234 H. S. Harrison, 



ditioii he speaks of the ductus as terminating blindly in the perilym- 

 phatic tissue. He thus failed to recognise that the perilymph of the 

 ductus is in a position to transmit vibrations imparted to it from the 

 perilymph of the spatium sacculare, which he overlooked. His suggestion 

 that the spaces are concerned in the conduction of sound is therefore 

 put forward very tentatively, and is based on the probability of vi- 

 brations being conveyed from the pharynx to the perilymphatic system, 

 and not from the fenestra vestibuli. 



In Rana fusca'^) we find that the most extensive portion of 

 the perilymphatic system is, as in the Urodela, the spatium sacculare. 

 Here also its cavity is separated from that of the sacculus by a thin 

 membrane, which embraces practically the whole of the outer lateral 

 wall of the sacculus (fig. 23). From the spatium sacculare the 

 ductus perilymphaticus arises, and pursues essentially the same course 

 as in the lower group. In the Anura generally, the spatium saccu- 

 lare extends further dorsally than is the case in the Urodela, and it 

 is from this upper part, which lies externally (laterally) to the sinus 

 superior utriculi, that the ductus takes its origin. It passes inwards to 

 the foramen perilymphaticum, running below the posterior end of the 

 horizontal canal, above the lagena and pars basilaris, below the pars 

 neglecta. Within the cranial cavity it joins the spatium meningeale. 

 The recessus partis neglectae has relations to the pars neglecta similar 

 to those existing in the Urodela. In the Anura however, only the 

 posterior part of the floor of the pars neglecta constitutes the tym- 

 panal area. The greater part of the wall of the pars basilaris is much 

 thickened by the development of "Spindelknorpel" (fig. 14: sic). This 

 brings into greater prominence the unthickened area in contact with 

 the recessus partis basilaris. It is not always possible to detect more 

 than one layer of cells between perilymph and endolymph, though it 

 is probable two layers are present. The appearance in fig. 14 is due 

 to the fact that the membrane is seen partly in surface view. The 

 tympanal area of the pars basilaris in the Anura is the most con- 

 spicuous example of a definite arrangement for the transmission of vi- 



') Prof. Gaupp informs me tluit lie believes tliis species to be identical with 

 Kana teniporaria. 



