July, 1889. 
A GRANITE CONTAINING LITHIA. 
155 
alkaline chlorides, which I proceeded to weigh. It was 
immediately obvious, however, that the mass was gaining 
weight somewhat rapidly, showing the presence of some 
deliquescent salt. An attempted purification left matters as they 
were before, and made the absence of calcium or magnesium 
chloride in any appreciable quantity quite certain. The point 
of a fine platinum wire was then dipped in the somewhat 
deliquesced salt, and when this was held in the flame of a 
Bunsen burner the momentary vivid red flame showed the 
presence of lithium chloride, which was at once abundantly 
confirmed, on repeating the experiment, by observing the flame 
through a pocket spectroscope. On separating the lithium 
chloride by Gooch’s method of using boiling amyl alcohol, the 
chlorides of potassium and sodium were quite normal in 
behaviour in the air, being only quite slightly hygroscopic 
under the then existing conditions of the atmosphere as to 
moisture and temperature. In this way the lithia in the 
rock was found to amount to as much as 0-7 per cent. 
Further SDecimens were obtained through the kindness of 
Messrs. Watts, Blake, Beanie and Co., the owners of the 
Meldon Quarry, near Okehampton, from which the rock was 
procured, and I have been able to make a more thorough 
examination of the chemical and microscopical characters of 
this peculiar granite. 
The full analysis gives the following results : — 
Silica ... ... ... ... ... 72-4 
Lime ... ... ... ... ... 0-9 
Manganese Oxide ... ... ... 0’2 
Magnesia ... ... ... ... 0-2 
Alumina ... ... ... ... 16-0 
Ferric Oxide ... ... ... ... 1*0 
Soda ... ... ... ... ... 4-6 
Potash ... ... ... ... ... 3-0 
Lithia ... ... ... ... ... 0-7 
Loss on ignition ... ... ... 0-9 
99-9 
In this analysis the amount of oxide of manganese is 
higher than is usual in bulk analyses of rocks. Attention was 
at once called to it by the bright green colour of the melt 
when decomposing by fusion with alkaline carbonates. It 
occurs apparently as the colouring substance of the tourmaline 
and possibly in part in the mica, 
