Page:A sketch of the physical structure of Australia.djvu/44

32, must yet historically have been of ancient date, as no traces of any crater are apparent. From the occurrence of pebbles of coral limestone, they are almost certainly of subsequent origin to the commencement of the coral reefs here, but may yet date back into some tertiary period.

The inspection of the "Chart of the Great Barrier reef and Torres Straits" will suggest to us some speculations worthy of note. First of all, the north and south central band of Torres Straits, or submarine continuation of the eastern chain, forms the western boundary of the Great Barrier reefs, just as much as does the dry land to the southward. No true independent coral reefs infringe on this band though they come close up to it. There are no wide spread coral reefs round its islets as there are round Darnley and Murray Islands, but only small fringing reefs clinging to the rocks. From this fact results two questions, 1. Are the "reefs seen by the Hormuzeer and Chesterfield" true coral reefs, or are they only banks of mud and sand with, perhaps, coral settled on them? 2. What are the islands called Duncan's Archipelago composed of? If not coral (and I believe they are not) good harbours with water 8 or 10 fathoms deep will probably be found among them.

The next point is the sudden ending of the Great Barrier reefs to the northward, and their reluctance