500 PROF. G B. HOWES ON AN UNRECOGNIZED FEATUHK [JlltlC 



d. Epiglottis and epilaryngeal folds both absent. 

 Tivfo typhonhis. $ 



Cortiufer vitianus. $ . 

 *Hijla meridionalis. 6,2- 



■ venu/osa. c? • 



*Iiylodes martinicensis. 5 j'l^'- 



Nannophrys cei/lonensis. cJ • 



Phrynisciis Itevis. 5 . 



rhyUomedv.sa bicolor. cS juv. 



*Runa esculenta. d , $ • 

 * temporaria. c? , ? • 



temporalis. J . 



Rappia murmoratn . $ . 



Rhacophorus maculatvs. S . 



Rhinoderma darivini. d . 



Xenophrys monfAcola. § . 



Perusal of the above brings into prominence a wide range of in- 

 dividual and specific variations. The former are, however, less 

 striking than appears at first sight, and little more remarkable than 

 those of the Amphibian manus and pes, or of the Batoid intestinal 

 valve' ; the latter find a near parallel in the modifications of the larynx 

 in certain fruit-eating Bats recorded by Dobson". That observer 

 has obtained a satisfactory explanation of the phenomena upon purely 

 physiological grounds, and the analogy suggests that an investiga- 

 tion into the habits of the living Annra inider consideration might 

 furnish a similar clue. If, as the analogy to Chiroleptes would 

 suggest, the folds in question are associated with the development of 

 vocal sacs, it is difficult to reconcile their absence in Hylu meri- 

 dionalis and //. venu/osa, with their presence in JI. dolichopsis and 

 Bombinator ; while the combination met with in Heleioporus would 

 appear to negative the suj)positioa. It may be held that the struc- 

 tures are developed periudically or with sexual maturation ; but, 

 if so, it becomes hard to account for their insignificance in the male 

 of Leptodacfylus (fig. 2), in which the accessory spurs had attained a 

 considerable development. The facts adduced in Chiroleptes aus- 

 tralis, while they appear to favour this b.^lief, are, in themselves, 

 capable of a different interpretation, for that the differences between 

 the two males which I have examined are due to age is a surmise. 

 The facts are very puzzling ; but who is to say that, in the most 

 marked cases recorded, we may not be dealing with a character of 

 taxonomic value, fit at least to rank with those customarily relied 

 upon*. 



Apart from the above considerations, examination of the figures 

 shows most conclusively that, in the specimens which I have described, 

 a gradational modification can be traced. Chiroleptes (fig. 6) and 

 Rana (fig. 1 6) stand at opposite ends of a series, the intermediate 

 steps in which are furnished by Ceratophrys, Leptodactylus, and 

 Rana pipiens, in succession. The question therefore arises as to 

 which of the two first-named typifies the more primitive arrangement. 



■ T. J. ParL-er, T. Z. S. 1880, pp. 49-61. 



« P. Z. S. 1881, pp. fiS5-93. 



» Cf. remarks by Huxley en the "Taxonomy of the Canidse" (P. Z. S. 1880, 

 p. 28(i), and by Herdnian, "On the Specific and Local Yariations in the 

 Tuuiciita (First Eejiort of the Liverpool Marine Eiological Committee, 188(i, 

 pp. 3.^>5, ?,m). 



