54 THE VOYAGE OF H.M.S. CHALLENGER. 



left correspond to an inter- and an intraseptal space, whilst those to the right correspond 

 to an intraseptal and two interseptal spaces. In order to explain the ultimate arrange- 

 ment on this plan, it would be necessary to suppose that a fusion takes place between 

 the two median transverse mesenteries, and that the interseptal space between them 

 is thus lost. There does not, however, appear to be any evidence in favour of this view. 

 It would also be difficult to understand the extremely rudimentary character of two pairs 

 out of the six, the tentacles corresponding to them being quite normal. Perhaps v. Koch's 

 theory of degeneration might suggest an explanation, but there does not at present appear 

 to be sufficient evidence in its favour. Again, the tentacles may be presumed to have 

 originally corresponded to intraseptal chambers, and those in the sagittal axis still continue 

 to do so on this view. It may further be supposed that by an imperfect development of 

 one member of a pair in other parts, the tentacles come to correspond below with a wider 

 section of the ccelenteron. An elongation of the body in the transverse axis should then 

 cause an increase in the size of the interseptal spaces, so that mesentery 2 would become 

 removed further from mesentery 1, and so on. Such, however, does not appear to be the 

 case. In Leiojxithes and several other forms, mesenteries 1 and 2, 5 and 6, 7 and 8, and 

 11 and 12 remain relatively close to one another so long as the "secondary" members 

 of each couplet are present (cf. fig. 2). The close relation between mesenteries 

 3-4 and 9-10 respectively is also very interesting. If one traces the course of these 

 mesenteries from above downwards, their mutual relationship is well seen. For instance, 

 number 3 approaches the transverse axis and becomes more important in proportion as 

 its fellow becomes reduced. 



An alternative explanation of the arrangement of mesenteries in Leiopathes may 

 now be mentioned, if only in order to exhaust the possible arrangements in pairs 

 consisting of adjacent mesenteries. I refer to the possibility of the mesenteries on 

 each side of the median transverse axis forming a pair. On this basis the mesenteries 

 numbered 1-2, 3-4, 5-6, &c, would form pairs, in which case each pair would consist 

 of a primary well-developed mesentery and a secondary imperfect one. There would 

 then be no pairs of "directives" corresponding to those of other Anthozoa, which 

 seems a great difficulty. Although I do not consider this a probable explanation of 

 the arrangement in Antipathidse, a consideration of its bearings brings out an interesting- 

 point. The pairs would be situated one on each side of the sagittal axis at each 

 extremity of the mouth, and a pair in the transverse axis on each side of the mouth. 

 The reduction in the number of mesenteries affects one member of every pah-, but 

 those are first to disappear which are situated in the transverse axis. A glance at 

 fig. 16 will show the effect of this arrangement ; the primary mesenteries are indicated in 

 thicker outline. The bilateral arrangement, on such an interpretation, is peculiar. Sup- 

 posing mesenteries 1 and 12 to occupy the "anterior" extremity, the mesenteries which are 

 incompletely developed are the posterior members of the first and second pairs on each 



