716 MIN ERA LOGIC Alj DESCRIPTION OF THE ISLAND OF BANKA. 
South of Pangk&l-Bulu a large tract of alluvial country is includ¬ 
ed between the hill Kappu and the mountain Mangkul: the large 
river of Marawang and many of its branches pervade it. With re¬ 
gard to the south-east division of Banka my mineralogical remarks 
are very limited I visited only part of the districts of Pangkal-pi- 
nang and Tirak. 
I shall commence the detail of the remarks made on a route in a 
south-west direction from the stockade. In the immediate environs 
the tract is still low and alluvial. The mines of Bakung, situated 
in the south-west, from the numerous remains, appear formerly to 
have been very extensive: the detached fragments which I noticed 
are of a mixed nature, and their origin can be traced to the remain¬ 
ing fixed substances. 
I found extensive tracts consisting of veins of rocks of Red-Iron¬ 
stone and the canals were often cut through deep layers of this sub¬ 
stance. In other parts the surface was covered with that mixture 
of clay and sand which has frequently been mentioned as forming 
strata of the mines; on this siliceous fragments were dispersed. I 
noticed principally; 
1. That species of crystallized rock (mentioned in the account 
of the mines of Sungie-liat and Robo-kli) consisting entirely of 
quartz, while the felspar appears on the fracture in a decomposed 
state as a white friable powder, 
2. The variegated, often marbled mass, formed of nearly equal 
portions of quartz and felspar. The promontories of the Gunong 
Mangkul extend in this direction and the siliceous fragments found 
at these mines have probably been derived from them. 
The layers of Red-Iron-stone at these mines (of Bakung) were 
similar to those above described at the mines of Sunwan at Mara¬ 
wang in arrangement and extent • I remarked here, however, the 
peculiarity that they were often intersected by extensive veins of 
fixed rock. The separate fragments resembled externally those 
found at the mine of Sunwan, but the fracture was different, as well 
in those fragments which compose the layers as in the fixed rocks; 
it was more close and compact and the colour varied from a deep 
reddish brown to nearly black; in some it exhibited a variegated 
metallic lustre. A few of the detached fragments remaining from 
the washing of the ore were amygdaloids ; those which composed 
the strata were tabular and externally corroded. In a few speci¬ 
mens I observed the same disposition to form nodules which has 
