238 
given up the correlation-idea. I do not think it will have to be 
the work of the experimental biologist to pass judgement upon 
the merits of a given type, his work is the study of the genetic 
factors, leaving the practical men to judge his combinations. 
The work of manipulating the genetic factors of the agri- 
cultural plants of all the different parts of one country can easily 
be centralized, and done for a sum of money, insignificant in 
comparison to the benefit the population in general would ultimately 
derive from it, and undoubtedly most governments will realize 
this importance and follow the example given by the government 
of the United States. But for obvious reasons it is impossible to 
do the necessary work with animals at one Station. The only way 
to have this work done is to leave it to the practical breeder, 
under supervision and advise of the experimental biologist. It is 
not to be desired that the advise of this man will take the form 
of showing the ideal towards which must be bred, for, more often 
than not, in the case of animals, the difficulty is not to produce 
a not yet existing combination of genetic factors, but rather to 
produce a homogeneous type, a strain homozygous for all the 
desired genetic factors. 
The practical men know very well what they want, and it 
remains for the experimental biologist to show them how they 
can attain the goal they have set themselves. Often also, it will 
be necessary for him, to devise experiments to find out, in how 
far the desired effect depends upon a certain combination of 
genetic factors, and in how far upon non-genetic factors, wliich 
might be easy of control. Such experiments with animals can be 
centralized in a certain way, by having the experiment-records 
kept at the central experiment-station. In work with animals 
everything has yet to be done from the bottom upwards. The 
existing system of-breeding animals, and above all of registering 
them has resulted in a great many cases in a population in which 
a minority have the desired genetic Constitution, the rest, though 
nearly all bred from individuals, Coming up to this Standard, being 
below it. In the first place it will in such cases have to be 
ascertained in how far the desired result depends upon genetic 
factors. It will be found that in most cases, such qualities as 
appear difficult to fix are such which need for their formation the 
Cooperation of a genetic factor, for which most of the individuals 
are heterozygous. It will have to be taught the practical breeders 
