also bear upon the subject of alterations; and definite alteration products are artificially obtained. So far, little has been done toward generalizing upon this class of observations; but, as the facts accumulate, new relations will appear and reasoning must follow. Each experiment suggests new experiments; each discovery points toward others, and the connecting theories will grow up around the conception of chemical structure. It is the only conception yet clearly recognized which is general enough to cover the whole field.
It has already been argued that the physical study of minerals is subordinate to their chemical investigation, for the reason that all properties depend upon composition. Physical researches, nevertheless, have great value in mineralogy, and a paper under the caption of this essay would be wretchedly incomplete if it failed to consider them. Physical data, moreover, aid in the discussion of chemical structure, and point out analogies of weighty significance. Specific gravity, for instance, is always an important datum in the study of a species; and the ratio between it and the molecular weight of a compound tells us something of the condensation which the elementary material has undergone in combining. One eminent mineralogist is now using this ratio as a basis for mineral classification, especially among the silicates; and his results are likely to emphasize the conclusions drawn from quite different sources. This method of study, however, presupposes a knowledge of true chemical composition. With the latter it means much; alone it signifies little.
Upon the thermal and electrical properties of minerals comparatively little has been done; and that little has slight reference to mineralogy in general. The optical constants, on the other hand, are elaborately studied by mineralogists, on account of their direct relations to crystalline form. Indeed, optical and crystallographic work is a dominant feature of modern mineralogical investigation, although a great part of it never rises above the plane of mere descriptive detail. In its higher aspect it deals with the internal molecular structure of crystals, and so furnishes data which may some day be connected with the broader general conceptions of the chemical field of research. The question of how the atoms are grouped has a mechanical as well as a chemical side; and some time it will be systematically attacked from both directions. At present we only see the future possibility of so handling the physical evidence; but the expectation is philosophically just. To-day a knowledge of crystalline form is mainly useful in the identification of minerals; for by it we may determine a species without destroying the specimen; but its deeper potential significance is none the less apparent. Along these lines we may safely prophesy progress, which can only end in the complete correlation of all mineralogic facts, and therefore in the solution of the fundamental problem.
Looked at from the descriptive side alone, mineralogy is a small