4f)S JOURNAL OF run WASHINGTON ACADEMY OF SCIENCES VOL. 11, NO. 19 



granite. Whether these are part of the same intrusive complex 

 as the granite-gneiss I am unable to say. In the northwestern corner 

 of the sheet we meet with a broad band of diorite, as well as a narrow 

 streak and what are apparently small outcrops cutting the Carolina 

 gneiss along Scott's Run. Diorite is also met with in Georgetown 

 and in the old tunnel of Rock Creek Park, analyses of which have 

 been published.^' A small exposure of "gabbro" occurs about 2 

 kilometers west of Bethesda, in the northwest part of the quadrangle. 

 Few exposures of diorite and none of gabbro seem to have been found 

 or recorded in the central granitic part of the igneous area. Thus, 

 although the structural relations are apparently somewhat complex, 

 and their study and interpretation in the field offer difficulties, as 

 Keith points out, the distribution of the various kinds of igneous 

 rock points to some magmatic differentiation in the intruded mass, 

 in the sense that it is more silicic and salic near the center and more 

 femic toward the borders. This is a mode of differentiational 

 arrangement which is commonly met with, and which is so well known 

 as not to call for the citation of corroborative examples here. 



The available chemical evidence bears out this suggestion, in 

 that the few analyses that have been made of the Washington igneous 

 rocks show serial relations between them, indicating a community 

 of origin. A very brief statement of this topic must suffice here. 



The serial and comagmatic relations are shown most succinctly 

 by the symbols denoting the positions of the various rocks in the 

 quantitative classification, which, it is assumed, are intelligible to 

 the reader. The analysis of the, fresh Broad Branch granite given 

 by Merrill is included. The series of rocks and symbols is given in 

 table 5. 



Inspection of these symbols shows clearly the gradual and pro- 

 gressive change in all the most essential chemical characters. As 

 regards the relation of salic to femic molecules, the highly silicic 

 binary granites are well within persalane, the less silicic biotite 

 granites are all on the border between persalane and dosalane, and 

 the diorites are in dosalane and salfemane. The amount of excess 

 silica, shown in the ordinal positions, remains almost constantly 

 quardofelic. There is a steady and almost uniformly regular decrease 

 in alkalies as compared with salic lime as we go down the list, and a 

 concomitant decrease in potash and increase in soda. 



The same story is told by plotting the analytical data in the usual 



12 U. S. Geol. Survey Bull. 591: 47 (E and F). 



