384 Dr. J. E. Duerden on the 



mesenteries, a single septum within each entocoele and each 

 exocoele. An order (primary, secondary, tertiary, &c.) of 

 entosepta corresponds with each order of mesenteries, while 

 the members of the last cycle of septa are all exosepta and 

 correspond in number with tlie total number of entosepta. In 

 Acropora the exosepta are frequently wanting. 



The tentacles also show a close correspondence with the 

 mesenteries and septa ; a single tentacle arises from each 

 mesenterial chamber, and corresponds with a septum below. 



The results above summarized are in harmony with those 

 of von Koch, Fowler, Bourne, and von Heider upon the 

 morphology of other coral polyps whose asexual reproduction 

 is by budding. There seems every reason for the assertion 

 that wherever coral polyps arise by gemmation they conform 

 to the hexameral cyclical plan as regards their mesenteries, 

 septa, and tentacles; likewise two pairs of directives which 

 denote the principal axis of the polyp are always present. 



The investigations of the late Lacaze-Duthiers (1873, 

 1897), von Koch (1897), and H. V. Wilson (1888) have 

 made us acquainted with the early larval development of 

 several corals. At the beginning the organs usually arise in 

 bilateral sequence, but in the end are found to exhibit a 

 hexameral plan, and the six primary pairs of mesenteries 

 include two pairs of directives. I have also followed the 

 development of larval polyps (that is, polyps originating 

 directly from sexually produced larvae as compared with bud- 

 ])olyps which arise asexually) of Siderastrcea radians (Pallas) 

 for four months, and the growth oi Favia fvagum (Esper) and 

 Manicina areolata (Linn.) less completely. These are like- 

 wise characterized by cyclical hexamerism. 



As gemmation usually takes place on the column wall, in 

 close association with the mesenteries of the adult, it has not 

 been possible to determine in many cases whether any of the 

 organs of the adult are included in the formation of the bud, 

 or whether the mesenterial pairs arise altogether as new 

 formations from the polypal wall of the parent. The latter 

 is certainly the case in buds of Acropora^ as I have sliown in 

 the paper preceding this (1902), and the very early stage at 

 which the polypal organs are found wholly separated in other 

 forms would seem to imply a similar origin elsewhere. 



Investigations have also been carried out upon the early 

 development of buds, more particularly in Acrojwra, Clado- 

 cora, and Solenastrcea, and as regards the order of appear- 

 ance and ultimate arrangement of the organs the sequence 

 followed is in close conformity with that of the larval polyps 

 whose history is known. 



