34 



ITTE STRUCTURK OF THE NUCLEUS. 



to ailruit of both losses of similar geometrical character. Finally the initial nucle- 

 atiou is not identical iu the different ex[)eriments, whence if JS'o = y"', 



N = \0 



(z^+z (i+^/))log7 



is the final form of the eqviation. 



If in two experiments beginning with two different nncleations, N^ and N'g, 

 tlie same corona or iVis reached for different times, t and /', and different numbers 

 of exhaustions, z and z\ of the same ratio, //, then N = iV', and therefore 



^^ a{,+it) log J ^ j()(2'o+2' U+6n ) log J' 



where for brevity ^^o = and A"y= 1. Thus 



i'-zt 



If two identical coronas are reached, the e'„ may be eliminated or 



•(/'-2/),-(/'-z/)e' 

 where the first identical coronas are seen for iYj, Zj, ?j; y\, 2\, t\ ; 



and the next for ^o^ ~2- ^2' ^^'2' ^'2' ^'2- 



The plan is, therefore, to obtain an even number of identical coronas in case of 

 tables 4. and 5, and then to comi)Ute I' from equidistant groups of two iu the way 



suggested. 



TABLE 6.— TIME RATES, i = jy,. 



27 

 61 



94 

 124 

 160 

 190 

 292 



50 



80 



1 12 



140 



o"' 



95 

 '55 

 '85 



*Xio» 



4 



8 



10 



14 



17 

 20 



86 



7> 

 87 



an 6 



.081 



8 



"•5 

 15 

 19 



140 

 140 



96 

 119 



.140 



. 107 



It. Applimtion. — The results of tables 4 and 5, if treated in this way, give 

 rise to tal)le (5. The time interval between the successive coronas in table 4 was 

 about i = 1.0 min. The other (piantities are fully given in the table. It will be 

 seen that for each of the series (1), (2), (3), the value of b is as nearly constant as is 

 possible under the method of observation. The lines of figure 5 are constructed to 



