GAS GLANDS O? SOME TELTi:OSTi;A^' FISHES. 



225 



structures tlier%,can be no doubt whatever (see Appendix B), 

 They have been exhaustively studied by Bykowski and Nusbauni 

 & Reis, and I have myself described and figured them in the 

 present paper. Jaeger himself, as I have already stated, admits 

 the presence of vacuoles in the gas gland cells, but, curiously 

 enough, denies that they contain gas, and proi:iounces them 

 similar in nature to the vacuoles in the cells of the liver. It is 

 difficult to understand why this comparison should be made, since, 

 from his own standpoint, it is the function of the gas gland cells 

 to pump oat gas and not to store up glycogen or fat. But even 

 adopting his suggestion that the supposed g-is bubbles are only 

 vacuoles, we may remark that several authorities have described 

 in the kidney " the formation of vesicles in the cells and appear- 

 ances which indicate the discharge of these vesicles into the 

 cavity of the tubules '' (37) ; and since kidney cells, like the gas 

 gland cells, are supposed by many modern physiologists to derive 

 most of their excreted substance by direct abstraction from the 

 blood, comjjaratively few of the constituents of urine being 



Text-fig. 60 (x 1000). 



Vacuoles in the kidney cells of Nerophis, some apparently being expelled into 

 the lumen of the tubule. 



manufactured in the cell, we may certainly conclude that even 

 liquid vacuoles can be originated and discharged in a manner 

 precisely comparable with that of the gas bubbles, and that if they 

 contained gas instead of liquid they would be indistinguishable. 

 I have figured some of the vacuoles present in the kidney cells of 

 the fish Nerophis (text-fig. 60). It is evident that in histological 

 preparations no sign of the liquid urine would be visible after 

 expulsion from the cell, whereas the gas bubble with its walls of 

 cytoplasmic material may, like a soap bubble in air, persist for 

 some time before bursting. Consistently with his denial of the 

 gaseous nature of the vacuoles in the gas cells, Jaeger was com- 

 pelled to regard the gramilar masses lying external to the gas 

 gland in the bladder lumen merely as a bye-product of the specific 

 activity of the gas gland. 



If, however, we fully agree with Kusbaum & Reis as to the 



