33 ANNALS NEW YORK ACADEMY OP SCIENGE,'4 



continues, "we may reasonably suppose the fore limbs were used for 

 grasping at the end of the leap. The use of the fore limb for this work 

 would naturally throw more work upon the inner digits — 1-3 — so that 

 the work of selection would rapidly tend to the increased development of 

 these, and the gradual decrease of the two outer and now useless mem- 

 bers. Correlated with this trend in the evolution, the axillary mem- 

 brane — the skin between the inner border of the arm and the body — 

 became drawn out into a fold, while a similar fold came to extend from 

 the shoulder to the wrist, as the fore limb, in adaptation to this ncAv func- 

 tion, became more and more Hexed. While the fingers, upon which safety 

 now depended, were increasing in length, and growing more and more 

 efficient, they were, at the same time, losing the power of lateral exten- 

 sion and becoming more and more flexed upon the fore-arm. And the 

 growth in this direction was probably accompanied by the development 

 of connective tissue and membrane along the hinder, post-axial border 

 of the whole limb, tending to increase the breadth of the limb when 

 extended pre|)aratory to parachuting through space from one tree to 

 another, long claws being used to effect a hold at the end of the leap. 



"The hind limbs, though to a less extent, were also affected by the 

 leaping motion, resulting in the reduction of the toes to four, and the 

 lengthening and approximation of the metatarsals 2-4 to form a "^cannon' 

 bone. 



"The body clothing at this time was probably scale-like, the scales 

 being of relatively large size and probably having a medium ridge, or 

 keel, recalling the keeled scales of many living reptiles. Those covering 

 the incipient wing, growing longer, would still retain their original over- 

 lapping arrangement, and hence those along the hinder border of the 

 wing would, in their arrangement, simulate in appearance and function 

 the quill feathers of their later descendants. As by selection their length 

 i" creased, so also they probably became fimbriated and more and more 

 tificient in the work of carrying the body through space. 



"There is less of imagination than might be supposed in this attempt 

 at reconstructing the primitive feather, inasmuch as there is a stage in 

 the development of the highly complex feather of to-day which may well 

 represent the first stage in this process of evolution. Creatures such as 

 are here conjured up would bear a somewhat close resemblance to Arclice- 

 opteryx, and it is contended that the discovery of earlier phases of avian 

 development, phases preceding Arcliwopteryx, will show that this fore- 

 cast was well founded. But in Arcliceopteryx, it is to be noted, the 

 feathers differ in no way from the most jDerf ectly developed feathers 

 known to us." 



