408 BELL SYSTEM TECHNICAL JOURNAL 



is beautifully adapted to certain problems as nature or art present 

 them, not so useful or altogether useless for others. So for instance, 

 the first is enormously the best for the spectroscopy of X-rays; the 

 second is the outstanding method for long electron-waves (of the 



/ 



V 





» 



•.^^' 



• 









t 



Fig. 10 — Laue pattern of a mica crystal. (R. M. Bozorth.) 



order of an Angstrom) and thin films of matter; the third is very 

 useful in the analysis of reasonably large crystals formed by intricate 

 chemical compounds, and the fourth is invaluable for substances like 

 the metals of which the crystals are often very small and tangled up 

 together. But the fourth does not carry the investigator so far into 

 the delicate details of crystal structure, as do the third and the first, 

 when large crystals are available; the third and the first are impotent, 

 when large crystals are not to be had; and the second would be 

 exceedingly toilsome with X-rays. The complete crystal-structure 

 laboratory now contains apparatus for the first method, the third, 

 and the fourth. Perhaps it will not be long before the second also 

 is demanded. 



Before entering into details I wish to comment on four assumptions 

 which have crept unsignalized into these deductions. 



(i) The assumption of the unlimited perfect crystal. As I have already 

 said sufificiently, a crystal composed of only a few atom-groups would 

 produce broad hazy spots instead of sharp ones, as a grating with 



