504 ME. H. M. BEEFAED ON THE AEFINITIES OF 



formatiou of the walls of Favosites somewliat differently from 

 those of Alveopora. With regard to the former of these, I do 

 not think this difficulty can override the morphological evidence, 

 first, that Alveopora in its earliest known developmental stage is 

 very primitive, and hence would naturally have its nearest allies 

 in Palaeozoic times ; and second, that no essential structural 

 difference can be proved to exist between it and the Palaeozoic 

 Favosites. The choice, it seems to me, lies between direct genetic 

 relationship and redevelopment of an archaic type. 



"With regard to the second point, it in no way affects the main 

 conclusion that Alveopora finds its nearest allies among the 

 Palaeozoic Favositidae. The working out of this point would 

 mean the settling of the exact affi.nity between Alveopora and the 

 genus Favosites. My investigations on this 'difficult point are 

 very incomplete ; see, however, the Postscript to this paper. 



On tlie supposed S elation ship o/" Alveopora ivitJi the Poritidae. 



This relationship, as we have seen, has been very generally 

 assumed. Milne-Edwards and Haime declared that Alveopora 

 was in a manner but an exaggeration of Goniopora. The 

 advance which has been made in our knowledge of the coral 

 skeleton leads me to believe that this view is wholly untenable — 

 Alveopora and the Poritidce standing at opposite extremes of 

 the coral system. I can best explain by a series of diagrams the 

 conclusions at which I have arrived after four years of close 

 study of hundreds of different coralla with the express object of 

 unravelling the lines of development along which the different 

 genera have travelled. 



1. (Pig. 6.) The primitive skeleton of the coral polyp was an 

 epithecal cup. The evidence for this I would find in its appa- 

 rently universal appearance in the earliest stages of development, 

 and always with a uniformity of shape, texture, and function, in 

 striking contrast to the marvellous variety of skeletal super- 

 structures which have been secondarily built upon it. In some 

 Palaeozoic corals it appears to form the greater part of the 

 skeleton and was, as a rule, a prominent structure. The very 

 slight systematic imjDortance which has been hitherto attached to 

 it is due to the fact that in recent corals it is to a great extent 

 vesstigial, the vestiges being difficult to explain; and, further, it 

 only appears in its original capacity as a recognizably mural 

 structure in the very earliest stages of colony formation, very 



