498 PROF. HOWES AND MR. A. M DAVIES ON THE [DeC. 4, 



like a conspicuous structure is interposed between the terminal and 

 penultimate phalanges) we meet (ex. the common Frog, Plate XXV. 

 fig. 1 6) with a pad of loose fibrous tissue, which is closely bound down 

 to the applied epiphysial surfaces, such as would appear to have 

 nothing to do with the undoubted phalanx under consideration. In 

 Nototrema, however (fig. 7), the pad, while never truly hyaline, has 

 all the fundamental relations described for the phalanx in the culmi- 

 nating term of the series ; and in Hyla j)eronii (fig. 13), while histo- 

 logically identical with that of Nototrema, it has the y)roportions, 

 detailed shape, and relationships seen in H. arborea. Hyla ccerulea 

 bridges over the gap histologically between Nototrema and Hyla ar- 

 borea ; and when to this it is added tiiat, in llioseiJa/nWfP the adults 

 of which bear the supernumerary phalanx, the same early becomes 

 hyaline as in Hyla arborea, it is clear that all stages between the two 

 extremes afore described are forthcoming. 



Minor modifications are met with, but these may, together with 

 the consideration of lesser detail, preferably be dealt with later. 



Seeing that, upon a purely structural analysis, the fibrous pad of 

 the Common Frog must be looked upon as the homologue of the 

 skeletal supernumerary jihalanx of the higher Ranidce and Hylidce, 

 we are next led to inquire what, if any, structural community the 

 two may possess in the embryo. On examination of the larvae oi Hyla 

 arborea and Rana temporaria, at a stage at which the hind limbs are 

 becoming differentiated, it is found that the places of the phalanx of 

 the former, and of the pad in the latter, are alike occupied by a 

 fibrous mass which is largely cellular and loosely interposed between 

 the applied epiphysial extremities. It is well known that, in most 

 Urodeles, there are interposed between the corresponding parts of the 

 limb-skeleton filjrous masses which take the place of the synovial 

 capsules of the higher Vertebrata ; and these are found, on examina- 

 tion, to be indistinguishable from those above described. They have 

 long been termed by Hyrtl (8) " syndesmoses." 



The foregoing is not all. Hyrtl, in describing the manus of 

 Salamandra maculosa, writes (p. Gl): "phalanges inter se, et cum 

 ossibus metacarpi, textu fibroso conjunguutur." AVe find that these 

 syndesmoses are structurally identical with the supernumerary 

 phalanx of the Anura in its least modified condition {cf. figs. 1 1, 12), 

 and, seeing that in many Anura similar pads are interposed between 

 the proximal phalanges themselves, it follows that the structures in 

 question in them are not in any sense to be regarded as peculiar to 

 the terminal segments of the digits, and that the supernumerary 

 phalanx would appear to be a specialized counterpart of tlie inter- 

 phalangeal syndesmosis. In other words, may not the community 

 of structure between the developing supernumerary phalanx and the 

 syndesmosial pad be indicative of a community of origin? Should 

 this be the case, there would open up a new and promising departure 

 for the reconsideration of the questions of origin and morphology of 

 supernumerary phalanges in general. To these we shall return. 



