r74 NEW YORK STATE MUSEUM 



ing cause can not be pyrite, since there is s'o little of it present 

 tbat the sulfur percentage does not in general reach .01;^. It is 

 difficult to see how carbonaceous matter other than graphite can 

 be present, and in an igneous rock any considerable amount of 

 graphite would be surpri.sing. The cause of the yitiation is as 

 yet undetected. 



The only analysis so far made of the augite syenite which 

 occurs cutting the anorthosite, analysis 8, indicates that to be 

 somewhat more basic than the usual rock, and this seems to be 

 true of all such syenite, so far as can be judged by the thin sec- 

 tions. Garnet is much more abundant than in the usual syenite, 

 and bronzite is lacking. The anailysis indicates a rock composed 

 of 24.25;^ orthoclase, 44.55,^ albite, 5M anorthite, l.lfo magnetite, 

 5.8^ garnet, 14.5^ augite and 3^ quartz. If the ferrous iron be 

 too high, and this is possible, though the discrepancy can not be 

 great in this case, the magnetite and anorthite percentages would 

 be slightly increased and those of augite and quartz diminished. 

 Except for a slight amount of acid plagioclase, the feldspar is all 

 of the intergrowth types, and cleavage fragments from the 

 crushed rock show the optical characters of anorthoclase, viz 

 a + 9° extinction on M, with an acute bisectrix in the center of 

 the field. 



The rock used for the next analysis, 9, is from near the anortho- 

 site boundary. Ferro-magnesian silicates are more prominent 

 than usual, considerable hornblende, augite and garnet being 

 present and some bronzite, all together constituting some 30^ of 

 the rock. The lower alkalis show the diminished feldspar per- 

 centage, but a calculation is rendered impossible by failure of 

 the ferrous iron determination. On the basis of 5^ of magnetite, 

 as indicated by the separation previously mentioned, the calcula- 

 tion gives a silica residue amounting to 13;^ of free quartz, which 

 is much too high, there being but little present. The remaining 

 analyses require little comment aside from no. 10. No. 11 is 

 regarded as giving the closest approximation to the mean com- 

 position of the rock, and is from the Loon lake type locality. 



