| 
Recombination of Ions in Gases. 209 
The agreement between the values of \ for any one pres- 
sure is very satisfactory. Butif a is independent of pressure, 
x should be constant. There is considerable deviation at the 
lower pressures. It cannot be held to be due to experimental 
error without detracting seriously from the value of the 
conclusion. It may be due to a difference in the sensibility 
of the electrometer ; and if so, Mr. McClung may be able to 
explain the deviation if he has a record of the sensibility. 
The next table is prepared from the same experimental 
results, and gives the values of a calculated by the three 
formule (4), (5), and (6). The values are expressed as 
multiples of e, the charge on an ion. 
Pe | JPA ay. [PMH od | RE | a. 
_ Atmospheres, | ; 1; Q My 
Panes | 250 236 1916e | 143 974¢ | 389 | 3940c 
OU ES aaah | 217 196 2060¢ 112 9372 | 343 | 4466¢ 
1 Rpts sa (Donel bs 181 | 1829e 51 1384e | 642 | 23876 
ae ae 147 MaeeigiGe' | 4t | 40708 | 578° \'| 26526 
5 eS | 94 | 136 | 1286¢ | 26 | are | TL | 21566 
| | 
The numbers are reduced by the data given on p. 300 of 
the same paper; and we may note that the values of N and 
Q given there give N?/Q=453. 
In the table the values of a for any given pressure are 
quite discordant, and the values of @ calculated by any one 
method cannot be said to be independent of pressure. 
It may be argued, since the values of N?/Q differ, that this 
is due to a change of sensibility of the electrometer If so, 
the values of N, Q, and a would have tc be changed in the 
same proportion for any one pressure. This, however, would 
not improve the relative values of ay and ag at that pressure. 
For example, at 3 atmospheres %y2 would have to be 
reduced while ay and ag would have to be increased in the 
same proportion, and thus @yjqg2 Would be brought nearer 
ay or aa, but ay and ag would not be brought nearer. 
Again, at *d atmos. ay, would have to be increased and 
ay and eg diminished in the same proportion. Thus I do not 
think that a change of sensibility would improve matters. 
Phil. Mag. 8. 6. Vol. 8. No. 44. Aug. 1904. ie 
