PENNATULIDA DREDGED BY H.M.S. "TRITON." 141 



as that here proposed, the axial surface being that which bears the long 

 mesenterial filaments. As these words express a real and an important rela- 

 tion, they would appear preferable to the very misleading terms dorsal and 

 ventral, which are commonly employed to denote the surfaces in question. 



The plane of symmetry of the zooids obeys exactly the same laws as that 

 of the polyps, the mesenterial filaments being placed on the axial wall. 



Concerning the arrangement of the zooids on the rachis, it will be seen from 

 fig. 28 that the reason of the existence of a short tract devoid of zooids imme- 

 diately below each polyp is that this tract is really part of the dbaxial wall of 

 the polyp; and as the zooids are developed on the rachis itself and not on the 

 polyps, there can clearly be no zooids on these tracts. 



Retraction of Polyps. — In spite of the great rigidity of the wall both of the 

 polyp itself and its tentacles, due to the enormous number of spicules contained 

 in it, the polyps can, as shown on the right-hand side of fig. 23, be withdrawn 

 almost completely into the rachis, the tentacles entirely disappearing from 

 sight in the fully retracted state. During the process of retraction the body 

 wall of the polyp is thrown into transverse folds, and one specially deep fold 

 at the junction of body and tentacles {vide the left-hand polyp of fig. 28) corre- 

 sponds to the calyx of other Pennatulida. 



Structure of Stalk. — This is well shown in fig. 27, representing a transverse 

 section taken about the middle of its length. The mesoderm is of great 

 thickness, and is divided into inner and outer zones by the well-developed 

 layer of longitudinal muscles Im, which forms a deeply corrugated sheath 

 extending round the whole stalk. Of the two zones the outer one is very 

 richly studded with calcareous spicules i, crossing one another in all possible 

 directions ; while the inner zone is devoid of spicules, and is traversed by a 

 dense network of nutrient canals. The stem c is quadrangular, with rounded 

 angles and grooved lateral surfaces. In the rachis, as we have seen (rig. 28), the 

 stem is cylindrical; but this change in shape is by no means exceptional, 

 occurring in Pennatula and several other genera, as well as in Kophobelemnon. 



The stem is invested by a mesodermal sheath, which is prolonged outwards 

 to the body wall as four vertical septa, which separate from one another the 

 four main longitudinal canals of the stalk, of which the dorsal dc, and ventral 

 vc, are considerably larger than the lateral ones Ic. If these canals, which are 

 lined by endoderm, be traced upwards towards the rachis, the two lateral ones 

 are soon found to disappear ; the dorsal one extends a short distance up the 

 rachis, and then in its turn disappears, while the ventral one (fig. 28, vc) 

 persists of considerable size throughout the whole length of the rachis. 



Of Kophobelemnon stelliferum Kolliker* distinguished at first two varieties, 

 which he named mollis and dura respectively, the difference consisting chiefly 



* Kollikbr, op. cit., p. 305. 



