454 GENETICS IN RELATION TO AGRICULTURE 
to be very simple and easy of solution. This deficiency will undoubtedly 
be remedied as the results of definitely planned experiments become 
known; for the present it is necessary to make the most of the meagre 
data now at hand. 
Selection as a Cause of Variation.—It is unnecessary to reopen here 
the question of the causal connection between variation and selection, 
for the arguments which have been presented in various places for the 
belief that modifying factors, rather than quantitative changes in a given 
Mendelian factor, are responsible for changes in a given character by 
selection apply just as well to domestic animals as to any other living 
beings. It is, however, true that a large number of the characters of 
farm animals are of a type such that they may be shifted with compara- 
tive ease in a given direction by selection. The characteristic white- 
face pattern of Hereford cattle behaves as a unit in heredity, but unques- 
tionably there is sufficient evidence to show that by continued selection 
the extent of the white area may be increased or decreased in exactly 
the same fashion that Castle by selection was able to increase or decrease 
the amount of the black pigmented area in hooded rats. No experiment, 
however, has yet been carried out, or is likely to be undertaken, in 
which over 30,000 Hereford cattle have been bred and raised for the 
purpose of determining definitely what are the limits of selection for 
this character. Similarly in such breeds of cattle as the Ayrshire and 
Holstein-Friesian the proportion of pigmented area to non-pigmented 
may be shifted at the will of the breeder from asolid pigmented condition to 
one almost entirely white. With more complex characters such as speed, 
milk production, and others of utilitarian value the evidence is even more 
convincing that selection does gradually shift the mean of the race. 
Whatever may be the true interpretation, there can be no question that 
the breeder may guide the variation of nearly every character in a definite 
direction by proper methods of selection. 
Variation by Modifiability—The success of the art of the caretaker 
and feeder in animal breeding depends upon the existence of a high degree 
of modifiability in domestic animals. Extreme cases such as the effect of 
starvation contrasted with the effect of liberal feeding are easily recog- 
nized, but in more obscure cases it is difficult to state how much of a given 
effect is due to inheritance and how much to modifiability. Thus the 
high standard of present racing records does not depend solely upon more 
careful attention to the selection of breeding stock, for improvement in 
methods of feeding, care, and training has been associated with this 
greater discrimination in the selection of breeding stock. The same fact 
is true in even greater degree with respect to the marvellous records which 
have been made by dairy cows during recent years, for elaborate methods 
of development and feeding have been devised to stimulate production to 
Digitized by Microsoft® 
