EXPERIMENTS WITH IONIZED AIH. 55 



such prisms may be imagiued travelling iu opposite directions. Tliey ti'avel to .•ind 

 from a bouudary. The motion of the prism I is an abstraction, but "if I accentuate 

 it hei'e, I do so because in the present investigation with phosphoi'us it may run 

 closely imi'allel to the actual state of things. When the phosphorus grid is placed 

 on a smooth, clean surface, the i:>osition of the discs is soon marked by apparent 

 gi'ease spots due to deliquescing pho8[)horic oxide which has diffused aci'oss. Vir- 

 tually therefore an outgoing current originating in the phosphorus is continually 

 kept up, whether electricity is demousti'ably conveyed or not. 



To summai'ize : Instead of operating with the velocity U{ V/x), I have used 

 the constant absorption velocity, I; found in the absence of an electric field.' There- 

 aftei' I have endeavored to account for the inevitable factor V/x, or d V/dx, by- 

 associating it with the conditions of discharge of the plate, A. 



The only other explanation which I can suggest would be an hypothesis 

 whereby the ions in cei-tain cases like the present are released out of nuclei by the 

 pi'esence of a field. In fact, this is, bluntly, about the drift of the computations in 

 the above pai-agraphs, 8 and 9. 



TRANSAtlSSTON TJIROUGII SOLID MEDIA. 



15. Method. — The remainder of the chapter will be devoted to a number of 

 promiscuous experiments. I shall endeavor to ascertain whether the ionized parti- 

 cles exhaled l)y phosphoi-us are accompanied by some form of obscure I'adiatiou, or 

 whethei' the reaction is restricted to an oxidation the pi'oducts of which at first es- 

 cape in gaseous form. This is best done by placing barriers of thin material be- 

 tween the plates of the condenser, care being taken to 2)revent the nuclei from 

 passing around the bari-ier, yet allowing sufficient space for access of air. It was my 

 plan to fold sheets box-like around the phosphorus grid, kept at potential zero, so 

 that escape around the edge would I'ecpiire a passage of 8-10 centims. to reach the 

 other plate. This suffices, and the plates may then be placed quite closely together 

 on opposite sides of the barrier. Energetic action is secui'ed at a distance apart of 

 about 1 centim. 



16. Hesults. — The results are given in the following tables, which will be in- 

 telligible at once, remembering that V= FolO"'^'', so thatc' varies as —{dV/dt)/ V. 

 The insulation of the condenser itself (absence of leakage when the phosphorus 

 grid is replaced by an ordinary plate) was frequently tested, often with the barriers 

 in place. 



' Remembering that k is to be replaced by 3 k, if but J of all the ions travel in a given cardinal 

 direction, roughly speaking. 



