WRIGHT AND VAN ORSTRAND : MINERAL ANALYSES 223 



MINERALOGY. The determination of the order of agreement 

 between observation and theory in mineral analyses.^ Fred. E. 

 Wright, Geophysical Laboratory, and C. E. Van Orstrand, 

 Geological Survey. 



In a recent number of this Journal, ^ Dr. W. T. Schaller 

 published an interesting note on ''The Calculation of Mineral 

 Formulas" in which he showed that the ordinary method of cal- 

 culation — by first reducing the weight percentages of an analysis 

 to mol numbers* and then dividing these by one of the mol num- 

 bers thus obtained — furnishes results whose degree of approxima- 

 tion to whole numbers is less than that of the values derived by a 

 new method of calculation which he described. In view of the 

 fact, however, that certain details of his method are open to criti- 

 cism, it has seemed appropriate to the present writers to consider 

 the principles underlying the subject, and to determine if possible 

 the most logical basis and method for comparison. From 

 Schaller's paper, the mineralogist or chemist may perhaps be led 

 to infer (1) that the ordinary methods of calculation furnish 

 only a rough measure of the degree of approximation of a chemical 

 analysis of a mineral or chemical compound to the values derived 

 from its chemical formula; and (2) that by a somewhat involved 

 calculation the analysis can be expressed in more accurate form. 



1 Published with the permission of the Director of the U. S. Geological Survey. 



2 Jour. Wash. Acad. Sci., 3: 97-98. 1913. 



* The term mol number is here suggested as an appropriate name for the number 

 obtained by dividing the weight percentage of a chemical substance in any given 

 chemical analysis by its proper molecular (respectively atomic) weight. Such 

 numbers have usually been called molecular ratios, but they are, rather, numbers 

 indicating the number of mols (gram-molecules) of each substance present in n 

 grams of the material analyzed where n is the sum of the analysis. The terms 

 molecular ratios or molecular proportions apply logically only to the numbers of the 

 chemical formula. Thus in the analysis cited below, the chemical formula num- 

 bers or molecular ratios are 



S : As : (Ag + Cu) = 11 : 2 : 8, while the mol numbers from the analysis are 

 S = 0.5444; As = 0.1008; (Ag + Cu) = 0.3975. To use the term molecular ratios 

 indiscriminately for both sets of numbers is not conducive to precise statement and 

 is not to be recommended. The term mol number seems preferable to molecular 

 nujnber or molecular quotient (Molekularzahl Molekularquotient) which are in 

 current use in German. The writers are indebted to Drs. Washington and 

 Niggli for detailed discussion of the proposed term. 



