438 PROCEEDINGS OF THE AMERICAN ACADEMY 



gives i?i .853 G, .75 G, and .853 G, and *^.853, .125, and .0244. Finally 

 »i = 10 gives ^^.916 6', .846 G, and .916 G; and G^ .916, .076, .0073. 

 It will be more convenient in general to give r, r", s', s", approxi- 

 mately the proper values, and then measure B^, R^, and ^g, by the 

 Wheatstone's bridge. Next interpose resistances in the wires attached 

 to the switch, so as to make the total resistance of the galvanometer 

 the same for all positions of the switch. Thus, in the above example, 

 when «= 10, if resistances of .084, .154, and ,084 ohms are inter- 

 posed, the galvanometer resistance will be an ohm in each case. The 

 values of Sj, *S„ and S^ are now found directly by comparing the de- 

 flections when the switch is moved. By this device, the range of a 

 tangent galvanometer may be increased indefinitely, and the strongest 

 as well as weak currents measured by it. Moreover, the resistance is 

 not altered, so that readuigs with different shunts are directly com- 

 parable. 



The first experiments made with the machine were for the purpose 

 of determining whether the current was constant under the same cir- 

 cumstances or not. It was feared that, as the magnetism was induced 

 by the current itself, variations would appear, dependent on the time 

 during which the circuit had been closed; but, on trial, it was found 

 that the magnet attained its full polarity sooner than the needle of 

 the galvanometer came to rest, and that, on making and breaking 

 the circuit, the successive deflections were almost precisely equal. The 

 next problem was to determine the effect upon the current of changing 

 the position of the commutator. This is so made as to be capable of 

 being revolved round the shaft of the Siemens' armature through an 

 angle of about 45^, thus taking off the current when the coil of the 

 armature is in different positions relatively to the electro-magnet. 

 Observations were taken with the commutator in the following eight 

 different positions : No. 1 is with the commutator turned farthest 

 down, or with its plane as nearly parallel with the plane of the electro- 

 magnets as possible. It is then turned up through an angle of about 

 6.5'* with each succeeding number. In No. 8, it is very nearly per- 

 pendicular to the i)laiie of the magnets. The results of several ex- 

 periments are given in Table I., in the 2d, 3d, 4th, and oth columns, 

 of which the current obtained in the various positions is given in 

 vebers per second. In the last four columns, the currents are given in 

 percentages of the maximum obtained. 



