Pirsson and Washington — Geology of New Hampshire. 443 



suffer, with the consequent masking in large degree of their 

 true character. In the quantitative classification this is 

 obviated to a great extent, for such rocks fall in peralkalic 

 rangs, which fact at once reveals their nature. 



At first thought all this may seem merely a rather forceful 

 way of putting well known facts, but the application of it 

 serves to bring out some points in a disputed field. In the 

 endeavor to use the genetic relations of igneous magmas for 

 purposes of classification, it is often implied, if not expressly 

 stated, v that we should not expect to find strictly alkalic rocks 

 associated with those of other series, with the granite-diorite 

 or diorite-gabbro families of Rosenbusch, alkalicalcic or docalcic 

 magmas in the new classification. If the idea of alkalic rocks 

 is however broad enough to include those greatly diluted with 

 silica but in which the other oxides are in the proper propor- 

 tions, as it rightfully should be, then numerous examples 

 which contradict the general supposition mentioned immedi- 

 ately come to mind. Thus in the Yellowstone Park region, 

 as shown by Iddings," the main lavas extruded have been 

 " rhyolites, andesites and plagioclase basalts," a really typical 

 granito-diorite-gabbro series. An examination of the analyses 

 of these lavasf shows that many of the rhyolites, such as those 

 of Obsidian Cliff, are of highly alkalic nature, belonging in 

 peralkalic rangs which with less dilution by silica would have 

 formed trachytes or phonolites. 



Liparase. — The type just described was selected as illustra- 

 tive of the granitic dikes in the area. There is some variation 

 among them in texture, but so far as one can tell by study of 

 the sections they are certainly peralkalic persalanes and appar- 

 ently are quardofelic. They fall therefore in liparase and 

 judging by analogy are probably sodipotassic and belong in 

 liparose. This last, however, could not be definitely determined 

 without making chemical analyses, a work whose results would 

 not justify the time and labor. A few words may be added 

 concerning the textural variations of the different occurrences. 



Gunstock Dike. — The summit of Mt. Gunstock is cut by a dike 

 three feet wide with the following megascopic characters : 

 phanerocrystalline ; inequiclimensional ; usually porphyri tic with 

 medium phenocrysts ; pale flesh color ; phenocrysts of ortho- 

 clase 2-5 mm , ill defined, cuboidal to subtabular, scattered ; 

 groundmass fine-grained l mm or less, dotted with minute gray- 

 black specks. Under the microscope the minerals are the 

 same as those mentioned in the type described, including 

 allanite ; the groundmass is similar but is thickly crowded with 



*Quar. Jour. Geol. Soc, vol. liii, 1896, pp. 606-617. 

 fBull. U. S. Geol. Surv., 228, p. 120, 1904. 



