82 MADREPOEAEIA. 



they have in common, while, on the other hand, the texture of the coenenchyma of Astrceopora 

 is so peculiar and specialised that it can hardly have been the original form of the ccenen- 

 chyma of Turbinaria. We are not, therefore, justified in classing Astrceopora with Turhinaria ; 

 thoy must, for the present, be regarded as independent developments of the ccenenchymatous 

 corals, and the striking resemblances between certain forms in the two genera are to be 

 considered as fresh instances of what is known as convergence* 



GEOGRAPHICAL DISTRIBUTION. 



Several species of Astrceopora are recorded from the Red Sea, while specimens have been 

 collected in Mauritius and in the Indian Ocean. In the Torres Straits, on the Great Barrier 

 Eeef, and along the north coasts of Australia they occur somewhat scantily, while only a 

 single specimen was discovered by Mr. SavUle-Kent on the west coast of AustraKa. They 

 have further been discovered as far east as the Solomon Islands. None liave so far been 

 recorded from the Malay or Chinese Seas. The small number hitherto collected is probably 

 indicative of rarity, but it may also be partly due to their inconspicuous forms. 



GENERAL REVIEW OF THE VARIATIONS IN FORM AND STRUCTURE 

 ASSUMED BY THE GENUS ASTRJIOPORA. 



Variations in the Form of the Stock. — The genus may be primarily divided according to 

 its methods of growth. Three more or less distinct types are recognisable : — 



1. Explanatc Type. — The young corallum in its saucer-shaped epitheca expands laterally 

 in all directions, showing no very marked growth in thickness. 



Under this head would come primarily Bruggemann's type Astrceopora expansa, the 

 growth of which appears to be entii'ely explanate, with hardly any marked growth in 

 thickness. Further, Lamarck's AstroiojJora myriophthalma forms, according to the original 

 description, great plate-like coralla, and therefore belongs here. This latter species may also, 

 however, form great ii'regular encrusting mounds. But this is apparently not due to regular 

 growths in thickness, but, whether stimulated by parasites or not, to the throwing up of irregular 

 humps. However massive these mounds appear, they can always be distinguished from the 

 following massive types of growth by their irregular surfaces and shape, and by the character of 

 their edges (Plates XXV. and XXVI.). 



2. Pulvinate Type. — The corallum grows chiefly in thickness, by the rapid lengthening of 

 the calicles. As they shoot up, those at the sides bend outwards, making room for new calicles 

 to appear. The top is thus always expanding, while the lower portions die. Ultimately the 

 expanding top, continually forming new layers, hangs over on all sides like a great bolster or 

 cushion. In transverse section, the calicles appear at first sight to be grouped somewhat like 



* One glomerate Turbinarian from the Tizard Bank (and associated with T. stellidata), has 

 even a somewhat similar texture of the coenenchyma. 



