]UG Thirty-Second Annual Meetiny 



the water spontaneously as soon as it emerges from the pipes. 

 It passes off insensibly at the surface but it also gathers in visible 

 bubbles on the sides of the tank and on the sides of the fishes, as 

 already described. A solid surface excites the release of gas from 

 a solution supersaturated Avith it, much as a crystal or foreign 

 body will cause precipitation from a supersaturated solution of 

 any readily soluble salt. The aquarium tank of water, holding 

 many gallons, will, if the flow is cut off, lose its excess of air, bui 

 it takes a number of days; two or three gallons, in a hatchery 

 jar, will lose it in two or three days ; a teaspoonf ul, probably in a 

 few minutes. If the water were warmed the escape of air would 

 be greatly facilitated. While the aquaria contain fishes and 

 there is a continuous flow of water, the supersaturation is con- 

 stant and nearly equal to that within the pipes. The spontan- 

 eous release is so small as to be negligible as far as the fishes are 

 concerned. 



These fishes find themselves in much the same situation as a 

 person who is subjected to a pressure of more than one atmos- 

 phere, as in a very deep mine, or as in the case of divers or work- 

 men in caissons in bridge building. In either case the breathing 

 apparatus has a task for which it is not adapted. The results 

 are more disastrous with fishes than with people. The gill fila- 

 ments of fishes are osmotic membranes, that is membranes which 

 allow substances in solution (in this case particularly gases) to 

 pass through them. The osmotic pressure is proportional to the 

 amount of gas in solution. With this water containing an ex- 

 cess of air, the osmotic pressure is high, higher than the fishes 

 experience in nature. The air passes rapidly into the blood and 

 tends to dissolve in it to the same degree of excess in which it is 

 present in the water. In other words, the osmotic pressiires on 

 the two sides of the gill membrane tend to equalize. The blood 

 as it streams through the gills becomes, like the water, super- 

 saturated with air, probably with nitrogen as well as with oxygen, 

 although the latter only is concerned in ordinary respiration. 

 But so far the air is still in solution and not free in the vessels. 

 What precipitates it? Two causes tend to this result, one the 

 presence of corpuscles, the other and probably more important 

 l)eing the higlier temperature of the systematic circulation of the 

 fish. While fishes are cold l^looded animals, thev nevertheless 



