THE AUSTRALIAN GREAT BARRIER REEF. 97 



land-locked throughout its entire extent by the three long narrow islands of Bribie, Moreton, and 

 Stradbroke, which, from north to south, form a natural mole or breakwater against the 

 incursions of the Pacific Ocean. The chief gap, that most extensively used for shipping, occurs 

 between the Moreton and Bribie islands ; it is some five or six miles wide, and is known as the 

 North Passage. The second-largest gap, distinguished as the South Passage, intervenes between 

 Moreton and Stradbroke islands. It is less than a mile wide, and is passable, for large vessels, 

 only under favourable conditions of the tide. The remaining gaps, between Bribie and the 

 mainland, in the north, and Stradbroke and the mainland, in the south, are narrow, shallow, 

 and available only as passages for the smallest craft. This categorical explanation of the 

 topographical features of Moreton Bay may prove of service when an intelligible interpretation 

 of its existing conditions of coral-growth is being sought. 



Mud Island, Moreton Bay, in whose vicinity the various corals just referred to are still 

 growing, or formerly grew, is situated at a point between Moreton Island and the mainland, close 

 to the centre of the bay ; no corals, either living or dead, occur to the south of it. Northwards, on 

 the other hand, in the vicinity of the township of Humpybong, the mainland beach for many miles 

 is made up of shells, sand, gravel, and a very considerable admixture of dead coral fragments. 

 Among the types identified, the species of Favia, and the two Madreporae previously reported 

 from Mud Island, constitute the most abundant varieties ; and, in the case of the Madreporae, the 

 branches thrown up by storms are often of considerable size. In addition to the foregoing, two 

 other genera, which were not found at Mud Island, are here represented. These include a Tur- 

 binaria, apparently identical with T. ciuerasccns, and a species of CyphastrEea. The original 

 growing-beds of the beach-stranded corals could not be precisely determined. They lay, as far 

 as could be ascertained, a mile or so off the coast. 



Two highly interesting and important questions arise in association with the facts just 

 related. In the first place, how comes it that, with these various typical reef-constructing 

 corals abundantly represented, in both the living and the dead conditions, no reefs or reef-rock 

 conglomerates are constructed ? Secondly, what causes or conditions have led to the destruction 

 of the coralla of the genus Madrepora in Moreton Bay, and this to such an extent that no living 

 specimens can now be found there ? The first of the questions raised is of special interest, since 

 upon its solution the interpretation of the most important phenomena of reef-construction 

 would appear to rest. The composition of ordinary reef-rock out of the consolidated debris 

 of dead and broken-down corals, and the local intermixture of the same with fragments of 

 broken shells and gravel, have been previously dealt with and illustrated in association with 

 the Photo-mezzotype plate No. XXXI I. The ingredients embodied in the specimens, which 

 are represented by Figs, i to 4, in the plate quoted, and were collected on the beach of Sweer's 

 Island, in the Gulf of Carpentaria, correspond in every essential detail with those of Humpy- 

 bong Beach, in Moreton Bay ; but, in this latter instance, they possess no element of coherence, 



