226 REPORT — 1889. 



the rod, as it were, by licking its surface, or, perhaps, as Van Rees ^ suggested, by 

 throwing ofi" under the influence of the charged rod, finite portions of itself charged 

 inductively with the opposite electriflcation, I made the experiment of surrounding 

 the rod by a strong blast of air from a foot blow-pipe, expecting thereby to prevent 

 the discharge by blowing away the conductive air before it reached the surface of 

 the rod ; but I found that the interposed blast of air, however violent, was quite 

 without influence on the discharge, which took place across it with perfect facility 

 when the flame was approached. It should be mentioned that the blast alone 

 without the flame was quite incompetent to discharge the rod. 



4. This observation makes it, I think, quite clear that the discharge is effected, 

 not by the flame or mass of air round it acting as a whole, but by a molecular 

 action, the velocity of propagation of which is very great compared with the 

 velocity of the air in the blast. 



5. All the phenomena above mentioned, and others to which I shall shortly 

 allude, appear to be explainable as follows, and, so far as I can see, in no other way. 



The flame contains a large number of dissociated atoms, each with +ve or — ve 

 atomic charges. A charged body, such as the rod, attracts towards itself those of 

 opposite sign and repels those of the same sign, and under this directive influence 

 an electrolytic action is propagated across the intervening air, the discharge being 

 effected by the liberation of free atoms on the surface of the body. When a glass 

 or ebonite plate is interposed this liberation takes place on the surface nearest the 

 flame, and the plate is thereby charged. 



The shortness of the mean free path of molecules, and presumably also of 

 atoms in the air at atmospheric pressure, and the incompetence, so far as I can 

 observe, of even the most violent interposed blast of air to aflect the discharge 

 seems to preclude the supposition that the free atoms which reach the rod are those 

 which were originally present in the flame. 



The flow iu the opposite direction of atoms bearing charges of the same sign 

 as that of the rod was put in evidence by the experiment of placing a glass or 

 ebonite plate near to an insulated flame, and then bringing up a charged rod to 

 the opposite side of the flame, when the plate was found to receive a charge of 

 the same sign as that of the rod. Care must be taken to withdraw the plate as 

 the rod is approached, otherwise it also will be discharged. A variation of this 

 experiment is to make the like charge flow on to and charge an electroscope by 

 means of an interposed insulated flame. For some experiments of this kind a 

 metallic conductor connected with an electrical machine by which the charge may 

 be maintained was found more convenient than the charged rod. 



6. When a charged insulated body is discharged by the proximity of an 

 insulated flame it is not at once obvious how the equal and opposite charge on the 

 walls of the room in which the experiment is made, has been neutralised. I 

 imagine that this must be efi'ected by an electrolytic discharge through the air 

 between the flame and the walls, originating in the excess of atoms charged 

 flimUarly to the body, which are left over when the charge on the body is 

 neutralised, and on which the lines of force proceeding inwards from the walls of 

 the room now terminate. Thus the discharge involves the shifting of the ends of 

 these lines of force from the original rigid body on to atoms which are free to 

 move under the directive stress. 



7. If this explanation of the discharge by a flame is correct, it should follow 

 that discharge should be facilitated whenever free atoms are present, and I found 

 on trial that when a rubbed ebonite rod was placed transversely in the path of the 

 spark discharge, between the terminals of a Wimshurst machine, it lost a large 

 amount of its charge at the passage of each spark, and was soon completely dis- 

 charged. 



This appears to be a special instance of the phenomenon brought into notice 

 by Dr. Schuster of a small electromotive force producing a current when the 

 electrolyte has been broken down by a greater electromotive force. 



8. For investigation of the temperature at which a piece of hot metal begins to 

 act like a flame in producir,g discharge, a storage cell was insulated and connected 



' Poffff. Ann., vol. Ixxiv., p. 379, 1849, 



