526 PROCEEDINGS OF THE AMERICAN ACADEMY. 



vapor passing through a tube similarly encircled by hot wire must be 

 free from superheating. 



The calorimeter, of about 1.5 liters capacity, was made of thin nickel- 

 plated sheet copper, highly burnished on the outside. It was almost 

 filled with water, and is shown in Figure 5 (D, D). The condenser (A) 

 within the calorimeter was constructed of block tin, the joints being 

 soldered by tin only ; it consisted of a spiral tube, 1 meter long and 

 3 millimeters in internal diameter, coiled in four turns, with a tin 

 cylinder 10 centimeters long and 3 centimeters in diameter at the 

 bottom to serve as a receptacle for the condensed liquid. The tin 

 cyHnder had an outlet tube, leading directly up to the air of the room. 

 The tin coil and cylinder together weighed 436.7 grams. The outlet 

 tube or beak of the vaporizer was attached to the worm by a short 

 piece of pure rubber tubing (B). Various preliminary experiments 

 were made as to the position of this joint in relation to the water of 

 the calorimeter, but the details need not be given. Finally, the 

 arrangements shown in the figure was adopted ; the silvered jacket of 

 the vaporizer was immersed to the depth of about a centimeter under 

 the water. The water-line was thus protected from heat by the 

 vacuum jacket, and abnormal evaporation and cooling at this point 

 were therefore avoided. 



The stirrer (C) within the calorimeter was of the propeller pattern, 

 having six blades, each 1 centimeter long. It was made of copper, and, 

 in order to prevent loss of heat by conduction, extended only to the 

 surface of the water, where it joined a shaft of hard rubber to which 

 was attached the driving mechanism. The scarcely perceptible evolu- 

 tion of heat from the stirrer, being directly proportional to the time of 

 the experiment, was included with the warming due to the proximity 

 of the vaporizer in a single time-correction, and thus eliminated from 

 the result. 



The heat capacity of the solid parts of the calorimeter was equiva- 

 lent to 53.4 grams of water, the separate parts amounting to the follow- 

 ing quantities. The finished copper calorimetric vessel weighed 299.95 

 grams. It was "plated" inside and out with a thin electrolytic film 

 of burnished nickel, which has a specific heat so near that of copper as 

 to cause no appreciable difference in the heat capacity of the whole. 

 The copper vessel was soldered with about 4 grams of solder, the solder 

 consisting half of lead and half of tin, and having therefore a specific 

 heat of about half that of copper. The calorimeter vessel with its 

 nickel and solder was therefore approximately equivalent to the pure 

 copper vessel, weighing 298 grams, and had a heat capacity equivalent 

 to 27.7 grams of water, the specific heat of copper at 21° being about 



