Kirk. — On Ascidioclava. 147 



portion greatly developed, and cavities completely obliterated. Medusa 

 buds shortly pedunculate, the hollow peduncle containing a prolongation 

 of the gastro-vascular cavity. Gonads at present unobserved. 

 . Ascidioclava parasitica presents in both trophosome and gonosome 

 anatomical characters that seem to me of interest. I note here the most 

 striking of these. 



Tiop/tosome. — The ectoderm shows great differences in the different 

 parts. On the tentacles and the hypostome it is cellular, and presents no 

 special features. Below the tentacles the ectoderm of the hydranth shows 

 nuclei and fragments of nuclei arranged indiscriminately, often three or 

 four deep. Cell-outlines have quite disappeared (fig. 2). In this part the 

 ectoderm somewhat resembles that of Clam squamata as figured by Allman.* 



Towards the base of the hydranth this thickened ectoderm thins out 

 to form the thin non-cellular membranous ectoderm of the hvdrocaulus. 

 Through this membrane the outline of the bases of the endoderm cells, 

 changed as noted below, can be seen, giving the appearance of a fine irregular 

 network. The ectoderm of the upper surface of the hydrorhiza is thin 

 and nuclei are sparse. On the lower surface, where it is in contact with the 

 epithelium of the host, the ectoderm consists of columnar cells with very 

 clearly marked outline (fig. 4). In no part of the organism have I been able 

 to discover any trace of a chitinous secretion. 



The mesogloea is everywhere thin, and almost membranous. 



The endoderm of the hypostome comprises a large number of glandular 

 cells, and these decrease in number towards the lower part of the hydranth 

 (fig. 2). In the hydrocaulus the endoderm cells are reduced to large thin- 

 walled vacuoles. Only occasionally can a shrivelled nucleus be found ad- 

 hering to the cell-membrane. Towards the base of the hydrocaulus these 

 gradually give place to living cells with highly granular protoplasm. These 

 encroach upon the cavity until, at the very base of the hydrocaulus, it is 

 completely obliterated. This obliteration of the cavity marks the hydrorhiza 

 also, and when the dividing walls become absorbed there is a continuous sheet 

 of endoderm surrounded by the thin mesogloea and the ectoderm (fig. 4). 



The large vacuolated endodermal core of the tentacles is very well de- 

 veloped. These cells, like the clear endoderm cells of the hydrocaulus, 

 usually contain each a single vacuole. Allmant says with regard to the 

 central endoderm of tentacles, " It would seem that the solid axial tissue 

 of the tentacles is in every instance separated by the mesosarc, not only 

 from the ectodermal layer of the tentacles, but, by a duplicature of the 

 mesosarc, from the endoderm which lines the body-cavity of the hydranth, 

 as was first pointed" out by von Koch in Tubularia." I can find no trace 

 of such an arrangement here. The axial endoderm of the tentacles appears 

 to me to rest unmistakably on the endoderm of the gastro-vascular cavity 

 of the hydranth (fig. 2). 



Gonosome. — Medusa buds are produced singly, or more often in groups 

 of from two to four, near the base of the hydranth. I have not found more 

 than one group on any hydranth. Material collected from October to 

 December does not enable me to say with certainty whether the medusa 

 becomes free. A single medusa better developed than the rest was found 

 unattached, but it is possible that it became detached in the removal of the 

 hydroid from its host. It still showed the scar at the point at which it had 

 been attached. 



* "A Monograph of the Gymnoblastic or Tubularian Hydroids," pt. i, pi. i, 1871. 

 t " Report on the Hydroida dredged by H.M.S. ' Challenger,' " p. x, 1888. 



