128 CARNEGIE INSTITUTION OF WASHINGTON. 



The principal minerals of igneous rocks can be arranged as reaction series, 

 such that any member of the series is produced from the preceding member 

 by reaction with the hquid magma in the manner above noted. The plagio- 

 clase feldspars, together with potash feldspar, constitute one series, and the 

 olivines, pyroxenes, amphiboles, micas, and quartz constitute another series. 

 A basic magma may give rise simply to calcic plagioclase and pyroxene, if 

 quickly cooled, whereas the same magma, slowly cooled, may yield a long 

 chain of products embracing all the minerals enumerated above. In other 

 words, diorite, granodiorite, granite, and other rock types may form by differ- 

 entiation from a basic magma in virtue of the existence of this unidirectional 

 reaction relation among the mineral products. 



Another possible cause of variation of igneous masses is that due to the 

 incorporation in magmas of foreign rocks. The importance of this action 

 has long been a subject of discussion, some assigning to it a dominant role 

 and others denying its very existence. The heat required for solution has 

 been considered too great to be supphed by the liquid magma; but a considera- 

 tion, from the viewpoint of the reaction principle, shows that magmas can 

 incorporate large quantities of foreign matter by reacting with it. If the 

 magmas and the various types of foreign material are considered with respect 

 to their position in the reaction series, it is found that a magma can not dis- 

 solve foreign inclusions belonging at an earlier stage of the reaction series 

 than the phases with which it is saturated, but can and will react with such 

 inclusions to convert them into the phases with which it is saturated. On the 

 other hand, foreign inclusions belonging to a later stage in the reaction series 

 than the phases with which the magma is saturated can be dissolved by the 

 magma by a sort of reactive solution, a certain amount of the phase or phases 

 with which the liquid is saturated being simultaneously precipitated. This 

 deduction from the results of experimental work is in accord with the obser- 

 vations of the field geologist and at the same time permits an evaluation of 

 the relative importance of solution as compared with spontaneous differentia- 

 tion. We have found that the material, in order to pass into solution in a 

 given magma, must belong at a later stage in the reaction series than the 

 dissolving liquid; in other words, must be of a composition toward which the 

 hquid changes by spontaneous crystallization-differentiation. Plainly, then, the 

 crystallization-differentiation is the dominant factor in bringing about varia- 

 tions in igneous rocks, for through it variations may come about spontane- 

 ously without the intervention of any foreign material, and at the same time 

 it completely controls variations due to the solution of foreign matter. (See 

 reviews of papers 451 and 459 following.) 



VOLCANO STUDIES. 



In recent years the attempt has been made not only to study and 

 reproduce in the laboratory the rock-forming processes through which the 

 great body of igneous rocks comprising the visible crust of the earth came 

 into being, but also to study the natural processes in the regions where 

 they are still going on, namely, in the volcanoes. Two researches in this 

 latter field have been completed recently, which add much to our knowledge 

 of these complicated natural phenomena. The first of these (Allen and Zies) 

 records the results of the field and laboratory work carried out by the chemists 



