482 MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 



structures which appear as processes of the mesoglcea. In sections the processes are striated 

 toward their free extremity, which in methyl blue and in carmine always stains much more deeply 

 than the remaining mesoglcea. Their function would seem to be to bind the soft tissues to the 

 corallum. Where the insertion of a mesentery on the column wall is seen in longitudinal section 

 the processes appear as represented in tig. 95, PI. XIII. 



The nature and origin of the mesoglceal processes has been specially studied by Bourne (1S99). 

 He shows that they are formed from special ectodermal cells which he terms desmocytes. These 

 take their origin from certain cells in the calicoblast layer, and become secondarily attached to the 

 mesoglcea. The pr-ocesses thus formed may be known as desmoidal processes, though Bourne 

 employs the term desmocyte for them, as well as for the cells by which they are produced. 



The desmoidal processes may occur at any part of the outer skeletotrophic tissues, but are 

 most numerous in areas along which the tissues may have to withstand, as it were, the strain 

 of any muscular activity of the polyp. The attachment of the mesenteries to the skeletotrophic 

 tissues represents such areas, and here desmoidal processes usually occur in numbers. 



Also, as shown in tig. 07, PI. IX, they are specially developed in colonies along the line of 

 separation of one polyp from those adjacent. It is obvious that at the point des. pr. the polyp 

 upon expansion will tend to raise the skeletal covering from its adherence to the edge of the 

 calice; hence to meet this the skeletotrophic tissues are provided with a special development 

 of desmoidal processes. However fully expanded a polyp may be, it is never able to detach 

 its basal wall from its adherence to the corallum. 



The whole manner of distribution of the desmoidal processes fully supports Fowler's (1899) 

 suggestion that the structures are special devices for maintaining the adherence of the polypal 

 tissues to the skeleton. Their purpose in many ways is comparable with that of the liga- 

 ments in the higher animals. They do not occur over the actively growing regions of polyps; 

 the skeletotrophic ectoderm is here a continuous epithelium (PI. II, tig. 8). 



ECTODERM OR CALICOBLAST LAYER. 



The basal ectoderm or calicoblast layer is of much importance in studies of the morphology 

 of the Madreporaria, seeing that by it is produced the entire skeleton, or coral as popularly 

 understood. The nature of the layer, and the mode of formation of the skeleton by it, have been 

 the subject of much controversy, with which the names of Milne Edwards and Haime. A. R. 

 von Heider, G. von Koch. Miss Ogilvie, and G. C. Bourne are associated. Bourne (1899) has 

 recently summarized the various views as to the structure and formation of the skeleton, and 

 has made a very thorough study of the. process as it takes place in different genera of Anthozoa. 

 With regard to the Madreporarian skeleton he finds, with von Koch, that the calicoblast layer is 

 everywhere a simple epithelium, the cells rounded, columnar, or fused together, and that the 

 calcareous matter is laid down wholly external to the polyp. He thus differs from von Heider 

 and Ogilvie, who concluded that the calicoblast ectoderm was a multilaminar layer, and that the 

 skeleton resulted from calcification within the cells. The calicoblasts described by von Heider 

 are shown to correspond with the desmocytes of Bourne, and are not concerned with the 

 secretion of the skeleton. 



The results from the present study fully confirm those of von Koch and Fowler as to the 

 unilaminar condition of the calicoblast layer and the ectoplastic formation of the skeleton. 

 1 siially the layer is only well developed within regions of active growth, as toward the 

 uppermost part of the corallites (PI. II, tig. S), or aborally. where dissepiments are in course 

 of formation (PI. X, tig. 73). Elsewhere the calicoblasts form an extremely flattened layer; 

 in Siderastrasa, however, the structure is the same practically throughout the whole of the 

 skeletal area (PI. XXIV). As a rule desmocytes are wanting where the calicoblasts are well 

 developed, but are plentiful where the cells are nonactive. especially along the line of attachment 

 of the mesenteries to the basal wall. 



In some instance- (Madrepora, tig. 16) the skeletogenic ectoderm shows distinct cell 

 limitations, as in ordinary columnar epithelium; but generally these are lost, and the contents 

 are arranged in a continuous manner and largely vacuolated. 



