126 GEORGE G. SCOTT. 



he carried on at the New York Aquarium he noted the beginning 

 of regeneration of scales six days after they had been removed, 

 and during which they had been kept in sea water with a specific 

 gravity of 1.025 having been transferred to this from the dilute 

 harbor water with a density of 1.007. Mr. Denyse, of the New 

 York Aquarium, informs me that it has been his experience that 

 death usually accompanies serious injury to skin of fishes. 

 Notwithstanding the difference of opinion cited above it is true 

 nevertheless that in none of the experiments about to be described 

 were scales removed. In fact, there was no visible evidence of 

 the coat over the scales being injured. Attention is here called 

 to a further consideration before the experiments are described. 



The specimens used in the following experiments bore no 

 surface abrasions and to the eye appeared normal. They were 

 taken as needed from a large storage tank supplied with running 

 sea water to which tank new specimens were added by the col- 

 lector from time to time. Not only was a rough selection em- 

 ployed here but a further examination was made before they 

 were used in the experiment. 153 specimens were used in 21 

 experiments and 988 weight determinations were made. 



Control experiments with Fundulus in sea water showed on 

 the whole a slight loss in weight as time went on. The results 

 were similar to those found by Sumner. Certain individuals 

 died in this control experiment showing that death in the follow- 

 ing cases is not always due to the experimental solutions. In an 

 experiment with six fishes in which an individual record of each 

 was kept the entire lot was dead about eighteen hours after the 

 experiment began. Variations in the individual weights were 

 evident. But since they were all dead at the end of the period in 

 question the average results will alone be given. Thus the 

 average gain in weight at the end of two hours was 5.2 per cent.; 

 at the end of five hours, 8.7 per cent.; at the end of eighteen 

 hours, 15.7 per cent. In a second lot, the results were similar, 

 five of a lot of six specimens were dead at the end of twenty hours. 

 The change of weight was as follows, at the end of three hours 

 4.4 per cent. ; at the end of five hours, 7.3 per cent. ; at the end of 

 seven hours, 9.6 per cent. This is the average of five specimens. 

 The sixth was dead and showed a gain of 12.4 per cent. The 



