1G<» THE STRUCTURE OF TIIE Nl'CLKUS. 



Tlie (juestiou as to the form in which the eiuanation enters the nucleus in the 

 rehition of solute ami solvent is beyond the scope of the present inquiry ; l)ut in 

 proportion as the emanation consists of fewer molecules, the nucleus will be 

 smaller. A fresh phosphorus nucleation conveyed by dry air may theiefoi'e be 

 very different as regai'ds fineness, from a stale nucleation conve3ed by dry air. 



lu the case of different vapors much additional information may be 

 gained from Kelvin's vapor pressure equation which in the above notation is 

 Sp X /■ = 'IT a/P^^ F, from whicli it appears that the surface tension (7") actu- 

 ally enters as a single factor among many, to be counterbalanced by the differences 

 of vapor (ff) and li([uid {p) density. The quantity 7'^ roughly computed for the 

 solvents of this (•ha[)ter shows the following succession of values: 



Water, F = 1440 



Hence the factor F, for water is only fifth in the order of size, while the extreme 

 ratios of these factors is smallei' than 1/3. 



Different solvents therefore introduce different relations into the pi'oduct F, 

 (/. e., the curve be of the figure in questi(m), as well as new^ chemical conditions, 

 while the occurrence of the factor T in F, renders the presence of traces of water 

 in evaporating nuclei like methyl alcohol, for instance, a serious consitleration. 

 The whole subject is therefore, very complicated, and it seems wise to refrain from 

 further inferences until completed series of the experimental constants needed have 

 been investigated for certain typical solvents. It is in pursuance of the work of 

 Chapter V that the nearest approach to definite conditions is to be anticipated. 



A final word shoidd be added in relation to ijravitation; clearlv the rate of 

 diffusion of the nucleus upward will be smaller, the diffusion of the nucleus down- 

 ward larger, than the diffusion in a horizontal direction. The value found for « is 

 thus too small. 



21. Ionization and nucleation in relation to doudy condensation. — Hitherto 

 tlie only irrefragaljle evidence showing that condensation is promoted by ionization, 

 or in other words that negative ions are somewhat more active as condensation 

 nuclei than are [lositive ions, is the brilliant experiment devised byC T. 11. Wilson.' 

 Nuclei are here produced by the X-rays in communicating condensation chambers, 

 on the two sides of a vertical earthed metal plate, which receives electrical current 

 normally on one side, through the ionized air, saturated with watei' vapor, and 

 transmits the current in the same way and through the same medium on the other 

 side. Necessarily there was an excess of negative ions on one side of the plate and 

 an excess of positive ions on the othei' side. It was found on pr^xlucing condensa- 

 tion by exhaustion simultaneously on both sides under like conditions, that the 



' /"////. Tnins., London, vol. cxciii, pp. 289-308, 1899. 



