124 A NBW FAMILY OF HYDROIDEA. 



from these other branches are given off on either side, which again divide in a roughly 

 but not constantly dichotomous manner, the smaller branches being cylindrical in 

 shape. All the branches in the living and spirit-preserved specimens usually lie in the 

 one plane, but when dried they often become slightly twisted, and thus thrown out of 

 the plane. There is thus a considerable resemblance in general appearance between 

 the form Clathrozoon on the one hand, and Dehitella and Ceratella on the other, 

 especially between the two first mentioned, owing to the rounded branches of 

 Dehitella. If, however, we come to examine the structure of the two more minutely, 

 we find that the differences are very marked ; whilst both have the skeleton in the 

 form of a horny network, that of Clathrozoon has more the appearance of being 

 composed of a scries of anastomosing tubes, and rather less of the appearance of an 

 open meshwork than is the case in the Ceratelladfe ; but, what is of much greater 

 importance, it presents on its surface a large number of circular openings at the 

 extremities of slightly projecting chitinous tubes, which form true hydrothecse. 

 These are very distinct from what Mr. Carter and Dr. Gray aptly described as scoop- 

 like projections of the chitinous network of the Ceratelladfe, though they 

 unfortunately gave to them the name of hydrothecse, that of hydrophore being 

 more correct and suggestive of their real function, as at most they can but afford a 

 support for the proximal part of the hydranth. 



Skeleton (Figs. 4, 5, 6, 7, 8, 9, 11, 12, 13, 16). 



When the branches of spirit-preserved specimens, with the soft parts present, 

 are viewed under the lens, the surface is seen to be covered with a great number of 

 tortuous grooves, filled with light yellowish coloured material — the coenosarcal 

 tubes. The edges of the grooves are formed by the chitinous perisarc of a dark-brown 

 colour (Fig. 6), and arranged in a spiral manner are circular projections, the external 

 opening of chitinous cylinders forming the hydrothecte and containing the retracted 

 polypeSi These hydrothecte are encircled by the tortuous grooves, and supported by 

 extensions of the ordinary perisarc. The whole surface is covered over by an 

 extremely thin and delicate colourless layer of perisarc common to the whole branch. 

 This layer is not itsually recognised until sections are cut, but the whole surface is 

 seen to be studded with small cylindrical tubes, which are really formed from this 

 thin layer and the spaces in which are continuous with the tortuous grooves below 

 the latter (Figs. 7, 8, 9). 



Sections show also that this thin layer rises up somewhat from the general 

 surface, and is attached to the lips of the hydrothecse ; beyond the margin of the 

 latter it is continued on so as to form a very thin collar-like extension, acting as an 

 operculum. The arrangement is represented in the figures, especially in Fig. 13, 

 which is meant to be a diagram showing the relationship of a hydrotheca to the 



