Galle lighthouse, ramparts and cockerel, by Muthu.
Last modified:
14 Dec 2001

Preliminary geological-petrological analysis of a ballast stone from site G

by Ananda Gunatailaka B.Sc. (Wellington, NZ), Ph.D. (Reading, UK)

The examined sample is a spherical mass about 10 cms in diameter. Hand examination of a freshly broken surface indicates that it is a coarse-grained mafic to ultamafic igneous rock, with distinct phenocrysts or xenoliths of olivine and pyroxene crystals (mostly). The olivine nodules are being altered to a yellowish-green mass of probably serpentine at the surface. The largest nodule is about 1 cm in length, while the pyroxene crystals were up to 4mm in length. The sample is of high specific gravity (>3.5).

A microscopic thin-section examination (three thin-sections were prepared) of the above rock shows mostly euhedral, well-zoned or unzoned pyroxene crystals and olivine nodules (forsterite composition) which are either fresh or in various alteration stages to serpentine. Pyroxene and olivine are the dominant minerals, making up about 75-80% of the rock with pyroxene > olivine. A few crystals of plagioclase felspar (labrodorlte-bytownite composition) are evident, but make up less than 2% of the rock. These three minerals show breakage or fracturing indicating that the cooling magma was subjected to an explosive event. The most distinctive aspect of the rock is the high proportion of iron-oxides (magnetite?), making up about 8-10% in volume. The groundmass was indeterminate, but may have substanital amounts of pryoxene. The three major silicate minerals form an interlocking texture under the microscope.

The above composition and texture, with the very small amount of felspar, appears to suggest a rock of peridotitic composition (ultramafic). On the other hand, the fractured nature of the crystals - suggesting an explosive episode during cooling - probably indicates an original mela-alkaline magma which filled up and crystallized in a volcanic vent or crater at depth, which is akin to a volcanic plug. The above composition indicates why the rock would have made ideal ballast material (high iron content in the minerals).

It is not possible to say whether the original rock was fashioned into spheres at a quarry for the specific purpose of using as ballast or whether the spherical masses represent some extreme example of natural spheroidal weathering. Or is it possible that the rock spheres were excess cannon balls from a previous historical period, which were later found to be ideal ballast material. In any case, sheres of identical diameter would make for easy packing, stacking and storage in a ship (c.f. the ideal cubic or hexagonal close packing of spheres in nature).

The rock could have come from anywhere else, but definitely not from Sri Lanka. The author is familiar with similar rocks of Carboniferous-Permian age in Glasgow, Scotland. Similar rocks are known from Italy, Norway, Saudi Arabia, Aden and, perhaps, from the Pyrenees region. They are not known from the Low Countries of Europe (Holland, Belgium or Brittany coast). A detailed chemical analysis of the rock could probably indicate its provenance by matching it to an existing petrochemical database of similar rocks.

Maritime Lanka homepage   Galle index