MUSEUM OF COMPARATIVE ZOOLOGY. 17 



In my preliminary analyses I bad found that a very thin stratum of 

 air under the inverted vessel was sufficient to allow the ordinary emis- 

 sion of gas. The first experiment in which the gas was accurately 

 analyzed was therefore conducted upon fishes confined under the in- 

 verted dish, beneath which was a bubble of air of known volume, 

 which had been deprived of all its carbon dioxide. The details of the 

 experiment were as follows. 



Experiment 1. — Dec. 13, 1883. At 11:45 a.m. six gar-pikes, varying 

 in length from 13 to 21 cm., were removed from running water ( + 12° C.) 

 to the experimental jar, containing recently distilled aerated water at 

 about the temperature of the room ( -I- 20° C), over which the glass 

 vessel, with a bubble of air (deprived of carbon dioxide) measuring 

 66.71 c.c. 1 had been previously inverted. The experiment was con- 

 tinued until 3:15 p. m. (3^ hours) without renewing the water. The 

 fishes were then taken out and the gas collected. In the final transfer 

 over the mercurial bath a portion of the gas was lost, so that its total 

 volume could not be ascertained. 2 The portion remaining was collected 

 over mercury, and found to measure 19.27 c.c. (reduced). Treated with 

 potassic hydrate, the volume was diminished to 19.07 c.c. Absorbed, 

 .20 c.c. = 1.04% = carbon dioxide. This volume (19.07) transferred 

 to water and treated with pyrogallate of potash measured over water 

 16.27 c.c. Absorbed, 2.80 c.c. = 14.5% = oxygen. 



Experiment 2. — Dec. 14, 12:30 p. m. Six gar-pikes (the same as in Ex- 

 periment 1) were removed from water ( + 11° C.) to the experimental jar 

 containing water at +21° C. The inverted vessel contained 65.20 c.c. 

 of air deprived of its carbon dioxide. The experiment was continued 

 until 4 p. m. (3 \ hours). During the last hour all the fishes remained 

 at the top of the water, and became somewhat swollen in appearance. 

 One of the six did not recover from the effects of the abnormal 

 conditions. The gas, re-collected and measured, was found to have 



1 The eudiometer employed was graduated to fifths of a cubic centimeter. 

 All the volumes given are the volumes reduced to 0° C. and 1 m. pressure. 



2 In a preliminary experiment it was found that under nearly the same condi- 

 tions there had been a slight increase in volume. The conditions were as follows. 

 Dec. 3, 1883, 12:30 p. m. Four fishes (17-20 cm.) were transferred from cold w.iter 

 into an experimental jar (Fig. 1, p. 16) containing water at +20° C. and a bubble 

 of ordinary air measuring 53 c.c. At 2:30 p. m. five other somewhat smaller fishes 

 were added. At 4:30 p. m. the experiment was discontinued, and it was found that 

 the bubble had increased to about 56.5 c.c, the increase being nearly 3.5 c.c. 



vol. xix. — no. 1. 2 



