Reviews — Dr. Washington — Igneous Rocks, Arkansas. 179 



but he offers some suggestive considerations, and presents them with 

 considerable originality of method. In setting forth the variations 

 in chemical composition in the series he makes use of the graphical 

 method, which, though it cannot of course lead to any result not 

 contained in the columns of figures, will often bring out points 

 which might otherwise escape notice. The abscissEe in his diagrams 

 represent the radial distance of the several rocks from the centre 

 or • epicentre ' of the complex, and it is considered that this expresses 

 best the essential relations. The a priori objection to taking the 

 silica for the abscissEe seems, however, to be without much force. 

 As in many other cases, the order of increasing silica-percentage 

 is manifestly the natural order in which to consider the several 

 rocks (in this case corresponding with their disposition in space) ; 

 if we arrange them in order of their content (say) of potash or 

 soda, the whole is thrown into confusion. This, even apart from 

 other considerations, places silica on a different footing from the 

 other oxides. The choice of ' independent variable ' cannot affect 

 any very essential relations of the series. If, for instance, the 

 silica and other oxides were all linear functions of some arbitrary 

 variable, the other oxides would necessarily be linear functions 

 of the silica. Washington's method, however, adapts itself well 

 to the special case of differentiation in place with symmetrical 

 arrangement. 



It is to differentiation in 'place of a presumably homogeneous 

 magma that the author ascribes this complex, notwithstanding the 

 fact that the several rock-types are sharply separated from each 

 other with little or no indication of transition. Taking into account 

 the volume of each rock in the complex, he calculates the composition 

 of the hypothetical initial magma, and finds that it does not agree 

 at all closely with any member of the series. This result connects 

 itself with the fact that the lines in the diagram of variation are not 

 straight, or, in other words, the series is not a ' linear ' one. Leaving 

 out of account one rock, the covite, the lines are all smooth curves. 

 The silica-line is inflected, being nearly horizontal in the middle 

 part of the diagram, but descending gently towards the basic end 

 and ascending rather steeply towards the acid end. The curves for 

 alumina and the alkalies ascend gently and steadily, being only 

 slightly curved. The curves for iron, magnesia, and lime, on the 

 other hand, descend, and are decidedly convex upward. Covite 

 does not fit into this comparatively simple scheme of variation, 

 and the author regards it as probably originating by a secondary 

 differentitaion, most likely from the foyaite magma. He attempts, 

 indeed, to discover the complementary (more leucocratic) product, 

 which may be represented by a variety of the foyaite. This latter 

 rock, like the jacupirangite, has not been found in the place required 

 by theory, and these irregularities, together with the generally sharp 

 delimitation of the several concentric zones and certain points in the 

 mineralogical constitution of the rocks, suggest to a critic that the 

 differentiation may have been effected, at least in the main, prior 

 to intrusion. In whichever way this minor question may be 



