Conducting Gases on the Direction of the Electric Field. 221 



to the fact that a small charged body moving through a gas 

 would experience a greater resistance than an uncharged body 

 of the same size. It does not seem, however, that this expla- 

 nation will suffice. 



On introducing water- or alcohol-vapour into the dry ga«, 

 the velocity of the negative ion decreases by gradual and not 

 sudden stages. If the ion were molecular in size, the addition 

 of a molecule of water-vapour would greatly alter its velocity. 

 The slow and gradual decrease of the velocity of the negative 

 ion with the addition of water- or alcohol-vapour, seems to 

 point to the conclusion that the ion on which it is con- 

 densing is already large, unless it is assumed that only a 

 fraction of the number of ions is affected by the presence of 

 the vapour. The addition of a molecule of water-vapour to 

 an ion of the same dimensions as a molecule would greatly 

 lower its velocity, while it would make only a slight difference 

 to an ion already much larger in comparison. 



There seems to be considerable evidence that the ion, some- 

 where between the pressure of an atmosphere and about 

 1 millim. of mercury, alters greatly in size. I have shown 

 that the velocity of the negative carrier produced by ultra- 

 violet light is about the same as that of an ion produced by 

 oc-rays; but the results of J. J. Thomson, Lenard, and others 

 show that it behaves at low pressures as if it were identical 

 with the cathode-ray carrier, which is apparently much 

 smaller than an atom. 



Townsend* has also shown that the negative ion produced 

 by Rontgen-rays at a pressure of about 1 millim. of mercury 

 is very small compared with a molecule. 



Experiments with Rontgen-Rays. 



Fig. 3 (p. 222) shows the general arrangement. A thin stratum 

 of Rontgen-rays was passed near the surface of the plate A but 

 not touching it. The current between the vertical plates A 

 and B was measured in the usual manner by an electrometer 

 with suitable capacity in parallel with it. The plates C and 

 D served as guard-plates for the plate B in order to produce 

 a uniform electric field between A and B. In order to keep 

 a check on the constancy of the rays, the current between two 

 standard plates, between which the rays passed, was observed 

 at the same time. 



The following results were obtained when a strong stratum 

 of rays passed through the fairly dry air of a Canadian 

 winter. 



* Phil. Mag. Feb. 1901. 



