362 PROCEEDINGS OF THE AMERICAN ACADEMY. 



calculated were used. To correct the ammeters for currents from 0.72 

 ampere to 2.2 amperes, four turns' of wire were wound ou the original 

 layer of the tangent galvanometer. As these turns did not fill the 

 breadth of the channels, they were kept in position by wedges of cork 

 placed at frequent distances around the circumferences. Since they 

 could not be made to lie as evenly as the 22 turns which filled the 

 breadth of the channel, the constant for these four turns was deter- 

 mined from the value of H^ measured as described below, by passing 

 a current through two copper voltameters in series with the galvanometer. 

 Two measurements of this constant, made with the copper voltameter, 

 using the value 0.1630 for H, gave 1.1920 and 1.192G respectively. An 

 attempt was made to determine the constant for the four turns from the 

 constant for the 22 turns, as calculated from the dimensions, as follows. 

 A current was sent through the 22 turns and the four turns in series, 

 first in such a way that their effects on the needle agreed, and second in 

 such a way that they were opposed. If G^ and G<._ be the respective 

 constants, 



II tan 6 = y (Ci + G.2) for the first case, 

 and II tan 6' = y ((?! — G.2) for the second case, 



where and 6' are the respective deflections, and y is the current in 

 c. g. s. units. 

 Hence 



+ 



ry /-Y w f tan 6 — tan 6' ) 

 I. tan 6* + tan 6' 3 



The values of and 6' were 47° 6' and 36° 13'. G^, calculated as 

 described below for the 22 turns, was 6.294. Hence (?o = 1.196, 

 which agrees well with the value for G^ given above, considering the 

 small difference between 6 and 0' and the possible error in the value of 

 ^in the previous determinations. 



This constant having been determined, the ammeters could be calibrated 

 for currents from 0.72 amperes up to 1.65 amperes for the Siemens and 

 Halske ammeter, and to 2.2 amperes for the Weston. The Weston 

 ammeter was provided with three shunts, the upper limit of the strength 

 of current which could be measured being 0.1 ampere, 1 ampere, and 10 

 amperes in the three cases. The Siemens and Halske instrument was 

 a milliamperemeter of 1 ohm resistance, reading up to 150 milliam- 

 peres. A shunt, of 0.1 ohm resistance, was made of manganin wire. 

 The lead-wires were soldered on the terminals of this shinit coil. By 

 means of this shunt, therefore, the range was extended to 1.65 amperes. 



