REPORT ON THE ANTIPATHARIA. 61 



again into the stoinodamni at a point only slightly below the upper aperture. From 

 the arrangement of the ectoderm cells lining this canal, Fowler thinks that a current 

 passes through it from the upper to the lower aperture, and that the modification is 

 probably connected with nutrition and the presence of symbiotic Algae. The remaining 

 six mesenteries are of the usual structure, and have no ectodermal canals. They cease at 

 the base of the stomodseum. The six modified mesenteries have a longer course, but those 

 numbered 4 and 9 are longest, and bear reproductive organs. In type B none of the 

 mesenteries are modified, but all have the same relative length as in type A. Numbers 

 4 and 9 are, in this type also, the only ones bearing reproductive organs. Both types 

 appear to be reproductive, and both are digestive, but type A seems to be more digestive 

 than type B, and may indicate a partly specialised gastrozooid. On the other hand 

 type B is more reproductive than type A; ova were only observed in a single instance 

 in the latter type. Fowler compares the elongation of mesenteries 4 and 9 to the 

 transverse mesenteries of Antipathidse and to the elongate ones in Alcyonaria. In 

 Madrepora aspera none of the mesenteries are specialised. The dimorphism in 

 Madrepora durvillei probably indicates a partial specialisation of certain zooids into 

 gastrozooid s and others into gonozooids, but the specialisation is not complete in either 

 case. Its tendency is evidently in the same direction as the dimorphism of the 

 Schizopathinse, but in the latter group the specialisation is complete, and is brought 

 about in an entirely different manner. 



In the Schizopathinse the dimorphism consists in the formation of gastrozooids and 

 gonozooids. This differentiation is not apparently brought about by the specialisation of 

 separate individuals as in other cases, but by a division of one primitive zooid into three 

 portions, a central one containing the stomodaaum, and two lateral portions bearing the 

 reproductive organs. The mode in which this is accomplished appears to be connected 

 with an elongation in the transverse axis of the typical zooid of the Antipathidse, and 

 the leading steps in the process can be made out from a study of the various genera 

 already known. The reproductive organs are in all genera of Antipathidse confined to the 

 transverse pair of mesenteries, which also bear the fully-developed mesenterial filaments. 

 The sagittal mesenteries sometimes have rudimentary mesenterial filaments, but these 

 apparently never occur on the secondary ones. In zooids with a rounded outline an 

 inequality in the breadth of the primary mesenteries, from body-wall to stomodasum, is 

 connected, in such types as Cirripathes, with an elongation of the stomodseum in the 

 sagittal axis. The transverse sections of the ccelenteron are therefore larger, and contain 

 the reproductive organs as somewhat hemispherical bodies applied to each side of the 

 mesentery. In Antipathella the zooids are small, and in those arranged on the more 

 delicate portions of the axis, which form the great majority, there is a slight elongation 

 of the body axis in the direction of the skeletal axis — that is in the line of growth. By 

 this means the previously rounded zooid becomes transformed into an oblong one, and 



