17 



genera, i. e. Juncella, Ellisella^ Vimenella and "-. Ifclicclla, a step in which he has not been 

 followed by later writers. 



Studer (1878) found that this perplexing group could be divided into two subordinate 

 groups on the basis of the spicules, one group having an outer layer of clubs and an inner 

 layer of doublé clubs, the other group being characterized by doublé clubs and spindles. By 

 using this feature and certain characters of the calyces he arrived at a definition of the genus 

 'Jiiucclla which may be translated as follows: 



"Stem simple or dichotomously branched; calyces moderately or strongly exserted, 

 club-shaped. Ccenenchyma thick, with an outer layer of clubs beneath which is a layer of 

 doublé clubs". This author includes in the genus as thus defined Juncella jitncea (Pallas), J. 

 flexilis Studer, J. gcmmacca (Valenciennes) and (provisionally) J. vimeti Val. and J. ccvis Verrill. 



Later (1887) the same writer uses practically the same definition, but adds that the 

 "Calyces are in two lateral rovvs", a character that does not hold, although repeated by Wright 

 and Studer (1889). 



HlCKSON (1905) discusses the genus at some length and points out clearly the great 

 difficulty in difining the genera of this family. He combines the genera Juncella and Ellisclla 

 of previous writers in the one genus Juncella. He would thus include about ten species in 

 the genus. Four of these (Juncella elongata Val., Ellisclla calamus Studer, E. maculata Studer, 

 and Juncella spiralis Hickson) the present writer would not admit in the genus, using the 

 very characters pointed out by Hickson, i. e. the presence or absence of club-shaped spicules. 



Thomson and Simpson (1909) do not attempt to define the genus or to place the large 

 collection of Juncella and its allies at their disposal in their proper genera or species. A 

 course which they explain as follows : 



"At the same time we refrain from multiplying species without some sound basis of 

 classification. This we hope to supply in a future study". 



The present writer, while naturally appalled by a task from which such able authorities 

 shrink, feels that it is possible to make at least some progress in defining this perplexing 

 group, and particularly that the genus Juncella itself is capable of being fairly well seperated 

 from allied genera such as Ellisclla and Scirpcarella. He therefore suggests the following 

 definition for the genus Juncella : 



Colony simple or branched, never anastomosing and usually dichotomous when branched. 

 Calyces usually on all sides of the colony, (except in young specimens where they are lateral) 

 often leaving a bare space on two opposite sides of the branches, papillate or club-shaped, inclined 

 distally, capable of partial retraction along with the strongly retracted polyp. Spicules in an 

 outer layer of club-stars with club end beset with distally directed spines underneath which is 

 a thick ccenenchyma with very numerous doublé heads and doublé stars. Axis cylinder with 

 concentric layers of horny and calcareous matter. 



The type species of this genus is Jimcella juncca (Pallas). Other known species are 

 J. barbadensis Wright and Studer, J. flagellum Johnston, J. flexilis Studer 5 , J. fragilis 

 Ridley, J. gcmmacca Valenciennes, J. racemosa Wright and Studer, J. trilincata Thomson 

 and Herderson, and the new species described beyond. 



S1BOGA-EXPEMTIE XIII Ifl. I 



