1896. THE WING OF ARCH^OPTERYX. 263 



of the base of the claw of digit II and the proximal end of the pollex 

 of the left wing. 



It is certainly open to question whether the metacarpal bones were 

 " free," as has hitherto been stated ; and this is a point which is 

 surely of considerable importance in the present discussion. As to 

 this we have but the evidence of the Berlin fossil. Metacarpals I, II 

 are so closely approximated that only a very faint trace of a suture can 

 be seen. The proximal end of digit III is a little damaged, but it is 

 nevertheless seen to be closely fitted to the base of Mc. II, Distally 

 the II and III metacarpals seem to be sHghtly separated one from 

 another. I am convinced, however, that this is but an apparent, not 

 a real, separation, and that the two bones are really joined by an 

 osseous bridge, which, being somewhat concave, has become filled up 

 with rock. 



If my interpretation of the manus of A ycJusoptcryx proves to be 

 correct, then it will, I think, be no longer possible to entertain the 

 hypothesis that these three digits were used for climbing. Thus it 

 will follow that they (or at least the II and III) supported the 

 remiges, so that we shall be able to return once again to our old love 

 the tridactyle avian manus. 



I have already urged (4) that if the digits I, II, III were used for 

 climbing, and if the remiges were supported by two additional digits, 

 then the manus of Arclueopteryx must have undergone an extreme 

 specialisation in two directions, to fit it for two different modes of 

 locomotion — flight and climbing. So far this objection has not been 

 met to my satisfaction. 



The fact that the digits of the Berlin fossil appear to lie on the 

 feathers of the wing, and not under them, as we should have expected, 

 has been held as proof that they could not have been concerned in the 

 support of these remiges. This is by no means so serious an objec- 

 tion as it appears. In the London fossil, the author of the above 

 objection assures us, the remiges are still in their normal position with 

 regard to the bones which supported them. Now since in both fossils 

 the bones of the digits lie above and not below the level of the 

 remiges, it is hard to see how the same relations can be made 10 

 prove both ways. 



Professor Dames, I believe, thinks it possible that what is pre- 

 served in the Berlin fossil is rather a combination of impressions of 

 the feathers of both dorsal and ventral surfaces of the wing, the 

 feathers themselves having quite disappeared. The remarkably 

 fine impression of the primary remiges, however, would seem to 

 oppose this view. The feathers are so perfectly preserved that 

 even the overlap can be distinctly made out. This, as I suggested, 

 and as Dr. Hurst afterwards maintained, is "distal" — i.e., the 

 outer edge of every remex is free and overlaps the inner edge of 

 the one next in front. This can be well made out in the fossil, 

 where, as photographs first suggested to me, the mud has drifted 



