[FjIj] embryography of osseous fishes. 509 



in Gadus, ou to that in which every mnscular segment has its corre- 

 sponding pair of nerve hills or eminences. Their function is evidently 

 a sensory one, and their serial relation to the auditory, optic, and olfac- 

 tory organs is at least suggestive, if nothing more. The structure of 

 these hills, however, bears a most remarkable resemblance to the end- 

 ing of the auditory nerve in the auditory vesicle, even as regards details. 



They appear to be mere local lenticular thickenings of the skin, con- 

 nected by a nerve filament with the spinal nervous axis, and with 

 careful illumination one may see that these bodies are surmounted 

 externallj^ by very fine, hyaline protoplasmic filaments, which extend 

 freely into the surrounding water, but which are perfectly rigid and 

 immobile. In this they exactly resemble the similar filaments which 

 are met with surmounting the macula or cushion-shaped termination 

 of the auditory nerve. There is much room here, for one disposed to 

 speculate, to suggest a probable explanation of such remarkable resem- 

 blances. Dercum has suggested that they may possibly serve to appre- 

 ciate vibrations not perceptible to the ear, serving i)erhaps to enable 

 the animal to detect the approach of another body, which starts the 

 surrounding fluid medium into slow vibration. Their columnar epithelial 

 structure has been determined by observers, but the nature of the 

 process by which thej' become converted into the covered system of 

 the lateral line, as found in the adult, still remains to be worked out. 

 In the adult, one or more rows of scales are often involved in the struct- 

 ure of the canals of the lateral line system, these scales having a tube 

 developed along their longitudinal axis, or it may even be branched. 

 Within these tubes, which also open in various ways to communicate 

 with the outside, a complex system of nerve buttons or eminences are 

 found, which are evidently akin to the nerve hills found in the larvae 

 as naked dermal elevations. Some of the mounted preparations of these 

 structures of the adult, treated with osmic acid and hsematoxylon, 

 prepared by Dr. Dercum, have a remarkable likeness in some respects 

 to the ending of the eighth nerve in the sf'iditory vesicle of larval fishes. 



Balfour found the lateral line system of Elasmobranchsto be innervated 

 from the ninth pair or vagus nerve. Such a relation of the first nerve 

 hill on the head of the young cod is conceivable, but I am assured that 

 nervous fibers pass inwards separately to the nervous axis from each 

 of the others behind it, so that such a relation to the vagus is here 

 scarcely possible for the latter. The inclosure of the lateral line system 

 is probably accomplished by the development of folds of the skin above 

 and below the series of nerve hills, these folds coalescing finally to form 

 a canal open to the exterior at iuterv^als. 



The single median barbel on the lower jaw of the adult cod is also a 

 sensory organ of a special kind, but is not developed until the young are 

 older than the oldest figured in the plates accompanying tbis memoir. 

 It is therefore developed during the post- larval stages. Leydig has in- 



