530 OGILVIE 



The petrographic descriptions of the various types which were 

 not analyzed are given in each case after the discussion of the 

 similar type which was analyzed. It is however realized that 

 there may be errors in the inferred relationships and quantitative 

 development of these unanalysed types. The accurate mathe- 

 matical estimation of quantities of minerals seen under the 

 microscope is not without difficulty, and is apt to be inaccurate 

 where there are variations in coarseness of grain and where there 

 is a gneissic arrangement of minerals in bands. The problem 

 too becomes the more involved, when the majority of the min- 

 erals present are not the ones on which the classification is based. 

 With full regard for the uncertainties involved, calculations were 

 made and the unanalyzed rocks placed with their nearest analyzed 

 allies. In spite of the manifold possibilities of error, it is con- 

 fidently believed that only by such estimations will the real 

 significance of the metamorphic rocks be appreciated. For 

 example, a dark hornblende schist, consisting mainly of horn- 

 blende and plagioclase would in the old system have been reck- 

 oned with the gabbros. If any proportions were recorded at 

 all, merely the ratio between hornblende and feldspar would 

 have been noted. In the light of the new system it becomes 

 evident that that ratio has no significance at all, except as regards 

 metamorphism, and the essential and significant conception of 

 the type involves the splitting up of the hornblende into an- 

 orthite, diopside and hypersthene molecules which possibly 

 never existed as minerals in the rock, and whose ratios may show 

 more acid affinity than would be supposed on casual inspection. 

 It is not possible to do this without any analysis of the rock, but 

 given one good analysis of a type, it is possible to consider 

 slightly dissimilar types by means of microscopical inspection 

 only. 



The acid rocks were the ones which received least attention 

 from us. These were very common and in no way notable, and 

 did not seem of sufficient importance to justify any great ex- 

 penditure of time on either mapping or analysis. The types will 

 be taken up proceeding from acid to basic within each of the two 

 series. It is, however, to be borne in mind that the acid types 

 are very much the most common. 



