412 .MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 



however, the ealieinul wall extends peripherally far upward within the cavity of the polyp, then 

 upon retraction of the latter the upper region of the cohiiiui becomes drawn within the ciilice, hut 

 the lower region, still with the mesenteries attached to it, remains outside. It is to this external 

 area of the column wall, often sharply marked off in retracted polyj)s, that the term '■ edge-zone" 

 is usually restricted. As a result of the same upgrowth of the calicinal wall, the coelenteron 

 likewise l)Pcomes separated into calicinal and pericalicinal or perithecal portions, each parti- 

 tioned into chambers by the ni(>senteries, and less so by the si^ptal and costa; (PI. Vll, fig. .'(4). 



Among corals like Portteii and Siderastriea, in which the calicinal wall is common to adjacent 

 polyi)s, and the septa are but little or not at all exsei't, there can possibly be no extrathecal, or 

 rather pericalicinal or perithecal, continuations of the tissues, and no edge-zone. 



Bv ■■ edge-zone" Dr. Ogilvie (p. 108) undei'stands " tliat the mesenteries of the interseptal 

 loculi are continued into the intercostal locidi," thus giving a more precise meaning to the term 

 than was done by von Heider. Among all the forms here studied, which are provided with a 

 perithecal continuation of the gastric cavity into intercostal loculi, Madrepora is the oidy one in 

 which the mesenteries also are not prolonged perithecally. In this genus the superficial covering 

 of the colony is continuous with the column wall of the polyp, and, as shown on PI. I, rig li, the 

 cadenteron is directly continuous over the edge of the theca with the superficial canals, liut there 

 is never any trace of external mesenteries. In the expanded polyp the mesenteries are seen to 

 pass from the extruded column wall directly into the calice, and the column wall below, unsup- 

 ported b}^ mesenteries, rests directly upon the skeletal echinulatious. 



The precise definition given to the edge-zone afl'ords Miss Ogilvie the opportunity of accom- 

 plishing the same for the somewhat loosely employed term '•(.Venosarc."' By this the aiithoi'ess 

 (p. 108), following Bourne (1888, p. 26), signifies " an extrathecal jiart into which the mesenteries 

 do not continue." Ccenosarc will, of course, consist of two distinct tissues: the skeletal cov(>ring 

 proper (base), and the superficial covering to the colony (column wall), the two separated 

 more or less !)}• a continuation of the gastro-ccelomic cavity." By universal acceptation, ''Cien- 

 enchvme" is the calcareous deposit originating from the ('(enosarc, and this is only laid down by 

 the skeletotrophic layer, the iimer of the two external tissues. According to the definition of 

 ctenosarc and ca'nench^'me just given, Madrnpora alone, among all the forms available foi- stiuly, 

 is characterized by these structures; that is to sa}', the only genus in which the perithecal walls of 

 the polyp are without mesenteries (PI. I). 



One of the most illustrative examples in this connection is Oculina. In all the definitions of 

 the genus one of the characteristics given is the presence of a solid cfenenchyme. Yet through- 

 out young colonies, and in the growing regions of others, the mesenteries are prolonged 

 perithecally, so as to extend as far as the spiral groove of separation of the superficial tissues of 

 the diflerent polyps, and the corresponding grooves on the skeleton are determinable throughout. 

 It is only in the older regions of large colonies that the me.senteries do not extend the whole 

 length of the column wall, and the skeletal surface then becomes perfectlj^ smooth, with an 

 a])sence of gi'ooves or co.stal ridges. Under such circumstances it becomes impossible to draw 

 any sharp line between edge-zone or column wall and c(enosarc. The latter is mereh' the extra- 

 calicular region of the polyp into which the mesenteries ar(> not prolonged. 



Bourne (p. :i6) states that "a common coenosarc is due to nothing more than a persistent 

 connection between the 'Randplatten' of adjacent polyps, and that the two structures are 

 homologous." This undoubtedly holds for some forms, e. g., Galaxea^ but the first portion of 

 the definition can scarceh' lie regarded as applicable to cases like 3fadrep&ra, where, by defini- 



"At the points where the cn'iioj^arc re.st.s upon the costal ridges or echinulations the two covering.s are combined, 

 and tlie skeleton is here overlaid only l.)y the superficial ectoderm, the mesoglaa, and the calicoblastic ectoderm 

 (Pis. I, II). The perithecal gastro-calomic cavity then becomes represented by canals, often reticular in char- 

 acter. Fowler (18SS, p. 7, PI. XXXII, figs. 2, :{) shows that in Ainphilidia rimim the direct adherence of the polypal 

 wall to tlie .skeleton may become very Ijroad, the canals being, as it were, pushed ajjart from one another and greatly 

 narrowed. 



Of the canals in Cccnopmmmia (iardini^r (1900, p. 361) observes: "The ccenosarcal canals in fact are simply 

 extrathecal portions of the cn-lc-ntera of the different polyps, which serve to connect their intrathecal or gastrovascular 

 portions." 



