602 Dr. Marine Siegbalm on 



crystal (and of the plate-bolder)*. Only in this case the- 

 exact angle of the crystal is without influence on the place 

 of the spectral lines on the plate. In this case also do the 

 same results come out when the crystal during the expo- 

 sure is continually turned over a little range of angles or 

 remains still. In using bad crystals small irregularities may 

 in this way be eliminated, if necessary. All the following 

 measurements are carried out with selected crystals where 

 such precautions were superfluous. 



For the determining of the lattice-constant of potassium 

 ferrocyanide only a bad specimen was obtainable. These 

 results are therefore not of the same accuracy as the others, 

 but are sufficient for the purpose in question. 



As the long waves which this investigation ranges over 

 are considerably if not wholly absorbed in atmospheric air, 

 the spectrograph was built for vacuum (O1-O01 mm. Hg). 

 Its construction may be seen in fig. 1. To the plate-holder 



Fiff. I. 



is attached a circle-scale of great accuracy which is read 

 off with two microscopes. Another circle-scale of smaller 

 weeuracy allows the approximate turning of the crystal in, 

 the desired position. 



* Bragg, ' X-Rays and Crystal Structure." 



