194 N. L. BOW EN 



members should crystallize first (witness the plagioclases), and 

 there is a no less distinct tendency in this direction in the discontinu- 

 ous reaction series. 



This explanation and vindication of the doctrine of the French 

 school is not, however, to be regarded as a proposal that we should 

 reverse the attitude arising from eutectic considerations and accept 

 unreservedly the statement that minerals separate in the reverse 

 order of their fusibilities. There are too many obvious exceptions 

 to such a rule. However, these exceptions are themselves such as 

 might be anticipated in a system dominated by reaction series. 

 We have seen on an earlier page that certain minerals that may be 

 called released minerals may appear in such a system. It is appar- 

 ently minerals that belong in this category that constitute the more 

 notable exceptions. The principal of these is quartz. It is a 

 released mineral of the reaction, oHvine-pyroxene and no doubt 

 has a similar relation in the reaction, potash feldspar-mica, a sort 

 of hydrolysis requiring an adequate concentration of water. It 

 is a mineral of high melting-point but separates very late — a fact 

 that is to be connected with its character as a released mineral. 

 In other words, it may be regarded as not actually present in the 

 liquid at an early stage, but as released later in consequence of 

 the reactions mentioned. 



Some minerals may separate early when intrinsically present 

 and may also appear as released minerals at later stages. Magne- 

 tite is a particularly good example. It may separate very early 

 from a gabbro and it may be formed very late as a result of, say, 

 the pyroxene-hornblende reaction and in particular of the mica- 

 feldspar reaction. 



PROGRESS TOWARD AN UNDERSTANDING OF REACTION 

 SERIES IN MA.GMAS 



At this place it may be well to point out where we are in par- 

 ticular need of increased knowledge in order to understand the 

 exact nature of the reactions occurring. It will be noted that each 

 member of the discontinuous reaction . series, olivines-pyroxenes- 

 amphiboles-biotites, is itself a continuous reaction series. Of the 

 oHvines and the pyroxenes we know a fair amount, but practically 



