accompanying the Electric Spark. 293 



and capable of giving a small spark of about 1| inch long. 

 When coupled with the Leyden jars, of course, the length of 

 spark was much diminished. In no case with either coil was 

 a longer spark employed than *75 inch. No change in the 

 appearance, regularity, or shape of the figures was caused by 

 replacing one coil by another. 



The effect of a difference in condensing-power was tried 

 by interposing more jars. Except that the vigour of the 

 spark caused a larger area of cleared space, no alteration could 

 be observed. This is, of course, confirmed by the fact of the 

 different coils, possessing condensers of different areas, giving 

 identical results. 



5. Influence of Contact- Breaker. — It might be supposed, as 

 w T as done by Grove in reference to the stratifications in partial 

 vacuo, that the intermittent character of the currents flowing 

 in the secondary wire caused by the interruptions in the 

 primary-coil by the contact-breaker w T ould have some influence 

 in producing or regulating the curves. This, however, is not 

 the case. The shapes did not alter no matter what coil was 

 employed, and when the mercury-break on the large coil was 

 used in place of the spring, no alteration in the resulting 

 curves was produced. Moreover the curves can be obtained 

 without the use of a coil with its necessary contact-breaker 

 at all. For this purpose it is only necessary to connect each 

 electrode of a Wimshurst machine to one or two Leyden jars, 

 and to allow the highly condensed spark thus obtained to act 

 upon the powder. In fact I have been able to produce the 

 curves in this way without employing any jars at all. With 

 the machine, however, they are far inferior in regularity to 

 those produced by the use of the coil. With a 17-inch plate 

 the best results have been obtained with jars having a total 

 area of foil of about 1500 square inches, a spark of about 3" 

 long, and the powder 6" below spark. 



6. Suture of Electrodes. — Varying the material forming 

 the points between which the sparks were taken made no 

 difference in the resulting shapes. Thus the same effects are 

 produced by using brass, iron, or carbon points. 



7. Influence of the Powdered Substance. — Soon after the 

 first results were obtained and upon trying different sub- 

 stances, it became evident that, although accidental irregu- 

 larities occurred, yet, upon the whole, the curves were 

 remarkably regular. It was also very soon noticed that the 

 frequency with which they followed each other altered with 

 an alteration of the substance of which they were composed. 

 This induced a desire to compare by measurement the dis- 

 tances between each successive curve. A large number of 



