PROCEEDINGS OF UNITED STATES NATIONAL MUSEUM. 317 



First. All fishes develope animal heat, its quantity varying according 

 to the organization rather than the habits of the family. 



Second. This heat results from the processes of nutrition (chemical) 

 and from the conversion of muscular motion (mechanical). That result- 

 ing from the oxidation of the blood is lost in the gills ; hence the venous 

 blood is decidedly warmer than the arterial. 



Third. Spawning and breeding fishes develope more heat than those 

 not carrying on these j^rocesses. 



Fourth. Elasmobranchs and, generally, fishes with a highly differen- 

 tiated digestive apparatus develoj)e more heat than those of simpler 

 organization, and (probably) very active surface fishes more than slug- 

 gish bottom fishes. 



Fifth. The intestinal canal and arterial blood do not correctly indicate 

 the animal heat of fishes. 



Sixth. The question of "normal range of temperature" remains un- 

 answered. 



SUGGESTIONS. 



Should you think it desirable to continue this investigation I would 

 suggest that the inquiry include the following details, indicated by last 

 summer's experiments : 



First. The range of temperature through which living fishes may be 

 carried. This miglit be observed by subjecting different species in 

 tanks to varying temperatures produced by ice or steam introduced into 

 the water, and noting the body temperature of the fish when it begins 

 to show signs of distress. Each experiment would expend a fish, but 

 the importance and practical bearing of this question of ''normal range" 

 of temperature might justify the expense. Much could be learned by 

 observing the temperature of the water at which the fish begins to show 

 signs of distress. In such a harbor as that of Provincetowu a consider- 

 able difference in the temi^erature of the water can be got by towing a 

 wooden-latticed tank into shallow water at ebb-tide and into suitable 

 positions at flood-tide. 



Second. Amount of oxygen required by tlifferent fishes. This may be 

 approximated by keeping different species under observation in separate 

 tanks without a fresh su])ply of water. 



Third. Length of life sifter being withdrawn from the water, and sub- 

 sequent duration of muscular irritability. Also the number of respira- 

 tions per minute in different species when at rest. 



Fourth. Influence of muscular 7novements on temperature. This may 

 be observed by tying a fish by the tail, in the water, until it exhausts 

 itself by struggling, and then taking the temperature, to be compared 

 with an observation upon another individual of the same species under 

 similar circumstances, but at rest. 



Fifth. Comparative activity of nutrition as indicated by the percent- 

 age of nitrogenous matter in the excreta. 



Sixth. Tlie rei)etition of similar observations on the plan of those 

 made last summer. 



