MATTHAI— RECENT COLONIAL ASTR^ID^ 



Subsequent examination revealed some nuclei and fibrous appearances, whereas Bourne 

 and Duerden did not observe any such structures in the coral-polyps they examined. 

 But owing to the uncertainty that still prevails over the true nature and origin of this 

 intermediate substance the term " mesoglsea " is used in this paper for descriptive pur- 

 poses without any of the significance that Bourne attached to it. 



Calicoblastic layer of ectoderm (PI. 1, fig. l, PI. 3, fig. 25 and PL 5, fig. 53). The 

 calicoblastic layer (von Heider) is discontinuous only where the soft parts are attached 

 to the corallum by means of the wedge-shaped processes described below. Its nuclei are 

 quite characteristic in being large oval sacs arranged in a single layer, in each of which 

 is a central darker spot, possibly the nucleolus. At the skeletal attachments of the 

 mesenteries the calicoblastic layer is considerably thickened and vacuolated in some 

 species (cf. Galaxea fascicularis, Linn.), its protoplasm appearing in transverse sections 

 as thin columnar strands. A structureless membrane covering the external surface of 

 the calicoblastic layer and separating it from the corallum, as described by Bourne, could 

 be made out, although torn in places. This membrane is similar in nature and appear- 

 ance to the structureless mesoglsea of the polyp. The calicoblastic layer is better de- 

 veloped in the edge-zone, coenosarc and the peristomial region of the body-wall, while 

 towards the base of the polyp it is attenuated. 



The soft parts ai'e attached to the corallum by means of wedge-shaped processes, 

 the latter being particularly conspicuous at the skeletal attachments of the mesenteries 

 (fig. 43). Later observations show that these processes are formed not in distinct cells 

 (Bourne's desmocytes) but by modification of the calicoblastic cytoplasm. They may be 

 formed anywhere in the calicoblastic ectoderm. Stages in their development have been 

 noticed, but these require further study. It is unlikely that the striae in these processes 

 mark the attachments of the muscular fibres of the mesenteries to the corallum as 

 Gardiner suggested for his species of Ccenopsammia, since they are found even at the 

 very base of the body-wall and at places other than at the skeletal attachments of the 

 mesenteries, where there are no traces of muscular fibres. 



Nematocysts have been sometimes observed in the calicoblastic layer. Duerden 

 (37, p. 32) also found " small oval nematocysts with a close spiral thread " scattered 

 throughout the calicoblastic ectoderm of Siderastrea radians. 



Dissepiments and polypal growth. Bourne, in his account of the anatomy of 

 Mussa corymbosa and Euphyllia glahrescens, thus described the formation of dissepi- 

 ments : ' ' There are probably periods of active coral secretion alternating with periods of 

 reproduction in these polyps. During the latter period the thin dissepiments are formed 

 by the basal tissues, whilst in the former period, the septa increase greatly in height, the 

 polyp is, as it were, moved higher up upon the stem, and deserts the old dissepiments 

 upon which it was resting. Then follows a new period of reproduction during which 

 new dissepiments are formed." 



It is difficult to see how such a view can be held when we consider that the polyp is 



firmly attached all along its height to the corallum by means of the wedge-shaped 



processes which would prevent any periodical moving up of the polyp. My observations 



show that there is always some tissue in the cavities of the corallites below the base of 



SECOND SERIES— ZOOLOGY, VOL. XVII. 2 



