446 MORRIS M. WELLS. 



increased to 76.9 grams. No way to account for this increase 

 is clear. Some food substance may have gotten into the 

 aquarium and this seems to be the most likely explanation, for 

 one week later the weight had decreased to 75.3 grams. The 

 fish was not dissected and the presence or absence of food 

 ascertained as it was thought best that the series be kept un- 

 broken. Fish no. 5 shows a similar increase in weight on Febru- 

 ary 8; again in a week the weight fell to a figure below that just 

 previous to the increase and no other rise occurred. 



3. Resistance to Lack of Oxygen. 



Table II. is a summary of experiments performed to determine 

 the resistance of starved fishes to lack of oxygen. 



Note that Table II. shows a rapid initial increase in the resis- 

 tance of the starved fishes to lack of oxygen and that this in- 

 creased resistance gradually diminishes till after 53 days without 

 food the starving fishes are considerably less resistant than the 

 control fishes. It is also interesting to note that the decrease in 

 resistance following the initial increase proceeds slowly for the 

 first 39 days and then rapidly, till on the fifty-third day the 

 starving fishes show a resistance that is not only lower than 

 that of the control but is also lower than that of fishes of the 

 same size but which had not gone without food for so long a 

 time. The increase in resistance upon the part of the starving 

 fishes is emphasized by the fact that the control fishes show a 

 markedly increased resistance with the progress of the season 

 as shown by the curve (Fig. i) and the control readings in 

 Table II. 



4. Resistance to KCN. 



When it was found that starvation results in an increase in 

 the resistance of the starving fishes to lack of oxygen, and that 

 this increase is followed later by a rapid and marked decrease in 

 resistance, it was decided to test the susceptibility of fishes of 

 the same species and in similar stages of starvation, to solutions 

 of KCN. In this way the "susceptibility to cyanide" method 

 that has been used so successfully with planaria by Child, was 

 applied to fishes. A comparison of the results obtained by the 

 two methods is interesting in that they show in general the same 



