MATTHAI— RECENT COLONIAL ASTR^ID^ 13 



axis shorter ; this variety may be termed type II c (PL 6, figs. 59 and 62); they are usually 

 arranged close together in the coils of the mesenterial filaments. In G. retiformis (Lam.) 

 the axis is about three-fifths the length of the sac and somewhat bent, whereas in 

 G. solida (Ed. & H.) and G. pectinata (Ed. & H.) it is about a quarter of the length of 

 the nematocyst and somewhat swollen in its middle; in the last two species there is also 

 a suggestion of a coiled thread in the sac. 



I have met with nematocysts of this type in the calicoblastic layer of certain well- 

 preserved polyps whose corallum resembles that oi MycecUum okeni Ed. & H.*. Obviously 

 they cannot be of any use in this layer as they are hidden by the corallum, but their 

 presence proves that the power of developing these structures is inherent in the ectoderm 

 everywhere. 



In certain species, e.g. Favia doreyensis, Ed. & H. (PL 5, fig. 45), I have seen this 

 type of nematocyst in the endoderm, though not in any great abundance as in the 

 ectoderm. Their presence in that layer may be explained in two ways, either that the 

 nematocysts have arisen in the endoderm, or more probably, that nematoblasts have 

 migrated into this layer from the ectoderm and have developed there t. 



It is highly probable that the dark-stained axis is a modified nucleus. Certain 

 appearances suggest that a nucleus wanders to the surface of the ectoderm, elongates and 

 becomes modified into the axis, its chromatin forming the axial core and the fine spiral 

 and the nuclear membrane constituting the surrounding sheath. This process is accom- 

 panied by the formation of a membranous wall in the surrounding cytoplasm, the latter 

 forming the contents of the sac. A nematocyst of this type would accordingly be a 

 complete cell (PL 4, figs. 37 a — -/). 



In Gardiner's polyps of Ccenopsammia nematocysts of this type are present in the 

 convolutions of the mesenterial filaments, but are somewhat narrow, and the dark-stained 

 axis of each extend usually to about a third of the length of the sac, rarely to its middle. 

 Around the axis (Gardiner's " eversible portion") is a continuous spiral as in my polyps, 

 not a " spiral row of short hairs." I have not seen any distinct coiled thread as Gardiner 

 has represented in his diagrammatical fig. 14, nor any definite nucleus in the nematocyst 

 as in his figs. 17 and 18. In a few cases the axis has been forcibly ejected (not everted), 

 as could be seen in Gardiner's section shown in fig. 13, leaving a corresponding space 

 within the sac J. 



Type III (PL 2, figs. 15, 16, 17, PL 4, fig. 40, PL 6, figs. 57, 60, and 61). The 

 largest nematocysts that I have examined belong to this type. Each of them is broader 

 in the middle than at the ends, and consists of a membranous wall, enclosing protoplasmic 

 contents of the same nature as in type II. Ptunning through the middle of the sac, along 

 its whole length, is a somewhat thick protoplasmic core, stained deeper pink than the rest 

 of the contents. In addition to this, there is a long thread spirally coiled along the inner 



* The description of this species is reserved for a subsequent paper. 



t For migration of nematocysts see Schneider, Hadzi and Boulenger. 



X The "larger nematocysts resembling those of the body-wall but usually much longer and narrower," 

 which Bourne (17, p. 224) observed in the tentacular batteries of Heterocyathus cequicontaius interspersed 

 among the "small nematocysts of the usual form," may, I venture to suggest, be similar to type II b, but 

 Bourne has not figured any of them. 



