MEMOIRS OF THE NATIONAL A.CADEMX OF SCIENCES. 531 



constitutes the ventral directives, and the fourth the dorsal directives, while the fifth and sixth 

 pairs, incomplete as yet, arise on the ventral aspect of the second and ti i--~ t pairs, respectively, the 

 fifth a little in advance of the sixth. (See diagrammatic figures on i>. 508.) 



A young polyp, settled mi the sa block of dead coral as a mature colony, affords the next 



stages required in the sequence the manner of appearance of the first metacnemes. The living 

 characters are shown in fig. 109, and a section through the decalcified polyp is diagrammatically 

 represented on p. 509. The protocnemes are in the same stage as in the previous figure — 

 the first four pairs are complete, hut the fifth and sixth are still incomplete. Within the dorsal 

 and middle exocoeles on each side a pair of mesenteries has appeared, the dorsal pairs being 

 better developed than the middle. In fig. 15/", showing the arrangement in another decalcified 

 young polyp, six pairs of metacnemes have appeared, completing the second cycle, and all the 

 members of the first cycle are united with the stomodseum. The six metacnemic pairs thus follow 

 a dorso-ventral, or antero-posterior, order in their appearance, but are now practically equal and 

 constitute the second cycle of mesenteries. 



Fi<;-. ley. p. .Mo. shows the manner of appearance of the first pairs of mesenteries which will 

 constitute the third cycle of twelve mesenteries, or second cycle of metacnemes, and it is at this 

 stage that fission is introduced (p. 511). 



A tangential vertical section through one of the larvae which had settled, but in which no 

 septa] formation had yet taken place, is represented by fig. 110, and the right half of the same 

 section, more highly magnified, is represented by fig. 111. The four complete mesenteries extend 

 from the base to the upper wall, and present a muscular development on each face, the fibers 

 being cut obliquely. The endoderm is still greatly thickened, especially basally, while the 

 superficial ectoderm has undergone but little change: zooxanthellse are altogether absent, 

 though present in abundance in the larvae. 



A great alteration has taken place in the basal ectoderm. It is no longer abroad columnar layer, 

 but is represented by little more than fragments, with h include a few nuclei and granular matter 

 which stains deeply. The mesogkea is likewise extremely narrow except mesenterially, where 

 it is much broadened. In these regions can be seen structures similar to the wedge-shaped, 

 striated, desmoidal processes characteristic of adult polyps, so that evidently these arise at a very- 

 early stage in the fixation of the larva: hints of the same processes also appear intermesenterially, 

 where the mesogkea is extremely narrow. The larva from which the sections were taken had 

 been adherent to a fragment of glass for over a week, and, though no septa were formed, it is 

 very probable that the basal plate had already been laid down, as this is one of the first parts 

 of the skeleton to appear. The skeletogenic ectoderm is in much the same condition as in 

 adult polyps, in regions where growth is not proceeding rapidly. 



The passage from the narrow basal ectoderm to the broad ectoderm of the column at the 

 margin of the section is abrupt. The cells around the indented vertical part at the right 

 e\tremit\ of tin'. Ill are somewhat modified compared with those beyond, and are probably 

 concerned in the formation of the epitheca. This is certainly the case in the slightly older 

 polyp of Manicina represented in tig. 137. 



YOUNG POLYPS OF MANICINA AEEOLATA. 

 I PI. XIX, dgs. L33-137.) 



To the very complete description of the early stages in the development of this species given 

 by Dr. H. V. Wilson, in 1888, I have nothing to add. and will therefore proceed to the point at 

 which Wilson's researches terminated, namely, the formation of the skeleton. The latest stage 

 reached in the growth of the Bahama specimens was one in which the twelve protocnemes 

 were present, only two pairs of which were connected with the stomodseum. The stage is 

 comparable with that represented in tig. 13-1, PI. XIX {of. Wilson's tig. 39), taken from a fixed 

 larva four days after extrusion from the parent colony. 



Out of many batches of larvse extruded from a small Jamaican colony only a few individuals 

 became fixed, and after several days these were reduced to two, which continued to live for nearly 

 three weeks, though under somewhat unfavorable conditions. The larvse were attached to frag- 



