216 ROSA, DORSEY AND MILLER: INTERNATIONAL AMPERE 



of the galvanometer constants of a small fixed coil (SI) and a moving 

 coil (M3). The values given have not been corrected for the length of 

 the needle but have had small corrections applied for temperature varia- 

 tions and the heating effect of the measuring current. 



The deviations from the mean in parts in a million are given in the 

 third column. The average deviation is 1.4 parts in a million. 



One of the features of the current balance which contributed 

 very strongly to the attainment of a high accuracy is the water 

 cooling system. In each of the fixed coils there is a channel 

 back of the windings through which water can be circulated to 

 carry away the heat generated by the electrical current. The 

 supply of water was held at a constant temperature by means of 

 a thermostat, thus maintaining the coils at a constant tempera- 

 ture and radius. Water circulated thru the fixed coils while 

 the constant was being determined and when weighings of the 

 force were being made. During the latter experiment the mov- 

 ing coil was surrounded by a copper water jacket thru which water 

 from the same source was circulated. This served the purpose of 

 carrying away the heat generated in the moving coil and also 

 made constant the convection currents of air which were set up 

 by the heating of the coil. The possible accuracy in the weigh- 

 ing was thereby considerably increased. By the method of weigh- 

 ing the constant lifting force of the currents of air was eliminated. 



It was found important to measure the temperatures of the 

 coils when being used. In the determination of the constant and 

 the weighing, the temperature coefficient of the apparatus is zero 

 since only the ratios of the dimensions enter; provided all portions 

 are made of the same material and are at the same temperature. 

 Since, however, heat is being generated in some portions and not 

 in other portions, it is evident that the whole apparatus will not 

 be at the same temperature and that the temperature differences 

 will depend upon the quantity of heat generated in the various 

 parts. Hence in this work the temperatures of the coils were 

 measured by measuring the resistance of the windings and cor- 

 rections were applied to bring all of the parts to the same temper- 

 ature. 



The rest points of the balance were read by means of a tele- 



