APRIL 21, 1904] 
NATURE 
591 
eof which for more than a hundred yards are crowded with | 
animal figures, there are, in addition to twenty-three nearly 
full-sized engravings of horses, numerous studies of equine 
heads. Some of the Combarelles horses decidedly differ 
from those of La Mouthe. There is, e.g., a large drawing 
of a heavily built horse (Fig. 2) with a coarse head, an 
arched muzzle, a thick under lip, rounded quarters, and a 
tail with long hair up to the root. 
pat 
\\: 
/s 
7 
f 
graving (Fig. 3) of a head 
ic. 1.—Horse with a!long head, 
which in outline is wonder- 
fully suggestive of an 
from an engraving in the cave of 
La Mouthe. 
Arab, and at yet another 
part of the cave a horse 
with a pony-like head is re- 
presented, behind which 
stands an animal with a 
head like that of a modern 
Shire horse. 
In addition to the types 
figured on the walls of 
caves, there are others 
carved on pieces of horn 
and other durable  sub- 
stances. The majority of 
the horses engraved on 
horn are characterised by a 
out ee * very large coarse head, 
(Munro’s “ Prehistoric Horses.’’) but a few (e.g. the horse 
from the Kesserloch cave 
mear Schaffhausen) are remarkable for the small size of the 
thead, the fine muzzle, and small ears. 
As already indicated, the men of the early Stone age 
haye left us drawings of some four or five different kinds 
of horses, some with large heads and stout limbs, some 
with fine heads and slender limbs, some with’ a nearly 
‘straight croup and a well-set-on tail, others with rounded 
‘quarters and the root of the tail far below the level of the 
croup. Drawings made at the present day will be of little 
use some centuries hence in providing an answer to the 
question, How many species of horses existed in Europe 
at the beginning of the twentieth century? They will con- 
fuse rather than enlighten future inquirers, because for 
several generations breeders of horses, like breeders of 
cattle and dogs, have with the help of selection and isola- 
tion succeeded in creating numerous artificial strains. Is 
there any reason for supposing the evidence afforded by the 
prehistoric drawings is more valuable to us than recent 
drawings will be to our successors thousands of years 
thence, should they desire to ascertain how many species of 
Britons possessed 
thorses at the end of the nineteenth 
| 
fic. 2.—Engraving of a heavily built horse, from the Combarelles cave (;';)- | 
(Munro’s ‘‘ Prehistoric Horses.’’) 
century? That depends on whether in Palzolithic times 
the horse was domesticated in Europe. 
It is extremely probable that the men of the early Stone 
age had now and again tame horses, just as nowadays we 
have at times tame zebras, but it is most unlikely that they 
thad herds of horses which they systematically bred and 
reared as stockmen now breed and rear cattle. 
That the domestication of the horse as now understcod 
avas not attempted in Paleolithic times may be inferred 
NO. 1799, VOL. 69 | 
At another part of the | 
from the fact that the majority of the horses in the Solutré 
bone-mounds were from five to seven years old. Had 
horses been bred for food as we nowadays breed cattle, 
young individuals would have been most abundant at 
Solutré. 
If it is admitted that the engravings on the walls of 
caves and on pieces of horn fairly accurately represent 
animals which actually existed at the end of the Ice age, 
and if it is also admitted that the creation of new varieties 
by artificial selection was never attempted until at the 
earliest the arrival of the Neoliths, it follows that in post- 
Glacial as in Pleistocene times there were several perfectly 
distinct wild species of horses in Europe. 
For some reason or other it has hitherto been very 
commonly assumed that, as in recent times the wild striped 
horses of South Africa—the quagga and zebras—have been 
gradually supplanted by occidental or oriental domesticated 
horses, the wild horses of Europe were gradually displaced 
by domesticated varieties introduced by the Neoliths. It 
seems to me quite unnecessary to assume that the indigenous 
varieties so long familiar to the Palaeolithic inhabitants 
were exterminated. 
The advent of the Neoliths, instead of implying the ex- 
termination of indigenous varieties, in all probability meant 
the introduction of yet other varieties. 
I may here repeat that now, as throughout the nineteenth 
century, it is generally assumed that all the domestic breeds 
—small as well as large—have sprung from a single wild 
species. The great French naturalist Cuvier not only believed 
that all living horses belonged to one species (the Equus 
caballus of Linnzus), but also that there was no specific 
difference between living breeds and the fossil horses of the 
Pleistocene period. Prof. Sanson, of the French National 
College of Agriculture, in 
his ewirartés ide -Zao0- 
technie ’’ (1901), assuming 
a single origin for domestic 
breeds, divides recent 
horses into two groups—a 
long-headed and a_short- 
headed group—each of 
which consists of several 
races, while Captain 
Hayes, in his recently pub- 
lished ‘‘ Points of the 
Horse ’’ (1904), says, ““ nO Fic. 3.—Head of a horse with a 
profile like that of an Arab, from 
the Combarelles cave (3). 
(Munro's ‘‘ Prehistoric Horses.’’) 
breed of horses possesses 
any distinctive character- 
istic which serves to distin- 
guish it from other breeds,” 
and adds that ‘‘ as a rule locality . . . and artificial selec- 
tion are the chief factors in the formation of breeds.”’ 
Elsewhere Captain Hayes states, “‘As far as I can 
learn, no attempt has been made to separate ponies from 
horses except on the purely artificial basis of height.’"* 
Even those who are prepared to admit that recent horses 
may have sprung from several wild species allege that 
owing to domestication, intercrossing, and artificial selec- 
tion it is no longer possible to indicate the distinguishing 
characters of the two or more wild species which took part 
in forming the present races and breeds. 
Tue Witp Horse (Equus caballus prjevalshkit). 
The wiid horse may be first considered. For many years 
the semi-wild Tarpan of the Russian steppes was regarded 
as the nearest living relative of the wild ancestor of the 
| domestic breeds, but in 1881 the existence of a true wild 
horse was announced by the Russian naturalist Polyakov. 
This horse occurs in the vicinity of the Gobi desert and 
the Great Altai Mountains, one variety living to the south- 
east, another to the west, and a third to the south of 
Kcbdo. All three varieties are of a yellow-dun colour, the 
scuth-eastern (Zagan-Nor) form being especially character- 
ised by a dark muzzle, dark points, and a dark mane and 
tail; in the western (Urungu) variety the muzzle is nearly 
white, the limbs are light down to the fetlocks, and the 
mane and tail are of a reddish-brown tint, the southern 
(Altai) form being nearly intermediate in its coloration. 
The markings consist of a narrow dorsal band, faint indica- 
tions of shoulder stripes, and indistinct bars in the region 
1 ** Points of the Horse,” pp. 422-425- 
