part of the pattern of the mixture, the lines belonging to the pattern of the 

 identified constituent are marked (on the pattern or a corresponding tabulation 

 of data). The procedure is now repeated for the remainder of the pattern, 

 again starting with the strongest remaining line. In this way all the constituents 

 of the mixture can be identified, provided their patterns are catalogued in the 

 index, when fluorescent scattering is small (recognized by light background in 

 the X-ray pattern). The relative amounts of the ingredients present are deduced 

 from the relative intensities of the lines in the pattern, as compared to the 

 intensities of the lines in the pattern of the pure constituents, the exposure times, 

 of course, being the same for all patterns. A series of underexposed patterns 

 (1, 21/2, 5, etc., percent of the total exposure time) of the pure constituent in 

 question will be of considerable help in estimating these intensities. 



If the absolute proportion of each compound in the mixture is to be deter- 

 mined, a synthetic specimen must be prepared from the identified pure materials 

 in such proportions that the synthetic mixture yields a pattern matching in 

 spacing and intensity all the lines of the original pattern, when both patterns 

 are prepared under identical conditions of exposure and processing. If line 

 shifts, fading of the pattern in general with increasing values of the 26 angle, 

 or other differences are observed in the patterns, irregularities of composition, 

 such as solid solutions, are indicated and the compound composition of the 

 specimen must be determined by calculation from a chemical analysis. The 

 chemical analysis frequently is best accomplished by means of standard spectro- 

 graphic procedures. For a thorough study of mixtures of silicates, the methods 

 of X-ray diffraction analysis (Clark, 1955; Ballard, 1946 and 1940) are 

 practically indispensable. These methods reveal the various chemical combina- 

 tions in which the silicon exists, whereas chemical or spectrographic methods 

 alone yield only the total amount of silicon in the unknown, giving no clue as 

 to its mode of combination. 



In the original and first supplementary sets of cards, the values of the d- 

 spacings corresponding to the three strongest lines, together with their corres- 

 ponding relative intensities, appear in the upper left-hand corner of each card 

 (fig. 9-9) . There are three cards in the file for each diffraction pattern; the first 

 card has the strongest line of the pattern at the extreme left and also contains 

 the complete pattern data and some crystallographic data where available. The 

 second card has the second strongest line in this position, and the third has the 

 third strongest line in this position. The cards with the second and third strong- 

 est lines at the extreme left position are only "follow" cards and do not contain 

 any data other than the c?-spacings corresponding to the three strongest lines. 

 The cards are filed in straight numerical order. 



The revised original and supplementary sets include only one card for 

 each pattern, so as to reduce the required number of cards. A book entitled 



167 



