Alluvial districts, evidently of recent origin, are noticed in several parts of the island. These are formed from the sediment and near the discharge of large rivers, and at the borders of the calcareous ridges, which are in many instances partially covered by them: their boundary can easily be traced, and most of them are still in a state of constant progression. Among other phenomena are mineral wells of various temperature and impregnation; wells of naphtha, or petroleum; and rivers arising, in a few instances, from the craters of volcanos, impregnated with sulphureous acid[24]. From these, and all other investigations yet made, the constitution of Java appears to be exclusively volcanic. From the vast Asiatic chain of mountains, one branch of which terminates in Ceylon, proceeds another, which traversing Arakan, Pegu, and the Malayan peninsula, extends to Sumatra, _Bánka_, and _Bíliton_, where it may be said to disappear. On Java no granite has been discovered. In its constitution, as in its direction, it may be considered as the first of a series of volcanic islands, which extend nearly eastward from the Straits of Sunda for about twenty-five degrees. At what period the island assumed its present shape, or whether it was once joined to Sumatra and _Báli_, is matter for conjecture. The violent convulsions which these islands have so often suffered, justify a conclusion that the face of the country has been frequently changed, and tradition mentions the periods when Java was separated from those islands;[25] but the essential difference which has been found in the mineralogical constitution of Java and Sumatra, would seem to indicate a different origin, and to support the opinion that those two islands were never united. Whether, at a period more remote, the whole Archipelago formed part of the continent of Asia, and was divided from it and shattered into islands; whether they were originally distinct from the main land, or whether they were formed at the same time, or subsequently, are questions we cannot resolve. Yet, when we reflect on the violence of those dreadful phenomena[26] which have occurred in our own times in the smaller islands of the volcanic series, and view this range, as it is now presented to us on the map of the world, a conjecture might perhaps be hazarded, that the whole may have once formed but the southern side of one large island or continent, within which much of the main land has fallen in, and subsequently disappeared on the influx of the sea.