84 BULLETIN OK THE Hl'KEAf OF FISHERIES. 



is progressive, lasting-, it iimy lie. for diiys. ;it llic cud of wliioli tiiiu- tlic Hsh lias a 

 decidedly water-soaked appeaiuiue. 



Loeb (1900) in experiinouts with pieces of muscle, which he immersed in various 

 solutions, found that such pieces might either gain or lose in weioht in solutions 

 which were equimolecular. but ditl'ered in respect to the salts employed. Chemical 

 action, rather than simple osmotic a(tii)ii. he believed to be responsible for tiie infiwx 

 or efflux of water. 



If this explanation be applical)le to the dead fishes used in my experiments, it 

 will be seen that the attempt by Fredericq (1904) to ascertain the osmotic pressure of 

 the fluids of the tissues of various fishes by similar weight determinations was not 

 likely to yield reliable results, and, indeed, he admits that the lig-urtvs thus obtained 

 differed widely from those obtained by the cryoscopic method. 



The case of those fishes which died in hypertonic solutions and likewise lost 

 weight in those solutions is perhaps to be explained on the ground that the loss of 

 weight occurred before and shortly after death, though this explanation does not 

 seem to cover all cases (experiment so. second day). 



PERMEABILITY OF THE MEMBRANES TO SALTS IN SOLUTION." 



Some of the preceding ex[)ei-iments seem to prove conclusively that considerable 

 amounts of water may enter or leave the liody of a living fish if the latter be tran— 

 ferred to water of a density much lower oi- much higher than that to which it has 

 ])reviously been accustomed. That in all probability this water does not enter or 

 leave the body by waj' of the alimentary canal has already been noted, in anticipation 

 of results to be discussed later. There is thus left, as tlie most likely alternative, an 

 osmotic passage through one or more of the membranes bounding the body. It has 

 ]»cen shown, likewise, that these changes occur, in many cases, witliout apparent 

 harm to the fish. 



The cpiestion next to be discussed is whether or not salts in .solution may likewise 

 dill Use through these membranes. If the gain or loss of weight shown in the preced- 

 ing experiments be due to osmotic action, it is of course impossible that the pas.sage 

 of salts should occur as readily as that of the water. Does it occur at alH 



This i)roblem I have attacked from ))otli si<le.>. In the first place, the passage of 

 salts (strictly speaking, of chlorides) into fresh water from fishes taken from salt or 

 brackish water was tested chemically. In the second place, the salt content of the 

 tissues of variou.s fishes which had li\ed in water of various degrees of .salinity 

 was likewise determined. It will be found that the results from these two methods 

 l)iesent some striking points of agreement. The former method will be discussed 

 first, although the results are as a whole less satisfactory than in the second case. 



DKTKinilNATIONS OF TUK Clll.oltlNK I'ASSED IXTO THE WATEi: BY THE FISHES. 



Fishes of known weight taken from salt or brackish water were transferred to 

 khdwn volumes of fresh (in a few cases of distilled) water. The fishes were, in all 

 ca-es. given a preliminary rinsing (Id to :!n minutes) in freshwater before being 

 plaied in the, water to be tested. This was ol)\ iously necessary in order that all salt 



.■ 1 liiivc spoken tlirougliout ol suUs nulu-r limn ioiie, bcuuuse I ciiu nol si-e ihe lulvantajje, (m- present pnrp..-.'. 

 ijor.tiHing Ihe language of the eleetrolyiie dissociation hyimUiesis. 



