178 Profs. Hagen and Rubens on some Relations 




TasieE VII. 
_| Mean deviation 36°5 
A. |C,=(100—R) V« | of the single pro- Q= = 
observed. ducts from the VX 
average Value. computed. 
4p 19-4 21:0 per cent. 18°25 
8 13:0 145 12°90 
12 11:0 9°6 10°54 
25°5 7°36 49 7°23 
| 
Moreover, this table contains the average deviation (ex- 
pressed in percents) which the single products (100—R) V« 
of the different metals show, as compared with the average 
value given in the second column, 
3. According to Maxwell’s theory the quantity 100—R 
undergoes a change corresponding to the change of re- 
sistance which the metals show with increasing temperature. 
This has been verified by our experiments. 
These two facts (2 and 3) form a new important proof of 
Maxwell’s theory, and that in a sphere, in which hitherto no 
connexion between the observed facts and the theoretically 
computed laws had been recognized. 
4. From our observations the conclusion must be drawn, 
that the periods of vibration of the molecules scarcely in- 
fluence the optical behaviour of metals in the region of long 
waves. 
5. No influence of the magnetic qualities of iron and 
nickel* on their behaviour towards these rays could be de- 
tected. Had such influence made itself manifest relative to 
these metals, the value (100—R) ought to have been con- 
siderably larger than it appears when computed from the 
conductivity for constant electric currentt+ by aid of 
formula (4). 
6. In the region of long waves we are, therefore, justified 
in assuming the agreement of the other optical constants 
with the values computed from Maxwell’s theory. This 
* This fact can be explained by Prof. Drude’s theory of magnetism 
(Cf. Verhandl. d. Deutschen Physikal. Geselisch. v. pp. 143 & 148, 1903). 
+ In the experiments of Mr. V. Bjerknes (Wied. Ann. xlvii. p. 69, 
1892) the influence of magnetism very strongly manifests itself for 
vibrations of about 10° per second. This is partly due to his method of 
observation, which greatly differs from ours. 

