284 BELL SYSTEM TECHNICAL JOURNAL 



experiments), but only 10% at GOA and only 3% at 40A. The per- 

 formance was much better at a still more oblique incidence; on the 

 other hand Holweck thinks that it gets rapidly worse as the incidence 

 is made more nearly normal, and if this is true the outlook for the 

 concave grating, with its condition of almost normally-incident light, 

 is most unpromising. 



Below the boundary 13A, the atomic constitution of solid sub- 

 stances turns from a hindrance into an advantage, and crystals serve 

 as natural diffraction-gratings of incomparable fineness — too fine, 

 indeed, for our convenience in this part of the spectrum, since the 

 boundary is fixed by the smallness of the distance d between suc- 

 cessive layers of atoms in the diffraction-grating. Rocksalt, one of the 

 standard crystals, for which (/ = 2.814A, has been used successfully 

 up at least to 4 A (by Fricke) and the rest of the way to 13A has been 

 explored with crystals of gypsum ((f = 7.58A) or sugar ((/ = 10.oOA); 

 in this region it was necessary to evacuate the light-path, precisely 

 as in the region beyond 136A. The only possibility of a new advance 

 depends on the utilization of crystals of still greater inter-atomic 

 spacings. Holweck mentions a crystal with a formidable name, for 

 which ff = 19A, and do Bniglie and Friedel found that the oleates of 

 sodium, potassium and ammonium prcsenttxl spacings of the order of 

 40A between consecutive molecule-layers. If these substances can be 

 adapted for use in crystal spectograjjhs, the boundary of the explored 

 region may be pushed far beyond its present place. It may be found, 

 howe\er, that the crystal absorbs the ra\s before they go deeply 

 enough to be diffracted. 



As for the region between 13A and 13GA, no one has ever measured 

 the wave-length of a radiation lying within it; but there is a method 

 which indicates the existence and something about the wa\e-lenglhs 

 of rays which almost certainly belong in it. In apjilying this method 

 the photographic plate is replaced by a metal electrode (usually of 

 platinum) which, when irradiated by rays of any wave-length (less 

 than a certain critical one which always lies far above this range) 

 emits eIectron>. Vtw this reason it is often known as the photoelectric 

 method. alth()iii;ii ihc suhslitulion of one kind of receix'er for another 

 is iinl ii> niii^t (li^tiiu'iis I' ili,ir.utiTi>i ic. .\ target made of the sub- 

 stance to i>e sluilied is si'.ded in in a I iilic, ii|)p(>sili' a source of electrons 

 (generalh' a filament for lIuTiniunic rmi^simi ) ; the photosensitive 

 electrode is placed somewhere in the tui)e where w!iale\er ra\"s are 

 excited at the surface of the target will fall direclh' upon it. It is all- 

 important to i)rotect this electrode from electrons and inns, negative 

 or positi\e. proceeding from target, lilanuiit, or anxwhere else. 



