Mr. Greville on the 
4*4 
wire, till, by sawing the crystal, the stretched wire is worn, 
and requires to be renewed from the coil. The twisted wire 
answers the purpose of a saw, and retains the powder of Co- 
rundum and water, which are used in this operation. Dr. Lind 
had before brought specimens similar to the above, from China. 
From Sir Joseph Banks I obtained Dr. Lind’s specimens, 
and some in powder, which Mr. Duncan, supercargo in Chi- 
na, had sent him, with the Chinese name, Pou-sa. The ma- 
trix, being mixed with a red and white sparry substance and 
mica, is generally called red granite ; but it appears to me of 
the same nature as the matrix of Corundum from India. The 
white is more fibrous, and like cyanite ; the red part of it is 
compact and opaque ; other parts appear to foliate, and pure 
mica is in considerable patches, and generally adheres to the 
crystals. This Corundum is of a darker brown, and more 
irregular on the surface than the Corundum of the Coast, and 
often mixed with black iron ore,* attractable by the magnet. 
It is described as the third modification of the Corundum crys- 
tal, in the analytical description which follows. The chatoyant 
or play of light, on these dark crystals, is very remarkable ; 
some are of a bright copper colour ; others exhibit the accident 
of reflection of light, which, in a polished state, gives varieties 
to the cat’s eye, star-stone, sun-stone, &c. ; which, as yet, are 
classed from such accident, without strict attention to their 
nature, which is various, and in general has not been ascer- 
tained. 
* A small group, consisting of three or four octoedral crystals, presents the least 
common variety of this kind of iron ore; the edges of the octoedra being replaced by 
planes which almost cover the triangular planes. Rome de I’lsle. Cristallog. 
Vol. IV. Plate 4. fig. 69. 
