592 MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 



AGARICIA FRAGIUS Dana. 

 (Pis. XXIV, XXV, figs. 1{>1-164.) 



Externa} cluiracterx. — Colonies form delicate, flattened, .MulxMrcular oi- irregular expansions, 

 attached to .some coral lilock hv a broad, irregular base. Young colonies may be wholly incrust- 

 ing, but later the thin peripheral regions liecome fi'ee. Typical examples are very regular in 

 form, the polyps arranged in incomplete concentric series. Where freedom of growth is not 

 permitted, the colony ma}' lie irregular in outline, and vertical exj)ansions may then arise from 

 its general surface, bearing polyps on both sides. The thickness of the central region varies 

 greatly, and a gradual thinning takes place toward the periphery, which is very delicate. In an 

 actively growing colony the periphery is a broad marginal zone without actual polyji.-^. and the 

 polypal tissues are continued on the under surface. 



The thecal ridges are arranged concentrically and I'adiateh', but the regularity is often 

 departed from; the concentric ridges are more pronounced than the radial, and some project 

 higher than others. Similarly with the radial ridges, some are nearly of the same height as the 

 concentric ridges, and may inclose two or more polyps of which the radial ridges remain lower. 

 In the retracted condition the central region of each polyp is deeply depressed within the calice, 

 so that each polypal area is distinctly separated from the others. The usual distance from one 

 concentric ridge to another is 3 nnn. and from one radial ridge to another 2 mm. 



As shown in the transverse section represented in tig. Iti4, the side of the polyp toward the 

 peripherv of the colony is more spreading than that toward the center, and thus the stomotheum is 

 not alwaj's in the middle of the disk. Owing to the arrangement of the thecal ridges in a roughly- 

 concentric and radial manner the form of the individual polyp becomes somewhat quadrangular. 



The edge-zone at the margin of the colony is very delicate, and closely adherent to the 

 corallum; sometimes it covers only a very limited peripheral portion of the under surface, the 

 remainder being hidden by various foreign growth.><; in other cases it may spread for some distance 

 over the surface of the foreign body to which the colony is adherent. 



The column wall of the individual polyp is very limited in extent, but i.s a little broade?- along 

 the concentric borders than on the lateral borders. The boundary l)etween the column wall of 

 one polyp and that of another is only approximately determinalile along the apex of the thecal 

 ridges; there is no dividing groove in the soft tissues limiting the individual polyps, such as is 

 found in most corals. 



The superficial polypal tissues are .smooth, and so thin as to allow the septo-costte to be seen 

 through. In retraction these give rise to somewhat prominent ridges on the column wall, those 

 of adjacent polyps corresponding and being continuous. The septal ridges are vi.sible from the 

 outside, and diti'erent orders are represented. In most places only alternately large and small 

 elevations are indicated, but elsewhere members less completely developed maj' denote later cycles, 

 or perhaps new septa in process of growth, which in time will attain the dimensions of the others. 

 The complete number of septal ridges on .seven polyps was found to be as follows: IS, 20. 22. 2-i, 

 26, 28, 30, numbers which possess no hexameral constancj'. 



The boundary between the column wall and disk also is not well defined, owing to the irreg- 

 ular arrangement, and. in some cases, apparent absence of the tentacles. During retraction no 

 overfolding of the wall takes place, so that the tentacles, disk, and mouth are always expo.sed. 

 On expansion of the polyps the superficial walls are raised ])utashort distance above the corallum. 

 and the column wall ])ecomes only approximately cylindrical in form, remaining attached along 

 the thecal edges. 



The tentacles are very rudimentary; indeed, in some living colonies they were indistin- 

 guishable even with the aid of lens, and such is often the case in preserved colonies. In other 

 instances the merest tubercular elevations over the larger septo-costas were the only indications; 

 none ever occur over the alternating small .septo-costfe, which on subsequent examination are 

 found to be exoccelic in position. The tentacles usually vary in number from ten to eighteen, 

 and are comparatively widely separated from one another. Where no tentacles ai'e api«irent 

 there is clearlv no line of demarcation to be established between the disk and the column wall of 



