PREPARATION OF PURE MINERALS 155 



components. This same vein quartz, when melted in the laboratory, is 

 among the most viscous of minerals, and takes the form of the containing 

 vessel only when placed under stress. It is, therefore, more than probable 

 that it once contained ingredients which contributed enormously to its 

 fluidity during solidification, but of which only a trace is left — that is, 

 volatile ingredients, like alkaline water solutions, or fluorine; the same 

 materials, in short, which produce the same result in the laboratory. 



Another difficulty arises through the tendency of rapidly crystallized lab- 

 oratory products to come out in a cryptocrystalline condition. If the short 

 time interval available for laboratory experiments is insufficient to provide 

 opportunity for the formation of large units, may it not also be true that 

 the size of grain in natural formations gives some indication of the rate 

 at which the formation went on? An inference of this kind must, of 

 course, be applied with care if volatile ingredients are present, for small 

 differences of composition often produce relatively very large differences 

 in the fluidity of the magma, and therefore in the size of the crystal 

 grains. In fact, the introduction of volatile ingredients is the expedient 

 used in the laboratory production of minerals which tend to crystallize 

 only in the most minute units. 



As a record of laboratory experience, it is unnecessary to do more than 

 call attention to the character of these difficulties which the preparation 

 of the pure minerals in the laboratory encounters, and to say that none 

 of them has j-et proved insuperable. J^early all our publications contain 

 illustrations of the practicability of obtaining pure mineral types of suffi- 

 cient perfection for competent study. Given these pure types, and there 

 is no difficulty of principle in establishing properties and relations and 

 erecting on them a quantitative system of solutions corresponding to the 

 typical natural rocks. In practice, the development of such a system is 

 slow, for the number of participating components in natural rocks is ver}^ 

 large, and the complications accordingly intricate. The solution of the 

 difficult cases must await the gathering of a wide range of laboratory ex- 

 perience and a considerable elaboration of existing theory. On the other 

 hand, the attempt to build a physico-chemical system on observations of 

 natural minerals alone will surely resemble an attempt at triangulation 

 without fixed triangulation stations. 



All GOVERNING Conditions must be knov^n 



If the first essential to the successful study of the properties and rela- 

 tions of the minerals in the laboratory is the ability to obtain pure types, 

 the second is certainly the necessity of defining the conditions under which 



