502 PROF. HOWES AND MR. A. M. DAVTES ON THE [DeC. 4, 



The supernumerary phalanx reaches an altogether special deve- 

 lopment in this family, and all stages in its histological structure are 

 represented {cf. ante). In H. peronii it attains the maximum bulk 

 observed in the entire Anuran order, the total length of its ventral 

 border (fig. 13) exceeding tliat of the terminal phalanx. It is very 

 surprising that, in spite of this, it does not even become hyaline in 

 this species. 



The remaining joints exhibit minor structural differences, but they 

 are all modifications of a condition fully exemphfied in Nototrema 

 (figs. .5, 6). The place of the syndesmodial pad is occupied by a 

 powerful annulus which, although generally dense and fibrous, never 

 becomes converted into hyaline cartilage. Looked at in tangential 

 sections this is seen (fig. 6) to be well defined and in close connection 

 with the synovial capsule; it furnishes articular surfaces for the 

 dorso-lateral portions of the apposed phalanges, and, as seen from 

 this aspect, it bears an astonishing resemblance to the supernumerary 

 phalanx of the same genus {p.s., fig. 7) as viewed in median section. 

 In median longitudinal section it is found to be perforated (fig. 5), 

 mainly for transmission of a strand of elastic tissue {inter-articular 

 ligament, I.), the fibres of which pass, in a somewhat complicated 

 manner, between the ventral borders of the adjacent epiphysial 

 cartilages. When analyzed in detail, this ligament is seen to be 

 largely cellular, and in a conditiou which admits of little doubt 

 that it represents the modified central portion of the original syn- 

 desmosis. 



b. BuFONID^. 



Examined : — 



Pseudoj)hryne bihronii, 23 millim. 



In the genus above named the syndesmodial pad is met with in 

 the distal joints alone. Structurally it is little modified, but it 

 gives the appearance of being subdivided mediallv into two 

 strands which are completely confluent ventro-laterally witli the 

 adjacent epiphysial cartilages and pass ventrally downwards, con- 

 verging, for attachment to the investing tunic. None but feeble 

 traces of the pads are met with in the other joints, and these may 

 be either converted into comparatively unimportant ligaments or 

 reduced to an absolutely vestigial condition. 



KoUiker, in describing the development of the Mammalia, writes ^, 

 on the authority of Henke and Reyher (6), " wandeln sich die 

 Gelenkstellen in ihren ausseren Theilen je liiuger urn so deutlicher 

 in Fasergewebe um, worauf dann in einem gewi!^sen Stadium auch 

 die Gelenkhohle in Form einer engen Spalte erscheint." Ilenke and 

 Reyher's " Zwischenscheiben," in which these changes go on, is 

 identical in the main with our syndesmosis ; and in view of these 

 facts it becomes exceedingly probable that the splitting above re- 

 corded in the Bufonid may represent a persistence, in a slightly 

 modified form, of the initial stage in ibrmation of the typical 

 synovial capsule. 



' Entwicklungsgescli. p. 493. 



