IO 



An examination of a Harvey specimen of //. gracilis^ which often resembles II. Opuntia 

 externally, shewed that though the filaments ;ilso unite in pairs at the apex of each joint, as 

 in //. Opuntia. the fusion is complete and the two filaments continue their course as one. This 

 filament remains unbranched in the node, but branches above, always trichotomously^ to form 

 the new joint. (fig. 30). 



Most of the forms ascribed to //. Tuna exhibited still another type of fusion, somewhat 

 resembling that of //. gracilis. The same complete fusion takes place between adjacent filaments 

 at the apex of each joint, but the number of filaments which fuse together may be two or 

 three, and instances of both occur in the same joint. The filament, resulting from the fusion, 

 branches later di-or trichotomously, no regularity being observed. (figs. 4% 4 1 '). 



In some plants however, owing to the large surface of the filaments in contact at the 

 apex of the joint, they remain attached to each other after treatment and do not easily drop 

 apart as in the plants of H. Tuna. 



Mam- forms ascribed to II. cuucata shewed a similar type of fusion to that of the plants 

 of IT. Tuna. 



An examination of all but two of the various species which have been described, shewed 

 that there was a definite number of types into which all the various methods of intercommu- 

 nication of the central filaments in Halimeda could be arranged. 



It is obvious then that such a character serves as a convenient basis for a classi- 

 fication of the genus and by its use (and by the help of the character of the peripheral cells 

 in two cases) I have been able to define seven distinct species. 



Classification by means of this character does not run counter to the old lines of classification, 

 many of the older species being, in the main, rightly defined, but their limits were far too narrow. 



The species having been clearly defined, it was possible to gain an idea of the range of 

 variation within the species limit. 



The great range of variation, which appears to be so common in marine forms of life, has 

 been a source of much confusion to workers at this genus. Owing to lack of opportunity of 

 examining large numbers of plants together, it was difficult for them to realise the existence 

 of many connecting links, and plants were described as new species which a large collection 

 would have shewn to be but varying forms of existing species. So complete is the series of 

 links connecting all the forms of variation that no form can be regarded as constitutiner a 

 definite "variety"; but for practical purposes it is advisable to indicate the chief lines of variation 

 within the species, defined by structural characters. The older species and varieties being based 

 on external characters serve fairly well for this purpose, so that for the present I have kept 

 them up as forms under the species as they now stand ; and where forms shewing a new line 

 of variation have not hitherto been clearly distinguished, I have placed them under a new 

 form name. By arranging thus the various forms under the species, I have tried to give some 

 idea of the extent and different lines of variation within the species limit. 



Under "forma typica" 1 have always given the characters of the first described form of 

 the species. 



