WILDFOWL DOGS— HOW TO BREAK THEM 63 



notions he may hear propounded on this subject by would-be 

 tutors, who in reality are ignorant misleaders. 



To begin with, the dog bred for a purpose has in it the 

 elements of its work. In the crossed animal there must be a 

 mingling which has some counteraction one way or the other. 

 Perfect dogs owe more to their breeding than to their training 

 and working. Several crude examples might be cited to show 

 my meaning, the most forcible of which I will put in the form of 

 a question : if we want white dogs, should we breed them from 

 black ones ? Of course, exceptions occur, but these in no way 

 concern the general fact. In addition, selection by man is 

 dependent to a considerable extent on the attainment of per- 

 fection in animal training, and on this point I will venture to 

 mention the old saying, "What the Lord hath left out no man 

 can put in." Here we have a point which very closely applies 

 to the dog-breaker's work, and one of extreme importance. 



A feature to be noted in all dogs, especially those which 

 are to be the working servants of man, is their disposition and 

 temper. I have generally found that when dogs have been 

 well bred (by this I mean carefully selected and confined in 

 breeding to one variety) their tempers have been mild, whereas 

 in cross-breds the reverse has occurred. I do not say this is 

 always so, I am only speaking generally and as far as my 

 experience goes. In show-dogs we occasionally meet with 

 bad-tempered champions. These animals, however, can only 

 stand as exceptional instances. 



Speaking of temper in dogs, my arguments may be better 

 understood from the following. If we accept the theory of 

 evolution, or, more correctly, variation under domestication 

 subject to selection by man, in dogs, we must naturally agree 

 that, according to their breeding, they are subject, if (!arelessly 

 or cross-bred, to reversion in some shape or form, whether in 

 outward appearance or temperament. By selection, we can 

 characterise to an unlimited extent in the modification of a form. 



