168 PEOCEEDINGS OF THE AMERICAN ACADEMY. 



others gives the four values 1613, 1613, 1612, 1610 for the apparent 

 resistance of the galvanometer circuit. These results, which agree 

 wonderfully with each other considering the slight uncertainty in the 

 readings, differ widely from the true resistance of the circuit, which 

 was 531. If, when the rheostat resistance was zero, the galvanometer 

 coil had been held still, the whole quantity of electricity set in motion 

 in the circuit when the inductor coil was moved between its stops 

 would have been more than three times the actual flux (rather less 

 than 5 microcoulombs) when the coil was allowed to swing. This fact 

 illustrates the magnitude of the error which might be made in a per- 

 meability measurement if the true resistance of the circuit instead of 

 the apparent resistance were used. As is evident from the numbers 

 given above, the apparent resistance is nearly independent of the 

 actual resistance in the circuit when the impulse is constant and 

 comes from a given source, but when the character of the impulse is 

 changed the apparent resistance changes slightly also sometimes, and 

 in a manner to be determined for each instrument by experiment. 

 For a certain impulse which caused a throw of 8.52 cm., as indicated on 

 the scale of the galvanometer just mentioned, the apparent resistance 

 of the circuit was about 1704. 



In such ballistic magnetic measurements as allow more than one 

 observation at each stage of the experiment, it is possible, by alter- 

 nately introducing and taking out of the circuit a known resistance 

 of suitable value wound non-inductively, to determine the apparent 

 resistance of the galvanometer circuit for each impulse. In other 

 cases it is feasible to put into the circuit once for all such a large 

 resistance, say 200,000 ohms, that changes in the apparent resistance 

 of the circuit and changes in the inductive impulses become relatively 

 unimportant. So far as my experience goes, a given instrument, as 

 was to be expected, holds its apparent resistance for a given impulse 

 unchanged for an indefinite time. 



If a condenser discharge be sent through a d'Arsonval galvanometer 

 which has no electromagnetic damping other than that which accom- 

 panies currents in its own coil, the form of the curve which shows 

 the position of the swinging coil as a function of the time is hardly 

 affected at all by the introduction of a large inductance or a resistance 

 of several thousand ohms into the discharging circuit ; but if the con- 

 denser be displaced by a magneto-inductor of small resistance, the 

 damping is much increased and the shape of the curve is much altered. 

 Figure 1, in which the horizontal divisions represent seconds, shows 

 a careful copy (MOPQN) on a small scale of a large photographic 

 record of the excursion of the coil of a certain d'Arsonval galvanometer 



