116 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY. 



iuto more or less rectangular blocks with rounded corners, often left 

 stranded almost like boulders on the summit of the ledge from which 

 they have originated. 



Retracing our steps, we may now take up the description of the great 

 coral reef fiats which we examined between the Hope Islands and Lizard 

 Island. 1 About five miles north of the Hope Islands we come upon b 

 Eeef, situated off the northwest extremity of Endeavour Reef (Plate V.). 

 It is a long flat sandy islet, with a conical heap of coral debris at the 

 southern extremity of the reef flat adjoining it ; at the northern end of 

 the flat there are a few negro heads, c and d Reefs are only sand 

 patches (Plates XXXII.-XXXIV.). 



For nearly an hour we steamed slowly along the western edge of 

 Cairns Reef, nothing but one mass of negro heads of all sizes, form- 

 ing a belt along the western edge of the reef flat (Plates XIII., XIV., 

 XV.). The western edge of this reef flat shows perhaps better than any 

 other flat we have examined the belt of dead corals on its surface and 

 the remnants of the old elevated reef on its lee side (Plates XIII., XIV.). 

 On the slopes extending to the main channel, or to the channels sepa- 

 rating Cairns Reef flat from the nearest flats, corals were growing abun- 

 dantly and in large masses and patches close to the upper edge of the 

 flat, rising from six to seven fathoms of water. At less than a cable's 

 length the bottom specimens brought up on the lead were already discol- 

 ored by admixtures of littoral deposits. On e Reef (South Turtle Reef) 

 we found one to two feet of water ; a third of a mile off the northwest 

 side of the centre of the reef we anchored in ten fathoms, muddy bottom. 



Coral heads begin to appear on the slope of the reef in from four to 

 five fathoms ; in deeper water, the bottom along the slope of the reef is 

 clean coral sand. The slope of the reef is covered with isolated patches 



1 On Plate XXX. Kent has represented the aspect of the surface of one of the 

 reef flats which, as he justly remarks, is in very striking contrast to those-coral flats 

 which are only rarely exposed at the lowest tides, and on which corals still flourish 

 in great profusion, and which form the suhjects of the many reef views illustrating 

 the "Great Barrier Reef." They are the only Plates which have been published 

 giving an idea of the appearance of a coral reef. Previous writers had only ex- 

 amined as far as they could be observed from the surface the upper belt of growing 

 corals in from seven to ten fathoms and upward, and from their descriptions 

 alone, no matter how vivid, it was impossible to give to the reader who had not 

 seen a coral reef an adequate conception of what a coral reef really looks like. 

 The illustrations of Haeckel in the "Arabische Korallen," while giving an ex- 

 cellent idea of the brilliancy and great variety of coloring in the masses of a coral 

 reef (Plate 3, Arab. Korallen, 1876), are of course not to be compared in their scope 

 with the photographs of Kent. 



