2i8 William E. Ford, 



In studying the analyses of these minerals that the writer has 

 collected, it is quite evident that there are certain definite restric- 

 tions to the amounts that the various molecules may replace each 

 other in the different minerals. Complete isomorphous replace- 

 ment in all proportions would only be expected in the case of 

 two molecules having naturally identical space lattices or with 

 lattices that may change progressively from one end of the series 

 to the other. In other instances the amount of replacement 

 would be conditional upon the similarity between the two lattices 

 and would proportionately diminish in amount as this similarity 

 decreased. The restrictions noted in the analyses clearly 

 conform to the demands of this theory. 



All of the analyses studied have been recalculated into per- 

 centages of the various carbonate groups and then reduced to a 

 100 per cent summation. Where close accuracy seemed unneces- 

 sary much of the calculating was done by means of graphical 

 methods, which, however, are accurate certainly within two 

 tenths of one per cent. In all the analyses which are quoted 

 in the tables given below the results were obtained by complete 

 arithmetical calculations. In the case of calcite the study of the 

 molecular volumes and angles given above would lead one to 

 expect that only comparatively small amounts of the other mole- 

 cules would be found replacing the CaCOo. Further, of the 

 various possible groups, IMnCO, would occur in largest amount, 

 FeCOa iT^ext and MgCOg in smallest amount, this being the order 

 in which their space lattices progressively differ from that of 

 calcite. 



In the analyses which have been collected there were some 

 twenty-nine analyses of calcite that showed isomorphous replace- 

 ments. All of these as far as a critical study was possible 

 seemed trustworthy analyses. A few others were rejected from 

 consideration because they showed distinct evidence of inac- 

 curacies in the analysis or of impurities in the material analyzed. 

 The average percentage of all the other 'carbonate molecules 

 present in these analyses was 5.2. In only six cases did this 

 percentage become equal to or greater than 10 per cent, the 

 highest of all being 15.06 per cent. In twelve cases this per- 

 centage was five per cent or above. The rhodochrosite molecule, 

 MnCOg, was present in eleven cases, usually in quite small 

 amounts but in three instances gave percentages of 11.2, 11.4 and 



