50 THE PRACTICAL rlUEON KEEPER. 



special value. We also learn, again, tliat the one necessai7 

 point in forming such a strain is never to lose a link in the 

 process, or groiind already gained ; but to keep definite objects 

 steadily in view. 



This, however, brings us back again to the difficulties of 

 such a course ; and, first of all, it is impossible to follow out 

 such a methodical system without very considerable in-breeding. 

 It is always found, practically, that a man who is constantly 

 buying other birds to cross with rarely succeeds. We have 

 seen the reason why : he is constantly introducing into his 

 strain tendencies which, being unknown to him, he can- take no 

 account of, and which crop up in the most unexpected manner. 

 Therefore, successful breeders always depend mainly upon suc- 

 cessive generations of their own stock. But imfortunately this 

 method has its limits, which cannot be passed over, on account 

 of the physical deterioration or weakness which results from too 

 close breeding of the same strain. Tliis may be carried very 

 much further, with care, than some authorities are disposed to 

 allow, without any perceptible injury ; but sooner or later 

 .symptoms occur which warn the breeder of the necessity for 

 <' fresh blood." It may be he finds perceptible loss of fertility, or 

 his young ones often fail to break the shell, or there is a marked 

 increase of liability to disease, or some other evident falling ofi' 

 in vigour or in size. Hence an occasional cross from at least 

 another family becomes a necessity ; and this at once brings us 

 back again to the old difficulty of developing not one only, but all 

 the various points which the breeder has in view. The problem 

 may be briefly summed up as follows : We have seen that 

 any given point can be bred with a certainty few beginners 

 Jiave any idea of; but that the means to this end are the con- 

 centration towards it of all the transmitting powers of many 

 successive generations, and never losing any step once gained by 

 going ofi' in some other direction or after some other point, in- 

 flomiif^li !Mi anv such step must undo more or less of what had 



