222 ANATOMY OF SPECIAL FORMS. 



hut wlicn in various states of contraction tlicy may assume the shape of thick cluljs, or Lc more 

 or less fusiform, or present a trumpet-shaped figure. The nioutli may be so completely closed 

 that it will be difficult to find any indication of it, or it may be kept widely open with the cavity 

 of the hypostonie actually everted, and the contracted tentacles thrown back and in great ])art 

 concealed beneath the reflected lip. So very mutaljlc, indeed, are tiie liydranths that the observer 

 ■can scarcely avoid attributing to the colony a greater amount of heteromorphisni than it really 

 possesses. In its most contracted state the Ijody of the hydranth is marked by very distinct 

 transverse rugae. 



Tiie tentacles vary much in number. Tliey are usually from fifteen to twenty, but I have 

 counted as many as thirty-one in a full-grown hydranth. Tlicy consist of two or more alternating 

 series, so closely approximated as to constitute a single circlet surrounding the base of a well- 

 developed hypostome. They are irregular in length, those belonging to the distal series of which 

 the circlet is composed being longer than those of the proximal series. The axis of the tentacle 

 is formed by the usual large-celled endodermal tissue, but the cells are of more unequal size, and' 

 give a less regularly septate appearance to the axis than is the case in many other hydroids. 



I have occasionally met with a hydranth which has become double-headed by a bifurcation 

 near its middle. Quatrefages has also noticed this condition, and Agassiz has noticed it in the 

 spiral zooids to be next described. 



Near that portion of the margin of the colony which lies over tiie mouth of the univalve 

 shell invested l)y it there occurs a set of very remarkable zooids (fig. 1, c, c, c). These are long- 

 cylindrical bodies, usually attaining, when fully extended, about two thirds of the height of the 

 ordinary liydranths when similarly extended. Their distal extremity is slightly enlarged, and 

 provided with a coronal of spherical clusters of thread-cells. Their axis is hollowed out into a 

 tubular cavity like that of the ordinary liydranths, but they seem to be entirely destitute of 

 mouth. They are provided with a strongly developed layer of muscular fibres between ectoderm 

 •and endoderai. Their most remarkable feature consists in their power of rolling themselves into 

 a spiral, whose coils are in vertical planes, somewhat recalling the circinate vernation of a fern. 

 On any irritation of the neighbouring parts they throv,' themselves into energetic and violent 

 action, uncoiling and coiling themselves, and lashing backwards and forwards on their basis. 

 They are quite constant in their occurrence. Among all the specimens of IL/dradinia I have 

 examined I have never met with one in which they were not present ; and it is difficult to 

 understand how Quatrefages and Van Beneden could have overlooked them. The first published 

 description of them is by Dr. Strethill Wright. 



T//e Gonosome. — The gonophores are always borne upon blastostyles (PI. NV, fig. 1, 5, h, h, 

 and fig. 3). These are scattered among the liydranths without any apparent order. They are 

 cyluidrical, slightly dilated at the distal extremity, and for the most part about half the height 

 and thickness of the hydranth. The summit is provided with globular projections, which take 

 the place of the tentacles in the liydranths. These projections are arranged in two or thi-ee 

 alternating, closely approximated verticils. Each of them is formed by a cluster of thread-cells, 

 thickly set in a projection of the ectoderm. I have not satisfied myself as to the com))lete 

 absence of a mouth in the blastostylc. If, however, it be present, it is only in the state of a 

 minute, scarcely perceptible puncture, which may occasionally be witnessed on the summit of the 

 blastostyle, and through which Wright has seen the contents of the somatic cavity voided under 

 pressure ; but it can hardly ser>'c the i)urpnse of ingestion of nutriment, though Agassiz has 



