THE TROPICAL AGRICULTURIST, [October * 1887 
elaborating this substance into beatiful shapes, 
from immense masses of brain coral, almost solid, 
to the most exquisite traceried branch corals, the 
hues of which in the water are varied and 
brillant, including green and gold and loveliest 
pink shades. From the violence of the waves 
*nd by processes of natural decomposition which 
doubtless accompany certain stages of life in the 
zoophytes, the corals thus formed have been 
broken up, comminuted by attrition and finally 
agglutinated together by the abundant carbonate 
of lime into breccia, into which dead and 
broken shells also entered. Gradually the frag- 
ments solidified into the form in which the 
" madrepore " now exists, and the elevating 
process already alluded to having raised por- 
tions of this formation to or a little above the 
suriace of the sea, the theory is that ourrents 
carrying sand and soil from the coast of India 
did the rest. Our correspondent entertains a 
different opinion regarding the origin of the 
red soil of Jaffna, and we shall be very glad to 
publish his views, if he will furnish them for 
information and discussion. In 1846 47, the late 
Kev. E. H. Hoisington sent us several analyses 
of Jaffna soil, amongst the rest that of the 
hills of white-antB, and if the records of Batti- 
cotta Seminary are searched we think it highly 
probable that copies of the analyses will be 
found included. Tbers is a natural formation 
of salt near the marine stream (haunted by 
crocodiles) at Tondaimannar, about seven miles 
west from Point Pedro. We do not reoollect 
seeing evidences of the presence of oxide of iron 
in connection with this formation, but we know 
that when we sent Mr. Steele, when he was 
Assistant Agent at Hambsntota, a notice of the 
great Sambur Salt Lake in Rajputana, showing 
that deep red contrasted with the silvery white 
crystals of the salt, Mr. Steele stated that the 
effect of peroxide of iron was also visible in 
the salt "lewayas" of Hambantota ; which 
by the way are in the neighbourhood of 
drifting sand "dunes" as formidable as those 
on the north east coast of Ceylon. Our corre- 
spondent will, of course, direct his attention 
to the limestone caves at Alvai, a little inland 
from Point Pedro, in which it is supposed is 
the source of the fresh water which bubbles 
up in the sea nea Kankesanturai. He will also 
study Tennent's extraodinary theory (founded on 
a popular belief), that besides fresh water seek- 
ing the sea (as is natural) in the Jaffna 
" madrepore " formation, the sea water goes inland, 
against natural laws, the substratum of salt 
water at bottom of the Puttur well forcing up 
the fresh ! A paper dealing fully with the pe- 
culiar physical conditions of the Jaffna ooral 
formations, overlaid by Indian sand (and soil?), 
could not fail to be very interesting. 
Dr. John Davy was himself never able to visit 
Jaffna, but he analyzed specimens of rock and 
soil received from his friend Mr. Finlayson, and 
this is what he wrote: — 
Tifis limestone contains numerous shells : it is gener- 
ally grey or light-brown, very hue-grained and c^m- 
piot, and breaks with a conchoidal fracture. The 
specimens I have tried, have boen very nearly pure 
carboiiiite of lime, exhibiting slight traces of the pre- 
sence of vegetable or animal matter, and containing a 
little water. It is not confined to the island of Jaffna- 
uatnni ; it occurs iu the district on the mainland, and 
li 18 been observed by Mr. Finlaynon as far as Pal- 
iverayenkatte, whore he fouud it with coral rock, in a 
gait-water iako, at a little distance from the sea, with 
which the lake communicates in the rainy season, 
f^here it occurs, the wholo of the country is similar 
ii vcl. without hills or even hillocks and elevated a 
very few feet only above the surface of the sea, by 
which, at no very remote period, there is good reason 
to suppose, it was once covered. The retiring of the 
sea from this district does not admits of a doubt. It 
is evident within the memory of man; many indivi- 
duals recollect the waves breaking where their spray 
now seldom reaches : Nor is it less evident, perhaps 
from the nature of the land as described, and from the 
circumstance of coral rock being found mixed with the 
limestone rock several miles from the sea. It is always 
more easy to observe the phenomena of nature, than 
to point outtheir causes, especially in geological changes, 
such as the present, which are not watched in the act, 
and are noticed only when completely accomplished. 
MiDute enquiry, on the spot, it is very likely, might 
afford a clue to an explanation of the formation of 
this rock, which, in all probability, is still going on in the 
shallows of the adjoining seas, and along the 6horesof 
Jaffnapatam. Its formation, it may be conjectured, 
maybe connected with coral which abounds greatly 
in t he narrow sea, between Ceylon and the continent 
of India, to such an extent, indeed, that most, if not 
all the islets in that sea, are composed of it. The 
difficulty is to find the cause of the solution of cal- 
careous matter in some places, and its precipitation in 
others adjoining. Perhaps, in the deeper, cooler water 
it is dissolved, and iu the warmer, shallow water it is 
precipitated. The solution is aided, perhaps, by th« 
presence of a little carbonic acid, and the precipit- 
ation is assisted or produced by the escape of the acid 
gas. This is mere conjecture, but of that kind that it 
admits of being tried by the test of experiment. 
Again, writing of Ceylon soils, he remarked : — 
The most striking iustance to be adduced, of soil 
destitute of calcareous matter, and incumbent on a 
bed of coral, is the common soil of the coral island 
Delft, off the coast of Jaffnapatam, celebrated for its 
excellent pasturage. It is a dirty-yellov, very fine 
sand, Slightly cohering which consists, in its dried 
state, of 
95.0 silicious sand coloured by iron, with 
perhaps a very little alumina 
2.5 vegetable matter 
2.5 water 
100.0 
Nor hardly less remarkable are some of the soils of 
Jattnapatam, for which as well as those of Delft, I 
am indebted to my friend, Mr. Finlayson. Two in. 
stances may be mentioned, — one of a soil of a to- 
bacco field, which is manured by means of sheep, like 
turnip-ground in England ; the other of rice-ground, 
whioh receives no manure, but is carefully irrigated. 
The tobacco soil, of a reddish-brown colour, collected 
when perhaps partially exhausted, the crop not hav- 
ing been long taken off the ground, consisted of 
95.5 silicious sand, coloured by iron with 
a few particles of calcareous matter 
2.0 vegetable matter 
2.5 water 
100.0 
The rice soil, of a light grey colour, containing a 
good deal of straw in a finely divided state, consist- 
ed of 
95.5 silicious sand, with traces of iron, 
carbonate of lime and alumine 
2.5 vegetable matter 
2.0 water 
100.0 
It seems exlraordinary that in slands, the founda- 
tions of which are calcareous, there should be so little 
calcareous matter, and so large a proportion of silici- 
ous matter, in the soil. It is a subject deserving of 
minute inquiry on the spot; perhaps, the fine . sili- 
cious sand is drifted there by strong wiuds from a 
