674 Scientific Proceedings, lloyal Dublin Society. 



The present apparatus was designed to overcome this defect. It is clear 

 that, by moving the condenser plates apart normally, as we decrease the 

 capacity of the system, and thus increase the potential difference, the spark- 

 gap between the plates is increased at the same rate, or indeed somewhat 

 more rapidly. Thus, if the plates are not within sparking distance when 

 close together, sparking cannot occur between them in any other position. 



The accompanying figure represents the machine as constructed. The 

 earth-plate B is mounted on the end of a steel rod A, sliding freely in two- 

 uprights E, F. A is given a reciprocating motion by an excentric I, working 

 between two guides G, H attached to the rod A. The inductor-plate C is 

 supported behind by an ebonite block, 1 attached to the upright D by four 

 adjustment screws, 1 whereby C can be rendered exactly parallel to B. The 

 opposing surfaces of B and are, of course, as nearly plane as possible. 



The contact-making device is attached to the excentric shaft, which 

 ensures that the contacts are made instantaneously at the correct moments, 

 viz., when the moving plate is momentarily at rest at each end of its 

 oscillation. The ebonite arm J carries a contact piece which passes 

 between pairs of contact springs Li, L 2 , M,, M 2 , mounted on ebonite pillars. 

 Contact is made between L^L™ when the plates are closest together, and 

 between M,M 2 when the plates are fully separated. LiMj are permanently 

 joined to C, while B is permanently earthed. One pole of a battery is 

 attached to L 2 , and the other pole is earthed. The high voltage is, of course,. 

 obtained at the terminal M 2 . The plates B, C are about 20 cms. in diameter, 

 and B has a reciprocating motion over a range of about 5 cms. By the 

 adjustment screws the minimum distance of C from B can be regulated, and 

 thus the multiplying factor of the machine varied. The extent of this varia- 

 tion is subject to two considerations. In the first place, the minimum 

 separation must exceed the sparking distance for the battery voltage. This 

 consideration is of minor importance in the case of a machine of the dimen- 

 sions given, because to obtain a voltage of 4000 or 5000 volts, using a battery 

 of 800 volts, the plates need not be brought closer than a couple of millimetres. 

 The other consideration affects the steadiness of working. This would be 

 impaired were the back-lash of the excentric motion comparable with the 

 distance of minimum separation of the plates. Here again a couple of 

 millimetres suffices to eliminate trouble. 



Finally, the speed of working which I find most suitable is abouL five 

 oscillations per second. At this rate the equivalent current, when multiply- 

 ing 800 to 4000 volts, is of the order of half a microampere, which, of 



1 These are not shown in the drawing, beina behind C. 



