April, 1895. THE HABITS OF ARCHMOPTERYX. 245 



the question of Archaopteryx, except in so far as it shows that so long 

 as the digits are useful for climbing, the development of quills upon 

 them is delayed ; and that points to a conclusion directly opposite to 

 his views. .He throughout " assumes " first one thing and then 

 another which he knows to be in direct antagonism with my views. 

 For instance, he " assumes " (vol. v., p. 352) that the recapitulation 

 theory in its crudest form is true ( ! ), that the ostrich is a suitable 

 species to be taken as type of " existing birds " in the matter of wing- 

 structure (p. 354) ; that some such arrangement as that he has de- 

 scribed from existing birds obtained in Archaopteryx (p. 356). This 

 last assumption might be all very well under other circumstances, but 

 in an argument raised in refutation of the contrary view it is not easy 

 to see what assumption could be less allowable. He assumes, with- 

 out offering a jot of evidence for it, that in the matter of the wing 

 Opisthocomus is "primitive" — therein begging the whole question 

 (P- 359)- On the same page he assumes that the common fowl is a 

 " more recently evolved form," and after stating that it is " probable 

 that the peculiar habits of the nestling [Opisthocomus'] may be the 

 survival of an order of things handed down from the very dawn of 

 avian development," he regards that as sufficient, and proceeds to 

 argue as if what he had stated, without a scrap of evidence, to 

 be probable, were in reality established fact. As to the powers of 

 flight of Arclueoptcryx, upon which (p. 359) he lays very great stress, 

 he expresses an opinion, but gives no grounds which would justify 

 our acceptance of it. He ignores the fact that the relative size of 

 wing and of body in equally good fliers is very different in the cases 

 of large and of small birds respectively. 



Mr. Pycraft's contention as to weak-jointedness of the phalanges 

 of the wing-digits is merely due to his having overlooked the fact 

 that resistance to torsional and not flexional stress was the point I 

 referred to in a previous paper. He says, " The phalanges of living 

 birds seem to be scarcely, if at all, more firmly bound together than 

 were those of Archaopteryx " (p. 445). My argument had reference, 

 not to the movement of one phalanx upon another in the same digit 

 (i.e., flexion), but to the movement of the metacarpal V on metacarpal 

 IV, or of metacarpals III on II, if we adopt his own views. I need 

 hardly urge that since III and II are free in Archaopteryx, and what 

 he regards as their equivalents are ankylosed and wholly immoveable 

 relatively to each other in modern birds, his argument fails to affect 

 my conclusion. As to the phalanges, using my own nomenclature, 

 the extraordinary breadth and flatness of the first phalanx of IV in 

 the plane of that of V, while they do not limit flexion, prevent their 

 independent movement and give that torsional rigidity on which I 

 insisted, while leaving unrestricted that flexional mobility upon which 

 the young Opisthocomus depends in climbing. The joint, however, is 

 stiffened very considerably in most birds (if not in all), but it is not a 

 point with any direct bearing on the question under consideration. 



