528 PEOF. GILBERT C. BOUENE ON EIVE 



The micromesenteries appear singly, and usually in the following order 

 on each side o£ the actinopharynx. First, a single micromesentery in the 

 lateral meoacoele ; then a single micromesentery successively in the dorso- 

 lateral and the ventro-lateral megacoele. In some cases, however (E. 

 claparedif), the micromesentery of the ventro-lateral megacoele is formed 

 before that of the dor so-lateral. 



Usuallv in the lateral and ventro-lateral megacoeles the more dorsal of the 

 two microcoeles formed by the first micromesentery bears the primary 

 tentacle, the more ventral the secondary tentacle ; but the opposite is the 

 case in E. claparedii and probably in E. maminillata. In the dorso-lateral 

 megacoele the primary tentacle is always in the ventral of the first two 

 microcoeles, the secondary tentacle in the dorsal. 



Succeedino- micromesenteries are formed singly and always on the side of 

 the primary tentacle furthest from the last-formed secondary tentacle. As 

 the microcoeles and the tentacles into which they are prolonged are always 

 formed first on one side, then on the other side of the primary tentacle, the 

 latter always occupies a central or subcentral position in the megacoele. 



The rate of growth is usually greatest in the dorso-lateral, next greatest in 

 the ventro-lateral, and least in the lateral megacoele ; but in E. claparedii 

 the rate of o-rowth in the ventro-lateral megacoele seems to outstrip that 

 in the dorso-lateral, and in some examples of E. carnea growth is most rapid 

 in the lateral megacoeles. 



It is clear that, after the stage with eight mesenteries is reached, the 

 sequence of mesenterial development in the Edwardsia? difPers altogether 

 from that of the ITexactinians, and, indeed, from that of any known 

 Actinian. The characteristic feature is that any two micromesenteries 

 of the same aoe in any given megacoele constitute a conpJe — that is to say, 

 they arise as singles on opposite sides of the actinopharynx, and not in 

 vairs in the exocoeles as in Hexactinians and Zoanthids. Edwardsia 

 therefore retains the bilateral mode of growth throughout its existence, 

 and exhibits no trace of the biradiality characteristic of the second stage 

 of growth of the Hexactinians. In this respect it shows some analogy 

 with the Cerianthidfe, but differs altogether from the latter in the fact 

 that both the ventral and dorsal directive megacoeles remain undivided 

 throughout life, and increase of growth takes place on either side of the 

 two directive chambers, and not in one of them. On the other hand, 

 the persistence of the two directive chambers in an undivided state is a 

 feature which the Edwardsise share in common with the Hexactinise and 

 the Zoanthese, and points to a fairly close alliance between the three groups. 

 There is, further, some analogy in the mode of growth of Edwardsise 

 and Zoanthe^, in that in both the addition of mesenteries after a certain 

 stao-e proceeds more rapidly in certain megacoeles ; but the Zoanthe^ are 



