
between Optical and Electrical Qualities of Metals. 173 
mercury with a very thin sheet of the same glass as had 
always served in our previous comparisons. 
No diffuse radiation of small wave-lengths could. possibly 
have influenced our results. This was proved by the insertion 
of a plate of fluorite or rock-salt, 1 cm. thick, that perfectly 
absorbed the whole radiation *. 
Columv 6 of Table LV. contains the values of (L00—R) 
computed by means of formula3, for %=25°5u, whereas 
column 7 shows the values of (100—R) obtained by ob- 
servation. The agreement between the two columns is so 
good as to forma sufficient proof for the correctness of the law 
(100 —R)W/«=const., 
found by our previous experiments on shorter waves. The same 
is shown by column 8, containing the product (100—R)\/«; 
taken from our observationst. The average value of this 
constant amounts to 7°33 for the pure metals, to 7°41 for 
the alloys; the theoretical value, calculated by means of 
formula 4, is 36°5/\/25°5=7:23. Aluminium alone gives a 
rather considerable deviation, and bismuth furnishes a com- 
plete exception to our law t. But we can scarcely wonder at 
this, since bismuth forms an exception in various other ways §. 
Dependence of the Emission-Power of Metals on 
Tenvperature. 
The good agreement of the emission-power (100—R) 
obtained from our experiments with that computed from 
Maxwell’s theory, justifies the conclusion that the variation 
of the conductibility of metals with temperature requires a 
* Cf. H. Rubens and A. Trowbridge, Wied. Ann. lx. p. 724 (1897). 
+ Ofthe numbers in column 8 those for Zn, Cd, Ni,Sn, Hg, manganin, and 
constantan agree particularly well with the theoretical value 7:23. This 
is most probably owing to the fact that for these metals the conductivity 
was very accurately known. It was certainly the case with the above- 
named three alloys and with mercury, which, in comparison with the 
solid metals, can easily be obtained very pure. As to gold, we must 
remark that the conductivity of this metal at 18° is only 41:3if the gold 
is absolutely pure. If it contains even ;;4,, of iron or copper, its conduc- 
rae Bs reduced to 24°7, and the temperature-coefficient sinks from 3°68 
o 2:08, 
{ Both the aluminium and the bismuth surfaces could not be kept free 
from oxide for a considerable length of time. With aluminium we are 
nearly sure that the observed deviation is due to this fact. But with 
bismuth the greater part of the deviation is apparently due to other 
causes. 
§ Cf. P, Lenard, Wied. Ann. xxxix. p. 626 (1890). 
