38 PllOFESSOH ALLMAN. 



hydranth-teiitacles in the great majority of the marine 

 hydroids, though of its existence in some others as well as 

 in the marginal tentacles of many Medusae there can be no 

 doubt. The peculiar septate non-tubular pith which is so 

 very characteristic of the tentacles of the marine hydroids 

 generally, has been already fully described and figured by 

 me (' Gymnoblastic Hydroids/ p. 126, pi. iv, fig. 5, woodcut, 

 fig. 48, &c.), but I still maintain that in most cases the axis 

 of the tentacle continues pervious for some distance from its 

 base, and that the large cells which, with their conjoined 

 walls like so many transverse septa, obliterate its cavity higher 

 up, constitute a tissue which has been formed from a direct 

 continuation by the endodermal cells of the Hydranth. In 

 the figures referred to I have represented this tissue com- 

 posed of a pile of cells, each with its nucleolated nucleus and 

 W'ith its contained protoplasm often sending off exactly as in 

 certain vegetable cells, radiating and branching processes. 



Our author further describes a transverse septum with a 

 central perforation at the base of each tentacle. This septum 

 he regards as an annular process of the " stiitzlamelle." I 

 have never seen it. 



We know that in Cordijlopliora and many other hydroids 

 the spadix or endodermal blind pouch which extends into 

 the axis of the gonophore from the somatic cavity, instead of 

 remaining simple becomes divided into branching tubes. In 

 my description of Cordylophora laciislris I represented 

 those branches as here and there inosculating with one ano- 

 ther. Schulze cannot confirm the existence here of a true 

 inosculation, and believes that the appearance of such is 

 deceptive, and caused by the close approximation of the 

 branches where they lie upon one another. This is possibly 

 true, and though I believed at the time that I had sufficient 

 evidence of a true inosculation I do not now desire to insist 

 on my own interpretation of the appearances presented. In 

 other cases where I have found a branching spadix the 

 branches certainly show no inosculation. 



But a more important statement is made by our author 

 with regard to these branches of the spadix in Cordijlopliora, 

 for he informs us that they are entirely invested by a pro- 

 longation of the hyaline " stiitzlamelle." If this hyaline 

 structureless tube lies between the generative elements and 

 the spadix, as asserted by Schulze, it has certainly escaped 

 me, not only in the spadix of Cordijlopliora but in that of all 

 the other hydroids which I have examined ; and yet, as I did 

 not employ the special reagents used by Schulze in tlie de- 

 monstration I should be Avilling to believe that I had over- 



