OF COENOPSAMMIA FROM LIFU. 371 



fourteen turns instead of three rows as described in Caryophyllia. I could not dis- 

 tinguish in the thread any spiral band as found in Canjophyllia, nor any peculiar 

 armature at the tip as in Euphyllia. The thi-ead is very thick, and lies in the 

 ripe condition close under the exterior membrane, which it may bulge out slightly 

 (Fig. 14). It is coiled from 8 to 10 times on itself, and ends below in a granular 

 mass of protoplasm at the base of the cell. The latter lies at some small distance 

 from the structureless membrane in a protoplasmic bed connected by one or more fibres 

 with the nervous layer. 



It is e.xtremely difficult to see the mode of eversion of the nematocyst, but 

 generally the thread appears to be extruded first, and to carry with its lower end the 

 eversible base (Fig. 15). Sometimes, the latter can, when the thread is partially 

 e.Ktruded, be seen to be partially everted, and in one case I found it completely 

 everted with the thread quite distinctly visible in the middle (Fig. IG). It does not, 

 however, in any way approach the condition described and figured by Mobius (20), 

 for the everted nematocysts of Caryophyllia. When the thread is partially, or but 

 recently extruded, there appears to be a distinct cavity left in the protoplasm (Fig. 16). 

 In sections too the thread appears to have a distinct sheath, which is quite separate 

 from the eversible basal portion. 



Besides these ripe nematocysts, a large number are found having the eversible 

 base with central thread well marked, the rest of the nematocyst staining very uni- 

 formly of a light brown colour in iron haematoxylin and eosin (Fig. 17). In a few 

 of these cells I found a large nucleus near the base, in one with a well marked 

 nucleolus. A differentiation of the protoplasm into a slightly darker band, running 

 spirally round the cell is next seen (Fig. 19), and from this up to the fully developed 

 thread every stage is clearly visible. The darker band first appears at the outer end — 

 in respect to the structureless membrane — of the cell, and the thread develops from 

 this end towards its base. 



In the filaments the nematocyst thread is often found completely extruded with 

 the cell in situ. In other cases the cell is partially or completely extruded from the 

 filament, but it is impossible to say how far this may not be due to pressure brought 

 about by the strong contraction of the polyps. I have been unable to find any 

 appearance of developing nematocysts other than those previously described, — with the 

 possible exception of the nematocyst represented in fig. 18 — and it seems to me to 

 be probable that the thread becomes detached somewhere beyond the eversible base, 

 which is then retracted, a new thread being formed. In some of the extruded ne- 

 matocysts the thread seems to have been broken off in this way, but there is no 

 appearance in the eversible base of myophan bands. In one extruded nematocyst there 

 is a well marked nucleus at the base, and in several ripe nematocysts I have found 

 a more darkly staining mass within the thread, but with sections only it is almost 

 impossible to speak with certainty on such points. 



Calicoblastic Ectoderm (c. ect. Fig. 20). The polyp next the corallum is every- 

 where covered by a thin layer of ectoderm, the calicoblast layer. This is not generally 

 very well marked, and in sections looks like a thin line of tissue, slightly swollen in 

 places where the small, round, homogeneously staining nuclei are situated. The nuclei 



