596 
to meet under the \orelock, differs greatly from the 
‘“ bumpy ”’ forehead of Prjeval ‘s horse and the flat fore- 
head of the Celtic pony. The space between the orbit and the 
nostril is relatively longer than in the Celtic pony, but 
shorter than in Prjevalsky’s horse. The eyes project beyond 
the level of the forehead. In the Celtic pony the eyes are 
large and adapted for a wide range, in the wild horse they 
are some distance from the front of the head, in the Norse 
horse they are small and look downwards rather than for- 
wards. The outline of the face becomes convex above the 
muzzle, and ends in a somewhat long upper lip, adapted, 
like the upper lip in the giraffe, for feeding on leaves and 
twigs. In the neck and shoulders, trunk and limbs, the 
Norse variety may be said to resemble a small cart-horse 
of the Suffolk type. 
Compared with the wild horse, the withers are lower and 
the hind-quarters more rounded, and the tail springs more 
abruptly and at a lower level, and hence fails to convey the 
impression that it is a direct continuation of the trunk. 
The dock is relatively longer than in the Celtic pony, but 
shorter than in the wild horse. Ihe limbs are short, but 
the joints are large and the hoofs fairly broad, hence in a 
side view of the foreleg a considerable increase is noticed 
as the thick fetlock joint is reached. 
It will be evident from what has been said that the Norse 
horse differs chiefly from the wild Gobi horse in being of a 
eyes 
Photograph by G 
Fic. 7.—A richly striped dark yellow-dun horse of the Norse type, which has a general resemblance 
to the Combarelles horse reproduced in Fig. 2. 
darker dun colour, in being far more richly striped, in the 
shape of the head, size of the ears, position of the eyes, 
and also in the muzzle, mane, tail, hind-quarters, joints, and 
hoofs. From the Celtic pony the Norse horse also differs 
in the colour and markings; but it especially differs in the 
tail and in the greater proportional length of the distance 
between the eye and the nostril, and in having a 
complete set of ergots and chestnuts. It is inconceivable 
that the Norse variety could revert to the Prjevalsky 
horse type, or be regarded as an offshoot from the Celtic 
pony. ; 
The question 1 
\y now be asked, Is there any evidence 
that the Palzolith 
of the south of Europe were familiar 
with horses of the Norse type? Fig. 7 gives an imperfect 
idea of a specimen of the Norse race from the west of 
Ross-shire. If this figure of a horse still living is com- 
pared with Fig. 2, which f ithfully reproduces an engraving 
made thousands of years ago in the Combarelles cave by 
one of the artist-hunters of the Early Stone age, it will, 
I think, be admitted the Norse horse probably belongs to a 
very ancient race. 
NO. 1799, VOL. 69 | 
NATURE 
[ApriL 21, 1904 | 
I need onty auu that I regard the Norse race as the 
foundation of what in the Highlands are known as garrons. } 
Horses of this type may very well have been originally 
obtained by blending the old indigenous yellow-dun striped 
race with Flemish and French breeds imported direct from | 
the Continent or introduced from England during the, 
middle ages. Further, it is extremely probable that the 
Norse race took part in forming the small active Clydesdales, 
of a former generation. | 
OtHER OccIDENTAL HORSES. 
In addition to oriental and African varieties, which doubt-' 
less include several wild species amongst their ancestors, ; 
there are two or more occidental varieties which in various! 
ways differ trom the Norse and Celtic races and from 
Prjevalsky’s horse. 
One of the latter races include long, low, heavily built | 
animals with unusually long heads, another consists of 
short-bodied animals with a large head and a pronounced 
Roman nose. 
The long-headed variety which occurs in the Hebrides 
and the Central Highlands reminds one of the horses en-- 
graved during the Stone age on a piece of reindeer horn. * 
In one specimen of thiS variety met with in Perthshire the’ 
profile is straight and the distance from the orbit to the 
nostril is 13 inches, t.e. 2 inches more than in a member 
, of the Norse breed of a like size, and‘ 
4 inches longer than in a 14 hands 
Connemara pony allied to the Celtic ' 
race. Some of these long-headed forms 
with a straight profile and a_ well 
moulded muzzle resemble the horses of 
the Parthenon. 
Horses with a 
pronounced Roman ‘ 
nose also occur in the western islands 
and Highlands of Scotland, and in 
Ireland, Austria, America, and other 
parts of the world into which breeds 
were introduced from Spain. One of 
this Roman-nosed type, of a yellow-dun 
colour, met with in the Outer Hebrides 
was especially interesting. It very 
decidedly differed from members of the 
Norse race in the same district, but, on 
the other hand, it agreed in the outline 
of the head with some of the engravings 
in the Dordogne caves. It is hence 
conceivable that the Roman-nosed 
variety (from which the modern Shire 
breed may be an offshoot) is a very old 
one—a variety which was firmly estab- 
lished centuries before domesticated 
breeds first made their appearance in 
Europe. 
Summary. 
A. Ewart. I have endeavoured to indicate that in 
post-Glacial as in  pre-Glacial times 
there were several distinct species of 
horses, and that it is extremely prob- 
able some of the prehistoric species and varieties have 
persisted almost unaltered to the present day. I have 
shortly described three distinct kinds of living horses, viz. 
the wild horse of the Gobi desert (E. c. prjevalskii); the 
Celtic pony, which though no longer wild, may be known 
as the E. c. celticus; and the Norse horse, which may very 
well stand as the type of one of the large occidental breeds, 
and be known as EF. c. typicus. I have also pointed out that 
in addition to these three very distinct types—two at least 
of which have taken part in forming quite a number of 
our British breeds—we have a long-headed, heavily built 
variety with a straight profile, and a long-headed, heavily 
built variety with a more or less pronounced Roman nose. 
I have also indicated that in addition to several occidental 
varieties there are several African and oriental varieties, 
and I might have added that, in so far as the English 
thoroughbred is a mixture of African and oriental varieties 
and of occidental light and heavy varieties, it might be 
cited as an excellent example of a breed which includes 
amongst its ancestors several wild species—a breed which 
has had a multiple origin. 
