GAS DISEASE IN FISHES. 853 



ature, for in its course th rough the gills it must be cooled to or nearly 

 to the temperature of the water. Its stream is too thin and it is too 

 intimately exposed to the water to maintain an appreciably higher 

 temperature. The blood, then, before it can return a second time to 

 the gills, undergoes a rise in temperature, and as the solvent power of 

 liquids decreases with increase of temperature, this rise tends toward 

 the release in gaseous form of some of the dissolved air. No doubt 

 the amount released is small per unit of time, but the free gas can not 

 be reabsorbed and the process of release is continuous, so that a fatal 

 embolism is only a question of time. 



This seems a fairl}^ satisfactory explanation of the means b}- which 

 the gas arrives free within the blood vessels. It requires the assump- 

 tion that' in water of normal condition with respect to dissolved air 

 the blood of fishes does not become completely saturated in the gills — 

 else gas would be thrown out constantly by the higher temperature of 

 the systemic circulation, w^hich is of course contrary to fact. There 

 is experimental proof that in mammals ordinary respiration does not 

 saturate the blood — that is, that all the oxygen which it is capable of 

 holding under the conditions does not enter it (L. Fredericq, 1896; 

 O. Hammarsten, 1901, 531). That the observation holds good for fishes 

 is extremely probable. It must further be assumed that under the 

 conditions of supersaturation existing in the Woods Hole water the 

 blood does take up all the air it will hold at its temperature in the 

 gills; or, if it falls short of this, that it takes up more than it can hold 

 at the maximum temperature to be encountered in its circulation 

 through the body. This latter supposition is the more probable and, 

 while no determihations support it, it is thoroughly in accord with 

 the facts and may be provisionall}^ accepted. 



temperature is not the sole cause which may play a part in the pre- 

 cipitation of the gas. For the separation of the solute, or dissolved 

 substance, from a supersaturated solvent, there nmst be a nucleus 

 about which the precipitating dissolved particles ma}" gather — an 

 excitant to start the process of precipitation. This is strikingh?^ 

 illustrated b}" supersaturated solutions of certain salts. A crj^stal of 

 the same salt as that dissolved when introduced into such a solution 

 will cause the immediate separation of this salt, which gathers about 

 the crystal as a nucleus. Likewise water may be heated, in a per- 

 fectly clean and smooth flask, above the boiling point without ebulli- 

 tion. If a solid foreign particle,*such as a fragment of pumice stone, 

 be dropped into the flask, boiling instanth^ begins. To apply this 

 principle to the present case, the minute floating corpuscles may be 

 considered as the nuclei for the separation of gas from blood, w hich 

 is supersaturated with it. The difference in temperature is the more 

 important and fundamental cause of the release of gas, while doubtless 

 the corpuscles at least provide loci for the change of state. 



F. C. 1904—23 



