62 PROF. ALLMAN ON STEPHANOSCYPHUS MIRAB1LIS. 



boscis, such as in the true hydroids we find intervening between the tentacles and the 

 mouth of the hydranth ; and it was the great width of the terminal orifice, with its mar- 

 ginal circlet of tentacles, so different from the narrow mouth of a hydroid elevated on 

 the summit of a hypostome, that first led me to doubt the justice of referring Stephano- 

 scyplms to the true hydroids. The appearance of the tentacular crown, indeed, was not 

 altogether unlike that of certain zoanthal Actinozoa; but all grounds for assigning to 

 Stephanoscyplms such actinozoal relations were removed by dissection. 



The tentacles form a single circlet, which is composed of two closely approximated 

 alternating series of about eighteen tentacles in each, the outer series being slightly 

 shorter than the inner, and usually depressed in extension, while the inner series is held 

 more elevated. Not unfrequently the whole circlet may be seen bent back upon the 

 outer side of the chitinous tube (fig. 3, c). The tentacles are very contractile : when 

 fully extended they are long and filiform ; but in certain states of contraction they 

 assume a very distinctly clavate form (fig. 3, b). Their structure is entirely that of the 

 typical hydroid tentacle : they present the septate condition so characteristic of such 

 tentacles ; and they carry along their length, just as in most hydroids, somewhat spirally 

 arranged slightly elevated clusters of thread-cells (fig. 4). In complete retraction the 

 tentacles are not only greatly shortened, but are thrown inwards and entirely with- 

 drawn within the terminal orifice of the animal, which is then nearly or altogether 

 closed over them (fig. 3, d). 



In the completely extended state of the animal, four longitudinal canals may be traced 

 from the base of the tentacular crown for some distance backwards in the walls of the 

 body ; but the general want of transparency, and the way in which the greater part of 

 the animal is concealed by the surrounding sponge, renders it impossible to obtain, 

 without dissection, a satisfactory view of the deeper-lying parts. Sections, however, were 

 easily made of specimens which had been preserved in spirits ; and some further im- 

 portant points of structure were thus determined. 



When a longitudinal section (fig. 5) of such specimens is made through the axis of 

 the body, it is seen that the terminal orifice opens into a wide thin-walled cavity, which 

 becomes continuous posteriorly with the canal which occupies the axis of the stem. 

 The contracted tentacles are now also brought into view with their extremities thrown 

 back, and with the whole tentacular crown entirely withdrawn into the cavity. 



By the same section another very important piece of structure becomes apparent ; for 

 it may be now seen that the knife has divided a circular canal which runs in the walls of 

 the body round the terminal orifice. This canal is very obvious ; it is remarkable for its 

 width, and easily admits the passage into it of a needle. It supports the tentacular 

 circlet ; and when a transverse section of the body is made along its course, it is seen to 

 be quite continuous and destitute of septa. It possesses a very distinct endodermal 

 lining. 



With the circular canal the distal extremities of the longitudinal canals are in con- 

 nexion, and they probably open into it. These canals can now be traced far back in the 

 walls of the body. They lie quite close to the inner surface of the walls ; and their sides 

 project into the cavity as four slightly elevated ridges. In the deeper parts of this 



