with dark streaks.  Mount Astrolabe is apparently of trap formation, as I have already stated.  Some conical hills scattered along the coast may possibly be of volcanic origin, especially one of that form rising to the height of 645 feet from the lowland behind Redscar Head.  It is in this neighbourhood also that we find the upraised calcareous rocks of modern date exhibited by the Pariwara Islands and the neighbouring headland, with which they were probably once continuous; near this, too, the barrier reef of the coast ceases at Low Island, which it encloses, although its line is continued under water, as a ridge of coral, as far as the South-west Cape, where the coral ends, unless the shoals apparently blocking up the channel south of Yule Island are of the same formation.

LOUISIADE ARCHIPELAGO.

Reference to the outline chart will enable the reader to follow me in some general remarks which did not properly enter into the narrative.  The Louisiade Archipelago, reduced to what I conceive to be its natural limits, includes that extensive group of islands comprised between the parallels of 10 degrees 40 minutes and 11 degrees 40 minutes South latitude, and the meridians of 151 degrees and 154 degrees 30 minutes East longitude.  About eighty are already known, and probably many others remain yet to be discovered in the north-west, a large space there being as yet a blank upon the chart.  All the islands of the group, with the exception of the low ones of coral formation to the westward, appear to be inhabited, but probably nowhere very densely, judging from the comparatively small number of natives which we saw, and the circumstance of the patches of cultivation being small and scattered, while the greater part of the large islands is either covered with dense forest, or exhibits extensive grassy tracts with lines and clumps of trees.  Such of the islands as were examined consisted of mica slate, the line of direction of the beds of which is nearly the same as that of the Archipelago itself, and the physical appearance of the other islands leads me to believe that the same rock prevails there also.

CORAL REEFS OF THE LOUISIADE ARCHIPELAGO DESCRIBED.

One of the most remarkable features connected with the Louisiade Archipelago is the manner in which its shores are protected by the coral reefs which have frequently been alluded to above.  The principal of these are good examples of that kind distinguished by the name of barrier reefs.  Rossel Reef has already been described, and the only other large one of this description which we saw more than a portion of, is that partially encircling South-east Island at a variable distance from the land, then passing to the westward as far as longitude 152 degrees 40 minutes, where it ceases to show itself above water; thence, however, the edge of a bank of soundings (represented on the chart by a dotted line) which is suddenly met with in coming from the deep blue unfathomed water