between Optical and Electrical Qualities of Metals. 161 
was so adjusted that its centre of curvature lay on the axis of 
the turning-table. By this means an image of equal size 
also appears on the slit H when the lamp is at B. .A proof 
for the correctness of our adjustment is given by the fact that 
the images of the two wire crosses on H are equally exact in 
both positions of the lamp. 
We need simply to add that the slit H was 6 mms. high 
and 1 mm. wide, and that a diaphragm of 24 x 24 mms. was 
placed in front of the mirror F’, in order to prevent any but 
the central part of the cone of rays from entering the slit. 
Thus it is evident that the rays follow the same path in both 
cases, the only difference being the additional distance BS A 
when the lamp is at B. On dividing the deflexion obtained 
in this position by that observed in the first case, the reflect- 
ing-power of the mirror 8 is at once obtained. 
We undertook measurements for about 15 wave-lengths of 
the spectrum. While the Nernst fibre was in the position A, 
the total distance traversed by the rays was 210 cms., in the 
position B it amounted to 270 cms. Thus it became neces- 
sary to limit the observations to wave-lengths for which the 
absorption produced by the water-vapour and the carbon 
dioxide * of the atmosphere is not considerable. For our 
purpose the absorption is small enough, in most parts of the 
spectrum, to prevent perceptible errors. An additional dis- 
tance, small compared with the whole path of the rays, scarcely 
increases the absorption already produced. Of all the wave- 
lengths concerning which measurements were made, X=7 fu 
is the only one forming an exception to this rule. Here a 
correction of about 31 per cent. was necessary, owing to the 
absorption produced by the water-vapour. 
The Mirrors. 
All the metals were used as concave mirrors of 30 cms. 
radius and 4 cms. aperture. They were made of very pure 
material. Three different modifications of silver, gold, and 
platinnm, and two of nickel were examined. All these 
mirrors, except that of cast bismuth, possessed perfect sur- 
faces and gave excellent images. 
Results of the Observations. 
Table 1. (p. 162) contains the results of our measurements for 
seven pure metals and for six alloys. The curves of fig. 2 
(p. 163) show the reflecting-power of ten of our mirrors for 
different wave-lengths. In the red, as well as in the neigh- 
bovring regions of the infra-red, our observations agree 
* Cf. F. Paschen, Wied. Anz. li. p. 51, li. p. 209, and lil, p. 3385 (1894) ; 
also H. Rubens & E. Aschkinass, Wied. Ann. lxiv. p 598 (1898). 
Phil. Mag. 8. 6. Vol. 7. No. 38. Feb. 1904. M 
