Vol. 66.1 
VOLCANO OF MATAVANU IN SAVAII. 
627 
there are points where the cone rises to a height of 350 feet above 
them. 1 The outside slope is in most places gradual; I did not 
notice any place where it appeared to exceed 33° (PI. XLY, fig. 2). 
The crater itself is oval, and apparently about 400 by 200 yards in 
area; its long' axis runs approximately south-west and north-east. 
Its inner walls are very steep throughout, and in many places 
precipitous and even overhanging. They are composed of the usual 
beds of lava and fragmentary material; but, except in the southern 
wall, where tuffs are abundant, the lavas predominate much more 
than in some other volcanoes that I have seen (PI. XLVI). 
When on the spot, I estimated the depth down to the lava-lake to 
be mentioned presently at about 400 feet, possibly somewhat less ; 
but, on subsequent comparison with Kilauea, I do not think that it 
exceeds 300 feet. At about half the depth is a well-defined bench 
on the walls showing a former height of lava. 2 The strata below 
this level are much obscured by this dense lava formation; there 
is also a less marked bench at a still higher level, say about a 
third of the depth from the lip (PI. XLVII). 
The bottom of the crater is of an oval form, proportionately 
narrower than the higher part, in fact with nearly parallel sides. 
I roughly estimate its length at 300 yards, and its breadth at 75 
to 100. It is entirely occupied by a lake of liquid lava all in rapid 
motion from south-west to north-east, and of such extreme fluidity, 
that it continually beats in surging waves against the walls, where 
splashes retain their heat and brilliant colour for sometime. The 
surface is in a constant state of ebullition, though not always to 
the same degree in different parts. Some of the boilings rise in 
veritable fountains of incandescent liquid basalt of 10, 20, or even, 
I think, possibly at times 50 feet high. The whole mass of the lava 
is at a brilliant white heat, visible as such even in bright sunlight, 
but a darker senrn is continually forming on the surface, especially 
when the trade-wind blows strongly on it. These pieces of scum, 
like ice-floes, break up and flow down to the north-eastern end of 
the crater, where they and the liquid lava disappear down a hole, or 
rather tunnel, at the foot of the cliff. The tunnel is perhaps 30 feet 
wide. Its roof is quite low, and is nearly touched by the surface of 
the lava, which rushes under it at a steep slope with the velocity of 
a cataract. There is also another smaller hole at the foot of a cliff 
forming the north-western wall of the crater, down which a stream 
of lava seems constantly to flow; but this is small in comparison with 
the other. The wall forming the north-eastern end of the crater 
above the outflow of the lava is of a structure different from that 
forming the adjacent parts on each side, and seems to be a great 
ebasm filled in with fragmentary material of a later date, and 
possibly partly with lava frozen on to the roof of the conduit. 3 
1 See H. I. Jensen, Proc. Linn. Soc. N.S.W. vol. xxxi (1906) p. 653. 
2 This may ha\e been the height of the lava in 1906, as mentioned by 
Jensen, op. cit. p. 653. 
3 I saw this clearly in the evening and night on two distinct visits; but, 
when I returned to photograph it by daylight, that part of the crater was so 
obscured by vapour that it was impossible to bring out the details. The chasm, 
may have been formed when the side of the crater fell out on Nov. 3rd, 1905. 
