Reflexion of X Rays by Crystals. 



893 



Sine of 

 glancing angle. 



1x0-1097 



v2x ;, 



V3X „ 



2X „ 



2V2x „ 



3X „ 



3V3X „ 



Observed 



values at 



15° C. 



Calculated, 

 taking 

 = 200. 



Calculated, 



taking 



0=280. 



i 



100 



100 



100 



41 



46'3 



47-6 



244 



28-5 



302 



18-7 



198 



21-6 



7-05 



7-2 



8-9 



6:25 



60 



7-5 



4-2 



3-6 



49 



Proportional 



change on 



raising to 



370° C. 



Calculated 

 change. 



1-07 



1-075 



1-20 



116 



1-26 



1-35 



207 



1-90 



1-94 



1-92 



— 



— 



On the whole there is certainly a surprising agreement. 

 I find also that fluorspar shows a small temperature effect : 

 this is also in agreement with Debije's theory, for the cha- 

 racteristic temperature is large and the constant B is small. 



The diamond is a very perfect crystal, and the intensity 

 curves which are obtained when the crystal is turned round 

 step bystep have none of the irregular forms given by rocksalt. 



In fig. 6 the points represent the results (uncorrected for 

 inequality of the electroscope scale) of such a series of 

 measurements. The steps are small, there being twelve for 

 every five minutes of arc, and the regularity of the curve 

 which may be drawn through the points is evidence of the 

 regularity of the crystal. The discontinuity on the right-hand 

 side of the curve is due to the double nature of the rhodium 

 line. Fig. 7 shows the first and third spectra side by side,, 

 each being the mean of several experiments. The readings 

 of current have been corrected for want of uniformity of the 

 readings of the electroscope. The readings of the third 

 order are magnified five times. The separation of the con- 

 stituents of the doublet is complete in the higher order 

 spectrum. The two constituents are then separated by 

 twelve minutes of arc. In the figure is shown the result oB 

 compounding the two when separated by only four minutes 

 of arc as they are in the first spectrum, and it is seen that 

 the result gives a curve of the experimental form. 



The areas of the two curves are to one another as 8*95 

 to 1. But it must be mentioned that the diamond employed 

 was a thin slip which did not completely absorb the primary 

 rays even when it made an angle with them of only eight 

 and a half degrees. It is not certain therefore that the 



