BY H. L. KESTEVEN. 685 



emlo-mesodeim ''anlacje" is early recognisable, but only in the early 

 nauplius-stage, does it become possible to separate endoderm and 

 mesoderm. '1^ 



It is worthy of note that histogenesis has not progressed so far 

 in Uhius at the nauplius stage as in Chondracanthus. This differ- 

 ence is most striking at the posterior end. In Ubius the ectoderm 

 is so extremely thin as to be actually doubtfully present at the 

 extreme posterior end, whilst the mesoderm-belt is composed of not 

 more than twenty-five large cells. In Chondracanthus, the ecto- 

 derm is so well developed that the rudiment of the ventral ganglion 

 (Bauchganglienanlage) may be identified. Mesoderm and endo- 

 derm, howexer, are not so far advanced in differentiation. 



Comparative. — A brief comparison with h->e halanoglossi Mayer, 

 has already been made in the generic description; in view of 

 their close relationship and similarity, it may be well to tabulate 

 the outstanding differences. 



Ive halanoglossi. Ubius hilli. 



Body presenting annular Body cylindrical, without 

 thickenings, suggesting segmen- trace of segmentation, 

 tation. 



Oesophagus chitin-lined. Oesophagus lined by colum- 



nar epithelium. 



Mid-gut strongly muscular. Mid-gut devoid of obvious 



muscles. 



End-gut lined by chitin. End-gut absent. 



Double ventral nerve-cord. No nerve-cord visible. 



Ovary single. Ovary paired. 



Testes large, reniform. Testes small, cylindrical. 



Systematic. — The systematic position of Vhius must remain more 

 or less in doubt, until the stages between the nauplius and the 

 adult are known; meanwhile, an attempt must be made to assign 

 it a position on the characters of the adult. Notwithstanding the 

 important differences between Ive and Uhius, it seems evident that 

 they both belong to the same family. I agree with Mayer that it is 

 not necessary to found a new family for the reception of Ive, but 

 unlike him, I am unable to see that its adult features point to a 

 relationship with Lerna^a. 



