290 



VAELiTION OF PLANTS AND ANIMALS. [Ch. XXXVI. 



have sometimes 

 the internal bo: 



modifications 



m 



framework 



Tlius while 



I 



they have given a longer body to the pouter, they have nnin- 

 tentionally augmented the number of its sacral and caudal 

 vertebra3, and the breadth of the ribs as well as the size of 

 the breast-bone. In the fan-tail they have increased con- 

 siderably the length and number of the caudal vertebrae ; and, 

 what is still more worthy of note, in several breeds the whole 

 skull differs in its proportions and outline from that of the 

 rock pigeon. 



So many passages have been traced between the most di- 

 vergent varieties above alluded to and the Avild Columha 

 Liviay that ornithologists do not hesitate to recognise this 

 species as the common progenitor of them all. Another 

 curious proof of such a derivation is afforded by crossing dis- 

 tinct breeds and finding in the offspring some peculiar cha- 

 racters of the rock pigeon, especially in the plumage, which 

 neither of the parent races possessed.^ Thus the blue slaty 

 colour, or dark bars, on the wings and tail, and the white 

 edging of the outer tail feathers of the original Columha 

 Livia^ are produced in the mongrel offspring of the carrier 

 and fan-tail, although all these characters have often been 

 in abeyance in both of the parent stocks for a hundred or 



more 



Mr 



singular principle of atavism, by experiment, in the case of 

 pigeons, and has also obtained analogous results, by pairing 

 some of the most distinct varieties of the common fowl; 

 as, for example, a black Spanish cock and a white silky hen, 

 two ancient and pure breeds in which there was not a trace 

 of the red colour proper to the plumage of the wild Galhcs 

 hanlciva^ a Himalayan species, which has always been sup- 

 posed to be the original of our domestic fowls. In many 

 of the young obtained from such a cross the peculiar orange- 



) 



red colour was 



t 



Revival of long-lost characters in the offsprings of cross-hreeds. 

 Why the act of crossing should tend to evoke characters 

 which had long been lost in each of the parent races, is 



^ Darwin * On Variation/ vol. i. p. 200, 



t Ibid. p. 241. 



