Simpson — A Revision of the Gorgonellidae. 283 



densis have proved to be but young stages of J. juncea, and must therefore be 

 inchided under the older name ; so that the number of species is now 

 restricted to two, viz., Jv.nceUa juncea and Juncella gemmacea. There can Vje 

 no doubt that these two names have been very loosely used in the identi- 

 fication of specimens, and with great justification; for after an examination of 

 the macroscopic and microscopic character of a very large number of each of 

 these, it must be confessed that it is almost impossible to distinguish between 

 a branch of Juncella gemmacea and a portion of a colony of Juncella juncea of 

 about the same size. 



Large specimens of J. juncea and complete colonies of J. gemmacea are 

 unmistakable on account of the great diiference in the nature of the 

 colonies. 



In the former the colonies are always simple or sub-simple, while in the 

 latter they are very much branched and markedly dendriform. Solely for 

 this reason has it been considered justifiable to maintain these as distinct 

 species. 



The spicules are identical both in type and measurements ; but the 

 extreme nature of the branching, which commences almost at the very base 

 of the colony in the case of J. gemmacea, and the normally simple character 

 of J. juncea, and the great length and size to which colonies of the latter 

 species may attain, seem to justify theii" recognition as distinct species. It 

 must be noted, however, that it would be extremely inad%dsable to attempt to 

 distinguish between one of the long terminal twigs of J. gemmacea and the 

 tip of a colony of about the same thickness as J. juncea. For this reason it is 

 difficult to decide exactly to what species certain records refer when these 

 have been based on fragments. 



In J. juncea there are two externally different tj^es, but morphologically 

 these are the same. In one of these the lateral bare tracts which correspond 

 to the two main longitudinal canals are evident throughout, but in the other 

 there is no trace of these. 



The importance of this has been discussed under the species ; and it has 

 been considered highly inadvisable to separate them, imless as varieties. 

 This step has been taken only to obviate any future misapprehension. 



I would therefore suggest the following classification : — 



Species of JunceUa. 

 A. Longitudinal main canals tico in number. 



(1) Colony simple, flagelliform — J. juncea Valenciennes, emend. 



(2) Colony much branched and somewhat bushy ; branches 



liform— J", gemvmcea Valenciennes, emend. 



