JUNE 20, 1884.] 
the same connection, one should consult Ver- 
beek’s earlier report on southern Sumatra, 
which contains descriptions of Krakatoa itself 
before the eruption.? 
. Fig. 1 is taken from a series of generalized 
profiles illustrating the geological history of 
Sumatra. .Archaean rocks are nowhere seen. 
3 ee 
: ee " eri 13 
a cS as Es sees 
2 = Be oe ‘ aS 
$9 c i ae ee 
5 iS ) EN Y S| 
ar: OO ee 
= 4 S SS 
Old slates. 
Old slates. 
S.W. Fie. 1. 
The oldest members of the series are non-fos- 
siliferous slates and limestones, in places hold- 
ing quartz veins that are sometimes aurifer- 
ous, and cut by eruptives of the granitic group : 
these are overiaid by limestones, well proved 
to be of carboniferous age, cut by diabasic 
eruptives. Mesozoic strata are absent, im- 
plying a general elevation to a broad land- 
surface, followed in eocene time by depression 
again, during which workable coals were 
formed. ‘There are other tertiary strata, such 
as the miocene beds of the small islands to the 
south-west, succeeded by broad quaternary 
deposits over the lowlands. ‘The early tertiary 
eruptives (basalt and hornblende andesite) are 
relatively scarce, and are but dwarfs among the 
gigantic cones that have been heaped up since 
the end of tertiary time. ‘These are chiefly 
augite andesite, mostly in the form of ashes 
and sand, holding larger blocks, but sometimes 
as dikes or lava-flows. They reach almost 3,000 
metres altitude, flattening from a slope of 30° 
or 35° at the summit, to an almost level plain 
at the base, with a curve of descent that is 
shown to be closely logarithmic in its form. 
Krakatoa (here called Rakata) is one of these 
cones, standing on the most south-eastern 
transverse group of the great range of Suma- 
tran volcanoes, of which sixty-six are given in 
a list, and seven among them (not including 
Krakatoa) are marked active. A considerable 
share of attention is given to lithology ; and on 
the atlas sheets, the different classes of eruptive 
rocks are distinguished. ‘There are also spe- 
cial descriptions of the several craters formed 
1 Topographische en geologische beschrijving van Zuid-Su- 
matra. Door R. D. M. VERBEEK (215 p., with geological map, 
profiles, etc.); Jaarboek van het mijnwezen in nederlandsch 
Oost-Indié, 1°, 1881. Our figures, 5, 6, 8, are from this work. 
SCIENCE. } 763 
successively about the great volcanic centres, 
—as on the summit of Merapi (fig. 2, ideal 
section), where four concentric walls, almost 
unbroken, stand one within the other, a gi- 
gantic cone-in-cone structure,—and also of 
the formation of volcanic lakes, from the small 
ones in the well-preserved craters, to the large 
basins of Maniendjoe 
(100 O kilometres in 
area), the result of a 
central caving-in of a 
great volcano whose 
remains are seen in the 
surrounding Danan 
Gebergte, or Lake 
Mountains; and the 
still larger Singkarah 
(112 oO kilometres), 
formed by eccentric 
subsidence. 
The theory illustrated by von Hochstetter! 
is quoted to account for the mechanism of these 
changes. His figures are therefore here repro- 
Basalt. 
2/ Granitic rocks. 
Old slates. 
N.E. 
duced, with slight alteration, as of additional 
Fig. 2. 
value from their acceptance by an observer 
practised in the study of volcanic phenomena 
on the largest scale. Fig. 3 shows the effect 
of continued eruption in melting the interior 
part of the cone previously formed: the vol- 
cano is here active. Fig. 4 shows the falling- 
~ 
ote ee, 
in of the cone when the molten interior is 
blown out, or allowed to sink, and, in this 
1 Ueber den inneren bau der vulkane. Neues jahrb., 1871, 469. 
