Morphology of the Madreporaria. 99 



tlie larger polyj3S of Madrepora is in marked contrast with 

 the same phenomenon in Porites *, for in numerous examples 

 of the latter with tentacles and mesenteries in excess of 

 twelve, often from twenty to twenty-four, I have only come 

 upon one or two instances in which two oral apertures were 

 displayed on a single disk. The problem of mesenterial 

 increase in polyps of Madrepora is therefore more complex 

 than in Porites^ for from the beginning the new mesenteries 

 in Madrepora are associated with two stomodseal tubes instead 

 of only one. 



A transverse section of an ordinary radial polyp of Madre- 

 pora through the storaodajal region is represented in fig. 1. 

 Attention will be directed only to the mesenteries. Twelve 

 of these are present, arranged in six bilateral pairs, of which 

 two pairs, ///, IV, situated at opposite extremities, are 

 directives, and two pairs, F, F/, are incomplete — that is, are 

 not attached to the stomodseum. The retractor muscles are 

 sufficiently well developed on mesogloeal plaitings to admit of 

 the paired arrangement of the mesenteries being established. 

 In the figures the mesenteries are numbered 1 to VI, in accord- 

 ance with the order in which the pairs are usually found to 

 •appear in the course of embryonic and larval development of 

 both corals and anemones. Throughout their course the fifth 

 and sixth mesenterial pairs never become inserted on the 

 stomodoeum, but disappear shortly below the lower termination 

 of the stomoda^um, and never bear mesenterial filaments. In 

 Actinological literature the remaining eight complete mesen- 

 teries are known as the Edwardsian mesenteries ; and the 

 polyps of Madrepora, like those of Porites, are said to retain 

 in the adult condition the Edwardsia-stage of mesenterial 

 development, a stage passed through in the growth of most 

 Madreporaria and Actiniaria. 



Nearly all Madreporarian polyps pass beyond the stage 

 with only twelve mesenteries, usually by the addition of 

 alternating isocnemic pairs within the six primary exocoeles, 

 wliich in the end constitute one or more distinct cycles. 



* In tlie course of my former investigations on the order of mesenterial 

 increase in Porites (1900) no evidence was presented tliat the additions 

 in the end resulted in polypal fission. From an examination of material 

 collected later it appears that when six new pairs of mesenteries have 

 been formed the enlarged polyp, now possessing twenty-four mesenteries, 

 may undergo division iuto equal halves, at any rate so far as concerns the 

 oral aperture, stomodseum, and mesenteries. The mesenteries in each 

 daughter polyp are arranged exactly as in ordinary polyps arising as 

 buds. The introduction of fission, however, in no way niodities the 

 morphological significance of the pinnate order in which the increase of 

 mesenteries takes place. 



