(127 ) 
qualification would only be given to those who 
annually sowed an amunam or ten parras which 
miglit be verified by the comparison of the roll 
of the labourers with the Hoofd Thombo. 
It is proper to remark here that the permission 
to be exempted from tlie Corvee by becoming a 
labourer is at the option of the Government and to 
be granted as an encrdachment to agriculture, 
otherwise the obligation to service remains, as it is 
immutably the duty of the individual, this en- 
couragement must however be extended for many 
years till the time when the population by being 
doubled will have cleared the land in the most 
favourable places. 
Attention must be given that the districts on 
the coasts of Ceylon whi^h most abound in &;rain 
are not favourably situated for the supply of 
Colombo, which lies under the wind in the very 
season in vvliich the small craft can navigate. 
This consideration should induce Government to 
employ its first efforts in the encouragement of 
agriculture in the district dependent on Negombo, 
Chilaw, Putlam, and above all the district of 
Mantotie which is now merely waste, though it 
furnished three or four centuries ago an immense 
quantity of grain under the dynasty of the 
Brahman Kings of Jafnapatam. 
{To be continued.) 
CONTRIBUTIONS TO CEYLON 
GEOLOGY: 
OCCURRENCE OF CORUNDUM IN SITU 
NEAR KANDY, CEYLON. 
(By A K Coomaraswamy B Sc., F.L.S., F.G.S., 
Director of the Mineral Survey of Ceyloii.) 
[Extracted from the GEOLOGICAL MAGAZINE, 
Decade IV", Vol, X, No. 470, August, 1903.] 
The present'notes are based on field observations 
made in 190O. Tlie section described is now 
obscured. 
Corundum Is abundant in the gem-bearing 
gravels of Ceylon, but with the exception of the 
case here described no localities are known where 
it occurs in sitH ; the prchcnt occurrence is there- 
fore of considerable interest, although not very 
eatisfactory in itself. Crystals of corundum were 
found in the surface soil on a piece of land known 
as Tenna Hena, and situated east of Kandy, and 
three quarters of a mile north-east of Talatnoya 
bridge. The exact spot is shown in a map accom- 
panying a paper on the crystalline limestones of 
Ceylon (Quart. Journ. Geol. Soc, 1903, vol. Iviii. 
pi. xiii). A small excavation had been made, 
and a few pounds of comnUum extracted and sold 
for use as emery, before my visit to the spot. All 
the rock exposed was decomposed, and crumbled 
in the fingers, being in a condition resembling 
sand.* I therefore carried on an excavation for 
two months, hoping to reach hard rock suitable 
for microscopic examination, but although a depth 
of about 30 feet was reached, no sufficiently hard 
rock was found. 
♦ It is very useful for the grannlitic rocks of 
Ceylon to be found in this friable, sandy condition, 
to a considerable depth. This mode of alteration is 
totally distinct from the formation of laterite, nor 
does it appear to be dae to the kaoliaisation of the 
felspam, as these are tranalncent, and the analysis 
shows that but little water is present. The change 
partakes perhaps rather of the nature of a physical 
di integration. 
At the corundum pit the ' beds' oi granulite 
dip northwards at a high angle. A conspicuous 
soft yellow micaceous band 7 — 3^ inches wide 
marks the position of the sapphire-bearing zone^ 
The sapphires occur in fair abundance in a less 
decomposed felspathic rock occupying a lew inches 
on either side of this yellow micaceous band 
in the upper part of the shaft, but on the south 
side only in the lower part. The associated types 
of granulite are chiefly acid leptynite. The corun- 
diferous band is about three yards from the 
northern boundary of a band of crystalline lime- 
stone about seven yards wide (in the lower part} 
of the pic the distance was apparently less). There 
is nothing to suggest any connection between the 
occurrences of corundum and limestone. It is a 
little strange that corundum has not so far been 
found in the crystalline limestones of Ceylon, 
although so characteristic of similar rocks in 
Burnmh, 
'1 he sapphires are of fair size, the largest 
about three quarters of an inch in diameter, and, 
though of a bright blue colour, are useless as 
gems owing to their opacity and well-developed 
cleavage, and often weathered, bleached, and 
hydrated condition, Rhombohedral cleavage and 
a basal parting are alike well displayed. Combi- 
nations of tlie hexagonal prism and basal 
plane are most usual, giving a columnar 
aspect ; the forms observed include c (OOOi), 
a (1120), r(lOll), n (2243); some double 
crystals with basal planes inclined at; 
a little over 90° resemble twins, but the basal 
planes (cc) are not quite in the zone cr, so that 
this resemblance appears to be deceptive. I am 
indebted to Mr L J Spencer, M.A., for these crys- 
tallographic details. 
The soft, yellow, micaceous band consists of 
biotite, pbigioclas (quite fresh and glassy), greenish- 
yellow, soft, serpentine-like decomposition»products 
after pyroxene (?), and minute quantities of garnet 
and iron ore. An analysis (No. I) by Mr W C 
Hancock, B A., shows that this yellow micaceous 
band contains a relatively slightly larger propoi-- 
tion of alumina than the corundiferous rock itself. 
The rorundiferous felspathic rock consists mainly 
of orthoclase-microperthite, with also plagioclase, 
biotite, coiundurn, and small quantities of garnet, 
green spinel, and zircon. It is not possible to 
make quite certain of the total absence of quartz ; 
a consideration of the amount of alkali, which is 
according to the analysis, available would indicate 
the presence of a small percentage of free quartz ; 
I have not, however, been able to detect any. 
The niicroperthite is in a very fresh condition, 
the plagioclase still more so. The corundum in 
the rock has usually a ' court ' of felspar free 
from biotite, separating it from the remainder of 
the rock, consisting of felspar with scattered 
biotite. 
Professor Sollas, F R S., has very kindly made a 
mineral analysis of the crumbled rock, with the 
following results : — 
Heavy minerals over 3"34 (cbiedy cornndnm)... 7.6 
Orthoclase, s.gr. 2'56 ... 64 2 
Oligoclase, (with possibly a little qnartz), B.gr.2*65 23'5 
Biotite, s gr. 2-8-2-92 .. ... 4.7 
100-0 
Mr Hancock has chemically analysed the same 
material (No, II); No. Ill shows the same with 
water removed and ratios calculated to 100; No. 
IV, the molecular rations. 
