'i6 Proceedings of the Royal Irish Academy. 



5. JuTicella gemmacfa Valenciennes. 



6. Jwiicella racemosa Wright and Studer. 



7. JunceUa miniacea Thomson and Henderson. 



8. JunceUa trilinenta Thomson and Henderson. 



But in addition to these the following species have been also referred 

 to this genus : — 



9. JunceUa santae-crucis Duch. and Mich. 



10. JunceUa funiculina Duch. and !Mich. 



11. JunceUa barbadensis^ Duch. and !Micli. 



12. JunceUa vim-en Ellis and Solander. 



13. JunceUa ccdricidata Ellis and Solander. 



14. Jtincena hystric Valenciennes. 



15. JunceUa surculus Johnson. 



16. JunceUa laevis Verrdl. 



17. JunceUa Cretans Verrill. 



Species 9-17 are, however, so imperfectly known that it is absolutely 

 impossible to include them in any scheme of classification. In several cases 

 they are names without descriptions ; and in the others, the descriptions are 

 extremely vague, and are based on characters which are now known to be of 

 no specific value. In no case have the spicides been in\estigated, so that it is 

 even impossible to say whether they actually belong to this genus or not ; 

 in fact, it is more than probable that they are not all referable to JunceUa. 



I have carefully searched through several old collections for authentic 

 specimens of any of these ; but the result has been negative, so that in the 

 absence of type-specimens, but for the sake of completeness, it has been 

 decided to place them in a group — " incertae sedis " — by themselves, and give 

 such references and descriptions as are available. 



An attempt, however, has been made to trace the affinities of species 1-8 

 Each of these is discussed in detail under its place in the emended classi- 

 fication suggested later, so that it is necessary here only to consider the 

 characters on which the classification is founded. 



The first and most important of these is " tits number of main longitudinal 

 canxfls," and this at once separates off JunceUa trUineata from the others. 



An examination of the spicules marks JunceUa raeemosa as distinct (see 

 figs. 14 and 23). In addition to this, however, the general nature of the 

 colony and the mode of branching are distinctive for this species, which 

 under the present system includes JunceUa minacea. There, therefore, remain 

 only species 1-5 to be considered. JunceUa flerilis, J. fragilis, and J. harha- 



