The eruption of vast quantities of acid pumice tuffs stands out as the major event in the initial stages of Quaternary volcanism on Sumatra. By their strict alignment along a trend of longitudinal fault-trough systems, which follow the entire length of the island and cut through Mesozoïc as well as Miocene fold belts, these rhyolitic volcanics strongly suggest that they are sheetlike deposits of paroxysmal fissure eruptions emptying magma chambers in an advanced stage of differentiation. During settling, the erupted material has evidently been similar to fiery clouds (“nuées ardentes”) of the Katmai type. Enormous gas pressures, gradually developed in the vault spaces of the original magma reservoirs, account for the sudden disrupture of the overlying roof portions along pre-existing lines of structural weakness.
As contrasted with later andesitic volcanism, these outbursts built up no central cones, but rather extensive sheets with plateau-like surfaces. In the Lake Toba and Pasoemah1 Highland rhyolite tuffs, the volcanic sheets appear to consist, in at least their lower portions, of welded tuffs with columnar structures bearing all characteristics, both macroscopic and microscopic, of the New Zealand ignimbrites. In extreme southern Sumatra, however, they are simply deposits of loose pumice tuffs, whereas similar rocks in the Padang Highlands in Central Sumatra still await further investigation.
At many places acid lavas of about the same age, and still preserved as partly mineralized rhyolitic plugs, rose along the same or adjacent channelways in South and Central Sumatra.
Later dacitic to andesitic, and in isolated cases also basaltic, volcanism built up central cones above longitudinal as well as prominent transverse faults or fissures. Basaltic lavas resembling true plateau basalts were poured out on the surface of the rhyolite tuffs in a limited area near Soekadana in eastern South Sumatra, well beyond the zone of Miocene orogenesis of the western coast ranges.
Repeated eruptions of acid pumice tuffs along the same longitudinal graben trends interrupted the andesitic period at the time of the great pre-historic Lake Ranau eruption in western South Sumatra, and, comparatively recently, at the occasion of the famous 1883 Krakatau pumice eruption, which apparently took place at the intersection of the submarine continuation of the Semangka graben in South Sumatra and an important transverse fault. Another young Quaternary explosion produced the acid block tuffs around the Pilomasin basin about 60 km northeast of Lake Ranau.
The trend of differentiation of Quaternary Pacific magma types on the Sumatran mainland is almost duplicated within the limited space of the various magma types of the Krakatau island group with its basaltic to rhyo-dadtic lavas and tuffs; the extremely acid effusiva of Sumatra, however, are without equivalents.