R. H. Rastall — The Genesis of Tungsten Ores. 369 



essentially a process of differentiation, which may be described as 

 fractional crystallization. Processes of this kind are generally 

 described as pneumatolytic, but there does not seem to be any need 

 for the use of the word in this connexion, since as commonly under- 

 stood it seems to imply something unusual and out of the common 

 order of events. This kind of differentiation is perfectly normal; its 

 final results depend mainly on the extent to which highly volatile 

 compounds and water were present in the original magma. The 

 more of these are present the lower will be the freezing-point of 

 the final product. The effect of alkaline tuugstates in lowering the 

 freezing-point of acid silicate melts, and thus enabling quartz and 

 acid felspars to crystallize, has long been known and used in petro- 

 logical and raiaeralogical research. It appears that during the 

 freezing of the residual magma cassiterite is formed fiist, then conies 

 wolframite, while last of all quartz and fluorite are formed, resulting 

 in veins of quartz and fluorite without metallic minerals, or even of 

 fluorite alone. 



The evidence afforded by the wolframite lodes without tinstone is 

 entirely in conformity with the foregoing considerations. There is 

 a continuous gradation between the two types by diminution of 

 tinstone; the series passes through the wolfram-quartz lodes, in 

 which molybdenum and arsenic tend to disappear, whereas other 

 sulphides tend to increase. As the final member of the series we 

 have the association of wolfram with siliceous gold ores, as in 

 Colorado, and of scheelite with gold ores in California. This 

 relationship of tungsten to siliceous gold ores is of much interest in 

 connexion with the classification of ore-deposits proposed by Spurr.^ 

 This autnor regards all ore-deposits as due to differentiation of 

 igneous magmas, the variations depending on the amount of water 

 and rarer elements originally present in the magma. According to 

 this scheme, the bulk of the tungsten deposits belong to the first or 

 pegmatitic type, whicb is characterized by tin, tungsten, molybdenum, 

 with tourmaline and topaz as gangue minerals. Some of the tungsten 

 ores, however, both wolfram and scheelite, extend into the second 

 group, " the free-gold pyrite zone with quartz." 



Spurr correlates these differences mainly with depth, since he 

 regards all ores as deposited by ascending solutions. Although this 

 idea is highly probable, it cannot yet be regarded as demonstrated 

 that the tin-tungsten lodes were formed at a greater depth from the 

 surface than the gold lodes or the cupriferous pyrite deposits. Tliis 

 point, however, is not of very great importance. It is a fact, 

 however, that the occurrence of tungsten ores Avhen studied in 

 detail lends strong support to the idea of a definite sequence of 

 differentiation of the rarer metallic contents of the granitic magmas. 



The tungsten ores are definitely oxidic in character, and in many 

 ways afford a strong contrast to the behaviour of the sulphidic ores. 

 They appear to belong characteristically to acid magmas, whereas 

 most of the great masses of sulphides are connected with basic 

 intrusions. There is also a strong contrast between the two classes 



^ Spurr, i?cono7Mic Geology, vol. ii, p. 781, 1907. 

 DECADE VI. — VOL. V. — NO. VIII. 24 



