450 MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 



The results on the nu'senterial arrangement in adult eonil polyps may be thus summarized: 

 The mesenteries in oonimiferous genera are regularly hexanierous, arranged in one. two. or 

 more alternating cycles, and two pairs of directives are present; tissiparous genera are devoid of 

 any hexaraeral mesenterial plan, are imperfectly multicyclic or acyclic, and without directives. 

 Only the members of the earlier cycles become inserted on the stomoiheum in the former, while 

 apparently anj' of the pairs mav become complete in the latttu-. 



DEVKLOPMENT OF PROTOCNEMES. 



From the time of the publication of the classic researches of Lacaze-Duthiers (1872-73) 

 onward, numerous embryological studies, upon both Actiniarian and Madreporarian polyps, 

 have demonstrated that the twelve primary mesenteries alw.aj's arise in bilateral jiairs. but in an 

 order which seems to v'ary somewhat in different species. The results of Lacaze-Duthiers were 

 ol)taiu(>d 1)V observation of the external appearances, apparently without confirmation by means 

 of sections, while the latter has been the method more usually followed in later investigations. 



In the Actinian larva^ studied by him, Lacaze-Duthiers found that the first mesenterial pair 

 divided the coelenteric cavity into two unequal compartments, known as dorsal and ventral. The 

 second pair appeared in the larger or dorsal of the two chambers, cutting off a middle chamber: 

 then within the primary smaller or ventral chamber the third pair was developed. According 

 to Lacaze-Duthiers, the fourth pair appeared between the lirst and second pairs; hut in most 

 subsequent researches, among which are those of H. V. Wilson on the coral Maniciiia areolata, 

 J. P. McMurrich on the Actinian Rliodactis mncd-Thompe, and G. von Koch on CaryophyUia 

 cyathus^ the fourth pair has been found to appear in the dorsal chamber beyond the second pair, 

 and its members become the dorsal or sulcular directives. The fifth and sixth pairs were found 

 to arise nearly simultaneously within the middle and ventro-lateral chambers on each side. 



According to the Hertwigs (1>S79), the fifth and sixth mesenterial pairs arise in Adamma 

 d'lapliana on opposite sides of the polyp between the first and second pairs. This has also been 

 confirmed by Boveri (1S89). In the light of subsequent results, such a condition must undoubt- 

 edly be looked upon as exceptional, having been met with in no other species, while the number 

 of forms in agreement with the relationships given above is continually increasing in the Acti- 

 niaria, and is the onlj' sequence yet met with in the jVIadreporaria. Appellof (1900), in connection 

 with his studies on the development of ZTrticina eram'icornls and Actinia equina, discusses at 

 some length the conclusions of Lacaze-Duthiers and later writers with regard to the mesenterial 

 sequence of the primary eight mesenteries. In contradistinction to the successive development 

 which Lacaze-Duthiers descril)es for A. equina, Appellof found that the first eight mesenteries 

 appeared for the most part simultaneously, and doubts the possibility of the order of appearance 

 beuig determined by external observation alone without the assistance of sections. Sometimes 

 only the strongly developed ventro-lateral pair of mesenteries would be visible from the outside, 

 while transverse sections would demonstrate four pairs. His results on llrficina showed consid 

 erable variability in the mesenterial sequence. Reviewing the statements of different Actinolog- 

 ical writers with regard to the appearance of the primary mesenteries, Appellof (p. 55) comes to 

 the conclusion: "Es ist wenigstens auf Basis des vorhandenen Materiales unmoglich eine 

 bestimmte Regel auszufinden." 



In comparison with the variable results obtained in the Actinite the protocnemic sequence in 

 the Madreporaria appears to be very uniform. 



Two most complete series of stages in the development of the protocnemes of Madreporarian 

 polyps are already known, thanks to the labors of Prof. H. V. Wilson and Prof. G. von Koch. 

 The former (1888) has traced their appearance in the West Indian coral, Maniciiia areolata, from 

 the stage in the larva with but one pair of mesenteries to the young polyp with three cycles of 

 mesenteries. His results as to the first c^ycle conform with those of Lacaze-Duthiers on various 

 Actinian types, the second and fourth pairs being transposed. 



G. von Koch (1897) also describes and figures the order of development in ( 'iiryajdiyllia 

 cyathus, from the stage with two pairs of mesenteries to the completion of the first cycle. In 

 this species the order of appearance, sulisequent developiU(>nt, and luiion with the stomodanuu 



