306 MK. C. T. R. WILSON ON THE EFFICIENCY AS CONDENSATION 



unaltered by such an increase in size as would reduce the velocity to anything like 

 such a small value as 1/4000 centim. per second for a potential gradient of 1 volt 

 per centimetre. In fact, if we can trust to obtaining approximately correct results by 

 applying to drops as small as 2 X 10~ 6 centim., the formulae which hold for larger 

 drops, we can see* that these nuclei must amount to at least 2 X 10~ 6 centim. in 

 radius ; the supersaturation required to cause condensation to take place on drops of 

 this radius being that produced by an expansion, ty'v, = about I'Ol, which is very 

 far removed from that actually required, v?jv l = 1'25. 



The ionisation may be a result of the expansion. This is the view I am inclined to 

 take. 



It is easy to understand, according to this view, how the number of drops produced 

 is entirely uninfluenced by even a strong electric field, for the whole time for which the 

 ions would be free to move under the action of the field, before the formation of drops 

 upon them, would be exceedingly short. 



Uncharged nuclei are probably present before the expansion. Some change is 

 produced in the air by weak ultra-violet light independently of the expansion, for, as 

 we have seen, the fogs may be obtained even when the expansion is made some 

 seconds after the light has been cut off. The behaviour of moist air exposed to 

 stronger ultra-violet light, and especially the extreme case where we get visible 

 particles produced without expansion, almost compel us to conclude that even weak 

 ultra-violet light produces nuclei before the expansion is made. I have already 

 suggested! that these nuclei consist of what we may regard as minute water drops 

 containing hydrogen peroxide in solution. The difference between the effects of the 

 strongest and weakest ultra-violet light would consist in a difference in the size of 

 these drops due to the larger quantity of hydrogen peroxide produced in each by the 

 stronger radiation. We may suppose the very minute molecular aggregates due to 

 weak ultra-violet light to be of themselves too small to act directly as condensation 

 nuclei with the expansion I'-j/v, = 1'25; in other words, the growth which results 

 from the supersaturation corresponding to this expansion may be insufficient to bring 

 them up to the critical size beyond which the unstable condition is reached, where 

 increase in size is accompanied by a diminution of the equilibrium vapour pressure. 

 But if, as a consequence of the increase in size which results from the supersaturation, 

 the nucleus becomes charged, an ion carrying electricity of opposite sign to that left 

 on the original nucleus being thrown off, the result actually met with would be 

 explained. Judging from the behaviour of hydrogen peroxide solutions, with respect 

 to the electricity developed by splashing, we would expect the original nucleus to 

 become negatively charged ; for the splashing results in a negative electrification of 

 the drops \ indicating that the inner coat of the double layer originally covering the 



* Using the value found by Professor THOMSON (loc. cit.) for the charge on one ion, and applying the 

 formula for the steady motion of a sphere through a viscous fluid (LAMB, ' Hydrodynamics,' p. 532). 

 t 'Phil. Trans.,' loc. cit. 

 \ J. J. THOMSON, 'Phil. Mag.,' vol. 37, p. 341, 1894. 



