M. W. Holtz on a new Electrical Machine. 431 



air this loss, after one or two minutes, is so great that the action 

 cannot be reproduced by rotation alone. 



To pass now to the action of the machine. I give, in the 

 first place, a few determinations of the maximum striking- 

 distance, and its quantitative effect. 



Between circular electrodes of J- inch diameter, the striking- 

 distance which could not be exceeded was with two elements 1 

 inch, with four elements in cross contact § inch; with two 

 elements and separated currents a striking-distance of 2 inches, 

 and with a larger conductor one of 3 to 4 inches could be ob- 

 tained. When a Leyden jar was used, in the first and second 

 case the maximum striking-distance was ^ inch less. 



Between the same electrodes, and with twelve turns in a second, 

 a jar with a square foot of coating, and ygth of an inch thick, 

 could be charged to a striking-distance of f inch by two ele- 

 ments in two seconds, and by four elements in one second. 



By diminishing the electrodes, the striking-distance could be 

 increased to a certain extent. With four elements, knobs half 

 an inch in diameter could be removed to |ths of an inch dis- 

 tance, and the jar discharged itself, regularly at this distance in 

 one second. Points, on the contrary, on account of the brush 

 formed, could only be brought to a distance of -Jths of an 

 inch, at which about three or four discharges took place in a 

 second. 



The expenditure of force which corresponded to this effect 

 was, as compared with other electrical machines, very small ; for 

 if the disk was suddenly left to itself, it generally continued to 

 rotate eight seconds with four elements, and fourteen with two 

 elements. 



The current of sparks took very different shapes, according 

 to the form and distance of the electrodes. Between points 

 it formed a small bundle, wider towards the middle, and consist- 

 ing of innumerable little sparks (fig. 4). Between knobs, at a 

 greater distance, it formed usually a thicker sinuous thread, of 

 small brightness and roseate colour. At a smaller distance it 

 divides, without losing its colour, into several separate threads, 

 which, for the most part parallel and lying in one plane, diverge 

 in continually greater arcs from the straight middle line. By 

 connecting a conductor with the ground, or by enlarging the 

 surface, the dark colour can be easily changed into a light one. 

 A very intensely luminous current of sparks, which is sur- 

 rounded by a bluish sheath, is obtained, between points, with a 

 jar of 3 to 4 square inches of coating. 



In order to vary most conveniently the luminous phenomena 

 in a Geissler's tube, it is interposed in the circuit instead of the 

 wire /. As long as m and n are in complete contact, it shines 



