Sizes of Water Particles producing Cloudy Condensation. 25 



The precipitation for adiabatic expansion was computed 

 from rx = sft — CQlnQ, where is the absolute temperature, 

 r the latent heat of evaporation, C the specific heat of the 

 liquid, S the constant entropy, and x/(l—x) the ratio of 

 vapour and of liquid in the chilled mixture. In each 

 exhaustion in the globe ™=79/10 8 gram was precipitated 

 per cub. centim. of air, and for the drurii similarly m = 36/10 8 

 grains of water per cub. centim. of air. 



Finally the size of the water globules was found from 

 m — N7rt/ 3 /6 = N7rD 3 /6N / , where N is the absolute and N' the 

 relative number of particles per cubic centim. In this way 

 for coronas in the globe, 



N=8'3xl0 4 xN / .... (punk nuclei), 

 N = 3'6 x 10 4 X W . . . . (sulphur nuclei) ; 



and for axial colours in the drum and weaker nucleation, 



JSr=r5xl0 4 xN / .... (punk nuclei). 



If N' = N ' = 1, or initially (w = 0), N = 83000 and 36Q00, 

 for punk and for sulphur nuclei, respectively, found in the 

 globe. This nucleation was made as dense as I conveniently 

 could. It is interesting to note, therefore, that these data 

 lie between 10 4 and 10 5 , the limits found for phosphorus, 

 nuclei in my earlier papers under very different conditions * 

 The table has been constructed for punk (glowing charcoal) 

 nuclei only, as these seem to be more uniform and give rise 

 to a greater wealth of coronal display. 



General Remarks and Deductions. 



With regard to the two sets of colours, it appears that the 

 coronas must in large measure be diffraction phenomena ; 

 the axial colours (which are rigorously axial) cannot be so 

 explained, but seem to be evoked by some unknown kind of 

 absorption. The contrast is sharply brought out by the 

 experiments. Only after about 20 partial exhaustions (76-58 

 centim.) are the ordinary white-centred co?onas encountered. 

 After n exhaustions all the preceding coronas and axial colours 

 from ton are flashed through, and after hovering over the 

 limit for a brief time show a tendency to a retrograde movement 

 (n to 0), though this soon vanishes beyond recognition. The 

 coronal centre (uniformly coloured patch or disk, in which lies 

 the coloured point-source of light) surrounding the first axial 

 blues and greens is not determinable as to colour, in the 



* Phi!. Mag. [6] vol. iii. p. 91 (1902). 



