1821.  August 1.

At half past five o’clock the next morning we were under sail but, the breeze being light, had only time to reach the anchorage under Greville Island in Halfway Bay, before the tide turned against us.  It was purposed to remain only during the flood; but, on examination, the place was found to be so well adapted for the purpose of procuring some lunar distances with the sun, to correspond with those taken last year at Careening Bay, that we determined upon seizing the opportunity; and as wood was abundant on the island and growing close to the shores, a party was formed to complete our holds with fuel, whilst Mr. Roe assisted me in taking observations upon a convenient station on the north point of the bay within Lammas Island, a small rocky islet covered with shrubs, and separated from the easternmost point of Greville Island by a very shoal and rocky channel.

During these occupations we examined Munster Water:  on our way to it we landed on the reef off the east end of the Midway Isles, which was found to be more extensive than had been suspected, and to embrace the group of small rocks, which at high-water only just show their summits above the water; at high-tide there is at least fifteen feet water over it, but being low-water when we landed, the reef was dry.  Upon it we found several varieties of coral, particularly Explanaria mesenterina, Lam.; Caryophylla fastigata, Lam.; and Porites subdigitata, Lam.:  the only shell that we observed upon the reef was a Delphinula laciniata, Lam.  (Turbo delphinus, Linn.).  After obtaining bearings from its extremity, as also from the summit of the outer dry rock, we landed upon a small verdant-looking grassy mound, the northernmost islet of the group; but we found the verdure of its appearance was caused only by the abundance of the spinifex, through which we had, as usual, much difficulty in travelling.  After procuring some bearings from its summit we re-embarked and pulled up Munster Water, supposing that it was connected with the strait at the back of Greville Island; but as the tide then flowing was running in a contrary direction to what was expected from the hypothesis we had formed, we began to suspect some other communication with the sea, and in this we were not deceived; for a narrow but a very deep strait opened suddenly to our view, at the bottom of the Water, through which some of the islands in the offing were recognised.  In pulling through we had kept close to the south shore, that we might not miss the communication with Hanover Bay, but notwithstanding all our care we passed by without noticing it, on account of the deceptious appearance of the land; indeed the strait which we discovered leading to sea was not seen until we were within two hundred yards of it, and would also have escaped our observation had not the channel been so direct that the sea horizon was exposed to our view.  At the bottom of this arm are two deep bays which were partially