MATTHAI— RECENT COLONIAL ASTR^ID^ 35 



pp. 535 — 541). According to him, "the young polyps of both gemreiiferous and fissiparous 

 genera are built upon exactly the same plan" (p. 539), viz., with the full complement of 

 primary mesenteries (" protocnemes ") which, in both cases, are formed as " bilateral pairs," 

 i.e. the two members of a pair arise, one on each side of the median axis, with the 

 " retractor muscle " turned towards the same aspect of the polyp, but that the succeeding 

 mesenteries (" metacnemes ") arise in two ways; in the first group (" Entocnemaria," 

 including only two genera Pontes and Madrepora) they arise in "bilateral pairs," as the 

 primaries, but only " within one or both of the directive entocoeles," whereas in the 

 second group (" Cyclonemaria," including the Astrajidse, Oculinidge, and Fungidye) they 

 arise as "unilateral pairs," i.e. the two members arise in the same exocoele, their retractor 

 muscles being vis-a-vis. The family Astrseidee is divided into two sub-families, Gemmantes 

 and Fissiparantes. 



Such a division of the Astrseidse is based upon the author's theories regarding 

 budding and fission, but from the above it is evident that the polyp symmetries are 

 not necessarily dependent upon those two processes. Whatever may be the factors 

 governing the symmetries, the presence or absence of two directive couples retains 

 a constant morphological value in all the species and genei-a I have studied, and hence 

 possesses an importance in classification over and above any physiological phenomena like 

 budding and fission. 



I have accordingly divided my genera into two groups, characterised (I) by the 

 presence of two couples of directive mesenteries and consequently possessing both the 

 bilateral and hexameral symmetries, and (II) by the absence of the directive couples, and 

 hence of the symmetries. 



Boveri (19) for Gyractis and Kwietniewski (88) for Thalassianthus regarded the 

 absence of directive mesenteries as of ordinal value, creating the Holactinise and 

 Thalassianthse for their reception. On the other hand, McMurrich (99) did not regard 

 their absence as having any phylogenetic significance, but viewed it as "a peculiarity which 

 is essentially individual, possibly rising secondarily in some cases to the value of a specific 

 or even a generic character — " (p. 121). 



From an examination of the sections of the coralla of the following genera I am led 

 to infer that in Group I the septa are not directly continuous from corallite to corallite, 

 the peritheca filling the inter-corallite spaces ; if any such septal continuity is seen, as in 

 Diploastrea heliopora, it appears to be due to a secondary rearrangement of the calcareous 

 elements of the peritheca. When two adjacent corallites are in contact, their walls are 

 still distinct, the costse of one fitting in between those of the other. On the other hand, 

 in Group II the septa of neighbouring corallites are usually continuous, the continuity 

 being primary. 



In all these genera the corallite-wall appears to be formed by the union of the outer 

 thickened parts of the septa, i.e. is a pseudotheca. In Galaxea a " eutheca," as described 

 by Ogilvie, is not seen in my sections* ; the wall in this genus is also composed of the 

 thickened peripheral parts of the septa, with the lines of suture quite distinct between 



* On the supposition of a true theca in Galaxea, Ogilvie separated this genus from the AstrEeidse and 

 placed it provisionally in a new family, Stylinidce (p. 162). 



5—2 



