no A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 



6. Diameters of jog particles. — If the diameters measured are plotted in 

 a chart together with the resvilts computed from successive exhaustions in 

 the older and in the more recent experiments, the present vaktes again lie be- 

 tween the two curves, but now much nearer the lower (recent) curve than before. 

 I shall not pause to interiDret the differences which remain, but only to remark 

 that the capillary forces at the area of contact of the droplet even with the liquid 

 oil film may transform it to an oblate spheroid, and that diffraction at the 

 circular edges of the drops is not excluded. If the nucleation, m„, obtained 

 from successive isothermal exhaustions and subsidence measurement, be ac- 

 cepted as correct (lower curve), the ratios of the nucleation found from the 

 different methods tested will then be 



From subsidence, a = .0029; d/d,, 



From lycopodium (c/=.oo3 cm.), a = .0034; " 

 From diffraction (blue), a = .0034; " 



From micrometer measurement, a = .0037 ; 

 Old results (adiabatic conditions assumed), " 



Since w is obtained from the cube of d, large differences of this kind are as yet 

 inevitable, particularly as the particles measured in these different cases are not 

 the same. 



7. Sizes of particles graded. — ^The point of particvilar interest which comes 

 ovtt on using the eccentric plate to catch the subsidence during 15 or 30 seconds, 

 and at once examining the deposit, is the result that particles of all sizes are 

 present. By far the greater number, however, have the maxinuim diameter. 

 These particles are caught from the fog without interference, and it is not prob- 

 able that coalescence or evaporation have been appreciably operative, so long 

 as the corona remains the same throughout the micrometer measurement. The 

 probable explanation is this : while the pressure decrement is growing from zero 

 to the maximum 6p, condensation is taking place on the greater numl)er of 

 particles throughout the whole of this interval. In other words, although the 

 nuclei are graded in size, the greater num.ber exceed a certain dimension and 

 rcquii-e almost no pressure decrement to induce condensation. These are the 

 particles (diameter exceeding a certain inferior linit) which give character to 

 the persistent corona. A minority of the graded particles are below the di- 

 mension in question, and upon these condensation does not take place until the 

 higher values of the pressure difference are reached; some may even require 

 the full decrement, 6p. Thus it is that in the deposit of fog particles, one finds 

 those of diameter .001 cu. intermixed with others of smaller diameter, even as 

 far as .0002 cm. or less, all shining like beads. When fi-esh phosphorus nuclei 

 are first introduced into the condensation chamber the result is a gray fog, but 

 a relatively sTall white reddish corona is nevertheless disceniiVjle. Accordingly, 

 the crop of droplets seen under the microscope contains not only surprisingly 

 small but also relatively large droplets, with all interm.ediate diameters. Hence 

 the indefinite fog and the small corona. The large olive (g b p) corona and 



