448 Edward Phelps Allis jr., 



joining portions of the main infraorbital, latero-sensory canals. This 

 part of the sacculus is the "schwimmblasenartigen Organe" of Erdl's 

 descriptions. The pit lodges no portion of the Pars superior of the 

 membranous ear, the utriculus lying wholly within the cranial cavity, 

 and the semicircular canals partly in that cavity and partly enclosed 

 in the bone or cartilage of the skull. The pit extends inward through 

 the space enclosed between the utriculus and the external semicircular 

 canal, the external opening of the pit accordingly lying dorsal or 

 dorso-mesial to the latter canal. This is all shown diagrammatically 

 in the accompanying cut. The posterior half of the ventral edge of 

 the pit is cut away somewhat, so that the pit may be said to 

 here lead onto, rather than to open onto, the posterior surface of 

 the skull. 



If now this large pit of Gymnarchus be compared with the tem- 

 poral-hole of Erythrinus, as shown in Sagemehl's [64] figures 3, 4, 

 and 11, it will be seen that if the hole in this latter fish were to be 

 deepened, in its posterior portion, it would necessarily open into, 

 and incorporate into itself the cavity of the bulla acustica lagenaris. 

 Slight changes in the position of the outer edge of the hole, inter- 

 mediate stages of which are shown in the Cyprinidae, would then pro- 

 duce the pit of Gymnarchus, which pit is thus certainly formed by 

 the fusion of the temporal groove (Temporalhöhle) of Sagemehl's de- 

 scriptions of other fishes, with that part of the labyrinth that lodges 

 the sacculus and lagena. The pit may accordingly be called the 

 acustico-temporal pit. The extrascapular bone of the fish would quite 

 naturally become a cover to it, for it in part covers the temporal 

 groove in Amia and certain other fishes. 



We thus have, in Gymnarchus, through the intermediation of the 

 latero-sensory canals and the acustico-temporal pit, an auditory appa- 

 ratus that is functionally quite exactly similar to the amphibian ear. 

 For it is plainly evident that whatever vibrations, or variations of 

 pressure, are impressed upon the fluid that fills the latero-sensory 

 canals of Gymnarchus must be at once transmitted to the endolymph 

 of the tightly distended sacculus, just as similar vibrations are trans- 

 mitted [44], in Amphibians, through the fluid that fills the perilymphatic 



