Arthur Hohnes — Lateritic Deiiosits, Afozambique. 533 



insoluble in water alone, their efficiency in the production of laterite 

 is doubtful.' Alkaline solutions readily dissolve the humus acids, 

 and it therefore appears that where dissolved silica is found in 

 association with the latter the alkalis have been the effective solvent 

 of the silica. 



Laterite was never observed in the neighbourhood of marshy land 

 or swamps, which appeared always to be comparatively free from 

 ferruginous deposits of any kind. 



In view of the evident importance of underground drainage in the 

 genesis of laterite, it may be of interest to describe a cave deposit 

 which was examined by Mr. Wayland and myself. Our first base 

 camp was built at Sawa, on a projecting spur in a deep valley 

 running northwards into the Kibawe Mountains. On each side of the 

 valley the mountains rose up precipitously, and the foliation of 

 the gneiss could be plainly seen dipping away at a gentle angle to the 

 north (Plate XXXVII). On the eastern side the nearly vertical face of 

 the gneiss was eaten away here and there into caves, the entrances to 

 which ran along tlie 'dip'. Those which we explored widened and 

 narrowed in the most remarkable way, defying explanation. In 

 each case a steady drip of water trickled from the roof and from 

 narrow fissures wiiich penetrated the rock parallel to the foliation, 

 thus demonstrating that a considerable drainage of water percolated 

 northwards down the * dip ' through the apparently impermeable 

 gneiss. The roofs of the caves were lined with a thick highly 

 polished deposit of limonite which also covered the upper parts of 

 the walls, gradually becoming thinner as the floor was reached and 

 giving place to bright pink rhodochrosite and black pyrolusite. 

 Where a true floor was visible it was also covered with these 

 minerals, the manganese oxide and carbonate coating the thicker 

 deposits of limonite. In the largest cave, however, much of the floor 

 Avas buried under the droppings of bats, and this was encrusted with 

 fine needles of vivianite. We had a hole dug through this material 

 in order to expose the gneiss below, and were astonished to find 

 a considerable thickness of sand, in which, at a depth of about a foot, 

 we found the ashes of a wood fire and fragments of coarse pottery, 

 proving a former human occupation of the cave and suggesting that, 

 at least in part, its excavation may have been due to human agencies. 

 The gneiss itself, when uncovered, was found to be perfectly fresh. 



The source of the deposits was not difficult to find, for the hill is 

 penetrated with numerous pegmatite veins rich in magnetite and 

 franklinite. Water, charged with carbon dioxide, would readily 

 leach out iron and manganese as carbonates. On reaching a cave, 

 oxidation and evaporation would cause a precipitation of iron- 

 hydroxide, followed by that of the more stable manganese carbonate, 

 some of which would slowly become oxidized to pyrolusite. 



This example has been described in order to demonstrate the 

 percolation of iron-bearing solutions along the fissures and foliation 

 planes of gneiss, a phenomenon which appears to be responsible for 

 much of the laterite of Mozambique. Similar cave deposits have been 

 recorded from Mt. Tugwi, near Nampula. On the western side of 

 ' Clarke, Data of Geoche?nistry (U.S.G.S. Bull. 491, p. 98. 1911). 



