THE ROSIWAL METHOD FOR MINERALS 



213 



microscopic measurements of the diameters of grains in thin section 

 made according to the Rosiwal method, the results thus obtained 

 being multiplied by the respective specific gravities to give weight 

 percentages. (Personally the present writers prefer to transfer 

 the specific gravity determinations into volume percentages rather 

 than the reverse, since one estimates the composition of a rock by 

 volumes and not by specific gravities.) The fourth column repre- 

 sents six different photomicrographic prints, made of different 

 parts of the slide, which were dissected and weighed according to 

 the Delesse-Sollas-Joly method. The weights of the fragments of 

 paper were assumed to be proportional to the areas, therefore to the 



TABLE I 



Calculated 



from 

 Chemical 



Analysis 



Heavy 

 Solution 



ni 



Measure- 

 ments of 

 Diameters 

 Rosiwal 

 Method 



IV 



Photomi- 

 crographs 

 of Areas 



V 

 Areas (d^) 

 Measured 



by 

 Microm. 

 Ocular 



VI 



Calcula- 

 tions of d3 



from 

 Column III 



VII 



Calcula- 

 tions of d> 



from 

 Column V 



Quartz. . . 

 Orthoclase . 

 Plagioclase 



Mica 



Magnetite . 

 Hematite. 



35-9° 



34 48 



24.99 



0.26 



2.60 



1 .10 



34-25 

 32.47 

 26.91 



19.02 



43-65 

 32.90 



4-43 



23-23 

 43-64 

 32.90 



0.25 



19.90 



43-67 

 32.90 



3-52 



1. 41 

 56-04 

 42.24 



0.31 



12 . 14 

 49.92 

 37-31 



0.63 



volumes of the minerals. The percentages were computed as 

 before. The fifth column represents measurements of areas made 

 by means of a net-micrometer ocular, as originally used by Rosiwal; 

 the number of squares included within the boundaries of the.differ- 

 ent minerals being counted. The areas again were assumed to be 

 proportional to volumes. The sixth column was obtained by cub- 

 ing the values in column III and reducing the sum to 100. The 

 last column represents d^ calculated from the value of d^ in column V. 

 Williams pointed out the close agreement between the composi- 

 tion calculated from the chemical analysis and the results obtained 

 with Sonstadt's solution on the one hand, and that obtained by the 

 Rosiwal method and the measurement of areas on the other. The 

 two sets of values, however, do not agree. He remarked that 

 "though the absolute volumes (of the microscopic methods) may 

 not be correct, using d as indicated, relative volumes as expressed 



