Chap. V. CORRELATION OF GROWTH. 169 



rock-pigeon. This greatly increased length of the orifice of the 

 nostrils seems to stand partly in correlation with the enlargement 

 of the wattled skin on the upper mandible and over the nostrils; 

 and this is a character which is selected by fanciers. So again, 

 the broad, naked, and wattled skin round the eyes of carriers 

 and barbs is a selected character ; and in obvious correlation with 

 this, the eyelids, measured longitudinally, are proportionally 

 more than double the length of those of the rock-pigeon. 



The great difference (see woodcut No. 27) in the curvature of 

 the lower jaw in the rock-pigeon, the tumbler, and Bagadotten 

 carrier, stands in obvious relation to the curvature of the upper 

 jaw, and more especially to the angle formed by the maxillo- 

 jugal arch with the premaxillary bones. But in carriers, runts, 

 and barbs the singular reflexion of the upper margin of the middle 

 part of the lower jaw (see woodcut No. 25) is not strictly corre- 

 lated with the width or divergence (as may be clearly seen in 

 woodcut No 26) of the premaxillary bones, but with the breadth 

 of the horny and soft parts of the upper mandible, which are 

 always overlapped by the edges of the lower mandible. 



In pouters, the elongation of the body is a selected character, 

 and the ribs, as we have seen, have generally become very broad, 

 with the seventh pair furnished with processes ; the sacral and 

 caudal vertebrae have been augmented in number ; the sternum 

 has likewise increased in length (but not in the depth of the 

 crest) by *4 of an inch more than would follow from the greater 

 bulk of the body in comparison with that of the rock-pigeon. 

 In fantails, the length and number of the caudal vertebrse 

 have increased. Hence, during the gradual progress of variation 

 and selection, the internal bony frame-work and the external 

 shape of the body have been, to a certain extent, modified in 

 a correlated manner. 



Although the wings and tail often vary in length inde- 

 pendently of each other, it is scarcely possible to doubt that 

 they generally tend to become elongated or shortened in 

 correlation. This is well seen in jacobins, and still more 

 plainly in runts, some varieties of which have their wings 

 and tail of great length, whilst others have both very 

 short. With jacobins, the remarkable length of the tail and 



