DowLiNG& Haughey — Electrification, Phosphorus SmoJce Nuclei. 53 



steps {i.e., the voltage increments) on the curves are equal — a result which we 

 might perhaps have expected; while, of course, the "rises" [i.e. the charge 

 increments) are also equal. ^ 



VI. — As we have said, the voltage " step," and also the " rise " in Ijh 

 at each completed step, are nearly constant in any one curve. On division 

 of the total voltage at the commencement of any step by the (average) 

 value of the " step " of voltage, we find that the quotient is nearly a whole 

 number. These numbers are quite small, rarely exceeding 20, and increase 

 by unity at each step. This is just what we would expect from our theory, 

 and, in fact, the numbers are to be identified with the numbers of 

 electron charges at each stage. Furthermore, the average increment in l/7i 

 (the " rise ") when divided into the corresponding value of the ordinate (l/A) 

 yields the same number. We feel very confident, therefore, that the 

 explanation put forward is correct, but it is desirable to subject the matter 

 to experimental test in another and more fundamental way, viz., to 

 investigate the actual variation of the charges by direct measurement. 



VII. — To achieve this we directed our attention to devising methods 

 in which the periodic alteration in the individual charges showed itself in 

 corresponding variations of the electric " transport " current, carried by the 

 smoke stream. It was found, rather unexpectedly, that the effect was 

 noticeable even with the very simple and direct method shown in fig. 3. 

 This arrangement, however, only enabled us to demonstrate the periodic 

 character of the current over about the same range as in the " visual " 

 method just described.' 



VIII. — In the apparatus shown in fig. 3 there is, of course, a large 

 admixture of ordinary ions. To avoid this, the apparatus (fig. 1) was 

 modified as follows : — The central partition was removed, and the plate C 

 divided into two parts. The upper part, Ci (fig. 4), was carefully insulated 

 and connected to a sensitive electroscope (10 divisions per volt) similar to 

 that shown in fig. 3. The lower part, Cj, was earthed. The plate A, to 

 which the phosphorus was attached, was connected to a suitable potential 

 dividing arrangement, so as to enable the field between A and C to be varied 

 as desired. Some of the observations thus obtained are given in fig. 5. 



' It was found necessary to introduce a correction for a gradual variation in the upward 

 air -current. This was done by repeating a measurement of h at intervals for a chosen 

 value of Vi. 



^ The phosphorus P is here placed on a small table connected to the fibre of the 

 "Bumstead" electroscope, and is exposed to a field of any desired intensity by means 

 of the upper plate. 



