86 Penfield and Pratt — Chemical 



For a better comparison of the above results the average 

 analyses are given below, after recalculating Fe 2 3 as A1 2 3 , 

 MnO and MgO as FeO and bringing the whole to one hun- 

 dred per cent. 



St. Gothard, Switz. 



"Windham, Me. 



Lisbon, N. H. 



Burnsville, N 



Si0 2 ._. -27-70 



27-60 



27-44 



27-47 



Al,O a __ 5504 



55*75 



55-16 



55-83 



FeO... 15-07 



14-43 



15-72 



14-74 



H 2 0___ 2-19 



2-22 



1-68 



1-96 



100-00 100-00 100-00 100-00 



The ratios in these analyses are as follows : 



Si0 2 : A1 2 3 : FeO : H 2 



St. Gothard -460 : -540 : -209 : -121 = 2-12 : 2-50: -967 : -560 



Windham -460 : -546 : -200 : -122 = 2-11 : 2-50 : -915 : -557 



Lisbon -457 : -540 : -218 : -093 = 2-11 : 2-50 : 1-01 : -430 



Burnsville -458 : -547 : -205 : -109 = 2-07 : 2-50 : -934: -497 



The above ratios approximate closely to 2 : 2 - 5 : 1 :0 - 5 which 

 would give the formula HAl 5 FeSi„0 ]S , in which the aluminium 

 is partly replaced by ferric iron and the ferrous iron by mag- 

 nesium and manganese. This is, moreover, the formula sug- 

 gested by Groth, and, as previously stated, may be written as 

 a basic ortho silicate, (A10) 4 (A10H)Fe(Si0 4 ) 2 or equally well, 

 (A10) 4 Al(FeOH)(Si0 4 ) 2 . The percentage composition re- 

 quired by the formula is the following : 



SiO„ 26-32 



A1 b O, 55-92 



FeO 15-79 



HO ■_ 1-97 



100-00 



From a comparison of the ratios, or of the analyses, as re- 

 duced, with the theory, it will be observed that the silica is 

 uniformly a trifle high, amounting to something over one per 

 cent. This cannot be referred to an analytical error, as the 

 distilled water and reagents were pure, and platinum vessels 

 were used for the evaporations. It was not derived from the 

 agate mortar in which the mineral was ground, for in the 

 analysis of the mineral from Lisbon a steel mortar was used, 

 the powder being afterwards purified by treatment with hydro- 

 chloric acid. From the careful selection of nearly pure min- 

 eral to start with, and the special precautions that were taken 

 to eliminate all heavier and lighter portions by the specific 

 gravity separation, it was not expected that the staurolite 

 grains would still contain inclusions of quartz, nor were they 



