62 Washington — Statement of Hock Analyses. 



done by some petrographers in the reduction of analyses for 

 certain lines of work, though in these cases oxygen is omitted. 

 This last seems to me a somewhat unjustifiable procedure, in 

 view of the fact that oxygen makes up by far the greatest per- 

 centage in all the known terrestrial rocks. 



It is true that we do not yet know whether the oxides exist 

 as such in the magma or not, though the theories of physical 

 chemistry seem to favor the view that they do. But apart 

 from such considerations, the possible benefits to be obtained 

 are so doubtful and meager that it seems scarcely justifiable at 

 present to make any such sweeping change. Experience has 

 shown the convenience and practicability of the present method 

 for the study of rocks, and as it is sanctioned by long and 

 universal usage, it had better be retained, and the results 

 stated in oxides. 



The statement of the analysis then will be divided into two 

 divisions, the main portion, in which are placed the principal 

 oxides, and the subordinate in which occur all the rest. 



The order in which the constituents may be put is the next 

 consideration. That which is here proposed is as follows : 



Si0 2 , A1 2 3 , Fe 2 0, FeO, MgO, CaO, Na 2 0, K 2 0, H 2 (ignit.), 

 H,O(110°), CO,, Ti0 2 , ZrO s , P 2 6 , S0 3 , (Jl, F, S (FeS f ), Cr,0„ 

 KiO, CoO, MnO, BaO, SrO, Li 2 0. 



As regards the main portion the sequence usually adopted 

 seems eminently proper. This is Si0 2 , A1 2 3 , Fe 2 3 , FeO, 

 MgO, CaO, Na 2 and K 2 0. In this we start out with the 

 chief acid radical and the constituent present in largest amount, 

 and go through successively lower orders of oxides to the most 

 positive radicals, the alkalis. At the same time they are pre- 

 sented in a way which brings the oxides together in their 

 natural petrographic and mineralogic relations. The sesqui- 

 oxides are together ; ferrous iron follows ferric, MgO is next to 

 FeO, as the two go hand in hand in the ferro-magnesian 

 minerals ; CaO is intermediate as is proper, since it is a con- 

 stituent both of these and of the feldspars, and it is next to 

 Na 2 0, as it is associated with it in plagioclase. 



H 2 should follow K 2 0, as it is a highly important and 

 regularly determined constituent. I follow Hillebrand* in 

 urging the separation of ignition and hygroscopic water, put- 

 ting the former first, as being an essential ingredient. Next 

 to these should come C0 2 , as this, with H 2 0, is a measure of 

 the freshness of the specimen, and this character can, there- 

 fore, be told at a glance. Together also they constitute the 

 " loss on ignition " so often given, and in this case can be con- 

 nected by a bracket. 



* Hillebrand, Bull. 148, U. S. a. S., p. 29, 1897. 



