142 Washington — Suggestion for Mineral Nomenclature. 



teristic crystal form, be extended to minerals in general and 

 made the basis of a mineral nomenclature. From this point of 

 view silicate minerals would not be considered simply as salts of 

 orthosilicic acid (H 4 Si0 4 ), metasilicic acid (H 2 Si0 3 ), disilicic acid 

 (H 2 Si 2 5 ), and so forth ; as substitution derivatives of normal 

 aluminum silicates according to Clarke, (which correspond to 

 the old theory of types) ; or of multiples of silica according to 

 Goldschmidt : but each group would be considered as salts of a 

 particular silico- or alumino-silico- acid characterized by the parti- 

 cular crystal form and symmetry of its salts, and capable of 

 ismorphous replacement either of its basic hydrogens or in the 

 acidic portion. 



This concept may be most appropriately applied to mineral 

 groups, distinguished as at present by identity of chemical 

 type and close similarity in crystal form, but it may be equally 

 well applied to monotypic "groups", represented, so far as now 

 known, by only one mineral and which exhibit no evident 

 near relationships, either chemical or crystallographic, as is the 

 case with beryl and calamine. It is also clear that it would 

 be applicable to minerals of simple as well as highly complex 

 constitution. A further point, and one of great importance, 

 is that a system of nomenclature based on this concept would 

 be applicable whether the structural or constitutional formula, 

 or even the exact chemical composition, of the mineral acid 

 were known or not, as the salts of a given mineral acid (in this 

 sense) would be identifiable and their relationship established 

 by their conforming to a certain empirical chemical formula 

 and crystal form. Thus the various pyroxenes and amphiboles 

 are members of two different, but well-characterized, and uni- 

 versally recognized "natural" groups, though we are as yet 

 almost wholly ignorant of their molecular constitution, know- 

 ing only that they may be referred, but do not necessarily be- 

 long, to the metasilicates, but are undoubtedly much more com- 

 plex than is indicated by the empirical formulas. We cannot 

 even determine which group is the more complex. 



Such a nomenclature would be rational and would be analo- 

 gous to that of inorganic chemistry, of which mineralogy may 

 be considered to be a branch, except that the crystallographic 

 character is implied in the name and is an integral part of the 

 definition. It might even be suggested that such a nomencla- 

 ture as is here suggested is also applicable to artificial salts, and 

 would be found especially useful with such highly complex 

 compounds as the silicotungstates, phosphomolybdates, cobalt- 

 ammine compounds, the various groups of which might be 

 named after chemists who have been especially identified with 

 their study. 



While in advance of the present inadequate nomenclature, 



