1920] 



on Low Temperature Studies 



273 



each silver chloride-alcohol surface, and 0*993 for each silver 

 chloride-methylene iodide surface, applied to the observations obtained 

 at each step. 



Rock-.jalt with its lower refractive index would be preferable to 

 silver chloride for such measurements, except that it has to be cooled 

 so slowly to avoid cracking that much more time is required, whereas 

 silver chloride in thin rolled plates is very easily handled. Two 

 superposed membranes with liquid films between are also quite 

 workable. 



Successful trials were made with a hollo wed-out flat capsule of 

 rock-salt fitted with a plane cover plate when necessary. By this 

 means plane layers of substances of definite thickness were obtained. 



'P&o 



Succession of 

 Plates 



2 3 4 5 



No. of Screens 



Fig. 18. 



With ordinary liquids the whole reflection correction is very small, 

 being practically the same as for a rock-salt plate alone — namely, 

 2 • 7 per cent, in liquid air (for the " D " line). This cell was to be used 

 for measuring the trans missive powers of liquid gases (such as 

 methane) with a " gas-cooled " thermoscope. With liquid immersion 

 in the ordinary way, the cover plate would be necessary. This, 

 however, was not carried out. 



The transmission through a succession of plates of silver chloride 

 alone was also measured for the purpose of studying the rate of 

 extinction of incident radiation by successive reflections. Thin 

 membranes and thin discs of mica and glass were also measured. 

 The results observed are shown in Fig. 18. 



A simplified method for several surfaces has been pointed out 

 Yol. XXIII. (No. 114) T 



