KKFKCTS UPON FlJ^HKS OF CHANG KS IN SALINITY OF WATER. Oi 



The lirst tlirc' .■xpcrimcnts deal with /•'. i,in},ii;x. 



K.rj,rrhnn,t 1." 



August, \W)± Thivo hundred :nid forty-four tishes put into t:iid< of salt 

 water and hitter displaeed l.y fresh. Fifty pei- cent of the lish died (hiring 

 lirst -24 hours: only 1 lish liviny- at ."ud of H days. 



August. I!tn4. Twenty-tiye pickcnl specimens phice<l alu-uptly in tank of 

 fresh \vater. All hut I dead in is hours; all dead within 24 hours. 

 E.qX'llm.nt S. 



September, iy04. Twenty-tive j)icked specimens placed ahruptly in fresh 

 water. Fish fed throughout expei-iment. Thirty-six ]ier cent dead at end of 

 first day; last fish dead in ti days. 



The li.shes here used had been taken in sea water of full strength, and kept, 

 prior to the experiments, in tanks of running salt water of a mean density of 

 abt)ut 1.023 (reduced to 4- C.) and a mean temperature i-anging from ^O^* to 

 22 C. The change of temperature in the transfer from salt to fresh water 

 was yery slight (probably neyer exceeding 3'- l!. ) and can haye play(Hl no j)art, 

 here or in later exjierinients, in causing the death of the lishes. 

 The next two experiments illustrate the relatiye mortality of /•: nm ),!/;.'< and /-'. 

 hrtrrorlltiis, and. in the second cas... of F. ih^/j^/nnn/s. 



July. l'.H)4. Thirty-three /•'. uiiij,ir,x and 71 /'. Intrro.'lif,,.. \mt into tank of 

 salt water, this being disjjlaceil by running fresh watei- !,s in ex])erinient 1. 



F. ii,<i},ii:.-<. No <leatlis during first day; 42 per cent dead within 2 days; 

 ss per cent dead within 3 days; the last in 4| days. 



/•: h.hroelifns. First death in 3i days; T."> per cent dead within (iA^ days; 

 all dead within 7i days. 



July. I!tn4. Twenty-eight ,„uj,il!s^ 32 hrlrrornfua.anA 2-2 <//,ij>/,(iini.^ \\in\' 

 put directly into a tank of running fresh water, at 20- C. The /ikiJ, ///'.■< and 

 /i(f</'(>r//'fii.'< came directly from sea water of full strength, tiie (Vuiiihuiiiix from 

 a tank of brackish water of specific- grayity l.OtH!'' (originally from a brackish 

 pond haying water of specilic grayity 1.00.5). 



/-: /,»(;<^/;x died in from less than 12 to 3G hours; ]\ hrtrmrl it ns died in from 

 12 hours to 10 days; /•'. ,i;„j'h,nnis died in from h to 14 days. All of i,iaj,ii:x 

 had died at a time when only 7 In I. rcrHfux (22 per cent) and when no iliiiplKiNux 

 had died; and all of h.fmx-lifHs had died at a time when only 5!> per cent of 



\hi',i;„j.Jnn,ii. had died. 



The or(hM- of death of these three si)ecies witen subjected to asphyxiation in stale 

 water was found to be (1) </;,q,h,nHis, (2) „ni}ul!s, (?,) hrtrmrlU us. 

 F,;„rnnr,.tG. 



Kighty-seyen per cent of the ilJa/iliinius had died wh(>n only '.\U per cent of 

 the w<i},iJ:s had died and when all of tlu> h.f.mrlltus wei-(> still liying. 



Frp,rn,„nt7. 



Results similar to experiment 6. 



<i Tlii-su i-xiuTiiiicnIs iirc! not numbered in i-lironolcigical (ihUt, but iire tlimiiRhiint iimmKed siilely with reference tii 

 snhjeet-iuatter. The lirst twelve iire imimM tliose cdnrtneteil -.a VVnniis Hole. I- .See p. (10. 



