334 Mr. W. Grant on the Curves of 



The measurements for the second set of curves were made ex- 

 perimentally in the same way as for the first set ; but as the 

 axes of the coils C and D were at right angles, there was a con- 

 siderable part of the field near the origin in which no measure- 

 ments could be obtained, on account of the coils coming into 

 contact with each other. 



The positive or negative induction corresponding to any one 

 of the plotted curves being equal to that between A and B 

 when placed at some definite distance from each other, it 

 became necessary, in order to place the curves so that the 

 numerical value of the induction-coefficient might have a 

 known relative for each, to determine the law according to 

 which the induction between the two coaxal coils A and B 

 varied with the distance between them. For this purpose the 

 coil A was connected with a contact-key and with a battery, 

 of 20 Grove's cells to form the primary circuit ; the secondary 

 circuit was formed by connecting the coil B with a contact- 

 key and with the coils C and D, which were placed close 

 together and used as a galvanometer. The needle and mirror 

 employed were rather heavy: this had the advantage of allow- 

 ing the induced currents to exert their full effect on the needle 

 before it had moved perceptibly from its position of equili- 

 brium. The key was a double-contact key, so arranged that 

 on depressing the key the primary circuit was completed; and 

 immediately afterwards, when the current was established, the 

 secondary circuit, which was already complete, was opened, 

 and was kept open by a spring acting on the lower key, while 

 the primary circuit was then broken. The time occupied in 

 depressing and releasing the key was in general about one 

 third of a second ; hence the time during which the primary 

 current circulated every time that contact was made was ap- 

 proximately that period. With this arrangement only the 

 induced current on making was allowed to act on the galva- 

 nometer, while that on breaking was prevented from circulating 

 in the secondary circuit. 



The temperature of the primary coil was indicated by a de- 

 licate thermometer divided to tenths of a degree Centigrade, 

 which had its bulb placed in contact with the silk covering of 

 the coil. The variation of the temperature of the coil, due to 

 the heating effect of the current during the experiments (as 

 indicated by this thermometer) was approximately l o, C. 

 The scale -readings obtained at the highest and lowest tempe- 

 ratures were found to agree so closely that it was unnecessary 

 to make any correction on account of the variation of the 

 temperature of the primary coil: its mean temperature during 

 the experiments was 11 0, 5 C. 



