2 I 



forma hederacca. 



Joints trilobed, not deeply cut, lobes spreading in the form of an ivy leaf, ribs distinct (fig. 23). 



forma elongata. 



H. cuneata var. elongata Bart. 1. c. 

 Joints trilobed, not deeply cut, not spreading, side lobes sometimes reduced to slight 

 enlargement at or above the middle of a joint (fig. 24). 



forma Renschii. 



H. Renschii Hauck 1. c. 

 Plant forming compact, circukir tuft, branches radiating from the centre. Upper joints 

 small, thin, irregularly round or transversely ovate, indistinctly ribbed (fig. 22 and 22a). 



This species from its extreme variability was formerly divided into several species, varieties 

 and forms, and indeed the variations are so great that viewed apart from intermediate forms 

 and without regard to internal characters, the extreme forms would certainly have to be regarded 

 as distinct species. But in consequence of being fortunate enough to have had in my hands 

 a large number of plants of this species, it has been possible to tracé the gradations which 

 lead from one form to another and so to link together even the most extreme forms. 



The manner of fusion of the central filaments at the apex of each joint forms an un- 

 failing test for H. Opuntia. It is the only species in which fusion of filaments takes place in 

 pairs and is at the same time incomplete, being limited to the surface of contact of the lateral 

 walls. Communication between each pair of filaments is effected by a breaking down of the 

 side walls at the points, where they become pressed together at the apex of each joint. Above 

 this point however the two fused filaments are again quite free and remain for a certain distance 

 unbranched, till they again give rise to the subcortical tissue to form the next joint (fig. 25). 



This method of fusion is also found in the type specimens of H. triloóa, H. cordata 

 and H. Renschii, which are therefore here placed under H. Opuntia ; but since they indicate 

 the various external directions of variation within the species, they are retained as forms in 

 accordance with the method already indicated. 



In a paper by myself, published in 1900 (Forms of Halimeda from Funafuti. Journ. etc), 

 a new variety is described under the name of H. cuneata Kütz. var. elongata. An examination 

 of the central strand of the plant here described shews that the manner of fusion is that cha- 

 racteristic of H. Opuntia and since the external form is fairly well marked it is included here 

 as H. Opuntia forma elongata. (The type specimen of H. cuneata Kütz. is shewn below to be 

 a plant of H. graei/is Harv.). 



In the same paper it is stated that the single joints of Halimeda, which were brought 

 up in the borings made into the floor of the lagoon, are fragments of H. Opuntia var. macropus. 

 This statement was made as the result of classification based on the size of the peripheral 

 cells and it was only after examination of the central strand that I was able to discover the 

 mistake. The joints in question shew filaments with a mode of fusion characteristic ol H. Opuntia 

 and in external form they come nearest to either forma triloba or forma cordata. 



