532 Henry C. RicJia/rds : 



in this way was seen to be largely the ojDaque mineral, hyper- 

 sthene containing large inclusions of the former mineral, leu- 

 coxene, and quartz and felspar containing magnetic inclusions. 

 The voltage was then increased to 10 (3 amps.), and all the 

 material was picked up with the exception of a few quartz and 

 felspar grains, and several small flakes of biotite. 



By means of Sonstadt's solution of Spec. Gr. 3, used with a 

 Sollas Separator, a practically pure hypersthene product was 

 obtained, the quartz and felspar being removed with the excep- 

 tion of a few composite grains, which were easily picked out 

 with a wet brush. 



An examination of the hypersthene under the microscope 

 detected opaque material in nearly every grain, so with a view 

 to getting rid of as much as possible of this it was ground 

 finer in an agate mortar, and subjected to 4 volts, with pole- 

 pieces 4 mm. apart. A good deal of magnetic opaque material 

 was removed thus. After picking over with a wet brush under 

 the microscope, this mineral was ready for chemical analysis. 



Ilmenite. — An attempt was made to separate this from the 

 rock powder by the magnet, but as it is present to the extent of 

 only one per cent., is largely included in other minerals, and is 

 in such small grains it would have been a lengthy process to 

 separate enough for a chemical analysis. The separation has 

 been done by nature, however, and some of the material as it 

 occurs in the stream sands in the dacite area was obtained. 

 It is remarkably fresh, and although the sample was not 

 obtained from exactly the same locality as the rock sample, it 

 is believed that its composition is similar to that contained in 

 the material used for separation and analysis. 



AxALY8Bs OF Rock and Minerals. 



Besides analyses of the three minerals of variable composition 

 in the rock, one was made of the rock itself. All the analyses 

 were made in duplicate with the exce})tion of that of the 

 ilmenite. Agreement between the duplicates was satisfactory, 

 and the analyses appended below are those of one of the dupli- 

 cates and not the mean of two deteriniiiations. 



