36 PROCEEDINGS OF THE ROYAL SOCIETY OF QUEENSLAND. 



Gadow's argument with reference to the pollex and four- 

 fingered Stegocephah is invahd, for five-fingered forms are 

 known, e.g. Keraterpefon crassum (Zittel, 1902, p. 127) has 

 five digits in both manus and pes. 



Beddard's suggestion regarding the prepollex and pre- 

 hallux as being accessory ossifications cannot apply to the 

 Anura, as such ossifications are not at all typical of the 

 Amphibia. The most characteristic accessory ossification in 

 the TetrajDoda is the pisiform of the carpus — a bone which does 

 not typically appear until the Reptilia are reached. 



In connection Avith this paper, we examined, by means 

 of X rays, the pes of the folloAving Anura : — Adelotis brevis, 

 Hyla ccerulea, H. peronii, Limnodynastes tasmaniensis, Notaden 

 bennetti, and Rana pajiua. From the plate it wiU be observed 

 that there is no trace of the prehallux in H. peronii, A^. ben- 

 netti,' and E. papua, while in the remaining species a prehallux 

 is present lying on the medial side of the pes near the tarso- 

 metatarsal articulation ; in ^. brevis and H. ccerulea it consists 

 of two components and maybe in L. tasmaniensis also, though 

 in regard to the latter the prehallux may prove to be only a 

 sing'e bone. 



In most textbooks it is usually stated that the prehallux 

 of the Anura consists of three or four pieces, of which the 

 proximal member represents the metacarpal, and the remaining 

 segments, the phalanges. 



After carefully considering the data we advance the theory, 

 that the pirepollex and prehallux are the degenerate remnants of 

 the true first metacarpal and metatarsal of the manus and pes 

 respectively. We also regard the frequent occurrence of the 

 prehallux in the Anura as a point of conciliance in favour of 

 our theory, for by the Law of Recapitulation traces of the 

 missing first metatarsal (and metacarpal) might be expected 

 to occur in the Amphibia, which are almost universally con- 

 sidered to be the parent stock of the Tetrapoda. We consider 

 that the pentadactyle condition became stereotyped very 

 early in the history of the Tetrapoda, and that in these early 

 forms all the digits were provided with true metacarpals or 

 metatarsals as the case may be {see Fig. C). It is interesting 

 to note that Bateson (1894) mentions a number of cases of 

 variations in the human hand in which the thumb is provided 

 with three phalanges — a condition which may be regarded as 

 a reversion to the primitive type. Furthermore, it is significant 



