S6 I'liOCEKDINGS OF THE 



jire forms wliidi jump as powerfully or swim as sti'onp;ly in the 

 proccelous Arcit'eru ;is in tlie diplasiocoelous Firmisternia, just as 

 tiiere are crawling, burrowing, or arboreal forms in each group. 

 A very similar statement might be made, comparing the opistho- 

 eoelous forms with the proc-oelons. Indeed, from the stand])oint 

 of function it would appear to be a matter of little import 

 whether the intervertebral spheres become attached to the 

 anterior or the posterior face of an adjoining vertebra, and, as 

 Gadow has remarked, we might imagine that the union of these 

 intervertebral spheres with centra before or behind would be a 

 mere matter of chance. Occasionally, as a very rare variation in 

 the Pelobatidae, the intervertebral spheres altogether fail to unite 

 with the centra and the amphicffilous (ancestral) condition of the 

 embryo persists into maturity {F. cultri_pes^ Duges, '34; F.fuscus, 

 IStannius, '54). 



When, therefore, such a comparatively trivial character as this 

 is found to occur over a wide range of genera in which there is 

 an almost complete absence of individual variation*, we are 

 surely justified in attacliing to it considerable value as a syste- 

 matic character, since its occurrence is to be explained only upon 

 the ground of community of descent. 



A division of the Phaueroglossa, based upon this character, 

 gives us, moreover, a grouping which (while it corresponds faii'ly 

 closely with that at present accepted as representing a natural 

 classification) separates the Discoglossidse a little more widely 

 from the remaining Arciferous forms. 



The breakiug-up of the Arcifera has already been proposed by 

 ►^tejneger ("07). That author divided the Anura into three sub- 

 orders : — Aslossa, Costata (the Discoglossidaj), and Linguata (the 

 remaining Phaueroglossa). 



Lataste (79) and Blanchard ('85) went still further and pro- 

 posed the inclusion of both the Aglossa and the Discoglossidte in 

 the fame Sub-oider (Opisthoto?la). 



The change which I propose in our present classification con- 

 sists simply in the grouping of the fan)ilies of the Plianeroglossa 

 into four larger divisions (tribes) in place of the two series, 

 Arcifera and Firmisternia, at present recognized. 



PHANEROGLOSSA. 



I. Opisthoccela. Sacral vertebra biconvex, free from coccyx ; 

 pre-sacral vertebrae eight, oi)isthocielous ; with ribs. 

 Includes but one family, the Discoglossidse. 

 11. Anomoc(ELA. Sacral vertebra anl<ylosed to coccyx or 

 articulating therewith by a single condyle ; vertebra? 

 procoelous (rarely opisthoccelous) ; without ribs. 

 Includes a single family, the Pelobatida^. 



* Even in tlie very variable Pelobatirla; only two cases o( a now a hits centra 

 (apart from the anij)hica'lj above uieiitio:ie(l) have been recorded in llie pro- 

 cu>lous species Tiicse were described by Stannius ('54) in Jrlolui/tt^ fiisciis and 

 by Boiilcuger ('82) in Aenaphriji nwiiiicoht. 



