THE CALORIMETER SYSTEM AND MEASUREMENT OP HEAT. 115 



which is dependent upon the difference in temperature. The intensity 

 of the current can be accurately noted with a delicate galvanometer, 

 and consequently the difference in temperature closely measured. Ob- 

 viously, in the case of noting the difference in temperature between 

 the copper and the zinc walls, it is necessary not only that the thermal 

 junctions be in the best possible thermal contact with the correspond- 

 ing metal walls, but also that there must be absolutely no electrical 

 contact to impair the accuracy of the measurements. The grooves in 

 the wooden rod are sunk sufficiently deep to bring the ends of the 

 junctions considerably below the surface of the wooden cylinder. As 

 will be seen on studying the method of installing the junctions, this 

 position of the ends of the junctions is necessary to secure electrical 

 insulation from the copper and zinc walls of the calorimeter. 



Distribution of elements. As it is desired to keep the temperature of 

 the zinc wall exactly the same as that of the copper wall, the elements 

 should be distributed over the whole six sides of the chamber in such a 

 manner that each one may assume the average temperature of a certain 

 portion of the total area of metal ; that is, the elements should be dis- 

 tributed all over the area of the calorimeter in points very closely pro- 

 portional to the area. As a matter of fact, the junctions are located 

 about 48 cm. apart, and each exercises the temperature control of an 

 area of metal about 48 by 48 cm. The distribution of the various ele- 

 ments in the rear end of the chamber is seen in figure 25, which shows 

 the respiration chamber inside the wooden house, the rear panels of 

 which have been removed. In this figure the location of ten elements 

 is distinctly shown. The wires attached to the calorimeter in a " zigzag 

 fashion ' ' are connected to elements at each point where the direction 

 of the wire is changed. Only two elements appear in the lowest zone, 

 and here the connecting wires do not take a zigzag course. 



In figure 29 a view of the distribution of the elements on one side of 

 the calorimeter is shown. There are here, as on the end of the calo- 

 rimeter, two zigzag rows of fourteen elements and a lower row of 

 three elements in a straight line. The number and distribution of the 

 elements on the wall of the metal chamber opposite to that shown in 

 figure 29 is precisely the same. Owing to difficulties in photographing 

 the apparatus, no figure is given showing the distribution of the eleven 

 elements on the top or the eleven on the bottom of the calorimeter. The 

 nine elements on the front end of the chamber are shown in figure 30. 

 There is also one additional element in a position that does not appear 

 in any of the photographs. This makes in all seventy-six elements, 

 each containing four pairs of thermal junctions. 



