Undijj-racted Long-waved Heat Rays by Wire Gratings. 333 



§ 11. On account of the very weak intensity of the radia- 

 tion we could not in our experiments upon Reststrahlen in 

 1904 carry out observations with inclined gratings. With 

 oar present arrangement and the increased sensitiveness of 

 our measuring instruments we were now able to determine 

 the transmissibility ratio as a function of the apparent slit- 

 width. The results are collected together in the following 

 tables and figures : — 



For fluorite Reststrahlen (\ = 24/a) : Table III. (p. 334) 

 and fig. 3. 



For rocksalt Reststrahlen (A, = 52 /a) : Table IV. (p. 335) 

 and fig. 4. 



For isolated long- waved rays (\ = 100yu.) : Table V. B and 

 fig. 5 (pp. 336-337). 



Fi°r. 3. 



Cu1_., Cu2 



,,~ 



^>--. 



"S-- 



^ = 24^ 



>»}Q 



Fipr. 4. 



Y y w" %o" y 



L Aui ' Ag1 _..__.. 



10° 20° 30° 40* 50 J ttf" 



For the 100 {jl waves the directly observed apparent values 

 Q,' and Q/ are cited in Table V. A (p. 337) ; these need 

 correction because the analysing grating Pt la, in this 

 case only, polarized up to 3 per cent. A simple recalculation 

 gives the true values Q s < Q s ' and Q^ > Q p ' in Table V. B. 



A glance shows that the curves are much more complicated 

 than our previous ones : we have therefore preferred to 

 express Q s and Q P directly as empirical functions of /3. The 



