A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION. 75 



is the ratio of the diameter of particles to the distance between particles, con- 

 stant throughout. The distance between centers is thus about 48 times the 

 diameter of particles for the temperature and pressure conditions prevailing 

 during the exhaustions. 



One may note that the diameters found are independent of in; for in the 

 above notation let z and D refer to the normal corona virtually used for stand- 

 ardization. Then, as the series stands, 



4 

 or after reduction, since the same equation also holds for Z, 



i/d'=(sJa)'io^'-^^'°^'''nli-S/s') 



z- 



where 5^ is the aperture of the normal corona numbered Z. Thvis d depends on 

 a, y, and S, and does not therefore differ much from my earlier values except in 

 so far as a and y were differently determined and S not observed. 



Finally, since «(i^ = 6m/ 7r = const., the relation of n and d are reciprocal, 

 and maxima in n thus correspond to minima in d. If d is determined too large, 

 n will be too small. The curves bear this out. The periods indicated by the 

 cusps in the d curves are more appropriately referred to below. They may be 

 placed as follows: in the first curve they lie at (i = . 00069, -oooSS- -ooosg; in 

 the second at about (i = . 00079, -oooSS- -00040 cm. In conformity with the 

 work below their mean position may be rated at d = . 00072, .00054, .00036, or 

 in the ratio of 4, 3, and 2. In other words, they are, roughly, multiples of the 

 cycle datum .00018 cm., and throughout large as compared with wave length. 



MONO-CHROMATIC LIGHT. 



18. Tables. — The coronas are too faint for effective observation with 

 mono-chromatic light obtained from siniple sources like the salt flame. I 

 therefore used the electric arc as a point source and obtained sufficient limita- 

 tion for the present with a double thickness of mby glass. This arrangement 

 has an iilterior advantage as it is thus possible to observe the colors of the 

 annuli as well as their red diameter on interposing the colored glass. This 

 greatly facihtates the reductions. The observations made with the large 

 cubical condensation chamber are given in the following table arranged on the 

 same plan as the preceding. The constant o" is as before definitely determined 

 from subsidence, while a' refers to preliminary standardization with lycopodium. 

 The time loss P is neglected. The subsidence loss has been separately 

 computed and differs slightly from the above, showing more expeditious 

 work. 



