Nov. 22, 191S Improved Respiration Calorimeter 337 



The temperature of the zinc wall is kept as nearly as possible like that 

 of the copper wall, so that the deflections of the galvanometer connected 

 with the differential thermoelements in the walls are as close as possible 

 to o. Even under the most favorable conditions it is hardly practicable 

 to keep the two walls so uniformly alike that there will be no deflection 

 at any time, because the temperature of the copper wall, however well 

 regulated, does vary to some extent, and it is not possible to anticipate 

 the change. It is possible, however, to keep the deflections most of the 

 time so small that any error introduced by the temperature differences 

 which they indicate would be insignificant. As explained on page 334, 

 the number of thermoelements in each section of the walls and the sen- 

 sitivity of the galvanometer are such that a very small difference between 

 the temperature of the copper wall and that of the zinc wall would cause 

 a fairly large deflection; hence, a very small deflection really means a 

 practical identity of temperature of the two walls. When the rate of 

 production of heat within the chamber is quite uniform and the rate of 

 abstraction of heat is so nearly like it that the temperature conditions 

 within the chamber are quite constant, the temperature of the zinc wall 

 may be kept so nearly like that of the copper wall that the deflection 

 will not exceed 5 mm. and will generally be less. A deflection of that 

 magnitude would indicate for the bottom section a difference not greater 

 than 0.005 degree between the average temperature of the copper floor of 

 the chamber and that of the zinc wall outside of it ; for the other sections 

 it would indicate still smaller differences. The amount of heat gained 

 by either wall from the other with such small differences is of little im- 

 portance in comparison with the total amounts usually measured in the 

 chamber. In an experiment with a variable heat production, as would 

 be the case with a man moving and quiet by turns, such a close balance 

 could hardly be maintained at all times, though the deviation need not 

 greatly exceed 5 mm. for any considerable periods. Furthermore, it is 

 possible to make the deflections in one direction equal to those in the 

 other direction for equal short periods, so that whatever heat may be 

 gained by the copper wall from the zinc wall during one period is counter- 

 balanced by that gained by the zinc wall from the copper wall during the 

 succeeding period, in which case there is no actual increase or decrease 

 of the quantity of heat in the chamber for the total time of the two 

 periods due to an exchange of heat between the walls. 



In order that the walls controlled in the manner described shall be 

 heatproof, their temperature and that of the iron structure between 

 them must be the same. The temperature of the copper wall, and con- 

 sequently that of the zinc wall, is governed by that of the air in the 

 chamber; but the two walls may be brought into thermal equilibrium 

 at a temperature above or below that of the framework, in which case 

 the quantity of heat in the chamber would probably be affected by the 



