196 CONDENSATION NUCLEI. 



by iv])eateilly fonniiig a cloud by exp;insioii, and allowiiii;; the drops 

 to fall to the bottom of the vessel. Both the facts and the ex})lana- 

 tion have been long known. The particles which serve as the nuclei 

 of the drops formed, when ordinary atmospheric air is allowed to 

 expand slii2:htly, are conveniently called "dust" particles; they are 

 generally too small to l)e themselves visible, and it would be difficult 

 to find a means of determining whether they consist of solid particles 

 or of minute drops of liquid. The number of these dust particles 

 per cubic centimeter of air in different localities and under different 

 weather conditions has been investigated by Aitken and by others 

 Avith the aid of the ingenious dust-counting ap})aratus invented by 

 him. 



It is not difficult to understand wdiy nuclei shoidd be necessary for 

 the condensation of water in the form of drops. Lord Kelvin proved 

 that the pressure of aqueous vapor necessary for equilibrium over a 

 convex or concave surface of water differed from that over a flat 

 surface, being less over a concave and greater over a convex surface. 

 He shows how we may calculate the difference. A very small drop of 

 l^ure water will, if we assume the surface tension to remain the same 

 for very small drops as for large ones, ev^aporate even when siu-- 

 rounded by vapor many times more dense than that in equilibrium 

 at the same temperature over a flat surface. Thus unless the initial 

 stages of the growth of the drops can be, as it were, omitted, owdng 

 to the })resence of not too minute nuclei, a high degree of supersatu- 

 ration may exis't without any condensation in the form of rain or 

 cloud resulting. Lord Kelvin shoAved that to alter the equilibrium 

 vapor pressure by one part in a thousand the radius of curvature of a 

 sj)herical dro}) must amount to about 10 — * cm. Thus very minute 

 nuclei will enable a cloud to be formed with a very slight degree of 

 snq)ersaturation ; in other words, as a result of a very slight expansion 

 of the air if this has \)een initially saturated with water vapor. 



Tjoi'd Kelvin refrained from extending his calculations to curva- 

 tures of greater amount, as the surface tension can not remain inde- 

 pendent of the radius much beyond that limit. It is convenient, 

 however, to extend the calculations to greater curvatures; for 

 although the results ol)tained can not be considered as quantitatively 

 correct, they enable us to form a picture of the mode of action of 

 nuclei of diH'ereiit kinds. Let us imagine* an ai'i'angement equivalent 

 to that considered by Lord Kelvin; but since we are oidy here con- 

 cerned with con\(>x surfaces, let the capillary tube be joined as a side 

 tube to the lower part of a tall vessel of water. The capillary must 

 be su])i)ose(I to have walls of such a nature as not to be wetted by 

 water, and let us suppose (he open end of it to be bent round, so that 

 it i)oints vertically upwaid. and that (he height of the vertical })or- 

 tion can be adjusted to bring the meniscus to the open end of the tube. 



