114 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 



motion of the submerged or partially submerged fog particles, to and from the 

 edge on one side. They approach and lea^'e in great armies, at first, gradually 

 dwindling in number as they pass out of the field of view, eventually to be lost 

 by evaporation. Sometimes both advancing and retreating fog particles are in 

 sharp focus at once. At other times the advancing set is obviously above or 

 below the other, to the extent of one or more tenths of a millimeter. Cases also 

 occur in which there are two edges or a sort of geographical strait in the field of 

 view, in which case particles are frequently seen moving from edge to edge 

 until they vanish in number, probably from evaporation. Vortical or involved 

 orbits of fog particles also occur. 



In explanation of these phenomena it is necessary to bear in mind that the 

 edge of retrogression mentioned is always nearly fixed, under the microscope. 

 Hence it is not probable that that motion of the mixed oil (varnish-turpentine) 

 dvie to concentration on evaporation can be the cause ; for in this case the liquid 

 would visibly gather itself up into a drop showing a shifting edge in the field. 

 The possible motion of a liquid on its surface skin or similar capillary phenom- 

 enon is equally hard to reconcile with the stationary edge. The most probable 

 explanation, it seems to me, is given by the annexed diagram, figure 9. If the 

 film on the plate of glass is microscopically uneven and slightly inclined, the 

 motion of a relatively liquid layer over a fixed layer may enclose a shallow 

 region of liquid, in which eddying is kept up, as shown in the diagram, remem- 

 bering, of course, that all motion is observ^ed under the microscope. The 

 particles indicate the motion of the skin circulation. 



14. Graded particles. — Suggestion may finally be made as to the cause of 

 the observed gradation of particles, w^here such gradation appears simultaneously 

 with clear-cut coronas. In case of the X-rays the coronas are vague and washed, 

 accompanied with copious rain. What is seen is a coarse red-rimmed fog. 

 Here gradation is obviously due to the corresponding gradation in the sizes of 

 the nuclei, as explained above. Something similar shows itself in the initial 

 phosphoiiis or sulphur fogs. 



The grading in question cannot be due to coalescence, not only because 

 such coalescence is but very rarely observ^ed, but because the volume increase 

 is as the third power of the diameter. Sizes as i to 3 being very common, this 

 would mean an equally frequent coalescence of 27 droplets, which could not 

 escape detection. Moreover, the coronas in air retain a nearly fixed diameter 

 tmtil they are lost by subsidence. 



Some difference of size must be due to a difference of surface adhesion; 

 but if the oil stratum is of the same character throughout, this is not liable to be 

 large without being detected in the picture. 



The final cause for gradation, apart from original differences of nuclei, is 

 evaporation. The occurrence of marked evaporation is at times beyond ques- 

 tion, and it is then impossible, or nearly so, to obtain a photograph. In general, 

 however, evaporation is obscure, and one may argue, as already suggested, 

 that in the time (about 60 seconds) needed for adjustment and photography. 



