210 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1920. 



destructively and sometimes constructively — depending upon the 

 absolute magnitude of the atom itself and upon the wave length of 

 the X rays — with those rays scattered by electrons on the opposite 

 side of the atom. This idea of atoms of appreciable size almost 

 certainly must be called into play in order to account for the peculiar 

 Way in which the intensity of reflection varies with the angle of 

 reflection. 



This field of studying the intimate structure of the atom is quite 

 untouched at the present time. The results which are to be obtained 

 from it are, as we have seen, of more or less doubtful value until 

 the determination of the structure of crystals has been put upon a 

 very certain basis. When this has been done — that is, when we have 

 succeeded in determining uniquely the structures of a few crystals — 

 these methods of X ray experimentation quite properly can be ex- 

 pected to yield very useful information. 



The method of poioders. — The second method of obtaining diffrac- 

 tion effects from crystals (the method of powders) is in a sense a 

 generalization of the one just discussed. If a parallel beam of mono- 

 chromatic X rays is caused to fall upon a fine crystalline powder 

 put in place of the crystal at C of figure 10 (p. 201), a number of 

 images of the slit produced by different planes of atoms of the 

 crystal will be obtained upon the photographic plate at D. For in- 

 stance, if the powder is that of a cubic crystal, some of the particles 

 of this powder will have the orientation necessary for reflection 

 from, let us say, the cube face, others for reflection from the octahe- 

 dral face, and still others from the dodecahedral face, and so on; 

 consequently images from each of these planes will appear upon the 

 film. The kind of photograph that is obtained in this way is shown 

 in figures 16 to 19. 17 



The information supplied by a powder photograph is about the 

 same as that obtained by spectrometer measurements, except that 

 now a single experiment furnishes a large number of reflections 

 which in the other case would have to be made upon each face 

 separately. This is obviously a great advantage. The procedure at 

 the same time has several serious disadvantages, however: (1) It is 

 greatly to be doubted if the photometered intensities of reflection, 

 such as are to be obtained from these photographs, are at the present 

 time as accurate as are the results derived from the use of the ioniza- 

 tion chamber; (2) the amount of material which is reflecting X rays 

 in the formation of any one of these images is extremely small, so that 

 relatively very long exposures are required; (3) partly because of 

 this small amount of diffracted energy, the images are of consider- 

 able width, so that planes having about the same relative spacings— 



11 Figs. 16 and 17 are given l>y A. W. Hull, op. cit. ; figs. 18 and 10 by H. Bohlin, Ann. 

 d. Phys., 51, 421, 1920. 



