paet 2.] Minemlogiwl notes on the Gneiss of South Mirzdpur. 43 
influences often stands up above the general surface in a low jagged ridge. The mineral 
has a greyish-white color and bright pearly lustre, and the approximate parallelism of the 
principal cleavage faces gives the rock a somewhat fissile structure. Tremolite is very 
abundant in the limestone of the Bichi nadi ;* but the above is the first instance I have met 
with of the occurrence of wollastonite. 
My work brought me again this year to the corundum quarries between Pipra and 
Kadopani, which I examined closely. The thickness of the bed 
cannot be determined with any degree of accuracy from the 
amount of debris lying about; but as a rough guess I thought I was more below than above 
the mark in estimating it at 30 yards at the quarries where it appears to be thickest, and it 
may be considerably more. The ground is too obscure for one to say with certainty that 
the above includes no subordinate layers of other rocks, but I observed no indication of such, 
and for anything I saw for the contrary, the bed may be a solid mass of corundum. It runs 
about east-north-east, west-south-west, the bedding being vertical or at a high angle. The 
section previously givenf is only true for the spot where it was made, for some of the 
associated beds die out rapidly. At the west end of the long low hillock which marks the 
position of the mineral, porphyritic gneiss and white quartz-schist are seen within 10 yards 
of each other with corundum in the space between. Prom this to the Behr, some 300 yards, 
is obscured by clay, and no trace of the corundum is to be found in the river. East of the 
quarries again, the bed can only be followed for a short distance, the entire length visible 
from west to east being, as laid down on the map, about half a mile. 
The corundum, where weathered, much resembles fine-grained hornblende-rock in a similar 
state, and might be easily overlooked. Its intense hardness is well shown by the way in 
which hammers which may have stood years of ordinary geological work are in a few 
minutes split and pounded out of shape on it. It seems strange that it should not form a 
more prominent physical feature. Pluvial mechanical erosion would apparently act very 
slowly indeed on it, in comparison with the softer rocks on cither side, and the absence of 
secondary minerals in considerable quantity does not point to important chemical alteration. 
Probably its weak point is the irregular jointing Ity which it is intersected. 
The quarrymen are, I was told, paid at the rate of one rupee per 31J kacha mands 
raised, hut the mineral is only worked now and then when a quantity is ordered by tbe 
mabajans who deal in it. Before commencing operations a kid is sacrificed to Devi, to insure 
good fortune, and protection from accident; fires are lighted against the large masses into 
which tbe corundum is divided by jointing, and when they have been rendered somewhat 
more brittle by this means, they are gradually smashed by heaving other pieces at them. 
Considering the thickness and length of the bed, it is clear that the supply may he consi¬ 
dered inexhaustible. 
I have described the minerals which arc associated with the corundum in my previous 
note. The only additional species I have observed this year is kyanite, which occurs in a 
radiating aggregate of a reddish color. It is a mineral, which as a simple silicate of aluminaj 
is a natural associate for corundum, and has been similarly met with elsewbere. There are 
also small bladed crystals with a bright pearly lustre, much like diaspora, but their small size 
and the impossibility of detaching them makes their examination difficult. They may be 
kyanite. 
Beds of magnetite interlaminated with granular silicious layers are met with not un- 
frequently, more noticeably in the crystalline inliers near Koelkat, 
Magnetite. a j g0 near (jairar and south of Kadopani. None of these, however, 
are as rich in iron as the magnetic band at Korche in Mirzapur.J Magnetic sand very 
* Vol. v, P. 20. 
t Vol. v, P. 20. 
t Vol. V, P. 22. 
