444 proceedinctS of the American academy 



other disturbing cause. Accordingly, the current would become zero 

 at two points midway between its two maxima, and this would be the 

 point where the commutator should be placed. In that case, no S[)ark 

 would be seen at the commutator, since the circuit would be broken 

 only when the current was zero. In practice, it was found that there 

 was no portion of the commutator where the spark could be entirely 

 avoided when the I'esistance was small, evidently owing to the fact 

 shown by these observations, that the current at no point is zero. 

 Moreover, on constructing tlie curves with coordinates equal to the 

 angles and currents, it will be seen that the inclination is much gi-eater 

 before than after the maximum ; so that the latter, as stated above, is 

 distant only about 55'^ from the minimum, instead of 90''. The cause 

 of the deviation from the curves of sines is probably the current in- 

 duced by the magnet, which adds or subtracts its elFect according as 

 the current is increasing or dimiuishinor. 



In trying experiments upon the light produced by the current, seve- 

 ral difficulties were encountered. One of the most serious of these 

 was from the slipping of the driving belts, when the machine was run- 

 ning at high rates of speed and the circuit was made through so small 

 a resistance as the regulator and light. From this cause, we were 

 unable to obtain a steady speed pf more than, 1,300 revolutions per 

 minute, which was not sufficient to give the best results. A further 

 difficulty was experienced from the great difference in power required 

 to run the machine when the current was passing, and when the car- 

 bons became so far separated that the current was unable to pass. A 

 change of probably 4 or 5 horse-power was thus almost instantly made, 

 whenever the current was made or broken, and the consequent shock 

 upon the machinery was very great. It also appeared that the form 

 of regulator used (Duboscq's) was not capable of controlling the cur- 

 rent so that the light should be steady. When the carbons were 

 brought in contact, the current was so great that the magnet acted 

 strongly, starting the reversing clock-work and separating them half an 

 inch or more. This broke the circuit, and the machine began to re- 

 volve very rapidly ; soon the carbons were brought together, throwing 

 a great strain on the engine, and thus they oscillated, producing a very 

 bright light for an instant and then extinguishing it. Better results 

 would i^robably be attained without the reversing arrangement, by a 

 change in the magnet of the regulator, or by increasing the electro- 

 motive force of the current. Some results were however obtained by 

 a very careful adjustment of the spring holding the armature. AVith a 

 velocity of 1,130 revolutions, a tolerably constant light was obtained. 



