LI^'NEAN SOCIETY OF LONDOIS". 87 



III. Proc(ela. Sacral vertebra free, articulating with the coccyx 



by a double condyle; vertebrae uniformly procoelous. 

 Includes the Bufonidae, Hylidse, and Cystignathidse. 



IV. DiPLASioc(ET.A (^Firmtsternia). Sacral vertebra bicon- 



vex, eighth vertebra biconcave; the first seven vertebrae 

 proccelous. 



Includes the Eanida? and Eugystomatidee. 



(i.) This arrangement involves, as will be seen, the subdivision 

 of the Arciferous forms into three groups. As already pointed out, 

 it has long been recognized that the Discoglossidae are somewhat 

 more widely removed than the existing classification would indicate, 

 (ii.) Concerning the Pelobatidae, Gadow remarks that they can 

 be separated from the Bufonidae only by the dentition. Never- 

 theless he advocates that they shall be retained as a distinct 

 family. In the condition of their vertebral articulation, however, 

 the Pelobatidae apparently possess a distinctive character which 

 amply justifies their separation from the Bufonidae. That both 

 procoelous and opisthocoelous vertebrae occur within a single 

 genus {Megalophrys) has been considered as sufficient justifica- 

 tion for altogether discarding this character for systematic 

 purposes. As a matter of fact, however, the whole family 

 appears to be strictly procoelous apart from this single genus. 

 As a family, moreover, it is notable for the extremely frequent 

 occurrence of variations in the vertebral column. Thus, Adolphi 

 found in Pelobates fuscus that abnormalities in the vertebral column 

 occurred in more than 23 per cent, of the specimens examined ! 

 As mentioned above, the two examples recorded in the Anura, in 

 which the aniphiccelous (embryonic) condition of the centra is 

 retained in mature animals, were both examples of Pelohates. 

 Nevertheless but one example (that recorded by Stannius) ap- 

 pears to be known of an irregularly abnormal condition of the 

 centra in this genus. 



In MegaJojihnjs, however, variation seems frequently to extend 

 to the centrum. In Megalnphrys parva [Xenojyhrys monticola) 

 Boulenger has described an anomalous arrangement of the 

 vertebral centra comparable to the case of P. ftiscus described 

 by Stannius. Moreover, 31. montana (the type species) was 

 originally described by Cope as having opisthocoelous centra, and 

 this type of vertebra has since been recorded in two other species 

 (if. nasuta, M. longipes). It is now known that the procoelous 

 condition is, even in these three species, at least equally frequent, 

 while in the remaining eleven species, hitherto described, tlie 

 vertebrae appear to be uniformly procoelous ! Thus there is little 

 reason to regard the occurrence ot' the opisthocoelous condition in 

 a few specimens of these three species as other than unusually 

 fi'equent cases of individual variation *. 



* Concerning Asferophrt/s little seems to be recorded. Like Meyalophrys, it 

 has been described as opisthoccelous, but its tongue is said to be entire and 

 adherent, and Blanchard assigned it to tlie Discoglossidsp. 



