82 GENERAL EVOLUTION. 



ally seasonal. Yet this is not always so, and, like tlie other char- 

 acters, has most likely had its period of shifting. Compare differ- 

 ence of time of development, for instance, of the frontal and pre- 

 frontal bones in Figs. 3 and 6. The comparison of the adult 

 stages of the less developed genera, at the tops of the columns, 

 with the larval conditions of those more fully developed, may be 

 traced in the absence of characters which appear in the latter. I 

 have convinced myself of the accuracy of the above relations by 

 the examination of many skeletons and wet preparations of adults 

 and larvae. 



Tlie diagrams* are representations of nature, and not ideal 

 sketches. It is to be noted as remarkable that the advance 

 throughout so many diverse groups is in the same direction, viz., 

 to complete or excessive ossification of the cranium ; and this 

 identity of progress might be readily shown by adding other char- 

 acters, were it not that the tables would become too complex for 

 convenience. 



Has any such transition from genus to genus ever been seen to 

 occur 9 



It must of course take place during the life of the individuals 

 of a species, and probably at different times during the lives of 

 different individuals, dependent on their relative vigor. In our 

 view, ordinary metamorphosis is such a change, and we have 

 stated its bearing in this form, that " every character distinguish- 

 ing suborders, families, and genera is to be found among the indi- 

 viduals of some species, living or extinct, to mark new varieties 

 or stages of growth." 



a. The Developmental Relation of Qeneric to Specific Characters. 



For the relation of the law of retardation and acceleration to 

 specific characters we will look to development again. While the 



* Notes on the diagrams : Fig. 4. Polypedates is here restricted to P. maculatus 

 and P. quadrilineatus. The other species are referred to Rhacophorus, which has 

 not hitherto rested on any proper basis ; the asserted character — the palmation of 

 the hands — being one quite graduated from species to species among Hylte. 

 Chiromantis, Peters, is referred to the same, as its character is not strongly 

 marked and is visible in other species. For similar reasons Lcpfomantis is referred 

 to Ixalus. 



Fig. 6. In each of series ii and iii, two series are mingled for the sake of com- 

 paring the structures of the prcfontal bones. Thus Heteroglossa, Staurois, Ilylo- 

 rana and Trypheropsis are one series, and Hyperolius and Hylambates members of 

 another. 



