A STUDY IN MO R DID AM) X () i; M A I. P II Y S I () I, (J V. 



19 



were examined not only for moisture but also for carbonic acid gas, so that the 

 cliniiuation of carbonic acid was studied along with the development of animal heat. 



There is, of course, loss of heat from the most perfect calorimeter. At tirst sight 

 this may appear of no importance in relative experiments. Experience, however, has 

 shown that the instrument, unless used with great care, yields the most fallacious 

 results. In testing the apparatus, of the two non-conducting packing materials em- 

 ployed, fine, perfectly dry sawdust seemed to yield the best results, lint even with 

 it, especially when the water was much above the temperature of tli(^ air, tlie loss of 

 heat was more or less irregular, unless great care was employetl. The first precaution 

 to be taken is to have the apparatus itself of a uniform temperature throughout; 

 so that water, iron, and sawdust should all be one in tliis respect. Over-warming 

 the apparatus previous to use, or using it when it is decidedly below tlie temper- 

 ature of the water, is equally fatal to accuracy. 



The amount of loss from the calorimeter is of course directly dependent upon the 

 relation between its temperature and that of the external air ; the greater the 

 difference the greater the loss. 



The question very early arose as to whether the temperature of the water should 

 approximate that of the air or that of the animal. A trial soon showed that when the 

 water was near the temperature of the animal, the loss of heat from the calorimeter 

 was not only enormous but also very irregular on account of the <lifhculty of heat- 

 ing the apparatus iniiformly through to such a temperature. Again, an animal 

 taken out of an ordinary room and thrust into a confined chandjer heatetl to 1()()° F. 

 or thereabouts, suffers violence in its environment, and is put under such unnatural 

 conditions as to vitiate more or less the result. The correct use of the calorimeter 

 evidently depends upon the keeping of its temperature equable, and as near as may 

 be to that of the external air. It should always, in winter, be used in a room heated 

 steadily to 70° F. 



In order to show in a measure the loss of heat from the apiiaratus, the follow- 

 ing series of trials were made with the two calorimeters with which nearly all the 

 work of the paper was performed. In these trials the temperature of the water 

 was a little above that of the external air, and the air was drawn through the inner 

 box at a steady rate. 



Trial No. 1. — Larrintl Calorimeter. 



' The apiia.ratiis had lieen runtiiii? several hours l)cfoi-e at a vorv hi.crh temperature, 100°, and 

 though effort was made to get it uniformlv cooled, tlio amount of loss may have licou all'eeted. 



