930 



SCIENCE 



[N. S. Vol. XXXIV. No. 887 



mission, contained in the same paper in 

 ■which he criticizes me, that all but five per 

 cent, of the adult Fundulus died in his 

 " double distilled " water. Why, then, as- 

 sume that the water employed by me contained 

 poisonous impurities? (4) Weighty evidence 

 upon this same question is already afEorded by 

 an experiment which I have only recently 

 commenced. At the date of writing this para- 

 graph (Dec. 12), I have kept in distilled water 

 for seven days eighteen specimens of Fundulus 

 heteroclitus, taken from a fresh water stream 

 and consequently habituated to the latter 

 medium. The distilled water was prepared in 

 an ordinary metal still. The fishes neverthe- 

 less all appear to be in perfect health, and no 

 deaths have occurred in the lot since the first 

 day, when two fishes died from causes having 

 no bearing upon the present problem. This 

 result is significant in comparison with that 

 which had been obtained when salt-water 

 specimens were used. The final outcome will 

 be reported upon later. 



As bearing on the question of the " pro- 

 tective " action of various salts, it may be 

 relevant for me to point out that in 1906 I de- 

 scribed experiments demonstrating the re- 

 markable efiicacy of even small percentages of 

 sea-water in counteracting the fatal effects of 

 fresh water upon Fundulus. I later showed' 

 that this proportion need not exceed one part 

 of sea-water to a hundred of ordinary fresh 

 water. Experiments in which pure NaCl was 

 employed, dissolved both in distilled water 

 and in fresh (" tap ") water, showed that this 

 salt, in concentrations of 3 to 15 grams per 

 liter (in some cases three tenths gram per 

 liter), may preserve the lives of the fishes for 

 three or four weeks or longer. Fishes kept in 

 pure tap water, under otherwise similar con- 

 ditions, all died within comparatively few 

 days. 



The fact that this fish will endure pure 

 NaCl, in " very weak solutions," is now fully 

 admitted by Loeb himself," but he still appears 

 'American Journal of Physiology {loc. cit., par- 

 ticularly pp. 68, 72, 73). 



' Archiv fiir Entioicklwigsmeclianilc (loc. cit.); 

 also Science, November 17, 1911. 



to overlook the fact that, in such low concen- 

 trations, the salt in question is far from being 

 a poison, but frequently preserves the fish from 

 destruction. 



Furthermore, we surely can not regard 16 

 grams per liter (a concentration tolerated by 

 many of my fishes) as a " very weak solu- 

 tion." Indeed, it is roughly an M/4: solution, 

 or one of more than half the concentration in 

 which this salt occurs in sea-water." Experi- 

 ments in which sodium chloride was used in 

 about the same concentration as in sea-water 

 resulted in the death of all the (30) fishes 

 used in from two to fifteen days. 



In summing up this part of the discussion, 

 I can but repeat my earlier statement that 

 " In addition to such a toxic effect, however, 

 the sodium chloride certainly has a potent 

 anti-toxic effect, since, even in solutions 

 which proved fatal, the rate of death was 

 usually much lower than in pure fresh water. 

 In the aggregate, these experiments may he held 

 to prove, therefore, that pure sodium chloride, 

 in certain proportions, has nearly (if not 

 quite) the same efficacy in counteracting the 

 fatal influence of fresh water upon Fundulus 

 heteroclitus as does the combination of salts 

 contained in sea-water. My previous experi- 

 ments have abundantly proved, I think, that 

 the action of this salt is not an osmotic but a 

 chemical one" (190Y, p. Y3). 



In a section of considerable length, entitled 

 " The Toxicity of Certain Poisons as Affected 

 by the Salinity and Osmotic Pressure of the 

 Medium," I pointed out, among other things, 

 that certain metallic salts (e. g., copper 

 chloride and sulphate, and mercuric chloride) 



" Referring to some of his experiments with 

 young Fundulus, Loeb tells us (Science, loc. cit.) : 

 ' ' I succeeded in showing that as long as the 

 sodium-chloride solution is very dilute and does 

 not exceed the concentration of M/S, the addition 

 of KCl and CaCL is not required. Only when the 

 solution of NaCl has a concentration above M/S 

 does it become harmful and does it require the 

 addition of KCl and CaCL. " The difference be- 

 tween Loeb's results and my own — of which last 

 Loeb does not seem to be aware — may be due to 

 the difference In the age of the fishes employed. 



