518 GILBERT 0. BOUHNE. 



entire destruction of the soft tissues, and for the study of the 

 corallum itself I made use of sections, minute transparent 

 fragments, crushed and '' etched " preparations. 



Most careful examination with high powers (Zeiss, -^ oil 

 immersion, with comp. oculars 4 and 8) convinced me that 

 there is no question of any '' spicular ^^ structure in Helio- 

 pora. The corallum is, as I have before stated, a secretion 

 product of a definite layer of cells derived from the ectoderm, 

 which I have called, like those of Madreporaria, calicoblasts. 



So far as the derivation and general arrangement of the 

 calicoblasts is concerned I have nothing to add to what is 

 contained in my previous paper, but improved histological 

 methods and the use of higher powers of the microscope 

 enable me to add a good deal in the way of detail. The cali- 

 coblasts form a somewhat thin covering over the coenenchyraal 

 tubes, and the proximal moieties of the zooids, and are every- 

 where present between the living tissues and the corallum. 

 But they seldom are arranged in a single layer. In some 

 places, it is true, only a single row of cubical or fusiform cells 

 of coarsely vacuolated structure with numerous granules can 

 be distinguished. In such cases the calicoblasts always appear 

 to have an external limiting membrane, turned towards the 

 corallum, and this membrane, like the mesogloea, stains bright 

 blue with picro-nigrosin. But in all places where the tissues 

 are well preserved, and show no indications of having been dis- 

 placed by the action of bubbles formed during decalcification, 

 a layer of flattened cells, fusiform in section, can be distin- 

 guished outside of the external limiting membrane, and there- 

 fore abutting on the corallum. These are shown in fig. 13, 

 ca. e. They are found in this particular condition only in 

 those places where coral secretion is least active. At the 

 bottom of the superficial canals, where active centrifugal 

 growth of the corallum is taking place, and at the lower blind 

 ends of the ccenenchymal tubes and zooids (the regions where 

 the tabulae are formed), the calicoblasts form a layer several 

 cells deep, as I have shown in my previous paper (loc. cit., pi. 

 10, fig. 4). In these regions tiie calicoblasts are grouped 



