[Annals N.Y. Acad. Sci., Vol. XVIII, No. 3, Part II, pp. 129-146. 4 April, 1908.] 



ON DETERMINATION OF MINERAL CONSTITUTION 

 THROUGH RECASTING OF ANALYSES.^ 



By Alexis A. Julien, Ph. D. 



Introduction. 



The recognition of the aggregate character of rock constitution, even in 

 varieties of aphanitic texture, has led the analyst in recent years to rearrange 

 the determined chemical components of a rock in the form and propor- 

 tion of its existing mineral constituents. The now well-known advantages 

 of this practice, in the bearing of its results on the true character and probable 

 origin of a rock, are bringing about a complete revolution in petrographical 

 science. The day of the representation of the material of a rock by a mere 

 report of its chemical analysis has now passed. 



The early mineralogists were accustomed frequently to transpose analyses 

 of a mineral substance into the proximate mineral constituents known at 

 that time, such as calcareous minerals and ores into various carbonates 

 and oxides. With the silicate minerals however the increasing list of known 

 minerals soon became burdened with an indefinite series of hypothetical 

 compounds, proposed by Rammelsberg, Tschermak, Knop and their suc- 

 cessors. The difiiculty and uncertainty attending the use of these, in 

 interpretation of chemical analyses, have perhaps served to discourage the 

 continuance of the ancient method; so that at present the discussion of the 

 chemical composition of a mineral generally ceases with presentation of its 

 analysis, accompanied by oxygen ratios and a formula. 



A chemical analysis alone, particularly of a complex compound, such as 

 a silicate, rarely conveys — even to the eye of an expert mineralogist — 

 much more than a vague guess or estimate of the distinctive character of the 

 combination. A glance, for example, over an analysis of a chlorite, sepa- 

 rately presented, would hardly enable him to assign it with any certainty 

 to the page-full of selected but widely varying analyses of penninite or to 

 those of clinochlore or to those of prochlorite comprised in every treatise 



1 Presented to the Academy at the meeting on 6 January, 1908. 



129 



