conveying the Conduction Currents an Metals. 155 
By introducing the value of o from (3) we find 
20tV 
| eet Me al aca 8 a VK. 
As Drude’s measurements of v« refer to sodium light we 
put A=5°89 x 10-5, and substituting for V its value 3 x 10”, 
we ultimately obtain 
p?=45°6 A?Cre. 
The following Table gives the value of p calculated in this 
way for all metals having a conductivity greater than lead ; 
for mercury, bismuth, antimony, and lead the complete 
equation (11) was used. 
| 
| ieee C. 





VK. Dp. 
LS 6-7 8:03x10-5 | 5:94 ‘99 | 
| Mercury ::.........: 14°8 1-06 | 858 FOr” 
| Bismuth............ p22 ‘76 | 6:96 iit 14 
33 67 9-92 8-54 131 | 
eter. 10°3 66:5 1-66 146 | 
Copper ...........- 71 | 626 | 160 | 154 | 
== ee 10-2 48:6 2-06 155 | 
Magnesium ...... 413:3— 23:7 1:64 1567. oa 
Inu... ...~- oe 11:0 878 EOL = 
2 ee 18-2 51 7-00 234 | 
Cadmium ......... 13-0 14:6 «566 9-53 | 
Antimony ......... A hiO 2°8 15-02 2°58 | 
ce ee 16-2 TE ever! 265 | 
ee sgt | 95 17°8 | 11-62 | 292 | 
Aluminium ...... | 10-1 B43 | 793 | 3:47 

The second and third columns give the atomic volumes and 
electric conductivities. The third column is Drude’s value 
of v« for sodium light and the last column gives the factor p, 
indicating the ratio of the number of electrons to the total 
number of atoms. It should be noticed that although the 
extreme electric conductivities differ in the ratio of 63:1, and 
the values of v« in the ratio of 23:1, the extremes of p are 
only in the ratio of 5°5 to 1,so that the product A?Cve, 
though not absolutely constant, is more so than two of its 
factors, the variability of the third (A) being about the 
same us that of p. For the first three metals on the list 
the number of electrons is practically that of the number 
of atoms, and p for all metals is less than three except 
in the case of aluminium, which in comparison with other 
metals is very opaque compared with its conductivity. It 
is always to be remembered that we have neglected all 
opacity due to selective molecular absorption so that our 
values of p are upper limits. Nevertheless, the tact that the 
