1888.] MORPHOLOGY OF SUPERNUMERARY PHALANGES. .'>03 



c. Cystignathid^e. 



Examined : — 



Hylodes martinicensisy 24 millini. 

 Lymnodynastes tasmaniensis, 4Q millim. 



Eylodes is especially interesting in the fact that, while possessed 

 of expanded digital extremities, it shows no trace of a distinct 

 supernumerary phalanx. 



_ This combination lends additional support to Boulenger's deduc- 

 tion, that the phalanx in question bears no definite relation to the 

 platydactyle type. 



d. RanidjE. 



Examined : — 



Ixalvs schmardanus (juv.), 1 1 millim. 



3Iegalixalus madagascarensis, 32 millim. 



Rana arvalis (juv.), 27 miUim. 



Rana ?naxmus, 90 milUm. 



Rana temporaria, adult, and 14 millim. (tail just gone). 



Rappia mnrmorata, 27 millim. 



Rhacop/iorus eques, 42 millim. 



We have already dwelt in sufficient detail (p. 500) upon the super- 

 numerary phalanx in this family ; all stages are represented, from 

 the completely ossified one (Peters and Boulenger, 11. cc.) to that of a 

 degraded syndesmosis (Rana temporaria, Plate XXV. fig. 1 6). 



The syndesmoses of the remaining joints are, in the adults of all 

 species exammed by us, with the exception of liana arvalis, modified 

 to form, more or less efficiently, the ligament and annulus apparatus 

 of the Hylidce. B. arvalis would appear to be less conspicuously 

 modified than the other members of this family, in that the ordinary 

 syndesmosis is retained for both the terminal and penultimate joints. 

 We have, however, only examined a young specimen. 



e. DlSCOGLOSSID^. 



Examined : — 



Alytes obstetricans, 30 millim. 

 Bombinaior bombinus, 4 1 millim. 

 Discoylossus pictus, 62 millim. 



In all three genera the phalanges are united, in both fore and hind 

 hmbs, by syndesmoses. Those of the terminal joints are in no 

 way in excess of at least that of the penultimate one. 



These syndesmoses are (cf. fig. U), for the most part, closely 

 bound down to the faces of the apposed epiphysial cartilages. In 

 the hind limb o{ Bombi?iator there is a suggestion of a loss of direct 

 connection between the two, with the development of a shallow 

 articular cavity, while, in the proximal joint of the same, the syn- 

 desmosis is in a loose and somewhat degenerate condition. These 

 modifications are, however, exceedingly feeble and unimportant, and 

 making all due allowance for them, the DiscoglossidcB are seen to 



