﻿of Electricity on Lou Wires. 433 



swing was, under like conditions, satisfactorily constant; and 

 a preliminary calibration of the instrument by passing currents 

 of known strength through it, showed that the square root of 

 the deflexion taken in this manner was in a constant ratio to 

 the current. 



The same copper wire (diameter 0*1201 centim.) that had 

 been used in the secondary without end capacities was used 

 for the rectangle KLMN in fig. 2. and, with the exploring 

 terminals close to L and N, the maximum point was found by 

 Lengthening and shortening the wire. Bolometer-readings 

 were taken for each length used, lo insure the constant 

 activity of the spark, a convenient length was taken as a point 

 of reference, and observations taken at this point before and 

 after a series of readings. If the spark had remained con- 

 stant, the readings were retained. A maximum point was 

 found when KL was 818 centim. The sides were fixed at 

 this length, and the form of the wave was obtained by sliding 

 the exploring terminals along the wire, and taking bolometer 

 observations for each position. The result is shown in fig. 3 

 of PI. XII. The critical points were determined several 

 times, and the steadiness of the spark insured by choosing, as 

 before, a point of reference. The curve shows three minima 

 at 240, 496, and 752 centim., starting from 0. These give 

 half wave-lengths of 2,55, 256, and 256 centim., with an average 

 of 255*6 centim. The third minimum at 752 centim., was 

 determined with care, as it was to be used as a basis for calcu- 

 lating the half wave-length. An error in determining the 

 position of this minimum would be divided by three, since 

 the distance from to this minimum was three half wave- 

 lengths. The total length of the circuit was seven half wave- 

 lengths, and it was the equivalent of one fourth of a wave- 

 length from the third minimum to the end. The actual 

 distance to the end was 818 — 752 = 66 centim. 127*8 — 66 = 

 61*8 centim. = the equivalent of the capacities in centimetres 

 of wire. 



A comparison of the curve (fig. 1, PI. XII.) obtained from 

 the plain wire circuit with the curve (fig. 3) obtained when 

 capacities were fixed on the free ends, shows a quite satis- 

 factory agreement, which tends to create confidence in both 

 methods. The half wave-length by the first is 254*3 centim. 

 and by the second it is 255*6, values which differ by about 

 one-half of 1 per cent. There is a marked difference, as 

 was to be expected, in the form of the curve next the free 

 ends. When end capacities were used, the accumulation of 

 charges seemed largely confined to them, out of reach of the 



