The Latero-Sensory Canals and Related Bones in Fishes. 449 



spaces to that that fills the saccnlus and lagena. And the enclosui-e of 

 the canals of G5'^mnarchus in open grooves instead of closed bony 

 canals, and the complete closnre of the canals themselves, by the 

 closing- of their primary pores, would certainly together facilitate the 

 reception and transmission of vibrations from the exterior to the 

 saccuhis. Have we then here simply two wholly unrelated though 

 functionally equivalent arrangements, evolved independently in the 

 vertebrate series? Or are they directly related, developmentally, one to 

 the other? An anatomical objection that at once suggests itself to 

 the latter supposition is that the latero-sensory spaces of Gj^mnarchus 

 lie wholly external to the perichondrial membranes, while in amphibians 

 the perilymphatic spaces lie wholly internal to those membranes, the 

 membrane of the fenestra vestibuli being said to be derived from the 

 perichondrium of the outer surface of the chondrocranium. But if 

 the acustico-temporal pit of Gymnarchus has the origin I ascribe to 

 it, the greatly distended saccnlus must lie in part wholly external to 

 the chondrocranium and its enclosing membranes, in the greatly en- 

 larged temporal groove of the fish. If then, in the further develop- 

 ment of the organ, the saccnlus should be withdrawn again within 

 the chondrocranium, or if it should become secondarily re-enclosed in 

 the skull by the closing of the external opening of the acustico-tem- 

 poral pit, the functionally associated portions of the latero-sensory 

 canals would naturally follow it and share a similar fate. This en- 

 closed section of latero-sensory canal could then acquire a considerable 

 inter-aural extension, that thin portion of the auditory wall that 

 separates the saccular and pharyngeal cavities could become secon- 

 darli}^ perforated by the fenestra vestibuli, as Harrison sets forth, 

 and the amphibian ear would be acquired. The functional and ana- 

 tomical agreement is much too complete not to warrant the suggestion 

 of a developmental relationship, and the origin and acquired function 

 of the perilymphatic spaces of the amphibian ear would be not only 

 naturally, but beautifully explained. For the perilymphatic spaces of 

 the ear would thus, as it were, be the result of a confluence and 

 subsequent condensation of canals produced by the involution of those 

 anterior and posterior portions of the latero-sensory system that were 



Internationale Monatsschrift für Anat. u. Phys. XXI. 29 



