498 ME. K. M. BEENAKD ON THE AFFINITIES OF 



between them seems to be demanded in order to account for such 

 a remarkable family likeness. 



The problem was at this point taken up by me on commencing 

 the fourth volume of the British Museum ' Catalogue of the 

 Madreporaria,' which is to deal with the Poritidae. As was to 

 be expected, the new light whicb in recent years has been 

 thrown on the morphology of the coral skeleton by the works 

 of Bourne, Fowler, v. Heider, von Koch, and Miss Ogilvie, 

 enables one to carry the question a step or two nearer solution 

 than was possible to the earlier workers. Further, the unrivalled 

 coral collection in the Natural History Museum, for my intro- 

 duction to which some three to four years ago I am indebted 

 to our President, Dr. Giinther, afforded opportunities of extended 

 observations and comparisons which are all important in the 

 solving of questions of affinity. 



The net result of my investigation has been to divorce Alveo- 

 pora wholly from the Poritidse, and to re-state Dana's claim to 

 find its nearest allies in the Palaeozoic Pavositidse ; and the best 

 way in which I can make my argument clear will be to describe 

 the essential features of the genus Alveopora, and then to compare 

 them, on the one hand, with those of Favosites, and on the other 

 with those of the Poritidse. I here take the opportunity of 

 thanking my friends Prof. Jeffrey Bell and Dr. Gregory — the 

 former, who is however in no way responsible for the opinions 

 expressed, for much valuable assistance and advice ; the latter for 

 giving me the freest access to the valuable collections under his 

 charge, and for the interest he showed in the whole subject I 

 was proposing to discuss. 



The Genus Alveopora. 



Pigs. 1, 2, and 3 (PI. 33) show three stages in the early growth of 

 Alveopora. I was fortunate enough to discover them grouped 

 together on the under surface of the type specimen of Montipora 

 pilosa collected by Dr. Willey at the Loyalty Islands. They are 

 too young to specify, but may perhaps belong to the original 

 species of Quoy and Gaimard, A. viridis from New Ireland. 



Pig. 1 shows an almost cylindrical ejjithecal cup 1 mm. across, 

 with a few septal spines projecting from its walls rather deep 

 down. The wall of the original polyp is thus wholly formed of 

 epitheca. 



