RELATIONS TO OTHER SCIENCES 593 



lends attractiveness to them, were to be withdrawn! That petro- 

 graphic-geologic theory, by means of which Bunsen would explain 

 the varied chemical compositions of the eruptive rocks, is reflected in 

 Tschermak's ingenious and fruitful conception that the triclinic feld- 

 spars consist of a series of mixtures of two chemically different but 

 isomorphous end-members. 



As a matter of course, in all these mineralogic petrographic studies 

 physical methods are continually employed. While it is true that the 

 optical appliances of physics have become the common property of 

 the petrographer, yet it must not be forgotten that the latter has 

 also invented new instruments after special patterns and has made 

 valuable improvements in others, all of which redounds to the advan- 

 tage of general physics. A further service to physical science arose 

 from the fact that a considerable portion of the laws of heat and 

 optics had to be first investigated or verified by means of substances 

 which belong to the mineral kingdom. Again the physical method of 

 procedure used to separate heterogeneous mixtures by means of heavy 

 solutions has been brought to yet greater perfection since its applica- 

 tion to petrographic problems. The investigations which are endeavor- 

 ing to apply the laws of mechanics in the study of rock-masses sub- 

 ject to deformation, torsion, or fracture are partly petrographic, but 

 chiefly geologic in character. 



We have long had lump chemical analyses of rocks, as well as partial 

 analyses dealing with those rock-constituents dissolved or decomposed 

 by acids, and those not attacked; and also analyses of the individual, 

 isolated, rock-forming minerals. To be sure, all such analyses were at 

 first considered as ornamental trimmings to the rock description, and 

 they were frequently executed by rather inexperienced novices. For 

 a while, also, the chemical analysis of rocks was neglected, because 

 the rapidly increasing study of the carbon compounds seemed to be 

 a more attractive and even lucrative field. At present the application 

 of the methods of chemical analyses to the study of petrographic 

 material is more general than ever before, and the undeniable signifi- 

 cance of the results cannot be too strongly pointed out. Very properly 

 the massive eruptive rocks and the crystalline schists continue to ex- 

 cite the most interest. Indeed, in recent years too much weight seems 

 to be given to special chemical peculiarities if one is thereby induced 

 to establish new and burdensome names for these rock-masses which 

 are certainly non-stoichiometric in composition, merely on the basis 

 of slight variations in the amounts of either the monovalent or the 

 bivalent metals, or of both. 



In recent years the United States Geological Survey also has made 

 many very valuable individual contributions to the science. Among 

 these are many hundreds of analyses executed with ever-increasing 

 completeness and accuracy, which have shown that such supposedly 



