THIRD HYPOTHESIS OF DERIVATION 



46' 



No fixed method can be adhered to in application of the process ; with 

 each rock the procedure should be influenced, to some extent, by its 

 chemical composition. In this case, to diminish error in connection 

 with the heaviest grains, the ores, on account of their irregular distribu- 

 tion (plate 62), three separate determinations of these were made on such 

 photographs, yielding the percentages (corrected by specific gravity for 

 weight) 5.58, 4.60, 6.75. From these it was inferred that the percentage 

 of FeoOj, 5.29, in the chemical analj^sis, represented chiefly that of ores 

 separated in form of ferric oxide. For the other constituents in the re- 

 maining 94.71 percent the following determinations were made: The 

 weight of paper pieces, representing hornblende, amounted to 1.9211 

 grams; the relation of areas in the actual and estimated plates (see A, 

 method IV), was found to he I X b : d: : : 0.95 X 0.32 (= 0.304) : 

 0.09078. Then, 0.304 : 0.09078 : : 1.9211 : .5737; that is, the estimated 

 area of plate in proper areal proportion to the true volume of the horn- 

 blende blades ought to be less than one-third of the actual area of the 

 plates, as seen in thin-section on the schist plane. 





« 2 









^ . 





03 



>. 



58 





5 ^ 





^32 



> 



■s^ 



£- 



2: 



Minerals. 



o O ^ 



^ 



^rs . 



^ 



^^ 





-ts •'-' OD 



Zi 



■i^ S OD 





-M <JJ 









;=; S s 



fi 



s P 





:^ss 



*3 



-gss 





Sl 





1-^ 



3; 





53 



i^ 



Quartz 



.1110 

 .2482 



2.65 

 2.70 



.2942 

 .6566 



9.01 

 20.12 



9.08 



Feldspar and calcite 



20.28 



Zoisite 



.0631 

 .0070 

 .5737 



3.20 

 3.00 

 3.30 



.2019 



.0210 



1.8932 



6.19 



.64 



58.01 



6.24 



Biotite 



.65 



Hornblende 



58.46 









.9980 



3.27 



3.2635 



93.97 



94.71 



Ores 











5 29 













100.00 



For discrimination of the ores, the application of a strong magnet to 

 2 grams of finely powdered rock of the original sample separated mag- 

 netite to the amount of 0.37 per cent. A few particles of pyrite also 

 had been recognized in the rock and in its thin-sections. The remain- 

 der, the greater portion, consisted of hematite. The figures of the pub- 

 lished anal3^sis of biotite from Monroe, Connecticut, were used in calcu- 

 lation for the following statement; for the plagioclase feldspar, those of 

 the analysis by von Rath of labradorite from diorite of Veltlin ; and for 



