UEP"O LU: 317 
of the basalt is occasionally divided by irregular fractures into small 
polygonal areas, six to twenty-four inches across, and these areas, 
owing to the filtration of some cementing material into the fissures, 
are separated by double walls, like the fissured sandstones of Aus- 
tralia. ‘The wear of the surface has left the harder walls prominent, 
and the appearance is much like that of the sandstone referred to. 
The infiltrating fluid may have contained silica. 
Stratification.—In this place we make a few remarks only on the 
older rocks of the central district, reserving many facts respecting the 
more recent basaltic lavas and tufas, till we have described the several 
craters of the island. On the shores, where alone the rocks are ex- 
posed to view, the basalt occurs in a series of layers, which appear 
to have been formed by successive flowings of the melted rock. The 
layers average ten feet in thickness, and are partially separated by 
small rugged caverns and blow-holes. The layers are nearly hori- 
zontal, or have a very gradual dip outward, not exceeding five or 
six degrees. ‘These are the ordinary characters of the rock between 
Laulii and Tiavea. Just east of Laulii the cliff is partly composed 
of a basaltic conglomerate. A layer ten feet thick intervenes between 
the two layers of basalt forming the cliff. ‘The conglomerate con- 
sists of rounded and ragged masses of basalt, cellular or compact, im- 
bedded in a fine earthy base, often of a reddish clayey aspect. At the 
rocky point, just west of Laulii, the same layer of conglomerate ex- 
tends down to the surface of the water. 
Dikes.—The only dikes observed among the older rocks of the 
island occur in the point east of Fangaloa Bay. There are two at this 
place ;—one follows a southeasterly direction, and dips 80° to the 
southward and eastward: it is about twenty inches wide. ‘The other 
is two and a half feet wide, and is mostly vertical: it follows nearly 
the same course with the preceding. ‘The small number noticed by 
us, 1s no evidence that the number may not be large: the rocks are 
rarely exposed for observation on this thickly-wooded island. 
3. EXTINCT CRATERS OF UPOLU. 
We describe separately the facts that have been collected respecting 
the eastern and western volcanic districts of Upolu. 
Western District.—This district includes one half the island of 
Upolu. The principal craters are situated along the summit of the 
80 
