ANALCITE CROUP OF IGNEOUS ROCKS 689 



fcrro-magnesian minerals, are very often vitiated by a failure to 

 "jM-operly separate alumina from magnesia. Many analyses 

 which would otherwise be good are spoiled by this error. The 

 writer's attention has been strongly called to this point on 

 examining, in connection with this article, various analyses of 

 basic rocks. The types are described as consisting chiefly or 

 largely of pyroxene with or without olivine and with the fels- 

 pathoid components in perhaps subordinate quantity, yet the 

 analyses may show very high alumina with very little magnesia. 

 An example might be quoted of an analysis of this class pub- 

 lished within the last few years : 



SiOo FeoOg.FeO AUG., CaO MgO KgO NagO P3O5 Ign. 



47.83 4.57 30-28 6.72 4.32 trace 1.30 2.19 2.05 = 99.26 



In the light of our present knowledge this may almost be said 

 to be an impossible composition for an igneous rock, and it is 

 ver}' clear that a large part of the magnesia must have been 

 thrown down with the alumina. The tendency of alumina to 

 drag down magnesia on precipitation with ammonia is very great 

 and only to be prevented by the presence of a liberal quantity 

 of ammonium salts, and the precipitation should always be 

 repeated, especially where both oxides are present in considera- 

 ble quantities, under which circumstances even a third precipita- 

 tion may be necessary. There are man_y excellent chemists who, 

 from a lack of experience in silicate analysis, fail properly to 

 appreciate this point, and it has been perhaps the most common 

 error in rock analyses. 



Returning to our analcite rocks, it is of interest to observe in 

 this connection that leucite and analcite have the same crystal 

 form and the same structural formula except the addition of the 

 molecule of water in the analcite ; it is not strange that this dif- 

 ference between them should exist when one reflects how com- 

 monly soda salts contain water of crystallization and how much 

 more rarely the potash compounds assume it. 



The result of this is then to show why leucitic rocks are 

 commonly effusive ones while the analcitic rocks would be more 



