32 



NEW YORK STATE MUSEUM 



when the analyses are recast according to the methods now much 

 in vogue and extremely useful.^ 



Si02 



ALO3 



FesOa 



FeO 



MgO 



CaO . 



NaaO' 



K2O . 



H2O 



P2O5 



MnO 



Sp. gr. 



I 



2 



3 



4 



5 



59-55 



54.62 



54-47 



53-43 



51.62 



25.62 



26.5 



26.45 



28.01 



24-45 



.75 



.75 



1.3 



-75 



1.65 





.56 



.67 





5-3 



tr. 



•74 



.69 



.63 



1. 21 



I-IZ 



9.88 



10.8 



11.24 



9-97 



5.09 



4.5 



4-37 



4.85 



3-49 



.96 



1.23 



.92 



•96 



1.27 



.45 



.91 



•53 



tr. • 



. .72 

 .01 



.1 



100.15 



99.69 



100.20 



99.87 



99-79 



2.66 



2.7 



2.72 



2.67 



2.798 



1 Chateau Richer, Quebec. T. S. Hunt. Geol. Sur. Can. 1863. 



2 Keene valley. A. R. Leeds. N. Y. State Mus. 30th An. Rep't. 1878. 

 p. 92. 



3 Summit of Mt Marcy. A. R. Leeds. N. Y. State Mus. 30th An. Rep't. 

 1878. p. 92. 



4 Nain. Labrador. A. Wichmann. Zeitschr. d.d. Geol. Gesellsch. 1884. 

 36:491. 



5 Games Quarry, Altona, Clinton co. E. W. Morley for H. P. Gushing. 

 N. Y. State Geol. 19th An. Rep't. 1901. p. 58. 



1 The recasting of analyses was first practised by W. C. Brogger of 

 Christiania about 1890, and has given a new significance to the chemistry 

 and mineralogy of rocks. A simple exposition of the methods employed 

 will be found in Kemp's Handbook of Rocks, the calculations being- 

 pursued only so far as they give results representing actual rock-making 

 minerals. A more elaborate method^ has been developed by Messrs 

 Cross, Iddings, Pirsson and Washington in The Quantitative ClassiAca- 

 tion of the Igneous Rocks, but in its application the authors are forced 

 because of the complicated mineralogy of many rocks to assume some 

 minerals or molecules which, so far as we know, are not in the rocks 

 under discussion. While the variation from the actual mineralogical 

 composition is oftentimes not necessarily great, yet hypothetical con- 

 ditions are unavoidably assumed. In the recasting here employed, only 

 those mineralogical molecules are used which we have reason to believe 

 are in the rock. While the results are not mathematically accurate and 

 while in some cases an excess or a deficit of a component has been 

 encountered, yet the results must be very near the truth. They have 

 their value in that they focus attention upon the percentages of the 

 several minerals rather than, as in chemical analyses, upon uncombined 

 oxids. 



