ROCK CLASSIFICATION 97 



does not occur with the alkali metasilicate or orthosilicate feld- 

 spathoid minerals. And these are found in rocks comparatively 

 low in silica, the orthosilicate being most abundant in the least 

 silic%'ous rocks. But quartz is found in variable proportions 

 with polysilicate feldspars, usually in rocks high in silica. So 

 that it appears to be a law that so far as the alkali-aluminous 

 silicates are concerned the highest silicate forms which is possi- 

 ble with the available silica in the magma. Thus a close rela- 

 tionship exists between the amount of silica in the magma and 

 the development of these ortho-, meta-, and polysilicates ; as 

 does also, quite naturally, the presence of free silica or quartz. 

 No such evident relationship has as yet been discovered between 

 the occurrence of the ferromagnesian minerals and the amount 

 of silica in the magma. 



Investigation also reveals the law, apparently of quite gen- 

 eral application that to a great extent the alkalis in a magma 

 control the alumina in such a manner as to compel it to enter a 

 feldspathic mineral if possible, 1 so that a reasonable estimate of 

 the amount of feldspathic constituents in a rock may be formed 

 by reckoning the alkalis and alumina together with the avail- 

 able silica as alkali-feldspathic constituents. And while it is 

 well known that the alkalis may enter abundantly into micas 

 and to a less extent into amphiboles and pyroxenes, it seems to 

 be the rule that they do not enter the two latter minerals to any 

 considerable extent unless the alumina present in the magma is 

 insufficient to form feldspathic minerals. The case of biotite is 

 different and does not conform to the rule, necessitating some 

 modification of it. 



It seems true also that in a great majority of cases aluminium 

 does not combine with calcium to form anorthite molecules 

 unless there is an excess of alumina over the alkalis in the 

 magma. Exceptions to this rule, of course, occur, but its appli- 



1 This assumption which has been used by the writer in his lectures as a working 

 hypothesis for several years has also been advocated recently by Michel-Levy in his 

 paper entitled Classification des Magmas des Roches Eruptives, Bull, de la Soc. 

 Geol. de France, 3d series, Vol. XXV, pp. 342-343, Paris, 1897. 



