\J LIBRARY UNIVER5ITY3>^ PENNSYL\^\NIA FAIRMAN ROGERS COLLECTION ON HORSEMANSHIP m Digitized by tine Internet Arciiive in 2009 witin funding from Lyrasis IVIembers and Sloan Foundation http://www.archive.org/details/horseitsrelaOOIyde THE HORSE AND ITS RELATIVES ffer^Ge ORG e -tfe 7q h b € FL '^ ^ s» OPrimnnWr - '^^TT^^^W^SZT' 0 b'i^^^^^^^^ ►3 W/I^nf^^^^^^ ^^> '1 w A mWSoRiES^P^^ ir 52 ^r^^M^«^5^ 5^1 © -3 wl K®wj]g£>-^^S^^^^^ . ^\ffcS H 3 s-^f- s^sfr-^ncz^s* •^1^1 a. ^3Si^!!5®2<^^^^^i 9 P''^RU5Kia-Tl;?T,OUj; e\^ i^ THE HORSE and ITS RELATIVES BY R. LYDEKKER, F.R.S. AUTHOR OF 'THE GAME ANIMALS OF AFRICA," "THE GAME ANIMALS OF INDIA, ETC.," "THE DEER OF ALL LANDS," "A GEOGRAPHICAL HISTORY OF MAMMALS," "THE OX AND ITS KINDRED," ETC. LONDON GEORGE ALLEN & COMPANY, LTD. 44 & 45 RATHBONE PLACE 1912 [All rights reserved] printed by Ballantyne, Hanson Sy Co. At the Ballantyne Press, Edinburgh PREFACE The following popular, and yet, I hope, scientifi- cally accurate, account of the natural history of the more important representatives of the horse family, inclusive of the older domesticated breeds and its extinct forerunners, will, I venture to think, appeal to a large circle of readers. For breeders, racing men, antiquarians, naturalists, and big-game hunters ought all to find something of interest. It should be emphasised that only the natural aspect of the subject is dealt with, such side-issues as the legendary history of the horse, horse-sacrifice, the acquisition and development of the art of riding and driving, the training and management of horses, being left untouched. ERRATUM Page 29, line 14 from top, for ' cement ' read ' dentine. The Horse and its Relatives. VK) '^EW BOLTON I i^' PREFACE ^ : The following popular, and yet, I hope, scientifi- cally accurate, account of the natural history of the more important representatives of the horse family, inclusive of the older domesticated breeds and its extinct forerunners, will, I venture to think, appeal to a large circle of readers. For breeders, racing men, antiquarians, naturalists, and big-game hunters ought all to find something of interest. It should be emphasised that only the natural aspect of the subject is dealt with, such side-issues as the legendary history of the horse, horse-sacrifice, the acquisition and development of the art of riding and driving, the training and management of horses, being left untouched. Several difficult and debatable points are pur- posely left undecided, as I have preferred to quote the various opinions expressed by different writers, rather than to assert my own views. For the illustrations I am especially indebted to the Trustees of the British Museum ; but I have also to express my acknowledgments to the Duchess of Bedford, Prof. R. S. Lull, Dr. E. L. vi PREFACE Trouessart, Mr. Theodore A. Cook — both in his private capacity and as editor of The Field — Prof. J. C. Ewart, and several other friends and corre- spondents. Since the text was in type, Mr. R. I. Pocock has pointed out i^The Field, Jan. 20, 191 2, p. 143) that the aperture of a scent-gland situated on the posterior aspect of the hind-foot of the Indian rhino- ceros occupies a position very nearly similar to that of the ergot in the foot of the horse (infra, p. 41). " The orifice of this gland," he writes, " is placed suggestively near the spot corresponding to that occupied by the ergot in horses, and despite the accepted view that the ergot is a sole of the foot, the possibility of its representing an aborted gland may be wisely borne in mind. Its general resem- blance to the warts or chestnuts on the leofs of horses, which most authorities regard as degener- ated glandular structures, is quite in keeping with this suggestion." R. LYDEKKER. Harpenden Lodge, Herts, January 191 2. CONTENTS I. The Zoological Position and Structure of THE Horse ..... II. The Wild Tarpan and its Relations III. Horses and Ponies of the British Islands IV. Some Foreign Breeds V. The Arab Stock VI. Feral Horses .... VII. The Kiang and Onager Group VIII. Zebras and Quaggas IX. The Ass X. Mules and other Hybrids XI. The Extinct Forerunners of the Horse Index I 71 117 136 150 170 176 187 215 225 239 283 ILLUSTRATIONS PLATES ♦ LATE FACING PAGE I. Fig. I. Cannon and Splint Bones of Shire Horse 14 (Brit. A/us. Gitide to Horse Fa7nily) Fig. 2. Bones of Fore and Hind Feet of " MiOHIPPUS "... [After Lull) II. Skeleton of Thoroughbred Stallion "Eclipse" i5 (Cook, " Eclipse and O' Kelly ") III. Fig. I. Skeleton of Fore-limbs of " Eclipse " 18 Fig. 2. Skeleton of Hind-limbs of "Eclipse" s, scapula ; h, humerus ; u, ulna ; r, radius ; c, carpus (knee) ; cati, cannon-bone \ pk, phalanges ; /, pelvis ; /, femur ; /, tibia ; ca, calcaneum (hock) (Both from Cook) IV. Fig. I. Skull of Shire Stallion ... 22 (Brit. Mus. Guide to Horse Family) Fig. 2. Skull of Ouagga . . V. Fig. I. Right Upper Cheek-teeth of Tarpan 34 Fig. 2. Right Upper Cheek-teeth of Arab „ p.7.-p.i,, premolars; »/.i-;«.3, molars ;/, anterior pillar ; hy, posterior pillar VI. Fig. I. Hind-foot of Horse showing Ergot 44 Fig. 2. Chestnuts on Legs of Horse . . „ (Both from " Brit. Mus. Guide to Horse Fafnily") ILLUSTRATIONS PI-ATE FACING PAGE VIL Fig. i. Frontlet of Horse with Horn-like Prominences Fig. 2. Prehistoric Tarpan or Wild Horse {Brit. AIiis. Guide to Horse Family) VHL Fig. i. A Mongolian Mare ... Fig. 2. A Tarpan Mare .... (Brit. Mils. Guide to Horse Family) IX. Fig. I. Skull of Tarpan Mare Fig. 2. Skull of Arab Mare. {Both from " Brit. A/us. Guide' to Horse Family ") X. Fig. I. A Norwegian Dun Stallion Fig. 2. A Mongolian Polo Pony . {From " The Field") XI. Fig. I. New Forest Ponies . Fig. 2. Shetland Ponies {Both from photos, by C. Reid) XII. Fig. I. A Suffolk Stallion . Fig. 2. A Shire Stallion {Photos, by C. Reid) XIII. Fig. I. A Percheron Stallion {From a French Journal) Fig. 2. A Belgian Stallion . {From a Brussels Journal) XIV. Fig. i. The Darley Arabian. {From the Picture at Aldby Park) Fig. 2. The Thoroughbred Stallion "Per Simmon" {From " The P'ield") 6o 8& 94 104 118 128 13& 152 ILLUSTRATIONS xi PLATE FACING PAGE XV. Fig. i. The Kiang Fig. 2. KoBDO Onager .... {Both from. ' ' Brit. Mus. Guide to Horse Family ") XVI. Fig. i. Grevy's Zebra .... [Brit. Mus. Guide to Horse Family) Fig. 2. Head of Grevy's Zebra XVIL Fig. i. The Quagga .... (Photo. Young) Fig. 2. Matabili Bontequagga (Pkoto. Prof. J. C. Ewart) XVIII. Fig. i. Kilimanjaro Bontequagga (Photo. The Duchess of Bedford) Fig. 2. Masai Bontequagga . (Brit. Mus. Guide to Horse Family) XIX. Fig. i. Skin of Foa's Zebra . Fig. 2. Skin of Kilimanjaro Bontequagga (Photos. Dr. E. L. Trouessart) XX. Fig. i. The Zebra 178 188 194 200 21O' 214 Fig. 2. Nubian Wild Ass ... [Both from " Brit. Mus. Guide to Horse Family") XXI. Fig. i. Hybrid Bontequagga and Pony Foal and Dam 234 (Photo. Prof. Ewart) Fig. 2. Hybrid Zebra and Onager . . ,, XXII. Bones of Fore-feet of Extinct Fore- runner OF the Horse .... 246' (Brit. Mus. Guide to Horse Family) xii ILLUSTRATIONS P'-ATE FACING PAGE XXIII. Skeleton of " Hippidium " .... 252 [Brit. AIus. Guide to Horse Family) XXIV, Fig. i. Skull of " Onohippidium" . . 260 Fig. 2. Molars of " Equus," "Hippidium," AND *' HiPPARION " „ {Both frotn " Brit, A/us. Guide to Horse Family") TEXT-FIGURES PAGE Bones of Fore-leg of Horse and Rhinoceros . . 6 (Brit. AIus. Guide to Great Game) Skull of Pig-like Animal, "Elotherium" , , 20 Molars of Single-toed and Thres-toed Horses . ^^ The Ancestors of the Horse and its Relatives compared ........ 240 (After Lull) Milk-molars and Premolars of the Extinct " Meryc- Hippus" 255 (After Lull) Bones of the Fore and Hind Feet of the Extinct "Hipparion" 257 (After Lull) Crown Surface of Upper Molar of the Extinct "Hipparion" 258 (After Lull) Bones of the Fore and Hind Feet of the Extinct "Hypohippus" 261 (After Lull) Upper Molar Tooth of " Anchitherium" . . 267 Bones of the Fore and Hind Feet of the Extinct "Eohippus" ........ 274 (After Lull) Bones of the Fore and Hind Feet of the Eocene "Phenacodus" 276 (After Lull) THE HORSE AND ITS RELATIVES CHAPTER I THE ZOOLOGICAL POSITION AND STRUCTURE OF THE HORSE The difficulty which occurs in the case of the ox^ as to what is the proper EngHsh designation of that animal does not arise in the present instance, for although we not infrequently speak of a horse, as distinct from a mare, there seems little doubt that the former term is really a species-name, and therefore applicable to both sexes of Eqtms cabalhts, as the domesticated horse of Europe was called by Linnaeus. As to the origin of the name horse — the equiva- lent of the Anglo-Saxon hors, the Frisian hars or hors, the German ross, the Italian rozza, the Old Saxon and Old German hros, and probably the Persian ghor and the Hindustani ghora — there has been some difference of opinion. It has been stated, for instance, to take origin from the Sanskrit ^ See The Ox audits Kindred, by R. Lydekker, London, 191 2. A 2 THE HORSE AND ITS RELATIVES hresk, signifying to neigh, so that the horse means the neighing animal.^ This derivation is, however, not admitted in the Century Dictionary, where the name is stated to be the equivalent of the Anglo-Saxon hors, which signifies swiftness, and is connected with the Latin currere, to run ; the English term horse thus meaning the running animal. The Sanskrit name of the species is agva, which appears to be the equivalent of the Hebrew sus, the Greek hippos (with its diminutives hipparion and kippidion), and the Latin equtcs. Another series of names for the horse is represented by the Greek kaballos, the Latin caballus, the Spanish caballo, the Italian cavallo, and the French cheval. In addition to these, we have the German pfe^'d and the Dutch /^^rflf. There is also the English name pony, for a small horse, which may possibly be connected with the undermentioned /^^//z^j-. As is commonly the case with domesticated animals, there is also a large series of names to denote the two sexes and the young of the horse. Stallion, for instance, the English name of the male of the species, is equivalent to the modern French ^talon, the old French estallon, and the Italian Stallone, or equus stallonis, the horse at stall. Mare, the designation of the female, is derived 1 See The Student's English Dictionary, by J. Ogilvie, London, 1865. POSITION AND STRUCTURE OF HORSE 3 from the Anglo-Saxon 7nyre or Tnerc, a word which appears to have been originally connected with increase, but in a later sense indicated the female of the strong animal. As regards the young, the name foal — equivalent to the Anglo-Saxon fole, the Latin pullus, and the Greek polos, all denoting originally a young animal — is applicable to both sexes. The word filly, on the other hand, which likewise apparently comes from the Anglo-Saxon fole, denotes a female foal ; whereas colt — an Anglo-Saxon derivative probably connected with cild, a child — is applied solely to a foal of the male sex. Finally, the term gelding signifies the castrated male. In this place it may be convenient to mention that although the name horse properly belongs only to the domesticated and wild representatives of Equus caballus, it is frequently employed by natu- ralists in a more extensive sense. We speak, for instance, of the Arabian horse ; and even if that be, as some suppose, specifically distinct from the ordinary horse of Western Europe, there is no question that such usage is perfectly legitimate and permissible. On the other hand, all the other ex- isting members of the horse tribe, or Eqiuidce, have distinctive names of their own, such as ass, zebra, and quagga. Nevertheless, all these are often called horses in works on natural history, although the practice has its inconveniences ; and the term 4 THE HORSE AND ITS RELATIVES members of the horse tribe is preferable. Similarly, the name horse, as the denomination of the typical member of the genus Bqzms, is very generally applied to the extinct representatives of the same genus or even of closely allied genera ; and there is less objection to this practice than to the one last mentioned, as there are no vernacular names for the animals in question. The term three-toed horses is, for instance, a convenient one for the members of the extinct genus Hipparion, as it is not likely to lead to confusion. On the other hand, the word horse must have some limitation ; and the American practice of applying it to diminutive ancestral types of the Equid<^ no larger than foxes is one that is not to be commended. Since the respective meanings of the terms species, genus, family, order, &c., are explained in most works on natural history, it will suffice in this place to state that Eqiius caballus, as represented by the ordinary domesticated horses of Western Europe, is the typical representative of both the genus Eqtms and the family Equidcu. Both that family and the Bovidce, or hollow-horned ruminants — of which the ox is the typical member — belong to the great order of hoofed mammals, or Ungulata, so called from the feet of its more typical representatives being encased in solid horny hoofs. These more typical groups are divided into two main sections or sub-orders, namely the even-toed POSITION AND STRUCTURE OF HORSE 5 ungulates, or Artiodactyla, as typified by the ox, and an equivalent, albeit at the present day much smaller section, known as the odd-toed ungulates, or Perissodactyla, of which the horse and its relatives are the most specialised members. Although the leading points of distinction between these sub- orders have been indicated in my volume on the ox, it is advisable that the characteristics of the second should be repeated, as well as somewhat amplified, in this place. The odd-toed, or perissodactyle, ungulates take their name from the circumstance that the toe corresponding to the middle finger of the human hand and its representative in the hind-limb, to- gether with the bone known as metacarpal in the fore, and metatarsal in the hind leg, respec- tively form the continuation of the main axis of the limb, and are symmetrical in themselves. In the horse and its immediate relatives this middle toe is alone functionally developed in both the front and hind legs ; but in the rhinoceroses, which belong to the same sub-order, although to a different family [Rkinocerotidcs), there is a pair of smaller lateral toes, each of which, together with its supporting metacarpal or metatarsal bone, is like- wise symmetrical. Just as the middle toe of the fore-leg corresponds to the middle or third finger of the human hand, so the lateral toes of the rhinoceros severally represent the second and fourth fingers 6 THE HORSE AND ITS RELATIVES of man ; a similar correspondence to the toes of the human foot obtaining, of course, in the hind-leg. A practically identical correspondence obtains in the hind-foot of the tapirs, which belong to the third The bones of the lower part of the left fore-leg of a Horse (A) and a Rhinoceros (B). r, lower end of radius or inner leg-bone ; u, do. of ulna or outer leg-bone ; c, carpus or wrist ; vie, meta- carpal bones; ph, phalanges or toe-bones; il., III., iv., second, third, and fourth toes, or (in A) the remnants of the metacarpals. In the horse mc is known as the cannon-bone, the two upper phalanges are termed pastern-bones, and the lower one is the coffin-bone. and last family ( Tapiricice) of living perissodactyles. The front-foot of the tapir has, however, four toes, of which the outermost represents the fifth, or little, fino^er of the human hand. The addition, or rather POSITION AND STRUCTURE OF HORSE 7 the retention, of the outermost toe does not, how- ever, affect the symmetry of the other three front- toes of the tapir, which are arranged in just the same manner as in the three-toed fore-foot of a rhinoceros. As a whole, however, the skeleton of the fore- foot of a tapir is obviously unsymmetrical, this being due to the loss of the toe corresponding to the human thumb, or first finger, as it should pro- perly be called. And it may be noted here that this first toe has disappeared from both feet in all members of the odd-toed group, extinct as well as living, although it is developed in certain primitive members of the ungulate order, of which mention is made in the sequel. This symmetrical development of the third toe (inclusive of the supporting metacarpal in the fore and the metatarsal in the hind limb) and its superiority in size over either of the lateral ones, when these are present, is the one great feature of the skeleton by which the odd-toed ungulates, or Perissodactyla, are distinguished from the even- toed group, or Artiodactyla. In the latter group, as is fully described in the volume on the ox, the third and fourth toes are equal in size, and developed symmetrically to one another on either side of the vertical line between them. Conse- quently, in that group the continuation of the main axis of the limb is formed by the vertical line 8 THE HORSE AND ITS RELATIVES dividing the third from the fourth toe, whereas in the odd-toed group it is constituted by a line running down the middle of the third toe. This, although by far the most important, is, however, by no means the sole character in which the skeleton of an odd-toed differs from that of an even-toed ungulate. One of the most easily recog- nised of these minor skeletal differences is the presence in the members of the odd-toed group of a more or less strongly marked projection or process on the outer side of the upper end of the shaft of the femur, or thigh-bone, which is totally absent in the even-toed group. Another, although less obvious, difference is to be found in the shape of the astragalus, or huckle-bone, of the tarsus, or ankle-joint — the so-called hock of the horse. In the Perissodactyla the lower surface of this bone is markedly flattened, whereas in the Artiodactyla the same surface is rounded and pulley-like : the vertical diameter of the whole bone being also relatively less in the former than in the latter group. Yet another difference is to be found in the number of joints, or vertebrae, in the backbone, or vertebral column, of the two groups. In perissodactyles the number of vertebrae between the skull and what is known as the sacrum (that is to say, the con- solidated mass of vertebrae to which the haunch- bone, or pelvis, is attached) is never less than 29 and is very generally 30, whereas in the artiodactyles POSITION AND STRUCTURE OF HORSE 9 it is invariably 26. If we exclude from this enu- meration the seven cervical vertebrae common to all ungulates, this may be expressed in another way by saying that whereas in the odd-toed group the number of trunk-vertebrae may be 22 or 23, in the even-toed group it is invariably 19. Other differences in the skeleton, as well as certain peculiarities in the teeth of the two groups, need not be mentioned here ; but it may be observed that no perissodactyle has the complex type of stomach characteristic of the ruminating artiodactyles. Although, as already mentioned, the Perisso- dactyla are represented at the present day only by three families — the EquidcE^ Rhinocerotidce, and TapiridcB — the living members of each of which may be included in a single genus, during the Tertiary period they comprised several extinct families and a large number of genera. Nor is this all, for whereas, with the exception of the tapirs, which are common to Malaya and Tropical America, the group is nowadays restricted to the Old World, in past times it was abundantly repre- sented in the New World, where the three existing families (together with certain extinct ones) occurred in North America, while the Equidce succeeded in effecting an entrance during the latter part of the Tertiary period into South America. Even this, however, does not represent the real 10 THE HORSE AND ITS RELATIVES poverty of the odd-toed group at the present day, for the three surviving families are remarkable for the small number of their existing representatives. The horse tribe, for instance, includes at the present day only about eight or nine species, while the rhinoceroses comprise five, and the tapirs another five, or possibly six, specific types. The whole number of living perissodactyles is thus well under a score. In this poverty of families, genera, and species the Perissodactyla present a remarkable contrast to the Artiodactyla of the existing epoch, whose specific representatives are between one and two hundred in number, and are classed in no less than nine or ten separate families, with a collectively world-wide distribution — exclusive, of course, of Australia. Despite the fact of its having lost a large number of generic and family types, the Artiodactyla may be regarded as a dominant type at the present day, whereas the Perissodactyla are as distinctly a waning group, so far, at least, as the numerical abundance of genera and species is con- cerned. What may have been the cause of this difference in the two groups cannot yet be determined. As regards the characters by which the family Equidce is distinguished from other groups of odd- toed ungulates, there is no difficulty at all when the existing representatives of the sub-order are alone taken into consideration, since the horse and POSITION AND STRUCTURE OF HORSE ii its immediate living relatives are broadly distin- guished from all other modern mammals by the reduction of the number of toes on each foot to one. During the later part of the Tertiary, or latest, period of geological history there existed, however, a number of animals agreeing in all essential characters with the modern horse and its relatives, but with three toes to each foot, although the lateral ones were so small as to be of no functional importance. These three-toed horses, of which there is more than one generic type, had tall-crowned cheek-teeth differing from those of the modern horse only in certain comparatively un- important details of structure, and in the somewhat inferior height of their crowns. Somewhat earlier in the Tertiary are found remains of other three- toed horse-like animals, which differ much more markedly from the modern type. Among their more salient differences are the relatively larger size of the lateral toes, which, in some cases at any rate, were at least partially of functional use ; and, more important still, the quite short crowns of their cheek-teeth. Now in defining the Equidcs it is found con- venient to take tallness of crown in the cheek- teeth as the main distinctive character, and to include in that family only those species in which this feature is distinctly developed. The short- crowned species are accordingly referred to separate 12 THE HORSE AND ITS RELATIVES families, of which all the members are extinct : to these fuller reference is made in the chapter on the extinct relatives of the horse. It will of course be obvious that if we had before us the whole of the ancestral series of the horse, we should find an absolutely complete gradation from types with tall-crowned to those with short-crowned cheek- teeth ; and it would consequently be impossible to draw a hard-and-fast line between the two. In that case it would be necessary to make an arbi- trary line of division, as it is sufficiently obvious that it is quite impossible to include all the ancestors of a given mammal in a single family group ; as, if this were attempted, we should find reptiles included in the horse family, mammals being undoubtedly descended from certain extinct, and to some extent primitive, groups of reptiles. The horse family, or Equidce, may, then, be briefly defined as including such odd-toed or perisso- dactyle ungulates as have tall-crowned cheek-teeth of a peculiar and characteristic pattern (described later on), and each foot terminating in a single large functional toe, which, in the case of some extinct species, may be flanked by a pair of much smaller functionless toes. In the course of this chapter other characters are noticed which may likewise be used in the definition of the family, although the two mentioned above are amply sufficient for the purpose. POSITION AND STRUCTURE OF HORSE 13 All the members, whether living or extinct, of the family, as thus restricted, come under the denomination of what naturalists term highly special- ised animals ; the specialisation in this instance taking the form of adaptation for the attainment of great speed in running, and likewise for grazing on grass or other herbage. Their specialisation is best observed in the skeleton and teeth ; the former departing almost as widely as possible from that of a generalised animal, such, so far as the structure of the limbs is concerned, as a bear. In the latter animal each foot terminates in five complete and functional toes, usually armed with claws ; and in walking the entire sole of the foot, inclusive of the heel-bone, or calcaneum, in the hind-pair, is applied to the ground. In the horse and its existing relatives, together with certain extinct species, such as the one of which the skeleton is shown in plate xxiii., each foot terminates in a single toe (encased during life in a hoof), upon which alone the animal walks ; this single toe, as already mentioned, and as shown in the figure on page 6, corre- sponding to the middle or third one of the generalised five-toed type. Consequently the heel, or ankle, technically known as the tarsus, in place of resting on the ground, as in the bear, is raised 14 THE HORSE AND ITS RELATIVES high above it — forming, in fact, in the horse the so-called hock. Similarly in the fore-limb the wrist- joint, or carpus, is raised to the same approximate level above the ground, and constitutes the so-called knee of the horse. The single metacarpal bone [mc in the figure on page 6) in the fore-leg of the horse is known as the cannon-bone ; the same name being also applied to the corresponding element in the hind- leg, that is to say, the metatarsal bone. The remnants of the lateral metacarpals (ii., iv. in the figure last cited) in the fore-leg, as well as those of the corresponding metatarsals in the hind-limb, are designated splint-bones. The degree of develop- ment, or rather of the degeneration, of these splint- bones varies considerably in different horses. In many instances, as in the figure on page 6, these bones represent merely the upper ends of the metacarpals and metatarsals. In other cases, as in the illustration of this part of the limb of a shire horse (pi. i. fig. i) the whole shafts of the splint-bones are retained, with remnants at the lower end of the first and second toe-bones, or phalanges (i, 2). This comparatively full develop- ment of the splint-bones appears to be not un- common in shire horses ; but remnants of the toe-bones, which in all cases are firmly welded with the splint-bones, are retained in the skeleton of the famous racehorse " Stockwell," of which the PLzVTE I Fig. 2 Fig. I. Back view of fore (A) and hind (B) cannon and splint bones of Shire Horse. Fig. 2. Front view of the bones of the left fore (A) and hind (B) feet of the extinct Three-toed American Alioliippjts anceps. {\ natural size.) POSITION AND STRUCTURE OF HORSE 15 limb-bones are exhibited in the Natural History- branch of the British Museum. The figure of the bones of the feet of an extinct three-toed horse is placed alongside that of the cannon-bones of the shire horse in order to show how the splint-bones and rudimentary toe-bones of the latter correspond with the same bones in a more fully developed condition in the former. So far as can be ascertained, the splint-bones of the horse and its existing relatives are of no use to their owners, although there is just a possibility that they may be of some slight service in mitigat- ing shock. On the other hand, in domesticated horses they are frequently harmful, since, through inflammation and subsequent exostosis, they give rise to the disease known as splint. In this connection it is interesting to note that a few years ago Professor La Van de Pas, of the Agricultural and Veterinary Institute of Buenos Aires, published an account^ of a prevalent type of degeneration in the splint-bones of Argentine horses. In 1904 the author received the left hind cannon-bone of a horse in which the outer splint- bone was only half the normal length, its lower portion being replaced by ligamentous tissue. The other cannon-bones of this horse were not forth- coming, but in the course of the next few years the author had the opportunity of examining a ^ Anales de Mus. Nacional de Buenos Aires, ser. 3, vol. x. i6 THE HORSE AND ITS RELATIVES considerable series of such bones. In a large proportion of these a similar degeneration was observed, sometimes in one and sometimes in both splint-bones ; and it may accordingly be considered that such atrophy is comparatively common in Argentine horses. It is further noticeable that the degeneration is more marked in the outer than in the inner splint-bone, alike in the fore and the hind limb. At the close of the descriptive portion of the communication the author arrives at the con- clusion that Argentine horses are endeavouring to discard these seemingly useless portions of the skeleton. It may be added that in veterinary anatomy the first and second phalanges, or toe-bones, of the horse's foot are respectively termed the upper and lower pastern-bones, while the enlarged terminal bone which carries the hoof is known as the coffin-bone. The last-named bone, it may be noted, is much wider in the fore-foot than in the hind-foot ; having almost the shape of a cheese- cutter in the fore-leg. Another term employed in veterinary works is fetlock, which denotes the joint between the lower end of the cannon-bone and the upper pastern. The middle segment of the skeleton of the horse's fore-leg, that is to say the one immediately above the carpus, or so-called knee, is formed mainly by the radius, or inner leg-bone ; the ulna POSITION AND STRUCTURE OF HORSE 17 or outer leg-bone — which is as long as, and separate from the radius in less specialised animals, such as rhinoceroses — being represented only by its upper end, corresponding to the human elbow, or olecranon, and this being immovably soldered to the radius. In consequence of this welding of two originally separate bones into a single compound element, the fore-leg of a horse is capable of no other movement than a backwards and forwards one ; this being all that is needed by a running animal. A similar consolidation and simplification of elements likewise obtains in the middle segment of the hind-leg of the horse, in which the originally distinct smaller bone known as the fibula is re- duced to its upper extremity, this being firmly welded to the upper end of the larger bone, or tibia. The reduction in the number of toes to a single large one in each foot, and of the metacarpal and metatarsal bones to the aforesaid splints, coupled with the elongation of the bones of the whole lower segment of the limb, the simplification and con- solidation of those of the middle segment, and the raising of the carpus (" knee ") and tarsus (hock) far above the level of the ground, so as to cause the animal to walk on the tips of its single toes, are the chief features in which the skeleton of the horse shows (as compared with that of more generalised animals) special adaptation for the attainment of a 1 8 THE HORSE AND ITS RELATIVES high speed. Such a type of limb is the one evi- dently best suited to this end, since if the short, many- toed, and many-boned limb had been elongated without special modification to the extent of that of the horse, it is perfectly certain that it would have been unequal to the strain of carrying the body of such a heavy animal at a high rate of speed over hard ground. In their limb-specialisation the horse and its relatives have attained practically the same evolu- tionary platform as the ruminant ungulates, only by a different line of development. In the horse group, as we have just seen, the development of a long cannon-bone in the lower segment of each limb has been brought about by the lengthening and strength- ening of the middle element of the primitive five-toed foot. In the ruminants, on the other hand, the same end has been attained by the lengthening and fusion of two adjacent elements, so as to form a compound, in place of a simple, cannon-bone. At the present day the members of the horse family are absolutely unique in the matter of limb-structure, no other living mammal (or, for that matter, no other living animal) having a single-toed, or monodactyle, foot. It is, however, not a little remarkable that during the middle, or Miocene, portion of the Tertiary period South America was the home of a genus of hoofed mammals known as Thoatherium, in which a monodactyle type of foot had likewise PLATE III Fk;. I Fig. 2 Skeleton of front (Fig. i) and hind (Fig. 2) limbs of " Eclipse." s, scapula ; /i, humerus; u, ulna; r, radius; c, carpus or wrist ("knee"); ca«, cannon-bone ; p/i, phalanges or toe-bones ; /, pelvis ; /, femur ; /, tibia ; ca, calcaneum or upper bone of tarsus or ankle ("hock "). POSITION AND STRUCTURE OF HORSE 19 been developed by the reduction of the lateral toes of a nearly allied tridactyle relative. In this case the specialisation was even greater than in the horse group, as the splint-bones were reduced to mere nodules of bone on either side of each cannon- bone. If Thoatherium had been a near relative of the horse, there would be no cause for surprise in its having attained the same remarkable and final stage of foot-development. As a matter of fact, it belongs, however, to a totally different and much more primitive group of ungulates, which appears to have been always restricted to South America, and although presenting certain structural resem- blances to the Perissodactyla, is in other respects so distinct that it is ranked, under the name of Lito- pterna, as an equivalent subordinal group of the great order Ungulata. The most remarkable thing connected with Thoatherium is that, despite the specialised toes, its carpus and tarsus are of an exceedingly primitive type.^ Several noteworthy features occur in the skull of the horse and its existing relatives. In the first place, it differs from the skulls of all other living perissodactyles — namely, tapirs and rhinoceroses — in the complete closure of the rim of the socket of the eye by means of a bridge of bone extending 1 See W. B. Scott, "The Litopterna," Rep. Princeton Univ. Exped. to Patagonia, vol. vii. pt. i. 1910. 20 THE HORSE AND ITS RELATIVES from the forehead downwards to the horizontal bar known as the temporal arch. Secondly, it is characterised by the inordinate length of the por- tion in front of the socket of the eye, or orbit, as compared with the part behind the same. " In the horse," writes Professor H. F. Osborn,^ ''lone- headedness is a very ancient character ; even the earliest known four-toed horses have quite elongate, or at least mesaticephalic [moderately long] skulls. Skull of a giant extinct Pig-like animal {Elothermm), to show the horse-like elongation of the facial portion. The progressive elongation of the skull in horses is apparently for two purposes : first, to facilitate reaching the ground with the row of incisor or cropping teeth ; second, and no less important, to allow space in front of the eye-sockets for the great rows of elongate, or hypsodont, grinding teeth, the marvellous dental battery of the horse. We might assume from these facts that long-headedness is correlated with long teeth, but the giant pigs * The Age of Mammals^ New York, 1910, p. 18. POSITION AND STRUCTURE OF HORSE 21 (elotheres) have still longer and narrower skulls than the horse, yet all the teeth are brachyodont, or short-crowned. Again, the elephant has extremely elongate, or hypsodont, molar teeth, yet it also possesses the shortest skull known among the Mammalia." Another feature in the skull of the existing members of the horse family is the comparative shortness of the slit separating the front end of the nasal bones, which form the roof of the nose- chamber, from those of the upper jaw, as is well shown in the figure of the skull of a shire horse (pi. iv. fig. i). The importance of this feature will be apparent when the extinct relatives of the horse are taken into consideration in a later chapter. Much has been made of the degree to which the facial portion of the skull of the horse is inclined to the basal axis of its hind part.^ Although there is undoubtedly great variation in this respect between different horse-skulls, it is far from certain that they are really of any special importance. For it has been suggested that this bending down of the fore- part of the skull on the basal axis, which occurs in many grass-eating mammals, is primarily due to the " pull " or strain caused by the act of grazing ; and if this be really the case, it is obvious that ^ See J. C. Ewart, Tratts. R. Soc. Edinburgh, vol. xlv. pp. 555, 587, 1907- 22 THE HORSE AND ITS RELATIVES horses whose ancestors have for many generations been accustomed to feed on hard and tough grasses will show a greater deg^ree of cranial deflection than those whose food has consisted either of softer and more easily yielding herbage or of grain. More- over, it seems highly probable that the degree of deflection may vary with age — the older the animal, the greater the degree of bending. This, indeed, is exemplified in the case of two skulls figured by Professor Ewart on plate ii. of the memoir cited in the footnote to illustrate this feature ; the one shown in the upper figure of that plate, in which the deflection is slight, being obviously that of a young animal, while the one in the lower figure, which displays the bending in a very marked degree, is as clearly that of an aged horse. Yet another cranial feature remains to be noticed. In the skulls of certain domesticated horses, especially Arabs, thoroughbreds, and shires, a more or less distinct oval or circular depression may be noticed a short distance in front of the socket of the eye, or orbit, and therefore con- veniently called the preorbital depression. It is shown faintly in the figure of the skull of a shire horse in plate iv. fig. i, and more clearly in the photo- graph of the skull of a quagga (plate iv. fig. 2) ; the latter instance shows that the feature is not confined to the horse itself. Of itself, this feature may seem PLATE IV Fig. I Fig. 3 Fig. I. Skull of a Shire Stallion ; //, preorbital depression. Fig. 2. Skull of a Quagga. to show preorbital depression. POSITION AND STRUCTURE OF HORSE 23 a very trivial one, but it happens, as will be noticed more fully in the sequel, that the skulls of certain extinct horses show a much more marked de- pression in the same region — a hollow so deep that it deserves the name of pit rather than depression. Writing of this preorbital hollow in the extinct three-toed Hipparion, Sir W. H. Flower in his volume on The Horse} observed that "although such a pit is not found in any of the existing species of horse, it was not infrequent in many extinct forms, and varied in them in size and depth. It so closely resembles a similar depression found in the same situation in many species of deer and antelopes, which lodges a glandular infolding or pouch of the skin called the 'suborbital gland,' 'crumen,' or in French 'larmier,' that there can be little doubt but that it had the same purpose in the hipparion. This gland in the existing animals that possess it secretes a peculiar oily odorous substance, the scent of which enables the animals provided with it to recognise each other even at immense distances, the faculty of smell being also developed to a wonderful degree. . . . " The presence of this gland in the hipparion and its absence in the more modern Equidce has been given as a reason for supposing that the latter are not the direct descendants of the former, ^ London, 1891, p. 64. 24 THE HORSE AND ITS RELATIVES but must have been derived from some other form in which such a specialisation had been developed. This, of course, is probable ; but it must not be forgotten that very slight changes in habits, or the increased uses of other senses than that of smell, may have diminished the value of the information afforded by this gland, and ultimately led to the elimination of the organ itself. It may be that a change from a life habitually passed in forests or scrub to one in open plains would be sufficient to account for such a modification in structure." Sir William Flower appears never to have noticed the presence of the slight preorbital de- pression in the skull of certain existing represen- tatives of the horse family ; but when it first came under my own observation the suggestion naturally arose that it was the last vestige of the decadent fore-gland of the three-toed hipparion. Confirmation of this was afforded by a statement made by Professor T. H. Huxley,^ that traces of a preorbital pit remain in the skulls of some of the fossil horses from the Siwalik Hills of Northern India. But this was not all, for some years ago I received a communication from a correspondent to the effect that he had seen a living horse, believed to be of Argentine origin, in which there was a distinct depression, although without any external orifice, just in front of each orbit. More * Quart. Journ. Geol. Soc, London, vol. xxvi. p. 2, 1870. POSITION AND STRUCTURE OF HORSE 25 important still was a letter from Mr. Wilfred Scawen Blunt, the well-known possessor of a stud of Arabs, in which it was stated that he once owned a horse of this breed in which there was a well-developed and functional gland on one side of the face in the same position as the larmier of a deer. The identification of the preorbital depression in the skulls of certain existing members of the horse family has been accepted by Sir E. Ray Lankester,^ who remarked that although dissection had not revealed any existence of glandular tissue in the structures overlying this structure in horses of Arab descent (in which the feature is very constant), yet it is not improbable that occasional instances of such survival will some day come to light. On the other hand, Mr. R. I. Pocock,^ after dissecting a number of horses' heads, came to a precisely opposite conclusion ; remarking that he had failed to discover any trace of glandular tissue in the soft parts overlying the depression, and that the depression itself is very variable in its degree of development. He then adds that " from this hollow or from the corresponding area of the skull [when it is absent] arises a long muscle which passes forwards to supply the upper lip and nose ; and I believe its sole significance is to give an increase of surface for muscular fibres. If this ^ Science from an Easy-Chair^ London, 1910, p. 87. - Ann. Mag. Nat. Hist.., London, ser. 7, vol. xv., p. 517, 1905. 26 THE HORSE AND ITS RELATIVES be so, variation in the extent to which the depres- sion is developed is exactly what would be ex- pected." Mr. Pocock then goes on to observe that the deeper pit observable in the skull of the extinct three-toed hipparion may possibly, although not probably, be also an area for muscular attach- ment. He adds that in the skull of the extinct South American Onokippidmm, where the preorbital pit (as is shown in the concluding chapter of this volume) is very large, there appears to be a division into two parts, of which one is shallower than the other, and may correspond to the depression found in some existing members of the horse family. Apparently, therefore, Mr. Pocock is of opinion that in the two extinct genera just mentioned lachrymal glands were probably developed. A later writer. Professor Studer,^ when describ- ing a species of hipparion from the upper Tertiary strata of Samos, goes much further than this, and ex- presses the opinion that in no case is the preorbital pit for the reception of a lachrymal gland, but that it is always solely for the purpose of muscular attachment, and attains its maximum development in species like Onohippidium and the Samos species of Hipparion which were probably furnished with a proboscis. The position of the pit, it is stated, differs somewhat from that of a true larmier, and ^ Verk. Deutsch. Zoo I. Ges. 1910-11, p. 11. POSITION AND STRUCTURE OF HORSE 27 the inframaxillary foramen is always some distance from the pit. Leaving the skull, attention may be directed to the teeth of the horse, which form an exceedingly important feature in its anatomy. As is well shown in the fiofure of the skull of a shire stallion in plate iv. fig. I, there is a considerable gap between the front teeth, which are mainly adapted for nipping and biting, and the long series on each side of the face, which are conveniently called cheek-teeth, and whose function is to grind up the grass or other herbage gathered by the front teeth. In the anterior half of the aforesaid gap there occurs, however, in stallions a tusk on each side of both upper and lower jaws, which corresponds to the canine of carnivorous mammals, and is separated, in each jaw, from the three pairs of incisor teeth, which occupy the front of the jaws. In mares the tusks are either very small or wanting, from which it may be inferred that in stallions these teeth are mainly, if not entirely, used in fighting, and not for gathering or masticating food, for which, indeed, they are obviously unsuited. The existence of this long gap between the front and the cheek teeth is a specialised feature ; mam- mals of a more primitive type having either the whole of the teeth in contact, or with relatively small intervals on each side of the canines when these are large. A similar long gap occurs in the lower 28 THE HORSE AND ITS RELATIVES dental series of the ox and its relatives, in which, however, the lower canines have become approxi- mated to the incisors, with which they form a regular series of spatulate-crowned teeth. In the upper jaw of the ox tribe specialisation has been carried to a much greater extent than in the horse, the canines and incisors having completely disappeared, and being replaced by a hard pad which takes the bite of the lower front teeth. In consequence of the retention of upper as well as lower front teeth, a horse is apparently able to graze closer than an ox. From the existence of the aforesaid long gap between the front and the cheek teeth in both the horse and the ox and their respective relatives, it would seem that such an arrangement is the one best suited for grazing or browsing animals ; and it is not improbably for the purpose of affording room for the play of the large tongue, which takes an impor- tant share in the action of grazing. The incisors of the horse, of which those of the lower jaw have somewhat less distinctly spatu- late crowns than their representatives in the ox, present a relatively complex type, met with among no other living mammals outside the Equidcs. In place of having simple conical crowns when they first emerge in an unworn condition from the gum, the incisor teeth of the horse have a kind of pit, or pocket, at the summit, which penetrates far POSITION AND STRUCTURE OF HORSE 29 down into the heart of the crown. The nature of this peculiar structure may be best understood by taking the finger of a kid glove, and, after filling it with soft wax, pushing in the summit by means of a fine pencil to a depth of an inch or so. This will give an exact representation of a horse's incisor, more especially if we smear the lower part of the outer surface of the finger with sealing-wax, as we shall then have representatives of the three constituents of the tooth. Thus the sealing-wax will represent the outer coat, or cement, the glove the middle element, or enamel, and the soft wax the inner constituent, technically known as the ivory or cement. In this condition the wax-filled glove-finger with the pit, or "mark," at the tip, will represent the unworn incisor of the horse ; but if we snip off with a pair of scissors half an inch from the summit, we shall have a model of the tooth after it has been in use for some time, and hns had its tip ground away by wearing against its fellow in the opposite jaw. On looking at the section of the summit of the cut glove-finger, it will be seen that the original pit now forms an island in the middle of the soft wax ( = dentine) bounded by a ring of kid ( = enamel). By carrying the experiment one stage further, and cutting off another three-quarters of an inch from the summit, we shall have the model of an incisor of an old horse. In this state the pit, or "mark," 30 THE HORSE AND ITS RELATIVES will have completely disappeared ; and the section will represent ( i ) a central core of soft wax, corre- sponding to the comparatively soft dentine, {2) a ring of kid, equivalent to the hard enamel, and (3) an irregular coat of sealing-wax, corresponding to the external layer of cement. A horse, like most mammals, grows two sets of incisor teeth in both jaws ; firstly, a baby, or milk set, and secondly, a permanent set. The three teeth of the first set are, however, not all shed at once, but one by one, when they are as gradually replaced by their permanent successors, which grow up beneath. And it is by knowing how this replacement occurs, and noting the extent to which the central mark is worn away, that the age of a horse can be approximately ascer- tained up to six or seven years old. The " mark," it should be added, is common to both the temporary and the permanent set of incisors ; but is deeper in those of the upper than in those of the lower jaw. The jaws of a quite young colt show only the first and second pairs of milk-incisors, both above and below ; but after a time the third pair appears on their outer sides. In a horse of about three years old the first pair of permanent incisors (recognisable by their larger size and unworn crowns) will have pushed out and replaced the corresponding baby- teeth. At an aofe of between three and a half and four years the second or middle pair of milk-incisors will have been similarly replaced by the permanent POSITION AND STRUCTURE OF HORSE 31 pair. About half a year later the permanent tusks, or canines, make their appearance in the case of stallions. By the end of the fifth year the third or outermost pair of permanent incisors will have re- placed the corresponding temporary pair ; and the dentition of the front of the mouth will consequently be complete. It will be obvious that of the three pairs of upper permanent incisors, the crowns of the first pair will, after all are in place, be more worn than those of the second, and the second more than those of the third. As a rule, the mark dis- appears in the first pair of lower permanent incisors when the horse is six years old ; in the second pair it is worn out a year later ; and in the third pair at eight years. In the corresponding upper teeth it persists about two years longer in each instance. In the case of a six-years' -old horse the third lower incisors retain large and conspicuous marks. Up to five years the age of a horse can be determined with comparative accuracy, and it can also be approximately ascertained for some years later. When the mark has been worn out in all the incisors, age-determination is no longer possible by means of the teeth. It appears, however, that in very old horses a kind of spurious mark is formed by the tooth becoming so worn down that the summit of the pulp-cavity at its base is exposed in the centre of the crown. Such a mark lacks, how- ever, the ring of enamel characteristic of the true 32 THE HORSE AND ITS RELATIVES mark. Moreover, when this false mark makes its appearance, the section of the crown of the incisor has become much more triangular than in the early stages of wear ; and in extreme old age, when the incisors are worn down to their very roots, these teeth become very narrow in the transverse direction, whereas in their earlier ages this diameter was considerably larger than the opposite one. From the foregoing description, it will be evi- dent that to the definition of the horse family given above may be added an additional characteristic, namely the presence in an early condition of wear, of a "mark," or pit, in the crowns of the incisor teeth. As regards the object of these pits, their function is probably to increase the grinding surface of the dentition as a whole during the period when the cheek-teeth have not attained their full develop- ment ; the number of functional teeth of the latter series in young colts being only three pairs in each jaw, while at a later stage, when some at least of the permanent incisors have come into use, the hinder teeth of the cheek-series are not fully developed. Before leaving this part of the subject, it may be mentioned as a remarkable circumstance that, in addition to the existing members of the horse family and some of their extinct forerunners, the only mammal which shows a pit in its incisors is POSITION AND STRUCTURE OF HORSE 33 the South American extinct Macrauckenia, a member of the same sub-order (Litopterna) as Thoatherium, a genus already referred to in connection with the skeleton of the horse's foot. This shows that the EquidcE are paralleled in two remarkable structural features — by Thoatherium in the monodactyle feet, and by Macrauchenia in the presence of pits in the crowns of the incisor teeth. Passing on to the cheek-teeth of the horse, it A left upper molar of a Single-toed {Equus), A, and a Three-toed Horse (Hipparion), B. p, anterior pillar, hy, posterior pillar. has to be noted in the first place that these are normally represented in the adult by six pairs in both the upper and lower jaw (pi. v.) ; of which the first three pairs, on account of being preceded by milk- teeth, are known as premolars, while the other three pairs, which have no such temporary predecessors, are the molars, or true molars, as they are often called. In some horses, however, there is a small and practically functionless tooth on the inner side of 34 THE HORSE AND ITS RELATIVES the front end of the anterior premolar ; and, more rarely, there may be a still smaller tooth in a corresponding position in the lower jaw. These small functionless teeth, which have no temporary predecessors, and are known to horse-dealers as "wolf-teeth," may be developed on only one side of the jaw. They are the vanishing representatives of teeth which were relatively large and functional in some of the extinct ancestors of the horse, and are of importance as showing that the three large functional premolars of the latter correspond to the three last of the typical mammalian series of four. Hence it is frequently found convenient to speak of these teeth as the second, third, and fourth pre- molars, instead of calling them the first, second, and third. On the other hand the three pairs of molars are respectively denominated the first, second, and third. In a young colt, if the "wolf-teeth" be not de- veloped, there are three pairs of milk-molars in each jaw ; those of the upper jaw having their crowns more elonofated from front to back than is the case with the premolars by which they are subsequently replaced. As the colt grows older, the first molar cuts the gum before the last premolar has replaced the corresponding milk-molar ; and, as a conse- quence of this, it will always be found in an adult horse that the crown of the first molar is rather more worn than that of the tooth immediately in U 2 3J 1) ^ ;i^ POSITION AND STRUCTURE OF HORSE 35 front of it, that is to say, the last premolar. With this exception, each cheek-tooth in an adult horse is always more worn than the tooth immediately behind it. With the exception of the first and last, which are more or less pointed at the free end, an upper cheek-tooth of a horse consists of a square prism rather more than an inch in diameter, and about three inches in height when unworn, with its lower extremity terminating in four roots. Both the outer and inner surfaces are marked by strong vertical flutings ; and when in use, only a small extent of the upper part of the crown is exposed above the gum. As these teeth are usually seen in a more or less worn condition, it is preferable to take such a partially worn tooth as the basis for a description of their leading characteristics. Such a tooth may be compared in structure to the incisors ; its apparent complexity of structure being due to the pushing-in, on the summit of the crown, of two pits compar- able to the single pit, or ** mark," in the incisor. These two pits are the two irregularly-shaped islands seen in the middle of the crown in A of the illustration on page 2>3- The centres of these pits, which extend right down to the base of the crown of the tooth, are filled with cement ; and the walls of enamel with which they are lined are thrown into a number of more or less complex foldings. 36 THE HORSE AND ITS RELATIVES These are most developed in the extinct three- toed hipparion, as shown in B of the same illustra- tion, in which the cement filling the pits is white. Further complexity is produced by vertical flutings on the inner side of the crown, which result in the production of the two semi-isolated pillars marked p and hy in the aforesaid figures. Of these two inner pillars by far the more important from a systematic point of view is the front or anterior one {J) in the illustrations), for it affords an import- ant character in the definition of the genus Eqiius. In all the existing members of the horse family this pillar is connected by a narrow isthmus with the main body of the tooth ; and the fore-and-aft diameter of its worn surface is considerably longer than the transverse one. A pillar of this type is termed a broad one ; but there are certain extinct horses in which, while the pillar remains connected with the body of the tooth, its two diameters are nearly equal ; in teeth of this type the anterior pillar is said to be narrow. In the three-toed hipparions (B of the illus- tration on page 2)1^) ^ totally different condition obtains, the anterior pillar, which is of the narrow type, being completely surrounded by a ring of enamel, so that its central core of dentine is cut off from the dentine of the main body of the tooth. To put the matter shordy, it may be said that while the upper molars of the horse and its immediate POSITION AND STRUCTURE OF HORSE 37 relatives have the anterior pillar of a peninsular type, in the three-toed hipparions it assumes a com- pletely insular form. The lower cheek-teeth of the horse have much narrower crowns than the upper ones ; but the foldings on their crowns are of the same general type, although in a reversed way, the portion corresponding to the inner pillars being on the outer side of the tooth. It will not be necessary to describe these teeth in detail. In a general way the three upper molars of the horse correspond in structure with those of the ox. In each, as is best seen when the teeth are' in an unworn condition, there is a pair of central pits or islands on the crown, around which are four sub-crescentic columns ; but whereas the pits are almost completely filled with cement in the horse, in the ox they remain more or less open. Such a tooth may be described as consisting of two lateral lobes, each with a single central pit. In the horse, as shown in the figures in plate v., the upper premolars are of the same structure as the molars ; the last premolar being in some cases even larofer than the first molar. In the ox, on the other hand, the upper premolars are smaller in size and simpler in structure than the molars ; each of the former consisting of only a single lobe, with one central pit, although this lobe is somewhat larger than one of those of the 38 THE HORSE AND ITS RELATIVES double-lobed ox-molars. In consequence of this greater complexity of its premolars (which is common to both upper and lower jaw), the dental mill of the horse, as Professor Osborn calls it, forms a more powerful and more efficient grinding instrument than that of the ox. And the reason for this greater masticating power in the dentition of the horse is, I think, not very difficult to discover. As is stated in my volume on that animal, the ox, in common with other ruminants, gathers its food quickly, swallows it, and subsequently, owing to the complex structure of its stomach, regurgitates and remasti- cates it at leisure when in repose in a position of more or less security. The horse, on the other hand, who has a stomach of ordinary structure, has to completely masticate his food and swallow it once for all as soon as it is gathered, and this, too, in places where he may be exposed to attack from enemies. Consequently, it is of vital importance that the process of mastication should be accom- plished not only in the most efficient manner, but likewise with the greatest possible rapidity. Hence the complexity and powerful grinding action of his cheek-teeth. There is, however, another point in connection with these same cheek-teeth to which attention may now be directed. As already mentioned, these teeth are characterised by the great vertical height POSITION AND STRUCTURE OF HORSE 39 of their crowns ; in which respect, as will be shown in the sequel, they differ from the corresponding teeth of the horse's early ancestors, which had quite short crowns. Now, if surrounding con- ditions be the same, tall-crowned teeth indicate the potentiality of much longer life on the part of their owner than is afforded by low-crowned ones ; as it is obvious that a tall tooth will take much longer to wear down than will a low one. In this particular instance it has, however, to be borne in mind that the early ancestors of the horse were swamp-dwelling animals living on soft, luscious vegetation which could be masticated without having much effect on their teeth. The horse and its relatives, on the other hand, when in a state of nature, live on open plains where the grass is often more or less hard and wiry, and thus calculated to wear away the teeth at a relatively rapid rate. In the case of domesticated horses the rate of wear is probably still further accelerated by the nature of the food. Still, after making due allowance for all this, there can be little doubt that the existing members of the horse tribe are longer-lived animals than were their early forerunners. And, as animals go, the domesticated horse may be considered to have a considerable pre-eminence in the matter of longevity, although in this respect it does not equal its distant cousins the rhinoceroses, some of 40 THE HORSE AND ITS RELATIVES which have been known to Hve for considerably more than half a century. One of the oldest — if not actually the oldest — horses on record was an Australian, referred to in the Field newspaper of March i8, 1905. This horse is stated to have been foaled on November 16, i860, and was still living at the date of the afore- said notice, when he would have been rising 45. I never saw a notice of his death. The celebrated Godolphin Arabian, or Barb, who died in 1753, was brought from Paris 25 years previously, where he is reputed to have been drawing a water-cart : from this it may be inferred that he was well over 30 at the time of his death. His grandson " Matchem " is believed to have reached -^-^ years, while " Diomed," the winner of the first Derby, is reputed to have attained the age of 30 or 31. " Pocahontas," again, died in 1870 at the age of -^Z 5 while " Touch- stone," who was foaled in 1831, died in 1861. Records of race-horses living to ages of between 23 and 27 years are comparatively common. Finally, it may be mentioned that the Duke of Bedford possesses the skull of a horse w^iich, owing to infirmity, was shot when close upon 38 years of age. " Prince," as this animal was called, was one of the old small and sturdy Galloway cart-horse breed, now nearly extinct. He was foaled in Wigtownshire, and taken over at a valuation with an estate pur- chased by the Duke. POSITION AND STRUCTURE OF HORSE 41 Although it does not come within the scope of the present volume to give a complete account of the structure of the horse, such as is to be found in the numerous works on veterinary anatomy, there are certain structural features, in addition to those already mentioned, which demand special notice on the ground of their morphological interest. The first of these is a wartlike structure buried in the tuft of long hair on the hind surface of each foot, which gives the name of fetlock {i.e. footlock or feet- lock) to this segment of the limb (pi. vi. fig. i). When the tuft of hair is cut away, there will be revealed on the summit of a fatty cushion a bare patch covered with a warty growth to which French veterinarians have oiven the name of er^ot, a word properly signifying the spur of a cock. The ergot is relatively larger in the ass than in the horse. Sir William Flower^ appears to have been the first to point out the true significance of the ergot, which really represents the large fatty pad or cushion on the sole of the foot of a dog situated above, and to a slight extent between, the four smaller toe- pads. Although now a rudimentary, or vestigial, structure, it was evidently functional in the early ancestors of the horse, which applied a considerable portion of the side of the foot to the ground, instead of resting only on the tip of the middle toe. Although the term hoof is generally applied ^ The Horse, Modern Science Series, London, 1891, p. 168. 42 THE HORSE AND ITS RELATIVES to the whole of the terminal segment of the limb, it properly denotes — at all events, in an anatomical sense — only the dense horny shell or wall investing the front and lateral surfaces, and corresponding to the nail of the human middle finger or toe and the claw of the same toe in a dog. To the trian- gular, less hard, horny structure projecting from the back into the centre of the lower surface of the foot the name " frog " is applied. In all members of the horse tribe the terminal segments of the fore and hind feet are remarkably alike ; more alike, in fact, than in any other animals. In the horse itself this similarity is, however, some- what less marked than in most of the other members of the group, the front hoofs being broader and rounder than the hind pair. In the kiang ^ of Tibet this difference is less marked, and in asses and zebras all the hoofs are relatively small and narrow, so that it is practically impossible to dis- tinguish the front from the hind ones when sepa- rated from the rest of the limb. In addition to the difference in the shape of the hoofs of the horse and kiang as compared with those of other members of the family, there are also specific differences in regard to the form of the frog. In the horse, for instance, the frog forms a long narrow ridge, deeply grooved posteriorly, ' The characteristics of this and other species mentioned in this chapter are given in the sequel. POSITION AND STRUCTURE OF HORSE 43 which does not extend behind the extremities of the outer case of the hoof, and is not applied to the ground in walking. In the North African Gravy's zebra, on the other hand, the frog is broader, with scarcely any trace of the posterior groove, and its hind-part touches the ground when the animal is standing. In the kiang, and probably also in the Asiatic onagers, the posterior development of the hoof becomes much more marked, so that a considerable portion projects behind the case of the hoof and touches the ground, the cleft being deep and narrow. Still greater development of the hind part of the frog occurs in the ass, in which, as in some zebras, it also becomes much thickened and somewhat spongy in structure. In the extinct South Ameri- can Onohippidium, the frog is somewhat inter- mediate between that ot the horse and that of the ass, being grooved, and not projecting behind the case of the hoof, but being of considerable breadth and thickness. Finally, in the bontequagga, and probably the quagga, the medium-sized and slightly- cleft frog is deeply sunk in the hoof, behind which it projects to a small degree, not touching the ground, except when the hoof is much worn down. These differences are probably correlated with differences in the nature of the habitat of the various species, and it is probable that species like the horse, in which the frog is narrow, are adapted for grassy or sandy plains ; while those in which it is 44 THE HORSE AND ITS RELATIVES broad, deep, and spongy, like the ass and the kiang, are better adapted for rocky ground. This is confirmed by the fact that the kiang does inhabit extremely stony and rocky country. By far the most remarkable and interesting structures in the limbs of the horse are those com- monly known as " chestnuts," although sometimes called " castors," but in old French veterinary works termed "sallenders" (from salendre) or " mal- lenders," from an idea that they were due to dis- ease. In the fore-limb of the horse the chestnut (pi. vi. fig. 2) or callosity, takes the form of an elongated patch of bare warty skin situated some little distance above the knee, or carpus ; while in the hind-limb there is a smaller patch — which may be absent in some cases — a short distance below the hock or tarsus, and likewise on the inner surface. In all the other members of the family the chest- nuts are wanting in the hind-limbs ; and in the ass and the zebras the front chestnut is larger, smoother, and softer than in the horse. The question as to what structures in other mammals are represented by these chestnuts has been much discussed. Practically all naturalists are in accord in regarding them as vestigial struc- tures ; Sir W. H. Flower,^ for instance, considering them to be decadent skin-glands. A qualified support to this theory is accorded by Mr. F. E. ' The Horse, p. 165. < (U t/) POSITION AND STRUCTURE OF HORSE 45 Beddard,^ who in one passage states that the chest- nuts on the fore-limbs probably correspond to glands found in the neighbourhood of the wrist or carpus in certain other mammals, although on a subsequent page the glandular nature of these structures is questioned. In another and apparently later com- munication the same writer ^ suggested that the front chestnuts may represent a carpal sense- organ, of which remnants are believed to exist in the bristles on the wrists of the African hyraxes, the sole survivors of a formerly numerous group of unofulates. This degfeneration of such a sense- organ might, it is suggested, result in the formation of a structure like a chestnut. On the other hand, there exists an idea that the chestnuts represent vanished toes or foot-pads. This theory has been supported by Professor J. C. Ewart^ of Edinburgh, and more recently by a German writer, Mr. R. Hintze,'* who compares the hind chestnuts to the pads on the foot of a kangaroo. If, however, the identification of the horse's ergot with the hind foot-pad of the tapir and the dog be admitted — and the evidence in its favour is very strong — it is practically certain that the chestnuts cannot represent foot-pads, much less ^ Cambridge Natural History — Mammalia, pp. 12, 13, and 240. * Proc.Zool. Soc. London, 1902, vol. i. p. 135. ^ See Nature, London, vol. Ivii. p. 239, 1903. * Zool. Anzeiger, Leipzig, vol. xxv. pp. 372-382, 1910. 46 THE HORSE AND ITS RELATIVES vanished toes. Pursuing this aspect of the subject still further, it should be borne in mind that, as I have pointed out in an article in the Proceedings of the Zoological Society of London for 1903, the chestnuts of the horse are situated on the inner surface, whereas, if they represented vestigial foot- pads, their position should be, prima facie, on the hind aspect, as is the case with the ergot. It might, indeed, be argued that they have changed their original position, but of such a shifting there is no evidence in the adult horse. A second, and perhaps more important, objection to the foot- pad theory may be drawn from the fact that the chestnuts in the fore-limb are situated above the so-called knee-joint (carpus), and are therefore altogether higher up than any of the foot-pads of plantigrade mammals. Unless, therefore, another shift of position has taken place, the fore-chest- nuts do not represent foot-pads. This argument was used by Sir W. H. Flower to disprove the theory that the chestnuts are remnants of lateral toes. The hind-chestnuts, on the contrary, are situated a short distance below the joint of the hock (tarsus), and are therefore on a part of the limb, although on its inner side, which is included in the foot of a plantigrade mammal. If, however, the front-chest- nut be regarded as corresponding in a general way with the hind one, it will be evident that in the POSITION AND STRUCTURE OF HORSE 47 event of the former not being a foot-pad, the same will hold good for the latter. A third, and perhaps stronger objection may be urged against the foot-pad theory. On the assumption that the chestnuts of the existing members of the Eqzndcs are vestiges of foot-pads, it is clear that these structures must have existed in the ancestors of that family since the time when such ancestors walked on the entire sole of the foot, in the plantigrade fashion ; but, so far as I know, no ungulate was ever wholly plantigrade in both feet ; the nearest approach to this condition obtaining in the Lower Eocene Coryphodon, in which the hind-limb was wholly plantigrade, while the front one was partially digitigrade. It has thus to be assumed, on the foot-pad hypothesis, that the front-chestnuts of the horse have been functionless structures from a period antedating the evolution of the Ungulata. Such a persistence, on exposed parts of the body, of a functionless structure seems im- probable, especially when the modifications are borne in mind which, on this hypothesis, the horse- line must have undergone since the time when the chestnuts were functional structures. Perhaps the case of the ergot may be cited against this argu- ment ; but it should be remembered that this structure probably acted as a functional pad at a later stage of evolution than could have been the case with the chestnuts. 48 THE HORSE AND ITS RELATIVES Having now stated some of the objections against identifying the horse's chestnuts with the foot-pads of polydactyle mammals, it remains to consider whether they can be identified with any other structures. Now a certain number of repre- sentatives of the deer family — notably the reindeer, the white- tailed deer, the mule-deer, and, in a rudimentary condition, the elk — are furnished on the inner side of the hock with a glandular tuft corresponding very closely in situation with the hind -chestnut of the horse. In fact, the only difference in the position of the two structures is that the tarsal tuft of the deer is placed rather lower on the hock. From the fact of its occurrence in deer so widely separated from one another as are the species mentioned, it seems evident that the tarsal gland (which is doubtless a scent-organ) is a very ancient structure, which was present in all the ancestors of the group, but has been lost, probably from disuse, in the great majority of Old World deer. Judging from their position, there would seem to be a certain probability that the hind-chestnuts of the horse and the tarsal glands of the deer are corresponding structures. With regard to the correspondence of the fore- chestnut of the horse, it may be mentioned that many gazelles have tufts of hair ("knee-brushes") at the knee (carpus), which are certainly glandular POSITION AND STRUCTURE OF HORSE 49 in nature. And it is possible that these may represent the fore-chestnuts of the horse, for there seems no good reason why the position of a gland should not have somewhat shifted in two widely separated groups of mammals. Then, again, we have the carpal bristles of certain mammals, such as the South American coatis and the hyraxes, already referred to as the remnants of a " scent- organ " — a structure probably not far removed in its nature from a gland. The occurrence of these bristles in the hyraxes is very important. Mr. Beddard states that these are the only ungulates in which he has found these bristles. Carpal callosities are, however, described by Dr. W, Leche^ as occurring in wart-hogs [Phacochoerus) ; although they are stated by their describer to be acquired, and not primitive structures. Of special importance is the occurrence of bristles in these structures, since, even if hairs be found to exist on the callosities of foetal Equidcs, this would be no bar to the supposition of their glandular nature. More recently, in the Btdletin de la Societe Scientijique et Medicale de I'Ottest for July 1909, Veterinary Surgeon J. Roger directed attention to the presence on the inner and hind surface of the fore-legs of pigs, in proximity to the wrist or carpal joint, of a patch of large pores, which in certain circumstances exude a transparent and slightly ^ Biol. Centralblatt, vol. xxii. p. 79, 1902. D 50 THE HORSE AND ITS RELATIVES glutinous fluid. After describing these sudoriparous glands in some detail, the writer suggested that they correspond to the chestnuts of the horse, or rather that the chestnuts represent the sudoriparous glands of the pigs in a decadent condition. The fact that those of the horse yield when cut a sticky, strong-smelling fluid favours this interpretation, which also accords fairly well with the theory that the chestnuts represent structures more or less similar to the foot-glands of deer. As regards the structure of the chestnuts them- selves, it may be noted that in the horse both pairs are of a distinctly warty nature, and that the hind pair is certainly in a more decadent condition than the other, being in fact on the verge of disappearing. In the zebras, on the other hand (in which the hind one has been lost), the fore-chestnut is larger and much less warty and also situated higher up. In dried skins it is, in fact, much more like the pale glandular patch of skin below the ear of a reedbuck. In this connection not only should Mr. Beddard's observations be borne in mind, but attention may be directed to others by Mr. Bland Sutton,^ in which it is pointed out that in certain lemurs decadent glands are actually converted into bunches of spines, which are practically almost the same as warts ; that is to say, they are excessive growths of some- what abnormal dermal tissue. Hence there seems ^ Proc. Zool. Soc, 1887, p. 369. POSITION AND STRUCTURE OF HORSE 51 no prima facie reason why the chestnuts of the Equidce should not be decadent glandular structures, the decadence being more marked in those of the horse than in the single pair of the asses and zebras. There is, however, another point which may have an important bearing on the subject. As already mentioned, the presence of a depression in the skulls of certain extinct three-toed horses renders it probable that primitive horses were furnished with face-glands comparable to those of deer ; such glands probably having a function somewhat ana- logous to that of the scent-glands on the limbs of the latter. If, then, the existing Equidce have got rid of their face-glands, as being (perhaps on account of change of habit) useless, it is conceivable that, for the same reason, they may have also discarded their limb-glands. And there is some reason to believe that such a change of habit has taken place during the evolution of the family. Mr. R. I. Pocock,^ for instance, in a paper on the primitive colouring of the members of the horse tribe, to which fuller reference is made later, has suggested that the ancestral animals were inhabitants of forests, in- stead of open plains, remarking that this primitive type of colouring " would lend itself especially to concealment in horses accustomed to shelter in ^ Ann. Mag. Nat. Hist., London, ser. 8, vol. iv. p. 409, 1909, 5 2 THE HORSE AND ITS RELATIVES woods through the foHage of which the sun-rays passed, dappHng the leaves and tree-trunks with spots of light." Now if such a change of habit has taken place, nothing would seem more likely than that it should have been accompanied by the loss of scent-glands, which would not be required among animals living in studs ^ on open plains in order to ascertain the whereabouts of their fellows. There is, however, yet one more point in favour of the view that the chestnuts represent decadent scent-glands. As already mentioned, these struc- tures exude, when cut, a strong-smelling fluid ; and I am informed that this fluid will not only attract other horses, but that it was formerly employed by burglars and poachers to keep dogs quiet. If this be true, the fluid must almost certainly represent the secretion of an ancestral scent-gland. Attention may now be directed to certain features connected with the colouring of the hair in the existing members of the horse family. It has long been noticed that dun-coloured domesticated horses frequently show a tendency to develop dark bars on the legs, and sometimes one or two transverse dark stripes across the shoulder and another along the middle line. And since similar markings occur ^ Although the word " stud " is now used to denote a stable of horses, it originally denoted (Anglo-Saxon s/ud, Slav, stod) a drove of wild horses, for which it is the proper term. See Heyn and Stally- brass, Wanderings of Plants and A?iinials, London, 1885, p. 39. POSITION AND STRUCTURE OF HORSE 53 in the wild asses of Africa, while the zebras and quaggas of that country are marked by dark and light stripes over the whole or a considerable portion of the head, body, and limbs, it has been considered probable that all the ancestral members of the family were fully striped. Consequently species like the Mongolian wild horse, the Asiatic kiang, chigetai, and onager, and the North African wild ass, which are more or less nearly self-coloured, are presumed to have lost their stripes in accord- ance with the special conditions of their natural surroundings ; and also that the tendency in dun horses, and it may be added mules, to the develop- ment of stripes is an instance of atavistic reversion. In addition to this tendency to develop stripes on the limbs and shoulders in dun-coloured horses and mules, there is a still stronger tendency among domesticated horses of all colours except duns, but more especially greys, to show dappled markings. Attention was directed to this feature by Darwin,^ who stated that it occasionally occurs among asses, and who expressed the opinion that it was probably connected in some way with the ancestral striping. At a much later date Dr. E. Bonavia, a brigade- surgeon in the Indian Medical Department, in a very remarkable book entitled Studies in the Evolution of Animals^ laid still greater stress on the frequent ^ Animals and Plants under Domestication^ 2nd ed. vol. i. p. 58, London, 1885. ^ London, 1895. 54 THE HORSE AND ITS RELATIVES occurrence of dappling" in horses of all colours, and suggested that it must be a very deep-seated and ancient character of the group. He even went so far as to suggest that the dappling of the horse represents the rosettes on the leopard's skin, and that the latter are derived from the pattern on the bony plates of the armour of the extinct giant arma- dillos, or glyptodonts, of South America. Now although this theory is certainly untenable it has the merit of recognising that dappling is a feature deeply implanted in the equine nature. Giving full credit to the discoverer of this fact, Mr. R. I. Pocock,^ who has done much to illustrate the meaning of animal coloration, has taken up the subject of the colours of domesticated horses, and concludes that their colouring may be classed under three main types. The first of these types com- prises bays, blacks, chestnuts, roans, piebalds, and skewbalds ; the second includes duns alone ; while the third is represented by greys and the majority of whites. From the fact that in all wild members of the horse family the mane and tail are darker in colour than the body, it is inferred that bay is the original phase of the first type. On this phase three modifications have been working, namely, blackness, or melanism, to give rise to browns and blacks ; redness, or erythrism, to produce chest- nuts ; and whiteness, or albinism, to develop (in ^ Ann. Mag. Nat. Hist., London, ser. 8, vol. iv. p. 404, 1909. POSITION AND STRUCTURE OF HORSE 55 conjunction with melanism and erythrism) piebalds and skewbalds, and in another direction greys, and finally whites. That this is not a mere fanciful suggestion is made evident by the fact that melanism, erythrism, and albinism are recog^nised features in the colour- development of wild animals. It is added that chestnuts, from the extension of the red to the mane and tail, may be regarded in the light of " sports." It has been suggested by Professor W. Ridge- way, in his work On the Origin of the Thorough- bred Horse^ that the white forehead-star and white "stockings" so often observed in chestnut thoroughbreds are hereditary features derived from the ancestor of that breed ; but this idea is rejected by Mr. Pocock, who shows that such markings are to be regarded as first steps in the direction of albinism, and are consequently in no sense ancestral. Dappling, as already mentioned, may occur in horses of all colours, but is most common in bays and greys and rarest of all in duns ; this prevalence being the main justification for Dr. Bonavia's view that it is an extremely ancient feature in the equine organisation. Mr. Pocock believes, indeed, that dap- pling is even older than striping, and he has also been led to conclude that the ancestral forms of the horse family were dark-coloured animals marked with white or yellowish flecks or spots arranged in ^ Cambridge, 1905. 56 THE HORSE AND ITS RELATIVES much the same manner as are those on the coats of young tapirs. A certain amount of support to this theory is afforded by the circumstance that it tends to bring two famihes of the odd-toed section of the hoofed mammals into line with one another in the matter of colouring. These hypothetical white markings of the ances- tral members of the horse tribe are presumed to have taken in the first instance the form of simple round spots, which subsequently tended to arrange them- selves in lines and then to coalesce into lono^itudinal streaks, but finally underwent a rearrangement, so as to produce light transverse bars or stripes on the greater portion of the head and body. On this view, zebras should be regarded as black or brown animals with white or buff stripes, and not, after the ordinary fashion, as white ones with dark stripes ; this idea having been suggested some years previous to the publication . of Mr. Pocock's paper by Sir H. H. Johnston. Similarly, there appear to be good grounds for considering that giraffes were originally brown or red animals with vertical white stripes arranged like those of elands and their cousin the bongo antelope of the forests of Equatorial Africa. It may be added that dappling occurs only in domesticated horses and occasionally in asses and mules, and that it is very rare in duns. From its absence in the wild Mongolian horse, which may POSITION AND STRUCTURE OF HORSE 57 safely be regarded as the ancestor of many of the domesticated European breeds, It may be suggested that dappling is an attribute of the Arab stock, which, as will be shown later, there is considerable reason to regard as being derived from a species different to the one which gave rise to the original domesticated horses of Western Europe. It is quite true that there are difficulties in regard to this suggestion ; one of them being that we have no information as to when Arab or Barb blood was first introduced among the horses of Western Europe, while we are equally in the dark as to whether dappling always occurred among the latter, or whether it is a foreign feature. It does, however, appear very strange that if dappling be a remnant of an ancient type of colour- ing at one period characteristic of the horse tribe in general, it should have completely died out in all wild species, to reappear in the domesticated breeds of Equtis cabalhis. And the only way out of the difficulty seems to be the above suggestion, that the progenitor of the Arab and Barb was a dappled bay horse. As regards the inheritance of coat-colour among horses, it appears, according to Mr. R. Bunsow,^ that in the case of thoroughbreds, bays (including browns) may be either pure as regards the power of transmitting their colour to their offspring, or ^ The Mendel Journaly London, No. 2, p. 74, 191 1. 58 THE HORSE AND ITS RELATIVES impure, when they may give rise to chestnuts. It thus follows that bays, as being capable of produc- ing offspring of a colour different to their own, are a dominant type (D), while chestnuts, which lack this capacity, are recessive (R). Chestnut horses, as having but one kind of sexual cells, may accord- ingly all be symbolised as RR, whereas bays may be classed either as DD or DR, according as to whether they are pure or whether they contain an admixture of chestnut cells. Now if a DD stallion be mated with an RR mare, all the foals will be DR bays. On the other hand, the foals of an RR mare by a DR stallion will, in the long run, consist of bays and chestnuts in nearly equal numbers. When chestnuts are bred together, their offspring should be all chestnuts (RR), but if chestnuts be crossed with bays, the foals may be either all bays or half chestnuts (RR) and half bays (DR), the former case, as mentioned above, being due to the fact that the parent bays were DD, and the latter to their being DR. Certain apparent exceptions to these conditions occurrincj in the Stud-book are shown to be due to incorrect registration of colour, and it is probable that the same is the case with all the rest. As regards greys, it is stated that in all cases one of the parents must be of this colour. At the meeting of the British Association held at Portsmouth in 191 1 Mr. C. C. Hurst discussed the question whether there is any connection POSITION AND STRUCTURE OF HORSE 59 between the coat-colour and the speed of thorough- breds. Although such a relationship appears to be generally lacking, evidence is gradually accumulat- ing to suggest that in certain strains there may be a partial coupling of coat-colour and racing power. For instance, the chestnut grand-children of the famous thoroughbred "St. Simon" have so far proved themselves to be much inferior in racing power to their bay and brown brothers and sisters. While these chestnuts have between them won only two first-class races, their bay and brown brothers and sisters have between them won fifteen classic races, and are only about twice as numerous. Another interesting point under investigation was the ap- parent partial conjunction of brown coats, high racing power, and female sex in St. Simon's own offspring. St. Simon's brown fillies proved them- selves to be strikingly superior in racing power to the bay fillies, the brown colts, and even the bay colts, a few individuals of which were extraordinarily good. This was the more remarkable when it is considered that in racing colts have many advantages over fillies. It seems possible that the elucidation of such an apparently trivial thing as coat-colour may help to throw light on the more complicated question of the breeding of a first-class winner. Certain abnormalities in the structure of the skeleton of the horse occur from time to time, of which it will suffice to refer to three. The first of 6o THE HORSE AND ITS RELATIVES these occurs in the bones of the foot, and takes the form of the more or less complete development of one or both of the lateral toes normally represented by the splint-bones. In certain instances, as has been pointed out by Sir William Flower, when only one of these lateral toes is developed, the abnor- mality does not apparently indicate reversion to an ancestral type, but seems due to a splitting of the bones of the main toe. In other cases, however, there seems little doubt that such supplemental toes are really a reversion to the condition obtaining in the extinct three-toed members of the family. An instance of this kind of reversion is exhibited in the foot of a shire colt formerly in the possession of Lord Wantage, by whom the specimen was pre- sented to the British Museum. The metacarpal bones of the fore-foot are complete, although vary- ing in size, and the terminal toe-bones carried complete hoofs. Apparently this reversion to the three-toed type occurs only among domesticated horses. A second abnormality among domesticated horses is displayed by the development of rudi- mentary horns, or rather horn-cores, on the fore- head (pi. vii. fig. i). In his Animals and Plants under Domestication^ Darwin wrote that "In various countries horn-like projections have been observed on the frontal bones of the horse : in 1 Vol. i. p. 52. PLATE VII Fig, I Fig. 2 Fig. I. Frontlet of Horse, showing horn-like processes. Fig. 2. Outline of the Prehistoric Tarpan or Wild Horse, incised on a piece of horn, from the Madelaine Rock- shelter in the Department of Dordogne, France. POSITION AND STRUCTURE OF HORSE 6i one case described by Mr, Percival ^ they arose about two inches above the orbital processes, and were very like those in a calf from five to six months old, being from half to three-quarters of an inch in length. Azara^ has described two cases in South America in which the projections were between three and four inches in length : other instances have occurred in Spain." The same abnormality is displayed in four specimens exhibited in the British Museum (Natural History). The first of these is the skull of an English horse presented by Mr. Hanbury Carlile ; while of the three other specimens of the same type, one is the frontal region of the skull of an English horse showing the pair of rudimentary horns in precisely the same position as in the first specimen, but of somewhat larger size. The other two are models of the foreheads of thoroughbreds, each showing a pair of similar horns, situated as in the preceding specimens. These are important as showing that the skin extends uniformly over the horn-like processes, without any trace of a dermal horn ; the same condition being observable in the other two examples. Baron Francis Nopcsa informs me that he knows of a horse in Transylvania with rudimentary horns. The significance of these horn-like growths is at present inexplicable, seeing ^ The Veterinary, vol. i. p. 224. ^ Quadrupedes du Paraguay, vol. ii. p. 313, 1801. 62 THE HORSE AND ITS RELATIVES that none of the ancestral horses, or even of the collateral branches of the horse-stock, were horned animals. Probably it is a kind of redundant deve- lopment, such as occurs in the later types of many groups of animals ; seeing that the members of the horse family have efficient fighting weapons in their hoofs, and that they are also largely pro- tected from foes by their speed. It would be of great interest if it were possible to ascertain whether all these "horned horses" were of Arab or Barb descent. One other abnormality, and this an individual one, may also be referred to. In the limb-bones of the celebrated thoroughbred " Stockwell " (1849- 1876), which are exhibited in the British Museum (Natural History) the projection on the hind border of the femur, or thigh-bone, known as the third trochanter icoide supi^a^ p. 8) is almost obsolete ; and it would be interesting to ascertain if the same feature characterises the skeleton of his descendants, if any of these have been preserved. It seems natural to suppose that the practical absence of the trochanter would have had some effect on the action of Stockwell ; and it is very note- worthy that in describing the sale of the Burghley Stud, a writer signing himself " The Druid " states in Post ajid Paddock, 1857, p. 296, that "Stockwell came ambling out in his peculiar style, with his Roman head and massive muscular points much POSITION AND STRUCTURE OF HORSE 63 fined down since he all but broke Teddlngton's heart." This seems to suggest that the horse had a distinctive action of his own. In this place a few lines may be devoted to the action and position of the limbs of a horse at the gallop, and the conventional modes of repre- senting the same. In this action the movements of the limbs are so rapid that, like the spokes of a quickly revolving wheel, the relative positions of the limbs at any particular moment cannot be appreciated by the human eye. Instantaneous pho- tography shows, however, that all the conventional modes of representing the galloping horse are untrue to nature. Sir E. Ray Lankester in one of his articles published in the Daily Telegraph under the title of " Science from an Easy-Chair," has, for instance, pointed out "that what has been drawn by artists and called ' the flying gallop,' in which the legs are fully extended and all the feet are off the ground, with the hind-hoofs turned upwards, never occurs at all in the galloping horse, nor anything in the least like it. There is a fraction of a second when all four legs of the galloping horse are off the ground, but they are not then extended, but, on the contrary, are drawn, the hind ones forward, and the front ones backward, under the horse's belly. A model showing this actual instantaneous attitude of the galloping horse has been placed 64 THE HORSE AND ITS RELATIVES in the Natural History Museum. When the hoofs touch the ground again after this instantaneous hfting and bending of the legs under the horse, the first to touch it is one of the hind-lees, which is pushed very far forward, forming an acute angle with the body. The shock of the horse's impact on the ground is thus received by the hind- leg which reaches obliquely forward beneath the body like an elastic spring. Since the instantaneous photographs have become generally known, artists have ceased to represent the galloping horse in the curious stretched pose which used to be familiar to every one in Herring's racing plates, with both fore- and hind-legs nearly horizontal, and the flat surface of the hind-hoofs actually turned upwards ! " Later on in the same article it is mentioned how M. Solomon Reinach has shown "that in Assyrian, Egyptian, Greek, Roman, mediaeval, and modern art up to the end of the eighteenth century * the flying gallop ' does not appear at all. The first example (so far as those schools are concerned) is an engraving by G. T. Stubbs in 1794 of a horse called Baronet. The essential points about ' the flying gallop ' are that the fore-limbs are fully stretched forward, the hind-limbs fully stretched back- ward, and that the flat surfaces of the hinder hoofs are facing upward. After this engraving of 1794 the attitude became generally adopted in English art to represent a galloping horse. . . . Reinach POSITION AND STRUCTURE OF HORSE 65 has shown that in the pre- Homeric art of Greece — that which is called * Mycenaean ' (of which so much was made known by the discoveries of that wonderful man Schliemann when he dug up the citadel of Agamemnon) — the figures of animals, horses, deer, bulls (see the beautiful gold cups of Vaphio ! ), dogs, lions, and griffins, in the exact conventional pose of 'the flying gallop,' are quite abundant ! There was an absolute break in the tradition of art between the early gold-workers of Myken^ (1800 to 1000 B.C.) and the Greeks of Homer's time (800 b.c). Europe never received it, nor did the Assyrians nor the Egyptians. Thirty centuries and more separate the reappearance in Europe of ' the flying gallop ' — through Stubbs — from the only other European example of it — the Mycenaean. What, then, had become of it, and how did it come to England? M. Reinach shows by actual specimens of art-work that the Mycenaean art tradition, and with it * the flying gallop,' passed slowly through Asia Minor into ancient Persia, thence by Southern Siberia to the Chinese Empire, as early as 150 B.C., and that 'the flying gallop,' so to speak, * flourished ' there for centuries, and was transmitted by the Chinese to the Japanese, in whose drawings it is frequent. It was at last finally brought back to Europe, and to the extreme west of it, namely, England, by the importation in the eighteenth century into England of large 66 THE HORSE AND ITS RELATIVES numbers of Japanese works of art. M. Reinach thinks that * the flying gallop ' was devised as an intentional expression of energy in movement. I venture to hold the opinion that it was observed by the Mycenseans in the dog, in which Muybridge's photographs (now before me) demonstrate that it occurs regularly as an attitude of that animal's quickest pace or gallop. It is easy to see 'the flying gallop* in the case of the dog, since the dog does not travel so fast as the galloping horse, and can be more readily brought under accurate vision on account of its smaller size. It is quite in accordance with probability that the early Mycenaean artists, having seen how the dog gallops, erroneously proceeded to put the galloping horse and all other animals which they wished * to make gallop ' into the same position." - This chapter may be brought to a close by a few remarks in regard to the past and present geographical distribution of the horse family and a brief reference to certain Hindu myths and customs relating to the horse. At the present day the group, in a wild state, is restricted to the Old World, where it is widely distributed in Asia and Africa, while, as is shown in the next chapter, it existed at a comparatively recent date in Eastern Europe. In Asia it does not, however, extend farther east than Western India and Mongolia, or further north than the latter country and Tibet. The Asiatic POSITION AND STRUCTURE OF HORSE 67 representatives of the group include the wild horse and the animals known as chigetais, kiangs, and onagers, which serve in some degree to connect the former with the ass. The latter, in a wild state, is restricted to North-eastern Africa ; while the east and south of the African continent form the home of the striped members of the family, commonly known as zebras and quaggas. During the Prehistoric Stone Ages, as is more fully indicated in later chapters, the range of some of the living Asiatic members of the group ex- tended to Western Europe. At an earlier date (Pliocene and Pleistocene) extinct species of horses roamed over Central and Northern India. And at some period horses doubtless ranged right across Asia to Bering Strait, as remains of a species closely allied to the existing horse occur on the opposite side of the Strait in the frozen soil of Alaska. Moreover, throughout the later portion of the Tertiary period numerous extinct representatives of the family inhabited North America, where horses or asses were quite unknown when that continent was discovered by Columbus. Nor is this all, for quite late in the Tertiary period, when South America, which had previously been isolated for a long epoch, became connected by land with North America, members of the family traversed the Isthmus of Darien, and made their way into Argentina and Patagonia, where they developed 68 THE HORSE AND ITS RELATIVES into several altogether peculiar types. It is generally believed that at the time South America was explored by the Spanish conquistadores the native wild horses had become exterminated, although it has been suo-p^ested that certain horses seen by John Cabot in Argentina in 1530 were remnants of the indigenous stock, as it is diffi- cult to see how introduced animals had reached that part of the country at such an early date. As regards the birthplace of the family, it was suggested by a famous American naturalist, Pro- fessor E. D. Cope, that the horse tribe had two independent centres of development from animals of a more primitive type, one in the Old World, and a second in North America. On the face of it, this is, however, a very improbable theory, and a more plausible suggestion has been subsequently made by another American naturalist, Dr. W. D. Matthew,^ who wrote as follows : " I assume that since the oreodonts [extinct American hog-like animals] and peccaries [the New World representatives of the swine] never reached the Old World, and the camels did not reach it till the Pliocene, their centres of dispersal were well to the south of the [ancient] Bering Sea connection with the Old World. I assume that since the horses are repre- sented by a double evolutionary series, one in ' "The Continuity of Development," Popular Science Monthly, New York, 1910, p. 473. POSITION AND STRUCTURE OF HORSE 69 Europe, a closer one in North America, their centre of dispersal lay far enough north to spread into Europe on the one hand, and North America on the other, but that the latter was nearer and more accessible ; i.e. that centre of dispersal was North-eastern Asia or Alaska." To deal adequately with the subject of the im- portant part played by the horse in the development of civilisation would require as much space as is contained in the whole of the present volume ; but it may be of interest to refer to certain Hindu myths relating to the origin of the horse, especially as these have a curious bearing on the classical fable of Pegasus, the flying horse. In the introduction to an Indian work on the horse Colonel D. C. Phillott ^ writes on this subject as follows : — "According to Hindu legends, the horse was created a winged animal, one that could fly and run, and no man or god could snare it. Indra wanted horses for his chariots, and requested the sage Salihotra to deprive the horses of their wings. Accordingly Salihotra, by yoga, a supernatural power, derived by his austerities, accomplished Indra's wish. The horses, now deprived of the ability to visit far-off jungles in search of medicinal herbs, approached Salihotra and entreated him to 1 The Faras-Nama-e Rait^in, or the Book of the Horse, London, 1911. 70 THE HORSE AND ITS RELATIVES write a book on the treatment of their diseases. Salihotra consented, and composed the greatest work on veterinary science known to the Hindus. This work was called Salihotra after him ; gradually this Sanskrit word came to mean veterinary science in general and also a horse. To-day every regiment of native cavalry has its salotris." Later on the same author observes that " Besides its use in war, the horse was important in Hindu eyes as an animal of sacrifice. ... In the Vedic period the sacrificial horse was first slain sacrificially {i.e. by severing the head at one blow), and then divided in portions, part being eaten by the attendant priests, and part being offered as a burnt- offering. In this age the object of the sacrifice was to obtain wealth, prosperity, and male offspring. " The Puranas, written several hundred years after the Vedas, describe the asvaniedha as a sacri- fice of the highest order. Performed a hundred times, it elevated the sacrificer to the throne of Svarga, Indra's dominion, deposing even the king of the gods. . . . "In the succeeding epic period, e'.^. after 1 200 B.C., this sacrifice was made by kings to demonstrate their claim to supremacy over neighbouring chiefs." What, if any, connection there may be between these ancient Hindu sacrifices and the slaughter of horses at the obsequies of the old Scandinavian chieftains, must be left for others to determine. CHAPTER II THE WILD TARPAN AND ITS RELATIONS Since the time of Cuvier it has been known that teeth and bones of horses are of common occurrence in the Prehistoric and other superficial formations of the continent of Europe and the British Isles, and consequently that wild horses inhabited this area contemporaneously with the mammoth, the woolly rhinoceros, and the men of the Stone Age. There is, moreover, abundant evidence to show that during that long-past age primitive man hunted these wild horses for the sake of their flesh ; the long bones of the limbs being split for the purpose of extracting the marrow, while in some instances the brain-chamber of the skull has likewise been fractured and opened in order that the brain itself might be used for food. Nor is this all, for some of the men of the Stone Age have left in the caves which afforded them shelter from the weather crude but life-like outline sketches of the horses they hunted for food, and subsequently domesticated ; these sketches serving to show that, in certain instances, at any rate, these Stone Age horses were closely allied to the existing tarpan, 72 THE HORSE AND ITS RELATIVES or wild horse, of the Zungarian district of Mongolia, having big heads, but relatively slender limbs. These Stone Aq-q horses will be described later on in this chapter ; and attention may now be directed to the evidence relating to the survival of wild horses in Western Europe during the historic period. In this connection it will be con- venient to quote the summary of the evidence given by Messrs. Heyn and Stallybrass in their work entitled TAe Wanderings of Pla7its and Animals} After pointing out that wild horses exist in Mongolia, the authors proceed as follows : — " That the horse in its original wildness also roamed westward to Turkestan, over the steppes of the present South-eastern and Southern Russia, and to the foot of the Carpathians, seems likely enough ; not so likely that even the forest regions of Central Europe once abounded in troops of that animal. And yet much historical testimony seems to put the fact beyond a doubt. Varro speaks of Spanish wild horses ; and Strabo writes, * In Iberia there are t many deer and wild horses.' Wild horses as well as wild bulls lived among the Alps, as we learn again from Strabo ; and Pliny tells us, not only in the Alps but in the north generally. Nor are the Middle Ages wanting in proofs of the existence of wild horses in Germany and the countries east of Germany. At the time of Venantius Fortunatus 1 London, 1885, p. 37. WILD TARPAN AND ITS RELATIONS 73 the onager — under which name may be understood the wild horse — was haunted in the Ardennes, as well as bears, stags, and wild boars. In Italy wild horses were seen for the first time during the rule of the Longobards, under King Agilulf. " In 732 Pope Gregory III. writes to St. Boni- face : ' Thou hast permitted to some the flesh of the wild horse, and to most that of the tame. Henceforward, holy brother, thou shalt in no wise allow it.' So, up to that time the Apostle of the Germans had been very liberal, perhaps because in his native island he had been accustomed from his youth to the habit which appeared so horrible to the Italian at Rome. Among the benedictions of Monk Ekkehard of St. Gallen (about a.d. iooo) to be pronounced over the meats to be served in the refectory of that monastery, one refers to the flesh of wild horses, which must therefore have been eaten by the pious brethren. An old German proverb says : ' A foal taken from a herd of wild horses will sooner be tamed than a depraved man learn to be ashamed.' In the Salksen-spiegel, where it treats of women's outfit and dowry, it is decreed that wild horses which have not always been guarded shall not be reckoned as part of such property. In a Westphalian document of 13 16, the fishing, game, and wild horses of a certain forest are apportioned to one Hermann. Not alone in the time of the Merovingians, but at the end of the 74 THE HORSE AND ITS RELATIVES sixteenth century, wild horses would seem to have lived in the Vosges Mountains, the wild borderland between two nationalities ; for Rosslin, in his account of Alsace and the Vosges (Strasburg, 1593), thus circumstantially describes them : ' Horses that be of their kind much wilder and shyer than the stag ; also much more difficult to take even in traps like the stag ; yet when they are tamed, which is accomplished with great toil and trouble, they make the very best horses, that equal those of Spain and Turkey, and surpass them in many things, and are hardier, for they are accustomed to cold and to coarse food, and are sure-footed, being as used to mountains and rocks as the chamois.' "If wild horses were thus found in the cultivated west and south of Germany, they must have existed still longer in the wild country on the Baltic, in Poland, and Russia. In fact, we find innumerable proofs of this down to modern times. At the time of Bishop Otto of Bamberg, in the first half of the twelfth century, Pomerania was rich in all kinds of game, including wild oxen and horses. At the same period wild horses are mentioned as extant in Silesia, whence Duke Sobeslaus in 1132 'carried away many captives, and herds of wild mares not a few.' It is known, and is confirmed by many literary allusions, that till the time of the Reformation, and even later, the woods of Prussia were inhabited by wild horses. Toppen's History of Masovia WILD TARPAN AND ITS RELATIONS 75 {Geschichte Masurens, Danzic, 1870) says : ' In the time of the Teutonic Knights, wild horses and other ofame were hunted for the sake of their skins. In 1543 Duke Albert sent an order to the com- mander at Lyck, bidding him take measures for the preservation of the wild horses.' Proofs of the horse being an object of chase in Poland and Lithuania are found far into the seventeenth century. As to Russia, it is sufficient to quote the remarkable words of Vladimar Monomach, Prince of Chernigoo, who lived from 1053 to 1125. He says of himself, in his posthumous exhortation to his sons (pre- served in the Lawrentkian Chronicle) : * But at Chernigoo I did this : I caught alive and bound with mine own hands from ten to twenty wild horses ; and as I rode along the river Ross (which formed a sort of boundary between the Russians and the wild Turkish Polovtsy), I caught similar horses with my own hands.' " Other writers, as quoted by Colonel Charles Hamilton Smith, ^ refer to the colour of these Central European wild horses. Erasmus Stella, for instance, writes of the wild horses of Prussia as being like the domesticated species, but with soft backs, unfit to be ridden, shy and difficult to capture, but very good venison. They were again referred to by Andrias Schneebergius, who ^ Jardine's Naturalises Library^ vol. xx. Horses^ 2nd ed. p. 158 ; the first edition was published in 1841. 76 THE HORSE AND ITS RELATIVES states that they were like the domesticated breeds, but mouse-coloured, with a dark streak along the spine, and dark mane and tail. They were not greatly alarmed at the sight of human beings ; but were extremely difficult to mount. Like other game, they were reserved for the table. The mention of dark manes and tails is very important, as it shows that these animals were not onagers, which appear to have ranged into Europe during the later part of the Tertiary period. Reverting to Messers Heyn and Stallybrass, it has to be mentioned that, after the account quoted above, they suggest that these horses were not really wild, but the descendants of horses escaped from captivity. "The fact," they write, "that in pre-civilised times Central Europe, as far as Spain, was covered with dense forests, makes the hypothesis that this region was one of the natural homes of the horse improbable, for this animal is a native of the steppes, needing wide grass-lands and space in which its speed could be of avail in escaping from the larger beasts of prey. The very way in which some of these facts are recorded seems to point to horses gone wild rather than to those originally wild. When the Vosges horses, though with difficulty, do get broken in ; when Duke Sobeslaus drives home herds of wild mares from Silesia ; when the fishing, the game, and the vagi equi of a Westphalian district WILD TARPAN AND ITS RELATIONS 77 are assigned to Hermann, and the untended horses of an estate are not to be included in a bride's outfit — in all these cases we may suppose that only fugitive horses are meant. So the animals found in Pomerania by St. Otto, and in Prussia by the Teutonic Knights, may have been in a wild state, and yet the progeny of merely fugitive mares ; and this becomes the more probable the longer those regions had been the scene of war and rapine." On the other hand, it should be borne in mind that during the Stone Age Western Europe, as already mentioned, was undoubtedly the home of a small big-headed race of horse ; and nothing is more likely than that herds of this or an allied race should have survived in certain districts till a much later epoch. As regards the argument that the whole of Central Europe was a forest-clad region during the Stone and Middle Ages, Dr. A. Nehring ^ has brought evidence to show that this is incorrect. After mentioning his belief that these small Prehistoric horses were the ancestors of the modern breeds of Western Europe, Dr. Nehring ^ proceeds as follows : — "The taming of the domesticated horse lasted, in my opinion, into the later Stone Age, and during that epoch there existed in Central and Western ^ Ueber Tundren und Steppen, derjetzt- und Vorzeit, Berlin, 1890. " Ibid.^ p. 189. 78 THE HORSE AND ITS RELATIVES Europe vast steppe-like tracts, which suppHed suitable nutriment to the wild horses, and afforded abundant space for their wanderings. . . . " Even during the Middle Ages wild horses may well have existed in Germany ; it is true that the objection has been made that the wild horse is a steppe-dwelling animal, and that in the Middle Ages there were no steppes in Germany. But this objection, in my opinion, is invalid. For admitting that the wild horse was originally a steppe-dwelling animal, it is by no means im- probable that during the post-glacial epoch, when the steppes were becoming constricted and the country overgrown with dense forest, small, or even large, herds of wild horses survived in many districts." These, it is added, may perfectly well have lived in the open tracts between the forest to a much later date without ever becoming forest animals ; although, as mentioned later, some may have become adapted to a forest life. Much the same view as to the nature of these horses was taken at an earlier date by Colonel Hamilton Smith, ^ who regarded them as the ancestors of the modern eel-backed duns — that is to say, duns with a dark spinal stripe. In the Middle Ages they had, however, probably become more or less crossed with escaped domestic horses, as will be shown to ^ Naturalist's Library^ vol. xx. Horses, p. 159. WILD TARPAN AND ITS RELATIONS 79 have been the case with the wild horses of Russia ; additional evidence in this direction being the above-mentioned statement as to their coats being mouse-colour, which appears to be an indication of cross-breeding. As regards the wild horses of the Volga-Ural steppes, commonly called tarpan, the best account is that given by the German traveller, Peter Simon Pallas,^ in the early part of the nineteenth century, who wrote as follows : — " The wild horses inhabit the steppes of Great Tatary and Mongolia, from the Dnieper to the Altai, and through the whole of Central Asia, in small herds, seldom fifty in number. Most of them are reddish grey or pale grey in colour, but the mane, the spinal stripe, and the tail are reddish brown, the muzzle is whitish, and the region of the mouth blackish. (There are, however, herds of different colours among them, which are due to wild horses having interbred with domesticated animals escaped from captivity.) They are inferior in stature to domesticated horses, and have larger heads, more slender limbs, and somewhat bigger ears, which are bent backwards at the tips in sickle- fashion. The forehead is swollen above the eyes, with a whorl of hair between the latter. The hoofs are small and almost cylindrical. The mane extends ^ Zoogeographia Rosso-Asiatica, St. Petersburg, vol. i. p. 260, 1811. 8o THE HORSE AND ITS RELATIVES from the space between the eyes [apparently a misprint for ears] to the shoulder-blades ; it is moderately long, and half upright. In winter the coat is rough, long, and waved on the back ; the tail is of moderate length. Young foals can be very easily tamed, but the adults are untamable. They gallop with wonderful speed, and scent human beings from a great distance, especially when they get their wind. . . . They frequent open undulating steppes, and avoid forests and mountainous locali- ties. If, as is almost certainly the case, the word eyes {augeit) is a misprint for ears (ohren) — for the author mentions the presence of a whorl of hair between the eyes — and if we except the absence of mention of the scanty hairing of the tail, the foregoing description might serve very well for that of the modern Mongolian tarpan. Other evidence was collected at an earlier date (1766) by S. G. Gmelin,^ who states that the Russian tarpan was a small, clumsy-headed, mouse- coloured horse, with a short wavy mane, and the fronts of the legs black from the knees and hocks downwards. In some examples the ears were short and horse-like, but in others longer and more ass-like. The tail was in some cases bushy, and in others scantily haired, but always shorter than in domesticated horses. This account appears to ^ Reise durch Russland, St. Petersburg (1770-84). WILD TARPAN AND ITS RELATIONS 8i refer largely, if not entirely, to hybrid tarpans. The Russian tarpan is now extinct, but one skeleton is preserved in the Zoological Museum at Moscow, and a second at St. Petersburg. One of the last survivors would appear to have been a gelding received in August 1884 at the Moscow Zoological Gardens, which had been taken as a foal in the government of Cherson in 1866. According to Dr. W. Salensky ^ this animal agreed in the matter of colour with Gmelin's description : but it had a forelock, and the mane fell over to the left side. There were no chestnuts on the hind legs. That this animal was a hybrid is practically certain. The great majority of naturalists have refused to admit the claims of the Russian tarpan in their earlier days to be regarded as truly wild animals. This view, however, Dr. Nehring,^ as already mentioned, considers to be erroneous ; and it seems most probable that even in Pallas's time there were some Russian studs of more or less nearly pure-bred tarpan, but that as time went on these became more and more mixed with escaped domes- ticated horses, so that to find pure-bred tarpan it became necessary to go further and further east- ward. A very important contribution to the history of 1 Wissenschaftliche Resultaie der von N. M. Przewahki nach Central- A sie7t untergennoinefien Reisen, Mammalia, pt. i., St. Peters- burg, 1902. * Ueber Tundren und Steppeii, pp. 92, 93. F 82 THE HORSE AND ITS RELATIVES the tarpan, although treated by naturalists with neglect, was furnished by Colonel Hamilton Smith.^ After referring to the doubts which had been enter- tained by naturalists with regard to the existence of truly wild horses, this author proceeds as follows : — " Whatever may be the lucubrations of naturalists in their cabinets, it does not appear that the Tatar or even the Cossack nations have any doubt upon the subject, for they assert that they can distin- guish a feral ^ breed from the wild by many tokens ; and naming the former takja and mtizin, denomi- nate the real wild horse tarpan and tarpani. We 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 Kalmuks, 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 and German, we drew the eeneral conclusion of their decided belief in a true wild and untameable species of horse, and in herds that were of mixed origin. Those most ac- quainted with a nomad life, and in particular an ^ Naturalises Library, Horses, p. i6o, 1841. ' "Feral" denotes animals originally escaped from captivity, as opposed to truly wild ones. WILD TARPAN AND ITS RELATIONS 83 orderly Cossack attached to a Tatar chief as Russian interpreter, furnished us with the substance of the following notice : — *' The tarpani form herds of several hundred, divided into smaller troops, each headed by a stal- lion ; 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 troop : 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. . . . "These animals are found in the greatest purity on the Karakoum (south of the lake of Aral) and the Syr Daria, near Kusnek, and on the banks 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 vicinity of the fixed settle- ments, distinguishable by the variety of their colour and a selection of residence less remote from human habitations. " Real tarpans are not larger than ordinary mules, their colour invariably tan, Isabella, or mouse, 84 THE HORSE AND ITS RELATIVES 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 in 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 malignant, 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 right up to the crupper ; the croup as high as the withers : the voice of the tarpan is loud, and shriller than that of a domestic horse ; and their action, standing, and general appearance resemble somewhat those of vicious 1 " mules. It is added that the genuine wild tarpans are migratory, wandering northward in summer, and returning south in autumn ; in this respect they differ markedly from the hybrid muzin. The above description, it is important to repeat, was drawn up after the Peace of Paris in 1814 by Colonel Smith from Cossack reports, and it is WILD TARPAN AND ITS RELATIONS 85 therefore only to be expected that in certain parti- culars, such as the alleged smallness of the head, the statement that the tail does not reach below the hocks, and the absence of definite mention of the white muzzle and the scanty hairing of the upper part of the tail, it should not accord precisely with the tarpan as now known to us. On the other hand, the reference to individual variation in colour and in the length of the ears is very noteworthy, and in accordance with the facts. Before proceeding further, reference may be made to the description and figure in Colonel Smith's book (p. 304, pi. xvii.) of an animal for which the author proposed the name of Asimts eqtmleus. This animal was kept some time previous to 184 1 in a livery stable in Park Lane, and was brought to the notice of Colonel Smith by Sir Joseph Banks, who had been informed by Lord Rivers of its existence, and that it indi- cated a new species brought from the Chinese frontier north-east of Calcutta. According to further information supplied at the stable, the animal was said to have come from some part of Chinese Tatary, that is to say, Mongolia. After mentioning that at first he had some doubt whether the animal might not be a variety of the chigetai or the kiang, its describer proceeds to state that, on examination, he was convinced it was much nearer to the horse, adding that he believed 86 THE HORSE AND ITS RELATIVES it to be identical with the breed or species known to the Chinese as "yo-to-tze." The animal, which was a male, was described as being not quite five years old, and standing 4 ft. (i2 hands) at the withers. In form it was distinctly "ewe-necked"; the mane, although longer than that of an ass, was upright ; the tail (which, from the picture, appears to have been cut) was scantily supplied with long hairs nearly to its root, resembling that of a rat-tailed horse ; and callosities were wanting on the hind-limbs. As regards colour. Colonel Smith writes that it was entirely of a yellowish red clay tint, "except- ing the black tips of the ears, the mane, and long hair on the tail, a well-defined line along the back extending down the middle of the tail, crossed by a broad bar of the same colour over the shoulders, three or four streaks very distinctly marked over the knees and hocks, the cannon-joints brown, and the fetlocks and pasterns down to the hoofs black, the hoofs and hide dark, the eyes brown." With the exception of certain remarks on an Arab-like appearance of the muzzle and nose, this description would apply fairly well to some of the impure tarpans characterised by having fawn- coloured instead of white muzzles. It is true that the absence of chestnuts on the hind-limbs appears to be a difference, but these are always small in the Mongolian tarpan, and were sometimes absent in WILD TARPAN AND ITS RELATIONS 87 the half-bred Russian animals. Bars frequently occur on the limbs of both types, and traces of a shoulder-stripe may be detected in some in- dividuals. If this animal was really a half-bred tarpan, it is important to notice that the name A sinus equuleus antedates the under-mentioned E quits przevalskii. In spite of Colonel Hamilton Smith's clear assertion that the true wild tarpan was a native of the borders of the Gobi Desert and the adjacent dis- tricts, naturalists persisted in applying that name to the Russian half-breeds, and most of them more or less completely ignored the evidence of the exist- ence of truly wild horses at the present day. There matters remained till the year 1881, when Mr. J. S. Poliakow ^ described the skin and skull of a reputed wild horse obtained a short time previously by the well-known Russian explorer Colonel N. M. Prze- walski, to whom it had been presented by an official at Zaisan, and in whose honour it was named Equus -przevalskii. Only a single specimen was obtained, and this was described as being intermediate in characters between the horse on the one hand and the kiang and onager on the other, having chestnuts on all four limbs as in the former, but only the lower half of the tail clothed ^ Proc. Imp. Russian Geographical Society, 1881, pp. 1-20 ; the paper is translated into English in the Ann. Mag. Nat. Hist., ser, 5, vol. viii. pp. 16, 26, 1 88 1. 88 THE HORSE AND ITS RELATIVES with long hair, as in the two latter. The general colour was described as dun, with a yellowish tinge on the back, becoming lighter towards the flanks and almost white on the belly, with no dark dorsal stripe. The short and upright mane, which was not continued forward as a forelock, was dark brown ; and the long coat was wavy on the head. The skull and hoofs were stated to be horse-like. In referring to this skin, Sir William Flower^ made no mention of Hamilton Smith's account of the Mongolian tarpan, but suggested that the speci- men might possibly prove to be an accidental hybrid between the horse and the kiang. Twenty years after the description of the type specimen of Equus przevalskii — that is to say, in 1907 — the Duke of Bedford received a number of tarpan colts from the Kobdo district of Western Mongolia, two of which were sent in the following year to the London Zoological Gardens, when one of them was figured by Dr. P. L. Sclater." About the same time living specimens and a large number of skins were received at St. Petersburg, which formed the basis of the monograph by Dr. Salensky published in 1902, of which the full title has been already quoted.^ Adult tarpan cannot apparently be captured, but, by taking certain precautions, the Kirghiz are able * The Horse, p. 79. * Proc. Zool. Soc, 1902, pi. xiii. ' Supra, p. 81. PLATE Vni FiGo I Fig. 2 Fig. I. A Mongolian Pony Alare, probably a half-bred Tarpan, used as a foster-mother to the Tarpan Colts brought to England for the Duke of Bedford. Fig. 2. A Tarpan :\Iare. WILD TARPAN AND ITS RELATIONS 89 to ride down and take the foals, which are snared in nooses. Those received in 1907 by the Duke of Bedford were brought from the Kobdo district of Western Mongolia by the agents of Mr. Carl Hagenbeck of Hamburg, who enlisted an army of Kirghiz for their capture. These foals were taken in three different areas in the neighbourhood of Kobdo ; those from each area showing certain colour- differences, into the consideration of which it will be unnecessary to enter in this place; and it will suffice to state that these differences suggest that there has been some admixture with domesticated breeds. The genuine wild tarpan may, however, be described as a big-headed pony, with a convex forehead, a short erect mane, and a tail covered with comparatively short hair on its basal portion, but terminating in a long tuft. Chestnuts and ergots are developed on all four limbs, as in most domesti- cated horses ; the limbs are moderately slender, and the front hoofs are not unduly broad. The general colour of the upper-parts is dun, but on the nose and under-parts it becomes more or less markedly whitish ; the mane (which does not extend on to the forehead to form a forelock), the tips of the ears, and the lower portions of the legs are black in front, and there is a distinct, although narrow dorsal stripe, while more or less defined shoulder- stripes and traces of barring on the upper parts of 90 THE HORSE AND ITS RELATIVES the limbs are frequently developed. Although short in summer, the coat becomes long and shaggy in winter, when the mane displays a slight tendency to fall to one side ; the hair on the fetlocks and lower jaw likewise showing a decided increase in length at the latter season. The dentition is characterised by the relative shortness of the interval between the outermost, or third, incisor and the first tooth of the cheek-series, and the absolutely and relatively large size of the cheek-teeth themselves, as shown in plate v. fig. i. This large size of the cheek-teeth is indicated by the circumstance that in a skull with a basal length of i8f the length of the row of six cheek- teeth is 7J inches, or only one-quarter of an inch less than that of the corresponding teeth in the skull of a shire mare, of which the basal length is 23 inches. Structurally the upper cheek-teeth are characterised by the absence of complex folding in the rings of enamel surrounding the central pits, and the relatively great length of the worn grinding surface of the anterior inner pillar, which is produced considerably in advance of the connection with the main body of the tooth, and is much flattened on the inner side ; this feature being more pronounced in the premolars than in the molars. In a Dartmoor pony, with a skull of about the same size as that of a tarpan, the length of the row of cheek-teeth was only 5I inches ; but this WILD TARPAN AND ITS RELATIONS 91 is not quite a fair comparison, as the pony was an older animal than the tarpan, and the length of the tooth-row shortens with age. It is noteworthy that there appears to be no greater tendency to develop the small first upper premolar, or wolf- tooth, in the tarpan than in ordinary domesti- cated horses. Relying on the local colour-differences referred to above, Dr. P. Matschie,^ of the Berlin Museum, has expressed the opinion that there are two kinds of tarpan, namely Equus przevalskii, with a dun- yellow colour, very dark mane, and black legs, which he considers to be restricted to the neighbourhood of Zagan-nor, a lake lying to the south-east of Kobdo. On the other hand, the tarpan of the Urungu district to the west of Kobdo, and the valley of the Ebi, which are lighter in colour, with no black on the front of the legs, and a lighter mane, have been named by him E. hagenbecki. There can, however, be little doubt that the difference between the two types is due to an admixture of blood ; and it is highly significant that the one which departs from the typical form occupies the western area or, in other words, is nearer to districts where there were formerly herds of half-bred tarpan. At the con- clusion of his memoir on the tarpan. Dr. Salensky ^ "Gibt es in Mittelasienmehrere Arten von Echten Wildpferden.?" Naturwissenschaftliche Wochenschrift, Berlin, vol. xviii. pp. 581-583, 1903. 92 THE HORSE AND ITS RELATIVES expressed the opinion that Equus przevalskii is certainly a distinct species, or race, although he declined to commit himself to any definite view as to its relationship to domesticated horses. In an appendix, dealing with a paper by Professor T. Noack, he admitted, however, that there is much to be said in favour of the view that a connectingf link between the tarpan and some of the domesti- cated breeds may have once existed, and that this link may have been formed by one of the small horses of the Stone Age. Before discussing the question as to whether the tarpan ought to be regarded as a species by itself or a race of the species typified by domesticated horses, it will be well to devote a few paragraphs to the Prehistoric horses of the Stone Age. Structurally the molar teeth of these Prehistoric horses are of the same type as those of domesti- cated horses ; but, nevertheless, in the early days of palseontological science a number of scientific names were given to these fossil horses on the evidence of isolated molar teeth and other specimens which are no longer available for comparison, and, if they were, would be quite insufficient for determining the particular type of horse to which they pertained. In 1832, for instance, a German palaeontologist, Hermann von Meyer,^ proposed the name Equus fossilis for a horse represented by remains from the ' PalcBologica, p. 79; Frankfurt-am-Maine, 1832. WILD TARPAN AND ITS RELATIONS 93 superficial (diluvial) formations of his own country ; and the title E. admniiicus had been given a dozen years earlier by Schlotheim ^ to the remains of the same or a closely allied type of horse. At a later date the English naturalist Sir Richard (then Professor) Owen,^ referred an upper molar of a horse from Kent's Hole Cavern, near Torquay, to von Meyer's E. fossilis ; stating that it differed from molars of domesticated horses by its narrower crown— a feature that may perhaps be due to its belonging to the deciduous, or milk, series. Other upper molars from the cavernous fissures in the Devonian limestone of Oreston, between Plymouth and Tavistock, were assigned by Sir Richard Owen^ to a second species, under the name of E. plicidens, in reference to the sup- posed more complex foldings, or pleatings, of the enamel in the central islands, or pits, of the grind- ing surface of the crown. Twenty-five years later the same naturalist* described a number of equine remains from the cavern of Bruniquel, in the department of Tarn-et- Garonne, France. These, in place of being isolated molars, comprised specimens of the complete denti- tion, as well as limb-bones ; and, from the relatively large size of the former as compared with the latter, ^ Petrefaktenkunde, p. 11 ; 1820. ^ British Fossil Matnmals and Birds, p. 383, London, 1846. ^ Op. cit., p. 392, and Rep. Brit. Assoc, for 1843, P- 281, 1844. * Owen, P/til. Trans. Roy. Soc. Londo?t, 1869, p. 544. 94 THE HORSE AND ITS RELATIVES Sir R. Owen estimated the shoulder-height of the Bruniquel horse — for which, ignoring the earHer names quoted above, he proposed the designation Equus spelcstis — ^at about 1 3 J- hands, or \\ feet. Now it can scarcely be doubted that this small Bruniquel Prehistoric horse was identical with the small big-headed horse drawn on horn by the Stone Age men of La Madelaine, in the department of Dordogne (pi. vii. fig. 2), and the name E. spelcsus will therefore be applicable to both. The bones and teeth of horses from the super- ficial formations of the Continent and Great Britain indicate, however, great differences in the bodily size of the animals to which they belonged ; and it has, therefore, been inferred that there were several distinct types of wild horses during the Stone Age. The evolution and differentiation of these types, it has been suggested, may have been due to the disappearance of the open tundras and steppes of Central Europe, and their replacement by forest, in consequence of which some of the wild horses took to a partially forest-life, which would lead to the development of a heavy and massive type of limb, while others, again, frequented the borders of deserts, where, it may be, they could exist only by the aid of man's cultivation of the soil in the oases. From the study of remains obtained from Anau, in Turkestan, Dr. J. U. Duerst^ considers that ^ In R. Pumpelly's Explorations in Turkestiut, vol. ii. p. 309; Washington, Carnegie Institute, 1908. PLATE IX Fig. I Fig. 2 Fin. I. Skull of a Young Tarpan Mare. Fig. 2. Skull of an Arab Alare, showing the sinuous profile ; //, preorbital depression. WILD TARPAN AND ITS RELATIONS 95 from the Prehistoric horse of the diluvial period, which he regards as a local race of the species typi- fied by domesticated horses, and therefore calls EquMS caballus fossilis, three other races were developed in the late Stone Age. These he desig- nates respectively as the desert-type i^E. caballus pumpellii), the steppe-type (^. c. germanicus or robustus), and the forest-type {E. c. nekringi). In a greater or less degree each of these, together with the rather earlier type identified by Dr. Duerst with E. c.fossilis, exhibits evidence of relationship with the Mongolian tarpan, which is regarded by the same writer as the direct descendant of the last- named. The desert-type, as represented by the Anau horse of the Prehistoric deposits of Turkestan, is regarded by Dr. Duerst^ as the direct descendant of E. c. fossilis. It was the smallest of all the Prehistoric domesticated horses, and may never have existed in a wild state. Having limbs of much more slender form than those of the tarpan, and very narrow hoofs, it was characterised by the medium width of the forehead as compared with the length of the base of the skull ; ^ some of these features affiliating it to the Arab type, of which it ^ Op. cit, p. 431. ^ Eastern and western horses are distinguished by a difference in the proportion of the width of the forehead to the basal length of the skull. As the proportion is low in the former, they are called " broad-fronted " ; whereas the western horses, in which it is high, are styled " narrow-fronted." 96 THE HORSE AND ITS RELATIVES is considered by Dr. Duerst ^ to have been the ancestor, although he also mentions near relation- ship to the tarpan. Bred, like the Arab, for speed and a desert-life, it is believed by the same writer to have been imported during the Bronze and Iron Ages into Europe, where it is well represented by the small horse of La Tene.^ Dr. Duerst thus regards all existing domesticated horses as derived from a single ancestral wild type, namely, the extinct forerunner of the tarpan ; such differences as characterise the various breeds being due to adaptive development. The subject is again referred to in the fifth chapter ; but it may be mentioned here that the evidence for the derivation of the Arab from the Anau horse does not appear conclusive, as no account is taken of the possibility of that animal having a mixture of Arab and tarpan blood. Very different from the last is the forest-type {E. c. ne/iringi), which was a small, stout horse, or pony, believed by Dr. Duerst ^ to have originated in the primeval forests of Germany, where it gradually became more and more stunted in size and thicker in the limbs, and where it was eventually domesticated. According, however, to Professor P. Matschie,* this small forest horse, or pony, is identical with a race from Wiirtemberg described * Loc. cit. ^ Ibid.^ P- 431- * Loc. cit. * " Allerlei aus der Geschichte der Einhufer," Monatshefte fiir Naturwiss. Unterricht^ vol. ii. p. 303, Leipsic, 1909. WILD TARPAN AND ITS RELATIONS 97 at an earlier date by Dr. Woldrich as E. c. fossilis latifi^ons, which, in the opinion of the former writer, may have been the ancestor of the European Bronze Age pony, identified by Dr. Duerst with his E. c. pwnpellii. The so-called Celtic pony, to which fuller reference is made in the sequel, is considered by Dr. Duerst to be a derivative from his forest type. The steppe-type of the same writer, as primarily represented by E. c. germanicus of Nehring, from Westeregeln, Thiede, and Quedlinburg, was a bigger animal than either of the preceding, with the proportionately narrow forehead characteristic of the heavy horses of Western Europe. How nearly this type was connected with the horse of Solutr^, in the Dordogne, north of Lyons, which was probably identical with the one depicted on the walls of the Madelaine Cave, in the same department (plate vii. fig. 2), and thus with the Bruniquel horse, is not clear, although it is probable that all these were of the same general character, and intimately allied to the Mongolian tarpan. Indeed Dr. Duerst states that some of the Solutre bones are absolutely indistinguishable from those of the latter. If this steppe-type be insepar- able from the Bruniquel, Madelaine, and Solutr^ horses, it will be obvious that the name E. c. spelcBiis^ as the earlier,^ should replace E. c. germanicus. Supra., p. 94. G 98 THE HORSE AND ITS RELATIVES Although the cave, or rock-shelter, of Solutrd could scarcely accommodate more than half-a-dozen families, however tightly packed, the entrance was protected by two walls of horse-bones, one a hundred and fifty feet long, ten high, and twelve thick, and the other forty feet long and five high. M. Toussaint, who explored this shelter of Prehis- toric man, roughly computed the number of animals whose bones were thus stacked as forty thousand. So many in one spot could hardly have been tame ; and, if they were, a large proportion would be old, but every one was quite young, many of them being foals, so that it is evident they had been killed in the chase, cut up, and brought home for eating. It would be natural to conclude, writes Mr. F. Boyle in the Cornhill Magazine for May 191 1, " that the hunters were horsemen. Boys would jump upon the back of a quarry wounded and overtaken ; the sport would teach them to ride, and presently they would take to catching foals. All the steps of the process follow logically. But perhaps the first did not occur to our remote fore- fathers. Asiatics never thought of riding till they were infinitely more advanced ; Gauls and Britons still clung to the chariot in Caesar's time. The lake-dwellers were horsemen certainly — we find their bits and accoutrements. And they used the same breed of horse which the men of Solutr^ ate, WILD TARPAN AND ITS RELATIONS 99 as the bones show. But that was a thousand years later, perhaps two or three or more." In this place it may be mentioned that much has been made of certain differences in shape and in the deeree of the hairiness of the head in the Prehistoric sketches of horses ; such differences being regarded as indicative of racial distinction. But Professor H. F. Osborn ^ has well remarked that it is quite probable these differences may be due to some of the animals having been depicted in the winter and others in the summer coat. A considerable amount of uncertainty and con- fusion exists, it will be noticed, in the foregoing determinations, especially in regard to the matter of scientific nomenclature ; but the confusion be- comes intensified when the views expressed above are contrasted with those held by Professor J. C. Ewart. According to one of the latest publications of that writer,^ at least three species or races of wild horses inhabited Western Europe in Pre- historic times. The first of these constitutes his steppe-type, which seems to be typified by the Mongolian tarpan, but is provisionally taken to in- clude the La Madelaine horse, for both of which the name E. przevalskii appears to be employed. This type, which is quite different from the one ^ Century Magazine, November 1904, p. 15. * " The Animal Remains at Newstead," in J. Curie's A Roman Frontier Post and its People, at Newstead, Melrose, p. 362, Glasgow 1911. loo THE HORSE AND ITS RELATIVES Dr. Duerst describes under the same name, is characterised by the large and heavy head and the relatively slender limbs, the face being long and narrow. The second of Mr. Ewart's groups is the so- called plateau-type, which appears to be typified by bones and teeth from French and English Pleistocene deposits, and is said to be a fine-headed, slender-limbed pony, standing from 12 to 13 hands at the shoulder, with short grinding surfaces to the anterior pillars of the upper cheek-teeth, and a forehead of medium proportionate breadth. For this plateau-type Professor Ewart, on page 2>^^ of the work just cited, adopts the name Equus agilis ; remarking in a foot-note that it includes a northern or "Celtic," and a southern or " Libyan," variety. The name E. agilis was proposed by him in 1910^ to replace E. gracilis, which he published in 1909,^ but subsequently found to be inadmissible on account of having been previously used in another sense. In the original publication of the last- mentioned name, it was stated that it was meant to replace the inadmissible Asinus fossilis of Owen, and it was likewise mentioned that it was intended to include, as varieties, the author's E. celticus and the E. libycus of Professor Ridgeway. Such nomen- clature is, however, totally inadmissible, the name ^ Ewart, Proc. Royal Soc. Edinburgh, vol. xxx. p. 299, 1910. ' Ewart, Proc. Royal Soc. London, vol. xxxi. p. 392, 1909. WILD TARPAN AND ITS RELATIONS loi Equus caballus celticus having been proposed by- Professor Ewart^ in 1903 for the so-called Celtic pony of the North of Ireland, the Hebrides, Faroes, and Iceland, and therefore antedating the name agilis. The third modification recognised by the same author " is the forest-type, which is said to be repre- sented by remains from the so-called "elephant- bed " at Kemp Town, near Brighton, and by the aforesaid horse from the Palaeolithic station at Solutre, northward of Lyons. This forest-type, which, it will be noticed, is different from the one so called by Dr. Duerst, is stated to have been a long low horse, probably characterised by a rela- tively broad and concave forehead, short, thick cannon-bones, wide hoofs, and long grinding surfaces to the anterior pillars of the upper cheek-teeth. The name Equus robustus (which Dr. Duerst regards as a synonym of germanicus) is adopted by Pro- fessor Ewart for his forest-type, as typified by the Solutr^ horses. The recognition by the writer last named of a so-called Siwalik type — that is to say, one related to the Pliocene Equus sivalensis of India — among the remains at Newstead scarcely demands serious notice. On the other hand, it is important to mention that Dr. Marcellin Boule^ has described ' Nature^ London, vol. Ixvii. p. 237, 1903. * "Animal Remains at Newstead," loc. cit., p. 363. ' Annales de Paliontologie, Paris, vol. v., 1910. 102 THE HORSE AND ITS RELATIVES from the Grottes de Grimaldi, in Monaco, the remains of a large form of horse, which he identifies with the existing Equus caballus, but to which, very judiciously, he does not assign a separate racial name. This Prehistoric horse approaches the modern Percheron breed, to which it may have been ancestral. Bones and teeth indicatinof horses of equally large size have been obtained from the Brighton "elephant-bed." From this long and somewhat wearisome survey of recent views in regard to the Prehistoric horses of Western Europe, which is essential in order to arrive at a satisfactory conclusion as to the syste- matic place of the Mongolian tarpan, it will be evident that during the period in question there were several more or less distinct types of wild European horses, differing from one another in bodily size, in the relative breadth of the skull, the degree of slenderness or stoutness of the cannon- bones, in the width of the hoofs, and to some degree perhaps in the conformation of the cheek-teeth. Some naturalists regard these different forms — or at all events a few of them — as distinct species ; but by Messrs. Duerst and Boule they appear to be all considered as races, or phases, of the species typified by the domesticated Equus cabalhis — a view in which I myself fully concur. These races are not, however, precisely comparable to the geographical races of existing mammals recognised by modern WILD TARPAN AND ITS RELATIONS 103 naturalists, no two of which ever occur in one and the same district. The races of Prehistoric horses, on the other hand, appear to have been dependent on environment, or "station," one being developed for a life on the open steppe, another in the forest, and another on grassy plateaus ; and their remains may accordingly be met with in one and the same deposit, or, at all events, in closely approximated localities. Furthermore, most, or all, of these Prehistoric types show more or less evident signs of near relationship to the Mongolian tarpan, while some of the existing Connemara ponies have been stated to bear the impress of descent from that animal,^ or rather, it should be said, from its Prehistoric representatives of the Madelaine and Bruniquel caverns. Before proceeding further it will, however, be advisable to refer to certain considerations in regard to domesticated horses. In the first place, attention should be directed to the fact that the name Equus caballus was given by the Swedish naturalist Linnaeus to domesticated horses in general, without mention of any particular breed to represent what naturalists call the type, that is to say, the typical form of that species. The same uncertainty obtains, however, with regard to certain species of European wild animals, ^ See R. I. Pocock, Harnisworth Natural History, vol. ii. p. 796. I04 THE HORSE AND ITS RELATIVES such as the red deer, the blue hare, and the fox, of which several local races are now known to exist ; but naturalists have agreed to solve the difficulty by taking the Scandinavian, or rather the Swedish, representatives of such species as the respective types. This being so, it is not only permissible, but likewise imperative, if consistency is to be main- tained, to follow the same course in the case of the domesticated horse. Scandinavian horses may therefore be regarded as the typical representatives of the J&qutis caballus of Linnaeus ; and since among these the "eel-backed dun" is a very common and characteristic breed in Norway, it may perhaps be permissible to take this as the actual type of the species. This course was, indeed, pro- posed some years ago by Professor Ewart, but subsequently abandoned on account of the circum- stance that dun horses may be produced by cross- ing two distinct Scottish breeds. This fact, in the professor's opinion, indicates that the dun is not a true breed ; but it may be pointed out that if this view is admitted the typical blue rock-pigeon is not a true breed, let alone a species, because, as was pointed out by Darwin in his Animals and Plants under Do^nestication^ several distinct breeds of pigeons will, when crossed, revert to that type. And as what holds good for pigeons will like- ' Vol. i. p. 64. PLATE X Fk;. I Fig. 2 Fig. I. A Norwegian Dun Stallion, showing dapple. Fig. 2. A Mongolian Polo Pony. WILD TARPAN AND ITS RELATIONS 105 wise obtain in the case of horses, it follows that the production of duns by crossing affords decisive evidence of the antiquity of that type, being, in fact, a case of reversion to the ancestral form. In colour the Norwegian so-called eel-dun is very like the tarpan, showing a narrow but distinct black dorsal stripe, and having the front surfaces of the limbs and the whole of the fetlocks black, while occasionally there may be traces of a shoulder- stripe and of barring on the upper part of the legs. In general form the breed is low in stature, but strongly built, with short, stout limbs. It should be added that, as mentioned later, two types of Nor- wegian duns are recognised ; one of these being shown in plate x. fig. i. On the other hand, the head and cheek-teeth are relatively smaller than in the tarpan, the front hoofs are broader in comparison to the hind pair, while the mane is comparatively long and pendent, with a forelock, and the tail is well haired up to the root. Such differences may, however, perfectly well be regarded as the results of domestication, due in part, it may be, to selection, or in part, perhaps, to crossing with a second wild type, or its descendants. This view in regard to the antiquity of the Norwegian dun type and its affinity to the wild tarpan accords exactly with the opinion of Darwin, who in the work already cited ^ wrote as follows : — ^ Page 63. io6 THE HORSE AND ITS RELATIVES ** With respect to the primitive colour of the horse having been dun, Colonel Hamilton Smith has collected a large body of evidence showing that this tint was common in the East as far back as the time of Alexander, and that the wild horses of Western Asia and Eastern Europe now are, or recently were, of various shades of dun. It seems that not very long ago a wild breed of dun-coloured horses with a spinal stripe was pre- served in the royal parks in Prussia. I hear from Hungary that the inhabitants of that country look at the duns with a spinal stripe as the aboriginal stock, and so it is in Norway." To this it may be added that dun horses, although not necessarily of pure blood, were formerly common in Spain, where, as in some other parts of Europe, they were considered to be the worst type of all ; the eel-backed being, however, a little better than the self-coloured dun.^ Perhaps this is the reason why in the sixth chapter of Revela- tion the " pale horse," Ttttto? -xXoopo^, that is to say, the horse of the colour of withered grass, or dun, is assigned to Death. Be this as it may, the general tendency of the foregoing evidence is to show that the eel-dun horses of Norway and other parts of Europe not only represent a very ancient type, but that they * See Ridgeway, Origin and Influence of the Thoroughbred Horse, pp. 260, 348. WILD TARPAN AND ITS RELATIONS 107 inherit their colour, either directly or by reversion, from the wild tarpan. This led Darwin ^ to conclude that all the existing breeds of horses are probably descended " from a single dun-coloured, more or less primitive stock, to which our horses occasion- ally revert." And although in the light of the foregoing evidence as to the existence of more than one type of Prehistoric wild horse in Europe, this conclusion requires some little modification, it is probably not far from the truth, though the Arab may perhaps form an exception to the generalisation. It being admitted, then, that the wild Mongolian tarpan is related not only to some of the existing horses and ponies of Western Europe, but likewise to their Prehistoric ancestors, it seems only logical that it should not be separated from the species typified by domesticated horses, and its name will therefore be Equus cabalhts przevalskii, or, at all events, until it is definitely proved to be entitled to a desio-nation of earlier date. Adult stallions of the Mongolian tarpan stand about 1 3. 1 hands (53 inches) at the shoulder; and, as might naturally be expected, its nearest domesti- cated representatives are the ponies of the same district, which measure from about 12.2 to 13.3 hands, and in their rough, untrimmed coats are very like their wild relatives, although they have developed long, flowing manes, with forelocks, and ^ op. ciL, p. 65. io8 THE HORSE AND ITS RELATIVES tails which sweep the ground and are thickly haired to the root. These ponies are kept by the Buriats and other Mongol tribes in millions, and are extremely hardy and enduring. Mr. C. W. Camp- bell ^ states that " a good specimen of the Mongol pony is perhaps the best of his size in the world for oreneral use. The head and shoulders will be too heavy for elegance, the eyes none too full, the muzzle and crest coarse, and the manners too often objectionable, but the quarters, loins, and legs are good, the barrel is deep and long, and there is no deficiency in bone. . . . The size and character vary with the locality. The commonest colour is grey, chestnut follows, and then come bay and sorrel. Stallions are selected animals, especially in North Mongolia, but the mares are not, and no special pains are taken anywhere to improve a breed. Along the China border the ponies are undersized, 12 to 13 hands, the result of the excessive demands of the China markets for all the larger beasts. As one travels northward, and the China market becomes more remote, the horse-flesh improves (12 to 14 hands), and the best specimens of the Mongol pony are found in the valley of the Kerulon." Mongolian ponies (pi. x. fig. 2) are brought down to China in large droves for racing and polo purposes; ' Report to Parliament on a Journey to Mongolia, London, 1904, p. 35. WILD TARPAN AND ITS RELATIONS 109 the majority going direct to Shanghai, although some are sold en route at Pekin and Tientsin. Accord- ing to a writer in the Field newspaper of April 8th, 191 1, "they are sold principally to the large race- owners for comparatively big prices, and are kept and trained for the races. As these ponies have to be bought without any trial entirely on their looks, with shaggy coats and totally unacquainted with the proper use of a brush, many are soon found to be hardly fast enough for racing purposes. Sometimes, however, the weirdest-looking proves himself to be very fast, while some of the good- looking ones turn out badly." The aforesaid variability in the colour of Mon- golian ponies, and the comparative rarity of dun, seem to indicate that they are not pure derivatives from the tarpan ; the same thing being, perhaps, indicated by the luxuriant hair of the mane and tail. According to Professor Ridge way ^ such colour- differences are known to have been in existence so early as the second century b.c. The mare represented in plate viii. fig. i is one of the dark types of Mongolian ponies. Although most of the so-called Chinese ponies are really Mongolian, the southern provinces of China do produce a native breed of pony, which appears to be nearly allied to those of Anam, Siam, and Burma (mentioned below), but is very small, ^ The Thoroughbred Horse, p. 132. no THE HORSE AND ITS RELATIVES scarcely reaching an average height of 1 2. i. These ponies are, however, very strong and hardy. It may be added that the nomad Mongols devote great attention to breeding ponies, of which they possess an immense number, although many of them are spoilt by having their hoofs and teeth abnormally worn down by the stony nature of the ground and the hard herbage on which they feed. Ponies more or less nearly related to the Mon- golian are to be found throughout the vast tracts of Central Asia lying between Siberia and the Himalaya, since, in the opinion of Captain Hayes, ^ the ponies of Bhutan, Nepal, Spiti, and Yarkand are of the same general type ; Yarkandis being not infrequently dun. The ponies of Corea, as already mentioned, are closely allied to those of Mongolia, and come equally close in general characters to the wild tarpan. Here it may be well to mention that there is no near resemblance between Mongolian ponies and the eel-backed dun horses of Norway, such as might be supposed to occur in breeds derived from the same ancestral stock. A moment's reflec- tion will, however, show that in this particular case no such resemblance is to be expected ; for if the wild horse was domesticated in regions so far apart from one another as Mongolia and Norway, it * The Points of the Horse, London, 3rcl ed., p. 599. WILD TARPAN AND ITS RELATIONS iii is only reasonable to presume that differences in climatic conditions and in the mode of treatment and selection, coupled in all probability by an admixture of different kinds of alien blood, would produce a marked difference between the eastern and western domesticated stock. Reverting to the Far East, it is important to observe that, as was long ago pointed out by Darwin, in most of the countries lying to the eastward of the Bay of Bengal, including Burma, Anam, Siam, the Malay Peninsula and Islands, the Liu Kiu Islands, and a large portion of China, the horse is represented only by small breeds which come under the designation of ponies. Among these, the Burmese or Shan ponies, which are mainly, if not exclusively, bred by the hill-tribes of the Shan States, in the interior of the country, are believed to be nearly related to the Mongolian breed, although probably modified by the infusion of foreign blood. In stature they are about the equals of the Mongolian, and are strong and active, although somewhat slow in their movements. On the other hand, the still smaller but closely allied Manipur ponies are much faster, and are used by their owners for polo, of which game Manipur is one of the original homes. Near akin to the Manipuris are the Batak or Deli ponies of Sumatra, which are bred in the Batak hills of that island, and are exported to 112 THE HORSE AND ITS RELATIVES Singapore from the port of Deli in large numbers. With their handsome, high-bred-looking heads, and high-crested necks, they differ, however, markedly from the Mongolian and Yarkandi types, which are often more or less decidedly ewe-necked ; this difference being due to a strong infusion of Arab blood. In stature they average only about 1 1.3 hands, although some reach 12. i or 12.2. Although most are brown, skewbalds are by no means uncommon. Sumatra also possesses a second breed of ponies, which take their name from the Gayoe hills, at the northern end of the island. According to Captain Hayes, ^ they are stouter in build than the Batak ponies, with shorter and thicker legs and heavier hind-quarters. They lack, however, the speed and fiery nature of the latter ; this being probably due to their having a smaller strain of Arab blood in their veins. On the other hand, the ponies of Java and some of the neighbouring islands, which, like those of Sumatra, not infrequently show distinct striping, appear to be, not only of very modern origin, but mainly of Arab descent, although it is quite probable that they may have some Mongolian blood. I am credited ^ with the statement that some Javanese and Sulu ponies show a large first upper premolar, or wolf-tooth, and as the same feature characterises the extinct Indian Equus sivalensis, the suggestion ^ op. cit., p. 633. ' Ridgeway, op. cit., p. 142. WILD TARPAN AND ITS RELATIONS 113 was thrown out that the former might be the direct descendants of the latter. Although I no longer maintain such a view, it may contain a certain element of truth, since Arabs, as mentioned later, may perhaps trace their origin to the aforesaid E(pms sivalensis. Before leaving this part of the subject it is impor- tant to observe that the Burmese and Malay countries have derived their ponies from Mongolian or Arab stocks, and had no indigenous breeds of their own. The tarpan has also had a share in the production of the Turkoman horses of Turkestan, which un- doubtedly have been produced by crossing Mongolian ponies with Arabs. In fact, the Turkoman horse passes insensibly through the Persian into the Arab. A very similar pedigree may be assigned to the well-known dun-coloured ponies of the Kathiawar district of North-western India, which frequently show transverse dark barrings on the legs, accompanied in some cases by traces of shoulder-stripes, and always by a narrow dark spinal stripe. The limbs are long and slender, and the ears large, with a decided tendency to turn inwards at the tips. In the opinion of Professor Ridge way,^ " There can be no doubt that the Kathiawar horse is a cross between the dun-coloured horse of Upper Asia and the Arab ; " there being historical evi- dence to show that so early as the commencement ^ Op. cit, p. 159. H 114 THE HORSE AND ITS RELATIVES of the Christian era large numbers of the dun horses of Northern Asia and Europe had been imported into the districts on the east side of the lower part of the Indus valley. The striping in the Kathiawar ponies, which is most marked in the best examples of the breed, appears to be a reversion to the ancestral type, as the result of crossing. The foregoing does not, however, by any means exhaust the extent of the influence of the tarpan on the domesticated horses of Eastern Central Asia, for Tibet is the home of a breed of ponies many of which are cream-fawn, or yellow dun, in colour. Many of these dun ponies, according to Mr. L. A. Waddell,^ are brindled, and in one particular indi- vidual a dorsal stripe, the tips of the ears, and stripes on the shoulders, flanks, and limbs were black, while there were dapplings on the haunches, as in many of the Mongolian ponies. More remark- able still are the so-called tanghans of Tibet, which are of larger size, and stated to derive their name from the Tanghastan district of Bhutan. They may be either piebald, or skewbald, with or without stripes, and, according to Colonel Hamilton Smith, the chestnuts on the hind-legs are extremely small. Some of the older travellers state that droves of tanghans were to be found in a wild condition on the Tibetan side of the Himalaya ; but, in one instance, at any rate, there appears to be a confusion ' Among the Himalayas, p. 248. WILD TARPAN AND ITS RELATIONS 115 between horses and kiangs, and in any case, if tanghans were found wild, they must have escaped from captivity. As regards their origin, Professor Ridgeway ^ states that "on the whole the balance of probability is in favour of the piebald colour of the tanghans ^ of Tibet being due to the crossing of the Mongolian and Arab stocks, as seems certainly the case with the piebalds of Sumatra." The ponies of the Mongolian type which formed the ancestral stock of the modern Kathiawaris were probably brought into Western India by the ancient Scythians from the neighbourhood of the Caspian ; and as these warriors also invaded Baluchistan and Afo^hanistan — the ancient Bactria — there is little doubt that the horses of these countries have a strong Mongolian element in their blood, although some of this may have been derived from the dis- tricts lying immediately to the north. On this point Captain Hayes ^ remarks that the Cabuli, Baluchi, and other trans-Indus horses so largely used in India, which, although stouter and shorter in the legs, are less smart in appearance and less suited to a hot climate than the so-called " country-breds," may be considered as intermediate between the latter and Mongolian ponies ; this being, in fact, equi- valent to saying that they are of mixed Arab and ^ Origin of the Thoroughbred Horse, p. 156. * I have ventured to alter Professor Ridgeway's spelling, " tangums," to accord with that adopted here. * Points of the Horse, 3rd ed. p. 603. ii6 THE HORSE AND ITS RELATIVES Mongolian blood. Stripes, it is said, are not infre- quent on the legs of horses in the Waziri districts of Afghanistan, as they certainly are on those of many of the ponies in the Punjab. When the origin of Punjabi and other Indian ponies is considered, we are, however, at once con- fronted by great difficulties. For during the Pleis- tocene period the Narbada district of Central India, and doubtless other parts of the country, were the home of a wild horse {Equus nainadicus) with long grinding surfaces to the anterior pillars of the upper cheek-teeth. And this horse — if indeed it is entitled to that name in the more restricted sense — may quite probably have had a share in the origin of the Indian country-bred and pony stock previous to the introduction of Arab and Mongolian blood. On the other hand, there is no proof that the extinct Narbada horse was not nearer to the onager than to the typical horse. In concluding this chapter reference may be made to the diet of Asiatic horses. In the work quoted on page 69 Colonel Phillott observes that : — " Indian country-breds will eat and thrive on food that would probably kill English horses. In the Persian Gulf and elsewhere locusts, fish, and dates are regarded as legitimate food for horses and cattle ; in Tibet the tanghans are given pig's blood and raw liver ; and in the cold regions of Central Asia meat is regarded as a necessity for horses." CHAPTER III HORSES AND PONIES OF THE BRITISH ISLANDS When Julius Cssar invaded Britain in the year 55 B.C. he found the natives in possession of swift and hardy horses, which they drove in their war- chariots with remarkable skill and adroitness. Although it has been stated that the horse is not indigenous to the British Islands,^ and the sugges- tion made that the original stock was introduced by the Phoenicians when they visited Cornwall for the purpose of obtaining tin, there seems no reason why the horses of the early Britains should not have been derived from the native Prehistoric breeds. The available evidence points to the con- clusion that these early British horses were of small size, so that at the present day they would come under the denomination of ponies ; this being another fact in favour of their descent from the small Prehistoric horse allied to the tarpan. In the opinion of Sir Walter Gilbey,^ it is doubtful whether the horses of Britain gained materially in size till the Saxons and Danes imported stallions belonging to larger breeds from the Continent. ^ Sir Walter Gilbey, Thoroughbred and other Ponies, London, 1903, p. 21. 2 Op.cit.,^. 22. ii8 THE HORSE AND ITS RELATIVES At the time when Domesday Book was written large droves of mares wandered at will through the forests of the great land-owners of England, and were only driven into enclosures occasionally when some of their number were selected for work- ing purposes, and doubtless also for breeding. And it is probable that from these forest mares (the EqucB silvest7'es or EqucB indoniit(2 of Domesday Book) were produced the first improved types of British horses. From the unimproved forest breeds are doubtless descended the modern forest and moorland breeds of ponies ; which, it has been suggested, have somewhat degenerated in size and quality owing to the poor fodder of the comparatively restricted areas on which they now survive. The first of these ponies for notice are those of the New Forest, in Hampshire (pi. xi. fig. i), which it has been suggested are descended from a stock found before the time of Knut (loi 7-1035) in the district formerly known as Ytene and afforested in the year 1072 by William the Conqueror.^ They are described by Low as ugly, large-headed, and short- necked, but hardy, sure-footed, and capable of bearing rough usage. In 1765, the breed was much improved by " Marske," the sire of " Eclipse," having been allowed to run with the herds for about four years. In 1889 the Forest ponies were again improved by thoroughbred blood ; and about the ' Gilbey, op. cit., p. 32. platp: XI Fig. I mmMm Fk;. 2 Fig. I. New Forest Ponies. Fig. 2. Shetland Ponies. HORSES OF THE BRITISH ISLANDS 119 same time stallions from the island of Rum, off the west coast of Scotland, were introduced. These black Galloways, as they are called, greatly improved the stock, which is now in demand as polo-ponies. Near akin to the New Forest breed are the ponies of Exmoor and Dartmoor, the former of which should average 12 and not exceed 13 hands in height, while stallions of the latter may run to 14 hands. In winter these ponies, which are left nearly wild until caught for use, are thickly covered with long hair. Exmoors are generally dark bay or brown in colour, with black points ; they have broad foreheads, sharp ears, well-formed shoulders, and short, sturdy legs. From the larger types of these ponies were bred the old pack-horses of the west of England, which were indispensable in former days to the farmers of the district, and were also largely used for riding. Of late years Exmoor ponies have been crossed with Dongola Arabs, and this, and perhaps earlier crossings, may account for the Arab-like character they now frequently possess. Welsh ponies, which are not confined to the principality, but range over the wilder parts of the adjacent counties of Salop, Hereford, and Mon- mouth, are more numerous than any other breed, and are a very ancient type. Early in the eighteenth century a famous thoroughbred stallion was turned out among the Welsh droves, so that in this case I20 THE HORSE AND ITS RELATIVES also the modern breed has a tinge of Arab blood. According to the modern standard, North Welsh ponies should not exceed i2| hands, but those of South Wales are allowed to measure 13 hands. The ponies of the Lake District (Cumberland and Westmorland) run larger, so that many of them are entitled to be called Galloways. According to Sir Walter Gilbey they possess no features en- titling them to be regarded as a distinct breed, and they do not therefore demand further notice in this volume. Of much orreater interest are the Connemara ponies of the west of Ireland, which inhabit the mountains of the Connemara district of Galway. It has been very generally asserted that these ponies were derived from horses saved from the wreck of the Spanish Armada in 1588. This, however. Sir Walter Gilbey ^ considers to be probably erroneous, and, in his opinion, the characteristics of the Conne- mara pony of the present day are due to the impor- tation of Spanish, i.e. Barb, horses from England during the period extending from the fourteenth to the seventeenth century. Low, in his Domesti- cated Animals of the British Islands, states that Connemara ponies "are from 12 to 14 hands high, generally of the prevailing colour of the Andalucian horses, delicate in their limbs, and possessed of the form of head characteristic of the Spanish 1 Op. cit., p. 86. HORSES OF THE BRITISH ISLANDS 121 race. . . . They are hardy, active, surefooted in a remarkable degree, and retain the peculiar amble of the Spanish jenet." The existence of a large amount of Barb blood in Connemara ponies is admitted by Professor Ridgeway,^ who gives reasons for believing that the ponies of Iceland, the Hebrides, and the Faroes were derived from Ireland, subsequently to the infusion of Barb blood into the latter country. Now, as mentioned in the last chapter, the ponies of Iceland, Finland, the Faroes, Shetland, Hebrides, Connemara, Wales, and parts of England are regarded by Professor Ewart as indicating a special type for which the name Equus caballus celticus was proposed in 1903.^ These ponies, of which the Iceland breed appears to be regarded as the typical representative, are collectively characterised, when pure bred, by the following features: — The height is about 12 hands, and the general colour very similar to that of the Mongolian tarpan. The head is small and delicately formed, the legs and hoofs are fine, and the hind chestnuts are lackino-, as are also the ergots on all four fetlocks. The coat in winter is long and thick, the mane and forelock are properly developed, and the tail is luxuriant, with short hairs at the side of the base, which form a pad to protect the inside of the buttocks, and are ' The Thoroughbred Horse, pp. 419, 420. ^ Vide supra, p. 10 1. 122 THE HORSE AND ITS RELATIVES shed annually. The mane is darker externally than internally, and the tail is not wholly black. Some Shetland ponies conform to this type, but others are more sturdily built, with little or no tail-pad, and the ergots and hind chestnuts de- veloped ; and much the same may be said of the Connemara ponies. As mentioned in the preceding chapter. Pro- fessor Ewart now regards the Celtic pony and the Barb (inclusive of the Arab) as divergent branches of a single primitive European stock. To the Celtic type Dr. L. Stejneger* refers the fjord-hest of Western Norway, which he regards as distinct from the doele-hest, or "valley-horse," of the interior," and as having been probably intro- duced into the country from Scotland. In his opinion the Celtic pony of Connemara and the Scottish Islands, the West Norwegian fjord-hest, and the now extinct Russian tarpan, all belong to the same stock. And he cites the evidence of a Russian naturalist, Professor Tscherski, who states that the hind chestnuts were frequently absent in the Russian tarpan ; he himself adding that the latter resembled the Celtic pony in size and colour, and what is equally to the point, that its skull 1 Naturen, Bergen, 1904, p. 161, and Smithsonian Miscell. Collec- tions, vol. xlviii. p. 467, 1907. ' It is uncertain to which type belongs the horse shown in pi. x. fig. I. HORSES OF THE BRITISH ISLANDS 123 agrees essentially in relative proportions with that of the Iceland pony. Assuming Dr. Stejneger's opinions to be trust- worthy, it follows, in view of the near affinity of the Russian tarpan to its truly wild Mongolian namesake to which reference has been made in the preceding chapter, that if the Celtic pony and the Barb are divergent branches of a common ancestral stock, the Arab is first cousin to the Mongolian tarpan — a relationship which few will be disposed to admit. But there is another way of looking at the matter, namely, that if we accept the evidence as to the infusion of Barb blood into the Connemara ponies, and also that the latter formed the source of the ponies of Iceland, the Hebrides, &c., the Celtic pony may apparently have derived its Arab characteristics from the same orig-inal infusion into an ancestral stock akin to the Prehistoric horses of La Madelaine and to the modern Mongolian tarpan. The case has been put very concisely by Dr. R. F. Sharffi in a paper on horse-skulls from Ireland, in which it is remarked that " the principal point of difference seems to me whether the Arab or Libyan features, as Professor Ridgeway would call them, in the Irish \_z.e. Celtic] horse are the result of introductions by mankind of Eastern ^ Froc. R. Irish Academy^ Dublin, vol. xxvii. ser. B, p. 85, 1909. 124 THE HORSE AND ITS RELATIVES or Spanish blood, or whether these features were inherited from a wild ancestor. I believe that the latter was the case." Others may, however, be permitted to hold the opposite opinion ; and it is significant that Professor Ridgeway ^ has experienced the difficulty of accepting Professor Ewart's view, and attempted to get over it by suggesting a dual origin for the Celtic pony. That the latter is a recognisable type may be accepted independently of the views taken as to its ancestry and relationship. In connection with the latter point, the reader may be reminded that Dr. Duerst, as stated on page 97, derives the Celtic pony from his "forest-type" {^Equus caballus nehringi)^ in which there appear to be no indications of Arab affinity. Reference may also be again made to the assertion (p. 103) that some Connemara ponies are very like the Mongolian tarpan ; and it may be added that the short hairs on the sides of the base of the tail in the Celtic type may be another indication of relationship to the true tarpan, in which the whole of this region is short-haired. Leaving theoretical matters, and resuming the consideration of the leading breeds of the British Isles, other than thoroughbreds, attention may be directed to the special features of the Shetland pony (pi. xi. fig. 2), which is the smallest of all. ^ The Thoroughbred Horse, p. 421. HORSES OF THE BRITISH ISLANDS 125 How and when these ponies, or rather the ancestral stock from which they are derived, reached the Shetlands is unknown, although some writers have suggested a Scandinavian and others a Scottish origin. From the circumstance that the " Bressay Stone," discovered at Bressay in 1864, includes among other designs the figure of a man on horse- back, it has been inferred that ponies were found in Shetland previous to the extermination of Celtic Christianity by the Norwegian invasion of 872,^ but the value of this evidence seems doubtful. The general characteristics of " Shelties " have been already indicated when discussing the so-called Celtic pony. The average height is about 10 hands, io| hands being the maximum show-standard ; but many do not exceed 9 hands. As regards colour, bay, brown, and dull black are the most prevalent shades, but these may be mingled with white, and in rare instances the whole coat may be white. Although in winter the coat is long, close, and shaggy, in summer the hair is quite short and sleek. It is stated that Shetland mares frequently have tusks as long- as those of the stallions. As already mentioned, some Shelties are cart-horse-like in make, while others are of a more slender and Arab-like type. The frequent presence of black in the colouring is considered to be indicative of Norwegian, and thus of Barb blood. ^ Gilbey, op. cil., p. 103. 126 THE HORSE AND ITS RELATIVES The ponies of Orkney are stated to be of a more mixed type, as well as larger in size and coarser in shape than those of Shetland. The Hebrides are also the home of numerous ponies ; those of the Outer Hebrides being small, round- shouldered, and muscular, with thick and rough winter coats, while those of the Inner Hebrides are usually larger ; Mull, Barra, I slay, Tiree, Skye, and Uist being the islands most noted for the good qualities of their ponies. Mr. Munro Mackenzie^ states that the small ponies of Barra and the outer islands stand from i2| to 13I hands, and, although having rather large and heavy heads and straight shoulders, are hardy, serviceable animals. The ponies of Mull, Tiree, Skye, and Uist, as well as some parts of the west coast of the Scottish main- land, are a larger type, running from 13I to 14I hands in height ; but are now very scarce. They are mostly blackish brown in colour, but some are brown, bay, or dun, others cream-colour, and a few grey. In this case tradition tells of infusion of Barb blood from horses saved from the wreck of the Spanish Armada ; while other reports refer to the introduction of discarded Arab chargers by military officers. Certain it is that Arab characters are prevalent among them ; and curiously enough, in the duns as well as in those of other colours. Whatever may have been the source of this Arab ^ Polo Pony Society's Stud Book, vol. vii. HORSES OF THE BRITISH ISLANDS 127 or Barb strain, all breeders and experts agree in attributing it to an introduced stock, and not to a primitive " Celtic " type. A larger type met with in the Highlands of Scotland is known as the garron, and is specially characteristic of Perthshire and the central Highlands. Horses of this breed may stand as much as 15 hands at the shoulder; in colour they range from black and brown to dun and grey, bay being rare. In the opinion of Mr. Mackenzie they are probably the offspring of ponies crossed with larger horses brought from the south during military expeditions. Intermediate between ponies and horses are the Galloways, so called from the district of Galloway, in the south of Scotland. Up to the end of the eighteenth century Galloways were generally under 14 hands, and were used alike for the transport of agricultural produce and for riding ; but after that date they were crossed with larger horses till they practically disappeared from the mainland, to survive only in remote islands like those of Mull and Rum. The general colour was bright bay or brown with black points ; but the Galloways of Rum, all of which were purchased in 1888 by Lord Arthur Cecil for the improvement of the New Forest breed, were black with hazel eyes. In all Galloways the head is small. The Welsh cob appears to be an allied breed. 128 THE HORSE AND ITS RELATIVES Since the present work purports to be an account of the natural history of the horse and not a treatise on horse-breeding, such groups as polo- ponies, hackneys (from the French haquenee), and hunters may be passed over without notice, since they constitute groups formed by selection from other breeds, rather than distinct breeds by themselves. Attention may accordingly be directed to the leading British types of horse employed for carriage and heavy draught. Among these, the first place may be assigned to the Cleveland bay, a magnificent stamp of powerful carriage-horses taking their name from the fertile district of Cleveland in the North Riding of Yorkshire, on the Tees, but now also largely bred in the East Riding, as well as in Durham and Northumberland. The name is, how- ever, a modern one, the original local breed having been known as the chapman or pack-horse. How it originated is not definitely ascertained, although it was not improbably produced by the infusion of foreign blood into the native stock of the district. The colour of the Cleveland is bay with black points ; and the height ranges from i6| to i6f hands. Heaviness of "bone" is one of the char- acteristics of the breed. Near akin to the Cleveland bay is the York- shire coach-horse, which tends to be smaller in size, with what breeders term more quality ; the latter being due to thoroughbred blood. PLATE XII Fig. I Fu; Fk;. I. A Suftblk Stallion. Fig. 2. A Shire Stallion. HORSES OF THE BRITISH ISLANDS 129 Of late years the tendency has been to render the Cleveland bay and the Yorkshire coach-horse lighter in make than formerly ; and in consequence of this it is difficult to find a sufficient number of horses of the type required for use in the royal stables of this country, so that the stud of carriage- horses has to be recruited from foreign sources. In this place it may be mentioned that the famous cream-coloured horses kept in the royal stables for use in processions of full state are a Hanoverian breed. Although frequently referred to as the ** cream ponies," they are in reality horses of large size and great muscular power, the biggest standing fully 16 hands, and the smaller ones an inch or two less. The black Drenthe horses em- ployed at royal funerals are another Hanoverian breed, mainly reared near Osnabruck. Of the heavy draught horses of Great Britain one of the most famous is the Suffolk (pi. xii. fig. i), frequently known, in allusion to its compact and "punchy" build, as the Suffolk punch, whose range extends from its native county into Norfolk and Essex. The original breed was noted for its hardi- ness and the capacity for exerting its utmost strength at a dead pull. A true Suffolk punch, it is written, would draw almost till he dropped ; and a team at a given signal would, without a whip, bend in a moment to their knees, and drag everything along. When Low wrote his Domesticated Animals of 130 THE HORSE AND ITS RELATIVES the Bi'itish Islands the colour of the Suffolk was light dun or sorrel, sometimes deepening into chest- nut, with a lighter mane and tail. The general shape was plain, with the head large, the neck short and arched, the shoulders low and heavy, the back straight, the haunches well developed, the loins wide, and the limbs short. Low observed that " the colour distinctive of this variety connects it with the race widely diffused throughout the North of Europe and Asia, from the Scandinavian Alps to the plains of Tatary, in which the dun colour prevails. It is believed to have been carried to the eastern counties of England from Normandy, which yet possesses many fine horses of this breed, introduced, it may be believed, by the Scandinavian invaders." Although there is no definite proof of such a Scandinavian origin, it is quite probable that it may contain an element of truth. In the latter part of the eighteenth and during the nineteenth century the Suffolk horse was modified by crossing — notably with the Lincolnshire trot- ting horse ; and nowadays the colour is generally either light or dark chestnut. From i6 to 162^ hands is the more general height, although some horses reach 17 hands. As the Suffolk is essenti- ally a farm breed, and not intended for heavy work in cities, the weight should be less than in Clydes- dales and shires. HORSES OF THE BRITISH ISLANDS 131 North of the Tweed the most famous breed of heavy cart-horse is the Clydesdale, so called from the district watered by the Clyde in its course through the county of Lanark. The breed appears to be of somewhat mixed origin, Scotch drovers who took cattle to England in the early part of the eighteenth century returning with horses which were used for the improvement of the native stock. What these horses were is, of course, unknown ; but it is certain that about the year 17 15 a farmer introduced into Clydesdale a black Flemish stallion from EnPfland which formed the foundation of the modern breed. The mares descended from this stallion were generally brown or black, with the face white, some white on the legs, and a white patch on the belly ; grey occurring abundantly in the tail, and occasionally on the body. Clydesdales are reared in Renfrew, Ayr, and Dumfries, although to the largest extent in their native Lanarkshire. The average height of the Clydesdale is about 16J hands for stallions and an inch or so less for mares. Breeders lay stress on the form of the feet, which should be large, round, and open, with abundance of "bone," and a free action. White feet, although common, are regarded as objection- able. A "dished" face, small ears, and a "pony" head are regarded as indicative of a strain of Galloway or garron blood ; while, on the other hand, a narrow face and Roman nose point as 132 THE HORSE AND ITS RELATIVES dearly to a shire cross. The feet should have long hair behind. As regards colour, bay or brown, with a blaze on the forehead, and the whole or part of the legs below the knees and hocks white, is the most prized ; but black, grey, or chestnut occasionally occurs, the last of these being regarded as indicative of a shire cross. The history of one of the most famous of the English heavy breeds, namely, the shire, or great, horse, has been fully worked out by Sir Walter Gilbey.^ This breed (pi. xii. fig. 2) was probably derived from the chariot-horses of the Britons of Caesar's time, and by the time of King John (11 99-1 2 1 6) had become the recognised English war-horse. With the increasing weight of armour, a heavier and larger type of horse became essential ; and accordingly breeders directed their attention to the production of such a type. During the Wars of the Roses (i 450-1 471) large numbers of great horses were exported, in order to escape being seized for military purposes ; but in the reign of Henry VII. (1485-1509) an Act was passed prohibiting the exportation of these and all other horses. In the succeeding reign — Henry VI n. — not only was this prohibition continued, but statutes were made for encouraging and improving the breed of shire horses. At this date the weight a charger had to carry (inclusive of his own armour) ^ The Great Horse, or Shire Horse, 2nd ed. London, 1899. HORSES OF THE BRITISH ISLANDS 133 was about 425 lbs., so that it is manifest a horse of great size and power was necessary. By Queen Elizabeth's time (1558- 1603) the exclusive re- striction of the great horse to military purposes was broken down, and these animals were in general use for farm and draught work. At this period it would seem that the colour of the great horse might range from black and bay to white. Accord- ing to Sir Walter Hungerford, who lived during the reign of Queen Mary, the British breed was at that time improved by the introduction of High Almaine (German), Flemish or Friesland, and, more rarely, Neapolitan, stallions. In the reign of James I. the great horse was still in use as a war-horse, as is proved by Vandyke's picture of the Duke of Arenburg, in the Earl of Leicester's collection at Holkam Hall ; and in spite of the introduction of lighter horses in succeeding reigns and during the Commonwealth, this usage continued till 1658, when a book was written by the Duke of Newcastle on the training and grooming of great horses for war purposes. By the latter half of the seventeenth century armour fell, however, into disuse, and the great horse, no longer required for war, or indeed for the saddle at all, took its place as an animal of draught, and eventually acquired the name of shire horse. It may be from the strain of North German and Flemish blood in the English great horse of best quality, that 134 THE HORSE AND ITS RELATIVES in the time of Paul Potter, who painted a portrait of a grey dappled stallion of the breed in 1652, the British and Continental representatives of the breed were practically identical. The frequent presence of black and grey at this period is indicative of Arab or Barb blood. The name shire horse seems to have been in use by about the end of the eighteenth century, for we find Arthur Young, in the description of a tour through England and Scotland, referring to two breeds of cart-horse as deserving of attention, namely the large old English black horse, " the produce principally of the shire counties in the heart of England, and the sorrel-coloured Suffolk punch, for which the sandy tract of country near Woodbrido'e is famous." During the last century, writes Sir Walter Gilbey, " the shire horse has played no mean part in building up size and massiveness in all the other drauofht-breeds in the kino^dom. That he has underofone or-reat chancres is certain ; but the characteristics of the breed — size, strength, substance, courage, and docility — have been perpetuated and developed by careful selection till we have now in our shire horse the ideal beast of draught." All the best characteristics of the breed were displayed by " Blythwood Conqueror," a famous stallion foaled in 1893, whose colour was bay, with a blaze on the forehead and all four feet white. HORSES OF THE BRITISH ISLANDS 135 These white markings are, indeed, very distinctive of the breed, which is further characterised by the relatively small size of the head, the short and heavy neck, thick and powerful shoulders, rounded and deep body, short and broad loins, massive hind-quarters, and enormously strong limbs, of which the lower portions are short and compressed, with an abundance of long hair on the fetlocks, the hoofs large and rounded, the frog well developed, and the lower surface of the hoof moderately arched. Nowadays grey is much less common in this breed than formerly ; the same being the case among thorouorhbreds. With such a build and size, shire horses are of course capable of performing extraordinary feats of power ; and they have the further advantage of possessing a very docile and tractable disposition, which, under careful training, renders them ex- tremely intelligent. Examples of this docility and intelligence are displayed by shires employed for hauling and shunting trucks on British railways. The skull of a shire stallion is shown in plate iv. fig. I, and a cannon-bone in plate i. fig. i. CHAPTER IV SOME FOREIGN BREEDS One of the most ancient and therefore one of the most interesting breeds on the Continent is the Schlettstadt horse, or pony, which has been described by Dr. Max Hilzheimer in an article entitled " Das Voso^esenrind und das Schlettstadter Pferd," published in the Mitteilungen der Philo77iat- ischen Gesellschaft in Elsass-Lothringen for 1906, vol. iii. pp. 368-380. This horse is to be met with in the neighbourhood of Schlettstadt, in upper Alsace, where it is locally known as Riedpferd (reed-horse) or Pickerle. Small in stature, and of all colours except grey, it frequently shows a dark dorsal stripe, while in one foal the last remnant of a transverse shoulder-stripe was observed, such a vestige being sometimes noticeable in the wild Mongolian tarpan. In its large and clumsy head, with a broad forehead, and a tendency to a con- cavity in the profile near the base of the nasal bones, the Schlettstadt horse likewise approaches the wild race, as it also does in its short ears and low withers. On the other hand, in its profuse mane and tail it makes an equally wide departure from 136 SOME FOREIGN BREEDS 137 the latter, although there is every probability that these features are the result of domestication. Dr. Hilzheimer is of opinion that this breed is the descendant of the wild horses — whether abo- riginally so or reverted to a wild condition from domestication, is more or less immaterial — described by Elisaeus Rosslin as having inhabited the Vosges at the end of the sixteenth century, and to which reference is made in an earlier chapter. ^ Moreover, he considers the Schlettstadt breed as nearly related to the horse of the Prehistoric Swiss lake-dwellings, or Pfahlbauten ; and that the latter, to which some writers have attributed an Eastern origin, was itself the offspring of the still earlier wild horses of the cave-epoch, such as the one represented in plate vii. fig, 2. On the other hand, it is suggested that the heavy so-called Eastern horse was not originally tamed to the west of the Alps or in the north of Europe, but that its ancestral home may have been in the neighbourhood of the Black Sea, whence it was carried by the Romans to Central and Western Europe. According to Dr. Conrad Keller,^ another ancient breed is to be found in the island of Majorca, in the Balearic group. These horses, which are most abundant in the Palma district, differ ' Supra, p. 74. * " Studien iiber die Haustiere der Mittelmeer-Inseln," Neue Denkschr. Schweiz. Naturfor. Gesellschaft, vol. xlvi. pp. 107-187, igii. 138 THE HORSE AND ITS RELATIVES markedly from all other breeds. Specially char- acterised by their slender limbs and free, graceful carriage, they vary in colour from dark to light brown, and have short, thick, and arched necks, with thick, upright manes, which are often clipped. The delicate head, with backwardly directed ears, is distinctly Roman-nosed, and when the animal is galloping, is carried sharply bent against the short neck. In this respect the Majorca breed differs markedly from Algerian and Andalucian horses, which carry their heads stretched out straight, nearly in the line of the neck. Dr. Keller compares the Majorca horses to those depicted on ancient vases and Greek coins, and believes the former to be the survivor of the ancient type. This identification, if trustworthy, is of great interest, as it serves to indicate that the hog-manes of the early Grecian horses, like those sculptured on the frieze of the Parthenon,^ were natural, although, as in the case of the Majorca breed, im- proved by trimming. This seems to be indicative of the affinity of both breeds to the wild tarpan ; and affords further evidence that the falling manes of modern horses (other than the Arab) are due to domestication. Of the heavy horses of France, perhaps the most famous is the Percheron breed (pi. xiii. fig. i), * See the cut on p. 297 of Ridgeway's Origin of the Thoroughbred Horse. PLATE XIII Fig. I Fig. 2 _'*--■.*•-. .-»* Fig. I. A Percheron Stallion. Fig. 2. A Belcian Stallion: SOME FOREIGN BREEDS 139 which takes its name from the district of Le Perche, in the south-east of Normandy. Although its origin is unknown, the Percheron is an ancient type ; and it appears to have been kirgely crossed with Arab and Barb blood during the Saracen invasion in the early decades of the eighth century. In 1755 the breed was crossed with Danish horses, and sub- sequently English and Belgian stock was introduced, while in 1820 other foreign blood was infused by means of two grey Arab stallions, which no doubt had a considerable share in inducing the grey colour now prevalent in the Percheron. This modern breed forms an ideal type of draught-horse, the height being relatively low (from 15^ to 17 hands in stallions, and the maximum half a hand less in mares), and the body compact and rounded, with a full chest and broad back. The rump is, however, short, and the tail set low — a feature showing little indication in this respect of the Arab cross — and there is also a lack of depth and fulness in the barrel. The shapely legs and feet, devoid of large hair on the pasterns, are very characteristic, and the cause of the free action in walking and trotting. In fact, next to the Clydesdale, the Percheron has the best action of all draught-horses. The colour is usually grey or black, although browns and bays are not unknown. The breed has been introduced into America, where it has become very popular. 140 THE HORSE AND ITS RELATIVES The Boulonnais breed, from the Boulogne district and the adjacent parts of Belgium, is a rather larger and coarser type than the Percheron, the neck being especially heavy, the rump steeper and more squared, and the colour frequently grey or white. The action is less free than in the Percheron. In Brittany the size of the draught- horses runs smaller, the normal height of the Breton breed being only from 14 to 15^ hands. Like the Percheron, to which they are allied, these horses are often grey, although bay is more common than among the former. They have been largely crossed with other breeds. Of less importance is the Nivernais, of the department of Nievre, in Central France, which is now mainly a black breed produced by crossing the native stock with Percheron stallions. Visitors on landing at Antwerp or other Belgian ports can scarcely fail to be struck with the intelli- gence, docility, and enormous power of the draught- horses employed on the quays. These horses (pi. xiii. fig. 2) belong to the Belgian breed, which is also used in the country for agricultural work, and appears to be of great antiquity ; Belgium having been noted as a horse-breeding country since the time of Diodorus Siculus, in the first century B.C. There is a certain amount of local variation in the height of this breed, the largest being the Flemish strain, in which it reaches from i6f to SOME FOREIGN BREEDS 141 17 hands, whereas in the Brabant type it falls to between 15J and i6f, while in the Ardennais stock, of the Ardennes, it is only from 15 to 15J hands. The Picardy horses of France form a fourth modifica- tion of the present breed. Perhaps the most striking feature of the Belgian is to be found in the great fulness of the chest and the depth and breadth of the back ; the girth of the body being relatively greater than in any other breed. The shortness and sharp inclination of the rump are more con- spicuous than in the Percheron, and constitute a distinct blemish. Another frequent fault is the lack of sufficient stoutness of bone in the legs, and the small and narrow feet, which, as in the Percheron, are devoid of long hair. The neat head is carried on a short neck, which is frequently of great depth, and thereby shows another indication of affinity with the French breeds. On the other hand, chestnut is the prevailing colour, bays, bay-browns, and roans being, however, by no means uncommon, although greys are rare. Despite its somewhat slow action, the Belgian, on account of its weight, enjoys an unusual capacity for moving heavy loads with the least possible amount of exertion and strain. As regards draught-horses of a lighter type, the rich lowlands drained by the Elbe, Weser, and Ems in North-western Germany have long been noted for the excellence of their breeds, among 142 THE HORSE AND ITS RELATIVES which special mention may be made of those of Hanover, Oldenburg, Schleswig-Holstein, and East Friesland. The original black Hanoverian appears, however, to have died out some time subsequent to 1746. Near Osnabruck is kept the celebrated stud of black Drenthe horses, originally from Drenthe, in Holland, to which reference is made on page 129, in the preceding chapter, where there is also mention of the royal cream-coloured horses, whose origin is Hanoverian. The name of German coach-horse is applied to horses of the above stamp, whose height ranges from 16 to i6| hands, and whose colour (exclusive of the creams) is almost invariably bay, brown, or black. The body is relatively large, with a high rump and well- set tail, the neck long and arched, the withers high, the legs relatively long, and the feet of excellent shape. The horses of Holland and Flanders are a heavier type of the same strain ; and, as already mentioned, it is from the big black horses of this part of the Continent that the English shire draws many of its present features. Germany possesses many other breeds ; but as these have been modified by crossing, they do not come within the scope of the present survey. Much the same may be said with regard to Danish horses ; the principal breeding-ground for these is on the Oldenburg border. Reference has been previously made (p. 122) to the dun and other horses of Norway, SOME FOREIGN BREEDS 143 which are the most characteristic type of that country. Hungary has long enjoyed a well -deserved reputation for its horses ; and it appears that the old Hungarian horse was usually bay, although grey, dun, and chestnut were also known. Early in the nineteenth century this type was, however, completely changed by the introduction of English thoroughbreds. On the other hand, it is important to mention that there is an indigenous Austrian breed of horses, standing about 14 hands in height, and in their angular make closely resembling the ponies of the Russian peasantry. " It seems certain," writes Professor Ridgeway,^ "that in these animals we have the descendants of the ancient ponies of the Danubian region, such as those driven by the Sigynnse, and their resem- blance to the country ponies of Russia confirms the conclusion that we have in them the old European- Asiatic horses more or less modified by crossing with superior blood." Switzerland possesses several breeds of heavy horses apparently derived from the mediaeval black war-horses, among which the Laumont breed of the Bernese Oberland forms an excellent draught- horse, whereas the black Erlenbuch is of lighter make. It would, however, be useless to devote more ^ The Thoroughbred Horse y p. 345. 144 THE HORSE AND ITS RELATIVES space to the horses of Europe (other than thorough- breds, which are discussed later), since nearly all of them have been more or less crossed with foreign blood. Professor Ridgeway, for instance, points out in one passage^ that "the black horses of Western Asia, Spain, and Italy all result from a mixture of the African bay [Barb] with the indigenous horses of Asia and Europe " ; while in a second ^ he mentions that " Wiirtemberg possessed a notable breed of horses, the best of which result from imported Arabs with an admixture in a small degree of the English thoroughbred and the Trakehaen " ; and in a third,^ after referring to the modern origin of the Russian Orlov trotter, he observes that " there is only one heavy Russian breed of draught-horses — the Beetewk, called after the river of that name. . . . In 17 12 Peter the Great was so struck by the good qualities of the horses of that locality that he imported Dutch stallions to improve the breed, and later on it was crossed with the Orlov trotter." In connection with Russia, it may be mentioned that the Kalmuks and Kirghiz own half-wild troops of coarse-bred horses, doubtless derived in great part from the original tarpan of the steppes, and much less altered by crossing with alien blood than the horses of Western Russia. Kalmuk horses, which are largely used in the Russian * op. cii., p. 320. "^ Ibid., p. 344. ^ Ibid., p. 350. SOME FOREIGN BREEDS 145 cavalry, are bred in the country lying between the Volga and the Ural, and stand about 15 hands in height. Kirghiz horses, on the other hand, are smaller, seldom exceeding 14I- hands, and are reared in the steppes to the north-east of the Caspian. The horses of Turkey can scarcely be said to form a distinct breed at the present day, being derived proximately from the ancient brown stock of Armenia, which itself originally came from Northern Asia, by crossing with Arab blood, which soon gained the predominance. Colonel Hamilton Smith ^ writes that " they have, from the ancient Turkoman blood, a tendency to Roman-nosed chaffrons and ewe-necks, but the head is finely set on ; they are delicate, have very tender and irritable skins ; but also they are docile, and grace- ful like gazelles." Before proceeding to notice some of the modern Asiatic breeds, other than Arabs, a few lines may be devoted to the early history of domesticated horses in the countries at the eastern end of the Mediterranean. With the exception of a solitary reference in Isaiah to their employment in threshing corn, horses are mentioned in the Bible only in connection with military operations. In Syria and Palestine they appear to have been unknown before the time of David ; but at that date they were used ^ Naturalist's Library, Horses, 2nd ed. p. 232, K 146 THE HORSE AND ITS RELATIVES both in war-chariots and for riding, the chariot-horse being termed sus, and the riding-horse jZ^^r^i^^. On the Egyptian monuments horses appear for the first time about 1600 B.C. harnessed to the chariot of the sun ; and it was not till a much later date that they were used for riding. Indeed in several Eastern countries the horse was employed for driving long previous to its use for riding. We find, for instance, in Assyria that the bas-reliefs portraying the conquests of Shalmanesir in Elam always show the Assyrians fighting in chariots while their enemies were mounted ; and it is stated that Sennacherib was the first to put cavalry in the field. These were mounted archers, each of whom required to be attended by a running foot- man, who had his hand on the bridle while the mounted man discharged his arrows. Later on the bowman learnt to manage his steed without assist- ance ; but there is stated to be no instance in the Assyrian sculptures of the use of the lance or sword by the cavalry of the period. It may be added that the heroes of the I/zad are always referred to as fighting from chariots, no mention being made of their mounting their horses before going into action. As regards the origin of these Eastern horses of the early historic period, Professor Ridgeway ^ is of opinion that those first introduced into Egypt * T/ie Thoroughbred Horse, p. 220. SOME FOREIGN BREEDS 147 came from the countries lying to the westward of the upper part of the Nile Valley, and were therefore of the Barb type. On the other hand, the horses of Syria and Palestine in Biblical times, which were of various colours, are considered to have probably come from Persia or the adjacent countries ; ^ the dun-coloured kadishes now found in the peninsula of Syria and Irak being a later introduction from Central or Northern Asia, and akin to the tarpan. The latter origin is likewise claimed by Professor Ridgeway ^ for the horses of ancient Babylonia (where they appear to have been introduced about 1500 B.C.) and Assyria; the first horses known in the Euphrates Valley thus being of the tarpan type. For centuries Persia has been noted for the ex- cellence of its horses, the typical breed being near akin to the Arab but rather taller and more slender in make. There are, however, other breeds in the country, such as the Turkoman and the half-bred Persian and Turkoman in the north-east; the Karadagh, a Cossack breed, in the north-west, near the Russian frontier ; the Kurdish breed in the province of Kurdistan ; and Arabs in the western and south-western districts. These Kurdistan ponies, which are usually grey or bay, came doubt- less in the first instance from Northern Asia, and therefore have the same origin as the Turkoman ; ^ The Thoroughbred Horse, p. 211. 2 Ibid., p. 198. 148 THE HORSE AND ITS RELATIVES when crossed with Arabs, they are largely exported into Turkey. They stand from 14 to I4-|^ hands. The true Persian may be regarded as a derivative from the Turkoman stock, specially modified by a orreater infusion of Arab blood. O The Turkoman, or Turki, horse takes its name from Turkestan, its original home, although it has spread into Persia, Armenia, and Asia Minor. There are several strains, of which the finest in- habit the country to the south of Lake Aral and the Sir Daria, or Oxus. Standing from 15 to 16 hands in height, and capable of great endurance, these horses have large, Roman-nosed heads, ewe-necks, slender bodies, and long limbs. Although gener- ally bay or grey in colour, some of them are black with white feet. The speed of these horses and such beauty as they possess are due to Arab parentage, grafted on an original stock doubtless more or less nearly akin to the Mongolian tarpan ; and it is noteworthy that the Turkoman horses to the north of the aforesaid line are much smaller and show much less evidence of Arab blood than those to the south of the same. Farther east the Turkoman gives place to the Mongolian pony, and the nearly related breeds of Bhutan, Nepal, Spiti, Ladak and Yarkand, to which allusion has been already made in the chapter on the tarpan, where mention is likewise made of the striped dun Kathiawar horses and the piebald SOME FOREIGN BREEDS 149 tanghans of Tibet, both of which appear to be de- rivatives from the tarpan stock crossed with Arab blood. Reference may, however, be made in this place to the horses of Northern Spain, which are quite distinct from those of the south. Many of them are grey and roan-grey, but in the sierras occurs a dun- coloured breed, in which the legs are frequently striped. That these are typical representatives of the old European dun stock may, as Professor Ridgeway^ remarks, be doubtful, as their striping may be the result of cross-breeding ; but even if this be so, their colouring is probably due to rever- sion to the original type. This is indeed practi- cally admitted by Professor Ridgeway, who on the page just cited remarks that " the horses of the Asturias and other mountainous areas of Spain are probably descended from the 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 Equidce should not have still survived in wild and mountainous districts." ^ The Thoroughbred Horse, p. 260. CHAPTER V THE ARAB STOCK By common consent the beautiful and hardy horses reared by the natives of the Nejd district of Central Arabia are acknowledged to display the finest de- velopment of which the equine type, as modified for speed alone, is capable. It is true, indeed, that they are outclassed both in the matter of stature and speed by the modern thoroughbred ; but that animal is but a derivative from the Arab crossed with the blood of the horses of Western Europe and Northern Africa, and its superiority is solely a matter of careful selection and breeding. The type was fully present in the Arab, and has merely been improved and developed. That the Arab type, with which may be grouped the Barb of Northern Africa, is markedly distinct from the original tarpan-like horses of Western Europe — the so-called cold-blooded horses of the Germans — is ad- mitted on all hands ; and the only question is as to the extent and degree of this difference. In other words, are the Arab and the Barb referable to Equus caballus, as typified by the original horses of Scandinavia, or do they represent a species by ISO THE ARAB STOCK 151 themselves ? Are they, in fact, the product and result of special selection and breeding, like the modern thoroughbred, or have they existed in practically their present form since the natural evolution of the horse tribe was completed ? To this question, which is by far the most impor- tant and far-reaching one connected with the history of the horse tribe, there is, unfortunately, no possi- bility of giving a decisive and indisputable answer. Consequently, extreme divergence of opinion, both on this point and in regard to the original place of origin of the Arab- Barb type, prevails among those who have studied and written on the subject. Colonel Hamilton Smith ^ seems to have adopted the view that the Arab is the product of breeding and selection, since he refers to it as the most artificial and the first of high-bred horses ; and he is followed in this view by General W. Tweedie,^ who remarks that " if special and exclusive breeding directed to a certain object explains our English race-horse, there is no need to go further for the secret of the Arab's foray-mare." No explana- tion is, however, given as to the stock from which the Arab horse was developed by selection, although it is pointed out that there is no evidence that Persia was the original home. Moreover, there is the difficulty that, in the first place Arabs are * Naturalist^ s Library, Horses, 2nd ed. p. 210. * The Arabian Horse, London, 1894, pp. 74, 241. 152 THE HORSE AND ITS RELATIVES unacquainted with the most elementary principles of horse-breeding, and secondly, that they did not begin to own and breed horses (at all events in the coast districts) till about the fifth or sixth century of our era. The latter fact would lead to the in- ference, on the assumption that Arabs were the originators of the breed to which they have given their name, that there were then no horses of this stamp in Palestine and Syria, which was not the case. Moreover, if some Eastern nation, Arab or otherwise, was able to produce high-bred horses from a stock akin to the Mongolian tarpan, there is not the slightest reason why some of the national- ities of Western Europe should not have accom- plished the same feat, which they certainly never did. And if the Arab was not thus evolved, it is clearly entitled to rank as a species apart from the original horses of Western Europe, which, as has been shown above, there is every reason to regard as descended from a tarpan-like stock. A somewhat different view is taken by Mr. Wilfred Scawen Blunt, ^ the well-known breeder of Arabs, who, after alluding to the fact that these horses have been maintained by the Bedouin for at least 1300 years, that is to say, from the sixth century of our era, goes on to observe that ' Article " Horse," EncydopCEdia of Sport, 2nd ed. vol. ii. p. 426, 191 1. PLATE XIV Fig. I Fk;. ~i Fig. I. The Darley Arabian. Fig. 2. "Persimmon,"' tht,- famous Thoroughbred Stallion owned by H.M. Kina Edward VIL THE ARAB STOCK 153 according to local tradition the Arab '* is a separate wild breed kept pure in the desert from the time of his first capture and domestication ; that his habitat was Nejd and the high plateaux of Yemen, and that he owes his distinguishing qualities to the fact that his original blood has never been mixed with that of breeds of inferior type. In physical science there is as yet nothing positively ascertained which would show this to be improbable. The high plateaux of Arabia, though all of them desert land, . . . are neither without pasture nor without water. It is unquestionable that the wild ass existed, if he does not still exist, in Yemen, and the wild horse, too, may have there existed." Later on it is added that " it is quite conceivable that in the gradual drying of the peninsula, of which we have geological proof, a section of the wild species may have found itself cut off in the south from the rest of its kind, and have developed there in isolation the special qualities we find in the Kehailan [Arab]." From the concluding passage it may be inferred that the wTiter considers the Arab to be descended from the same species as that which gave rise to the ordinary horses of Western Europe ; the argument will, however, stand just as well, and be even stronger, if the ancestral stock were regarded as a species by itself. In either case, some of the objections raised against the views previously 154 THE HORSE AND ITS RELATIVES referred to will be applicable ; and if the Arabs had these wild horses in their midst from Prehistoric time it is difficult to see why they did not domesticate them till the fifth or sixth century, previous to which they appear to have used camels. As to the argument that we have no evidence of the former existence of wild horses in Central Arabia, the same negative testimony might once have been cited to prove that there were never elephants in Mesopotamia, whereas there is geologi- cal and historical evidence to show that a species closely allied to, if not identical with, the Indian elephant inhabited that country during the early part of the human era. Professor Ridge way, i on the other hand, be- lieves Northern Africa to have been the birthplace of the Arab stock. "It is now clear," he writes, "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 Thoroughbred Horse, p. 246. THE ARAB STOCK 155 the ancestors of the kohl breed of Arabia came from North Africa." While admitting that the statement with regard to the early date at which the Libyans were in possession of horses may be perfectly true, it by no means follows that these horses were derived from an aboriginally wild African stock. In the first place, all the existing wild African representa- tives of the horse family are either asses (quite distinct from the so-called wild asses of Asia), zebras, or quaggas. It is true, indeed, that remains of extinct Equidce have been obtained from the superficial deposits of Algeria, and that these may be referable to true horses, more especially since many of the animals of North Africa are of a European type, and therefore quite distinct from those characteristic of the rest of the African conti- nent. There is, however, no evidence that this was the case ; and the point is apparently considered of no importance by Professor Ridgeway, who has devoted a whole chapter of his oft-cited work to a discussion of the origin of the Libyan horse. For he labours to show that the latter is related to the zebra-quagga group. Indeed he once went the length of suggesting that the North African horse is derived from Grevy's zebra. As indica- tions of zebra-affinities he affirms that African horses show a more marked tendency to be striped than breeds which have not the same blood in 156 THE HORSE AND ITS RELATIVES their veins ; and he also suggests that the white frontal star and stockings so commonly seen in African and Arab horses are remnants of the mark- ings of zebras and quaggas. Whatever value may attach to the first argument, the second, as already mentioned in the first chapter, has been shown by Mr. Pocock^ to be utterly untrust- worthy, such white markings being merely one of the first stages in the development of albinism. Moreover, the absence of the hind chestnuts and the nature of the front ones in zebras and asses show wide divergence from the horse. As a matter of fact, the theory of the origin of the Arab type from the zebra-quagga group has not a leg to stand upon ; and this being so, there remains little to be said in favour of an African birthplace for the former. That horses of the Arab type have existed in Libya from a very early date may be freely admitted. But the same is the case with humped cattle and their derivatives, which, as has been shown in the present writer's volume on the ox, originally came from South-western Asia, and are probably derived from the wild bantin of the Malay countries. Moreover, there is good reason to believe that the other domesticated ungulates of Africa, such as sheep, goats, and swine, are of Asiatic origin ; and there ' Annals and Magazine of Ntt, History., ser. > 8, vol. iv. '"p. 406, 1909. THE ARAB STOCK 157 is accordingly a prima facie probability that a similar origin may be attributed to the domesticated horses of Africa. This view is adopted by Mr. T. A. Cook/ who, after referring to Professor Ridgeway's Libyan theory, and stating that the Barb is as different from the true Arab as is the Turk from either, proceeds to observe that "as a matter of much greater probability, the kehailan, or Arabian, was the original type from which both Barb and Turk were early derivatives, and it was from the East, and not from the West, that ancient Egypt took her best breed, as eighteenth-century England took it later on." As mentioned in an earlier chapter,^ Dr. Duerst holds that the Arab had an Asiatic origin, and that, like the horses of Western Europe, it was a derivative from the tarpan stock ; the intermediate form in the case of the Arab being his so-called desert type. " The wild ancestral form," he writes, '* was the same for both [that is, the Arab and the horse of Western Europe] ; it was the diluvial horse of the ancient world, which roamed as far as the loess-steppes and tundra-plains extended ; and which, surviving in separate groups the disappear- ance of the tundras, was transformed, according to 1 Eclipse ajid C Kelly, London, 1907, p. 13. 2 Sufij'a, p. 96 ; the passage here quoted is from page 399 of Dr. Duerst's work. 158 THE HORSE AND ITS RELATIVES the newly-developing regional physiographic in- fluences, into the desert type, the steppe type, and the forest type." It is now time to devote attention to the physical attributes of the Arab, of which, so far as general external features are concerned, an excellent summary is given by Prof. David Law,^ who adopts the view that these are due to adaptation to sur- roundings. " Arabs," he writes, "are more compact than the horses of Barbary, having a rounder body, shorter limbs, with more of sinew, or what is termed bone. Yet they are of the smaller class of horses, very little exceeding, on a medium, fourteen hands, or fifty-six inches, in height. As compared with the horses of countries abounding in the grasses, their aspect is lean, their form slender, and their chest narrow. But the slimness of figure of these horses is not inconsistent with muscular force ; and their movements are agile, their natural paces swift, and their spirit is unmatched. The power of their delicate limbs is indicated by the well-marked muscles of the fore-arm, and the starting sinews of the leg. The shoulder is sufficiently oblique ; the withers are elevated : the back is moderately short ; and the quarters are good. The head is well formed ; the forehead is broad ; the ears are ' Domesticated Animals of the British Islands, 2ncl ed. p. 476. THE ARAB STOCK 159 somewhat long, but alert ; the eyes full and clear ; the veins prominent. . . . '* These desert horses subsist on the scantiest food, and are patient of hunger and thirst in a degree unknown in any other races except the African. They feed on the scanty plants which the borders of the desert supply, and when these are wanting, on a little barley, with chopped straw, withered herbs, roots, dates, and, in cases of need, the milk of the camel. They drink at long intervals and in moderate quantities. They bear continued exposure to the fiercest heat, and day after day perform marches of incredible toil through the burning sands of the wilderness." In the foregoing account sufficient emphasis is not laid on the absence of slope in the rump, and the consequent high setting of the tail ; features differing markedly from those obtaining in the wild tarpan and its relatives. The average height is also under-estimated, many Arabs standing 14I or 14! hands. The colour of Arabs is commonly bay or chest- nut, frequently with a star on the forehead and one or more of the fetlocks white ; but it may be black or grey, although never dun. For the various strains of Arabs, the reader may be referred to other works ; and it will suffice to mention that the highest or pure-bred is the one known as kohl or kehailan ; both names referring apparently to i6o THE HORSE AND ITS RELATIVES the blackness of the skin, which is compared to antimony [kohl). Five strains of the kohl breed are generally recognised by the Bedouin, of which the kehailan is the first and best. Here it may be mentioned that the low-caste horses of the towns are termed khadishes by the Bedouin. Takine both external and anatomical characters into consideration, Professor H. F. Osborn^ has formulated the following features as distinctive of the Arab : — The skull is relatively short, very wide between the eye-sockets, which are high and pro- minent, giving the eyes a wide range of vision, while the profile of the face is concave (pi. ix. fig. 2) and the lower jaw slender in front and deep and wide-set behind. The chest is rounded, and the back and the loins are well " ribbed up," due to the fact that there are only five (in place of the normal six) lumbar or ribless vertebrae. The pelvis has a nearly horizontal position — a character connected with speed ; the croup, or tail-region, is relatively high, and the number of caudal, or tail, vertebrae few. In the limbs the shaft of the ulna, or small bone of the lower part of the fore-leg, is complete ; the cannon-bones are elongated and slender, and the pasterns long and sloping. Allusion is also made to the occurrence of a slight depression in front of the eye-socket, and to the statement that ^ Bull. Amer. Mus. Nat. Hist. vol. xxiii, p. 259, 1907. THE ARAB STOCK 161 the bones are denser than in ordinary horses. The latter feature was not, however, observable in an American skeleton, although it may occur in desert- bred Arabs. The features to which the greatest importance are attributed comprise the sinuous facial profile (due to a relatively large brain), the absence of a sixth lumbar vertebra, the complete shaft of the ulna, and the shortness of the tail, which has sixteen in place of eighteen vertebrae. As regards the completeness of the ulna, it is notice- able that the same feature was observed in a skeleton of Grevy's zebra. Taken together, the foregoing distinctive features, in the opinion of Professor Osborn, are sufficient to justify the specific separation of the Arab, which appears to be de- scended from ancestors distinct from those which gave rise to the ordinary northern and western horses. Fuller allusion has been made in a previous chapter ^ to the preorbital depression in the skull which appears to be characteristic of horses of the Arab and Barb stock. It should be added that in Arabs the cheek-teeth (pi. v. fig. 2) are relatively small in comparison with the skull, and that the upper premolars have their transverse diameter as large as or larger than the longitudinal one, which is not the case with the horses of Western Europe,^ ' Supra, p. 22. 2 See Duerst, Animal Remains from Anau, p. 386. L t62 the horse and ITS RELATIVES while the foldings of the enamel in all the cheek- teeth are rather more complex than in the latter, and the longitudinal diameter of the anterior pillar is proportionately small. The first upper premolar, or "wolf-tooth," is not infrequently developed in the upper jaw. Professor Osborn adds that if the Arab is to be considered specifically distinct from the horse of Western Europe, it should bear the name Equus africanus, which was applied by Sanson ^ in the year 1869 to the Dongola breed. Professor Ridge- way ^ has taken objection to that name, and pro- posed to replace it by E. libycus or E. cabalhts libycus, on the ground that the Dongola horse is a half- breed, but such an objection is invalid. On the other hand, there is an insurmountable bar to the use of the name africanus owing to the fact that it was employed by Fitzinger in 1857 as the designa- tion of the African wild ass, for one of the races of which, as will be seen in the sequel, it is still used. There is, however, the name E. asiaticus, proposed for the Arab by Sanson in the passage cited, which is free from objection ; and that animal may therefore be known either as E. asiaticus or E. caballus asiaticus, according as to whether it is regarded as a species or a race. In discussing the origin of the Arab, Professor ^ Comptes Rendiis Acad. Set. Paris, vol. clxix. p. 1205. * The Tho7-oughbrcd Iloi'se, p. 477, 1905. THE ARAB STOCK 163 Ridgeway claims that India cannot have been the home of the ancestral stock, owing to the fact that during the historical period that country has been unsuitable for horse-breeding. The Narbada valley of Central India was, however, inhabited during the Pleistocene, or latest geological epoch, and pro- bably within the human period, by an extinct horse {Eqtius namadicus), while a second species i^E. sivalensis) has left its remains in the some- what older (Pliocene) deposits at the foot of the Himalaya. Obviously, then, the argument that India (which at the present day nurtures the onager in Sind and Cutch) is unsuited to horses applies only to part of the existing epoch. Now the Siwalik horse agrees with the Arab in the degree to which the facial part of the skull is bent down on the basal axis, in the presence of a preorbital depression, in the great relative width of the upper premolars, and in the complexity of the enamel-foldings in the centres of all the upper cheek-teeth, and the shortness of the p-rindino-- surfaces of their anterior inner pillars ; the two latter features being, indeed, more developed in the extinct species than in the Arab, and thereby approximating to the condition obtaining in the extinct three-toed hipparion, as described in a later chapter. The extinct species has also a large upper wolf-tooth. I have therefore suggested the possibility of i64 THE HORSE AND ITS RELATIVES the Arab being the descendant of the Siwalik horse or some nearly alhed species from Southern Asia, not necessarily India. It has been suggested that as the extinct Equiis stenonis has likewise a pre- orbital depression and short grinding surfaces to the anterior inner pillars of the upper cheek-teeth, it may claim to be regarded as the ancestor of the Arab or the Siwalik species. It was not, however, a native of the countries where, in my own opinion, the Arab probably originated. Turning to the Barb type, this breed has its native home in Morocco and Algeria, and in its original form stands from fourteen to fifteen hands at the withers. It is characterised by the flat shoulders, rounded chest, relatively long head, and, as compared with the Arab, the lower setting of the tail, the hair of which, like that of the mane, is profusely developed. The prevailing colours are dark bay, brown, chestnut, black, and grey. The skull has the same sinuous profile as that of the Arab; but the teeth have not been described. Formerly the Barb was extensively crossed with Syrian Arabs, while in Algeria it has of late years been much mingled with European horses, so that pure-bred animals are not easy to obtain. The Barb will thrive on as poor fare as the Arab, and is equally hardy in constitution and docile in temper, although somewhat less spirited. Several strains THE ARAB STOCK 165 of the Barb type are recognised by Col. Hamilton Smith, the first of which is reared by the Mograblns on the western side of the plains south of the Atlas, to whom it is known as shrubat-ur-rich (drinker of the wind). These horses, which may be either grey or brown in colour, are low and greyhound- like in shape, and carry very little flesh. More remarkable is the Bornu breed, from the district south of Lake Chad, which is stated to be greyish- white with black legs. The tail is set rather low, the legs and feet are beautifully made, and the body is relatively short. A third breed occurs typically in the Dongola district of Nubia, but is also found in Alfaia and Gerri. Typical horses of this breed are stated to be very similar in make to the Bornu type, but those of Alfaia and Gerri are smaller. The normal colours are bay, black, and white, with white legs in the two former. Professor Ridgeway regards the black and grey Dongola horses as half-breds, but the evidence for this does not appear very con- clusive. The near relationship of the Turkish horse to the Arab has been already mentioned ; and in Spain the jennet presents an equally close affinity to the Barb, from which it has undoubtedly been derived. Jennets, as might be expected, occur in their purest form in the southern provinces of Spain, especially 1 66 THE HORSE AND ITS RELATIVES Andalucia, Granada, and Estremadura. Bay- appears to be the predominating colour, next to which come black and grey. It has been very generally considered that Barb blood was first introduced into Spain during the Saracen Conquest, but Professor Ridgeway adduces evidence to show that the introduction occurred about a thousand years earlier, although a fresh infusion of the same blood was brought in by the Moors at the time they overran the country. Jennets are characterised by their easy pacing amble. The horses of Northern Spain, which are smaller than jennets, but may have a certain infusion of Barb blood, are referred to at the end of the fourth chapter. The influence which Arab and Barb blood has had on the indigenous breed of European and Asiatic horses has been incidentally mentioned in the course of the preceding chapters. The triumph of the Arab-Barb stock, when mated with the best indigenous breed, has been the development of the English thoroughbred, although, as the latter is essentially a modern type, it can receive but brief notice in the present volume. It will accordingly suffice to state that although the English breed of fast horses had been undergoing a slow but steady improvement for centuries, and that an Arab stallion (the " Markham Arabian ") was purchased for King James I. in 1616, it was to three THE ARAB STOCK 167 horses, the " Byerly Turk," the " Darley Arabian" (pi. xiv. fig. I), and the " Godolphin Barb," that the evolution of the modern thoroughbred is mainly due. The " Byerly Turk," taking his name from Captain Byerly, his owner, was imported in 1689; and from him was descended " Herod," who gave his name to one of the three orreat lines of English racing stock. The " Darley Arabian " was pur- chased in Aleppo during the reign of Queen Anne by the brother of his owner, Mr. Darley, of Aldby Park, Yorkshire in 1 702.^ He gave rise to " Flying Childers," and " Bartlett's Childers," from the latter of whom the famous "Eclipse," the great-great- grandson of the '• Darley Arabian," and the founder of the Eclipsi line, was descended. " Persimmon " (pi. xiv. fig. 2), owned by King Edward VH., was a direct descendant of " Eclipse, " and affords an example of the great increase which has taken place in the stature of racehorses ; his shoulder- height being 16I hands, whereas that of •' Eclipse," who was considered an unusually big horse for his time, was only about 15J hands. "Persimmon" was foaled in 1893; his sire being "St. Simon," and his dam " Perdita H." He was winner of the Derby and the St. Leger in 1896, and of the Ascot Gold Cup and ^ The date is usually given as 17 10, but the picture at Aldby Park from which pi. xiv. fig. i is copied, is inscribed 1702. i68 THE HORSE AND ITS RELATIVES 1897. His pedigree is Eclipse Stakes in follows : — as PERSIMMON, 1893 .St. Simon, 1881 Perdita II. Galopin, [Vedette, 1854 1872 \Flying Duchess, 1853 St. Angela, fKing Tom, 185 1 1865 [Adeline, 185 1 Hampton, (Lord Clifden, i860 1872 \Lady Langden, 1868 Hermione, /YoungMelbourne,i855 1875 '-La Belle Helene, 1866 Lastly, we have the " Godolphin Barb," a dark bay horse with some white on the off hind-fetlock, who was purchased in Paris about 1784, and pre- sented to Lord Godolphin (by whom he was regarded as an Arab). From his grandson, " Matchem," the third great line of English thoroughbreds derives its title. "Herod," "Eclipse," and "Matchem," it should be mentioned, were closely related ; and it is to their descendants that the term thorough- bred should be restricted. As the thoroughbred has been developed solely from the point of view of speed combined with staying power, it is only natural that he should not conform to the ideal type of equine beauty ; and, as a matter of fact, the frequent presence of ewe-neck detracts from perfect symmetry. Neither are these horses safe to ride. They have the broad forehead, brilliant eyes, delicate muzzle, expanded nostrils, and wide throat of the Arab and the Barb, with the body long and light, and the last rib THE ARAB STOCK 169 rather widely separated from the pelvis. The chest is deep but narrow, thus affording due space for the lungs without making the fore-limbs too wide apart. The obliquity of the shoulder gives full play to the upper part of the leg; while the extreme length of the haunch, and the elongated hind-limbs, with their long, sloping pasterns, are essentially adapted for the maximum development of speed. The most common colour is bright bay or brown, with black legs, mane, and tail, although chestnut is not infrequent; but black and grey (especially at the present time) are less common. Although, as stated above, the development of the English thoroughbred did not take place till the seventeenth and eighteenth centuries, it is important to mention that Irish hunters, which have long been celebrated, were derived from Barb horses imported into Ireland from Spain several centuries earlier. Remarks on the inheritance of coat-colour in thoroughbreds and on a possible connection be- tween colour and speed will be found in the first chapter. CHAPTER VI FERAL HORSES Although the comparatively modern domesticated breeds produced in America, Australia, and other countries to which the horse is not indigenous do not come within the purview of this volume, reference must be made to horses which have run wild in various parts of the world, since some of these display features of considerable interest in connection with the history and evolution of the family. For domesticated animals that have escaped from captivity and reverted to a more or less completely wild condition, it is frequently convenient to employ the term " feral " ; for, al- though this term, which is derived from the Latin ferus, is etymologically equivalent to the English "wild," it has acquired a special restricted meaning, which can be expressed by no other word in our language. To North America horses were introduced during the Spanish Conquest, and their feral descendants, like those of South America, are consequently of Spanish origin, and therefore of the Barb type, just as the feral cattle were originally FERAL HORSES 171 of the zebu stock. Writing in 1829, Sir John Richardson stated that at that time herds of feral horses were to be met with on the plains to the west of the Mississippi ; and that at an earlier date they were common in the Kutannie country near the northern sources of the Columbia River, to the east of the Rocky Mountains. The young stallions, which were expelled from the main herds by their seniors, formed troops by themselves. Early in the eighteenth century feral horses abounded in Virginia ; and as these enticed away the domesticated horses of the English settlers, the Spanish type became gradually modified. There are also herds in Texas, where they are known as mustangs, and likewise in Mexico ; many of the former being piebald or skewbald. The feral horses which formerly abounded on the pampas of Argentina appear to have been descended from five stallions and seven mares of Andalucian origin which escaped when the city of Buenos Aires was suddenly abandoned by its inhabitants about the year 1535. These rapidly multiplied, and gave origin to the herds on the pampas to the south and west of the Rio de la Plata ; but the troops to the north of that river, in Paraguay, were derived from another stock. Although these horses frequently went about in small troops, each led by a stallion, these troops sometimes combined into herds com- prising thousands of individuals. When the herds 172 THE HORSE AND ITS RELATIVES sought fresh pastures, they were led by a few of the older stallions, who gave warning of impending danger. Like kiang^ in Ladak, the herds come up to gaze at novel objects, and run in circles round bands of mounted travellers, who on such occasions find it difficult to prevent their own animals from escaping. When once caught, the wild horses of the pampas, like those of the North American prairies, soon re-acquire domesti- cated habits. It is stated that these pampas horses, or baguals, as they are called, have acquired larger heads, longer ears, and more muscular limbs than their domesticated ancestors ; although, on account of the mildness of the climate, there has been no marked increase in the length of their coats. Very noteworthy is the statement" that "their colour is always of a chestnut-brown, and never dun, as in the Tatar races ; and whenever a bay, a black, or other colour appears, it is inferred that the indi- vidual is of the domesticated race, and has made its escape and joined the wild herds." This affords further evidence in favour of the view that the Barb, or Andalucian stock, is descended from a species distinct from the wild dun tarpan of Mongolia. Horses from La Plata, and therefore probably of the Andalucian type, were introduced by the French in 1764 to the Falkland Islands, where * See chapter vii. * See Low, Domesticated Animals of British Isiamis, p. 499. FERAL HORSES 173 they subsequently ran wild. According to Darwin, their predominating colours are roan and grey : like the domesticated cattle of the same inclement islands, they have become stunted in size, their average height being only about 14J hands. It is stated that the Falkland horses, like the tarpan, have the habit of scraping away snow with their hoofs in order to get at the herbage beneath ; and it is generally considered that this is an instance of re- version, although this can scarcely be the case if they are all of Andalucian, i.e. Barb, stock. The Puno ponies of the high Cordillera of Chile afford another instance of a dwarf feral race apparently derived from the same stock. In A Naturalist on the " Challenger'' Professor H. F. Moseley states that in the peninsula of Lafonia, where the Falkland horses run larger than elsewhere, the stallions guard their own herds of mares. "They keep the closest watch over them, and if one strays at all, drive her back into the herd by kicking her. The younger horses live in herds apart, but the more vigorous ones are always on the look-out to pick up a mare from the herds of the older ones, and drive her off with them, and they sometimes gather a few mares for a short time and hold them till they are recaptured. When they think they are strong enough, they try the strength of the old horses in battle, and even- tually each old horse is beaten by some rival and 174 THE HORSE AND ITS RELATIVES displaced. The fighting is done mainly with the tusks, and front to front, not with the heels. Thus the most active and strongest males are constantly selected naturally for the continuation of the herds." In most parts of America the feral horses appear to have no special difficulty in defending themselves from the attacks of predatory carnivora, such as jaguars, pumas, lynxes, wolves, coyotes, and bears. The case is, however, different in certain parts of Patagonia, where pumas are so numerous that wild horses seem unable to exist. It has accordingly been suggested by Mr. W. H. Hudson that these carnivores were the cause of the extermination of the indigenous American repre- sentatives of the equine family. The suggestion, however, was made at a time when the possibility of extermination being frequently due to bacterial agencies was not generally recognised. The following extracts from a letter published by Professor Ridgeway^ afford valuable informa- tion with regard to feral horses in Australia : — •' Wild horses have been running in the mountain- ous country of East Gippsland, in which are the sources of the Buchan River, and through which flow the Snowy River and its tributary the Deddik. 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 ^ The Thoroughbred Horse, '^. 431. FERAL HORSES 175 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 hi^h mountain plateau which lies between the upper Tambo River and the sources of the Buchan River I have seen horses which can best be 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 . . . the horses were of a much better stamp, in many cases showing good breeding, partly owing to the excellent stamp of the New South Wales horses of about fifty years ago, but also to the fact that a Persian horse . . . escaped and lived for many years in the Tubbut country." In some of the above districts these brumbies, as they are locally called, became a nuisance to the settlers, by whom they were eventually exterminated, and a similar extirpation of feral horses has taken place, for the same reason, in other parts of the world. CHAPTER VII THE KIANG AND ONAGER GROUP The wild Asiatic representatives of the Equidce^ other than the tarpan, are commonly known as wild asses, but they have really no right to that name, any more than they have to be called horses, in the literal sense of that term. Indeed, this does not really express the truth of the matter, for some of them are more nearly related to the horse than they are to the ass. Consequently it is far preferable to employ their native or classical names, such as chigetai, kulan, kiang, onager, and ghor-khar. The near relationship of these animals to the horse, and more especially the wild tarpan, may be made apparent by the following table : — I. — AH the hoofs broad, the front much broader than the hind pair, i. — Front hoofs very broad, chestnuts usually on all the legs, ears small, tail more or less completely haired to root, front of fore-legs usually black in bay or dun- coloured individuals. The horse, Equus cabalhis. a. — Mane long and pendent, with a forelock, tail long and fully haired, normally no dorsal or shoulder stripe. E. c. typicus. Domesticated breeds, typified by those of Sweden and Norway (exclusive of the Arab and Barb stock). 176 KIANG AND ONAGER GROUP 177 b. — Mane short and erect, basal portion of tail short- haired, a dorsal and a shoulder stripe in the summer coat, muzzle usually white or yellowish. The Mongolian tarpan, E. c. przevahkii. ii. — Front hoofs less broad, chestnuts on fore-legs only, ears large, tail short-haired for a considerable distance from the root, front of fore-legs yellowish or white. The kiang and onager group, E. hemionus, &zc., of Asia. II. — All the hoofs narrow and nearly alike in form. This group includes domesticated and wild asses, zebras, and quaggas, all of which are African. In addition to their broader front hoofs, the members of the kiang group are characterised by the absence of striping on the head, body, and limbs, the general rufous or sandy colour of the upper-parts, and the lighter tint of the under surface of the body, part of the buttocks, and the limbs, all of which may indeed be white. In all cases the spinal region is traversed by a dark dorsal stripe, varying in width in the different species ; and occasionally faint bands are noticeable on the shoulders, knees, and hocks. As a rule, the mane, which is upright, and the terminal tuft of the tail, are black. To a considerable extent the cry, which in the case of the kiang has been described as a shrieking bray, is intermediate between the neigh of the horse and the bray of the ass, although apparently nearer to the former than to the latter. Kiangs and onagers are, however, distinctly less noisy animals than the ass ; and in this respect present, perhaps, another point of resemblance to M 178 THE HORSE AND ITS RELATIVES the horse. The ears, although larger than in the horse, lack the excessive length and breadth dis- tinguishing those of the ass. On the other hand, the members of the kiang group display affinity to zebras and asses, not only in the absence of hind-chestnuts, and the large size and smooth surface of the front ones, but likewise in the length of the period of gestation, which is about a twelve- month, whereas the mare only goes with young for eleven months. Of all the members of the group the largest, and in some respects the finest, is the kiang (Equus kiang) of the elevated plateaux of Ladak and Tibet, where it goes about in small troops, which gallop in circles round the mounted traveller or his camp in such a manner as to completely prevent in many instances the successful pursuit of nobler game, or, I might say, game of any kind, as kiang are scarcely entitled to that designa- tion. Curiosity is a marked trait of the kiang ; so strongly developed in some instances that young individuals, as has happened to myself, will walk almost into the camp. These animals are free movers, going at a fine, springy trot, and the manner in which they traverse the most rocky ground, and this, too, at an elevation of between 13,000 and 16,000 feet, is marvellous; their hoofs must be like flint, and their lungs as strong as bellows. Fig. cr'f ■^ Fig. I. The Kiang in summer coat. Fig. 2. Kobdo Onager, KIANG AND ONAGER GROUP 179 The kiang (pi. xv. fig. i), which has a shoulder- height of about 13 hands, has been regarded as a local race of the chigetai, but it differs from that animal by the redder colour of the upper-parts, and the sharply-defined demarcation between this red area and the white of the muzzle, under-parts, buttocks, and limbs, thus giving a kind of skewbald appearance, which is most marked when the animal is in its short summer coat ; the long and shaggy winter dress tending to obscure the difference between the dark and the light areas. The ears are characterised by the presence of a dark patch at the base, and another at the tip. The kianof was first brous^ht to scientific notice by Moorcroft, one of the early explorers of Kashmir and Ladak, whose travels, which contain an excellent account of the habits of the animal, were published in London in 184 1. In Ladak the kiang is to be met with a few marches to the eastward of the city of Leh, and abounds in the great Chang-Chenmo plain and the arid country around the wonderful Pangong lake, the home of the chiru or Tibetan antelope, and formerly, the yak. Thence it extends northwards to the Kuen-Lun and eastwards into Tibet, where the limits of its range are still un- known. Scant as appears to be the nutriment in these barren countries, which in summer are scorched at midday by a burning sun, but become bitterly cold at night, it suffices to keep these i8o THE HORSE AND ITS RELATIVES animals, as well as hares and marmots, in prime condition. Unlike the African wild ass, which displays a holy horror of water, the kiang is very partial to that element, and never lives far away from some lake, river, or stream, into which, when occasion requires, it plunges without hesitation to take a longer or shorter swim, despite the icy coldness of Tibetan rivers. In Mongolia the kiang is replaced by its cousin the chigetai (or dziggetai, as the name is spelt in German fashion), E. heinioniis, which is a rather smaller and more uniformly coloured animal, of lighter make, and more rounded hoofs. The dif- ference in general appearance is due in the first place to the less rufous tint of the darker areas in the summer coat, and secondly to the fact that this shades off almost imperceptibly into the dirty white of the under surface of the body and the paler fawn of the throat and limbs. Having the same narrow dorsal stripe and dark tips to the ears as the kiang, the chigetai lacks the dark patch at the base of the ears distinctive of the latter. Information is still required with regard to the extent of the range of the chigetai ; but the animal is generally believed to be identical with the kulan of the Kirghiz Tatars. Nearly allied to the chigetai is the species known in Persia as the ghor-khar, i.e. horse-ass, KIANG AND ONAGER GROUP i8i and to the ancients as onager, i.e. wild ass (Greek onos, ass, and agrios, wild), which inhabits the deserts of Asia from Syria and Persia in the west to North-western India and Mongolia in the east. The onager [Equus onager), which is the wild ass of the Bible, is a rather smaller and paler-coloured animal than the chigetai, with nearly as well marked a contrast between the dark and light areas as in the kiang. Standing from ii to 1 1|- hands at the shoulder, it has ears of much the same relative length as in the chigetai, but the hoofs narrower and more ass-like, this being especially the case with the front pair, which are scarcely wider than the hind ones. The profile of the face may be either nearly straight or markedly sinuous ; the tail-tuft is of moderate size ; and the dark dorsal stripe, which is always much wider than in either the kiang or the chigetai, stops in some cases short of the tail-tuft, and is flanked on either side, at least in the posterior half of its length, by a white or whitish band, joining the white on the buttocks and the backs of the thighs. In the summer coat the general colour of the upper- parts is usually some shade of pale reddish fawn or sandy, while the light areas, which vary from pure white to whitey brown, are much the same in extent as those of the kiang, but embrace more of the buttocks, from which they spread along the margins of the dorsal stripe, and in some cases 1 82 THE HORSE AND ITS RELATIVES occupy more of the body and head. In winter, when it grows much longer and rougher, the coat becomes more or less decidedly grey, and in one race is distinctly mouse-grey, with sharply-defined white areas. With such a wide geographical distribution, it is not surprising to find that the species is divis- ible into a number of more or less well-defined local races. One of the best-known of these is the Indian ghor-khar {E. onager indicus) of the desert districts of Sind, Cutch, Baluchistan, Eastern Persia, Afghanistan, and thence as far north as Bokhara, which is stated to attain a height of ii^ hands, and comes nearest to the chigetai. In Baluchistan the ghor-khar is most abundant near Mithankot, on the Punjab frontier. These dis- tricts lie to the west of the Indus ; to the east of that river the chief districts frequented by this animal are Bikanir, Jeysulmere, and the saline tract known as the Rann of Cutch. With a straight facial profile, this ghor-khar has the general colour of the upper-parts sandy in summer, with the light band on each side of the dorsal stripe narrow, ill-defined, and whitey brown, and the white on the rump and thighs not pure. The broad dorsal stripe does not reach as far as the tail-tuft in the Indian representatives of this race, although it is stated to do so in Persian examples. The second race.^^^. o. cas^aneus{p\. xv. fig. 2), is at KIANG AND ONAGER GROUP 183 present known only by a single example, reported to have come from the Kirghis-Nor, in the Kobdo district of Western Mongolia.^ It is characterised by the straight profile of the face, the rufous isabella- colour of the summer and the full greyish brown of the long winter coat, as well as by the large amount of white on the buttocks, and the distinct- ness of the pure white band on each side of the broad chocolate-coloured dorsal stripe, which reaches to the tail -tuft; the lateral white bands uniting with a broad white blaze on the buttocks, which is larger than in the other races. The short ears have more black at the tips than in the Indian race. In the typical or Persian ghor-khar {E. 0. typicus), from Western Persia, to the southward of the Caspian, the white areas, as compared with those of the Indian race, have become so enlarged as to give the appearance of a white animal with three large fawn-coloured patches on each side. The general colour is silvery white; the dorsal stripe does not reach the tail-tuft; and the head, the sides of the neck, a small ill-defined band on the front of the shoulder, a larger quadrangular patch on the side of the body, and the middle of the hip, are isabella-colour, or pale sandy fawn. The facial profile is distinctly convex, and the ears are relatively small. Owing to our imperfect acquaintance with the ^ The locality is doubtful. i84 THE HORSE AND ITS RELATIVES animals of Syria and Palestine, the fourth race {E. 0. hemippiis), the wild ass of Scripture, cannot yet be properly described. It inhabits the deserts between Bagdad and Palmyra, Mesopotamia, and Northern Arabia. Nearly related to the last, it appears to be reddish isabelline above, with the throat, under-parts, and a band on each side of the dorsal stripe silvery white, the dorsal stripe not reaching the root of the tail, which is moderately haired, and the profile of the skull sinuous. The extremely light colour of the Persian race appears to be an adaptation to a purely desert existence, being paralleled in Africa by the mohr gazelle and the white oryx of the Sahara. Other adaptations to surroundings are shown by the thick- ness and length of the winter coat of the reputed Kobdo race, which is evidently an inhabitant of a country with a cold winter, as compared with that of the Indian race, which is quite short. Like all large desert herbivorous animals, the ghor-khar is famed for its speed ; this being so great that adults in good condition can neither be ridden down (unless, perchance by relays of horse- men) nor taken with greyhounds. Baluchis are indeed stated to have accomplished the former feat, but this was probably only when mares heavy in foal were the objects of pursuit. On the other hand, ghor-khar foals are commonly captured in summer in the Bikanir desert by parties of mounted KIANG AND ONAGER GROUP 185 Baluchis, who relieve one another, and are thus enabled to continue the pursuit till the victims are completely exhausted. Ghor-khars are mostly found on the fringe of desert plains, where they usually associate in small troops, although in Afghanistan herds containing sometimes a thousand head have been observed ; such large herds being apparently composed of mares and foals, the old stallions collecting in smaller troops by themselves. In Baluchistan, as probably also in Persia, the foals are born in June, July, and August. On the plains the food of the onager consists of desert grasses, which are often in a parched and withered condition. In Baluchistan ghor-khars migrate to the hills in early summer when the plains are practically devoid of grass and water. Unlike the kiang, they are exceedingly shy and suspicious, and consequently difficult to approach within rifle-range. In former days kulan and onagers appear to have ranged much further westward than is the case at the present day. It is stated, for instance, by the Russian naturalist Rytschkov ^ that in the eighteenth century kulan abounded on the eastern side of the Volga, and from time to time troops swam that river and made their appearance in the Waldinsel Steppe. Then, again, in spite of the difficulty of specifically distinguishing many of the ^ See Nehring, Uber Tiindren iind Steppen., p. 187. i86 THE HORSE AND ITS RELATIVES fossil remains of the genus Eqmis, it is generally considered that during the Pleistocene period either the kulan or the onager (for it is impossible to distinguish between them in the fossil state) ranged into Central and Western Europe. Dr. Nehring,^ indeed, believed that all the remains of small equines from France and Germany were referable to these animals ; but Prof. M. Boule,^ while admitting the occurrence of the onager in Western Europe, refers, as will be more fully noticed in the sequel, the majority of such remains to the ass. 1 Op. cit., p. 187. ' Annales de Paliontologie, vol. v. p. 132, 1910. CHAPTER VIII ZEBRAS AND QUAGGAS Africa south of the northern tropic, that is to say, the Ethiopian Africa of naturalists, is the home of a group of more or less fully striped members of the horse-family, commonly known as zebras and quaggas. As to the derivation of the latter name there is no sort of doubt, quagga being the Dutch corruption of the Hottentot title — qua-cha—oi one of the southern species, which is taken from the animal's cry. With regard to the origin of the name zebra, there is a difference of opinion. In one dictionary ^ it is stated, for instance, that the name comes from the Hebrew tzebi, meaning splendour or beauty, and connected with the verb tzdbd, to shine, and the equivalent of the Arabic zib, beauty. An old writer, Job Ludolphus, in his Historia ^thiopica, published at Frankfort-on- Maine in 1681, states ^ that these animals are called zecora in Abyssinia and zebra on the Congo; while Colonel Hamilton Smith ^ shows that zebra " seems to be the Negro mutation of the Abyssinian ^ Ogilvie's Student" s English Dictionary, London, 1865. ^ Book i. chap. x. ^ Naturalisfs Library, Horses, 2nd ed. p. 321. 1 187 i88 THE HORSE AND ITS RELATIVES zeuru of Lobo and the Galla zeora or zecora, according to Ludolphus." To the Greeks and Romans zebras were known as hippo tigris, a word meaning horse-tiger, although in Liddell and Scott's lexicon it is translated "large- tiger," from a mistaken idea that it is analogous in meaning to such words as horse-chestnut and horse-radish. There is evidence that one at least of these hippotigres was exhibited in the Roman amphi- theatre, by Caracalla ; ^ and there is little doubt that this belonged to the species now known as Grevy's zebra, which is therefore not only the true hippotigris, but likewise the true zebra, although the latter designation is now applied to the South African species. In early days these zebras were considered royal gifts by the Abyssinian emperors, and it is stated by Mr. H. Scherren ^ that some were sent by King Assaghedus to the Governor of the Dutch East India Company at Batavia, by whom they were presented to the Emperor of Japan. " News of these facts was sent to Ludolphus by Emanuel Nawendorff, a native of Altenburg resident in Batavia. In return the Emperor sent 10,000 silver taels and thirty Japanese dresses, so that, as this correspondent of Ludolphus says, they were amply paid for. And the latter notes that these animals might be sent by sea if they would only bear the * Hamilton Smith, 'h 5 '^he hollows between the enamel-foldings being completely filled with cement. In the skull the socket of the eye is surrounded by a complete ring of bone, there is no deep depression or pit immediately in front of the same, and the slit between the fore part of the upper jaw and the nasal bones is short. A long gap, or diastema, in the front half of which are implanted the tusks, or canines (when these are present), separates the incisor from the cheek-teeth ; and the crowns of the incisors themselves are penetrated, in early life, by the pit, or "mark," which has already been sufficiently described. Each limb terminates in a single hoof, upon which alone the animal walks ; but the lateral toes — second and fourth — of the extinct three -toed members of the family are represented by splints, which may either remain free or be welded to the adjacent cannon-bone. 246 THE HORSE AND ITS RELATIVES As mentioned in earlier chapters, the Prehistoric and Pleistocene deposits of Europe and Turkestan have yielded remains inseparable for the most part at any rate from the modern horse {Eqims caballus), of which they probably represent several phases or races, while others have been assigned to the ass {E. asinus), and yet others to the onager {E. onager). From the Pleistocene gravels of the Narbada Valley, in Central India, have been obtained skulls and other remains of a horse (^. namadicus) characterised by the elongation of the o-rinding surface of the anterior pillars on the inner side of the upper cheek-teeth ; the same species also occurring in the topmost beds of the Siwaliks of Northern India. In North America the Pleistocene and Upper Pliocene formations have yielded remains of at least nine extinct members of the modern genus ; one of these, E. fraternus, closely resembling E. caballus, while a second, E, giganteus, from South- western Texas, appears to have been the largest of the whole group ; the cheek-teeth exceeding those of the biggest cart-horses by more than one-third the diameter of the latter. In the Upper Pliocene deposits of the Val d'Arno and other parts of Europe, including the so-called Forest Bed of the coast of Norfolk, occurs a horse i^E. stenonis) with molars of a somewhat more primitive type than those of the existing PLATE XXII Bones of fore-feet of extinct forerunners of the Horse. A, Hyracothcriuin, or Eohippus; B, Mesohippus ; Q, Meiychifpiis ox Protohippits ; D, Hipparion. THE FORERUNNERS OF THE HORSE 247 members of the genus ; the primitive feature being the shortness of the antero-internal pillar of those of the upper jaw. There is also a slight depression in the wall of the skull immediately in advance of the socket of the eye. Nearly allied to Steno's horse is Equus sivalensis, of the Pliocene deposits of the Indian Siwaliks, which, as already mentioned, exhibits both the aforesaid features. So far as can be determined, this Siwalik horse seems to have stood about 15 hands at the shoulder, and to have had a relatively big head, and slender cannon-bones, with proportionately large splints. After mentioning that in respect of their upper molars Equus stenonis and E. sivalensis occupy an intermediate position between the modern members of the genus and the extinct Pliocene Protohippus, Prof. Marcellin Boule' proceeds to observe that when a large series of the remains of the first-named species is studied, " it will be found that in respect of stature, of the plications of the enamel of the upper molars, and of the shape of their inner pillars, Equus stenonis presents individual variations com- parable to those which occur in the different varieties or races of the horse, both of the Pleisto- cene and modern epochs. Certain teeth, of relatively small size, display remarkable resemblances to those of the ass ; and I have shown that milk-teeth of ^ "Les Chevaux fossiles des Grottes de Grimaldi,'M««. de Palion- tologie, vol. V. p. 130, 1 9 10. 248 THE HORSE AND ITS RELATIVES E, stenonis from certain Pliocene formations in the Auvergne and Velay present certain features now found only among zebras. I have also shown that other teeth, from a higher horizon in the Pliocene, differ from the former by their larger size, the more complicated folding of the enamel, the greater length of the anterior pillar, and certain other features connectingf them with the Pleistocene representatives of E. cabalhcs, more especially those molars with very complicated enamel-foldings, like those from England described by Owen as E. plicidenSy and those from certain French caverns, such as Cindr6 and Bruniquel. "It appears, then, that the stenonis type was extremely variable during the Pliocene, being represented by forms already showing tendencies towards various modern groups of EqtiidcB. Among these forms, all of which are remarkable on account of the shortness of the anterior pillar of the upper cheek-teeth, some were of small size, and apparently related to the asses, having the enamel-folds of the cheek-teeth relatively simple, and their external lobes dilated ; these seem to have developed directly into the zebras of modern Africa. Others, of larger size, with the enamel of the molars more plicated, and the external lobes forming more pronounced crescents, appear to have passed insensibly into some of the larger forms of Equus caballus found in certain Pleistocene formations." THE FORERUNNERS OF THE HORSE 249 After remarking that the modern horse dates back to the Pleistocene, or Quaternary, epoch, but is unknown in the Pliocene, Prof. Boule observes that "it is found at Chelles and in most of the deposits of the great interglacial period. It persists right through the Quaternary, everywhere in abundance. It is represented in the Middle Pleisto- cene by races or varieties, to which it is possible to affiliate certain modern races or varieties. Finally, it seems that the evolution of this useful and interesting animal continues to the present day, and that under human influence it will still continue to progress." That the extinct Indian E, sivalensis and E. namadicMS, which, as already mentioned, differ from one another in the length of the grinding surface of the anterior pillar of the upper cheek- teeth (this being short in the former and long in the latter), gave rise to successors seems almost certain. At one time I suggested ' that the kiang might be derived from E. sivalensis ; but the cheek- teeth of the former are so much smaller than those of the latter, that this seems unlikely. A latter suggestion is that if the Arab be specificially dis- tinct from E. caballus it may have originated from the Siwalik species, probably through the later Narbada horse. ^ PalcEontologia Indica {Mem. Geo/. Sur7>. India), ser. 10, vol. ii. p. 89, 1882. 250 THE HORSE AND ITS RELATIVES In North America the immediate precursor of the modern Equus is the genus Pliohippus, whose remains occur in deposits belonging to the Lower Pliocene and to the antecedent Miocene epoch. Whether this was a completely single-toed animal, or whether there were small remnants of the lateral toes, appears uncertain. If these were present they must, however, have been extremely small and quite functionless, as the splint-bones are scarcely larger than in the modern horse/ The cheek-teeth are laro^er than those of the under- mentioned ProtohippiiSy but owing to the relative shallowness of the jaws, as compared with those of Eqtms, their crowns were still more sharply curved. Pliohippus, as typified by P. pernix of the Loup Fork beds of Nebraska, was the largest of the con- temporary horses, standing about 12 hands at the withers, and thus equalling a good-sized pony. The range of the genus included the Western United States, especially Nebraska and Oregon. Not improbably this or a closely allied form was the direct ancestor of Equus. It has likewise been suggested that Pliohippus gave rise to the remarkable Hippidium {^Hippidion) and Onohippidiimt of the Pleistocene deposits of South America : this is considered problematical by Prof. Lull, although more favourably received 1 See R. S. Lull, "The Evolution of the Horse Family," ^wU6f ^ (^^ y ^ 7: 4 AM ~U^ ^— ^ .v^^^^ ^ ^^ ««c :>