490 MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 



Where the mesenteries are pierced and tixed by synapticuhi it is manifest that their retractile 

 power will bo lost; hence, only centrally and above will they still be able to extend and retract. 

 \\'ith this, perhaps, may be associated the fact that the polyps of l)oth Siderastrsea and Agaricia 

 are able to expand above the coralliim the least of any species coming- under observation. 

 Colonies of the former have been Icept in aquaria for months, and the polyps raise themselves 

 but little above the corallum. 



COLUMELLA. 



If the aboral surface of a coral polyp be examined after decalcification, either as a whole or 

 by means of sections, its middle aifords certain indications as to the nature of tlie columella. If 

 the calcareous tissue in the center of the corallite ))(> elevated and solid, it intrudes, as it were, 

 into the central basal part of the polyp, just as the septa intrude radially, and upon its 

 removal a conical chamber remains. The soft tissues of the intersoptal loculi are then wholly 

 distinct from one another, for a greater or less vertical distance from the base, and in transverse 

 sections the loculi are separated both laterally and centrally (figs. 55, 84). The calcareous 

 deposit in these cases extends radially from without the boundary of the polypal tissues to the 

 central deposit. If, on the other hand, the middle of the corallite be occupied l>y a spongy or 

 merely convoluted calcareous mass, irregular chambers remain, which represent so much of the 

 gastro-ctelomic cavity of the polyp, and serve to maintain the difl'ercnt interseptal loculi in 

 counnunication (fig. 142). 



The details represented by polypal sections, however, atford no evidence as to how the 

 columella has been produced, whether as an upgrowth of the floor of the corallite, or by fusion 

 of the free edges of the septa, or by both combined. This can best be ascertained from an 

 examination of the skeleton itself. 



ORDER OF APPEARANCE OF SEPTA. 

 PROTOSEPTA. 



In accordance with the division of the mesenteries into Trotocnemes and Metacnemes, the 

 septa appearing in association with them will be described as "Protosepta" and "Metasepta" 

 respectively. The protosepta will include the six primary septa appearing within the primary 

 entoco^les, and the six septa which appear within the primary exocoeles; all the succeeding 

 septa, arising within the entocu?les oi exocceles of the later mesenteries, will be metasepta. The 

 septa within the directive entocceles are known as " Directive Septa," and are axial in position. 

 Sometimes, as in Mad/rpom, these can l)e distinguished from the other protosepta by their 

 greater or less size, and thus give a marked bilateral character to the calices. 



Prof. H. de Lacaze-Duthiers (1873) was the first to observe the origin of the primary septa 

 in coral larvie, his researches lieing conducted upon the simple, perforate coral, Adroldcs 

 calijcidans. The description and figures indicate that twelve independent septa appeared simul- 

 taneously, one in each mesenterial chaniber, while the young polyp was at the protocnemic 

 stage of development. Apparently from the beginning each septum was constituted of three 

 pieces, arranged in a Y-shaped manner, the bifurcated end being peripheral. Prof. G. von Koch 

 (1882) subsequently obtained similar results for this species, at the same tim(> fully establishing 

 the ectodermal origin of the skeleton. In a paper, "Evolution du Polypier da Fiidjellimi 

 aiifJiophyllunu'" Lacaze-Duthiers (189-1) refers to young specimens of this species with only six 

 septa and six tentacles, and in later coralla, where twelve septa occur, the members of the second 

 cycle are much smaller than those of the first. G. von Koch (IS.SS) had previously demonstrated, 

 by means of serial sections of adult coralla, that Flahellum at its earliest stage is provided with 

 only six septa, and that an alternating cycle of six i\ppi^ars later. 



In one of his last works on corals, Lacaze-Duthiers (18!»T) makes further important contribu- 

 tions to our knowledge of the early stages in the development of the septa, so far as the process 

 can be followed by ol)servations on the living polyp and macerated coralla. A very complete 

 series are given, illustrating the appearance of the septa in BalunophylUa regia, as well as notes 



