DARWIN'S SPECIFIC VARIATION. 125 



a white barb with a red tail and a red 

 spot on the forehead, and which notori- 

 ously breeds very true : the mongrels were 

 dusky and mottled. I then crossed one of 

 the mongrel barb-fantails with a mongrel 

 barb- spot, and they produced a bird of as 

 beautiful a blue colour, with the white 

 loins, double black wing-bar between, and 

 barred and white-edged between tail feathers, 

 as any wild rock-pigeon. We can under- 

 stand these facts on a well-known principle 

 of reversion to ancestral characters, if all 

 the domestic breeds are descended from the 

 rock-pigeon." 1 



Here a curious anomaly strikes us. Iff 

 birds that bred very true came from the 

 rock-pigeon, it is singular that uniformity or 

 fixity of type should come through variation. 



That the barb-spots of Darwin's experi- 

 ment bred very true to their various peculiar 

 markings indicates that this hereditary re- 

 production must be wrought out by some 

 potent if delicate organisation, and it is 

 at least mysterious how a delicate organisa- 

 tion so stable could be the outcome of long- 

 continued variation. If Darwin had told us 

 whether other characteristics . of the breeds 

 were reproduced in these hybrids, we should 



1 Origin of Species, Ed. vi., p. 18. 



