REPORT ON THE ANTIPATHARIA. 59 



the hypothesis one step further, it will be seen that the Antipathidce would thus need to 

 be regarded as very primitive forms, which have not lost their bilateral symmetry, but 

 which, in other respects, have gradually become specialised in one direction leading to 

 dimorphism. Edwardsia, Cerianthus, Zoanthus, Madracis, &c, would also represent 

 types in which the bilateral symmetry of parts is preserved, and the fact that Edwardsia 

 is generally regarded as a very primitive form tends further to support this view. 



The general plan of development may be stated as follows : — The mesenteries have 

 a radiate arrangement in forms with a round storuodseum; this arrangement becomes 

 bilateral by an elongation of the stomodseum in one axis, the sagittal. In this case the 

 anterior and posterior pairs (directives) come to consist of adjoining mesenteries, whilst 

 the intermediate pairs consist of opposite mesenteries. New mesenteries are added 

 between any or all of the lateral pairs, the space between two existing mesenteries being 

 divided in two on the formation of a new one. So long as the folds of the body-wall 

 give rise to only one mesentery each, the simple bilateral arrrangement of parts is 

 retained, as in Cerianthidse. In case the mesenterial rudiments give rise (after the 

 formation of the first twelve mesenteries) to two mesenteries instead of one, the 

 Hexactinian type is reached. In certain Madreporaria (e.g., Lophohelia, Mussa, and 

 Euphyllia) the radiate arrangement appears never to be lost. 



The different position of the retractor muscles in Alcyonaria and Edwardsia requires 

 explanation ; their intraseptal situation in the lateral pairs of Hexactinise is more easily 

 understood. The fact that in Antipathidse, Cerianthidse, and Zoanthiche the septal 

 musculature is more or less rudimentary may indicate that the special differentiation in 

 other types is of later origin. 



A further discussion of the subject must be deferred until I have been enabled to 

 study the whole of the material at my disposal. 



Complete and Incomplete Mesenteries. 



A comparison of the relations of the mesenteries in Leiopathes with the structure of a 

 typical Actinian will show an important point of divergence. In Leiopathes and other 

 Antipathidse, in which the number of mesenteries in the oral cone is greater than in the 

 lower section of the ccelenteron, the following points have an important bearing on their 

 origin. The mesenteries, which have for convenience been termed " secondary," are those 

 which do not reach the lower section of the ccelenteron, and which in certain genera appear 

 not to be developed at all. The behaviour of these mesenteries is most interesting. In 

 the upper portion of the oral cone they constitute short partitions, stretching from the 

 stomodseum to the outer wall. A little lower down they lose their connection with the 

 oral cone, and persist for some time as mesoglceal processes, clothed with entoderm, which 

 project from the ccelenteric surface of the stomodaeum. The mesogloea of the projection 



