242 H. S. Harrison, 



bates. The parts in this region in both Rana and Bufo show greater 

 evidence of secondary modifications. The recess of the cranial cavity 

 in Bufo, into which the recessus partis basilaris passes, on its way to 

 the fissura metotica, may be explained as the result of a secondary 

 restriction of the fissura metotica. That considerable changes have 

 occurred at this point, and are perhaps still in progress, is proved by 

 the various modifications occurring during development in all the forms 

 — e. g. the temporary formation of the foramen perilymphaticum ac- 

 cessorium in Rana, and the late forward extension of the fissura meto- 

 tica in Pelobates. In every respect the structural arrangements in 

 the larval Pelobates can be deduced from those found in Urodeles 

 with greater simplicity than is the case with Bufo. If we accept the 

 latter as primitive, then the course of events in Pelobates can only 

 be explained as an independent line of progress from the simple Uro- 

 delan condition. Taking all the facts into consideration, I am convinced 

 that the development of Pelobates gives us the key to the difficulties 

 found in the other two forms, and that, in Bufo as well as Eana, the 

 foramen perilymphaticum inferius is an independent aperture which 

 originally opened on to the exterior of the capsule, and that therefore 

 Basse's expression "geteiltes foramen rotundum" is unjustifiable in a 

 double sense. 



The single perilymphatic foramen of Bufo at a latter stage beco- 

 mes divided into the two we find in .the adult. Ontogeny in this case, 

 as in Rana fusca, is not a true record of phylogeny, since there were 

 as we have just seen originally two apertures which are at present in 

 early stages continuous with each other. The dividing bar is formed 

 in such a position that the passage from the recessus partis basilaris 

 to the ductus perilymphaticus is on its medial side; the condition after 

 metamorphosis is therefore similar to that in Rana and Pelobates at 

 the same stage. The chief difierence lies in the fact that in Bufo the 

 recessus partis basilaris still passes into the cavum cranii before ente- 

 ring the fissura metotica. The saccus perilymphaticus appears to be 

 of less importance than in the other two cases, if we may judge from 

 its relationships to the perichondi'ium of the outer surface of the chon- 

 drocranium. 



