Nolan and Eneight — Experiments on Large Ions i7i Air. Ill 



determined, and, therefore, the mean charge per ion. The following table 



gives the results of some observations on the ordinary flame-gas : — 



Total nuclei UncLarged nuclei Ions of one sign Electronic Electronic 



per c.c. perc.c. perc.c. chnrges charges 



per c.c. per ion 



38-6 X 10' 22-6 x 10' 8 x 10' 25-4 x 10' 3-2 



70-5 34-5 18-0 53'5 3-0 



840 41-0 21-5 61-7 2-9 



325 192 66-5 171 2-6 



230 105 62-5 214 3-4 



225 133 46 ■ 116 2-5 



391 213 89 289 3'2 



500 229 135 344 2-6 



567 317 125 360 2-9 



It will be seen that usually about half of the nuclei are charged, and that 

 the average charge per ion is three electrons. That this condition is quickly 

 arrived at is seen from the following observations, which refer to air com- 

 pletely freed from ions and exposed to the X-rays : — 



n 



"1 



n — ni 

 2 



P 



X - 



240 



117 



62-5 



182 



2-9 



256 



144 



56-0 



185 



3-3 



532 



245 



143-5 



381 



2-7 



322 



190 



66-0 



174 



2-6 



370 



290 



40-0 



137 



3-4 



In the last two observations the ions were removed by applying a field 

 close up to the flame. 



We tested the effect of weakening the ionisation by interposing sheets of 

 tinfoil in the path of the X-rays. The following observations correspond to 

 successive diminutions in the intensity of the ionisation : — 



V — ill 



11 rii ■ V X 



190 X 10' 112 X 10' 39 X 10' 81 x 10' 2-1 



130 60 35 61 1-7 



130 58 36 39 11 



132 110 11-0 very small — 



132 120 6-0 „ „ — 



At first the effect is to lower the charge per ion. The half-and-half 

 distribution of the nuclei between charged and uncharged persists up to the 



