clays are found accumulated on either granite or some hornblende rock.
Near the Tacuparu creek a decomposed granite appears, also at the rapids of Maritote, and others below them, where the granite and syenite are extensively exposed.
After passing several rapids, including that of Alabamaloo, the syenite is more general. A remarkable dyke of porphyry, 4 feet wide, which has passed through the granite, occurs at the falls, which were about 12 feet in height at the time I descended.
The granite series, with trap dykes, are more largely developed. At Huamata they cross the entire river. Most of this granite is very coarse-grained, with green or red felspar.
At the rapids and the island of Taparu, also at the rapids of Presatu, white granite occurs ; and near Cubangan, gneiss, which extends down to the Ourasas-sartur rapids.
The land rises here, on each side of the river, to from 150 to 200 feet. At the base granite, syenite, trap, and quartzose rocks spread out over a much larger surface, in comparison with those on the upper part of this river.
After passing the great falls at the junction of the Puruni with the Mazuruni, there is so great an extent of bare granite separated into innumerable islands, and such a large expanse of water within the low and distant confines of the river, that it presents a scene rarely to be met with.
On entering the Mazuruni I found the river unusually low, and consequently many rocks exposed that would not have been seen had the river been in flood ; but all are so uniform in their character (granite, syenite, gneiss, and porphyry, with trap dykes) as to render the labours of the geologist very monotonous.
These rocks continue up the river without any important modification as far as the Wanamu and Maribisi rivers, where a fine conglomerate occurs. Above this there is an altered sedimentary rock, forming a mass.
At the falls of Tebuco there are granite, diorite, and very dark porphyry, also veins of dark quartzite, followed by greenstone, and schistose rock with quartz veins, both black and white.
At Carinang Creek the same rocks occur, and continue some distance further up the river. Some of the quartz veins are bold, being several feet across ; but no gold could be detected in them with the aid of a strong magnifying power.
About this part of the river I obtained the first view of the Maribiacru cliffs, which form a part of the Mereme range, rising from 1000 to 2000 feet over the forest-trees growing on the banks of the river.
At a point turning south, towards the cliffs, granite underlies 8 and 10 feet of recent conglomerate and sand interstratified ; then the granite again rises to the surface.
At this point the heights of Caribisi are seen to rise in a peak about 18 to 20 miles distant ; and not far above, conglomerate is again seen lying on white granite with quartz veins. This is fol-
