M. W. Holtz on a new Electrical Machine. 427 



the expense of their number, either the striking- distance, the 

 electrodes, or the surface of the conductors must be increased : 

 the latter is best effected by a Ley den jar, the coatings of which 

 are connected with a and c. 



By increasing the striking- distance, the efficiency of the ma- 

 chine is not in general diminished. This is best seen in charging 

 a jar, for the time then varies in the same ratio as the intensity 

 of the charge. Beyond a certain point, the striking-distance can 

 be no more increased. If it is, the current of sparks begins gra- 

 dually to disappear, and, indeed, most rapidly if both conductors 

 are insulated. A complete connexion must then be quickly re- 

 established for a short time, if the electricity is not to be quite 

 lost from the coatings. If it is increased, when a jar is in the 

 circuit, the direction of the current changes when a certain 

 intensity is exceeded : the jar is discharged by the apparatus itself, 

 to be recharged in the opposite direction, and this is continued 

 until the striking- distance is lessened. But the striking- distance 

 which can be attained is in each case somewhat greater if both 

 conductors are insulated. 



The resistance of the body interposed between b and c must 

 not exceed a certain limit -, for instance, it must not be a semi- 

 conductor, if the machine is to work. On the other hand, the 

 conductors, if they are to fulfil their object, must not be con- 

 nected with semiconductors alone, as these act, as regards the 

 intensity of the current, more or less like insulators. 



If the machine is to cease working for a short time, the cur- 

 rent is closed and the disk then allowed to come to rest. The 

 glass surfaces remain then (usually for half an hour) sufficiently 

 electrical for its working to be reestablished by rotation. In dry 

 air the apparatus can be left for hours without the electricity 

 being quite removed from the glass. 



The connexion of these phenomena may be seen from the fol- 

 lowing considerations. 



Assuming that the coating denoted + is charged feebly with 

 positive electricity, it acts inductively on the rotating insulator. 

 The repelled -fE passes into the conductor e, while — E simul- 

 taneously collects on the glass disk. The latter becomes partially 

 free as soon as it passes beyond the limits of the coating ; but it 

 still remains partially bound, as it acts inductively on the outside 

 of the fixed disk, causing here a gradual accumulation of + E. 

 The binding influence of the latter is interrupted by the notch 

 in the glass. The — E still on the insulator, and which has 

 now become free, then enters at the points of the second coat- 

 ing, and the part which is not taken up by it disappears by the 

 conductor /. In the degree in which the second coating is 

 charged, it exerts an inducing action; negative electricity is 



2F2 



