28 INCREASING OUR SUFPLIES OF CAVALRY HORSES. 
2 inches to 16 hands is the best size for cavalry horses, as for 
hunters. 
With regard to the use of Arabs as sires for cavalry 
horses, I wish to speak with considerable diffidence, because 
my own personal experience of them is but limited. As far, 
however, as it extends, it is decidedly adverse to their em- 
ployment, unless in exceptional cases. The few Arabs I 
have seen were characterized by the upright shoulders which 
the writer to whom I have alluded attributed to our thorough- 
bred horses. They were, moreover, low in the forehand (an 
unpleasant conformation to the rider), and apt to be too 
drooping at the pastern. The progeny of Arabian sires, out 
of English mares, is usually small and light of bone, though 
pretty, and possessed of showy action. Their character is 
that of park-hacks or of ladies 5 horses; and they would at 
once be rejected by any officer purchasing troop-horses, as 
unfit to carry even the lightest of our so-called 66 light cavalry 5 ’ 
troopers. Unless, then, for the exceptional case of an over- 
sized or enormously powerful mare, it is useless to expect 
cavalry remounts from the use of Arab sires. 
A much higher authority has recommended that recourse 
should be had to sires like the weight-carrying hunter, with 
the view of perpetuating the breed of horses under conside- 
ration. 1 am convinced that such advice, if largely carried 
out, would lead to the most fatal results. Your readers may 
perhaps recollect that I have always strongly insisted upon 
the necessity of purity of race on the part of the sire, what- 
ever may be the class of animal which it is desired to produce. 
That the male ought to be thorough-bred, or of an accredited 
pedigree, and of a higher caste if possible than the female, is 
a maxim I believe unanimously upheld alike by the highest 
theoretical and practical authorities in breeding. For my 
own part, I never knew T it departed from without signal 
failure. Taking only one or tw T o of the more obvious consi- 
derations connected with such a course into account, it is 
obvious that such must be the almost inevitable consequence. 
On what ground does any man, w 7 ho reflects at all, select a 
sire? Why, that he wishes his offspring to resemble him. 
But it is well know 7 n that the power possessed by either 
parent of imprinting their own type upon their offspring 
depends upon the purity of blood and antiquity of race of 
each. Thus the offspring of a thorough-bred short-horn 
bull and a common cow will frequently resemble very closely 
the character of the pure short-horn. In like manner, when 
a hackney mare or a Welsh pony is put to a thorough-bred 
horse, the offspring show's indications of being much more 
