1 8^ Lord Armstrong. On a Multiple Induction 



cient quantity of pure carbon to give a dark slate colour to the com- 

 bined mass. The carbon was in the form of an impalpable powder, 

 and, I believe, consisted of purified lamp black. I tried every variety 

 of discharge I could think of sometimes weak and sometimes strong 

 sometimes from a single coil and sometimes from several in com- 

 bination. The most powerful sparks were obtained by using all the 

 six coils in parallel and discharging them simultaneously by the 

 action of a single alcoholic contact breaker common to the whole. I 

 also used a Ley den battery, consisting of four -J-gallon jars, which 

 1 joined up in various ways both in parallel and in series, and in 

 some cases introduced a wet string to soften the discharge. 



The circular lines which surround the main discharge, both hori- 

 zontally and transversely, first demand attention. They appear to 

 represent sections of concentric layers or shells, of which the spark is 

 the nucleus. Their great similitude to the lines of force represented 

 by iron filings under the influence of a magnet is suggestive of similar 

 causation, but I could not find any proof of their identity. The dis- 

 tance to which this circular action extended was far beyond the limit 

 exhibited by the photographs. By laying patches of extremely light 

 dust on paper at various outside distances, I distinctly traced these 

 lines at a distance of 18 inches from the centre of dispersion, and 

 there can be little question of the action existing in a lessened degree 

 at a still greater distance. It can hardly be doubted that the particles 

 of dust are linked together by polar attraction, and they probably 

 represent similar conditions of the air jerked out by the discharge, 

 but whether it be possible to regard them as indications of tracks of 

 diffused discharge of a different character from that of disruption is 

 more than I dare venture to say. I am satisfied that they are not 

 mere ripples resulting from pulsatory vibrations of air, for I find 

 that radial pulsations produced by mechanical means merely clear a 

 circle without the least tendency to form similar rings. I made the 

 experiment by stretching a membrane over a tin cylindrical box, 

 with a small air tube leading from the bottom to near the surface of 

 a dusted card. The membrane was then set in motion by a rapid 

 succession of light taps, which dispersed the dust in the same manner 

 the spark did, but not the least indication of circular lines could 

 seen, and yet air puffs play an important part in the process, 

 tor the effects are greatly modified by screens, as will presently 

 appear. J 



The figures also show in the most unmistakable manner that the 

 T' t S .r U I" the Spark ' 6XerCise a dis P ersi force. I am inclined 



. 



the dispersive action of the wire differs only in degree 

 om that of the spark. We see in the lines emanating from the 

 wires evidence of a molecular disturbance in the material which 

 >ots off the molecules of air in contact with the wire. When the 



