264 POLYNESIAN RESEARCHES. 
the edges of the crater, had been wafted by the 
winds over the adjacent plain ; for we also found 
quantities of it at least seven miles distant from 
the large crater. We entered several small craters, 
that had been in vigorous action but a very short 
period before, marks of most recent fusion pre¬ 
senting themselves on every side. Their size and 
height were various., and many, which from the 
top had appeared insignificant as mole-hills, we 
now found twelve or twenty feet high. The out¬ 
side was composed of bright shining scoriaceous 
lava, heaped up in piles of most singular form. 
The lava on the inside was of a light or dark red 
colour, with a glazed surface, and in several 
places, where the heat had evidently been intense, 
we saw a deposit of small and beautifully white 
crystals. We also entered several covered chan¬ 
nels, or tunnels, through which the lava had flowed 
into the large abyss. They had been formed by 
the cooling of the lava on the sides and surface oi 
the stream, while it had continued to flow on un¬ 
derneath. As the size of the current diminished, 
it had left a hard crust of lava of unequal thick¬ 
ness over the top, supported by walls of the same 
material on each side. Their interior was beau¬ 
tiful beyond description. In many places they 
were ten or twelve feet high, and as many wide at 
the bottom. The roofs formed a regular arch, 
hung with red and brown stalactitic lava, in every 
diversified shape, while the floor appeared like one 
continued glassy stream. The winding of its cur¬ 
rent and the ripple of its surface were so entire, 
that it seemed as if, while in rapid motion, the 
stream of lava had suddenly stopped, and become 
indurated, even before the undulations of the sur~ 
face bad subsided. 
