NUCLEI OF POSITIVELY AND NEGATIVELY CHARGED IONS. 805 



terminals, at a distance of 55 centime, from the quartz plate in one series of 

 experiments, at 180 centims. in another. In neither case was the radiation strong 

 enough to cause drops to be produced with expansions appreciably below v 2 /v ) = 1'25. 

 The spark -terminals were placed in the plane of the central plate, so that the air in 

 lx>th halves was equally exposed to the rays. The wet filter paper which covered 

 the brass plates prevented any surface effect from the light which might reach the 

 plates. As before, the air on one side was between plates at the same potential, that 

 on the other side in a strong field. The central plate was earthed and one of the 

 side plates kept at a positive potential of 320 volts, the other being earthed ; the 

 connections were interchanged after each observation, so that the air in each half 

 alternately was subjected to the action of the field. 



When the expansions were made while the air was exposed to the rays, no 

 difference could be detected between the fogs in the two halves of the apparatus. 

 Moreover, even when the rays were turned off 10 seconds before the expansion was 

 made, a slight fog was obtained, equally dense on both sides of the central plate. 

 Thus some of the nuclei appear to persist for 10 seconds, and even in that time the 

 field has had no sensible effect in reducing the number of the nuclei. 



A field, therefore, of 400 volts per centimetre causes the nuclei to move in 10 seconds, 

 a distance small compared with 8 millims., the distance between the plates. This 

 gives, for the velocity under a potential gradient of 1 volt per centimetre, less than 

 1/4000 centim. per second, whereas the ions produced by Rontgen rays travel under 

 these conditions between 1 and 2 centims. per second. 



The slight rain-like condensation which takes place, when v.,jv l lies between T25 

 and 1'38, in the absence of all radiation, as well as the much denser condensation 

 produced by the same expansions when the air is exposed to weak ultra-violet light, 

 are thus essentially different phenomena from the apparently similar condensation 

 produced in air ionised by Rontgen rays. 



We might perhaps most naturally conclude that we are in these cases not 

 concerned with ions at all. There is, however, the difficulty of the unlikelihood of 

 two entirely different classes of nuclei being so exactly identical in the degree of 

 supersaturation necessary to cause water to condense on them. The apparent 

 existence of a second coincidence (an increase of the number of drops when v 2 /r, 

 exceeds l - 31) is still harder to explain on this view. 



It is possible that condensation in these cases really does take place on ions 

 carrying the same charge as those produced by Rontgen rays. There are, in fact, 

 several ways in which we may account for the fact that an electric field does not 

 remove them. 



We might suppose that the nuclei differ from those produced by Rontgen rays 



merely in being so much larger, that their velocity in a given field is diminished 



enormously, the charge in each nucleus remaining the same. It is difficult, however, 



to believe that the efficiency of the nucleus in helping condensation would remain 



VOL. cxcin. A. 2 R 



