No. I.] THE VERTEBRATE EAR. 315 



21. The above conclusion is rendered certain by my discov- 

 ery that in ElasmobrancJis the stnictiwal connection betzveen the 

 ear organs and the surface canal organs is for a long time inaiji- 

 tained after the ear has migrated to its internal home, and in 

 some forms may be said with truth to persist during the life of 

 the individual. 



22. In the Mammalia the anterior and posterior canals, by 

 uniting, fuse their two half-pores into one pore, which maintains 

 the open communication of the semicircular canals with the 

 morphological surface of the head. Thus we have in the human 

 ear involution on involution ; for the main chambers of the ear, 

 utriculus, and sacculus, are the result of an involution of a por- 

 tion of the surface of the head into the tissue below, while the 

 semicircular canals forming later are the result of involutions 

 from this ear surface. 



Recapitulation D. 



23. The three semicircular canals arise in all cases which have 

 been adequately studied from three depressions, and their up- 

 growing walls are formed in connection with the differentiation 

 of the canal sense organs. They appear about the ampullary 

 cristas first, and only later grow toward each other to produce 

 a common pocket. Consequently the conclusions of Krause 

 and others who state that the posterior and anterior canals arise 

 from one pocket are to be accepted only after such modifications 

 as will bring them into harmony with the course of development 

 known for them in bony fishes. 



24. The semicircular canals may be said to be completely 

 formed so soon as the lips of the folds have met and fused. 

 The separation of the canal body from the utriculo-sacculus is 

 a secondary process due to the resorption of the fused part of 

 the lips of the groove. In most vertebrates this condition is 

 soon assumed after the lips meet, and persists throughout life ; 

 but in Petromyzon the tissue intervening between the utriculo- 

 saccular walls and the canals is not so soon nor so completely 

 resorbed, which accounts for the closeness with which the 

 canals are applied to the body of the ear in this animal. 



25. A more exact determination of the absolute position of the 

 ear with respect to the structures of the side of the head is now 

 possible, and we find that the auditory vesicle is always developed 



