we trace out the left-over material which was last deposited; we find secondary growths built up from more primitive substances. Throughout we gather evidence bearing upon the life-history of each mineral, and this may be directly correlated with the conception of chemical structure. When we can determine the conditions under which a compound can be formed, we shall have made a long step forward in understanding the nature of the substance.
The second of our three lines of investigation is closely allied to the first, and, indeed, overlaps it somewhat. It is the study of alterations. A mineral has not only an origin and growth, but also a process of decay, during which its material, disintegrated, is made over into new forms. It is very common to find a crystal with its nucleus unchanged, and its surface transformed into some other species. Some of these alterations are easy to understand; as when, by oxidation, a cube of iron pyrites becomes a cube of the brown hydroxide, limonite; or when an arseniate or sulphate is derived from an arsenide or sulphide. Other changes, however, are less simple, such as the transformation of topaz into mica, or of corundum into margarite; but all of them tell something as to the nature of the substance altered, and help to elucidate the problems of structure and function. An alteration product is the record of a chemical reaction, which may be traced and reasoned about; and the evidence which it offers is quite analogous to that used by the organic chemist to determine the structure of a carbon compound. In the latter case alteration products—that is, derivatives—are produced artificially; in the former the mineralogist finds them ready formed in Nature. Unfortunately, however, such alteration products are not attractive specimens; and the ordinary collector throws them aside as worthless. An altered crystal has lost its perfection and beauty, and is variable only for what it signifies. But, from a scientific point of view, its value is real and considerable, if only it be studied thoroughly, apart from superficial appearances, and without jumping at conclusions. Here, again, the microscope and the chemical analysis are necessary coadjutors.
One line of research yet remains to be considered. The two already disposed of deal with material as gathered in the field; the third is an affair of the laboratory. Of late years many mineralogists have been actively at work upon the synthesis of mineral?, building up their crystals by artificial means, and reproducing in a rapid way the slower processes of Nature. Many species have thus been formed in well-defined condition; and other compounds, different from but closely analogous to well-known minerals, have also been produced. Every year there are great advances in this field of work, and every step which is taken is in the direction of the main problem. Sometimes results are attained by methods unlike those which grew the native crystal; but even then new light is shed upon its nature, and we know more of its possible modes of genesis. Some experiments