SCOTT, STUDY OF CHANGES IN MUSTELUS CAN IS 25 



having an inside diameter of about a centimeter. It seems to me that the 

 normal A was regained in this case, because the external media produced 

 but a small degree of departure from it. It is interesting to note that the 

 maximum change produced in A is about equal to the normal range in A 

 of normal dog-fish blood as described in the first section of this paper. 

 The results with the other fishes indicate, however, that osmotic phe- 

 nomena are complicated by the presence of other factors. 



EOLE OF THE GiLLS IN THE MODIFICATIONS 



Considerable difference of opinion exists as to the part of the body that 

 is concerned in the osmotic changes in the blood brought about by changes 

 in the osmotic pressure of the surrounding medium. As stated above, 

 there are three structures that may be the seat of this phenomenon, 

 namely, the skin of the body, the lining of the alimentary tract and the 

 gill membranes. Any one or all of these structures may be conceived to 

 share in the above processes. The surface of the body of the dog-fish is 

 covered with a closely associated system of dermal plates forming, with 

 other structures of the skin, a tough coat through which it would appear 

 that fluids could pass with the greatest difficulty if at all. The cells of 

 the intestinal tract are known to exert a selective action on materials 

 present in the intestine, and therefore we should expect that solutions 

 more or less concentrated than sea-water which would possibly accom- 

 pany the swallowed food would be passed out through the cloaca before 

 osmotic changes of any account would take place. Furthermore, my ob- 

 servations indicate that the oesophagus and the cloacal aperture are kept 

 closed during the greater part of the time, and are probably opened only 

 during the taking in of food and the getting rid of waste. Therefore the 

 wall of the gut would not ordinarily be exposed to solutions differing in 

 density from that of sea-water, even though the whole fish were entirely 

 immersed in such solutions. 



The gills, however, are always freely exposed to the external medium. 

 Each gill filament contains a fine capillary loop composed of an afferent 

 vessel and an efferent vessel supported by connective tissue. Covering the 

 capillary apparatus is an extremely thin epithelial membrane, so that 

 there are but two thin layers of cells between the water and the blood 

 Btream, namely, the gill membrane and the endothelial wall of the capil- 

 lary. If the rich capillary supply of the gills be taken into account, there 

 is in effect a large, broad sheet of circulating blood separated from the 

 water by an extremely thin membrane known to be permeable to gases. 

 A priori, therefore, it would seem that the osmotic changes in the blood 

 described above might take place through the gills. 



