that the Persians had invaded his dominions, sent Belisarius to oppose them. " Riding on the public horses which are commonly known as veredi (German Pferd), inasmuch as he had no army with him, with great speed he reached Euphratesia I" As we may assume that Belisarius' well-tried charger is a fair representative of the best war-horse of the time, it is now clear that already by the beginning of the sixth century a dark-coloured animal, probably either dull black or dark- brown with a white blaze on the face, — features which we have seen to characterise the large cross-bred horses of Asia, North Africa, and Spain — was already the typical war-horse of Europe; and it is not improbable that the Thuringian, Burgundian, and Frisian horses, so highly praised as war-horses by Vegetius, may well have been of a similar dark colour, especially in view of the fact that from before the Christian era the fine cross-bred horses of Northern Spain were iron-grey, a colour which easily passes into black. The Roman contorniates (Fig. 92) of the 1 Procopius, de bello Gothico, i. 18 : #ri>xf 5^ 'Lirjrii} T7)VLKavTa oxovfievos, TToX^yUWi' re \lav ifxirelpo} koX diaauaaadai tov iTn^drriv iTricTTafiivq), 6? drj SXav p.€v TO cQfxa ^^^ ^jsvi- depicted in green (Fig- f\{^^^i%C^- 110)1. Neither has this '© W^^^^^^ ll%5^-. horseman nor another on •5^'is^^^^^^'=i^^ t .■ t- 77 ' '^ liG. 110. Irish Horseman ; Book oj Reus. saddles, and the use of bridles or halters we may recognize in the horsemen of the Book of Kells the same equipment as that used by the Irish in the time of Giraldus some four centuries later. It has now been proved that the typical Irish horse as described by Thomas Blundeville in 1580 was no recent outcome of Spanish sires, as believed by the members of the Royal Commission, but was already in general use in Ireland by the tenth century, if not earlier. But this is not all ; three horse-skulls lately discovered in a crannog (lake-dwelling) are of the highest importance in proving that not only were horses of the North African type used in Ireland as early as the tenth century, but possibly at a considerably earlier date. 1 Wilde's Cat. of the Antiq. of the Roy. Irish Academy, p. 300. 392 THE HORSES OF PREHISTORIC [CH. These three skulls (now in the National Museum in Dublin) were found associated with various antiquities, the character of which has led Mr George Coffey, the keeper of the National Museum of Irish Antiquities, to date the deposit as not later than the tenth century and possibly as early as the sixth century A.D. Dr Scharff, the head of the department of the Irish National Museum of Natural History, has pointed out to me that the skulls, which are beautifully preserved, have the distinctive features of the Arab, i.e. North African horse. The various kinds of evidence here adduced j)ut it beyond doubt that at a period long anterior to the supposed intro- duction of Spanish stallions into Ireland in Tudor times that country already possessed a breed of horses closely related to the North African. The question now arises. Is the typical Irish horse an in- digenous development from an Equus caballus celticus or from E. c. eurupeus typicus or from both, or is it the outcome of a very early and oft-repeated admixture of North African- blood with that of the old European horses, whether an E. c. celticus or E. c. typicus, or with both ? Both Col. St Quentin and Sir Walter Gilbey hold that the superiority of the Irish horses over all others is due largely to the limestone formation of the great central plain of Ireland and the excellence of the pasturage, but probably neither of these writers would maintain that the present Irish horse could have been evolved either out of the typical Great Horse of the Continent and England in the lapse of some centuries, or from the ancient horses of Upper Europe without any blending of other blood. Indeed the very value attached to ' blood ' by Col. St Quentin, when writing of the Irish horse, shows that he, like Mr Kenny, Dr Cox, and the members of the Royal Commission, believes that the Irish horses have in their veins a very large proportion of Arab, i.e. North African blood. It will therefoi-e be hardly asserted by anyone that the Irish Hobbie of the time of Thomas Blundeville had been specialized merely under favourable conditions of soil and climate from the heavy-headed, thick-set horses of Europe and Asia, especially in view of the three skulls from a lake-dwelling just cited, which Ill] AND HISTORIC TIMES 393 alone are sufficient to indicate that horses of the North African type or saturated with North African blood had reached Ireland at a very early period. It is even possible to find some evidence of a far earlier date which has a direct bearing on the origin of the horses whose skulls have been preserved in the peat-buried crannog. I have already pointed out (p. 98) that in the oldest Irish Epic cycle — that of Cuchulainn — the combatants never ride on horseback, but always fight from chariots, whereas the reverse is the case in the poems of the later or Ossianic cycle, as was apparently the established practice when the Book of Kells was written. Of course various dates have been assigned by scholars to the Cuchulainn Saga, but though it may have been revised and augmented at a later period the main elements in the poems belong to pagan times. The events commemorated in these poems are supposed to have taken place in the first century before Christ, and even though it may not be admitted that they were first composed at so early a date it will be generally conceded that the main body of the poems was composed in pre-Christian times, for there is good evidence that some of them were already regarded as of great antiquity in the seventh century A.D. In The Wooing of Emer (who lived at Lusk in co. Dublin) we are told that Cuchulainn went to Alba, i.e. Albion, to perfect himself in feats of arms, and that he learned there the use of the scythed chariot, and in such a chariot he set out to see Emer, after his return from Alba. When Caesar invaded Britain in 55 B.C. he found the Belgic tribes still using chariots, although they also possessed cavalry, whilst for a considerable period later the tribes of the north of Britain continued to use chariots, as we have already seen (p. 95). But it seems most unlikely that they continued to do so for very long after the Roman Conquest. It is not very probable that the use of the chariot for war would have continued in Ireland very long after it had been replaced by men mounted on horseback in Britain. Though this is not the place to deal in detail with the armature and dress described in the poems, it may however be pointed out that both dress and arms seem to belong to 394 THE HORSES OF PREHISTORIC [CH. what is known as the La Tene, or ' late Celtic ' period, so-called from the Gaulish settlement at La Tene, on Lake Neuchatel, where many remains typical of the Gallic culture of the centuries immediately preceding the Christian era were dis- covered. Fig. 111. Sword of La T^ne type in its sheath ; Connantre, Marne. Fig. 112. Iron Sword (La Tene type) in bronze scabbard; Hallstatti. Fig. 113. Brouze Shield; Bingen-. The Gauls at that period used long iron swords of a well- defined type (Figs. Ill, 112), carried oblong shields (Fig. 112) instead of the round shields (Fig. 118) of the previous period, 1 W. Ridgeway, The Early Age of Greece, Vol. i. p. 410. ^ Id., Vol. I. p. 478. AND HISTORIC TIMES 395 III] and fastened their cloaks with a peculiar form of brooch (Fig. 114). Now relics of this La Tene culture are not wanting in Fig. 114. Bronze Fibula; Marne^ Fig. 115. Alderwood Shield; Ireland^. Ireland, for there is a shield of the Gallic type in the Museum of the Royal Irish Academy (Fig. 115) differing completely from 1 Bidgeway, op. cit. Vol. i. p. 456. 2 Id., Vol. I. p. 580. 396 THE HORSES OF PREHISTORIC [CH. the earlier round shields used in Ireland (Fig. 118), and from the round shields used by the Irish at a later period, such as that still in the possession of the O'Donovan (Fig. 119). Similarly, brooches of undoubted La Tene or 'late Celtic' types have Fig. 116. Bronze Fibula; Ireland^ occasionally been found in Ireland (Fig. 116), and this type of brooch is undoubtedly described in at least one passage of the Cucliulamn Saga, where Cuscraid son of Conchobar, the tall, yellow-haired, grey-eyed king of Ulster, is represented as wearing a dark-grey cloak fastened round him with " a leaf- shaped brooch {delg ndaillech) of white metal over his breast l" This epithet exactly fits such a brooch as that here figured from Navan Rath, co. Armagh (Fig. 116), but could not possibly be applied to the pen- anular (Fig. 117) brooches used in Ireland at a later period, and the genesis of which I have elsewhere traced^ As leaf-shaped brooches closely allied to the Irish brooch here shown, were in use by the Belgic tribes of East Anglia at the time of the Roman conquest of Britain^, there are good grounds for believing that the poems of the Cuchu- lainn cycle took shape in the period in which are laid the scenes described. Iron swords of the La Tene type have been discovered in Ireland ^ ^ Ridgeway, ojy. cit. Vol. i. p. 464. - The Cattle liaid of Cualnge, translated by L. Winifred Faraday, M.A. (1904), p. 120; Yellow Book of Lecan, fol. 46 b, 28. '•^ The Earlji Age of Greece, Vol. i. pp. 589 sqq., Figs. 146—152. ^ Ibid., p. 581, Fig. 131. ^ E. Munro, The Lake-dwellings of Europe, pp. 382-4. Fig. 117. Bronze Fibula ; Ireland. Ill] AND HISTORIC TIMES 397 Such relics seem to indicate that the Irish traditions of the invasion of fair-haired strangers from Gaul, in the centuries preceding the birth of Christ, are based on actual facts, and are not the mere outcome of a romancist's brain. In any case they demonstrate that there was constant intercourse and trade between the Continent and Ireland at that period. Fig. 118. Bronze Shield; co. Limerick ^ Fortunately, for our immediate purpose, a description of Cuchulainn's horses is given in The Wooing of Emer. They were " alike in size, beauty, fierceness and speed. Their manes were long and curly, and they had curling tails. The right- hand horse was a grey horse, broad in the haunches, fierce, swift, and wild ; the other horse jet-black ; his head firmly knit, 1 The Earhj Age of Greece, Vol. i. p. 479. 398 THE HORSES OF PREHISTORIC [CH. his feet broad-hoofed and slender ; long and curly are his mane and tail. Down his broad forehead hang heavy curls of hair." " That was the one chariot which the host of the horses of the chariots of Ulster could not follow on account of the swiftness and speed of the chariot and of the chariot-chief who sat in it." The horses were guided by " two firm-plaited yellow reins," which shows that only a single rein was used for the pair. From the results obtained in our previous investigations the reader will at once see that Cuchulainn's horses were well bred, the result of crossing the European- Asiatic horse with Libyan blood, as horses of grey and black have been proved to be, whether found among the Turcomans of Central Asia, in Asia Minor, Mesopotamia, Syria, Nubia, the Barbary States, Spain, or France. How came this breed into Ireland ? When we remember that we traced the black horse of Ariege through Auvergne, Central France, along the Loire to Brittany, and when we remember that so great an authority as M. Sanson holds that there is a close kinship between the Breton pony and the ponies of Ireland and Great Britain, and when it is likewise borne in mind that all authorities are agreed in de- riving the pedigree of the horses of Ariege, Auvergne, Morvan, and the Breton ponies from an ' Oriental,' i.e. Libyan, origin, there is at once a strong presumption that Cuchulainn's black steed was of Spanish or Gaulish blood. Again, when we remember that the little horses of Provence, which are grey in colour, are held by the best French authorities to be derived from Libyan blood at least a century before our era, and we also consider the fact that the Percheron, the most famous half-bred horse of France, is thought to have been already in the valley of the Seine from an early period, it seems equally probable that Cuchulainn's grey steed was also derived from either Northern Spain or Gaul. A recent discovery confirms this argument. In 1903 Mr G. Coffey^ found in the centre of a small tumulus near Loughrea, co. Galway, a cremated burial " on the level of the old surface of the ground. It rested on a rude block of stone, and consisted of an almost plain urn ^ Proc. Itoy. Irish Acad. Vol. xxv. sec. C, no. 2, p. 14. Ill] AND HISTORIC TIMES 399 inverted over the burnt bones. Directly above the bones lay the skeleton of a woman with its head to the west, and beside it were the remains of a small horse, which lay on its left side with the head to the west," and which had been probably buried along with the human body. The woman was probably a slave killed to be the guardian of her master's graved Dr Scharff^, who examined the horse bones, states that they all belonged to one individual — a seven-year old stallion of small size. " To judge from the length of the humerus, radius, and metacarpal, the forelimb belonged to a small race of horse or pony. The measurements of these bones are somewhat larger than those given by Dr Marek of an Exmoor pony, but they Fio. 119. The O'Donovau Shield; Skibbereen'*. are almost identical with those of the largest of the horses found at La Tene," which I have identified with the Ligurian gimii, the ancestors of the grey Camargue. As cremation only came late into Ireland and was never general, being chiefly used by the chieftain class, and as it was practised by the Gauls in Caesar's day and by the Belgic tribes of Kent^ there is a high probability that it came into Ireland with the La Tene, or ' late Celtic ' culture. It is certain that the Loughrea tumulus is pre-Christian and it probably belongs to the period when the La Tene culture came into Ireland. Accordingly 1 Kidgeway, Early Age of Greece, Vol. i. pp. 497, 505. - Proc. Roy. Irish Acad. loc. cit. ^ Ridgeway, op. cit. Vol. i. p. 462. ^ Id., Vol. I. pp. 503, 505. 400 THE HORSES OF PftEHISTORIC [CH. the horse found in the barrow whose measurements are iden- tical with those of the Helveto-Gallic horses probably either sprung from imported ancestors or had been itself imported. Nor is there any difficulty in proving that constant trade with France and the Peninsula was carried on in the days when the Cuchulainn Saga was composed. It is absolutely certain that already in the early centuries of the Christian era there existed a very considerable trade between Ireland and the Continent. According to the Confession of St Patrick, the apostle of Ireland sailed from that island to France in a ship whose cargo com- prised those famous Celtic dogs\ which in the time of Augustus (29 B.C. — 14 A.D.) also formed part of the exports of Britain""' to the Continent, where they were highly valued both for war (cf. p. 419) and the chase, more especially as the Greeks and other Mediterranean peoples possessed few dogs that could run down a hare until they came to know the Celtic vertragus in the third century B.C.^ Indeed Arrian — in what appears to be the oldest reference to the sport of coursing — calls special attention to the fact that some of the Celts who did not live by the chase, but hunted for sport, did not use nets (as did the Greeks)-, but gave the hare fair play. If dogs were exported from Ireland, there is no reason why the Irish should not have imported horses from France and Spain, for the Irish legends also point to early intercourse between the latter country and Ireland. Thus in The Wooing of Emer^ — the same poem which contains the de- scription of Cuchulainn's steeds — we read how Emer's father, Forgall the Wily, " went to Emain Macha (the modern Navan Rath) in the garb of a foreigner, as it were an embassy from the king of the Gauls, that had come to confer with Conchobar with an offering to him of golden treasures, and the wine of Gaul, and all sorts of good things besides." This passage occurs in the older redaction of the poem, which is assigned to the eighth century by Prof. Kuno Meyer. From this it is perfectly clear that when the poem was composed there was constant intercourse between Ireland and Gaul. Elsewhere in the sagas we hear of Irishmen going to the coast of Spain and Portugal. ' Tripartite Life of St Patrick, ed. Whitley Stokes, pt. ii. pp. 362-3. - Strabo, 166. ^ Arrian, Cijnegeticus, 3. Ill] AND HISTORIC TIMES 401 The very accurate knowledge of the geography of Ireland shown by Ptolemy (120 A.D.), puts it beyond doubt that traders from Britain and the Continent frequented the mouths of the Irish rivers in the first century after Christ. It is thus possible that before the beginning of our era the Irish had obtained some horses superior to those which they were then using, the latter being doubtless similar to those diminutive animals, still used, as we have seen, by the tribes of northern Britain to draw their war chariots at least two centuries after the Roman conquest. There is no doubt that the trade between Gaul and Ireland and the British Isles in the first century B.C. was almost entirely in the hands of the Veneti, a tribe of Armorica (Brittany), who excelled in shipbuilding and seacraft, and whom Caesar had great difficulty in reducing to subjection^ They seem to have had the complete control of the Channel, and, as I have shown elsewhere^ it was these people who were carrying on the tin trade between Cornwall, the Isle of Wight, and the mouth of the Loire when Pytheas of Marseilles made his memorable voyage into our northern seas about 350 B.C., and it was probably the same people who gave the Britons of Kent the news that Julius Caesar was preparing to invade their island home. But, as it was probably to the mouth of the Loire that St Patrick sailed in comj^any with the wolf-hounds, in the fifth century A.D., and as it was to the same haven that the tin of Cornwall was carried when Posidonius visited Britain about 90 B.C., it is exceedingly probable that it also was the port for the trade between Ireland and France at the time when the Cuchulainn Saga first took shape. But as, according to Sanson, the Breton pony, so similar to the ponies of Britain and Ireland, was already in Brittany before Roman times, and as we have proved the Libyan origin of that animal, it is highly probable that superior horses of Libyan blood, such as those of Cuchulainn, were imported into Ireland from Brittany. We have thus a satisfactorj^ explanation of the origin 1 B. G., V. 8. 1. - Eidgeway, ' Greek Trade Routes to Britain,' Folklore Jour. Vol. i. R. H. 26 402 THE HORSES OF PREHISTORIC [CH. of the fine-headed horses of Libyan type, whose skulls have been discovered by Mr Coffey. From this it would follow that Sanson's E. c. hihernicus is not a separate species, or variety of horse, but only a derivative from the Libyan. In the black Irish Hobbie that beat the best Barbs of the day, in the black Connemara ponies, in the black Hebridean ponies without hind callosities, and in the dark-coloured High- land ponies we seem to have lineal descendants from the black horses, like that of Cuchulainn, which were almost certainly sprung from the same stock as the black horses of Brittany, Auvergne, and Ariege. On the other hand, as there is good evidence of the presence of grey horses in Northern Spain at least two centuries B.C. (p. 256), and it has been made probable that the grey Camargues of Provence are descended from the ginni, which were identical with the horses of the Gauls of the La Tene period, and as there is also good reason to believe that the grey Percheron is of great antiquity in France, and that both it and the Camargue owe their excellence to Libyan blood (pp. 822, 325), we may reasonably conclude that Cuchulainn's grey steed was, like its black comrade, of Libyan ancestry derived through Gaul. These imported horses would be larger and stronger than the native Irish horses, whose own ancestors had been brought to Ireland at an earlier period. As already stated, the Irish horses of the present day are the best in the world, and as it is in horses of the hunter type that Ireland especially excels, and as these are certainly not sprung from the unalloyed Irish Hobby, it is most desirable to trace the history of modern Irish horses. Without doubt the best representatives of the old Irish Hobbies at the present hour are those ponies of Connemara which have not been adulterated by the Clydesdale blood introduced by Scotch farmers during the second half of the last century, or by the Hackney stallions which, by a blunder begotten of historical ignorance, were introduced by the Congested Districts Board some years ago. As we have already seen, it has been generally held that the Irish Hobby was the outcome of Spanish blood (either derived from horses saved from the wrecked ships of the Armada) or directly through England (as believed by Sir W. Gilbey) or Ill] AND HISTORIC TIMES 403 directly from Spain in Tudor times. Accordingly it has been the fashion to recognise an Andalusian type in the Connemara ponies, just as it is popularly supposed that a well-known type amongst the peasantry of the west and south of Ireland is due to Spanish blood derived from Spaniards escaped from the Armada, though history shows that, with very few exceptions, these unfortunates were despatched on the shore and left no time to perpetuate their race. In the chestnut colour, which Fig. 120. Yellow-dun Connemara Pony (so-called 'Andalusian' type). some have taken to be the most usual, writers have seen further proof of an Andalusian origin. But the Connemara ponies of to-day are dun, white, grey, black, chestnut, and bay, the most typical specimens of the so-called Andalusian type being yellow-dun (Figs. 120, 121)\ whilst the most common 1 Ewart, Journal of the Department of Agriculture for Ireland, Nov. 1900, pp. 181 sqq. ; 'The Ponies of Connemara,' in Ireland, Industrial a)id Agricultural, 1902, pp. 332 sqq. I am indebted to the Irish Department of Agriculture for 26—2 Ill] AND HISTORIC TIMES 405 colour seems to be grey (Fig. 122). Though the old Connemara or 'Andalusian' type resembles in many respects ponies still to be seen in Andalusia, Prof. Ewart has pointed out that they bear a still more striking resemblance to some of the New Forest ponies (Fig. 123), which, as we have seen, are probably related to the Breton ponies and thus have in their veins Libyan blood derived through Spain and France (pp. 322-4), ii^HI Fig. 122. Light-grey Cunneuiaia Pony (so-called ' Andalusian ' type). and possibly also that of the ' Celtic ' pony. They vary from 13 to 13'2 hands : some are black, others grey or chestnut, but the most characteristic specimens are of a yellow-dun colour. Some of these are very fine in the bone with the long pasterns often seen in the New Forest ponies. The ears the loan of all the blocks of the illustrations of Irish horses here shown (through the kindness of Mr T. P. Gill, Under-Secretary for Agriculture and Technical Education, Ireland). 406 THE HORSES OF PREHISTORIC [CH. measure about five and a half inches ; they have a tendency to be roach-backed, as is sometimes the case with Barbs ; when compared with a Spanish jennet or with a Barb, these ponies are relatively shorter in the neck and legs, deeper in the ribs, shorter in the ears, and provided with more powerful jaws. "The Connemara is a slightly altered Barb on pony's legs\" After having followed the pedigree of the Irish Hobby in the previous pages, the reader will not be surprised to find the Fig. 123. New Forest Ponies-. resemblance and, at the same time, the difference between the Connemara pony and the North African horse. The prevalence of yellow ponies in the west of Ireland is due to the fact that the ancient horses, whether ' Celtic ponies ' or the heavy-built old European horses, were of a dun colour, and were sufficiently prepotent to transmit their colour to a considerable proportion of their descendants, as has been done by the Asiatic element in the horses of Kattywar (p. 139) and by the ancient Spanish ^ Ewart, loc. cit. - The illustration is from a photograph taken by Mr G. E. Low. Ill] AND HISTORIC TIMES 407 horses to their descendants in the sierras of Spain. The yellow- dun Connemara ponies are highly prized in some districts, not only because they are hardy and easily kept, but also because in staying power and vitality they are more like mules than pure- bred horses. Others of these ponies look more like Syrian Fig. 124. Light-grey Counemara Filly (so-called 'Eastern' type). Arabs than Aiidalusians, and they are frequently grey (Fig. 124). Intelligence, good-temper and courage seem to characterise the majority of the Connemara ponies. In the district of Clifden are bred ponies but little larger than the old Connemara, yet of very different build. The head is beautifully moulded and the face is very intelligent, the ribs are well arched, the shoulders 408 THE HORSES OF PREHISTORIC [CH. good, the loins and hind-quarters well developed, while the short legs stand an enormous amount of wear. A typical specimen of this strain measures 13"2 hands, has ears Arab-like in form measuring six and a half inches, and is of a grey colour (Fig. 12.5). It is quite possible that in this type Ave have a blend between the old European horses of the Solutre type, the Libyan Fig. 1'2o. Connemara Pony ; Clifden district. horse and 'Celtic' pony, the first mentioned giving it its char- acteristic strength of leg. There are also Connemara ponies which resemble more the Irish hunter type, and which probably have been crossed with large horses from Roscommon or from further east. Some of these make good hunters (Fig. 127). Finally there are ponies sprung from the crossing of the native mares with Clydesdale blood, which comes out in their hairy fetlocks, their small heads, and occasionally very bald faces. in] AND HISTORIC TIMES 409 In the so-called 'Andalusian' and 'Eastern' types we may recognise the descendants of the old Irish Hobbies, which in their turn were descended from the black and grey horses im- ported from Gaul and which no doubt had blended very largely with the ponies previously introduced. We have seen that Connemara ponies not unfrequently lack hock callosities (p. 18), a fact which may point to their 'Celtic' ancestry, though, since many North African horses have a similar characteristic, the Fig. 126. Connemara Gelding of larger type; Cashel. absence of the hind chestnuts in Connemara ponies may be due rather to the Libyan strain. In the stout-legged ponies of the Clifden district we have animals which have probably in their veins more of the blood of the old thick-set European horses of the Solutre type, and we have found at least two other classes of Connemara ponies which are still more horse-like in their conformation : of these, one is due to the recent admixture of Clydesdale blood, whilst the others resemble the Irish hunter, and in fact some- 410 THE HORSES OF PREHISTORIC [CH. times are used as hunters, and we saw that these probably result from blending the large horses of Roscommon or of the east of Ireland with the native ponies. This gradual infiltration of heavier strains shows us how the old Irish Hobby disappeared by degrees from all the eastern, north-eastern and south- eastern parts of Ireland, finally surviving only in the extreme west, and even there hardly quite pure. In other words, in the l^f^ ^ .'. 14,"-"":. ,^^ ^ 2J . ■'''-■■- ''■ BmHmp''. Fig. 127. Connemara Pony used as a hunter. parts of Ireland which first passed under English influence the Hobb}' gave place to or was contaminated by horses of a larger type brought over from Great Britain. This is a fact of great importance, for it offers the true explanation of the origin of the Irish cart-horses (Fig. 128), the crossing of which with thorough- bred stallions since the eighteenth century has produced the Irish hunters. It is not improbable that there were practically no horses save those of the Hobby type in Ireland until after Ill] AND HISTORIC TIMES 411 412 THE HORSES OF PREHISTORIC [CH. the Norman invasion in 1172, but there can be no doubt that from that time onwards Ireland has never been without horses of a large size, at least in the English Pale. The passage cited above (p. 389) from Giraldus Cambrensis shows that the Nor- mans brought with them to Ireland the same breed of large horses which their ancestors had carried into England, and on which they charged the Saxons at Hastings. Nor can it be doubted that similar 'great horses' were from time to time imported as well as bred in Ireland, for the Nor- mans, though in many respects becoming ipsis Hihernis Hiher- niores, retained their own method of warfare, and consequently required horses of large size to mount their men-at-arms. Thus when the Pale was troubled by an irruption of the O'Byrnes and O'Moores in 1372, who burned the priory of Athy, John Colton\ the first Master of Gonville Hall (now Gonville and Caius College), and successively Dean of St Patrick's, Chan- cellor of Ireland, and Archbishop of Armagh, raised a force of 26 knights and a large body of men-at-arms, and fell upon the Irish and defeated them with great slaughter. Even within the then narrow limits of the Pale there must have been a con- siderable number of horses capable of carrying heavily-armed men. With the gradual extension of the area occupied by Norman, and later English settlers, the 'great horses ' must have spread likewise and the native Hobbies must have con- stantly been crossed with the larger strain. The same process continued under the Tudors and the Stuarts. A single instance will suffice. Dr Winter, an Englishman, and a graduate of Emmanuel College, Cambridge, who became Provost of Trinity College, Dublin (1652 — 1660), was a great lover of horses and had brought over horses of peculiar merit from England, some of Avhich were stolen from him by the Irish army on one of his journeys with Cromwell's Commissioners^ This loss was amply compensated by substantial grants of land in King's Co., and 1 J. Venn, Hutortj of Gonville and Caius College, Vol. iii. p. 9. Bishop Keeves (Irish Arch. Soc. 1850) has published Colton's account of his Visitation of the Diocese of Derry, 1397, with a valuable introduction. 2 J. P. Mahaffy, An Epoch of Irish History.- Trinity College, Dublin, Its Foundation and Early Fortunes (1903), p. 301. Ill] AND HISTORIC TIMES 413 also at Agher, co. Meath, which his direct descendants still enjoy and where a very large breed of horses said to be de- scended from the Provost's own animals is still maintained^ It is therefore probable that Dr Winter's horses were the 'great' English horses of his own clay (p. 366). This example suffices to show how in Meath and Westmeath the native Irish horses were gradually saturated with the blood of the large English horses, from which the Shires and Suffolk Punches are de- scended. It is a well-known fact that horses whose ancestors have long been bred in Ireland are distinguished for their great de- velopment of bone, and for their clean, flat, hard legs, free from the spongy softness of bone so characteristic of British horses. As the infiltration of the blood of the English 'great horse' had thus gone on slowly for a very long period, the Irish horses of this mixed strain, owing to the extraordinary effects of the Irish soil and climate, did not inherit the softness and flabbiness found in the progeny of the Shires and Suffolk Punches intro- duced towards the end of the eighteenth century, and of the Clydesdales brought in about half a century later. As the English draught-horses imported towards the end of the eighteenth century produced an unsatisfactory progeny, no further attempts were made to improve the breed in this wa}^ The heavier class of native animals, the genesis of which I have just sketched, produced a better type of horse for the needs of Irish farmers. The?;e animals made good roadsters and serviceable harness horses, and, though too coarse for hunting, they had a high spirit and a natural turn for jumping, inherited doubtless from tiieir Hobby ancestors. It was the crossing of such mares with the thoroughbred (Fig. 129) that produced the Irish hunters. By the end of the seventeenth century horses of so-called 'Oriental' blood had reached Ireland, for example the Byerley Turk, brought to Ireland by his owner, Captain Byerley, who served in the army of William III, and 1 Mahaffy, loc. cit. My old friend Eev. T. T. Gray, Senior Fellow of Trinity College, Dublin, of Cam Park, co. Westmeath, tells me that these animals, on which he himself has ridden, are of great size, and that the breed seems certainly derived from the Provost's horses. 414 THE HORSES OF PREHISTORIC [CH. probably had that famous horse at the battle of the Boyne. The mixture of such horses with the native mares laid the foundation of the Irish thoroughbred strain, and fresh thorough- bred blood was continually being introduced, until by the middle of the eighteenth century there were in Ireland up- wards of one hundred imported stallions. From that time onwards the history of the Irish thoroughbred is bound up with that of the English racing stock, but the Irish thorough- bred has continued to preserve traits derived doubtless from the Irish mares first mated with the imported sires of Libyan blood. As the Irish hunters (Fig. 130) are the progeny of the Irish thoroughbred and the Irish cart-mares, on the number and quality of the latter must depend one of the most precious productions of Ireland. The hunters bred in Ros- common are especially noted for their size and great develop- ment of bone, though they are occasionally coarse, whilst, though the horses reared on the rich grass lands of Meath and Westmeath are not more fleshy and are scarcely equal in bone, they are, however, more shapely than those of Ros- common. The increase of tillage during the French war at the close of the eighteenth and the beginning of the nineteenth century gave a great impetus to the breeding of agricultural horses, as the bullock rapidly ceased to be the tiller of the soil, although in certain localities he has lingered on to our own day, both for the plough and the cart. The growth in the number of cart-mares naturally increased the production of weight- carrying hunters, but unfortunately since the repeal of the Corn Laws the shrinkage of Irish tillage has gone on steadily and there has been a corresponding diminution in the number and quality of Irish cart-mares. The larger farmers were the first to lay down their lands in permanent grass, and gradually the cart-mares only remained in the hands of smaller farmers who continued to till their ground, but, as was naturally to be expected, the mares kept by the latter were, as a rule, not of so good a quality as those formerly maintained by the wealthy farmers. The agricultural depression of the last quarter of a century has too often compelled the small farmers to sell their Ill] AND HISTORIC TIMES 415 416 THE HORSES OF PREHISTORIC [CH. best fillies, and at the present moment Ireland possesses only some very old and degenerate specimens of an invaluable breeds the loss of which would be little short of a national disaster. Every year the demand for Irish hunters becomes greater and their value increases, and no matter what may be the fate of other classes of horses owing to the competition of motor-cars, the high-class Irish hunter is not likely to suffer through the rivalry of any mechanical contrivance. It is earnestly to be hoped that in view of the fatal damage done to the fine old breeds of France by the unwise admixture of foreign strains, and the like injury wrought to Irish horses by contamination with Shire, Suffolk Punch, Clydesdale, Cleveland Bay, and Hackney blood, no further rash experiments of this kind will be tried, but that steps will be taken to rebuild the old breed of Irish cart-horses by the careful selection for stud purposes of the best of those which still survive. Let us now return to the horses of Scandinavia, the Faroes, and Iceland. We saw above (p. 18) that from the absence of hock callosities, the presence of short hair on the upper part of the tail, the shortness of the ears (in which they differ from Arabs), and the fineness of the head and limbs in certain ponies found in Connemara, the North of Ireland, the Outer Hebrides, Faroe Isles, and Iceland, Prof. Ewart was led to his doctrine that a separate variety of horse, which he names E. c. celticus, had survived from the Palaeolithic period in the north-west of Europe. When and whence horses first reached Iceland we have ample evidence. After Harold Fairhair had made himself sole king in Norway in 870 A.D., many of the turbulent Norwegian jarls preferred exile to submission. Among these were Ingolf and Leif, who set forth to Iceland in 871 a.d., and finally settled there in 874 a.d. Three years later Kettle Haeng led a further body of settlers to the same island. Gradually Harold Fairhair began to extend his authority and to root out the Vikings from the Western Isles, and after the fall of Thorstein the Red in Scotland there was a rush of settlers from the British Isles to Iceland (890—900). 1 Joiirn. Dept. of Agriculture (Ireland), Oct. 1904, pp. 25 sqq. m] AND HISTORIC TIMES 417 P4 27 418 THE HORSES OF PREHISTORIC [CH. When Ingolf and his folk settled in Iceland that island had never known any human inhabitants save a few Irish anchorites, who cannot be deemed to have contributed anything towards the population, or to the domestic animals found there later on. If these ecclesiastics had brought horses with them — a thing in itself not very likely — these animals would have been Irish or Hebridean in origin. On the other hand the Norse settlers brought ^vith them their families, house- hold goods, and domestic animals, amongst which horses were almost certainly included. It is important to bear in mind that although all these colonists were Norsemen, more than half of them, as is definitely proved by the Landnamaboc — an ancient record of the names, ancestors, and holdings of the early settlers — had been living in the British Isles before they removed to Iceland, and that only a minority went direct from Scandinavia. It is therefore highly probable that the original stock of the ponies of Iceland and the Faroes came partly from Ireland, and partly from Norwa}^ but as has been well pointed out the proportion that came from the former country was probably greater, inasmuch as some of the settlers from Scan- dinavia did not go directly to Iceland, but first went and sojourned for a while in the Western Isles. It is therefore not improbable that if these colonists brought ponies with them, the latter would be of the Hebridean or Irish breeds. That the ponies of the Hebrides had been brought from Ireland by the Irish monks who settled at loua (lona) is ren- dered almost certain by a famous passage in Adamnan's Life of St Columba'^. In the evening of his life the old man, worn out with age, went about lona in a cart to visit the brethren who were at work on the other side of the island. On the day he died, he and his attendant Diormit " went to bless the barn which was near at hand, and after having blessed two heaps of winnowed corn that lay therein, the saint left the barn, and in going back to the monastery rested half-way at a place where a cross, which was afterwards erected, and is standing to this day, fixed into a mill-stone, may be observed on the roadside. 1 Adamnan's Life of St Columha (Reeves' ed.), pp. 95-fi. Ill] AND HISTORIC TIMES 419 While the saint, bowed down with old age, sat there to rest a little, behold, there came up to him a white pack-horse, the same that used, as a willing servant, to carry the milk- vessels from the cow-shed to the monastery. It came up to the saint and, strange to say, laid its head on his bosom, inspired, I believe, by God to do so, as each animal is gifted with the knowledge of things according to the will of the Creator : and knowing that its master was soon about to leave it, and that it would see him no more, began to utter plaintive cries, and, like a human being, to shed copious tears on the saint's bosom, frothing and greatly wailing. The attendant seeing this began to drive the weeping mourner away, but the saint forbade him, saying : ' Let it alone, as it is so fond of me — let it pour out its bitter grief into my bosom.' Then the saint blessed the work- horse, which turned away from him in sadness." From what we have learned about the sanctity of white horses, it may not be rash to suggest that the prescience (p. 114) of St Columba's horse was not unconnected with its white colour in the mind of good Abbot Adamnan. Moreover it is clear from the sagas that there was constant intercourse between Ireland and Iceland in the tenth and eleventh centuries, and that Irish slaves and Irish wolf-dogs were known in the latter island. Thus Otkell of Kirby had an Irish thrall named Malcolm \ whom his brother Hallbjorn the White had brought out to Iceland. Again, when Gunnar of Lithend, then in sore peril, was parting from his friend Olaf the Peacock, the latter said : " I will give thee three things of price, a gold ring, and a cloak, which Moorkjardtan the Erse king owned, and a hound that was given me in Ireland ; he is big, and no worse follower than a sturdy man. Besides, it is part of his nature that he has man's wit, and he will bay at every man, whom he knows is thy foe, but never at thy friends ; he can see too in any man's face, whether he means thee well or ill, and he will lay down his life to be true to thee." After that he said to the hound, " Now shalt thou follow Gunnar, and do him all the service thou canst." The hound went at once to 1 The Saga of Burnt Njal, xlvii. 420 THE HORSES OF PREHISTORIC [CH. Gunnar, and laid himself down at his feet^ It will be re- membered that Otkell of Kirby was the owner of the two dun horses with stripes on their backs — the two best in all the country. If slaves and dogs were brought from Ireland, it is not unreasonable to suppose that horses also would occasionally form part of the cargo of the homeward-bound Icelander. The reader will also remember that two other horses of exceptional merit are described in Burnt Njal — the chestnut stallion of Starkad and Gunnar's brown one. But our investigations have made it clear that the colours chestnut and brown in horses are a sure indication of a large proportion of Libyan blood. It is also to be noted that in Otkell's dun horses with dorsal stripes we may have an outcome of Libyan blood, similar to that so well exemplified in the horses of Kattywar — but with this feature we shall have to deal at greater length in the next chapter. We have had good evidence for believing that there was a considerable element of Libyan blood, derived through Spain and France, in the best horses of Ireland, the country from which a large proportion of the original stock of Iceland was derived. It will also be remembered that, according to Sanson, some North African horses lack hock callosities, the absence of which often characterises Iceland and Faroe ponies. It cannot, therefore, be held that the fine heads and limbs, small joints and the absence of hock callosities in the ponies of Iceland and the Faroes, as well as those of the Outer Hebrides and some found in Ireland, are solely due to these animals being descended from a small horse which inhabited the north-west of Europe in late palaeolithic times. But as it is possible that the ' Celtic ' pony in the north-west of Europe and the small Libyan horse of North Africa are both descended from the small horse of the Brighton ' elephant bed,' the Libyan blood which passed into Ireland through Spain and France may well have been there combined not only with the coarse European type, but also with a closer relative, and accordingly not a few of those (especially the bay, black, ^ The Saga of Burnt Njal, lxix. Ill] AND HISTORIC TIMES 421 brown, and chestnut) ' Celtic ' ponies which lack hock callosities may derive this feature from a double line of ancestry. On the other hand, it is plain, as we have shown (p. 346), from the mention of black horses in the early literature that the best horses of Scandinavia were saturated with Libyan blood long before the colonisation of Iceland. This blood was probably derived through Germany, where the Frisians already possessed, in the fourth century, the famous breed of black horses still known by their name, and whose pedigree can be traced back to the horses owned by the Tencteri in the first century A.d., and still earlier to the horses imported by the Gauls from Mediterranean lands. It is therefore clear that in the best horses brought from Norway to Iceland and the Faroes there may have been a fair proportion of Libyan blood. The presence of black, brown, chestnut, or striped dun animals in Iceland and also in the Faroes may therefore be due not only to the horses brought from Ireland, but also to those direct from Scandinavia. But Prof Ewart informs me that, as far as he has seen, " Iceland and Faroe ponies in their ears, heads, manes, and tails are never like the large Barbs and Arabs, though Hebridean horses in which there is Spanish blood (introduced in the eighteenth century) frequently are like long-eared Arabs." But it by no means follows that all the horses brought by the first colonists, or even those which came in later, were of a superior breed. On the contrary, the descriptions of the black and grey steeds of Cuchulainn, of Otkell's dun horses with dorsal stripes, and of the chestnut and brown horses of Starkad and Gunnar, all of which are commended for their peculiar ex- cellence, show clearly that there were plenty of inferior animals, either of the old thick-set, heavy-limbed, large-jointed European-Asiatic horse, or possibly derived from an ancient horse of lighter build — Ewart's 'Celtic' pony itself, the de- scendant of the slender-built horse whose remains are found in the Brighton ' elephant bed.' That ponies with large heads and of a heavy build were familiar in Iceland in medieval times is proved by an old picture preserved in Iceland and reproduced by Bruun. But as 422 THE HORSES OF PREHISTORIC [CH. it is clear that the chestnut and brown ponies of Iceland had certainly Libyan blood, and as, on the other hand, the heavy- headed ponies of the picture seem almost certainly sprung from the old heavy type of European horse, it may be urged that it is difficult at first sight to show that any Icelandic or Faroe ponies owe their characteristics to the slender- built horse of palaeolithic times, though it is quite possible that in some Icelandic ponies the presence of short hair on the upper part of the tail, instead of the long hairs so characteristic of the Libyan stock, is a feature derived from this ancient light-built horse. But it seems difficult to derive the 'Celtic' 'tail-lock ' (p. 18) from the Libyan infusion, whilst the fact that Icelandic and Faroe ponies more frequently lack hock callosities than is as yet proved to be the case with North African horses, indicates that this feature may not be wholly derived from Libyan ancestors, but may be due in some measure to an old ' Celtic ' pony. We have now passed in review all the chief breeds of horses of prehistoric and historical times, and the evidence has led us to the following conclusions : (1) that the horses of Upper Europe and Upper Asia were always dun or white, the vast majority of them being thick-set, slow animals, though in the north-west of Europe there was the ' Celtic ' pony, an animal of much lighter build, more elegant shape, and probably greater speed; (2) that these coarse, thick-set horses of Upper Asia and Upper Europe have continually kept making their way into the regions lying south of the great mountain chains which cross the Asia-European Continent ; (8) that these horses first made their appearance in Babylonia not long before 1.500 B.C., and about the same time in Palestine and Greece ; (4) that the Arabs of the Peninsula did not become possessors of this or any other horse until after the Christian era ; (5) that at some date not long prior to 1500 B.C. the kings of the XVillth Egyptian dynasty were already in possession of horses of a type com- pletely different in shape, colour and manner of carrying their tails, from the Asiatic horses, though closely resembling in these particulars the best Arabian and Barbary horses of modern times ; (6) that these horses are regularly depicted on Egyptian Ill] AND HISTORIC TIMES 423 monuments as of a dark colour ; (7) that by B.C. 1000, and we know not how much earlier, these horses were so highly esteemed by the horse-driving peoples of Western Asia that King Solomon imported them from Egypt at a great price, not only for himself but also for all the kings of Syria and for the kings of the Hittites ; (8) that this bay breed was already in the Troad before B.C. 1000, and was there regarded as not only divine, but as quite different in origin from the dun and white horses of Greece and Thrace ; (9) that it was even then distinguished by a star in its forehead, a feature which to this day is characteristic of the pure-bred Arab horses and their deriva- tives ; (10) that already these horses were swifter than all others ; (11) that the Greek legends regarded the horses bred near the Atlantic as the swiftest; (12) that Pegasus, the fabled winged steed, was said to have been born in Libya ; (13) that the Libyan horses were the swiftest in the Roman Circus ; (14) that these African horses have been continually sought for by the peoples of Asia and Europe for the purpose of improving the quality of their own indigenous horses, and that consequently their blood has influenced the horses of Asia as far as China and those of Europe as far as Ireland ; (15) that it is to the blending of the bay horses of North Africa with the dun or white horses of Asia and Europe in varying degrees that are due not only all the improved breeds of the world, but also the various shades of grey, rufous-grey, roan, skewbalds, piebalds, chestnut, brown, and black ; (16) that the white bracelets or stockings so frequent not only in bay, but also in black horses, are due to the Libyan strain, and that the white star or blaze on the forehead frequently found in domestic horses is due to the same cause ; (17) that the peculiar blend of the two strains which results in black often produces great strength combined with fair speed; (18) and that for this reason the black horse through the ages has been especially valued for war from Turkestan to England, and from Morocco to Sweden ; (19) we have likewise seen that the horse has everywhere been driven under chariots before he was ridden, and that it was the Libyans in Africa and the Turko-Tartaric tribes in Asia who first began to ride habitually on horseback, 424 PREHISTORIC AND HISTORIC HORSES [CH. Ill and that the Greeks }3robably first learned this art from North Africa, Finally, we have learned a lesson that all the races which have in their turn held the mastery in Asia, Africa, and Europe, have owed the extension of their power, or the preserva- tion of their liberty, to the possession of horses, whether they were Egyptians, Syrians. Libyans, Medes, Persians, Scythians, Macedonians, Carthaginians, Numidians ; that the lack of horses till after the conquest of Gaul was the great weakness of Rome ; that the acquisition of the horse by the Arabs was a main factor in the spreading of Islam ; and that had not the Franks owned good horses by 732 a.d. Western Europe might have been enslaved by the Saracens ; that the possession of horses enabled the Normans to conquer at Hastings, and the possession of great war-horses was the sure means of preserving one's own country or conquering that of others throughout the Middle Ages ; even when armour was discarded the new cavalry mounted on light horses became an engine of war more formidable than any yet known; whilst Marlborough's great victories were largely due to his cavalry. With the improve- ment of fire-arms and the ability of infantry to resist cavalry it seemed as though the days of the horse-soldier were over, but the recent war in South Africa has shown that in the future struggles of the nations mounted infantry are likely to play an all-important part. Accordingly, though the place of the horse under the carriage, the omnibus, the waggon, and even the plough itself, may be taken by automobile engines, never in the history of the world was there greater need of horses to draw artillery and to carry infantry. It is therefore imperative that this country should not shut its eyes to the need of breeding horses suitable for war, and that careful steps should be taken to preserve our good breeds and not permit them to be contaminated and destroyed by rash experiments in breeding. CHAPTER IV. THE ORIGIN OF THE LIBYAN HORSE. The King, sir, hath wagered with him six Barbaiy horses. Shakespeare, Hamlet^ v. 2. We have now briefly surveyed the history of all the chief breeds of horses — modern, medieval, and ancient — and we have been led steadily to the conclusion that the best horses are sprung wholly or in part from a North African stock, the lightly built horses excelling in speed being pure or almost pure derivatives, whilst the large heavy cart-horse breed owes much to the same blood when blended with the coarse, large-headed, short-necked stock. Of the origin of the latter we have full knowledge, for we have seen that it is the indigenous horse of Upper Asia and Europe, and we have found a wild species of Equidae — Prejvalsky horse or the tarpan — existing down to our own time in Central Asia and Eastern Europe; it has been shown not unlikely that the ordinary Equus cahallvs of Europe and Asia and the Prejvalsky horse have sprung from a common immediate ancestor, or, what is less likely, that the former has developed out of the latter. But it is altogether different with the thoroughbred stock, inasmuch as Africa does not possess any wild horses such as the tarpan, although she is the mother of the Abyssinian ass, and at least three species of zebras (not including the recently lost quagga). We have now to face the problem of the origin of this Libyan horse, which I shall provisionally term Equus caballus libycus. We have seen that it has been commonly held that all our domestic horses have come from a single stock, a view y 426 THE ORIGIN OF THE LIBYAN HORSE [CH. largely based on the fact that all interbreed with fertility. But, as I have already pointed out, this is no criterion, since animals which are regarded as distinct species, such as the dog, wolf, and jackal amongst carnivores, and the common ox, the zebu, and grunting yak amongst herbivores, also interbreed and produce fertile offspring'. As I have shown that M. Sanson's E. c. cifriccmtis, which he supposed to be a primeval stock of North-east Africa, has no historical basis, the Dongola breed, on which he principally based his argument, being merely a cross-bred animal of comparatively recent date, we must there- fore look elsewhere for the source of the Libyan horse. As it is found among the Libyan tribes from the dawn of history, it is obvious that it was either developed from the ordinary horses of Asia and Europe, which they had obtained already domesti- cated, or else it was an indigenous species which had reached Africa either through Asia or through Asia and Europe, and had been there highly specialised by its peculiar environment, and domesticated by the Libyans themselves. The first alter- native seems very improbable, since we have traced Libyans with chariots and horses up to a period almost contemporary with the first appearance of the horse on Egyptian monuments, and we have presented reasons for believing that the Egyptians, who admittedly borrowed the horse and the chariot from some other people or peoples, were using Libyan chariots about 1400 B.C., that they do not appear to have had the horse much before 1500 B.C., and that along with the light chariot with four-spoked wheels they had obtained the horse from the Libyans (p. 227). We have seen that by Homeric days it was a general belief in Greece that the swiftest horses came from the Western Ocean, and that horses of a bay colour — the ever constant livery of the Libyan horse and its derivatives — were already kn6wn in Asia 1 Report by C. W. Campbell, H.M. Consul at Wuchow, On a Journey in Mongolia, Jan. 1904, p. 36: "The yak (Bos grunniens), here (Mongolia) called sarlik, is kept in the place of cattle to a considerable extent throughout the mountainous parts of North Mongoha. Hybrids of the yak and ordinary cattle are common, and their milk is much esteemed." See also Blanford, Indian Mammalia, p. 491. IV] THE ORIGIN OF THE LIBYAN HORSE 427 Minor as far superior to the very best blood from the north, whilst by the tenth century B.C., although the Hittites and Syrians had plenty of horses, they were eager to purchase horses from Egypt at a great price, whilst in the post-Homeric myth Pegasus, the most famous of all steeds, is represented as born in the western Libyan desert. All these facts prove that before the end of the second millennium B.C. the Libyan possessed horses far superior in speed to those of Europe and Asia, and also that these horses were already distinguished by their bay colour with a star in the forehead, which characterises the Libyan stock and its posterity to this very hour. As the Egyptians did not obtain horses until the sixteenth century B.C., it is obviously impossible that the Libyans, supposing them to have obtained the horse for the first time from the Egyptians, could have developed by Homeric times a race of horse so absolutely distinct from all others bred since, even in times when men have expended much care and skill in pro- ducing artificial varieties. But as it has been pointed out (p. 209) that the Arabs, who take such pride in their horses and their pedigrees, are ignorant of the very first principles of breeding, it is most unlikely that the nomad Libyans, who never kept any account of strains, such as Al-Khamseh, or practised castration, would have produced artificially in a com- paratively short time the most wonderful breed of horses that the world has known, and with characteristics so indelibly fixed that they can permanently modify the form and colour of all other breeds. If it be objected that there is no record of wild horses in Libya, either in ancient or medieval times, and that conse- quently there never was any indigenous horse for the Libyans to domesticate, I reply that remains of fossil horses have been found in North Africa, that though at present there is not any kind of ass either striped or unstriped in that same region, yet undoubted evidence has been given that there was in that area some kind of wild ass in the time of Herodotus, that many- striped asses existed in the Great Oasis down at least to the tenth century A.D., and we know not how much later, and finally that, according to Pigafetta (p. 60), there were zebras 428 THE ORIGIN OF THE LIBYAN HORSE [CH. in Barbary in the sixteenth century. That some parts of the region lying between Egypt and the Atlantic were admirably adapted for horses is proved by the accounts of Numidian and Cyrenean horse-breeding and by the fact that the humrahs, the descendants of horses that have run vv^ild, flourish in Northern Nigeria, where they are occasionally captured and broken in. How easy it is to exterminate the Equidae and other larger mammals has unhappily been too well demonstrated in our own time by the extinction of the quagga and all but complete destruction of the mountain zebra in South Africa, and the reduction of the vast herds of the North American bison to a few hundreds, which can only maintain a precarious existence under artificial protection. There are various reasons which lead especially to the destruction of the wild horse. In the first place it has been valued as food by many tribes, who have not domesticated it ; secondly, when once a community has learned to tame and utilise the horse, the wild horse becomes a valuable prize when taken alive; thirdly, when the chief wealth of a community, such as those of the Tartars, the Gauchos, and the Pampas Indians, consists of horses, herds of wild horses are a constant nuisance and danger, as the domestic animals are often enticed away by their wild relations ; fourthly, where pasture is often scarce, as in various parts of Africa, pastoral peoples, such as the Boers and Australian stock keepers, are always anxious to exter- minate, or at least to lessen, the numbers of the large her- bivores. If it be said that it is only by the use of fire-arms that such extermination takes place and that it could not be effected by men armed merely with spears, javelins, and lassoes, and that it is not at all likely that domestic horses handicapped by a rider's weight could overtake their unweighted wild brethren, these assertions can be at once disproved by modern instances. How men on horses and without fire-arms are able to ride down wild horses and to kill or capture them is made clear by Azara', who relates that the Gauchos of the Pampas constantly 1 Quadrupeds of Paraguay, p. 14 (Eng. trans.). IV] THE ORIGIN OF THE LIBYAN HORSE 429 persecute the wild droves and drive them away from their neighbourhood to prevent the domesticated horses from joining the wild ones. " Men on horseback drive the haguals before them until they are tired ; when fresh men and fresh horses continue the chase, press upon and urge them on both sides, killing vast quantities of them with chuzos or spears, without ceasing to gallop or slackening their pace." The same observer^ points out that " all horses run swifter when mounted than when galloping loose, especially if ridden without a saddle." North America offers very valuable instances of a similar kind. Dr Richardson - states that " herds of wild horses, the off- spring of those which have escaped from the Spanish possessions in Mexico, are not uncommon on the extensive prairies that lie to the west of the Mississippi. They were once numerous on the Kootannie Lands, near the northern sources of the Columbia, on the eastern side of the Rocky Mountain ridge, but of late years they have almost been eradicated in that quarter. They are not known to exist in a wild state to the northward of the fifty-second or fifty-third parallel of latitude. The young stallions live in separate herds, being driven away by the old ones, and are easily snared by using domestic mares as a decoy. The Kootannies are acquainted with the Spanish-American mode of taking them with the lasso." Major Long^ says that the Osages hunted the wild horses, which are exceedingly fleet. " They go in large parties to the country of the Red Canadian River, where they are to be found in considerable numbers. When they discover a gang of the horses they distribute themselves into three parties, two of which take their stations at different and proper distances on the route, which by previous experience they know the horses will most probably take when endeavouring to escape. This arrangement being completed, the first party commences the pursuit in the direction of their colleagues, at whose position they at length arrive. The second party then continues the 1 Azara, op. cit. p. 29. 2 Fauna Boreali- Americana : or the Zoology of the Northern Parts of British America (1829), pp. 231-2. ■* Fauna Boreali- Americana , pp. 232-3. 430 THE ORIGIN OF THE LIBYAN HORSE [CH. cbase with fresh horses, and pursues the fugitives to the third party, which generally succeeds in so far running them down, as to noose and capture a considerable number of them. The domestic horse is an object of great value to the nomadic tribes of Indians that frequent the extensive plains of the Saskatchewan and Missouri, for they are not only useful in transporting their tents and families from place to place, but one of the highest objects of the ambition of a young Indian is to possess a good horse for the chase of the buffalo, an exercise of which they are passionately fond. To steal the horses of an adverse tribe is considered to be nearly as heroic an exploit as killing an enemy on the field of battle, and the distance to which they occasionally travel and the privations they undergo on their horse-stealing excursions are almost incredible. An Indian who owns a horse scarcely ever ventures to sleep at nightfall, but sits at his tent door with the halter in one hand and his gun in the other, the horse's fore-legs being at the same time tied together with thongs of leather. Notwith- standing all this care, however, it often happens that the hunter, suffering himself to be overpowered by sleep for only a few minutes, awakes from the noise made by the thief galloping off with the animal. The Spokans, who inhabit the country lying between the forks of the Columbia, as well as some other tribes of Indians, are fond of horse-flesh as an article of food ; and the residents at some of the Hudson Bay Company's posts on that river, are under the necessity of making it their principal article of diet." At the beginning of the eighteenth century there were many wild horses in Virginia^ and they became a great nuisance to outlying settlers, by enticing away their domestic horses. As the latter were English in origin, the wild horses were modified to some extent, though the Spanish traits still pre- dominated. Similarly the feral horses of East Victoria in Australia became a constant source of annoyance and loss to the settlers in that region, until they were at length all captured. For the following account of these animals I am indebted to my friend Dr A. W. Howitt, of Metung, Victoria, the famous ^ Wallace, The Horse of America, p. 204. IV] THE ORIGIN OF THE LIBYAN HORSE 431 Australian ethnologist^ : " Wild horses have been running in the mountainous country of East Gippsland, in which are the sources of the Buchan River, and through which the Snowy River and its tributary the Deddik River flow. To this I must add the dividing range from Omeo to Mount Kosciusko. These wild horses probably date back in places to a time antecedent to the discovery of Gippsland in 1842, On the Manero table- land which lies on the New South Wales side of the border, and extends up to Kosciusko and Kiandra, and Sunit, as also from the country to the heel of the dividing range, I have no doubt that horses escaped and became wild. Of course these have been of all kinds. On the high mountain plateau which lies between the upper Tambo River and the sources of the Buchan River I have seen horses which can be best described as dwarfed cart-horses, and probably were the descendants of light draught stock used by prospectors and miners in the early times of gold discovery — after 1850. The country they lived in is very high and cold, being covered in winter with snow, and altogether ill adapted to feral horses. In the warmer but very hilly country which lies to the east of the Snowy River in Victoria, for instance at Gatemurra, Deddik, and Tubbut, the horses were of a much better stamp, in many cases showing good breeding, partly due to the excellent stamp of the New South Wales horses of about 50 years ago, but also to the fact that a Persian horse, imported by Benjamin Boyd, of Twoford Bay, escaped and lived for many years after in the Deddik, Gatemurra, and Tubbut country. The grey horses which occurred there may be attributed to his influence. " The horses of this district were in many cases very good, being especially sure-footed, but frequently were broken down by galloping when driven over the mountainous and exceed- ingly rough country which they inhabited. I have often seen one of these ' mobs,' as they are called, coming down the mountain side when disturbed at a gallop. I remember one instance on the western side of the Snowy River, where a number were killed by running against trees, or by being crushed against them by others of the horses in their flight, 1 Dr Hewitt's letter is dated September 14; (1904). 432 THE ORIGIN OF THE LIBYAN HORSE [CH. At one time about twenty years ago the horses became such a nuisance in the cattle stations in the district that steps were taken to get rid of them. This was most successfully done by building a yard of tree trunks on one of their main tracks, which led from one valley across a range to another. This was left open at each end for a time until the horses became accustomed to pass through it. Then a slip of white calico about 18 to 24 inches wide was fastened at each side of the entrance to the yard, continued from tree to tree for a consider- able distance, diverging from the track on either hand. The wild horses being then driven from their feeding-grounds in the valley started to escape along the track across the range, and having entered between the two converging lines of calico were driven into the yard, the farther opening having been closed. The other entrance could be closed behind them. It was a usual practice to catch them and hobble them with plaited strips of green hide (salted and dried) before they were let out of the yard to prevent them escaping again and to admit of their being watched while grazing in the neigh- bourhood. Finally, they were driven to some market. I know of cases where they were driven to the coast, say at Bega, and sold at half-a-crown a head to people to repay themselves with the skin and hair, the flesh being boiled to feed pigs. I may add that so far as I remember it was not found to be possible to run in one of those ' mobs ' to a yard, as they separated, the stronger ones outstripping the weaker. I re- member another plan being tried. The ' mobs ' on the western side of the country — near the Snowy River — were all started by a number of men on horseback, who drove them at full speed to the other side of the country, not themselves going all the way, but making as much noise by shouting and cracking whips as possible. The horses having gone even as much as eight miles, were turned back by other men ; after a couple of days most of the horses had separated, and were scattered all over the country in twos and threes and even singly, many were lame, and some were then caught, but the expenditure of horseflesh in doing this was not compensated by the result. After this the method by a yard and calico ' wings ' was IV] THE ORIGIN OF THE LIBYAN HORSE 433 adopted. Much of the country is now fenced and stocked with sheep. I imagine that the feral horses are not to be found excepting in the higher, more rugged, or scrubby country outside the fencing." It is now clear that for various reasons herds of wild horses are almost certain to be exterminated in any region where the inhabitants are at all numerous and keep domestic horses. Having rejected the view that the Libyan horse is a purely artificial breed developed by the Libyans out of the ordinary domestic horses of Asia and Europe, let us examine the evidence in favour of the second alternative — that the Libyan horses were descended from wild horses which passed into North Africa from Asia or from Europe or from both. Colonel Hamilton Smith held that our domestic horses are descended from five original stirpes — the bay of Western Asia, i.e. Arab, the white, the black, the dun, and the pied — and he believed that dark stripes were a special characteristic of the dun stirps. Though Darwin rejected Hamilton Smith's doctrine of five original stocks, he found in the latter's dun-coloured stirps with a tendency to stripes the bases of his own hypothesis. From the facts that " horses in various parts of the world often have a dark stripe extending along the spine, from the mane to the tail," and that "occasionally horses are transversely barred on the legs, chiefly on the under side," and "more rarely have a distinct stripe on the shoulder, like that on the shoulder of the ass, or a broad dark patch representing a stripe," and from a consideration of the general tendency of horses to revert to a yellow-dun hue, Darwin^ was led to the conclusion that " the seven or eight species of Equidae now existing are all descended from an ancestor of a dun colour more or less striped," and elsewhere^ he argued from the results of his experience in crossing pigeons and fowls that "the progenitor of the group was striped on the legs, shoulders, face, and probably over the whole body like a zebra." He thus assumed that the first horses domesticated by man (from which he held that all ' Variation of Animals and Plants under Domestication, Vol. i. pp. 58-9. - Id., Vol. II. p. 17. R. H. 28 434 THE ORIGIN OF THE LIBYAN HORSE [CH. our domestic races, no matter what their form and colour, are descended) were of a dun colour. But he himself points out that though "all English breeds however unlike in size and appearance, and several of those in India and the Malay Archipelago, present a similar range and diversity of colour, the English race-horse, however, is said never to be dun-coloured." He thought^ that this might be explained by the fact that " as dun and cream-coloured horses are con- sidered by the Arabs worthless and fit only for Jews to ride, these tints may have been removed by long-continued se- lection." But as it has been demonstrated that the Arabs did not develope their famous horses by selection solely from the dun-coloured horses of ancient Persia or any other part of Asia, but, several centuries after Christ, obtained it from North Africa, where it was already of a dark colour at least a thousand years before the Arabs ever possessed a horse, it is obvious that Darwin's theory of the origin of the bay colour is based on a false assumption. Again, as we have shown that whenever horses of a light colour are met with in Arabia, and in other parts of Western Asia, they are always kadishes, or in other words the coarse, thick-set, slow type of Upper Asia, the contempt of the Arabs for such horses is due to the fact that horses so coloured are always inferior, and not to any dislike of their livery, since they hold, as we do, that "a good horse is never of a bad colour," and we have also seen (p. 186) that the Arabs, like many other peoples, from religious motives have a preference for white and grey horses. It is then clear that Darwin's explanation of the dark colour of the English race-horse and his progenitors does not supply a vera causa. If the bay colour of the Libyan horse has only been acquired by artificial breeding, and if, as Darwin held-, "colour is a fleeting characteristic," the descendants of such horses when they become feral and return to a state of nature ought to revert to dun colour, or at least show a tendency to do so. Yet we have seen that although the Pampas horses descended from some 1 Variation, Vol. i. p. 58. ^ Variation, Vol. i. p. 53. 1V~\ THE ORIGIN OF THE LIBYAN HORSE 435 dozen Andalusian stallions and mares turned loose in 1535, have been living under perfectly natural conditions for more than three centuries and a half, yet, as already noticed (p. 263), Azara^ states that amongst the "numerous herds of wild horses that had passed under his observation, he had never seen any colour but bay, some inclining to brown, in others more or less to red, and whenever a piebald, black, or any other coloured horse is seen, it is immediately known to have been a domesticated individual, which had made its escape or had been carried off by the wild herds." Elsewhere- he says that it is "unaccountable, the wild horses being, as I have observed, all bay, how so great a variety of colours is found amongst the tame, although black and dark-coloured are extremely scarce ; I must also mention that white, bay, and greyish-coloured horses, and above all chestnut piebalds (sabinos), the ground colour of whose coat is white, with an infinity of obscure and cinnamon-coloured specks or spots, pass for the best swimmers." I have pointed out already (p. 263) that the white, grey, and other light-coloured horses of South America are derived from the horses of Upper Spain, whilst the Pampas horses are sprung purely from those of Andalusia. But as the Andalusians are in great part of Libyan blood, and as their wild descendants of the Pampas obstinately refuse to revert to a dun colour, as they ought to do if their bay livery has simply been developed by artificial breeding, it is reasonable to infer that, whilst the Asiatic ancestors of the Libyan horses were dun-coloured, the bay colour of the Libyan horse was gradually acquired in North 1 Quadntpeds nf Paraguay, p. 14 (English trans.). As I pointed out before, Col. Hamilton Smith, who seems to have worked direct from Azara's original Spanish, translates as 'bay' what the English translator renders 'chestnut,' but as Col. Smith had given great attention to the vast variety of names for different shades of colour used by Spanish writers on horses, and as the Pampas horses seem really to be what we term ' bay,' this rendering seems to be the true one. Mr Darwin (Variation, Vol. i. p. 64), who cites Azara's French edition of his work, agrees with Col. Smith, for he cites Azara (Les Quadrupeds du Paraguay, Tom. ii. p. 307) as stating that "90 out of 100 horses were 'bai- chiltains,' the remainder were ' zains,' that is without any white ; not more than 1 in 2000 being black." - Quadrupeds of Paraguay, p. 31 (Engl, trans.). 28—2 436 THE ORIGIN OF THE LIBYAN HORSE [CH. Africa. We have seen that the quagga, whose habitat was bounded by the Vaal River on the north and extended as far as lat. 32° S., had not only divested itself of its stripes to a still greater degree than the Burchell zebra, but had assumed a general bay colour except on its lower parts. Mr Pocock^ has admirably traced the various gradations in coloration from Grant's zebra in North-east Africa to that of the Quaggas in the Cape Colony. The first of these is striped all over down to the very hoofs with black in strong contrast to its white ground colour, but even in British and German East Africa " the pale interspaces begin to be washed with brown and to be filled in with narrow intervening stripes, and such forms are difficult to distinguish from E. selousi of the Mashonaland plateau. From these may be traced a series of gradations represented by the local races named after Chapman, Wahlberg, and Burchell, in which the stripes gradually dis- appear and thin upwards from the fetlocks to the shoulders and haunches, while those on the body lose their connection with the mid-ventral band and becoming shorter leave the belly unstriped. Concomitantly the intervening ' shadow ' stripes increase in number and definition as they extend for- wards towards the neck, then the normal stripes themselves turn brown, and the ochre-stained brown colour deepens in hue. In Burchell's the shadow stripes reach the head, and the last of the complete stripes is the one that extends backwards from the stifles to the root of the tail, the hindquarters and the legs being practically, and the belly actually, stripeless. It is but a step from this to the extinct Grey's quagga, in which the stripes of the body were fused together and blended to a great extent with the brown of the intervening areas, those on the neck being exceedingly broad, and broken up by paler tracts of hair." This process is admirably illustrated by the head and neck (Figs. 131 — 3) of a quagga, hitherto unnoticed by the students of the Equidae, to which my attention was called by my friend Dr W. L. H. Duckworth-, M.A., Fellow of Jesus 1 ' The Coloration of the Quapgas,' Nature, 1903, pp. 356-7. '■* Dr Duckworth (who has laid me under many obligations) on a hasty visit IV] THE ORIGIN OF THE LIBYAN HORSE 437 College, Cambridge, and which is here for the first time pub- lished. It is in the museum of the Elgin and Morayshire Society at Elgin, to which it was "presented in 1861 by John Maclean, Esq., of King Williamstown, British Kaffraria," according to a document hunted up for me by the most obliging curator, Mr Gordon Tajdor, to whom I am indebted for the following description. " There are nine stripes on the face (Fig. 131), spreading out and meeting in a point on the fore- head about three inches below the ears and meeting in the same way about four inches above the nose, with four or five running upward and inwards above each eye. The ears are smaller than those of the zebra. The ground colour of the head and face is a dark fawn, the neck (of which there are about 15 inches attached to the head) inclines to dull brown above, passing into a fawn colour below. On the portion of the neck attached to the head there are six dull white interspaces mixed with brown hairs measuring from half-an-inch up to an inch in width, the widest being nearest the body, the stripes themselves being very irregular and wavy and being completely fused on the throat. The hair of the almost erect mane is mostly reddish-brown tipped with a darker shade, having patches of white, the largest bunches being where the interspaces meet the mane. The stripes on the face are white, the spaces between them a dull brown, showing a decided contrast to the ground colour of the face. On the side of the face, running down across the jaw, are six or seven stripes of the same shade of brown, but no white." Of the described specimens of the quagga (pp. 70 — 7), Mr Pocock thinks that the Elgin specimen comes much nearest to E. greyi, but he thinks it may belong to a new type. The importance of an additional fragment of quagga is increased by the fact that it is the only one of the existing specimens the provenance of which is known. The change in coloration from E. granti (pp. 63-9) to Daniell's quagga (p. 78) was probably due to "a change of habitat from bush to open plain. A new method of conceal- to the Elgin Museum last September noticed a specimen labelled 'Quagga' and at once called my attention to it. 438 THE ORIGIN OF THE LIBYAN HORSE [CH. meat by means of shadow contraction was required and was gradually perfected by the toning down of the stripes on the upper side and the suppression of those on the hind-quarters and legs^" Mr Pocock thinks that the ruddiness of Cape Fig. 131. The Elgin Quagga. Colony quaggas was acquired by two different processes — " in the case of Daniell's quagga by a deepening of the red of the ground colour or interspaces and the reduction in the length 1 Ann. Mag. Natural History, 1901, p. 328. IV] THE ORIGIN OF THE LIBYAN HORSE 439 4^ *?.; Fig. 133. The Elgin Quagga (left side). 440 THE ORIGIN OF THE LIBYAN HORSE [CH. and the width of the black stripes on the neck and shoulder concomitantly with their suppression on the body, the result being a chestnut or bay-coloured animal with narrow black stripes on its fore-parts " ; in the case of Grey's quagga (and the Elgin specimen) by an increase in the width and brownness of the stripes, followed by their fusion and loss of definition on the body, the result being a ruddy-brown animal marked with narrow pale bands (the interspaces) upon the head, neck, and shoulder, as in the Elgin specimen here figured. Now as the Libyan horse makes its first appearance amongst the nomadic tribes of Libya whose territories were about lat. 32" N. it is fair to suggest that similar conditions of climate and food and a like need of a protective colour suited for life in open plains had produced the bay colour in both the quagga and Libyan horse. It will also be remembered that the quagga closely resembled the horse in the character of its tail. It has likewise been shown that the feral horses of North America sprung from Spanish horses chiefly from Northern Spain after three centuries wore liveries of black, grey, roan, roan pied with dun, and dun frequently with dorsal stripes derived from their cross-bred ancestors, the dun with stripes not being a reversion to a primal ancestor, but simply the coat inherited from the dun-coloured striped horses of the sierras of Spain. The swiftest horse known in Homeric days was a bay with a star in his forehead ; in Greek classical times, the dark horses of Libya were the swiftest known and the same horses bore away the palm from all others in the Roman circus in the first century of our era ; and the Saracens, so soon as they got possession of these horses, became the swiftest riders in the world, and the best Anazah horses of the present day are bay with a star in the forehead and 'bracelets.' If then it could bo shown that in a definite series of cases where very well bred, but not quite pure-bred, horses have been bred solely with a view to speed, the increase of speed in the stock has been steadily accompanied with the gradual disappearance of all other colours except bay (or chestnut), we should get a IV] THE ORIGIN OF THE LIBYAN HORSE 441 very substantial corroboration for our conclusion that the primeval colour of the Libyan horse was bay. The English race-horse supplies exactly the test case that we seek. Major-General Tweedie^ when discussing the colours of Arab horses, makes the following valuable observation on those of the English thoroughbred : "The production by man through methodical selection of breeds of horses of one colour, is as intelligible as the distribution by nature of troops of wild horses, every individual of which resembles the surface of the ground. But another fact here presents itself, which seems still to await explanation. Except in so far as statistics show that there have been more winners of one colour than of another colour, English breeders for the turf may safely be acquitted of all preference, or fancy, respecting colour. And yet, equally in our islands and at the Antipodes, the long course of scientific breeding, of which our racing stock is the product, has practi- cally resulted in its becoming a family of bays and chestnuts — two colours essentially one. In olden times when England was full of fresh Eastern blood, greys were as often seen at the starting-post as they were down to a much later period in New South Wales and Victoria. In the thirty years preceding 1S66 it was estimated- that the Derby had been won by seven chestnuts, seven browns, and sixteen bays ; the St Leger by five chestnuts, eight browns, and seventeen bays, and the Oaks in like proportion. The tendency of the highest breeding in latitudes far separated is to wipe out in horses all colours save bay and chestnut." Now, as it will be remembered that Major- General Tweedie is a firm advocate of the theory that the Arab horse is a purely artificial product, his testimony is all the more valuable as it is that of a hostile witness. An examination of the colours of the winners of the Derby, Oaks, and St Leger^ for the three decades from 1870 to 1899 inclusive proves that not merely has grey disappeared altogether 1 The Arabian Horse, p. 267. 2 R. H. Copperthwaite, 'The Turf and the Race-horse, ed. 2 (1866), p. 144. * As the Racing Calendar does not attach colour marks to the lists of winners of the Derby, Oaks, and St Leger, I had to compile my table by going through the volumes of the Racing Calendar for the years tabulated. 442 THE ORIGIN OF THE LIBYAN HORSE [CH. and that black has ahnost gone, but that chestnut (which Major-General Tweedie holds to be the same colour as bay) is also disappearing as well as brown. Table I, of the colours of the winning horses in the Derby, Oaks, and St Leger from 1870 to 1899 : Decade Bay Chestnut Brown Black or Brown Black 1870-9 15 12 2 1 0 1880-9 16 8 5 0 1 1890-9 17 6 7 0 0 From the table it is clear that during the last third of the past century ba}'^ has been slowly gaining upon both brown and chestnut combined, and that brown has been gaining upon chestnut. Thus the dun element, which as we already believed from our previous investigations when mixed with bay gives chestnut, is steadily being eliminated and our racing stock is becoming a breed of bays and browns with a steady tendency to become eventually purely bay. The same tendency is shown still more emphatically if we take the colours of the first three horses in each of the three great races just named. Table II\ showing the colours of the three first horses in the Derby, the Oaks, and the St Leger from 1870 to 1899 : Decade Bay Chestnut Brown Brown or Black Black Bay or Brown 1870-9 36 34 13 4 2 1 1880-9 42 28 16 0 2 1 1890-9 54 16 17 1 0 2 Out of 90 horses bay only had 36 in 1870-9, but rose to 54 in the last decade; whilst chestnut ^ which was repre- 1 This table is not quite complete, for the colour was omitted in one instance in the Calendar, though those of the winners are given without fail. '^ With reference to the proportions of chestnut found in the two sexes the results are as follows: Totals for thirty years: colts 41, fillies 36. 1870-9: colts 20, fillies 13; 1880-9: colts 9, fillies 19; 1890-9: colts 12, fillies 3. The great decrease in chestnut in the last decade is therefore especially seen in the mares. IV] THE ORIGIN OF THE LIBYAN HORSE 443 sented by only 2 less than bay in the first decade, has sunk to 16 in the last ; brown has gained slightly at the expense of chestnut, rising from 13 to 17. There can be little doubt in view of these facts that the English racing stock is steadily becoming bay. Table II shows that the intermediate steps between black and bay are (1) a hue which cannot be described as a really black or really brown, (2) true brown, (3) bay-brown, true bay being finally reached. Chestnut seems to pass into bay either directly or indirectly through brown. Thus dun and white (the cliaracteristic colours of the old European- Asiatic horses), first disappeared, then grey went, black is almost gone, chestnut is following it slowly, and brown at a still slower rate. I have shown in the course of our investigation that black, dull black, brown, various shades of grey, and probably chest- nut, are due to the intermixture of Asiatic horses with Libyan in various proportions. But as the Asiatic hoi'se of historical times is thick-set, coarse, and slow in pace, and inevitably injures the speed of the pure Libyan strain when crossed with it (as in the case of the 'Gulf Arabs,' p. ]7o) every attempt to improve the speed of such a mixed breed will inevitably tend to eliminate every Upper Asiatic element. But as the original Arabs, Barbs, and Turks which formed the basis of our thoroughbred stock (p.' 382) were, with the single ex- ception of the Darley Arabian, all more or less contaminated with European and Asiatic blood {e.g. the Yellow Turk, the White Turk, Button's Grey Barb, Grey Wilkes, etc.), the unceasing efforts of breeders to obtain greater lightness and speed are continually eliminating the Asiatic and European element, and accordingly dun and white have long disappeared from, and grey, black, chestnut and brown horses are gradually ceasing to be found in, our blood stock. But as increase of speed is gradually rendering the English thoroughbred a purely bay stock, and as from the earliest times of which we have any record the Libyan horse has been not only the swiftest horse known, but also has been of a bay colour, we are justified in concluding that his bay colour is as funda- mental a characteristic as his speed, and that it is due not to artificial selection, but to natural specialisation. 444 THE ORIGIN OF THE LIBYAN HORSE [CH. Nor must Darwin's second conclusion — that the ancestor of all the Equidae was striped like a zebra — be accepted in its entirety. We have already seen (p. 78) that the tendency to stripes is least in the northern latitudes, where the genus first made its appearance in Asia, that this tendency gradually in- creases as we advance southwards, that it reaches its maximum in the tropical and sub-tropical regions of Africa, and that the stripes show a tendency to disappear in Burchell's zebra of the Transvaal and Orange River Colony, and to a still greater degree in the quagga, whose range seems always to have been south of the Vaal River. I showed that if the Equidae as a whole or in part are gradually divesting themselves of stripes, those of Africa have retained their stripes much more tenaciously than those of Asia, or if as a whole, or only in the case of certain species, they are gradually assuming stripes, those of Africa have far outstripped their congeners of the northern latitudes, and I argued that the presence of manifold stripes all over the body in any member of the genus Equus is a strong indication that it has been long domiciled in Africa, where its progenitors for protection or recognition or for some other purpose, either retained and modified the coat of a common ancestor of all the Equidae, or else put on striping differing in different species and varieties according to the nature of their environment or for other reasons. These considerations suggest that the tendency to zebra stripes in certain domestic horses may be less due to reversion to the colour of a remote ancestor, than to their being descended from the Libyan horse. It is therefore worth examining the evidence collected by Col. Hamilton Smith, Darwin, Ewart and others for the exist- ence of such markings in horses. If it should turn out that all such animals are of undoubted North African origin, or at least very probably have some of that strain in them, our contention that the Libyan has been domiciled and highly specialised in North Africa from a very remote period, and that he is distinct from the Asiatic horses of recent epochs, will have received very substantial corroboration. Let us now examine all the instances on which Darwin based his conclusion, supplementing them as we proceed with fresh examples : — IV] THE ORIGIN OF THE LIBYAN HORSE 445 1. Mr Edwards examined twenty-two foals of race-horses, and twelve had the spinal stripe more or less plain. 2. A bay mare (belonging to Mr Darwin), descended from a dark brown Flemish mare by a light grey Turcoman horse, was put to Hercules, a thoroughbred dark bay, whose sire Kingston and dam were both bays; she had a colt which ultimately turned out brown, " but when only a fortnight old it was a dirty bay, shaded with mouse-grey, and in parts with a yellowish tint : it had not only a trace of the spinal stripe, with a few obscure transverse bars on the legs, but almost the whole body was marked with very narrow dark stripes, in most parts so obscure as to be visible only in certain lights, like the stripes which may be seen on black kittens. These stripes were distinct on the hind-quarters, where they diverged from the spine and pointed a little forward : many of them as they diverged became a little branched, exactly in the same manner as in some zebrine species. The stripes were plainest on the forehead between the ears, where they formed a set of pointed arches, one under the other, decreasing in size downwards towards the muzzle ; exactly similar marks may be seen on the forehead of the quagga and Burchell's zebra. When this foal was two or three months old all the stripes entirely disappeared ^" 3. Prof. Ewart's'^ high-caste Arab filly Fatima, bred by Mr Wilfrid Blunt, even when full-grown, shows as distinct stripes in the region of the knee and hock as are to be found in Norwegian dun-coloured ponies, and she has in addition to a dorsal band faint indications of markings across the withers. 4. Major Upton noticed very frequently among colt foals (though not in fillies) of the pure-bred Al-Khamseh horses of the Anazah tribes of Central Arabia "a line somewhat darker than the general colour of the animal running in continuation of the mane along the spine, and to be traced for some way even among the long hair of the tail. It is not obliterated with age ; it can be traced in old horses and in those of a very dark colour*." 1 Variation, Vol. i. pp. 60-1. '^ Ewart, Pemjcuik Experiments, p. Ixx. 3 Op. cit. p. 339. 446 THE ORIGIN OF THE LIBYAN HORSE [CH. 5. Mr Edwards had seen a nearly thoroughbred chestnut horse which had the spinal stripe and distinct bars on the legs\ 6. Col. Hamilton Smith- speaks of dun horses in the sierras of Spain which have a spinal stripe, and we have found the descendants of Spanish horses in Mexico and the Western States frequently of a dun colour marked with stripes (p. 265). But we have seen that all the horses of Spain from before the Christian era were more or less impregnated with Libyan blood (p. 256). 7. Out of 300 South American horses imported into Madras many had transverse stripes on the legs and short shoulder stripes I The most strongly marked individual was a mouse- dun with the shoulder stripe slightly forked. We have shown that the Andalusian horses are almost wholly Libyan in blood, and that other Spanish horses are largely imbued with the same blood, and also that the South American Pampas horses are descended from some Andalusian horses introduced by the early Spanish settlers. 8. The Karadagh horses, which are the best native horses of Armenia to-day, were originally the dun horses of the southern slopes of the Caucasus. Many of them have been recently crossed with Russian blood (p. 193), which is of course largely Libyan, and probably have had much of the same strain through Arab and other channels. They are usually bay or chestnut with black manes and tails. They all have a dorsal stripe about an inch broad from the mane to the taiP. 9. In the north-western parts of India striped horses of more than one breed are very common. In Kattywar the native horses are usually of a rufous-grey or khaki colour. At one time Kattywar horses were not considered pure unless decorated with a dorsal band and bars across the legs. Some- times in addition there were stripes on the neck, forehead and withers^ The Kattywar horses are often fifteen or sixteen hands in height, and are well, but lightly built, 1 Darwin, Variation, Vol. i. p. 60. '^ llie Horse, p. 275. 3 Darwin, Variation, Vol. i. p. 61. •• Hayes, Points of the Horse, pp. 610-11. * Ewart, Exper. Contributions, p. 21. IV] THE ORIGIN OF THE LIBYAN HORSE 447 " There are sometimes stripes on the grey and bay Katty- wars when first foaled, but they soon fade away^" But we have seen (p. 157) that the horses of all North-western India are a blend of the dun-coloured aboriginal horse of Upper Asia and Arab horses which have been imported in enormous quantities annually into India, and whose blood has saturated the Turcoman horses. The fact that the Kattywar horses are dun, rufous-grey, and bay is sufficient proof in the light of our investigations that they are largely mixed with Arab blood, a cooclusiou strongly corroborated by their slender build, and the fact that " a horse is not considered pure unless he shows stripes " indicates that these stripes are a special characteristic of horses which have the greatest amount of good Arabian, i.e. North African, blood in their veins. 10. The horses of the Waziri are said to be not uncommonly decorated with stripes on the legs. But, as we have seen above, the horses of Afghanistan and Baluchistan (p. 159) are cross- breds between the Mongolian and the Arab, and there is some evidence for the existence of Arab blood in Afghanistan at least as early as the thirteenth century A.D. For Marco Polo'', speaking of Badakshan, says that " it produces numbers of excellent horses remarkable for their speed. They are not shod at all, although constantly used in a mountainous country and on very bad roads. They go at a great pace, even down steep descents, where other horses neither would nor could do the like." And Messer Marco was told that " not long ago they possessed in that province a breed of horses from the strain of Alexander's horse Bucephalus, all of which had from their birth a particular mark on the forehead. This breed was entirely in the hands of an uncle of the King's ; and in con- sequence of his refusal to let the King have any of them the latter put him to death. The widow then in despite destroyed the whole breed, and it is now extinct." We have already repeatedly seen that a star in the forehead is especially characteristic of the North African horse and its various 1 Darwin, Variation, etc.. Vol, i. p. Gl. - Vol. i. p. 15G (Yule). 448 THE ORIGIN OF THE LIBYAN HORSE [CH. derivatives. The statement that the best breed of Badakshan had a peculiar mark in the forehead curiously confirms the native belief that the strain had come from the westward. 11. We saw (p. 154) that the Tibetan ponies are the most richly decorated horses in Asia, and that they were mainly of Mongolian stock, but had probably been crossed with Libyan blood, like all the horses of western, north-western, and northern India. They are frequently piebald and skewbald, like the Sumatra ponies, sprung from the crossing of Arab horses with native ponies of a mixed Arab and Upper Asiatic stock, and as the Tibetan ponies also show dorsal and other stripes, like the ponies of Sumatra and Java just mentioned, and the horses of Kattywar, which are saturated with Arab blood, we may reasonably conclude that the colouring and striping of the Tibetan ponies are due to the blending of Arab blood with that of the Mongolian pony. 12. The Shan^ ponies have not unfrequently spinal, leg, and shoulder stripes ; 13 the Burmese, and 14 Javanese ponies are frequently dun-coloured, and have three kinds of stripes " in the same degree as in England'-." 15. Two bay Pegu ponies had leg stripes^, and 16 two Chinese ponies — one of the Amoy, the other of the Shanghai breed — in colour light-dun, had each the spinal stripe, the latter an indistinct shoulder stripe*. But I have already shown that all the horses of India, both Hither and Further, have been largely mixed with Arab blood, which has streamed for at least a thousand years into all parts of Hindustan, and thence into the lands beyond, and along with manifold other Arab influences into the great islands of the Malay Archipelago, which previously possessed no horses (pp. 142-6). We have seen (p. 141) that the Shan (Burmese) or Pegu ponies, the Manipur ponies, and those of Sumatra and Java resemble each other. But as the Javanese and Sumatran ponies are largely impregnated with Arab blood, and as all the horses of India are saturated with the same strain, we need not 1 Darwin, Variation, Vol. i. p. 61. ^ Ihid. » Ibid. * Op. cit. Vol. I. p. 62. IV] THE ORIGIN OF THE LIBYAN HORSE 449 hesitate to believe that the Shan and Manipur ponies have been similarly crossed. Nor need we be surprised to find in China ponies showing traces of stripes, when we recall the vigorous horse-trade between China and India in the days of Marco Polo, and when we bear in mind the repeated evidence of the steady determination of every tribe in Asia and Europe from the dawn of history to obtain for their mares the services of sires of the Libyan stock. Moreover we know from the Chinese docu- ments that about B.C. 100 the emperor sent westwards to obtain horses of an improved breed from Turkestan \ whither it is highly probable horses of a mixed Libyan strain had found their way from Bactria. The fact that the two Pegu ponies cited by Darwin were of a bay colour in itself shows that they had a large infusion of Arab blood. The occurrence in various parts of Eastern Asia of dun- coloured horses with stripes can be completely accounted for by the historical fact that in the horses of all those regions there is a large substratum of Upper Asiatic blood, which has been more or less improved by the repeated introduction of Arab sires, as has been the case in Western Asia also (p. 183). And as in the latter region the predominant colour is that of the native Turcoman and Kurdish mares, which has been but slightly influenced by the dark colour of the Libyan sires, so too in Eastern Asia, whilst the structure and quality of the breed have been improved by the Arab sires, the latter have not been able to modify to any great extent the light colours of the native mares, though not unfrequently transmitting to their progeny their own zebra-like markings. Let us now return to Europe. 17. " On the Continent the offspring of black sires satu- rated with Arab blood are often more or less distinctly striped-." 18. The Cleveland Bays which resulted from crossing Yorkshire cart-mares with a Barb, had almost invariably a dorsal stripe (p. 886). This appears probably due to the Barb blood. 1 H. F. Osborn, Century Magazine, Nov., 1904, p. 16. " Ewart, Exp. Contributions, p. 21. R. H. 29 450 THE ORIGIN OF THE LIBYAN HORSE [CH. 19. Two bay English carriage-horses had each a black spinal stripe ; one of them had on each shoulder a light shoulder stripe, and the other had a broad, black, ill-defined stripe running obliquely half-way down each shoulder ; neither had leg stripes ^ 20. A bay Highland cob belonging to Prof Ewart had a shoulder stripe nearly a foot in length, and a neck stripe that extended quite two-thirds across the neck, and six short stripes on the body behind the shoulder striped 21. The offspring of dark Highland pony sires not unfre- quently show markiugs. 22. Darwin had seen marks on the forehead of a fully-grown fallow-dun, cob-like horse (which had also a conspicuous spinal stripe and its front legs well barred), marked similar to those on the forehead of his own foal. 23. A bright fallow-dun cob had its front legs transversely barred on the under side in the most conspicuous manner. 24. A bright fallow-dun colt, fully three parts thorough- bred, with very plain transverse stripes on the legs. 25. A small, purely-bred, light fallow-dun Welsh pony had a spinal stripe, a single transverse stripe on each leg, and three shoulder stripes ; the posterior stripe corresponding with that on the shoulder of the ass was the longest, whilst the two anterior parallel stripes arising from the mane decreased in length in a reverse manner to that in the next instance. 26. A rather large, lightly-built, fallow-dun, Devonshire pony^ with a conspicuous stripe along the back, and with light transverse stripes on the under sides of its front legs, and with four parallel stripes on each shoulder. Of these four stripes the posterior one was very minute and faint : the anterior one, on the other hand, was long and broad, but interrupted in the middle and truncated at its lower extremity, with the anterior angle produced into a long tapering point. "The shoulder stripe of the ass occasionally presents exactly the same appearance." 27. A dark leaden, mouse-coloured pony had leg stripes, but not very conspicuous. 1 Darwin, Variation, Vol. i. p. 60. ^ Oj)- cit. p. 105. * Variation, Vol. i. pp. 610-11. IV] THE ORIGIN OF THE LIBYAN HORSE 451 As no one will dispute the fact that our carriage-horses, riding horses and trappers are the result of a judicious blend of the coarse, thick-set Asio-European horse and the thoroughbred, and as it is also admitted that many of our native ponies, such as the Welsh and Exmoor, have been largely mixed with North African blood, the occurrence of stripes on such horses as those cited by Darwin can be readily accounted for on the hypothesis that the stripes are due to thoroughbred blood. In several of the instances given he himself states that the animal so marked was well bred, or else his description shows that such was the case, so there can be no doubt of the presence of Libyan blood in these animals. Our investigations have likewise shown that all shades of black and chestnut are due to a blending of the Libyan with the northern blood. 28. A pony said to be Welsh, in colour mouse-black or dark slate-grey, with a dorsal stripe, four stripes across the withers, and one or two indistinct ones on the leg\ 29. A chestnut-dun cart-horse had a conspicuous spinal stripe, with distinct traces of shoulder stripes but none on the legs. 30. A large, heavy Belgian cart-horse, of a fallow-dun colour, had a conspicuous spinal stripe and traces of leg stripes, and two parallel stripes on both shoulders. 31. Another rather light cart-horse of a dirty dark cream colour, with striped legs, and on one shoulder a large ill-defined dark cloudy patch, and on the opposite shoulder two parallel faint stripes. But I have already shown that the large cart-horses of England have been imported from the Continent, and that these large breeds were evolved by infusing the blood of the Libyan horse into the little horses of Upper Europe. It has also been shown that dark sires saturated with Arabian blood constantly beget progeny marked with stripes. It is therefore not a matter for surprise if stripes occasionally appear on cart- horses both in this country and on the Continent. 32. We have seen that the colour of the horses of the 1 This animal was examined by my friend Dr W. L. H. Duckworth, M.A., Fellow of Jesus College, Cambridge, at the Cambridge Midsummer Fair, 1904. 29—2 452 THE ORIGIN OF THE LIBYAN HORSE [CH. North in early days was dun. lu modern Norway the colour is dun, varying from cream colour to dark mouse-dun ; and an animal is not considered purely bred unless it has the spinal and leg stripes. A large proportion (one-third) have striped legs, one pony having seven on the fore-legs and two on the hind-legs, but only a few exhibit shoulder stripes^ 33. A sand-coloured (yellow-dun) Norwegian pony, with black mane and tail, in the possession of Prof Ewart, has stripes on the face, neck, body, and legs. " On the forehead there are two all but complete frontal arches and portions of five others. Being of a reddish-brown colour, these stripes are easily seen when the forelock is thrown back. The uppermost (orbital) arch ends in the frontal tuft, but instead of forming an actually pointed arch as in Matopo (his Burchell's zebra), it forms a somewhat rounded arch, as in the Amsterdam quagga" and in one of Prof. Ewart 's zebra hybrids. Fragments of the other arches are most distinct in the middle of the forehead. " In having seven more or less complete frontal arches," says Prof. Ewart", " this pony differs from my Burchell zebra, in which there are only three distinct arches in the corresponding posi- tion. When the incomplete arches are ' restored,' a pattern is formed which is almost intermediate between that of the Amsterdam quagga and the Somali zebra. From within the lowest arch several obscure lines, such as occur in zebra-ass hybrids, can be traced along the face. The stripes doubtless originally ended in or near a mealy-coloured muzzle, such as we find to-day in typical Exmoor ponies and some Somali zebras." In this Norwegian pony " there are only a few light hairs at the tip of the ears, but immediately below there is a broad black band, and an indistinct band near the base. Had the tip been lighter in colour, the ear of this pony would have not a little resembled in its decoration that of my Burchell zebra." Ewart has seen two other ponies, in colour light dun — 1 Darwin, Variation, Vol. i. p. 61. It is right to add that during the present summer (1904) my friend Dr Venn, F.R.S., and his son Mr J. A. Venn have examined for me a large number of Norwegian ponies at various coast towns in Norway, without having met a single instance of striping. - Penycuik Exp., pp. 102-3, fig. 36. IV] THE ORIGIN OF THE LIBYAN HORSE 453 one from Shetland, the other from Norway, " with the tips for quite half an inch almost white. In both cases a broad dark band extended across beneath the light tip\" Prof. Ewart adds that he has seen only two ponies with stripes on the face, and that he thinks that facial stripes are extremely rare. The same authority^ has " only seen faint indications of stripes on the side of the face, but from what he has seen he has no hesitation in saying that were a sketch made showing all the stripes of which fragments have been observed on the face of the horse during the present generation, the sketch would closely resemble the head of one of my (Ewart 's) hybrids and less closely the Amsterdam quagga." We shall see very shortly that Ewart's hybrids show the decoration not of their sire Matopo, a Burchell zebra of the Chapman variety, but that of the Somali zebra. , Prof. Ewart's Norwegian pony has only a short shoulder stripe, and she has a number of ill-defined stripes in front of the shoulder stripe^ She has three short stripes on the body behind the shoulder stripe, and an extremely well-developed dorsal band "as distinct and as broad as it crosses the croup as in my Burchell zebra." The edges of this band "give off short processes — rudiments of developing stripes, or vestiges of dwindling ones, such as are seen in some of the quaggas and in zebra-ass hybrids." "In another light dun-coloured pony there are ten cer- vical stripes. As these ten stripes only extend about half- way along the neck, and as stripes are sometimes present immediately behind the ears, there may have been quite twenty stripes in the ancestors." Prof Ewart^ remarks that "sometimes the shoulder stripes bifurcated some distance above the shoulder-joint," and thus suggests not so much the zebra as the quagga and zebra- hybrids, and that " as a rule the neck stripes are short and in- distinct, but that in some cases he had seen three or four cervical stripes nearly as well defined as in the zebras, whilst in one case he had observed several stripes extending into the mane." 1 Ewart, op. cit. p. 104. ^ Qp. cit. pp. 103-4. 3 Op. cit. p. 105. * Op. cit. pp. 105-6. 454 THE ORIGIN OF THE LIBYAN HORSE [CH. He adds that once only had he seen " as many as five distinct stripes extending from the dorsal band across the back." First let us notice that stripes are in Norway, as in Kattywar, the marks of good breeding. But, as in the latter case, we know absolutely that the native horses have been saturated with Arab blood, the stripes in Norwegian ponies may be similarly explained. We have traced the improve- ment of the breeds of horses in Upper Europe from the second century B.C., when the Gauls began to import from the south a superior type of animal, and we saw that by Caesar's time Gaul was well supplied with southern horses. We also ascribed the superiority of the horses of the Tencteri, the only German tribe of the time of Tacitus which appears to have been well- mounted, to their close proximity to the Gauls. It is thus clear that by the second century a.d. a good deal of Libyan blood had made its way into Central and Upper Europe. It is not then surprising if by the tenth century excellent horses imbued with southern blood were found in Norway and Iceland. We have seen that the best horses mentioned in Burnt Njal are a chestnut, a brown, and two dun-coloured horses with black stripes down their backs, these last " were the best steeds to ride in all the country round." Here, again, the combina- tion of black stripes with special excellence points to the same explanation as that given for the striped horses of Kattywar. If the stripes were but a characteristic feature of the dun stock of Northern Europe, why should horses of this description be superior to other dims, both in ancient and modern times ? On the other hand, as soon as we recognize that the stripes are due to the presence of North African blood, the cause of the superiority of such animals is at once obvious. This is fully confirmed by Glaus Magnus, who says (p. 348) that the light dun is especially the livery of the wild horses of Europe, and that horses of this colour were always the worst (a view held by "Virgil many centuries earlier), but that the best of them were those with a dorsal stripe. But since dun without stripes was the colour of the unimproved wild or feral horses of Europe, and the best duns whether in medieval Iceland and IV] THE ORIGIN OF THE LIBYAN HORSE 455 Europe or in modern India are commonly striped, and as we have shown that the striped horses of Kattywar owe their superiority to the infusion of Libyan blood, and since the striped dun horses of Spain are certainly a blend of the old dun European horses and the Libyan, we may conclude with high probability that stripes in all such cases are mainly due to the infusion of Libyan blood. An analysis of the instances of striped horses here adduced proves that the great majority were pure Arabians, English thoroughbreds, Spanish horses derived from North Africa, Pampas horses descended from Andalusian horses, three-quarter bred and half-bred horses, English carriage-horses and cobs, well-bred Welsh and Devonshire ponies, Kattywar, Burmese, and Javanese ponies largely infused with Arab blood, the progeny of dark sires saturated Avith Arab blood, and the off- spring of dark Highland ponies. Prof. Ewart has lately informed me that he has seen three Mongolian ponies which show markings. They are particularly well defined in one in his own possession, but, as she is very fleet and shows clear indications of a good deal of Arab blood, we must not hastily ascribe the stripes to her Asiatic rather than to her Libyan blood. The account given on an earlier page (p. 138) of the Mongolian ponies shows that they are much mixed in blood, and as there is good reason to believe that the Chinese became acquainted with the fleet horses of the West at a compara- tively early period, we need not be surprised if the Mongols, like the Turcomans, Kurds, and other horse-keeping tribes of Asia, were always anxious to infuse Arab blood into their own slow ponies. Mongolian ponies have not come under the observation of scientific men to the same degree as the horses of other parts of Asia, and, consequently, the absence of more instances may be simply due to lack of information. Prof. Ewart has recently received from Mr Hagenbeck a Prejvalsky two-year old mare which has a distinct dorsal band, faint markings across the withers, and conspicuous bars in the vicinity of the knees and hocks. We have already seen that the Prejvalsky horses are thus marked, but unfortunately as the question of the purity of 456 THE ORIGIN OF THE LIBYAN HORSE [CH. the specimens is very uncertain (p. 34), we cannot venture to say whether such markings are primitive in the stock or due to crossing with feral horses. On the other hand stripes are rare in cart-horses, which, as we know, are chiefly sprung from the old coarse, thick-set horses of Asia and Europe. We must, therefore, conclude that such stripings are in a special degree characteristic of the Libyan horse and his derivatives, and are not equally common in all breeds of horses. But as manifold striping is especially characteristic of the Equidae of Africa, we may not unreasonably infer that the ancestors of the Libyan horse had long lived in Africa. Prof. Ewart^ has pointed out that in " highly bred foals with very fine coats there are often at birth across the sides and croup, and especially in the vicinity of the flank feather, narrow markings that might be mistaken for stripes. These markings are caused by the hair being arranged in well-marked tracks or ridges, separated by almost hairless spaces. In these tracks, which were very distinct at birth in a cinnamon-coloured foal I bred this year, out of a bay half-Arab mare — the sire was a chestnut thoroughbred horse — we have, it may be, a restoration for a time of an ancestral condition. Sometimes along with these hair-tracks or ridges there are faint stripes, seen only in certain lights, but evidently in part due to subtle colouring. Stripes of this nature I noticed plentifully scattered over a reddish-grey foal out of my flea-bitten New Forest pony by the grey Arab Benazrek. More common and more evident are comparatively broad wavy bands often seen across the croup and on the brow of half-bred bay foals. These bands may occupy the position of ancestral stripes — stripes out of which the colour has been completely washed since they ceased to count in the struggle for existence." Ewart based his sup- position on the fact that " they occupy the position of stripes in a yellow-dun Norwegian pony, and of the stripes over the croup of one of Lady Meux's [zebra] hybrids (p. 463), which may have been inherited either from the American trotting horse or from a remote common ancestor." ^ Penycuik Experiments, pp. Ixxvi-vii. IV] THE ORIGIN OF THE LIBYAN HORSE 457 The result then of our examination of the occurrence of stripes in horses has led us to the conclusion that such stripes are very often to be traced to North African blood. But as we have seen that it is in Africa the Equidae show a special tendency to stripes, and that it is there alone they are found with stripes all over, as was the case with Darwin's colt, whose markings, as the reader will remember, resembled those of certain zebras, we are justified in inferring that the tendency to stripes which is so marked a feature of the North African horses and their derivatives is due to the circumstance that that strain is not a mere recent differentiation under domesti- cation and artificial breeding of an Asiatic domestic breed, but is rather to be considered a species specialised during a long lapse of time in Libya under conditions somewhat similar to those which have produced the zebras. It has been supposed that certain experiments conducted by Prof Cossar Ewart in order to test the truth or falsity of the commonly received doctrine of Telegony, confirm Darwin's hypothesis that the common ancestor of the Equidae was a striped animal. The theory of Telegony gained much support from the famous letter of Lord Morton to Dr Wollaston, of Gonville and Caius College, Cambridge, then President of the Royal Society. In it Lord Morton stated that he had mated his quagga stallion (Fig. 39) " with a young chestnut mare of seven-eighths Arabian blood, and which had never been bred from." The offspring was a filly, striped not only on the body but also on the legs, which were not so marked in her quagga sire. Lord Morton sold the chestnut mare to Sir Gore Ouseley, who bred from her by a very fine black Arabian horse a filly and a colt, which according to Lord Morton, in their colour and the hair of their mane had " a striking resemblance to the quagga." The filly and colt were in some respects more striped than either the quagga or quagga hybrids From this it was inferred that the quagga had so infected the mare that her union with him influenced the offspring of her subsequent matings. In order to test this. Prof. Ewart 1 Phil. Transactions, 1821, p. 21 ; Ewart, Pemjcuik Experiments, p. 59. 458 THE ORIGIN OF THE LIBYAN HORSE [CH. crossed Matopo (Fig. .36), a stallion of the Chapman variety of the Burchell zebra, with mares of different kinds and colours as a preliminary, intending after they had foals to put them to ordinary horses. The results of these subsequent matings seem to put it beyond all doubt that there is no sound evidence for Telegony, at least as far as the Equidae are concerned. But the results of the mating of the mares with Matopo are in themselves of the highest interest apart from any question of Telegony. The hybrids, as was to be expected, were striped indeed like zebras, but instead of reproducing the broad charac- teristic markings of the Burchell species, to which their sire belongs, they showed numerous stripes not only narrow like those of the Somali zebra (Fig. 28), but showing the same arrangement, exhibiting on the forehead the round arches seen on the forehead of the Somali zebra instead of the pointed arches of Matopo, and bearing marks on the croup unknown on the Burchell zebra, but peculiar to the Soinali and Mountain zebras. The oldest of the hybrids, Romulus \ was out of a thirteen hands, black, Island of Rum pony. " The well-bred, nearly black ponies of the Scottish Western Highlands and Islands, which have long been under observation, form a dis- tinct breed, well adapted in many ways for crossing with zebras. Their resemblance to Eastern horses has been accounted for by saying that they have descended from sires which escaped from the ships of the Spanish Armada." But I have shown above (p. 400) that these dark-coloured ponies of the Western Isles and Highlands of Scotland are to be traced back to North Africa through France and Spain from before the Christian ei'a, and their origin is no more to be ascribed to the horses from the Spanish Armada, than are the Connemara ponies to Spanish horses similarly obtained or imported during the Tudor period. " In the plan of his striping Romulus was utterly unlike his sire, and when a careful examination was made it became evident that in the number and arrangement of the markings he was not unlike a Somali zebra." "Instead of the four or five acutely pointed frontal arches of his sire, there are fourteen 1 Penyciiik E.xper., pp. 29-33 (figs. 9-11). IV] THE ORIGIN OF THE LIBYAN HORSE 459 rounded arches, that remind one of the face of a Somali zebra. Instead of twelve cervical stripes as in Matopo, there are in Romulus twenty-four cervical stripes, all of which can be traced into the mane. In having so many cervical stripes he seems to be more primitive than even the Somali zebra (in which I have never seen more than fourteen cervical stripes), but closely agrees with one of my zebra mares, when the shadow stripes are included." Romulus likewise had at birth numerous spots arranged in nearly transverse rows over the loins and rump, which as he grew older united to form somewhat zigzag narrow bands, almost identical in their direction with the narrow stripes over the hind-quarters of the Somali zebras. "Counting from the shoulder stripe to the root of the tail, there are forty- three stripes in the hybrid — about the same number as in the Somali zebra." Of all the hybrids Remus is most like a zebra. His mother was Biddy, an Irish mare, three-parts bred, in colour bay, with black points and black mane and tail. " Remus's ground colour is light bay, the stripes — numerous and distinct, except over the croup which has a mottled appearance — are of a dark bay or brown tint. In his hoofs, mane, and tail, and in the body- hair Remus is of all the hybrids most like a zebra\" Baron de Parana has made experiments by crossing a zebra stallion of the true Burchell type (Fig. 37), a white-legged variety with distinct shadow stripes, in build not far removed from the extinct quagga (Fig. 38), with South American mares. All the Brazil hybrids out of ordinary mares very closely resemble " Romulus in their markings — the legs being well striped, notwithstanding absence of markings on the legs of their sire — but they have rounder quarters and are apparently more cob-like in build'-." It has been commonly held that these hybrids revert in their decoration to a remote common ancestor of the Equidae. But it is also possible, and I venture to think more probable 1 'Experimental Contributions to the theory of Heredity, Reversion, and Telegony in tlie Equidae,' Trans. Highland and Agricultural Soc. of Scotland, ser. 5, Vol. xiv. (1902), p. 41. - Ewart, Experimental Contrib., pp. 47-8. 460 THE ORIGIN OF THE LIBYAN HORSE [CH. that the h3'brids have not reverted to a remote ancestor, but rather to the immediate ancestor of the North African horses, with the blood of which Biddy, the three-quarters bred Irish mare, the well-bred Island of Rum pony, and the South American mares were saturated. We have just seen that beyond reasonable doubt there is a special tendency to exhibit zebra stripes in the Libyan horse and its descendants, whether they be in the Pampas of South America or in Kattywar, Tibet, or the Malay Archipelago. It is therefore not im- probable that Prof. Evvart's hybrids show stripings closely resembling those of the zebra of North-eastern Africa instead of those of their sire because their mothers had in their veins more or less of the blood of the North African horse, which has retained or developed a tendency to stripes resembling closely in size and distribution those of the Somali zebra under con- ditions somewhat similar to those under which have arisen the stripes in the latter animal. The reader however will remember that it was shown (p. 80) that although the plan of marking in the Somali zebra represents the oldest type amongst zebras, it by no means follows that the peculiar markings of the Somali zebra represent the original livery of the common ancestor of horses, asses and zebras. A simple experiment seems to confirm this view. I mated a Muscovy drake ' Hans ' (a gift from my friend Dr Gadow) with a common white Aylesbury duck. The Muscovy is, in spite of its name (which is a mere corruption for musk), a South American species, whilst, as is well known, the Aylesbury is derived from the Mallard or common wild duck. The progeny', eight in number, resemble clearly in their colouring the Mallard, each showing the colouring of the corresponding sex in the wild ancestors (Fig. 134). But it will hardly be maintained that the colouring of the hybrids, which proved 1 My hybrids have proved absolutely sterile, though they paired off at an early age. The ducks did not lay a single egg, and a Pekin and an Aylesbury duck mated with one of the hybrid drakes did not produce a single fertile egg. One of the ducks was examined after death, and showed only vestigial ovaries, though a drake similarly examined had apparently the organs of reproduction fully developed. IV] THE ORIGIN OF THE LIBYAN HORSE 461 sterile (like Ewart's zebra hybrids), is a reversion to a remote common ancestor of both Muscovy and Mallard, and not rather to the Mallard, the immediate ancestor of the white Aylesbury duck. But as the markings on the zebra hybrids resemble in character the narrow markings of the Somali zebra, we are led to conclude that this peculiar characteristic of the North African horse is not due to a comparatively recent differentia- tion under domestication of the domestic Asiatic horse in Fig. 134. The Muscovy drake ' Hans ' and his hybrids by a white Aylesbury duck. Libya, but rather that living for a vast period under conditions similar to those which have produced the peculiar stripings of the Somali zebra, it has been so highly specialised as to constitute a separate species. The circumstance that large functional first pre-molars are found in some of the horses of South-eastern Asia (as for instance in Javanese and Sulu ponies), which I have shown to be almost wholly of Libyan blood, and also in the Somali zebra and some members of the Burchell group (p. 142), points to a similar conclusion. 462 THE ORIGIN OF THE LIBYAN HORSE [CH. All these considerations point to the conclusion that when the land-bridges were still unbroken between Europe and Africa, at least one variety of Equidae, probably Equus stenonis or some allied species (p. 10), had crossed over into the region lying between that occupied later on by Equus caballus in Europe on the one hand, and that by the Somali zebra on the other, had there been still more specialised, had differentiated its stripes and subsequently almost completely lost them, and pari passu had assumed a nearly uniform bay colour. As Prof Ewart's experiments have shown that it is most improbable that the stripes on the foals of Sir Gore Ouseley's chestnut mare, by the black Arabian, were due to infection from her former mate the quagga, the stripes in these foals are to be ascribed to the fact that the chestnut being seven- eighths Arab had an inherent tendency to such markings, and as she is said to have come from India, the frequency of the occurrence of such stripes on well-bred Kattywar horses renders still more probable such a tendency in her. Further- more, as we have seen that on the Continent black sires saturated with Arab blood frequently beget offspring with stripes, a fortiori, there must have been a very strong tendency to produce offspring with stripes in the black Arabian stallion as well as in the chestnut mare. The fact that the quagga hybrid was more striped than its quagga sire seems to indicate that the stripes on the hybrid, especially those on its legs, were due not merely to the quagga, but also to its dam ; the further fact that her subsequent foals by the black Arabian were in some respects more striped than the hybrid suggest that the striping in their case may have been due to the dark black Arabian as well as to the dam. The presence of stripes on the legs of the quagga hybrid, though such were absent in the quagga, is completely paralleled by the occurrence of stripes on the legs of hybrids bred by Baron de Parana from a true Burchell zebra (with white legs) and South American mares. As the latter are largely of Andalusian, and consequently of North African blood, and as South American horses constantly show stripes, the markings on the legs of the hybrids may in part be due to their dams, IV] THE ORIGIN OF THE LIBYAN HORSE 463 and need not be referred to a remote common ancestor of the Burchell zebra and the South American mares. Further confirmation of my doctrine that the tendency to stripes is due to the presence of North African blood is afforded by Prof. Ewart's experiments, which point to the conclusion that the less Libyan blood there was in the mares mated with his zebra stallion, the less defined were the stripes in the offspring. Lady Douglas, a young fifteen-hands bay cart-mare by Matopo produced Brenda, who in make and dis- position is quite unlike all the other hybrids : she is of a bay Fig. 135. Chapman's variety of the Burchell Zebra. colour and not very distinctly striped. As a foal she was less intelligent than her hybrid half-brothers and sisters. Lady Douglas next by Matopo had Black Agnes, who " is almost black, so black that the stripes, though abundant, are hardly visible at a distance of a few yards. Black Agnes may have derived her colour from a recent maternal ancestor." Contrast the description of the offspring of the cart-mare with that of Remus (p. 459), the son of the three-quarters bred Irish mare Biddy. Again, the hybrids bred at Theobalds by Sir H. Meux out of a Chapman zebra mare lend some support -to this contention. The eldest, by an English pony, is a yellowish- 464 THE ORIGIN OF THE LIBYAN HORSE [CH. brown, but faintly striped ; the second, by an American trotting- horse, is brilliantly and richly decorated with brown stripes over a bright bay background ; the third hybrid, by a Shetland or Highland pony, is only very faintly and partially striped \ It is almost certain that the American trotting-horse Avas better bred, i.e. had more North African blood in him, for most of the blood in such trotting-horses is thoroughbred"'', than either of the other sires, a supposition rendered all the more likely by the fact that his progeny was bay, a sure index, as we have seen, of the presence of Libyan blood. This cir- cumstance may well account (prepotency apart) for the fact that the hybrid by the American horse is much more richly striped than those by the other sires. If then the hybrids of horses and zebras are the more striped, in proportion as the horse parent, whether it be sire or dam, is better bred, there is certainly a prima facie pro- bability that stripes are more connected with North African than with Asiatic blood. Darwin's view that the original ancestor of the Equidae was a dun-coloured animal striped all over was based, not merely on the occurrence of stripes in horses, which we have just discussed, but on his belief that such stripes were common in dun-coloured horses, and that there was a tendency in horses to revert to dun colour. But it must be confessed that the facts do not warrant his conclusion. In the first place, we have just seen that stripes are specially characteristic of the North African horse and its descendants. But, as we have shown on an earlier page, that the North African horse is invariably dark in colour unless there has been admixture from Europe or Asia, it follows that stripes, so far from being more closely connected with dun colour, are in reality as constant a feature of dark-coloured horses, such as the pure Arab of the Anazah breed, the English thoroughbred, and the South American pampas horses, the last mentioned being, as we saw above (p. 435), universally of a dark colour, bay largely predominating. "■ Ewart, Guide to Zebra Hybrids, pp. 33-4, figs. 28, 29 (Shetland pony sire). 2 Wallace, The Horse of America, pp. 456-7. IV] THE OEIGIN OF THE LIBYAN HORSE 465 In the four regions — Spain, Mexico, Kattywar and Norway — where stripes frequently occur in dun-coloured horses, we were able to make it highly probable that there was a large admixture of North African blood. From this it is clear that stripes are at least as often a concomitant of dark as of dun colour. Moreover, if Darwin's hypothesis of a dun-coloured ancestor with stripes is sound, dark colours such as bay and brown must be of more recent origin, and accordingly there ought to be a great readiness on the part of the progeny of a light-coloured animal when mated with a dark to revert to the light. But Prof. Ewart's zebra stallion "has never been able to stamp his own peculiar pattern or his own colours on his hybrid offspring. The ground colour has been determined by the dams of the hybrids. The hybrids of the better bred mares are of a bay or chestnut hue — the prevailing colour of Arab foals; the hybrids of the Highland, Shetland, and Iceland mares are of a dun colour, and thus they probably take after the horses that in olden times inhabited the north temperate regions \" There is a general belief that stai'S and blazes are found amongst all kinds of horses. But we have now seen that (1) the bay horse has come from North Africa, (2) that it is specially prone to stripes on the head, legs, and back, and (3) that it and its derivatives are frequently characterized by having a star or blaze on the forehead and by white ' bracelets ' or ' stockings.' These considerations suggest that the star in the forehead of the Libyan horse corresponds to the light- coloured space included by the central arch in the forehead of the zebra, whilst the white ' stockings ' are an extension of the light-coloured band which intervenes between the hoof and the first dark band on the leg of the Somali and Mountain zebras. But as this tendency to stars and ' stockings ' cor- responding to certain markings in the zebras is characteristic of the Libyan hoi-se, it is another indication that that animal has been differentiated in the same region as the zebras. Of course others would explain the presence of ' stockings ' and ^ Ewart, Exper. Contr., p. 41. R. H. 30 466 THE ORIGIN OF THE LIBYAN HORSE [CH. stars as due to in-breeding, and doubtless such markings may well arise from that cause. But there is no reason why they should not likewise occur under natural conditions, for the Quagga which turned bay in an environment similar to that in which the Libyan horses lived, had " a little white in the fore- head " (p. 77), and the Elgin specimen also has white on the face. Prof. Ewart has a photograph of a Mountain zebra with a ' stocking,' whilst the white legs of the Burchell zebra and the Quagga are only more thoroughgoing examples of the same tendency. As the North African horse had a star in the forehead before 1000 B.C., and as the Libyans do not seem to have given any thought to artificial breeding down to the time of Christ, it is difficult to suppose that the stars and ' stockings ' in Libyan horses were the result of in-breeding. Nor is this all, for another argument of great weight is supplied by the colour of the skin. We have already shown that from the beginning of written history white horses are found all across Upper Europe and Upper Asia, and we found reasons for believing that wherever white horses make their appearance in Mediterranean lands, such as Greece, Italy, Sicily, Egypt, and North Africa, they are the result of importations from more northern regions, as was certainly the case with the famous white horses of Dionysius of Syracuse, the strain having been imported by that despot from the Veneti at the head of the Adriatic, whilst the white and grey Arabs found in Egypt and North Africa in modern times are imported thither from Syria and other parts of Asia Minor. It has also been shown that there is no tendency to revert to dun or white colour among the thousands of Pampas horses which are descended from the North African horse. Thus a light colour — dun, skewbald, or white — is an essential charac- teristic of Upper European and Upper Asiatic horses, whilst a dark bay or brown, with a constant tendency to stripes, is the stamp of the Barbary horse and the true Arab of Al-Khamseh. Moreover it lias been clearly shown that the blue-black, antimony-like colour of the skin of Al-Khamseh horses is so marked a feature that it has furnished the generic name for the breed (Kohl) It will be remembered that the skin is of this IV] THE ORIGIN OF THE LIBYAN HORSE 467 dark colour, even on those parts of the body which are covered with white hair, as for instance in the star on the forehead. On the other hand, the skin of white horses is usually white, though in grey horses it is generally dark. But it has been shown that grey horses are especially common where the North African blood is mixed with the Asiatic, whilst the white horse is essentially the child of the regions lying north of the great central mountain ranges of Europe and Asia, where it has been specialised under conditions analogous to those which make the stoat and the ptarmigan turn white in winter, and have perma- nently clothed in white the Arctic hare and the Polar bear. Moreover, it is a fact well known in India that white horses have not the same power of enduring heat as bay and grey. But, as the latter have dark skins, while the former have a white skin, we may reasonably infer that the pigmentation of the skin helps to give the bay and grey horses their power of withstand- ing tropical heat. It follows that the dark skin of the African horse, like that of the negro, has been developed in a hot climate. But this specialisation cannot have been produced by the residence of the Asiatic horse in North Africa from the date of its importation already domesticated into Egypt about B.C. 1500, for it might just as well be argued that a fair-haired, light-skinned people from Europe, if transported to Africa would in an equally short period develope the peculiar skin of the negro. It is of the highest importance to note that the zebras, which have admittedly been specialised in Africa, have dark skins like that of the Libyan horse. The skin of the Burchell zebra Matopo is described by Prof Ewart^ " as dark throughout ; under the white hair the skin is of an iron-grey colour, elsewhere it is nearly black, owing to the pigment in the hair roots." Thus the skin of the zebra is dark beneath his light as well as his dark parts, exactly as the skin of the Al-Khamseh horse is dark under its white markings as well as under the ground colour. We are therefore led in- evitably to conclude that the dark skin of the Libyan horse is the result of its having been domiciled in North Africa for long ages before it Avas ever domesticated. ^ Penycuik Experiments, p. 75. 30—2 468 THE ORIGIN OF THE LIBYAN HORSE [CH. We have seen that according to Sanson many North African horses and ponies frequently lack hock callosities, which have been generally regarded as one of the chief distinctions between Equus caballus and the asses and zebras, and Prof Ewart has pointed out the same peculiarity in the ' Celtic ' ponies. But we have seen that the dark-coloured cross-bred ' Celtic ' ponies are certainly closely related to the dark breeds of Brittany, Auvergne, and Ariege, which beyond doubt owe their form and colour to the mixture of Libyan blood with the indigenous horses of France. We have also seen that the Libyan horses have been crossed from several centuries befoi-e Christ with horses from Europe in order to give them greater strength, whilst it is absolutely certain that a vast proportion of the so- called Arabs, especially those of a grey colour and large size, such as the Gulf Arabs bred by the Montefic tribes of South Arabia, and the horses of Babylonia, have a very large propor- tion of Asiatic blood in their veins. But as the hock callosities are the special feature of both the domestic Asiatic horses and also of Prejvalsky's horse, it would indeed be strange if Arab horses of a coarser type and many North African horses, which have much of the old European strain (derived through Spain) in their veins, should not have inherited hock callosities from their Asiatic and European ancestors. Of course it is possible that, as in Upper Europe there were two distinct types of horse — the slender and the coarse — so in North Africa there may have been a heavy type, corresponding to the old European horses of Solutre as well as the small slender horse, for it would indeed be rash to maintain (especially in view of Prof Osborn's researches) that only one variety of horse had roamed the plains of Libya in remote epochs. But there is not a scrap of historical evidence to show that the Libyan tribes from the Nile to the Atlantic originally possessed any domestic horse except the small slender type, which I term E. c. libycus, whilst the presence of horses of a heavier type in those parts of North Africa which were in contact with Asia, Greece, Italy, and Spain, is fully explained by the abundant evidence of the introduction of the heavy horses of Asia and Europe into that region. IV] THE ORIGIN OF THE LIBYAN HORSE 469 Again, although ergots (fetlock callosities) are generally present in all domestic horses, Prof. Ewart has shown that they are frequently absent in Icelandic and occasionally in Hebridean and Connemara ponies, whilst Captain Hayes ^ has " noticed their frequent absence in pure-bred Arab horses and in thoroughbreds." The same great authority observes that "the nearer a horse approaches the heavy draught type, the thicker is the growth of the callosities on his legs." In view of the complete absence of hock callosities and also of ergots in many horses of the same race, and the fact that such callosities seem universal in Prejvalsky's horse and the Mongolian pony, and that the more nearly a horse approaches the coarse type, the larger are such callosities, and the nearer he approaches the Libyan and Celtic types (in which they are sometimes completely absent), the smaller they become, the evidence indicates that E. c. lihycus either had completely discarded or had a general tendency to get rid of both hock and fetlock callosities. In the Libyan horse and its derivatives — the Arab, the Andalusian, and the English thoroughbred — the tail is different in structure, in its covering and in the manner in which it is carried (Figs. 58, 68, 73, 75) from that of the Prejvalsky's horse and the Mongolian pony (Fig.s. 18, 53). Yet this is no more a mere outcome of artificial breeding since the Christian era than is the bay colour and the star in the forehead, for we have found the same feature in the horses driven by Seti I (p. 217), in those under Cypriote chariots on vases dating from 1000 B.C. (p. 288) and in those ridden by Libyans (p. 243) pourtrayed on the pottery found at Daphnae and dating from 600 B.C. Look at the well-bred Sicilian horse on the coins of Panormus (p. 255). The animal carries his tail in the characteristic fashion that we associate with Arabs, Barbs, and thoroughbreds. We have already seen (p. 143) that since my paper appeared Mr Lydekker, in view of the facts that Hipparion had a deep pre-orbital pit for a gland, that E. sivalensis, an Indian fossil horse, had a rudimentary pre-orbital pit, and that he himself 1 The Points of the Horse, pp. 319-20. 470 THE ORIGIN OF THE LIBYAN HORSE [CH. had found a similar depression in the skull of ' an Indian domesticated horse ' and also in that of Bend Or, had sug- gested that the 'blood-horse,' unlike the 'cold-blooded' horse of Western Europe, may possibly have been the descendant of E. sivalensis. As these pages are passing through the press Mr Lydekker announces^ that he and Dr Ray Lankester have found that a like depression occurs not only in the skulls of the racers Bend Or and Stockwell, but also in those of Eclipse, Orlando, and Hermit, as well as in that of an Arab horse, and that "at present they fail to detect it in any of the ordinary English and Continental horses. It appears to be also lacking in horse-skulls from the drift and turbary of Europe. On the other hand it exists, in a less rudimentary condition, in the fossil horses of India," and Mr Lydekker repeats his suggestion that the ' blood-horse ' is of Indian origin. But I have already shown (p. 143) that it is most unlikely that ' the Indian domesticated horse ' on whose skull Mr Lydekker's argument depends was of pure Indian or Asiatic origin, since all Indian country-bred horses are saturated with so-called Arab blood, and accordingly this skull cannot be taken as a link between E. sivalensis and the Arab. On the other hand we have seen that Hipparion was common in Europe and Africa, that E. stenonis (a species closely related to E. sivalensis), which is found both in Europe and Northern Africa, had a deep pre-orbital depression, and that its later ally, E. quaggoides, had a similar feature, that Mr Lydekker himself (following Dr Forsyth Major) has pointed out the existence of this depression in the now extinct quagga, and also in the skull of an ass, and that Mr Pocock (p. 76) has shown a similar depression in the skull of a male Grant's zebra. Now as all the living Equidae which show this feature and whose origin is known — the quagga. Grant's zebra, and the ass (Nubian) — are African species or subspecies, the occurrence of such a character- istic in any of the living Equidae is a prima facie indication that it is African in origin. But as Mr Lydekker and Dr Ray Lankester have now shown that such a depression occurs in all ^ Times, 14 February, 1905. IV] THE ORIGIN OF THE LIBYAN HORSE 471 the skulls of Arabs and thoroughbred horses which they have examined, we have another clear indication that the ancestors of the Libyan horse had long lived on African soil. But as E. stenonis of Southern Europe and Northern Africa and its later ally L\ quaggoides had a pre-orbital depression as well as E. sivalensis, there is no need to go to India for the fossil ancestors of the Libyan horse, and the true explanation of the presence of such a depression in ' an Indian domesticated horse ' is to be found in the historical facts that the Arabs got their ' blood-horse ' from North Africa, and that for ages these so-called Arab horses have been pouring annually into India and are there crossed with the dun-coloured Asiatic horses. If we could rely on the statement or rather on the reading of the text of Strabo in the passage where he declares that the Libyan horses have longer hoofs than those in any other region, it would further support the view that that animal has been specialised in Africa, where all the Equidae have hoofs of a longer conformation than the horses of Asia. It is certainly a fact of considerable interest that in some high caste Arabs the hoofs are longer than in the quagga\ It would seem there- fore that Strabo's statement had a basis in fact. Nor is it only in colour and other external respects that the Libyan differs from the Asiatic horse. As the cry of the quagga, from which that animal derived its name, was distinct from that of the zebra (p. 73 n.), so the voice of the Libyan horse differs from that of his vulgar Asiatic brother. This is rendered clear by the evidence of Major-General Tweedie already cited (p. 180), who, as before remarked, may be regarded as a hostile witness. In speaking of the Kuhailan horse he thus writes^: "The stallion picketed beside the tent is as good as a sentinel. The first sound of an intruder brings him to attention. Generally he will stamp with one fore-foot and challenge ; not braying like a kadish, but sounding one or two short and sharp notes, to intimate that he will make no terms."..." His gentle salutations of 1 Ewart, E.rper. Contr., p. 21. ^ The Arabian Home, pp. 267-8. 472 THE ORIGIN OF THE LIBYAN HORSE [CH. passing mares are widely different sounds from the bagpipe- like squeals of the Iraki stallion." We have thus the un- impeachable testimony of a first-rate observer who has had exceptional opportunities over a long period of years for study- ing the Kuhailan and the kadishes (common Turkish horses) and the half-bred horses of the Euphrates region, and who, in spite of his preconceived notion that the Kuhailan has been produced by purely artificial breeding from the common horses of Asia, has nevertheless been forced to point out the remark- able difference in voice between the pure-bred Arabian and the horses of undoubted Asiatic lineage. Not only physical characteristics, but also temperament must be taken into consideration in discriminating between species and sub-species, as is clearly shown in the cases of the intractable Mountain zebra and the more docile Burchell species. The difference in disposition between the Asiatic- European horses and the Libyan and its derivatives has been noticed, as we passed in review the breeds of various regions. The Libyan down to tlie Christian era and probably long after rode his horse without a bit, simply guiding it with a nose-band or a switch (p. 240), the Egyptians seem to have used the former contrivance for controlling their chariot-horses (p. 228), the ancient Andalusian horses were noted for their docility (p. 256), and their descendants the Pampas horses of South America, after having regained their liberty for three centuries and a half, are the most docile in South America; the ancient and medieval Irish rode their Hobbies with a mere halter unprovided with a bit, and the Arab to this hour employs only a nose-band to steer his foray steed ; Col. Hamilton Smith has pointed out that frequent crossing of the Turcoman with Arab blood has rendered the well-bred Turkish horse almost as docile as his Arab ancestors, and the extraordinary tractability of the Prussian Trakehnen breed derived from English thoroughbreds and Arabians is a well known feature at the present moment. The ease with which Arabs and thoroughbred horses are broken in compared with cross-bred and inferior horses is a matter of common notoriety amongst horse-breakers, and we observed that horses of Libyan blood have been constantly taught to IV] THE ORIGIN OF THE LIBYAN HORSE 473 dance to music and to perform tricks. On the other hand Prejvalsky's horse is remarkable for its indomitable temper, and the Mongolian pony is famous for its bad manners ; as the horses of Libya were proverbial for their gentleness before the Christian era, so on the contrary the horses of North- western India are specially mentioned by Aelian on account of their violent tempers and the difficulty of riding them, which necessitated the use not merely of bits but of muzzles to control them (p. 153), a fact in itself sufficient to disprove Mr Lydekker's theory of the Indian origin of the Arab horse ; that the horses of Eastern Europe were of a similar tempera- ment is rendered highly probable by the statement of Strabo that the Scythians and Sarmatians were the only peoples who habitually castrated their horses, which they did to make them more easy to manage (p. 25), and we have seen that the term kadish applied to common Turcoman horses seems primarily to have meant a gelding. The legend of the flesh- eating mares of Diomede of Thrace points to a general belief in the savage nature of the ancient horses of that region, whilst in Roman times the horses of Dalmatia and Epirus, which were heavy horses fit for war, were noted for their bad tempers \ and the cross-bred horses of Persia descended from the Upper Asiatic stock were noted for their intractability-. Finally we have seen that cantherius, the Roman term for a gelding, meant originally a pack-horse, and therefore an inferior animal of the Upper European type. We have seen the Libyans, Egyptians, medieval Irish and modern Arabs all riding and driving the Libyan horse without a bit, but on the other hand the Homeric Acheans were using bronze bits to control their dun-coloured horses before 1000 B.C., and bits of a primitive kind made not only of bronze or copper but also of horn and bone are found in the Lake- dwellings of Switzerland and in the pre-historic graves of Russia and Central Asia, whilst the Massagetae in the fifth century B.C. rode their horses with copper bits (p. 130), and ^ Vegetius, Ars Veterin., iv. 6. 6: postea Epirotas, Samaricos, ac Dalmatas, licet contumaces ad frena. - Ibid., nisi labore subigetur assiduo, advei'sum equitem contumax. 474 THE ORIGIN OF THE LIBYAN HORSE [CH. the peoples of North-western India used not nnerely bits but muzzles to control the Upper Asiatic horses. It seems there- fore certain that the invention of the bit at so early a period by the peoples of Asia and Europe was due to the intractable nature of the indigenous horses, whereas the Libyan horse and his descendant the Arabian is ridden to this hour with nothing more than a nose-band. We have in this another specific difference between the two animals. If the Libyans had obtained the horse already in a state of domestication from Asia or Europe, they would probably have borrowed the bit, and it is inconceivable that they could in a short time have influenced the stubborn temper of the Asiatic horse to such a degree that not only their own horses, but all the descendants of these animals down to the present, even after they have been feral for centuries, are stamped with extraordinary docility and good temper. It is clear that the difference in temperament between the Libyan and Asiatic horse has not been acquired under domestication, but is fundamental, and this of itself is a sufficient indication that the Libyan horse is a naturally differentiated species or variety. It is significant that the ' Celtic ' pony, which may be in part descended from a northern branch of the same variety as the Libyan horse, is also remarkable for its docility. We have noted the well known belief that chestnut horses are frequently bad tempered, even when well bred, and reason has been given for thinking that chestnut colour in English thoroughbreds and even in Anazah horses is the outcome of a small strain of Asiatic blood. Now that we realise the funda- mental difference in temper between the Asiatic and the Libyan horses, we at once understand why a cross temper should be a concomitant of chestnut colour. Moreover if, as is commonly held, the ' Arab ' horse is more prolific in Barbary than in other regions where it at present exists — from India to the British Isles — we may reasonably infer that as North Africa is best suited for its propagation, it was there that tlie stock was originally differentiated \ ^ Col. Hamilton Smith, "The Horse," Naturalist's Library, Vol. xii. jd. 214. IV] THE ORIGIN OF THE LIBYAN HORSE 475 Finally, how comes it that no breed of Asiatic horse has ever been so improved by careful selection, or even by the admixture of African blood, as to be able to contend in speed with the latter, unless it be that the North African horse has been differentiated from the Asiatic horses during a very long- lapse of time ? It has been pointed out that the English thoroughbred only really arose when mares as well as horses of North African blood were imported by Charles II, and it is a well-known fact that no three-quarters bred horse has ever beaten a thoroughbred. The astonishing superiority in speed of North African horses over all others seems to indicate that that strain is the outcome of natural specialisation carried on through countless generations. We have now examined the available data for tracing the history of the thoroughbred horse, and we found that the historical evidence put it beyond doubt that it originated in North Africa, from whence it has gradually kept spreading northward and eastward from at least 1000 B.C. The evidence of its characteristic bay colour, the not unfrequent occurrence of stripes on its head, body and legs, its dark skin resembling that of the zebras, its special fecundity in North Africa, all point to its being no merely artificial breed formed under domestication by careful selection by man, but indicate clearly that it is a distinct variety developed during a long succession of time in Libya, under conditions similar to those which have produced some of the zebras with their finely-formed limbs, their dark skin, and striped bodies. The only other conceivable alter- native is that domestic horses from Asia were crossed in North Africa with some variety of striped African Equidae. I men- tioned this as a not wholly impossible alternative when writing in 1902, for the fecundity of zebra-horse hybrids had been held as not impossible by leading experts \ and as an animal deposited by the King in the Zoological Garden, Regent's Park'-, in that year was alleged by some to be the offspring of a horse and a zebra-hybrid. But as Prof. Ewart has now demonstrated the sterility of zebra-horse hybrids, and since the animal sent home 1 Ewart, Guide to the Zebra Hybrids (Edinburgh, 1900), p. 34. " Froc. Zool. Soc, 1902, Vol. ii. p. 225. 476 THE ORIGIN OF THE LIBYAN HORSE [CH. to England from South Africa by Lord Kitchener as the re- puted offspring of a horse and a zebra-horse hybrid, has been proved to be only a zebra-horse hybrid (Fig. 136) S this alterna- tive must be summarily rejected. All zoologists are agreed in regarding the African wild ass as a species distinct from the Asiatic group of asses on the grounds that it is grey instead of being rufous-brown, that it has a shoulder stripe, that its ears are a little longer, and that it has more frequently dark bars on its lower limbs. Mr Sclater holds that the Somali ass is a species separate from the Nubian because it is more grey in colour, has no shoulder stripe, has numerous black markings on the legs, smaller ears and a longer mane, and some make the Asiatic ass into three separate species, whilst those who do not, make them into three or more sub-species or varieties; and some have even made four valid species out of the Burchell group of zebras. Mr Lydekker holds that the Burchell zebra and the Quagga are specifically distinct on the grounds that (1) the pattern on the forehead of the Quagga forms a shorter and more regular diamond than in the Burchell zebra (Bonte quagga) and that in the former the centre of the diamond is a pale stripe with four or five dark stripes on either side of it, whereas in all Bonte quaggas or Burchell zebras the diamond is made up of from five to nine stripes, the middle line being black with from two to four stripes on each side, and (2) on the ground that quaggas may be distinguished from Burchell zebras by the presence on the skull in front of the orbit of a depression (p. 76); and the same authority regards as " subspecifically distinct from the kiang of Tibet " a wild ass from Mongolia which differs simply in colour from the kiang (pp. 44-5). Now as the Libyan horse differs from the Asiatic by being bay instead of yellow-dun, by the shape of its head, by a pre-orbital depression in the skull, by the set of its ears, by frequent tendency to stripes on the back, legs, shoulders and face, by having typically white ' bracelets,' by having usually a white star or blaze on the forehead, by its dark skin, by the absence of hock callosities, by the absence ^ Proc. Zool. Soc, 1903, Vol. i. p. 2, fig. 1. The animal is the offspring of a male zebra and a common pony. IF] THE ORIGIN OF THE LIBYAN HORSE 477 or small size of the ergots, by the length of the hoof, by the covering and set-on of the tail, by its voice, by its disposition, and by its speed, I submit that if the African ass is a distinct i Vfll .t>-e» 1 1 ■ d i i U M I 1 1 ^R f r^^ iP m >J^Bi 1 w^^ u ^^.w^ 1 W wM % P f fflt i. J ^ p^ i: w^ * .♦•• ■ [ ) Fig. 136. Zebra-pony Hybrid. species from the Asiatic, a fortiori, Equus caballiis lihycus must be considered a distinct species, or at least a distinct sub-species. It is for others to decide on the cogency of my arguments. CHAPTER V. SUPPLEMENTARY. THE DEVELOPMENT OF EQUITATION. The antique Persians taught three useful things, To draw the bow, to ride, and speak the truth. Byron, Don Juan. Although the art of eqn itation does not fall strictly within the scope of the present work, yet, as it has been necessary in the course of our historical survey and in discriminating between the Asiatic and Libyan horses to mention the various methods of capturing, controlling and utilising the steed em- ployed by the horse-owning peoples of the ancient world, it will not be out of place if we sketch briefly the chief steps in the evolution of equitation. It is not improbable that amongst the Turko-Tartaric tribes the horse was first domesticated not for locomotion, but, like the ox amongst other races, for the sake of its milk and flesh, and just as at a later stage the cow-keeping peoples began to use the ox to draw the plough and cart, so the Turko-Tartaric race began gradually to use their horses as a means of transport and locomotion. The deeply-rooted love of mares' milk which still characterises Kalmucks and other Tartars seems to indicate that it has formed a substantial part of the nutrition of their race through long ages. The horse was ready to hand on all the vast plains of Upper Asia, where neither wild sheep, goats, nor cattle were to be had. On the other hand neither the Aryans of the Rig-veda nor the Libyans seem ever to have drunk mares' milk, probably because they had possessed cows, sheep and goats, and had been accus- CH. V] THE DEVELOPMENT OF EQUITATION 479 tomed to drink the milk of these animals before they had mastered the horse. The Lasso. The capture of the wild horse was of course the first step towards its domestication. This must have been accomplished either by the capture of foals or of animals not yet full-grown. This could hardly have been effected without employing a rope or cord of some kind, and as in modern times when man desires to domesticate either zebras or feral horses, he always resorts to the lasso, and as I have offered abundant proof of the use of the lasso amongst various peoples of the ancient world (pp. 49, 117, 130, 192), it seems certain that when man first essayed to tame the steed he used a rope with a running noose to ensnare his victim. The Whip. From the inherent tendency in mankind, especially in the lower stages of civilization, to beat unmerci- fully domestic animals, we may without hesitation assume that the lassoed horse was well belaboured with stick and cudgel to cow and subdue him, and as all forms of the whip have grown out of the primitive stick or switch, we are justified in giving the whip precedency over the halter. This is rendered all the more probable by the fact that the Libyans frequently guided their docile horses solely by a switch (p. 240) and that the medieval Irish often controlled the descendants of the Libyan horse b}^ a rod with a crook at one end (p. 389). The Bridle. In each region where the horse was domes- ticated, it seems certain that the first device which can be properly termed a piece of harness was the halter or headstall. For it is most unlikely that man after capturing the horse with the lasso, would have ventured either to mount on the back of his new possession or to yoke him to any kind of wheeled car without some means of controlling him. Thus though the Libyan horses were so docile that the rider could guide them with a switch, yet their masters regularly used halters (p. 240), as did also the medieval Irish. Indeed the straw halters still to be met in some remote parts of Ireland remind us of the rush halters of the Libyans and may be regarded as the most primitive representative still surviving of the earliest step in horse trappings. 480 THE DEVELOPMENT OF EQUITATION [CH. The Bit. It has been shown that the Libyan horse, whether driven under the chariots of Egyptian kings in the second millennium B.C., or ridden by the native Libyans in the centuries before Christ, or in the present day when ridden by Arabs or driven under the Neapolitan carrozzella, was and is controlled by a noseband without any bit, and the evidence is equally clear that from the earliest times the Asiatic-European horses have had to be controlled by a bit at first made of horn and bone and later of copper, bronze and iron, whilst in at least one case it was found necessary in ancient days to muzzle the horses of North-western India. The primitive bits found in Asia, Russia, and in the Swiss Lake-dwellings consist of two side pieces and a cross piece, a type which survived in the bits brought by the Huns into Europe. The earliest literary evidence for the use of bits is furnished by the Iliad, for in one passage the bronze bits are placed between the jaws of the horses. As regards the shape of those bits we have no means of judging, but as they were used to control the dun- coloured horses of Upper Europe brought down by the Acheans into Greece, there is a prima facie probability that they were of the type found in Central Europe. Bits of this type Avere pro- bably known to Xenophon\ for though he holds that it was necessary for a horseman to have two kinds of bits — one with smooth and moderate-sized links, the other with heavy links, with sharp points (in order that when the horse takes the latter into his mouth he may be offended with its roughness and con- sequently let it go), and after he has been trained with the rough bit, he may be ridden with the smooth, yet he emphati- cally urges "that whatever sorts of bits may be used, they should all be flexible, for wherever a horse seizes a rigid bit, he has the whole of it fast between his teeth, as a person when he takes up a stick wherever he lays hold of it, raises up the whole. But the other sort of bit is similar to a chain, for of whatever part of it a person takes hold, that part alone remains unbent, but the rest hangs." The bits used in North-western Europe in the early Iron Age (pp. 96, 98) were more or less flexible, for in the middle they were either single jointed or double jointed (Fig. 45). ' De re equestri, 10. 6. V] THE DEVELOPMENT OF EQUITATION 481 The Chariot. We have already seen that in most regions men employed the horse for draught before they habitually rode upon his back, not because they were afraid to mount him, but because he was either too small to be used effectively as a charger, or because where he was of sufficient size to carry a man easily, it took a long time before men were able to devise weapons and methods of warfare suitable for a man mounted on horseback. For example, peoples who carried large oblong shields, like the Egyptians, the Greeks of the Mycenean (Bronze) Age, or the Assyrians, would have to discard their national shields and adopt a new shape better adapted for a horseman. How unsuited the large oblong shield was for cavalry, is proved by the fact that although the Roman infantry carried the scutum, the cavalry carried the round shield, though even the latter was not the best possible shape for a horse- soldier. Accordingly the Teutonic peoples, such as the Normans, who had once used circular shields, when they began to fight on horseback, devised a shield large at the top and tapering towards the bottom somewhat like a boy's kite. Such are the shields carried by the Norman knights on the Bayeux tapestry, and from this type came the later medieval shield, which through its importance in heraldry has become the conventional idea of a shield in modern times. This shield tapering towards its lower end was admirably suited for horseback, its broad upper part protecting the bearer's body, whilst the tapering lower part fitted down along his thigh, thus obviating the incon- venience arising from a circular shield of any size, the lowest part of which, if it covered the wearer's body, would have had its bottom resting on the front of the saddle ; if to obviate this it was worn to one side, it would leave a considerable portion of the body exposed. Again, tribes whose chief weapon was the bow, like the Scythians and many other nations, would have to learn to shoot from horseback before they could use their horses effec- tively in warfare. On the other hand the archer had little difficulty in shooting with precision from the chariot, as was the practice with the Egyptians (p. 217) and the Hittites (p. 215). R. H. 31 482 THE DEVELOPMENT OF EQUITATION [CH. Where the bow was not the national weapon it was long before man was able to develop weapons adapted for horseback, the warrior simply used the chariot and horses as means of rapid locomotion to meet the foe, whom once reached, he dis- mounted in order to do battle with the arms long used before the advent of the horse. Such seems to have been the case not only in Europe but also in Africa, whilst it seems equally true of the peoples of Asia Minor and of the Vedic Indians, though it is possible that the Turko-Tartaric tribes of Upper Asia may have ridden the horse from the outset. Yet as the Scythians down to the fifth century and later lived in waggons drawn by oxen, it is not improbable that they once lived in waggons drawn by horses, and that it was only when they got cattle at a later time they yoked the more patient and steady-going ox instead. Though indeed the Sarmatians, both men and women, rode on horseback it must not be assumed that they never had passed through a previous waggon-living stage like that of the Scythians, for although the Libyans, men and women alike, all rode on horseback in later times, yet it is certain that in the earlier period they habitually used chariots. The Sarma- tians may therefore once have used the horse under the chariot, as did the Vedic Indians and the Libyans. The Sledge. Hitherto it has been a generally received article of faith that wheeled vehicles and the modern spoked- wheel have had an evolutionary history much as follows. First men fastened to poles their scanty household goods and either themselves dragged them along (or more probabl}^ made their wives do so), when they shifted from one camp to another ; in some cases they may have utilised their dogs for this purpose, as was perhaps the practice of certain North American Indians before they had tamed the feral horses of the prairies, an event which wrought a marvellous revolution in the social life of the Indians of the West, who from being feeble com- munities, dwelling along the banks of the great rivers, which yielded them abundance of fish, and who but rarely could kill a bison, were suddenly metamorphosed into powerful tribes of horsemen faring well on the fiesh of the vast herds of bison. V] THE DEVELOPMENT OF EQUITATION 483 and who made their tents from, and clad themselves and their families in, the warm robes of their noble quarry. After these Indians had become possessed of horses it cer- tainly became their practice on striking camp to pack up all their goods and chattels in the skin tents and tie the bundle on the pole, and trail the poles behind their horses, whilst the dogs were even employed to draw smaller loads on trailing stakes. From such a rude beginning as that last mentioned it seems fairly certain that the Eskimo of the Far North developed their famous dog-sledges. There can be no doubt that the sledge is the first step in the evolution of the wheeled vehicle. The sledge or slide-car^ has played a considerable part in the life of the more remote districts of these Islands down to our days, for such were still in use in Strathglass, Kintail, and elsewhere in Scotland in the years 1863 and 1864. It con- sisted of two shafts, the body being formed by two pieces of wood bent in a semicircle, the ends of which were fastened to the shafts, the one close behind the pony, and the other a little distance behind ; the arches were steadied at the top by a piece of wood running from the one to the other. Thin slats of wood formed the bottom of this primitive contrivance. This vehicle is still in use in the glens of Antrim under the Gaelic name of carr sliimain. The Wheel. It is assumed that the next step was to place beneath such a sledge or slide-car a roller formed out of the cylindrical trunk of a tree, but as Dr Haddon well remarks, " there does not appear to be any positive evidence to render this view absolutely certain." Herr Stephan described a very primitive car that he saw in Portugal : a log is cut from a large tree, the central portion is hacked away so as to leave a solid disc at each end, joined together by an axle. The next step was to form the block wheels of two separate cross- sections of a tree trunk, but fixed firmly on a separate axle- 1 For the following account of the slide-car and other primitive vehicles, as well as block wheels, I am indebted to the admirable statement of the traditional view given by Dr Haddon in The Study of Man (London 1898), pp. 161—199. 31—2 484 THE DEVELOPMENT OF EQUITATION [CH. tree, the wheels not yet revolving on the axle. Then to get greater lightness two or more holes were cut in the solid wheel, the solid portions left being the precursors (as is supposed) of the felloe and spokes made of separate pieces. Finally, the axletree no longer revolves, but is firmly fixed to the cart, and the wheels, now made of spokes, revolve upon it, being kept in their places by linch-pins inserted into a hole in each extremity of the axle. At first sight nothing can be more plausible than this hypo- thesis, but when it is closely examined it must be confessed that there are but few facts to support it, and that those few are capable of other explanations. In the preceding pages we have passed in review the earliest vehicles used by the horse-keeping peoples of the world, and in every case where we had any evidence — in Egypt under the XVlllth dynasty, in India under the Vedic Aryans, amongst Hittites, Assyrians, Persians, Libyans, Mycenean Greeks (Bronze Age), Homeric Acheans (Iron Age), the Gauls of Northern Italy, as well as those of Gaul itself, ancient Britons and ancient Irish — everywhere the chariot wheel is formed of a felloe, a hub or nave, and of spokes ranging in number from four to ten or even twelve. It is therefore clear that the chariot is never found with solid wheels such as are supposed to have been the forerunners of the spoked wheel. Nor is this a matter for surprise. The horse, as we have seen, was throughout early and medieval times used almost solely for war. As speed and mobility were the grand requisite in the war-chariot, it is obvious that solid wheels, such as those used under Portuguese and Chinese ox-carts, would have rendered the vehicle useless for war. We may, therefore, safely con- clude that from the first the war-chariot never had block wheels. But it may be said that although the war-chariot from the first was fitted with spoked wheels, nevertheless the solid wheel had long preceded it, having been invented for purposes of agriculture, and that doubtless the ancient ox-cart, which was not built for lightness and speed like the chariot, was furnished with clog wheels. Yet if the reader will look at the picture of V] THE DEVELOPMENT OF EQUITATION 485 the Thracian ox-cart (p. 106), the oldest representation of any such vehicle that has reached us, his faith in this second hypo- thesis will be shaken on discovering that this ox-cart runs on four-spoked wheels. If it be said that this cart belongs to a comparatively advanced period, and that in earlier days, when agriculture was in its infancy, the carts used had solid wheels, I at once reply that amongst two at least of the great races which at the dawn of history had domestic horses — the Libyans and the Turko-Tartaric peoples — agriculture was scarcely, if at all, practised, for both were essentially nomadic ; whilst though the Scythians in later times at least used four- wheeled waggons to convey their families, the Libyans never used either ox-cart or ox-waggon for that purpose. It must also be clearly borne in mind that primeval agriculture had no need for the cart. Corn was not bound in sheaves as with us and carried home on carts or waggons. The ears of corn only were snipped off, gathered into baskets and carried to the threshing-floor or garner. Indeed, in the days when North Africa was one of the chief granaries of Rome, a basketful of corn-ears was placed as the symbol of Africa on a coin of Hadrian. Again, as there was no manuring in the common field system, there was no need of a cart for manure. The functions which in our minds are so inseparably associated with carts and waggons were in the earliest stages of society discharged by human beasts of burden, as they still are over a large part of Africa, and later on by pack-animals, as they were in medieval Europe and are in wide regions of the earth down to this very hour. These facts sufficiently refute Dr Hahn's' theory that wheeled vehicles did not arise from the sledge fitted with a roller, for in that case (said he) wheeled vehicles would have arisen wherever rollers have been employed. He main- tains that the waggon arose only in the district from which agriculture originally spread, which he assumes to be Greece. He believes that the waggon was primitively a holy imple- ment, consecrated to Demeter, the great goddess of agriculture and fertility, and that it only subsequently became a secular 1 Demeter und Baubo (1896, Liibeck); Haddon, op. cit. pp. 170-1. 486 THE DEVELOPMENT OF EQUITATION [CH. farm implement. He avers that the waggon came into being because miniature wheels in the form of the common spindle whorls were already in general use. Dr Hahn points out that he deals only with the four-wheeled ox- waggon, which was used for religious purposes. But, as it has just been shown, there is no evidence for the use of either the two-wheeled ox- cart or the four-wheeled ox-waggon in primitive agricultural communities, whilst it seems certain that the Libyans, who never used ox-carts at all, had invented a very light form of spoked wheel at least by B.C. 1500. Dr Haddon' has already pointed out that there is no reason to believe that agriculture was discovered only in some area of Eurasia, and that the art thence spread over the greater part of the habitable globe, and " it seems more in consonance with what we know of the history of sacred institutions and implements that the waggon had an industrial origin and that it may well be that it arose in close connection with agriculture." But though it may well have arisen in connection with agriculture, as Dr Haddon says, yet it may have come into use at a comparatively late period, and long after the invention of the war-chariot. The ox-cart or ox- waggon indeed was certainly in use in Greece in the seventh century B.C., for a certain Argive lady wished to go to the festival at the Argive Heraeum in a waggon and pair. When the oxen did not arrive in time from the pasture, her two sons, both distinguished athletes, yoked themselves to the waggon and drew their mother to the temple". But if Dr Hahn's theory is sound, we ought to find the ox-waggon not merely in the early classical period, but in Homer. Yet neither the two- wheeled ox-cart nor the four-wheeled ox-waggon appear in the poems, though the four-wheeled mule-waggon plays a conspicuous part. In such a vehicle Priam brought with him the rich gifts with which he set forth to the camp of the Acheans to ransom Hector's body from Achilles. " Thus having spoken fleet Iris departed from him and he bade his sons make ready the smooth-wheeled mule-waggon {amaxa) and bind the wicker carriage thereon ^ 1 The Study of Man, pp. 170-2. - Herod, i. 30. 3 II XXIV. 188-90. v] THE DEVELOPMENT OF EQUITATION 487 " Thus spake he and they fearing their father's voice brought forth the smooth-running car, fair and new, and bound the body thereof on the frame ; and from its peg they took down the mule yoke, a boxwood yoke with knob and well fitted with guiding-rings ; and they brought forth the yoke-strap, of nine cubits with the yoke. The yoke they set firmly on the polished pole on the rest at the end thereof, and slipped the ring over the upright pin, which with three turns of the strap they lashed to the knob, and then belayed it close round the pole, and turned the tongue thereunder. " Then they brought from the chamber and heaped on the polished car (apene) the countless ransom of Hector's head and yoked strong-hoofed mules that work in harness, which on a time the Mysians gave to Priam, a splendid gift, but for Priam they yoked the horses that the old man kept for his use and tended carefully at the polished crib\" That Priam's mule-car had four wheels is shown by another passaged In the Odyssey^ we hear of " two-and-twenty excellent four-wheeled waggons (cwiawai)," and apparently such too was the vehicle in which Nausicaa set out with her maidens to wash linen in the rivers It was " a high waggon (apene) with good wheels and fitted with an upper frame." " Without the palace they made ready the smooth-running mule-wain (avmcca) and led the mules beneath the yoke and harnessed them under the car, whilst the maiden brought forth from her bower the shining raiment." From these extracts it is, clear (1) that the amaxa (waggon) is identical with the apene or car commonly drawn by mules also in the classical period. Thus in B.C. 500 a race for mule- cars was established at Olympia (abolished. in B.C. 444). Anaxi- las, the despot of Rhegium, not many years after won the prize with his mule-car, and commemorated this event, as well as the fact that he was the first to introduce the hare into Sicily, by placing his victorious mule-car and the hare as types on his coins (Fig. 137). It will be seen that this mule-car has four-spoked wheels, and, as the epithet "easily running" is applied to the 1 II. XXIV. 265-80. 2 jfojVZ. 324. 3 Od. IX. 242. * Od. vi. 68 sqq. 488 THE DEVELOPMENT OF EQUITATION [CH. mule-car in the Homeric passages just cited, and we also learn that Nausicaa's mules trotted briskly off with the waggon, it is most unlikely that the Homeric waggon had block wheels. Now, as the Homeric mule-car, the classical mule-car, and the Thracian ox-cart about B.C. 500 all have spoked wheels, there is therefore no evidence for the existence of the use of solid wheels under either mule-car or ox-cart in early times in the countries round the Aegean. If such did once exist there before the invention of the spoked wheel, it must have been at a time anterior to the appearance of the horse both on the monuments of -Egypt and on the tombstones of Mycenae. But it has just been shown that the use of the ox-cart for agriculture at so early a period is extremely unlikely. Accordingly, so far from the ox-cart with solid wheels being the precursor of the chariot, it is most likely that the latter was the first to be in- vented for purposes of war, and that later a stronger and cheaper form of vehicle for oxen was modelled after the chariot for everyday use — a simple lalatform on which a wicker creel or crate like that of the Homeric mule-car could be placed. Of course it will be said that block wheels survived in the British Isles down to our own time, — that in 1775 goods were conveyed about Dublin on carts furnished with solid wheels about 20 inches in diameter, that solid-wheeled carts may still be seen in the North of Ireland from Donegal to Down, and that two kinds of block-wheeled carts were in use in Inverness about 1730, both of them being simply modifications of the slide-car still surviving in Antrim, with wheels about a foot and a half high, but which were soon worn very small. Yet it must not be assumed that such wheels were the first kinds known in all these localities. It is most improbable that any kind of wheeled cart was in use in Ireland in early times, yet there were war-chariots with spoked wheels in Ireland at the time of Christ. Again, it is on record that in Borrowdale wheeled vehicles did not make their appearance until about 1770; and when these novelties did reach the lakes they were clumsy and awkward in character. Clog- wheels were the first type used on farm carts, yet spoked wheels had long been in general use all over England, and had been known in the island from V] THE DEVELOPMENT OF EQUITATION 489 before the Christian era. It is most important that cheap sub- stitutes be not taken for genuine survivals of a primitive type. To the former category we may assign not only the block-wheels of Borrowdale, Inverness, and Ireland, but also certain wheels seen on Greek vases consisting of a felloe and two parallel cross-pieces, crossed by another at right-angles (a variety in which some have recognised the first step towards a spoked wheel). To the same category I would likewise refer the Portuguese wheels made out of a single piece of wood in which two elliptical holes are cut, the wheel itself being clamped with bands of iron. This wheel and others like it have been supposed to be a first step towards a spoked wheel, but they are rather to be regarded as cheap and clumsy substitutes, as are also the solid wheels built up of three pieces of plank common Fig. 137. Coin of Messana. in Galicia, the Canaries and amongst the Zuni Indians in Mexico, who have borrowed them from the Spaniards. These wheels are regularly clamped together by iron bands, although in Mexico they are said to be sometimes unshod. In all cases of solid wheels the wheel is fixed on the axle, and does not revolve on it. Yet in the Florentine and Homeric chariot the wheels play freely on the axle. It can therefore hardly be maintained that we have genuine survivals of the first stage in the evolution of the wheel in the Portuguese and Spanish waggons with their revolving axles and wheels fixed to the axle, for it is clear that the principle of the wheel revolving on the axle has been known from an age far anterior to any evidence of the existence of an ox-cart with solid wheels. But it is not in itself probable that solid wheels were evolved at a date when iron was not yet known, and copper was com- 490 THE DEVELOPMENT OF EQUITATION [CH. paratively scarce in most parts of the ancient world. The hewing of a section of a tree trunk two feet in diameter to serve as a solid wheel rather indicates a period when tools of a superior kind were available, or otherwise the task would have been so difficult that man would probably have resorted to some other method of shaping discs on which to set the frame of his car, though of course it is not utterly impossible that man by dint of hacking with a bronze, copper, or even a stone axe, could have managed to rough-hew a pair of solid wheels con- nected by an axletree out of a tree-truuk. But as such wheels could never have been of practical use for war-chariots, and the Libyans, who had never any ox-cart, had devised for themselves long before they had metal in any quantity beautifully light chariots, in the structure of which no metal was employed, it is most unlikely that their spoked wheels were evolved from an antecedent block wheel. In the Florentine chariot the wheels are four-spoked and are 38 inches in diameter, both felloe and spokes being made of rods about one and a half inches in diameter. The spokes fit into a hollow hub formed of a wooden cylinder about nine inches long, with fairly thick walls through which the axle runs. The whole structure of this chariot is that of wicker and meshwork (p. 225). It is therefore far more probable that the spoked wheel was an adapta- tion from a circular piece of wicker-work, such as might be used for a shield or for some other purpose. The simplest form of such a circular frame consists of a rim strengthened and kept in shape by two other rods crossing each other at right-angles, thus forming four radii or spokes. The four-spoked wheel is found in the oldest representations of the chariot in Egypt, in Crete, Cyprus, and on the mainland of Greece. That it was considered by the Greeks the most ancient form of wheel is shown by the fact that in the myth of Ixion that miscreant is represented as bound to a four-spoked wheel. As the wheels of the Florentine chariot revolve on the axle, there is no reason to believe that such wheels only came into existence after a long period during which block wheels fixed to the axle had been in continuous use. We may reasonably conclude that the light war-chariot was invented long before the ox-cart or mule- V] THE DEVELOPMENT OF EQUITATION 491 cart, and that so far from the spoked wheel having been evolved from the block wheel, the converse is really the case. In the tomb of the father and mother of queen Teie, the wife of Amon-hotep III and the mother of Amon-hotep IV, Mr Theodore M. Davis has just found a pleasure chariot broad enough to hold two persons, richly painted and encrusted with gold. The leather work belonging to it is still as fresh as when it was first made. It is fitted with six-spoked wheels still covered with their wooden tires*. This fresh discovery shows that the conjecture of Mr Carter and Mr Newberry that the wheels of the chariot of Thothmes IV had been fitted with metal tires is wrong. There is no reason now to doubt that it, like the newly discovered chariot and that at Florence, had no metal shoeing on its wheels. We have seen that the mule-cart in Homer is fitted with wheels of lightness and elegance, and there is really nothing to distinguish them from those of the chariot. We have a full description of the Achean chariot in the Iliad, for we cannot doubt that the chariot of the goddess Hera is a faithful copy of those used by her worshippers, save that the car of the immortals is represented as being made of precious metals. "So Hera the goddess queen, daughter of great Cronos, went her way to harness the gold-frontleted steeds ; and Hebe quickly put to the car (ochos) the curved wheels of bronze, eight-spoked, upon their axletree of iron. Golden is their felloe, imperishable, and tires of bronze are fitted there- over, a marvel to look upon ; and the naves are of silver, to turn about on either side. And the body of the car {diphros) is plaited tight with gold and silver straps, and two rails (antux) run round about it. And a silver pole stood out therefrom ; upon the end she bound the fair golden yoke, and set thereon the fair breast-straps of gold, and Hera led beneath the yoke the horses, fleet of foot, and hungered for strife and the battle- cry'-" This description when compared with actual specimens found in Egypt gives us a very clear view of the structure of the chariot, the plaiting with straps of gold and silver at once 1 Times, 10 March, 1905. - II. v. 721 sqq. 492 THE DEVELOPMENT OF EQUITATION [CH. recalling the floor of the Egyptian chariots with their plaited leather meshvvork (p. 225) in which we may recognise the first step towards leathern springs. The Harness. The elaborate account of the harnessing of Priam's mules above cited when taken in conjunction with other passages in Homer, with the harness found in the tomb of Thothmes IV, and representations of Egyptian chariots (Fig. 68), with the description of the Assyrian chariot-harness (pp. 195-196), with the Hittite (Fig. 67), with the Persian (Fig. 61), and with numerous representations of Greek chariots in classical times, enables us to form a clear idea of the nature of the harness used in early times and the method of attaching the horses to the car. Whilst the harness seen on the monument of Seti I and found in the tomb of Thothmes IV belongs to a date anterior to Homer, that seen on the Assyrian monuments falls several centuries after that period. It is therefore but natural to find that whilst the Egyptian and Homeric horses are attached only by breast-straps, the Assyrian have also elaborate body bands which maybe taken as the forerunners of the saddle or straddle of modern harness. The Rein-rings. Attached to the yoke of the Florentine chariot (Fig. 69) are seen two Y-shaped objects, which must have hung down from the yoke or some other part of the harness or chariot. They are of wood, and the arms of each are pierced with holes near the extremity. The height of the whole is seven inches, the width from hole to hole not being more than four inches, and the tail ends in a large round bone stud, one and a half inches in diameter, which shows that the object was not meant to be inserted into a hole in any part of the yoke or chariot. They cannot have been collars (as com- monly supposed) to rest on any part of the horses' necks, for they are obviously too small for that purpose, and are quite different from the actual collars found in the tomb of Thothmes IV figured in Mr Davis's publication (p. 227). Associated with the pairs of bronze bits not unfrequently found in Ireland are usually pairs of spur-shaped bronze objects (Figs. 138, 139). V] THE DEVELOPMENT OF EQUITATION 493 Each of" the specimens here shown (Figs. 138, 139) is one of a pair found along with the pairs of bronze bits, examples of which are given (p. 98, Fig. 45). These objects could not have stood upright on the yoke or anything else, for the two arms are neatly rounded off (see Fig. 138 B, where an end is shown in detail): again, the tail could not have been inserted into a hole in the yoke or anything else, for in both the examples Fig. 138. Ancient Irish Rein-ring (all-dual) ? figured the end is ornamented, more especially in Fig. 138, where, as will be seen (Fig. 138 a), it is beautifully ornamented in the ' late Celtic ' style, as are also the ends of both arms, one of which, as already stated, is shown in detail (Fig. 138 b). As this curious piece of metal work could not have stood up on either end, and as neither end was meant to be inserted into any other object, clearly these mysterious implements were suspended with the tail hanging free, as is demonstrated by the 494 THE DEVELOPMENT OF EQUITATION [CH. fact that the extremities of the two arms are always furnished with means of suspension. Some are perforated close to the extremity (Fig. 138 b), as if meant to be suspended. In most specimens, however (as in Fig. 139), there is a groove on the inside of the upper portion of each arm extending for some distance, and this groove is crossed a little below the end by a bronze loop (Fig. 139 b) which could be slipped on a hook, thus enabling the whole to be attached to the lower side of some object such as the yoke. In specimens of the second class where the loop has been accidentally broken off, its place has been supplied by a hole bored right through, which is plainly meant to admit the passage of a hook. When once the hooks fastened into the under side of the yoke were slipped through the loops or holes in the extremities of the spur-shaped objects they could not easily jump off. The analogy between these Irish bronzes and the primitive wooden pair found along with the Florentine chariot is very close, and they would both seem to have fulfilled a like function. We may take it that the harness of the Homeric mule-car was practically identical with that of the Homeric chariot, for otherwise the harness of Priam's chariot would probably have been described. Now, we are told that the yoke was 'well- fitted' with oiekes, literally 'steerers,' which are explained by the ancients themselves as "a kind of rings through which the reins were passed." This fact shows that from very early times it was found necessary to have some kind of rings attached to some part of the harness through which the reins might be passed, and thus kept them in place and free from entanglement with the horses' manes, and give more power to the charioteer. The reins in the Assyrian chariot seem to have passed through some such contrivance fastened to the under-side of the yoke (Fig. 62). In the Egyptian, Mycenean, Hittite, and Assyrian chariot the yoke, as was the case in Homer, was the only part to which such 'steering-rings' could be attached, though the reins might have been passed under a strap going round the horse's body, as is the case on the monument of Seti I (Fig. 68). As apparently no such rings were ever fitted into the yoke of the Florentine chariot, any such contrivance for guiding the V] THE DEVELOPMENT OF EQUITATION 495 reins must have been either suspended from the yoke or attached to it by strings. The two Y-shaped objects were probably suspended from the yoke, much as they are at present (Fig. 69), and through them the reins were passed and kept in their place. The more elaborate bronze objects found in Ireland probably served a like purpose, for as already argued they must certainly Fig. 139. Ancient Irish Rein-ring (rtZ/-fZi/a/) ? have hung down. This view can be supported from the oldest Irish texts — in these the name for reins is all : thus Laeghaire Buadach's chariot " had two pliable beautiful alls," and it had likewise "two rich yellow all-dualach," literally, "two rich yellow rein-loops^' or "rein-rings." Dr Sullivan thought that these rings were on a straddle, but as it is very unlikely that the Irish chariot-horses wore straddles, it is far more 496 THE DEVELOPMENT OF EQUITATION [CH. probable that the all-dualach were attached in some way to the yoke, as was the case with the Homeric steerers. We may therefore conclude that the curious wooden objects on the Florentine yoke were really a primitive contrivance for keeping the reins in place, and that the Irish implements are simply more elaborate forms of the same type. The Scythed Chariot. The addition of scythes probably prolonged the use of the chariot for war, as such " scythe-bear- ing chariots " became a formidable arm when driven against bodies of footmen. Thus, although owing to 'villainous salt- petre' the medieval knight with lance in rest has long departed, nevertheless lancer regiments still linger on in the armies of modern Europe, partly as a survival, and partly because they are found useful in certain conditions of modern warfare. Certainly, whenever we hear of the employment of war- chariots at a late period they are usually described as 'scythe- bearing.' The chariots used by the Persians at the battle of Cunaxa were so equipped, and chariots still more elaborately armed were employed in Syria at a much later date. Thus in the great battle between Eumenes of Pergamus and Antiochus of Syria, the latter placed in the front of his line four-horse chariots, furnished both with scythes and spears. Spears fastened round the pole projected like horns ten cubits in front of the yoke to transfix everything that came in the way; two scythes were attached to each end of the yoke, one fixed on a level with the latter, the other sloping towards the ground, the former being meant to cut away every obstacle from the side, the latter to strike foes already prostrate, or endeavouring to escape by passing under the more elevated blade, whilst from each axle two other scythes extended set at different angles like those attached to the yoke. But Eumenes managed to stampede the chariot-horses of his adversary, which turned round and dashed into their own ranks, and Antiochus suddenly found his army panic-stricken and routed by the engines which he had devised for the destruction of his foes^ Doubtless such occurrences as this were always liable to 1 Livy, XXXVII, 41, V] THE DEVELOPMENT OF EQUITATION 497 occur and accordingly we hear little of scythed-chariots in the armies of the civilised peoples of the Mediterranean from this time forth. The war-chariot only lingered on among the more barbarous peoples of the north-west of Europe, and amongst the remote Libyan Pharusii, by the latter of whom also it was now furnished with scythes. Riding'. When men began to ride regularly on horseback at first they sat simply on the bare back of the steed, which in Asia and Europe, as we have seen, was from the first controlled by some form of bit, though the Libyan used at most but a noseband. The Greeks of the fifth century paid great attention to the shape of the bit, as is made clear by the elaborate directions respecting it given by Xenophon ^ The Horse-cloth. The first step towards a saddle was naturally some kind of cloth placed on the horse's back for the greater comfort of the rider. The Assyrians had already made this first advance by the eighth century B.C. (Fig. G4), and it was certainly known to the Greek settlers in Egypt by B.C. 600 (Fig. 72), whilst it had become a fully recognised part of the equipment of the Greek and Macedonian horse-soldier (Fig. 87) by the beginning of the fourth century B.C., if not earlier. The earliest literary testimony is that of Antiphanes-, the comic poet, who began to exhibit plays in 387 B.C., and his contemporary Xenophon^ The former speaks of "the coverlet for a horse." But bare-backed riding was still regularly practised, as we know from the latter writer'*, and apparently the jockeys in the races at the great festivals of Greece rode bare-backed (Fig. 86), as is the case in Mongolian horse-races held at temple feasts at the present hour (p. 139). The cloth known as an ephippion (horse-cover) had come into universal use amongst the Romans by the time of Caesar (cf. p. 114), though the German tribes considered it disgraceful and a mark of laziness to use it, and were always ready, riding ^ Be re equestri, 10, 6. - Meineke, Com. Fragin., iii. p. 3, to ecpiirwiov aTpQ/xa. ' Eq. VII. 5, TO e