236 



has been domesticated is impossible to say. I have 

 searched all the oldest literature we have on the 

 subject, with the result of finding the earliest allusion to 

 it in 1556.* But in these early days it seems to have 

 been long enough domesticated to have already shown 

 distinct and more or less fixed variations of plumage 

 through the agency of artificial selection. In 17 13 

 there were no fewer than twenty-nine distinct 

 varieties, including one that must have required very 

 man\' generations to produce — the crested bird. If, 

 therefore, we sa\^ that the Canary has been domesti- 

 cated for the last five hundred years we shall more 

 than probably be much within the mark; and since 

 we have definite proof that two hundred years ago egg 

 was always being given to breeding birds, and that 

 judging from various indications in the literature of 

 the da}^ it was even then being given in deference to 

 old traditions, it is fair to assume that the Canary has 

 been exposed for at least four or five hundred 

 generations to the influence, not only of the septic 

 bacillus as found in the ordinary way in cage life, but 

 also of that virulent form which is generated at the 

 high temperature of a bird's bod}^ in the favouring 

 nidus of egg food. So it is quite easy to see why it 

 survives under those ordinar}^ conditions which render 

 the fate of the newly-caught wild bird an almost 

 certain death, and wh}- even with egg food its death 

 rate is still much less than that of the other bird is 

 without it. 



Continuing the argument, we shall also realize that 

 although our breeders of Canaries annually lose great 

 numbers of birds, both old and young, from septic 

 enteritis, and moreover succeed in setting up now and 

 then dangerous and fatal epidemics by giving the 

 favourite egg, yet it must be also conceded that by so; 



Buffon, Hist. Nat. cies Oiseajix, Tome IV., page 26, and Histone des Incas 

 Tome II., page 239. 



