INTRODUCTION. XXXlll 



in the various tissues. In the genus Sagartia, however, it 

 is by far the rarer form, while in Actinia and Antliea, 

 it seems to be the only one. 



The pretty little Corynactis viridis is the best species 

 that I am acquainted with for studying this kind of cnida. 

 Their figure is near that of a perfect oval (Plate XI. fig. 9), 

 but a little flattened in one aspect, about -004 inch in the 

 longer, and -0015 in the shorter diameter. Their size, 

 therefore, makes them peculiarly suitable for observations 

 on the structure and functions of these curious organs. 

 Within the cavity is a thread (ectJwraum) of great length 

 and tenuity, coiled up in some instances with an approach 

 to regularity, but much more commonly in loose contor- 

 tions, like an end of thread rudely rolled into a bundle with 

 the fingers. 



The armature of this kind does not differ essentially from 

 that already described. It is true, I have detected it only in 

 Corynactis, where the short ecthoraum of the tangled cnida 

 is surrounded throughout its length by a barbed strebla of 

 three bands. The barbs are visible under very favourable 

 conditions for observation, even while the tangled wire 

 remains enclosed in the cnida, but their optical expression 

 is that of serratures of the walls, without the least appear- 

 ance of a screw. This is the only species in which I have 

 actually seen the armature of the ecthoraum in this kind of 

 cnida, but I infer its existence from analogy, in other 

 species, where the conditions that can be recognised agree 

 with those in this, though the excessive attenuation of the 

 parts precludes actual observation of the structure in 

 question. 



(3.) Spiral Cnidoz {Cnida cochleatce). In a few species, as 

 S. parasitica, T. crassicornis, and Cerianthus Lloydii, I have 

 found very elongated fusiform cnida which seem composed 

 of a slender cylindrical thread, coiled into a very close and 

 regular spiral. In some cases the extremities are obtuse, but 

 in others, as in T. crassicornis, the posterior extremity 

 runs off to a finely attenuated point, the whole of the spire 

 visible even to the last, the whole bearing no small resem- 

 blance to a multispiral shell, as one of the Cerithiada or 

 Turritellada (Plate XI. fig. 10). The ecthoraum is dis- 

 charged reluctantly from this form, and I have never seen 

 an example in which the whole had been run off. So ex- 

 cessively subtle are the walls of the cnida, that it was not 



c 



