GAS DISEASE IN FISHES. 357 



atmosphere. But in this mechanically induced supersaturation frag- 

 ments of the atmosphere are forced bodily into solution in their 

 entiret}", and the dissolved content is increased b}- nitrogen and oxy- 

 gen in atmospheric proportions, 79+21, instead of in dissolved pro- 

 portions, 67+33. When the excess of these two gases escapes spon- 

 taneously from the water the oxygen has about the atmospheric 

 relation to the nitrogen, i. e. , about 21 per cent of the total, notwith- 

 standing that while in solution the oxygen is more than 30 per cent of 

 the total of these two. In other words the excess goes in as air and 

 comes out as air. Thus the actual analyses alread}^ cited (p. 35-1) of 

 precipitated gas from Woods Hole water, show the proportion of oxy- 

 gen to be about as in air. 



Since the blood does not release its supersaturation in this wa}', it is 

 at once suggested that the hemoglobin capacity for oxygen modifies 

 the effect of the water so far as the supersaturation with oxygen is 

 concerned. It would appear that the corpuscles can take up more 

 than the usual amount of oxygen and that the increment is not thrown 

 out by the rise in temperature. It remains to studj^ experimentally 

 the effect upon fishes of water in which the supersaturation is with 

 oxygen alone. Some evidence is afforded by an instance of such a 

 supersaturation, naturally occurring, in a pond containing trout. At 

 the Cold Spring Harbor Station of the New York Forest, Fish, and 

 Game Commission, the springs which chiefl}^ supply the station make 

 immediately a shallow pond of considerable size. In the spring of 

 1901 the bottom of this pond became heavil.y overgrown with green 

 algRj, chieily with a species of Sjylrogyra. Presumably from these 

 alga3, the water about the middle of the pond acquired an excess of 

 oxygen of 3 c. c. per liter, while the nitrogen content remained nor- 

 mal, or but slightly in excess. Remote portions of the pond were 

 normal in oxN'gen. Large trout lived in it in good condition and 

 showed no gas symptoms, but the fact lacks conclusiveness since they 

 had access to normal water, which they doubtless frequented. It is 

 probable, however, that a large excess of oxygen is required to pro- 

 duce untoward results from this gas alone. In the conditions at Woods 

 Hole, while the excess was of both ox3'gen and nitrogen, it is prob- 

 able that the damage was done b}^ the latter gas alone. 



RELATION OF GAS DISEASE TO TEMPERATURE AND PRESSURE. 



When water is here described as containing an excess of air, or an 

 excess of oxygen or nitrogen, a definite relation of the quantity of gas 

 to temperature and pressure is of course connoted. It is hardly 

 necessary to insist that dissolved gas only is referred to, for loose 

 bubbles present are not really in the water, though they may be 

 beneath its surface or within its volume. The gas-disease process, 

 then, bears an intimate relation to temperature and pressure. If a 



