172 POLYNESIAN RESEARCHES. 
between the ancient rocks and more recently 
formed lava, was from six to twelve feet. On one 
side, the lava rose perpendicular and smooth, 
shewing distinctly the different and variously 
coloured masses of ancient lava of which it was 
composed ; some of a bright scarlet, others brown 
and purple. The whole pile appeared to have 
undergone, since its formation, the effects of violent 
heat. The cracks and hollows, horizontally be¬ 
tween the different strata, or obliquely through 
them, were filled with lava of a florid red colour, 
and much less porous than the general mass. 
This last kind of lava must have been brought to a 
state of most perfect liquefaction, as it had filled 
up every crevice that was more than half an inch 
wide. It appeared highly glazed, and in some 
places we could discover small round pebbles, from 
the size of a hazel-nut to that of a hen’s egg, of 
the same colour, and having the same vitreous 
covering, yet seeming to have remained solid, 
while the liquid lava, with which they were mixed, 
had been forced by subterranean fire into all the 
fissures of the ancient rock. 
The pile on the other side, formed by the drip- 
ing of the liquid lava from the upper edge of the 
rocks, presented a striking contrast, but not a less 
interesting sight. It was generally of a dark 
purple or jet black colour, glittering in the rays 
of the sun, as if glazed over with a beautiful vitre¬ 
ous varnish. 
On breaking off any fragments, we found them 
very porous, and considerably lighter than the 
ancient lava on the other side. Its varied forms 
baffled description, and were equal to the concep¬ 
tions of the most fertile imagination. The archway 
thus formed continued for about half a mile, occa- 
