510 Pirsson and Washington — Geology of New Hampshire. 



fragments of the enclosing rocks, and that these appear at 

 times, when very small, to be more or less absorbed, bat gen- 

 erally even when minute they retain all their original sharpness 

 of outline and angular form though completely metamorphosed. 

 Moreover there is no evidence that the enclosing rocks are in 

 general more siliceous than the syenite and in some cases they 

 are distinctly less so. Where small fragments appear to have 

 gone into solution the rock is not more but less siliceous, the 

 small spot locally having more the character of a monzonase as 

 already described in the inclusions in lassenose above Point 

 Belknap. 



Again, it should be considered that if such an action took 

 place one would expect it to be general and everywhere present, 

 while as previously shown the syenite on the southwest foot of 

 Locke's Hill comes itself directly in contact with the schists. 

 The consideration of the chemical character of the two rocks 

 furnishes, however, the strongest argument against this view. 

 In the syenite the percentages of the alkalies are Na 3 = 4*89, 

 K 2 = 5-90, while in the granite border they are Na 3 = 4*06, 

 K 2 = 2*06. It is clear from this that to reduce the alkalies of 

 the former to those of the latter an enormous amount of some 

 rock very rich in soda must have gone into solution. It could 

 be easily demonstrated what the composition of such a rock 

 would have to be by calculating from the two analyses, but this 

 is unnecessary as the enclosing schists are clearly not of such a 

 composition as could produce this change. 



This explanation of the granite border being untenable, we 

 are forced to fall back on causes endogenous to the fluid mass 

 itself and to conclude that it is a case of magmatic differentia- 

 tion. Cases where a massif of igneous rock has a differentiated 

 border facies are becoming more numerous and several have 

 been described where the border is more salic than the main 

 portion. 



Chemical Characters of the Belknap Magmas. 



The data for studying the chemical characters of the mag- 

 mas which furnished the rocks described in the preceding pages 

 are found in the analyses presented in the annexed table. 



Of these II and Y have been calculated from the measured 

 modes and, while not entitled to quite the weight of the others, 

 may yet be regarded as representing fairly well the chemical 

 compositions of the rocks in question. 



The chemical range of the rocks is wide, especially as regards 

 SiO„ A1,0 3 , FeO and CaO, with Fe 3 3 , MgO, K u O, TiO„ 

 and P a 6 showing less variation, while the amount of Na 2 is 

 remarkably uniform. In general the variations of the several 



