96 Proceedings of the Royal Physical Society. 



successive whorls, and have no months. They surround 

 the laxger zooids in definite systems, and perform all the 

 functions of prehension, supplying the latter with food. The 

 structure of the various zooids of the colony is such as to 

 leave no doubt as to the fact that Millepora is not an 

 Actinozoan at all, but that the genus is referable to the 

 Hydrozoa, in which class it constitutes the type of the new 

 order Hydrocorallince. Millepora is a Tertiary and living 

 genus, but the Cretaceous Porosjphcera, Steinm., appears to 

 be closely related to it, and therefore also to belong to the 

 Hydrocorallince. The Tertiary Axopora, E. and H., with its 

 fasciculate columella, may also belong here, but its affinities 

 are not quite clear. 



II. PociLLOPORiDiE. — The type of this group is the Eecent 

 Pocillopora, Lam., in which there is a dendroid or foliaceous 

 corallum, composed of numerous tubular corallites, sur- 

 rounded by an imperforate compact coenenchyma, provided 

 with twelve to twenty-four septa (sometimes obsolete), and 

 intersected by complete tabulse. Professor Verrill, as before 

 stated, has shown that the polypes in Pocillojpora are pro- 

 vided with twelve nearly equal tentacles, and are in all 

 respects identical with the polypes of the ordinary Zoan- 

 tharia aporosa. The genus must therefore be retained 

 among the true " corals " in the neighbourhood of the Ocu- 

 linidce. From the close general resemblance between the 

 corallum of the recent Seriatopora and that of Pocillopora, 

 we may conclude that the former is also one of the Zoan- 

 tharia aporosa, so that we may abolish the family of the 

 Seriatoporidce, as understood by Milne-Edwards and Haime. 



The genus Pocillopora is Eecent and Kainozoic, and it is 

 questionable if any fossil forms of Seriatopora are known. It 

 may, at any rate, be taken as certain that the alleged forms 

 of Seriatopora in the Pala30zoic deposits will be ultimately 

 found to have different affinities. The Palceozoic genera, 

 Trachypora, Dendropora, and Bhabdopora, usually associated 

 with Seriatopora, Lam., are, again, truly perforate corals, desti- 

 tute of a true coenenchyma, and referable to the Favositidce. 



III. Favositidce. — The corallum in this family is composed 

 typically of polygonal, more or less closely contiguous, coral- 



