WHURRY: reply to dr. TUTI'ON S DIvSCUSSION 99 



results that cobaltite may belong to the tetrahedral pentagonal 

 dodecahedral class 28, like barium nitrate. The Braggs have 

 shown that the alkali chlorides are most probably holohedral. 

 The rutile group requires much more research, there being no 

 satisfaction in building conclusions on contradictory data. 



The present moment is a dangerous one in the history of the 

 use of X-rays in unravelHng crystal structure. No more specula- 

 tions built on incorrect crystallography are desirable. What 

 is needed is solid, well and carefully carried out, prolonged and 

 thoroughly tested experimental work, and a complete revision 

 of the principles on which results are based, with the view of 

 rendering them both more fully trustworthy and of definite 

 application. Two fundamental problems are especially urgently 

 requiring solution before much further progress can be made, 

 namely, the falling away of reflection intensity with increase 

 of order of spectrum, and the quantitative relationship between 

 reflection-intensity and atomic number or atomic weight (mass). 

 While we cannot hope to get much more information from the 

 Laue radiograms than at present, the Bragg spectrometric 

 method is full of promise, and when these root-problems are 

 satisfactorily settled much more progress may be expected to 

 be made with the finer details of the structure of the more im- 

 portant crystalline substances. 



Yelverton, S. Devon, England. 

 November 18, 1918. 



CRYSTALLOGRAPHY.— i?^:^/,)' to Dr. Tutton's discussion of 

 the assignment of crystals to symmetry classes. Edgar T. 

 Wherry, Bureau of Chemistry. 



In the course of his scientific study of natural phenomena, 

 man is continually devising pigeon-holes into which to distrib- 

 ute given series of facts. Nature, however, often refuses to be 

 pigeon-holed, and persists in bringing to the attention of all 

 who will stop, look, and listen numerous facts which do not 

 accord with the classification in vogue at a particular time. 

 New classifications must therefore be continually worked out 



