PEIRCE. — CORRECTION FOR COUNTER ELECTROMOTIVE FORCE. 163 



the solenoid and decreasing the resistance of P so that the throws could 

 be determined. For the puri)ose of this experiment it was allowable to 

 shunt P, and when a one ninth shunt was used, the " throw " caused by 

 the reversal of a certain current in the solenoid was 73.9 mms. or 

 45.0 mms. according as the suspended coil was clamped or free. 



(B) A certain d' Arson val ballistic galvanometer the total resistance 

 of which at the room temperature was almost exactly 123.7 B. A. units, 

 was connected in series with a rheostat and an earth inductor of 11.7 

 B. A. units resistance, the frame of which was fixed at a certain station. 

 The moving coil of the instrument was furnished with a horizontal wire 

 which projected about 23 mms. beyond the coil on each side and car- 

 ried on each end a brass weight. The object of this device was to 

 increase materially the moment of inertia of the suspended system and 

 thus to make the time of swing larger without increasing the already 

 inconveniently great sensitiveness of the galvanometer. When the 

 inductor ring, which lay in a vertical plane perpendicular to the merid- 

 ian, was reversed, the galvanometer throw, as measured on the scale 

 at a distance of 120 cms. from the mirror, was 9.41 cms., 5.51 cms., or 

 3.95 cms., according as the resistance of the rheostat in B. A. units 

 was 0, 100, or 200 : if no account were taken of the counter electro- 

 motive force induced in the suspended coil, these figures would point 

 to a circuit resistance of something more than 141 units instead of 

 135.4 units. In the condition just described the periodic time of the 

 suspended system was 34.5 seconds : when that time was reduced to 

 12.4 seconds by the removal of the weights attached to the horizontal 

 wire, the throw caused by reversing the inductor ring was 12.98, 7.72, 

 or 5.49 according as the rheostat resistance was 100, 300, or 500 ; these 

 figures point to an apparent resistance of 193 units in the circuit or 

 42 per cent more than the real resistance. 



(C) A certain extremely sensitive d'Arsonval ballistic galvanometer, 

 the coil of which was about twice as long as is usual in such in- 

 struments, was connected in series with a rheostat and a "magneto- 

 inductor," i. e., a short coil of very fine msulated wire wound on a 

 spool which could be slipped between near, fixed stops on a long mag- 

 netized rod. The permanent magnet of the galvanometer was built 

 up of a large number of thin horseshoe plates hardened separately. 

 The resistance of the galvanometer current when the rheostat resist- 

 ance was zero was about 8691 at room temperature. When the ex- 

 ploring coil was slipped from one stop to the other, the throw of the 

 galvanometer, as measured on the scale, was 67.0, 58.1, 51.3, 48.3, or 

 40.05, according as the resistance of the rheostat was 0, 2000, 4000, 

 5000, or 8500. Of course it is not possible to compute the apparent 



