Formation of New Breeds. 
63 
type would remain pure. It is questionable, though, how far 
it would be practicable to make such a combination in the 
case of the larger animals, though theoretically there is no 
difficulty about it. Where, however, a number of offspring 
are produced annually one season’s testing would be sufficient 
to determine the purity. 
From the foregoing account it will be seen that breeding 
has entered upon a definite stage, and that order can be traced 
in a subject which hitherto has appeared to be chaotic. The 
breeder has now to recognise that new breeds can be built 
up with certainty by recombining characters already existing 
in other breeds, and improvements in type will result from 
picking out a desirable feature here, another there, and so on, 
and combining them together. These characters are handed 
on in their full intensity, and as far as it can be determined 
there is no possibility of definitely diminishing or intensifying 
any character by breeding. The intermediates which result 
from certain combinations which might at first sight appear 
to contradict this statement are unstable. In some few cases 
evidence has been found which shows that characters are 
sometimes coupled together in some way and are incapable 
of being separated on crossing, but this phenomenon does 
not appear to be a general one, and at present it cannot be 
discussed in detail. 
Whether any fresh features not already existing in the parents 
will ever appear among the descendants of cross-bred animals 
is still a matter for future experiments to decide. Indications 
that such may be the case have been found, and though 
unexpected at first, Mendelian explanations of their occurrence 
are not wanting. In this direction, then, there is still the 
chance which always appeals to the breeder of producing 
a type totally distinct from the forms he set out with. 
R. H. Biffen. 
Cambridge. 
