1 88 N. L. BOW EN 



examined, which fact renders it more Hkely, but by no means 

 certain, that the reaction relation obtains in magmatic systems/ 

 The service rendered by experimental investigation, so long as it 

 is confined to a limited number of components, must He in its 

 indicating where a reaction relation is to be expected. We are, 

 moreover, instructed as to what we may expect in the way of 

 indications of reaction and thus enabled to extend our inferences to 

 phases not formed under laboratory conditions. This brings us to 

 the question of the criteria of the reaction relation. 



A criterion of the reaction series, common to both the continu- 

 ous and discontinuous type, and serving to show their fundamental 

 likeness, is simply the tendency of one mineral to grow around 

 another as nucleus. In the case of the continuous series this is 

 commonly known as zoning of mix-crystals and in the discontinu- 

 ous series as the formation of reaction rims, coronas, etc. Thus we 

 have plain evidence of this kind, from a wide range of rocks, that 

 that the plagioclases constitute a continuous reaction series and 

 that pyroxene, amphibole, and mica form a discontinuous series. 

 The development of these special structures is, however, dependent 

 on particular conditions of consolidation and the lack of such struc- 

 ture in an individual case should not be regarded as indicating a 

 lack of the reaction relation in that case. It must be, in part, 

 from a general survey of mineral relations in igneous rocks that the 

 reaction series are deduced. 



Fortunately the continuous reaction series are easily picked 

 out, for the mere existence of sohd solution or variability of composi- 

 tion in a crystal phase is sufficient to estabhsh that phase as a 

 continuous reaction series. Their number is legion, all the impor- 

 tant igneous rock minerals with the single exception of quartz 

 being members of soHd solution series. The detection of the 

 discontinuous reaction series is not always so easy, and the element 

 of judgment enters to some extent. 



As an example of the information to be obtained on this point 

 from a general survey of an igneous sequence let us reproduce a 



^ Vogt believes there is no reaction relation but a eutectic relation at higher 

 pressures between olivine and pyroxene. Jour. GeoL, Vol. XXIX (1921), p. 528. 



