CONDENSATION NUCLEI." 



liy C. T. U. Wii,s()\'. F. 11. S., 

 Fellow of Sidiici/ Siissc.r CoUv<i<\ Ciniihridf/c, IJiu/land. 



If we tako the ordinary air of a room and inclose it in a glass vessel 

 containin<>- some water and provided with some means of increasing 

 or diminishing the volnme at Avill, we are able to observe the follow- 

 ing phenomena : If the air has been alloAved to stand sufficiently long 

 to become saturated with water vapor, any increase of volume, even 

 if very slight, causes the formation of a fog throughout the volume 

 of the moist air. This is easily made visible by concentrating a 

 powerful beam of light on the contents of the vessel; or, by placing 

 a small source of light behind the vessel, brilliant-colored rings or 

 coronas may be seen surrounding the source. If the air be made to 

 contract again to its original volume, a second expansion like the first 

 will again give a similar fog, but when this process has been several 

 times repeated the fogs become thinner, the drops being fewer and 

 larger; we get at length a fine rain on expansion rather than a fog, 

 the drops falling to the bottom of the vessel within a fev/ seconds 

 instead of remaining in suspension for many minutes like the first- 

 formed fog particles. When this stage has been reached, the next 

 and all succeeding expansions produce no drops at all, no condensation 

 resulting elsewhere than on the w'alls of the vessel. If ordinary air 

 be now admitted into the vessel, drops will again l)e seen on expan- 

 sion, uidess the air introduced has entered through a tightly pressed 

 plug of cotton wool, or has been otherwise filtered, in which case no 

 drops are seen. 



The phenomena are readily explained if we suppose that water can 

 not under ordinary circumstances condense in the form of drops 

 unless suitable nuclei are present to serve as starting points for the 

 drops. These nticlei are present in very varying numbers in ordinary 

 atmospheric air, from which they may be removed by filtering, or 



9 A paper presented at the International Electrical Congress of St. Louis, 1904. 

 Reprinted from author's revised copj'. 



195 



