21 



it in their produce. Do the same as to colour^ decide 

 at the outset what this shall be, and do all you can to 

 get it pure and distinct. Thus, whether it be pure 

 ■white, pale lemon or buff, bright yellow or deep orange, 

 grey or green, cinnamon or mottled ; in short, whether 

 it be the Lizard, the London Fancy, the Yorkshire 

 Spangle, or the prize Belgian, do not mix them, but 

 keep them separate, selecting such birds only as are 

 likely in your judgment to produce ofFspriug still mora 

 excellent than their parents. With regard to shape, 

 there is little or no difficulty to contend with, the rule 

 of "like producing like" invariably holding good. 

 With regard to colour, however, owing, I imagine, to 

 the promiscuous breeding and crossing hitherto prac- 

 tised, it is nothing like so certain. Could you only 

 obtain birds of any given colour, whose ancestors had 

 never been crossed with those of any other for eight 

 or ten generations previously, I have little doubt that 

 it would be as constant and unchanging as any other 

 quality you could name. But owing to the want of 

 the systematic breeding we recommend, this rule can 

 hardly be depended upon at present, or indeed any 

 other, for all are liable to be marred by vagaries which 

 the breeder could not possibly anticipate. Hence all 

 the discrepancies and conflicting advice given by 

 various writers upon the subject, such being true 

 probably as regards his own experience, but scarcely 



