REPORT ON THE PETROLOGY OF OCEANIC ISLANDS. 



47 



minute green needles of the same nature. In very vitreous varieties it is not uncommon 

 to see the microliths grouping themselves in a manner resembling the arborescent 

 forms which certain pitchstones exhibit. 



Accidental constituents play a very small part in the rocks we have just described. 

 Magnetite occurs pretty frequently, titanite and apatite more rarely, and sometimes 

 sections of quartz ; but these are probably of secondary origin, as are also the grains 

 and veins of hematite and limonite. 



■ We give below an analysis by Dr. Klement of one of the trachytic rocks ; the 

 specimen came from Weather Post Hill, and its characters correspond with those 

 described above. 



I. 1*0401 grammes of the substance, dried at 110°C and fused with carbonates of 

 soda and potash, gave 0'7384 gramme of silica, 0*1543 gramme of alumina, 0*0430 

 gramme of ferric oxide, 0'0062 gramme of lime, 0*041 gramme of magnesium p3 T ro- 

 phosphate, and traces of manganese. 



II. 0*8480 gramme of substance treated with hydrofluoric acid gave 0*1871 

 gramme of sodium and potassium chlorides, and 0*1048 gramme of potassium 

 chloroplatinate. 



III. 1*0950 grammes of substance treated in a sealed tube with hydrofluoric and 

 sulphuric acids was titrated by potassium permanganate. 0*7 cubic centimetre of 

 solution (1 c.c. = 0*005405 gramme of ferrous oxide) was required to oxidise the 

 ferrous oxide. 



IV. 1*0370 grammes of substance fused with sodium-potassium carbonate, according 

 to the method of Sipocz, gave 0*0041 gramme of water. 



Percentage Composition of the Specimen. 



99-42 



The percentage of silica given by this analysis is too high for normal trachyte ; in fact 

 in unaltered specimens it only amounts to 65 per cent, which corresponds to the 

 amount of silica in sanidine. In exceptional cases certain trachytes may contain as 



