Measurements in the Extreme Infra-red Spectrum. 781 



As is to be expected the reflectivity of potassium bromide 

 for its own Reststrahlen is very great ; however, the 

 value here cited is certainly too low, as the plate used for 

 the reflexion measurements was showing cracks. The 

 reflectivities of rock-salt and sylvine for these Reststrahlen 

 are still much higher than is calculable from their dielectric 

 constants. On the other hand, the value of the reflectivity 

 obtained for fluorite corresponds approximately to the value 

 thus calculated; that is, for the dielectric constant K = 6'8 

 of fluorite for infinitely long waves, the reflectivity amounts 

 to 19'9 per cent.* 



In the measurements on water, a small crystallization 

 vessel, filled to the required height, was placed upon the 

 small standard of fig. 12. The observed value is not much 

 different from that obtained from the Reststrahlen of rock- 

 salt f. From the value of g = 0*6* 6, and the reflectivity 

 R. = 9*fi, one obtains the index of refraction of water according 

 to the formula 



100 + R 

 lOU-l-i^ 



/. 100+R \ 2 



V lioo^Hj 



as equal to 1*41 ±. According to the same method w = l'41 

 was obtained for the Reststrahlen of fluorite and n=136 

 for the Reststrahlen of rock-salt. Though this method gives 

 only approximate values, it proves beyond doubt that the index 

 of refraction of water for waves 200 times as long as those of 

 blue light is nearly the same as in the visible spectrum. 



Synopsis. 



The contents of this investigation may be summarized as 

 follows : — 



The wave-lengths and energy distributions of the Rest- 

 strahlen of rock-salt, sylvine, potassium bromide, and 



* A value differing but little from this was obtained previously for 

 the Iieststrahlen of sylvine, i. e. 204 per cent. (II. llubens and 

 E. Aschkinass, loc. cit. p. 253.) 



f By means of an experiment we satisfied ourselves that the layer of 

 water vapour which hovers directly over a water surface at a given 

 temperature, could not affect the measurement by way of absorption. 

 When the relative humidity in the research room was 29 per cent., a 

 cylinder covered on the inner surface with saturated filter-paper was 

 brought into the optical path, but no diminution in the intensity of the 

 radiation was noted. It can therefore be said that no error due to 

 witter vapour absorption crept in. 



\ The second value which the formula gives is 1*01 : this does not 

 come into consideration. 



