MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 413 



tioii, there is iiu Kaiidpkitte, and one can hardly employ the tenu honiolooous in conueotion with 

 structures which are inerelj^ continuations of one another. 



Ccenosarc then, no more than Randplatte, is a pol^'pal structure "sui generis; " the two are 

 merely special regions of the coluuui wall and underlying skeletotrophic layer, in the latter case 

 provided with mesenterial continuations, and in the other devoid of them. Cojnenchyme likewise 

 is inseparable from the portion of the thecal wall laid down by the extrathecal layer of the 

 morphological basal disk, under whatever name it may be known. The terms have merely a 

 topographical, not a morphological, significance." In the following pages column wall will 

 generally include the whole of the external body wall, from the line at which it passes below into 

 the skeletogenic tissues to the outer margin of the tentacular zone above. 



Fowler, in his studies of various species of corals, has given much attention to the relation- 

 ship of the peripheral part of the column wall to the skeleton, particularly to the manner in 

 which it may be said to be supported. At first it appeared that in species without ctenenchyme 

 the cobunn wall was supported upon only the perithecal continuations of the mesenteries 

 (" perijiheral lamelln? ''), while in species with ctenenchyme the wall was directly supported 

 upon only echinulations of the skeleton. PI. VII, fig. o-t, and PI. XIX. fig. 132. will serve 

 as examples of the former, and PI. I, figs. 2-6, taken from Madrepora^ are instances of the 

 latter method. Later, however. Fowler found that no such rule could be maintained; that the 

 two methods of support — mesenterial and echinulato — might co-exist in the same form, e. g., 

 2f(icli'acls, Atiiphihelia. 



Where mesenterial continuations occur, the perithecal portion of the j)olypal cavity exists as 

 a series of simple vertical canals; but where mesenteries are absent, and the coluuui wall rests 

 directly upon skeletal ridges or echinulations {Madrepura), the cavity is usually l)roken up into a 

 complicated system of canals. 



FORM AND ANATOMY. 



Externally the column wall of coral ])olpys presents few structural modifications compared 

 with the same regian in the Actiniaria. There is an entire absence of the simple or complicated 

 columnar outgrowths often displayed in the latter group, and nothing comparable with a capitidum 

 or cj'Cle of acrorhagi has been oljserved, tho column always passing uninterruptedly into the 

 tentacles. Pi'actically the onty external distinction in this direction concerns the surface of the 

 column, whether smooth or verrucose. The latter condition is lirought about by the presence of 

 teeth or spines on the edges of the costse and septa. AVhere these occur the poljpal walls on 

 retraction come to rest upon them, and the areas over the projections become slightly raised 

 above the general surface, assuming a warty appearance; and even on fullest expansion, when 

 free from the corallum, the tubercle-liko character rarel}- entirely disappears. Sometimes the 

 verrucaj are indicated by a slight color distinction, and often give a coai'.se appearance to the 

 polyps. Where the edges of the costffi and septa are smooth, or only finely toothed, the surface 

 of the outer polypal tissues is likewise smooth. Histologicallj' the verruca' present no difl'erences 

 from the rest of the column wall, except that their constituent layers are generally thinner. 

 They are thus to be distinguished from the verrucw of Actiniif , which are slightly modified 

 evaginations of the wall, or more often take the form of vertical rows of suckers, with a 

 strongly marked histological modification. The verruca- in corals are characteristics dependent 

 upon the form of the skeleton, rather than a structural difierentiation of the soft tissues. 



Corresponding with the costa? and septa, the verruciv are arranged in vertical intermesente- 

 rial rows, larger and smaller rows often alternating, in agreement with the large and small skeletal 

 partitions. This is readily seen in species of Orhicella, Faci<t, and Jlmu'r/'/i/i, while in Jlxandrina 

 all the rows are equal. The veriaicte in anj- single row are somewhat irregular in size and height 



« The .study of the Comopsammia from Lifu has lead Gardiner (1900, p. 361) to define ccenosarc in such a way as 

 to make it much more emljracive tlian would either Bourne or Miss Ogilvie. Tims: "The Ccenosarc U that jmrt of the 

 ydijps in a colony irh ich lies outside but not above { i. e. in expondeil state ) the thecw. of the several corallites. The ' Randjilatte ' 

 of von Heider and von Koch, the 'edge-zone' of Miss Ogilvie, is tlien that part of the ccenosarc which lies over the 

 free portions of the corallites." 



Vol. S— No. 7 2 



