Pirsson and Washington — Geology of New Hampshire. 513 



one-twentieth, and that of the complementary dikes the same, 

 those of liparose and of camptonose being in the proportion of 3 

 to 2. The analysis of the akerose (VI) may be neglected. 

 Taking then 80 parts of IX, 10 of III, 5 of VII, 3 of I and 

 2 of VIII we obtain the following results : 



Si0 2 60-94 Qz 4'50 "| 



Al 2 O s .___ 17-57 Or ._.. 32-80 I * 



Fe 2 .0 3 .._. 2-13 Ab.__. 38-25 f 8b b7 



FeO 3-91 An 11-12 J 



MgO 1-40 Di 0-92 ] 



CaO 2-65 Hv ... 6*90 | 



Na 2 4-52 Mt 3-02 }> 13*31 



K 2 5-53 II 2-13 | 



TiO a , 1-15 Ap 0-34 J 



P 2 O fi ..__ 0-20 



99-98 



100-00 



These figures are practically identical with those furnished 

 by the main syenite, though SiG 2 and CaO are a trine higher 

 and K 2 a little lower, and the classificatory position of the 

 rock is the same, pulaskose-monzonose (I— II. 5.2.3.). That the 

 divergence in composition of the average magma from that of 

 the main syenite should be small follows from the fact that the 

 combined weight of the modifying factors, that is the liparose, 

 lassenose and camptonose, is only one-fourth that of the main 

 mass. But the very close agreement shows that the several 

 constituents in the more salic and the more femic magmas bal- 

 ance each other to a very great extent, and indicates that they 

 are in the nature of complementary differentiates. 



It is to be noted, however, that the main pulaskose-monzo- 

 nose is higher in alumina, soda and potash than any of the 

 smaller rock bodies,* while the figures for the other constitu- 

 ents lie between the extremes. 



It is perhaps justifiable to infer from this .the existence in the 

 complex of a differentiate relatively higher in these constituents. 

 Such a rock would be composed almost wholly of soda-ortho- 

 clase about Oi^Abj, with relatively insignificant amounts of 

 quartz, plagioclase and al ferric or femic minerals. We may 

 possibly look for this in the syenitic cement of the breccia de- 

 scribed above, which has not been examined chemically, but 

 which we have shown to be markedly deficient in plagioclase 

 and colored minerals, though the amount of quartz is consider- 

 able. 



In conclusion a few words may be devoted to a comparison 



of the Belknap rocks with those of other igneous areas in New 



England and Canada. The best analogues are met with in the 



larger and more complex igneous district of Essex County, 



* Except as regards soda in the akerose (VI). 



