142 ME. H. M. BEENAltl) ON EECEl^T POEITID^. 



but is most often well developed and with a fourth cycle indicated. 

 The lamellate character of the septa is more evident in the larger 

 septa of Goniopora than in the smaller septa of Porifes, the 

 perforations being about the same size. Thus the perforations 

 as such are of less account in Gioniopora than in Porites, and 

 the vertical section is more of an irregular reticulum than a 

 regular lattice-work, as it frequently is in Porites. The close 

 relationship between Porites and Goniopora suggested by Dana 

 has been universally accepted, Milne-Edwarda and Haime aban- 

 doning their former position as soon as possible after Dana's 

 work appeared. There has, however, been a tendency to limit 

 the genus too much to forms which have tall thin walls and 

 consequently deep calicles. As a matter of fact, the range of 

 variation is very great ; and the collection in the Natural History 

 Museum contains many new and beautiful forms. 



Admitting this genus, then, as a near ally of Porites, the much 

 greater size of its calicles raises an objection to our conclusions. 

 Por^Yes, by the small size of its calicles, might easily be accounted 

 for in the way above suggested as fixed young forms. But 

 how shall we explain the much larger size of the calicles of 

 many Goniopores ? ;■- ■ - 



It seems to me that these need not present any great diflSculty . 

 Passing over the possibility above suggested, that in these 

 Poritidae we m.ay have a group made iip of fixed young forma 

 of several different corals, whose separate ancestries it would 

 now be extremely difficult to unravel *, there need be no 

 difficulty in deducing the Goniopores from Porites directly; and 

 this seems, for the practical purposes of classification, the simplest 

 course to pursue, provided, however, we do not lose sight of the 

 above-mentioned possible polyphyletic origin. 



I propose, then, to regard the Goniopores as merely enlarged 

 Porites, a kind of giant race which retains tbe skeletal habit of 

 Porites. If once that habit became fixed, there is no reason 

 why further growth should not simply enlarge it without 

 necessarily running it into ancestral Madreporidan lines. 



In the present state of our knowledge, I regard anything like 

 certainty in these relationships as unattainable. "What I have 



* Here it is of gi-eat interest to note that Dana himself suggested that 

 Goniopora miglit occupy a position in the Caryophyllacea corresponding to that 

 which Porites occupies in the Madreporacea (Zoophytes, p. 407). 



