244 H. S. Harrison, 



In both forms, during, or perhaps a little before metamorphosis, 

 this peculiar connection between lung and perilymphatic system dis- 

 appears, so that whatever its physiological significance may be in the 

 larva, in the metamorphosed animal the function is either fulfilled by 

 other structural arrangements, or is one which disappears in the change 

 from a purely aquatic to a chiefly terrestrial mode of life. I would 

 venture to suggest that the arrangement may be compared with the 

 connection of the air bladder with the auditory capsule which is found 

 in some fishes (Weberian ossicles). At any rate there is obviously 

 here a very efficient apparatus for passing on changes of pressure 

 within the lung, along the dense fibrous column to the membrane of 

 the Saccus perilymphaticus, and thence by the recessus partis basi- 

 laris to the interior of the pars basilaris itself. It seems probable that 

 the function of the apparatus is intimately connected with the aquatic 

 mode of life. The lungs at this stage are functional in part as swim- 

 bladders, and the variations of pressure within their cavities is no 

 doubt partly dependent on the hydrostatic pressure.^) 



The phylogenetic origin of the fibrous strand, in its definite form, 

 I should attribute to the fact that the recessus partis basilaris passed 

 out of the auditory capsule at such a point that it lay directly above 

 the line of attachment of the ligament suspending the lung to the base 

 of the skull. The perilymphatic system thereby accidentally acquired a 

 connection with the lung, and it is to a gradual increase in the in- 

 timacy and functional efficiency of this connection, that the existing 

 arrangements in the larva are to be attributed. 



With regard to the portion of the perilymphatic system, to which 

 Eetzius gave the name of saccus fenestrae ovalis, I have little to add. 

 It is sufficient here to mention that it is a part of the spatium sac- 

 culare having much the same relations to the anterior border of the 

 fenestra vestibuli, as we found to obtain in Urodeles between the 

 posterior border of the fenestra and the recessus fenestra vestibuli. In 

 the Anura however it is larger, and the portion of the fenestral mem- 



') It has been surmised that tlie air-bladder, in tislies posscssinj^ a Weberian 

 chain of ossicles, is functional as a resonator. Tiiis suggestion sliould not l)e lost 

 sight of in the present case. 



