122 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY. 
not raised beyond low-water mark, while where the surface of some of 
the large heads is topped with dead coral or with worn masses of 
the larger species, we may feel reasonably sure that the patch once 
formed a part of an outlying patch of the elevated reef which has 
been саб down to its present height by the combined action of the 
sea and of atmospheric agencies. 
We could not fail to notice how rapidly the water became clearer as 
we steamed eastward toward the outer edge of the Barrier Reef. Both 
its color and purity were in great contrast to the almost turbid water 
met with in the channel between the mainland and the most westerly of 
the inner reef flats, where the silt from the numerous islands and the 
wash from the mainland is thrown in very large quantities into the sea. 
But while we could not fail to notice the great contrast in the purity 
of the water, it may be due to its distance from shore, and perhaps also to 
its greater depth allowing the silt to subside ; for the mud brought up at 
our anchorage,only a third of a mile from the inner edge of the outer 
reef flat, was not coral sand, as might have been expected, but mud 
nearly as dark as that found in the inner more westerly main channels. 
Returning to the inner roef patches of this part of the Great Darrier 
Reef, we find Lark Reef, which forms the southwestern edge of Lark 
Passage, a huge reef flat differing in no way from the inner reef flats 
already described, except in the comparative absence of negro heads. 
The patches on the north of the track to Lark Passage are compara- 
tively small reef patches. On Swinger Reef, a small sand key has been 
formed on it since the time of the last survey. A sand key will soon be 
formed on the northern edge of Turtle Key (Plate XXXIII.). 
On the east half of Marx’s Reef, negro heads occur well on the reef 
flat itself. Negro heads are also found on the reef flat to the west of 
Marx’s Reef. The only other reef flats we examined were those lying 
to the westward of the Lizard Islands (Plates X VI, XVIL, XXXIV.). 
Eagle Islet is a low coral gand islet thrown up from the decomposed 
beach rock found on the northwestern side of the islet. There is in the 
extension of the islet a line of sandbars along the edge of Eagle Reef. 
The summit of Eagle Islet is covered with a little low vegetation. As 
far as we examined Eagle Reef flat, it appears to be made up of 
decomposed beach rock sand (Plate XVIL). We could see but few 
patches of this on the interior of the flat, only here and there an occa- 
sional fragment covered with Algo. Near the edge of the Eagle Reef 
flat, in from two to three feet of water at low tide, small patches of 
corals began to make their appearance. The patches rapidly increased 
