70 BULLETIN OP THE BUREAU OF FISHERIES. 



Dr. Exfnnanii. liowcNcr, that in the case of the Pacitic salmon the passage into 

 the t'rosli water iiiav Ix' made (|\iite altruptly at the mouths of the smaller streams. 



Mather (ISSl) publishes a list, based mainly upon the authority of J. W. Milner 

 and (i. B. Goode, of "■ fishes which can live in both fresh and salt water." This list 

 aggregates 33 species, and it would probably be possible to enlarge it very greatly." 

 Nothing is said, however, regarding the duration or circumstances of the change of 

 medium, and physiologically these are of course highly important. By sufficiently 

 slow acclimatiisation almost any change of habitat seems possible, and iiid('(>d such a 

 process must have occuri'ed on an enormous scale in the evolution of lishes. 



CHANGES OF WEIGHT RESULTING FROM CHANGES IN WATER DENSITY. 

 CIIANCES OF WF.ICIIT IN LIVIXU FISHES. 



Tlius far tlic (lisriiNsi<in has been lontiiied mainl}' to the death or survival of the 

 various li-^lics under the conditions of e.Kperimentation, but little analy.sis being 

 attempted of the pheiiouieiia iu\ol\e(l. It is evident that such changes of density 

 as we ha\e dealt with in\-ol\-e enormous ehaiiges in the osmotic pressure of the 

 medium with wliich th(> lish is bathed. In relation to the latter, the limiting mem- 

 branes of a given fish might be (1) impermeable both to water and to salts tlicrem 

 dissolved; {2) semipermeable, permitting the passage of water, ))ut barring the sulr>: 

 (3) permeable to both in greater or lesser degree. If the first condition were real- 

 ized, no change in weight ought to occur following the transfer of the fish from one 

 "medium to another; if the second condition o))tained, we should expect an increase 

 of weight in hypotonic .solutions and a decrea.se in hypertonic ones; while in the thii-d 

 case we should expect the same results in a somewhat lesser degi-ee, unless, indeed, 

 the membranes were equall}^ or almost equally permeable to water and salts, in which 

 case the conditions necessai'y for osmotic action would be wanting. 



It must be remembered, however, that we are dealing with living matter, 

 bounded by living membi-anes, and that the conditions may in consequence be much 

 more complicated than outlined in tlie foregoing scheme. Various so-called "vital"' 

 (i. e., as yet unexplained) factors may intervene. Thus the degree of "perme- 

 ability" of a given membrane may not be a constant quantity; it maybe found to be 

 in some way under the regulative contrt)l of the organi.sm, and to vary greatly at 

 ditfei'ent moments and under different conditions. Again, it may be that chemical 

 factors intervene, and that an imbil)ition of water may at times occur which is not 

 accounted for by tiie laws of osmotic pressure alone; and indeed we have already 

 noted a vast ditlerenee between the physiological eflFectsof fresh water and of slightly 

 saline water, a jihenomenon not to be accounted for by their difference of osmotic 

 pressure. 



If the \\-eight of a given tisji i-emained constant in a given medium'', it might, 

 then, be concluded either (1) that the medium was nearl}- or quite isotonic with the 

 body fluids of the fish, or, if not, (2) that the membranes of the fish were for the 

 time being impermeable to water, or (3) that they were more or less permeable to 

 both salts and water, but to one in nearly the same degree as to the othei-. 



a Dr. Gill has informed me of several examples not incliicied in Mather's list. 

 ''Always allowing for loss due to waste. 



