74 



DE. A. HOLMES O^ THE PRE-CAMBRIA1* [vol. lxxiv, 



rock a striking and handsome appearance. In places, the rutile 

 is bordered by minute grains of quartz ; but the total proportion 

 of the latter mineral in the specimen collected is not more than 

 ■6 per cent. (PL X, fig. 4). 



The following table gives the mineral composition of the rock 

 (by weight) as determined by the Rosiwal method, and also hj 

 heavy-liquid separation of the minerals : — 





*als. 



Percentages of Minerals 

 ■ by toeight. 



Specific 

 Gravities. 





Separation 

 Method. 



Rosiival 

 Method. 



Muscovite. 



! 

 Rutile .... 





70-9 

 23-3 



5'8 



68 



27 



5 



2-83 





5*12 



Quartz 



2-66 







Specific gravity , , , , 



Calculated. 

 3-11 



Calculated. 

 321 



Determined. 

 317 



An analysis of the separated muscovite was made by Dr. H. F. 



Harwood, to whom I express my sincere thanks. The results are 



as follows : — 



Percentages. Molecular Proportions. 



SiO., 47-15 -786 



Al,6 3 33-57 -3291 qq>7 



Fe" 3 1-35 -008J 66i 



Feb 0-32 -004 i 



MgO 0-45 -001 1-007 



CaO ' 0-14 -002 J 



Na.,0 1-04 -017] 



K.,6 9-27 -099 1-405 



H~0 above 110° C. . . . 5'20 -289 J 

 H."0 below 110° C ... 0'48 



Ti0 2 1-31 -016 



F 0-05 -003 



MnO none — 



100-33 

 Less O for F -02 



Total 100-31 



The molecular proportions indicate that the ordinary formula 

 for muscovite does not adequately express the composition of 

 this variety. The chief discrepancies are due to an excess of water 

 and of silica. The latter may be explained by assuming that 

 2*5 per cent, of the mineral contains silica as trisilicate instead 

 of the normal orthosilicate ; but this assumption is probably 

 unnecessary, as it seems more likely that a small quantity of 

 quartz remained with the muscovite after separation. Microscopic 



