360 PROCEEDINGS OF THE AMERICAN ACADEMY. 



Two measurements of the resistance of the galvanometer were taken 

 with an interval of four months. The first gave 2.243 ohms at 17.2° C. ; 

 the second, 2.223 ohms at 16.3°C. The difference between these values 

 when correction is made for difference of temperature is 0.011 ohms. 

 The second value was used, that having been taken during the period 

 when the determinations of the Hall effect were made. During these 

 determinations a thermometer was kept with its bulb in the interior of 

 the galvanometer case, and the temperature recorded. In making cor- 

 rection for the temperature, the value 0.0039 was used as the tempera- 

 ture coefficient of the copper wire of the coils. 



The resistance of the remainder of the secondary circuit was measured 

 frequently, and, in the case where the effect in gold leaf with weak cur- 

 rents was being determined, before and after each measurement of the 

 transverse current. The measurement of this resistance was also made 

 on the Carey Foster bridge. With gold of thickness of 0.0002 cm. 

 and above, the variation of the resistance of the gold and of the connect- 

 ing wires outside the galvanometer did not amount to more than 0.03 

 ohms. The variations of the resistance in the gold leaf will be referred 

 to below. In these resistance measurements much use was made of a 

 box of coils made by Morris E. Leeds and Co. of Philadelphia. The 

 temperature of these coils was noted. They were compared with standard 

 coils. A measurement of the temperature coefficient of the Leeds coils 

 gave + 0.03 per cent per degree centigrade. The other coils used were 

 manganin coils, whose temperature coefficient was given by Messrs. 

 Persons and Coulson as 0.00002 per degree centigrade. 



Measurement of Sensitiveness of Galvanometer. 



The sensitiveness of the astatic galvanometer was in almost every case 

 determined before and after each set of consecutive observations of the 

 transverse effect. The following diagram (Figure 3) will explain the 

 arrangement for the measurement of the sensitiveness. 



The resistances whose values are given in this figure were standard 

 resistances. Connections were made between them by means of mercury 

 cups set in heavy bronze blocks. A double commutator was used to re- 

 verse the current from the storage cell and that from the Carhart-Clark 

 cell simultaneously. 



The certificate of the Carhart-Clark cell used as a standard dated 

 from 1896. A similar cell, which had been put up in the same case 

 with it, at the same time, agreed with it to about one-tenth of one per 

 cent. On comparison with a cadmium cell and another Carhart-Clark 



