XXXIV INTRODUCTION. 



until after many observations that I detected them, in an 

 example from T. crass icor-nis, which had discharged about 

 half of the wire ; I have not seen the slightest sign of arma- 

 ture on the cethorceum. So far as my investigations go, 

 these spiral cnidce are confined to the walls of the tentacles, 

 in which, however, they are the dominant form. 



(4.) Ohbate Cnidce (cnidce globatce) ? In the acontium 

 of T. parasitica flattened under pressure, and finally ex- 

 pressed from its substance, are numerous more or less 

 globose or ovate vesicles, which' gradually push out a 

 cylindrical protuberance at each end, sometimes to a length 

 equal to that of the original form (figs. 11, 12). These 

 vesicles appear filled with a fluid of different refractive 

 power from that of the clear sarcode in which they are 

 lodged ; but no sign of contained thread have I been able 

 to detect, nor have I seen any discharge beyond the pro- 

 trusion above spoken of. I am not at all sure that these 

 vesicles are consimilar in function with the true cnidce; 

 and I am still more doubtful about the bacillar bodies 

 ound in the acontioid filaments of T. crassicornis. 



In the indubitable cnidce, — those which I have distin- 

 guished as (1) Chambered and (2) Tangled, — the emission 

 of the ecthorceum is a process of distinct eversion. This is 

 not a solid but a tubular prolongation of the walls of the 

 cnidce, turned in, during its primal condition, like the finger 

 of a glove drawn into the cavity. Some of the observa- 

 tions on which I ground this conclusion I have already 

 published, but it may not be impertinent to repeat them 

 here, with others which have since occurred to me, all 

 proving the same fact. In the discharge of the ecthorceum 

 of the tangled cnidce, it frequently runs out, not in a right 

 line, but in a spiral form ; whenever this is the case, each 

 band of the spire is made, and stereotyped, so to speak, in 

 succession, while the tips go on lengthening : the tip only 

 progresses, the whole of the portion actually discharged 

 remains perfectly fixed ; which could not be on any other 

 supposition than that of evolution. In the discharge of the 

 chambered kind, the ventricose or basal portion first 

 appears ; the lower barbs fly out before the upper ones, 

 and all are fully expanded before the attenuated portion 

 begins to lengthen. This again is consistent only with the 

 fact of the evolution of the whole. On several occasions of 

 observation on the chambered cnidce of Caryophyllia, I 



