238 H. S. Harrison, 



tlie internal ventral angle of the capsule, this groove for a very short 

 distance even becoming a canal (fig. 19). Here then we have a con- 

 dition resembling in essential features that found in the Urodela, the 

 most important difference lying in the fact that the greatly enlarged 

 recessus partis basilaris has acquired an independent outlet, and has 

 therefore entered upon new and direct relationships with the soft parts 

 surrounding the skull. The membrane forming the ventral boundary 

 of the Saccus and recessus is probably at this stage of as great functio- 

 nal importance as that closing the fenestra vestibuli. It is indeed pos- 

 sibly of greater importance, if we may trust the evidence afforded by 

 a structural feature shown in fig. 21 {st), and discussed in the sequel. 

 We should naturally anticipate that, since Pelobates fuscus is a 

 lower type than Rana fusca, the former would exhibit the more primi- 

 tive features, if no secondary modifications had interfered with the 

 course of development. This consideration, taken in conjunction with 

 facts derived from the study of the Urodela, would lead us to con- 

 clude that the arrangement of parts just seen is less modified than 

 that found in R fusca, and that therefore the foramen perilymphati- 

 cum inferius is a newly acquired aperture, independent of the foramen 

 perilymphaticum superius, and that the recessus partis basilaris origi- 

 nally communicated with the ductus perilymphaticus within the audi- 

 tory capsule. This conclusion is strengthened by the course of deve- 

 lopment in Pelobates itself. From the condition just described, the 

 parts come to assume relationships almost identical with those found 

 in R, fusca. In a young Pelobates, just after metamorphosis (length 

 40 mm) the saccus perilymphaticus lies within the anterior end of the 

 fissura metotica, and the ductus reuniens is no longer within the audi- 

 tory capsule (fig. 22). There has in fact been an extension forwards 

 of the flssui-a metotica, and in addition the nairow bar of cartilage 

 formerly lying between the ductus reuniens and the cavum cranii has 

 disappeared, giving place to another which separates the ductus reu- 

 niens from the cavum perilymphaticum. It therefore appears at this 

 stage as though the recessus partis basilaris (or saccus perilympha- 

 ticus) sent a short canal into the ci-anial cavity, through the anterior 

 end of the fissura metotica, to join the spatiun» meningeale. What is 



