120 THE ITATUEAL HISTOET EEYIEW. 



Finally, the nature of the supports of the terminal digital dilata- 

 tions, which adapt the Tree-Frogs to their mode of life, may be 

 compared. 



I. Claw-like, with globular t)ase . Htlid^. 



Leptopelis. 

 II. Simple, obtuse-depressed at tip . Eanidje, 1. aa & III. a. 

 III. With a terminal transverse limb . EANiniE, Hylarana et aif. 



CaloJiyla. 

 BracJiymerus. 

 Sylodes, 



IV. Bifurcate Batracliyla. 



Dendrohates. 

 Polypedates. 

 Rhacophorus. 



A glance at what precedes will show that there is no complete- 

 ness of generic diagnosis anywhere attempted. My object having 

 been to point out the importance of characters hitherto much over- 

 looked, I have dwelt but lightly upon those now sufficiently well 

 known, especially through the labours of Giinther, in whose memoirs* 

 will be found explained also the relations between such and the geo- 

 graphical distribution of the species. 



The arrangement and definitions of the higher groups differ con- 

 siderably from those hitherto adopted. There are two of the four 

 families of Dumeril and Bibron accepted as suborders ; but those 

 authors had not completely investigated many important types, at 

 the time their system was proposed. A complete and practically 

 useful system is that of Giinther. I have already f demurred to the 

 recognition of the Tree-Frogs as a natural division, accepted by this 

 author as well as by the preceding ; and while agreeing with him in 

 attaching less value to the condition of the cavum tympani than did 

 Miiller, I find it even less frequently characteristic of otherwise ho- 

 mogeneous groups of genera. Thus, while accepting some of his 

 families, others are rejected. 



My thanks are due to Prof. Dumeril and Dr. Giinther for the 

 great advantages I have enjoyed from the examination of the speci- 

 mens under their respective charges. 



* Proc. Zool. Soc, 1858, pp. 339, 390. 



t Proc. Acad. Nat. Sci. Phil. 1S63, p. 50. Similar conclusions have been 

 aiTived at by Prof. Peters, Monatsber. Preuss. Acad. 1864, p. 455. 



