452 MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 



( f) The sixth pair becomes the ventral moiety of the ventro-lateral pairs of mesenteries on 

 eaeli side of the polyp. 



Among- the many larva' sectionized very few irregularities have been met with. Sometimes 

 one member of a pair will appear in advance of the other: in oik 1 larva only five complete mesen- 

 teries occurred, one member of the third pair having lagged behind. A young polyp settled for 

 some time still presented only three complete pairs along with three incomplete pairs. 



Although none of the other larva? which have been studied present so complete a series as 

 F.fragum . yet all the evidence from them goes to support the sequence just established. They each 

 represent Favia at one or other of its developmental phases. Newly-hatched larvse of Tsophyllia 

 dipsacea contain three pairs of mesenteries, the pairs differing greatly in the extent of their 

 development (Pis. XVII, XVIII). The middle pair again extends nearly the whole length of the 

 larva, and bears filaments which are strongly developed, especially at their lower extremity. Of 

 the two smaller pairs, one is very rudimentary, while the other extends a short distance below the 

 stomodseum. The stage very closely corresponds with that in the earliest available larva of 

 F.fragum (fig. 112). 



The non-extruded larva? of P<>ritt* clavaria also reveal a phase with three pairs of mesente- 

 ries, while the mature polyps never get beyond the mesenterial stage with four pairs complete 

 and two pairs incomplete — a stage represented by F. fragum at the. time of fixation. 



The larvae of Agarieia <t<i<<ric!t<x on extrusion already possess the six pairs of primary 

 mesenteiies. all extending nearly the full length of the larva, but only the first four pairs are 

 united with the stomoda?um (PI. XXV). Below the stomocheal region all the, twelve mesenteries 

 also bear well-developed mesenterial filaments. In this species, then, the stage reached by 

 the larva on hatching is directly comparable with that in Faviafragum, Mimicina areolata, and 

 Caryophyllia oyathus at or about the time of fixation, as well as with the adult polyps of Porites 

 and Madrepora. 



The earliest larva- of Sidt rastrsea radians sectionized reveal eight mesenteries arranged in 

 four bilateral pairs. The two lateral pairs, representing the first and second in the sequence, are 

 united with the stomodseum, while the dorsal and ventral axial pairs, representing the directives, 

 are free. Of the two directive pairs, the ventral pair (III) is slightly larger than the dorsal pair 

 (IV), and in larva a little, older the former becomes united with the. stomodiem, while the latter 

 is still free. In larva 1 of about this age the fifth and sixth pairs make their appearance, and the 

 dorsal directives uniting with the stomodseum the larva has reached the Fdwardsia-si&ge of 

 mesenterial development. At about this stage the larva undergoes fixation. Filaments do not 

 appear on any of the mesenteries until their connection with the stomodseum has been fully 

 established, but in most other species they are formed while the mesentery is still free. 



The order of appearance and subsequent development of the primary twelve mesenteries, 

 within the sexually produced larva? and young polyps of the Madreporaria, thus appears to be 

 very uniform, for no exception to the sequence first established by Wilson and von Koch has yet 

 occurred. In the extent of its development, and also in its strong mesenterial filaments, the first 

 pair to arise usually assumes predominance, and retains it until most of the other mesenteries 

 become fully established. While the second and third pairs are scarcely apparent the first pair 

 may have grown nearly the full length of the larva, each member tipped with the mesenterial 

 filament all the way. The second, third, and fourth pairs follow one another in regular succes- 

 sion, uniting with the stoinodamm in the order of their appearance. 



In most Actinological studies the fifth and sixth pairs are stated to arise simultaneously, and 

 H. V. Wilson observes tin 1 same for Manicina. Though such may often be the case, instances 

 occur in which one pair appears in advance of the other, and where the same pair becomes united 

 with the stomodseum before the other. Young polyps of a Sagartia from Beaufort, for which 

 I am indebted to Dr. C. Grave, all show in section that the bilateral pair between the first and 

 second Edwardsian pairs becomes inserted on the stomodseum in advance of the pair between the 

 first and third Edwardsian pairs. Another such instance occurs in the bud polyp of Chultx'orn 

 arhuscula, represented in transverse section on PI. VIII, fig. 60. The polyp was preserved in 

 a fully distended condition, the disk protruding in a cone-like manner above the zone of tentacles. 



