C ml ijfr acted Long-waved Heat Rays by Wive Gratings. 341 



aperture B in the screen C ; the rest of the arrangement 

 was identical with that depicted in fig. 2 for rays of wave- 

 length 100 ft. 



The deflexions were determined for vertical and horizontal 

 directions of the wires of the platinum grating Pt 1, which, 

 as before, acted as analyser; on account of the polarized 

 emission of the Auer mantle, to be referred to below, a certain 

 difference between the two deflexions was noticeable. 



The silver grating Ag 1 was next placed — with wires ver- 

 tical — so that its front surface was exactly in the same position 

 as the above-mentioned silver surface, and the deflexions 

 again measured. From these the reflecting powers R s and 

 Rp of polarized radiations normal and parallel to the wires, 

 respectively, could be easily calculated ; after correcting for 

 the somewhat incomplete polarization of the platinum grating, 

 they gave the following values : — 



R s = 0-766 

 R„ = 0*301 



Q s = 0-022 R, + Q, = 0-79 

 Q„= 0-439 R P + Q„=0-74 



in which the values for Q s and Q^ at 45° for Ag 1 are taken 

 from Table V. B. 



It is thus seen that the sum of the reflecting and trans- 

 mitting powers is somewhat different in the two cases and 

 less than unity. Regular diffraction will hardly occur here, 

 while. \ = 100 fi, a = 91 m 2 fij a / = G4'4 pu; a perceptible diffusion 

 will, however, take place. 



§ 16. A polarized temperature radiation of the gratings is, 

 as a matter of course, very probable : we could not, however, 

 apply white heat to such costly and sensitive appliances. 



Some success was obtained in this respect with a specially 

 woven grating-like Auer mantle, which exhibited a broad 

 zone of exclusively vertically stretched threads : its emission 

 was partially polarized. The radiation polarized || to the 

 threads amounted at 100 p to about 0"60 of that polarized at 

 right angles. Unpolarized radiation was sent out by the 

 glowing magnesia mantle-holder which is superposed on that 

 observed ; on correcting for this, the above fraction would 

 become still smaller and eventually approximate to the value 

 given above, n 2 = 0'55, which the same mantle acting as 

 a grating gave for the transmissibility ratio. The above 

 ratio at 52 /ul was, on the other hand, only of the order - 95. 

 A quantitative comparison can, however, scarcely come into 

 consideration with such undefined structures ; their properties 

 at white heat would also no doubt differ from those at 

 ordinary temperature. 



