Dec. 2, 1901.] THE TROPICAL 
AGRICULTURIST. 
411 
extinctions from 18 dg.— 18 deg. to 22 ilg.— 22 dg. 
Athin sectionhas been cut from a specimen taken 
near tlie small wayside station of Balana, on tlie 
Colombo side of Kadugannawa. Here the darker 
rock is in excess, but the white bands are still 
clearly seen in places. Under the microscope we 
see it to be closely related to the series already 
described. It contains a considerable quantity 
of hornblende and mica ia approxiraateiy equal 
proportions. 
A thin section has been cut from a very similar 
rock cropping out on the path to Lanka Tileka 
near Karidy. It is a black and white speckled, 
rnck, the white constituents being granular and 
.saccharoidal. [J'lder tiie miscroscope the resem- 
blance to the Halana rook is very great : the fer- 
romagnesian silicates are somewhat less strongly 
jeveloped, but iron-ores are not uncommon, 
(c) Railway to the north of Mahaiyawa 
^ANDY). — A weil-pre>erved banded rock is found 
the north of Mahaiyawa. It sometimes con- 
•^^ins quartz-grains of considerable size and, 
**cally, garnet. Not far from this is a small 
^°uarry near the railway, excavated in a grey bio- 
lite-wneiss, streaked rather regularly by a red 
oarse granite, which occasionally contains large 
black uiica plates. The rock is often very coarse, 
and exhibits big eye-s!iaped telspars such as were 
seen near Rambukkana. The bands which this 
granite forms run parallel with the general folia- 
tion, and from the iield-evidence seem clearly to 
be an integral part of it. A slide cut from a fairly 
coarse specimen of the granite shows that it is 
essentially the same rock as that wdiicli forms 
the lighter-coloured part in the cuttings near 
Rambukkana (p. 199). Both rocks are character- 
ized by microcline and micropertliite, and by 
the presence ot large, very irregular quartz- 
grains, while the ferromagnesian silicates are 
represented in both by a few flakes of mica. A 
thin section taken from a specimen shorving the 
banding of this rock with the lionite-gneiss, 
demonstrates that no line of demai'cation can be 
drawn bstweea the two varieties, that is, that 
whole must have solidified at tlie same time. Tlie 
darker part, that called biotite-gneiss, consists of 
an interlocking masaic of quartz and felspar, 
with orthoclase and plagioelase. A little micro- 
cline is found, and one or two grains of ortho- 
clase contain microperthite. Biotite with a slight 
foliation is the sole ferro-uiagnesian silicate, and 
possesses the same characters as before. The 
specimen is almost identical with one from near 
Rambukkana. 
A little farther north we come to a very slabby, 
rather tine-gtaiued, and uniform pinkish ' granite,' 
the planes of parting being determined by mica- 
flakes. This ' granite' forms the greater part of the 
cutting, but the gneiss also appaai s, and speeiaiens 
were taken showing the junction of the two. A 
thin section proves a perfect gradation between 
the two rocks and both are characterized by the 
same type of structure. 
The pink rock contains no ferromagnesian sili- 
cate at all ^^except the black mica just referred to), 
and is distinguished by a quantity of o"tiioclase 
with microperthitic intergrowtli, though the 
mineral does occur without it, and a little 
plagioelase. The rest of this rock consists of quartz 
in Tmall and very irregular grains, which fre- 
quently occur as inclusions in the felspar and 
vice verm. The other part of tiie slide (that is, 
the gneiss) contains orthoclase without a micro- 
perthitic intergrowth, large and very irregular 
quartzes, some plagioelase, and no mircrocline in 
addition, we find a few flakes of mica. In passing 
from the pinker rock to that just described the 
first sign of change is the appearance of the larger 
grains of quartz, and then the gradual disappear- 
ance of the micropertliitic structure. It seems 
clear that the pinkish ' granite' is a fine-grained 
representative of the coarse rock which streaks 
the biotite-gneiss in the small quarry above des- 
cribed. 
The rocks of Kadugannawa, Rambukkana, and 
Mahaiyawa are usually well foliated^ and com- 
monly banded. The dark bands are characterised 
by green hornblende in varying quantity, and the 
association with this of a brown mica. Garnets 
are found locally. Field-evidence shows that the 
inter-relationship of the light and dark bands may 
be best explained by the streaking together of the 
component parts of a magma which had undergone 
differentiation, and that occasionally this process 
was carried to a. rather unusual extent. 
III. CONTACr OF LIMESTONE AND PYROXENE- 
GRANULITE. 
Railway between Matale and Ukuwela. — Oa 
leaving Matale Station we see, first, a few 
outcrops of white crystalline limestone, followed 
and overlain by hard red soil. This exhibits a 
number of small ferruginous concretions, which 
are, however merely the outcome of a weathered 
condition, for internally it is mottled red and 
yellow, audis apparently an argillaceous soil rich 
in iron. 
In a short distance we come to a level crossing, 
and on the left hand side or the track a small 
section is disclosed by the railway-cutting. The 
dominant rock is a crystalline limestone, but in 
this, and looking like an approximately horizontal 
dye, is a fine-grained rock rather saccharoidal 
on a fractured surface, but with a greasy lustre. 
Tills very closely resembles the pyroxene- 
granulites exposed on the road to Hakga'la from 
Nuwara Eliya (see p. 207). This rock which I 
will call ' the Band,' in order not to prejudge the 
question of its intrusive nature, is abonc 4 inches 
thick at the only place where its upper and lower 
surfaces are seen. It can be traced for about 5 
yards, the base not been seen ; and in one place it 
forks, and partly includes a mass of limestone 
about a yard square. 
After this exposure, we find a good cutting of 
well-crystalized limestone (specific gravity=2-86), 
followed by another of red rook similar to that 
seen before, butless concretionary on the weathered 
surface. This is succeeded in turn by limestone, 
and again by the red soil, which must be a kind 
of laterite. 
Thin sections from the ' Band' and limestone will 
now be described. In a slide cut from a specimen 
of the former (specific gravity=2-9G) the dominant 
mineral is a pleoeliroic augite. In an orthopina- 
coidal section the colour for vibrations normal to 
the prismatic cleaveage is a pale pink with a 
sliuht tinae of crimson ; at right angles a pale, 
rather bluish, green. The pleochroism thus closely 
resmbles that of hypersthene. The polarization- 
tints are brilliant, rather recalling those of olivine. 
Occasionally this augite is of a medium shade of 
sage-green, non-pleochroic or but feebly so, and 
proliably merely a variety. Mixed in with this 
augite area few irregular grains of pale red garnet, 
which are absent in a second slide. The augite ia 
altering to brown hornblende, pleochroism ot which 
