INTRODUCTION. 
xxxin 
in the various tissues. In tlie genus Sagartia, however, it 
is by far the rarer form, while in Actinia and Anthea, 
it seems to be the onl}^ one. 
The pretty little Corynactis viridis is the best species 
that I am acquainted with for studying this kind of cnidcc. 
Their figure is near that of a perfect oval {Plate XT. fig. 9), 
but a little flattened in one aspect, about '004 inch in the 
longer, and "OOIS in the shorter diameter. Their size, 
therefore, makes them peculiarly suitable for observations 
on the structm'e and functions of these curious organs. 
AVithin the cavity is a thread {ecthorceiim) of gi-eat 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 eciliorceum of the tangled cnida 
is sun-ounded 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 Cnidee {Cnidee cocldeatce). In a few species, as 
S. parasitica, T. crassicornis, and Cerianthus Lloydii, I have 
found very elongated fusiform 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 Cerithiadee or 
Turritellad(B (Plate XI. fig. 10). The cctliorcciim 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 
