On the Striicture aud Affiuities of Mnestra parasites Krohn etc. 41 



the ventral margin of the neck of the Ph//IlirJioe and is generally 

 so tirmly implauted that the manubrium will sooner be torn in two 

 than let go its hold on tlie host. In one or two instances, I have 

 thought I could detect an indication of oral lobes at tire end of the 

 manubrium, but the indication was too slight aud the modification 

 of the manubrium too great to support any definite comparison with 

 Gemmaria in this respect. 



The mouth and "oesophageal" portion of the manubrium exliil)its 

 au extraordinary modification of structure, which although paralleled 

 by what is known among certain Acrasped medusae, lias not hitherto 

 been deseribed in any Craspedote form. The oral passage, instead 

 of being freely open, has beeome blocked by a spongy growth of 

 endoderm traversed by minute intercellular Spaces, just as the central 

 mouth of the Rhizostomae has been replaced by a multitude of 

 tiny suctorial passages (red in fig. 41, shown in transverse section 

 in fig. 19). It is a case of like cause producing like effect. Both 

 Miiestra aud the Rhizostomae have adopted a suctorial method of 

 nutrition for which a spacious open mouth, perhaps for purely 

 physical reasous, is not well adapted and has been replaced by a 

 number of small passages. In Rhizostomae these passages are of a 

 well defined nature, in Mnestra they are mere intercellular Spaces in 

 the endoderm. 



The intercellular passages lead into a spacious axial cavity from 

 which the radial cauals lead on to the circular canal. The central 

 cavity is usually irregulär in shape and offen has pouches communi- 

 cating with it. The radial canals, though typically four, are variable 

 both in number and distribution (figs. 39, 42), Two radial canals 

 may unite aud open by a common Channel into the stomach; a 

 fifth canal may appear betweeu two radii etc. On the whole the 

 variations in the arrangement of the radial canals seem to be more 

 numerous than is usuai amongst free swimming medusae, but it 

 should be noted that in the closely allied Eleutherla dichotoma the 

 number of radial canals seems also to be highly variable (Haeckel). 



A ciliated epifhelium in the radial canals causes the contained 

 liquid to flow in two opposite directions. Aloug the upper part 

 of the canals there is a centrifugai current; along the lower part, 

 the current is ceutripetal. Thus a very complete circulation is 

 maintained. 



