196 THE MICHIGAN ACADEMY OF SCIENCE. 



With regard to the distribution in time of the Hexacoralhi, Zittel, in 

 his Text-Book of Palaeontology (Eng. Ed., page 96), states: "The group 

 Aporosa of the Hexacoralla appears to have originated from the Tetra- 

 coralla, and to form the direct continuation of their line. They begin as 

 the latter disappear; they develop a great variety of forms in the Trias, 

 and from the Mesozoic down to the present day they have continued to 

 play a leading part in the construction of coral-reefs. . . . The Per- 

 forata constitute a well-defined branch of the HcxacoraUa, whose ances- 

 try may perhaps be sought in the remarkable Archaeoci/athidae of the 

 Cambrian. The Eupsammidae and Poritidae occur sporadically in the 

 Silurian and Carboniferous, while it is not until the Trias that the Tham- 

 nastraeidac and Poritidae develop a large variety of forms; from the 

 Trias to the Tertiary, however, these genera continue to be imi)ortant reef- 

 builders." 



As would be expected there is much uncertainty concerning the true 

 character and relations of many of the earliest corals. This is especially so 

 as regards the relationships of the Hexacoralla to the Tetracoralla. 

 According to my researches the two groups are each characterized by six 

 equal septa separated by six interseptal chambers during their primary 

 (protoseptal) stage, but beyond this they diverge. Independently of 

 this, however, what seem to be true hexacorallids are found among the 

 earliest faunae known to us, perhaps in the Cambrian, but more cer- 

 tainly in the Silurian. If these ancient corals be true Hexacoralla their 

 polyps in all essential morphological characters must have resembled 

 the coral polyps of to-day, whether they belong to the perforate or im- 

 perforate group. Hexacoralla have been shown to be but skeleton-produc- 

 ing actinians, hence we may with good reason conclude that coral-form- 

 ing and skeletonless hexactinians have lived together side by side since 

 early Palaeozoic times. Further, the skeleton-producing forms must have 

 tiad soft-bodied ancestors. 



So far as fundamental polypal characteristics are concerned living 

 Madreporaria, notwithstanding their geological antiquity, are really a 

 very homogeneous group; variation and complexity have resulted from 

 the modification of parts which are of small morphological value, namely 

 the skeleton-producing tissues. To the predominance of asexual growth 

 by budding and fission is largely due the great diversity of form of 

 corals as we know them to-day. Soft-bodied actinians on the other hand 

 present a greater variety of polypal modifications. Some, like Rliodactis, 

 Ricordea, and Corynactis suggest the actual colonial types from which 

 corals may have arisen, or possibly they are forms which have relin- 

 quished the skeleton-producing habit. 



A similar line of argument can now be applied to the zoanthid actin- 

 ians, though here the conditions are not quite so conclusive. If we find 

 corals whose structure and septal plan suggest that they were formed by 

 polyps similar to those of zoanthids we have good grounds for assuming 

 that the skeletonless zoanthids have an ancestry like that of the corals. 



No living corals, so far as known, have a method of addition of mes- 

 enteries and septa which at all suggests the manner already mentioned 

 as characteristic of the Zoantheae. In all living niadreporarian corals 

 yet studied the mesenteries and septa follow the hexameral polycyclic 

 plan. However, in a paper recently published (A^in. Mag. Nat. Hist.. May, 

 1.902), I have given good reasons for concluding that living zoanthids are 



