332 Shepherd, Rankin, Wright — Binary Systems of 



isolated grains *02-'03 mm in diameter, although for ease of 

 manipulation and general accuracy larger grains and sections 

 are preferable. It may be stated as a general rule that most 

 of the optic properties can be determined with sufficient 

 accuracy on grains measuring -02-*05 mm in diameter, and certain 

 optical properties on still smaller particles. 



The great advantage of examining a preparation in powder 

 form rather than in the thin section is two-fold : (1) by the use 

 of refractive liquids, the refractive indices can be determined 

 at once and traces of inhomogeneity in a compound appear 

 most clearly if it be immersed in a liquid of the same refrac- 

 tive index ; (2) the individual grains in the powder are isolated 

 and can be rolled about in the liquid and examined along 

 different directions if necessary. These two conditions are 

 difficult to obtain in the thin section. The chief disadvantage 

 of the examination of preparations in the powder form is the 

 loss of texture. In the thin section, the texture or relation of 

 the different crystallites to each other is much more clearly 

 marked than in the haphazard particles of a powder prepara- 

 tion. 



As a general rule, the morphologic development of crystal- 

 lites from artificial melts is poor and crystals suitable for 

 goniometric measurement are very rarely obtained. The 

 crystal system of any compound has to be inferred, therefore, 

 from the optic and crystallographic properties obtained by the 

 microscopic investigation alone. 



Briefly summarized, the optical investigation is adapted 

 primarily to ascertain the mineral composition of the prepara- 

 tions of any given series, while the thermal work serves chiefly 

 to establish the stability ranges of these compounds, both alone 

 and in the aggregate, at different temperatures throughout the 

 series. Other evidence — specific gravity determinations, crys- 

 tallographic features, chemical behavior, etc. — tends further to 

 supplement and to substantiate the thermal and optical data. 



Evidence of this nature is in part chemical and part phys- 

 ical, and properly falls in the domain of physical chemistry, 

 for its interpretation — wherefore the extreme importance of 

 this science in the study of rock and ore formation. 



Geologic data. — In the study of rocks their microscopic 

 examination and bulk chemical analysis have heretofore 

 received the most attention, and in fact petrography, which 

 has primarily to do with rock description and classification, is 

 a result of this study. The general science of rocks, however, 

 demands not only rock description and classification, but also 

 inquires into their formation and genesis, and this fact postu- 

 lates thermal evidence. In the petrologic treatment of rocks, 

 exact thermal evidence is just as essential as precise optical 



