"QCIKTOCUBITALISM " IN THE WIXQ OF BIRDS. 229 



allies o£ pigeons. But, unless one knew tlie size of the ancestral 

 pigeon, it would be impossible to draw any strong argument 

 from size ; in a general way, it may be said that a large number 

 of very primitive birds are large, while many of the most 

 extremely specialized forms are minute, as among the Passeres. 



Theory of the Origin of the Diastataxic Condition. 



As my general alignment involves tlie supposition that eutaty 

 is a derivative of diastataxy, it may be worth while to advance a 

 speculation as to the reason why there should have occurred a 

 gap in the wing of birds, and why that gap should be in a fixed 

 position. 



The first point upon which I wish to insist is that the quills 

 are not different in kind but only in degree of development, 

 as compared with other feathers. From a uniform covering 

 certain individual feathers have become enlarged to serve the 

 purpose of flight, instead of the more primitive purpose of pro- 

 tection. Precisely in the same way in the case of Elasmobranch 

 fishes, from the general protective covering of toothed scales all 

 over the body, certain individual scales or rows of scales become 

 specially enlarged, and form series of exaggerated teeth, either 

 at special points, as in the case of the dorsal spines of the spiny 

 dogfish, or in rows along the dorsal surface of the tail, as in 

 some skates, or in one or more rows along the edges of the jaws 

 where the external skin folds in to form the stomatodeal lining 

 of the mouth, when such rows are developed. It may fairly be 

 taken for granted that birds had feathers or feather-like scales 

 before they had quills, as quills are simply exaggerated feathers, 

 and that the quills are simply rows of extremely developed 

 feathers. 



When rows of large structures follow body contours, as in the 

 case of true teeth, caudal spines, or quills, the series appear to 

 possess a longitudinal coherence and integrity which may be 

 fallacious. If pins be placed vertically on diagonal lines crossing 

 the surface of one of the irregular stuffed cushions familiar on 

 old fashioned toilet-tables, the pins which happen to lie along 

 the contour edge appear to have a special symmetry with regard 

 to these contours, and could we imagine these pins to increase 

 in length because of their position, it would be difiicult to avoid 

 supposing that the long contour pins were planted specially in a 



