igS PRIMARY FACTORS OF ORGANIC EVOLUTION. 



fonidae) we have a group which is imperfect in two 

 points instead of one. 



The genera of these salientian suborders exhibited 

 on a preceding page as forming indentical series in five 

 different families (pp. 66-67) are related to each other 

 as developmental stages in the history of the genera 

 that attain the extreme development on each line. 

 For example we select the family Hylidae of which 

 the terminal genus is Trachycephalus. Nearly allied 



Fig. 51.— Shoulder girdles of Anura ; a, of the arciferous type {Phyllomcdusa 

 bicolor) ; b, Rana tempararia, tadpole with budding limbs; r, do., adult, firmis- 

 ternial type ; b and c from Parker. 



to it is the genus Osteocephalus, which differs in the 

 normal exostosis of the cranium not involving the 

 derm, as it does in the former. Close to this is Scy- 

 topis, where the fully ossified cranium is not covered 

 by an exostosis. Next below Scytopis is Hyla, where 

 the upper surface of the cranium is not ossified at all, 

 but is a membranous roof over a great fontanelle. Still 

 more imperfect is Hylella, which differs from Hyla in 

 the absence of vomerine teeth. Now, the genus Trachy- 

 cephalus, after losing its tail and branchiae, possesses 



