MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 545 



color than the others. The polyps oa the sides of galls are also very diminutive, apparently 

 undergoing retrogression (tig. If). 



The form assumed by the tentacles is somewhat dependent upon the degree of expansion of 

 the polyp. When enlarged to their utmost they are digitiform and overhanging, occasionally 

 distinctly swollen at the apex; at other times, when not fully extruded, they are erect and 

 more subulate. 



The disk is very small and flat, and, as just mentioned, may look in almost an j' direction with 

 regard to the axis of the branch, according to the plane in which the aperture of the coi'allite is 

 placed. The mouth is circular or oval, without thickened lips and gonidial grooves; looked 

 at from above, the aperture often appears excentric, being situated nearer the axis of the branch, 

 the axial tentacle overhanging (lig. lil). 



The movements of the polyps are rather slow, and the tentacles rarely ever wave about like 

 those of anemones; in a colonj' one polyp may retract independentlj' of another. "White 

 mesenterial filaments may be extruded through the wall of any part of the polvp, but the 

 phenomenon does not readily take place. 



The full complement of twelve tentacles is reached at a very early stage in the development 

 of new polyps; even in the second or third bud visible below the axial polyp they can all be 

 recognized. In some cases, however, developing polyps occur with only eight tentacles; the 

 two abaxial lateral and middle lateral of the smaller series, on each side, are the last to be 

 developed. 



Slight variations may be noted as regards the polyps of the different forms of colonies. In 

 general the poljqjs of palmata are somewhat smaller than those of proUfera or cervicornis, but 

 they exhibit exactly the same relations in regard to the size of the tentacles. In the older parts 

 of all colonies the anterior tentacle is scarcely larger than the others, the polyps attaining a more 

 approximate radial sj'mmetry. Usually the tentacles seem more pointed in jMlinata, digitiform 

 examples not often occurring, but they ma}" assume this form when fulh' extended. 



The corallites are often larger and more tubular in jxth/tata than in the other two/w/vwa', and 

 the disk of the polyps is directed toward the free growing edge. When fully expanded only 

 the disk and tentacles extend beyond the corallum, but practically no free portion of the column 

 wall can be seen except in the apical polyps. The disk is circular or slightly oval; the mouth 

 very small and circular or oval. 



Immersed polyps are usually very numerous on palmate colonies, especially on the upper 

 surface of a colony, and away from the free growing regions. They are somewhat smaller than 

 the others, but all gradations can be traced toward the fully developed individuals. When alive 

 the immersed polyps very rarely extend bej^ond the corallum, and the tips of the tentacles only 

 are visible. Their smallness did not allow of any difi'erences in the size of the tentacles being 

 determined; in some examples less than twelve were present, while in others a greater number 

 were found. The scattered polyps on the under surface of palmate colonies are usually smaller 

 than those on the ujjper surface. Doul)le polyps with two oral apertures sometimes occur (fig. 1*'). 



The color varies but little in all the Jamaican Madrepores. Colonies as a whole are lighter 

 or darker shades of brown, sometimes becoming green, yellow, or orange. Different regions, 

 however, vary in intensity, some l)eing quite colorless and others darker than the general 

 surface. The coloration is wholly due to the presence in the endoderm of S3'mbiotic yellow cells 

 or zooxanthella"; where these arc few in number the coloration is faint, while where they are 

 entire!}' aVisent the area is altogether colorless, the white corallum showing through the thin, 

 transparent tissues. The tissues directly along the edge of the striae and apex of the echinulations 

 are devoid of color, there being over these, as shown in fig. 7, only a very thin endoderm without 

 zooxanthella". In the intervening spaces the yellow cells are abundant. The tissues at the apical 

 region of the corallites are usually colorless, and microscopic examination reveals that, though 

 the endoderm is well developed, zooxanthella? are absent (fig. S); also, as a rule, the axial polyps in 

 palmate colonies are colorless. The distal region of the jwlyps on the upper surface may be 

 colorless, while the under polyps are uniformly brown. 



