56 BULLKTIN OF THK BUREAU OF FISHERIKS. 



Tlu' ('\i)eriments hero (Icsnihcd were ciuricd on ihiriiiu- the .summer.s of 11)04 

 iuid lli<»5 at the biological ialKU'atorv of tiic Hiircaii of Fisheries at Woofls Hole, 

 Mass., and during the spring of 1!I05 at the New York A(iiiarinni. In the earlier 

 portion of this work I was aided by Mr. D. W. Davis, assistant in this labora- 

 tory. Tiie work at the New 'Vork A(|uariniii was rendered possible throiigii the 

 kindness of tiic director. Mr. (". II. Towiisend, wlio placed at my disposal a room 

 e<pnppe(l for research, and pros ided nie with ahundant nialeiial throughout' the 

 course of the experiments. 1 must acknowledge, also, the ever-ready help of several 

 members of his stati'. My thanks are likewise due to Prof. W. C. Sabine, of the 

 department of physics of Harvard University, for valuable ei-iticism. 



THE DEATH OR SURVIV.'^L OF VARIOUS SPECIES AFTER CHANGES IN THE 

 SALINITY OF THE WATER. 



TIIK FATAL EFFECT OF FRESH WATER UPON VARIOUS SALT-WATER FTSHES. 



My experiments were originally undertaken from a biometric rather than a 

 physiological point of view. It was my object to determine measurable indices of 

 fitness and unfitness among the individuals of a species, somewhat as Bunipus had 

 done for sparrows and Wcddoii for crabs. Different mod(!s of elimination were 

 chosen, chiefly asphyxiation and change from salt to fn^sh water. The individuals 

 of the more and the less resistant halves of each lot of lish >o treated were subjected 

 to statistical study and differences in type and in \ arialiility were noted. These 

 biometric studies are still far from complete and no statement of the results is here 

 attempted. My attention has for a time been diverted to the physiological questions 

 which form the basis of the present paper. 



The tirst experiments recorded illustrate the fatal effects of transfer to fresh 

 water upon our three local species of killitish {J^u».dtdus). These are all small fishes, 

 restricted to shallower waters. J^. hidjalis and 7^ hcteruclitux are marine Littoral 

 forms, occurring particularly on weedy shores, though both species, and especially 

 the latter, pass into the brackish waters. Bean (lt>0;3) says of hcterocUtns that it 

 sometimes ascends streams l)eyond tidewater, and regarding imijalis he sees "no rea- 

 son to dou}>t its occurrence, even in fresh water," while Eugene Smith (1897) states 

 of ](ii, rod i tux thiit it "is often found landlocked in ice or quarry ponds." Dr. H. M. 

 Smith informs me that the latter species "is found permanently in the vicinity of 

 Washington, in the Potomac and its tributaries, and also in ponds." Mr. Vinal 

 Edwards likewise reports a cas(> where this hsh was found in fresh water above a 

 milldam. Fivsh water is not, however, tlie more usual habitat of F. hetrroditm. 

 F. di(i.j>li(inu:<. on the contrary, is predominantly a brackish and fresh water fish, 

 extending from the coast far into th(^ interior, and occurring in lakes and streams as 

 far west as the Mississippi River, and northward to MiiuK^sota." These three species 

 are readily dislini^ui-hable from one another in form and color. They are very dif- 

 ferent in their haliit^and in \arious physiological characters. Accordingly, the .species 

 should always be spccilii'd in any work dealing with a member of this genus.* 



"The u. viiiii r. |,r,v,!it,itH c-^ .if llir spcirv imvu huen given the rank iif ii (iistinet viiriety. mcnuiia. 



^Fnr (liMiii.tiMiis ;iii<l illi!>n-:itioiis (»f ilirsi' tislu's llie reader is referred to .Jordan and Everniann's •• Fishes of North 



