THE EVOLUTION OF PETROLoaiCAL IDEAS. 297 



iuul potiissiuin or of ii'on and ])otassiuiii (l)iotit<>). calciuin silicate 

 (aiiorthite). silicates of sodium and calcium (plaii'ioclasc). sodium sili- 

 cates (uci)heliuc and albitc). and lastly, potassium silicates (orthoclasc) 

 in conjunction with ([uaitz. 



In his important work on slaos. Professor Vo^t has (dearly estab- 

 lished the intluenceof the ridative ])r()portions of the bases to each other 

 and to silica in determinintj;- the nature of the comi)ounds which sep- 

 arate out. Thus in sla<4S in which the ratio of bases to silica corre- 

 sponds approximately to that found in ])isilicates the ratio of C^a():MgO 

 determines the foi-mation of such minerals as enstatite, auoit(\ and 

 wollastonite. When th(^ ratio of MgO-fFeO : C'aO is o-reat(>r than 

 2.44 : 1, enstatite forms: when the same ]-atio is less than 1.4 : 1, augite 

 separates out. and continues to do so until this ratio ])ecomes less than 

 .;^5 : 1; with a still further diminution in the ratio of mauiu^sia to lime, 

 wollastonite is formed. 



In slags having approximat(dy the composition of nionosilicates the 

 ratio of MgO+Mn04-FeO : CaO determines {\\o formation of olivine 

 or melilite. When the above ratio is greater than J : 1.1 (in slags with 

 al)out 2*) per cent of ahunina), olivine is formed; but when it is less 

 than 1 : 1.2.5, melilite is ])roduced. 



The general concdusion arriv(Hl at as a result of the work of \ ogt, 

 Lagorio, and others is that mass action and the affinities of the bases 

 to ea(di other and to silica are the two factors of primary importance 

 in determining the mohn-ular grouping, so long as the pressure 

 remains constant. The action of alumina may 1)e espiM-ially referred 

 to as illustrating tlu^ intiuencc^ of the mutual affinities of th(^ so-called 

 bases. In the sorting of partners in accordance with the law of mass 

 action this substance, when present in sufficient ([uantity, practically 

 takes the whole of the alkalies and as much of the lime as is necessary 

 to make felspathoid molecules. So marked is this action that 

 M. Michel Levy and M. Osann. in calculating the results of 'analyses, 

 cond)ine the whole of the alumina with the alkalies, when the latter 

 are present in sufficient <|uantity, and associate any excess of alumina 

 with lime in the form of fidspathoid molecules. It is only in those 

 rocks that contain an abnormal percentage of alkalies that minerals 

 like ivgirine and riel)eckite occur. 



This controlling influence of alumina, which has also been empha- 

 sized ]>y Professor Iddings, has the most far-reaching eti'ects in determ- 

 ining petrographical s|)ecies. It is as if there were a kind of repulsion 

 between the ferromagnesian and alumino-alkaline constitutents. Dark 

 rocks rich in the former, and light ro(d<:s ri(di in tla^ latter, represent 

 the extreme forms of many internu'diate types; and Professor 

 Brogger has recently proposed that this should receive expression by 



