NATURAL SCIENCES OF PHILADELPHIA. 247 



group as sufficiently defined to be so regarded, except Ranula, as the adults of 

 some species appear not to be constant in possessing them. Thusaverj- large 

 Rana catesbyana sometimes exhibits prefrontals in contact en the median 

 line, while it is difficult to say whether R. a r e o 1 a t a of North America is of 

 the Nearctic type so much as of the ^Ethiopian. Nevertheless the groups are 

 generally quite geographically restricted. 



3. A similar relation exists between the genera Hyperolius, Staurois and 

 Heteroglossa in respect to the prefrontal bones and the separation of the outer 

 metatarsi, and — 



4. Between Ixalus, Rhacophorus and Polypedates also, in reference to vo- 

 merine teeth, bifurcation of last phalange, and dermoossification of the cra- 

 nium. 



5. When the larvje of certain species of Spelerpes possess brauchife, thev 

 also lack one digit of the hind foot, also the maxillary, nasal and prefrontal 

 bones, and exhibit a broad continuous palatopterygoid arch, in close contact 

 with the parasphenoid. The prootic is separated from the exoccipital by a 

 membranous space, and the exoccipitals themselves are not yet united above 

 the foramen magnum. There is at the same time a series of splenial teeth. 

 Both ceratohyals are confluent, the posterior is present, and there are but 

 three superior hyoid arches. After they lose the brancliia?, the hinder foot, 

 which has four toes only lor a time, gradually adds another at first rudimental 

 digit, in the Mexican species ; in most North American species the fifth digit ap- 

 pears at an early larval stage. Five digits are finally present in all Spelerpes. 



We have thus four combinations of the above characters, at different periods 

 of the life history of certain (but not of all) of the species of Spelerpes, There 

 exist four permanent series of species or genera, equivalent to these stages. 

 The well-known " perennibranchiate " Nectnrns is nearly identical with the 

 first, Batrachoseps with the second, the half-toed Spelerpes with the third, and 

 the typical Spelerpes is the last. 



In one character of generic value only, do I find that Necturus differs from 

 the early larval Spelerpes. It closes the premaxillary fontanelle with which 

 it commences, by an approximation of the premaxillary spines, but not by a 

 sutural union, as takes place in Amblystoma. It thus, in this one point, ad- 

 vances a stage beyond the condition to which Spelerpes attains, though it 

 maybe a question whether such a closure without union should not be classed 

 among the specific characters by which N. maculatus differs from the young 

 of the various Spelerpes, as they do from each other. Characters of the latter 

 kind are the following: in N. maculatus the frontals are more deeply 

 emarginate behind; it has little or no ala on the inferior keel of the caudal 

 vertebne, which is prominent in Spelerpes larvoe. 



It may be that the parallelism in the case of Spelerpes is inexact by one 

 character, and that a strictly developmental one ; or it may be regarded other- 

 wise. 



6. It is well known that the Cervidre of the old world develope a liasal snag 

 of the antler (see Cuvier, Ossem. Fossiles ; Gray, Catal. Brit. Mus.) at the third 

 year; a majority of those of the New World (genera Cariacus, Subulo) never 

 develop it except in "abnormal" cases in the most vigorous maturitj- of the 

 most northern Cariacus (C. virginianus); while the South American 

 Subulo retains to adult age the simple horn 'of the second year of Ceivus. 



Among the higher CervidjB, Rusa and Axis never assume characters beyond 

 an equivalent -of the fourth year of Cervus. In Damathe characters are on the 

 other hand assumed more rapidly than in Cervus, its third year corrosjionding 

 to the fourth of the latter, and the development in after years of a broad plate 

 of bone, with points, being substituted for the addition of the corresponding 

 snags, thus commencing another series. 



Returning to the American deer, we have Blastocerus, whose antlers are 

 identical with those of the fourth year of Cariacus. 



1868.] 



