14 PERCY SLADEN TRUST EXPEDITION 



surface of the membranous wall, much thicker than in type I, but with far fewer turns of 

 the spiral. In the unextruded condition, it is difficult to make out the exact nature of 

 the thread, but, as in some of my polyps, e.g. Favia favus (Forsk.), most of these nema- 

 tocysts are found with the thread partially ejected, I have been able to study its structure. 

 It consists of an axial strand stained homogeneously dark, enclosed in a thin membranous 

 sheath against the inner surface of which lies a fine closely-wound spiral stained dark. 

 The tip of the axial strand is pointed, and projects in spike-like fashion beyond the 

 sheath. In some of the nematocysts the sheath is partly torn and the spiral broken 

 here and there along the thread. The central protoplasmic core is never everted; indeed 

 it does not appear to have any connection Avith the thread at all. As in type I, the 

 thread ends at the base of the nematocysts in a small granular mass. This type varies 

 like I and II in the different species, an extreme case being that of Goniastrea retiformis, 

 in which the sac is much narrower (PL 5, fig. 49). 



There is little doubt that the so-called " ripe nematocysts " of Bourne from Euphyllia 

 glabrescens, judging from his figures (PL 4, figs. 10 and 12) really belong to this type; 

 its " axial body," which the author believes to be eversible, being probably the central 

 protoplasmic core. I have not met with any peculiar armature at the extremity of the 

 thread, as in Bourne's fig. 11, in any of my nematocysts. 



In the three species of Echinopora examined, this type has undergone a characteristic 

 modification, in that the sac is not only narrow, as in Goniastrea retiformis, but the 

 longitudinal protoplasmic core is absent and the turns of the coiled thread more numerous. 

 This variety may be termed type III h (PL 2, figs. 16 and 17). 



The large nematocysts with the long coiled thread which Jourdan (74) found in the 

 ectoderm of the body-wall of Corynactis viridis (PL 8, figs. 54 and 55) and in the 

 mesenterial filament of Balanophyllia regia (PL 15, fig. 113), come under my type. 

 Resembling the latter are also the two nematocysts which Bourne (17) has figured from 

 the ectoderm of the body-wall of Heterocyathus mquicostatus , but as in Gardiner's 

 representations {Coenopsammia, figs. 15 and 16), the everted thread is covered with long 

 barbs arranged in a spiral. The " medium-sized elongate " nematocysts which Bourne 

 describes in the same paper from the ectoderm of Dendrophyllia gracilis (PL 4, figs. 25 

 and 26 6), as well as the nematocysts from the mesenterial filaments of an anthocyathus 

 of Fungia (13, fig. 28), resemble III h (PL 2, figs. 16 and 17); I have little doubt that the 

 thread in these cases, if carefully examined, will be found to possess a closely wound 

 spiral. On the other hand, I have not seen the " lai-ge elongate oval " nematocysts 

 (PL 4, figs. 26 c — e, which are the same as PL 3, fig. 18); they are similar to Gardiner's 

 fig. 14 or 19, but have a longer "eversible portion." 



Type III is essentially similar to type II, consisting of a membranous wall, proto- 

 plasmic contents and a thread which, like the axis of II, has a dark-stained central core, 

 a thin membranous sheath, and a surrounding fine spiral. The difierent parts appear 

 also to have the same origin as those of type II, the thread being probably a modified 

 nucleus (the axial strand and the spiral formed perhaps of the chromatin and the sheath 

 of the nuclear membrane) and the membranous wall developed independently in the 

 surrounding cytoplasm. The gi'anular mass at the base of the nematocyst into which the 



