18 BULLETIN OF THE 



diminished nearly 18.3% in volume ; it measured only 53.29 c.c. This 

 diminution, however, is evidently not to be accounted for as the result 

 of actual absorption by the fishes, but rather as the result of their over- 

 distention with gas swallowed, — probably in the vain endeavor to com- 

 pensate themselves for the altered condition of the atmosphere. The 

 recovered gas (53.29 c.c.) was divided into two portions (A, B), and 

 treated separately. 



A = 43.20 c.c. (reduced) measured over water. This was treated 

 with potassic hydrate solution, but the volume remained unchanged. 

 Therefore carbon dioxide = zero. This volume (43.20 c.c.) was then 

 treated with pyrogallate, which reduced it to 38.88 c.c. The diminu- 

 tion, 4.32 cc, = 10% = oxygen. 



B. The average of three readings made this volume = 10.07 c.c. 

 Treated with pyrogallate, it was reduced to 9.12 c.c, the reduction in 

 volume, 0.95 c.c, being equivalent to 9.4+ % = oxygen. 



Experiment 3. — The gas for this experiment was secured by collecting 

 the bubbles as they were given off by the fishes. A number of the gar- 

 pikes were placed in an uncovered experimental jar containing recently 

 distilled water at + 20° C. It was found that by using a small glass fun- 

 nel, — the one employed was about 65 mm. in diameter, — held near the 

 surface of the water, the fishes were not prevented from the ordinary 

 movements accompanying the emission of bubbles, as they were when 

 confined under a larger funnel. The funnel was held so that the tip of 

 the snout, but not the gill region, projected beyond its edge. The 

 otherwise very slow process of collection is somewhat hastened by em- 

 ploying a number of fishes in the same jar. As they rise one after 

 another, at comparatively short intervals, the collector, anticipating 

 their movements, holds the funnel in the proper position, and seldom 

 fails to secure the desired bubbles. But even in this manner the 

 amount to be collected in the course of a few hours is not large. In 

 this experiment the (reduced) volume of gas collected in about two 

 hours was 9.54 c.c. Treated with potassic hydrate, the volume was 

 reduced to 9.38 cc. The difference = 0.16 c.c. = 1.7—% = carbon 

 dioxide. Treated with pyrogallate, the volume was further reduced to 

 8.20 c.c. The diminution, 1.18 cc, = 12.4-% = oxygen. 



For convenience of comparison the results of measurements may be 

 tabulated as follows ; — 



