than in forma typica. Walls of peripheral cells directly in contact or lying in cellulose 

 matrix (üg. 29). 



This species often resembles externally certain forms of H. Opuntia and of H. iticrassata, 

 but in internal structure it is at once to be distinguished from both. The filaments of the central 

 strand in H. gracilis fuse in pairs as in H. Opuntia, but in H. gracilis the fusion is complete, 

 the two filaments continuing for a short distance as one and branching later trichotomously 

 (fig. 30). In H. incrassata there is no fusion of filaments, but communication is established by 

 means of pits; so the difference between it and H. gracilis is at once obvious in a preparation. 



There is a certain resemblance in the marmer of fusion between H. gracilis and H. Tuna ; 

 but in H. gracilis all the filaments fuse in pairs and the resulting filament branches later tri- 

 chotomously; while in H. Tuna the filaments fuse in twos or threes indiscriminately, and the 

 resulting filament branches later di-or trichotomously with equal irregularity. H. Tuna is exter- 

 nally also quite distinct from H. gracilis. 



In my paper on "Forms of Halimeda from Funafuti", (1. c.) I described, on characters 

 of external form and peripheral cells, a new species, H. laxa. The great length of the speci- 

 mens, the diffuse habit and the width of cellulose separating the peripheral cells caused me to 

 reo-ard the Funafuti plants as distinct from all hitherto described species. However, a later 

 examination of the central filaments shewed the type of fusion characteristic of H. gracilis. 

 It was difficult at first to admit that this plant which resembled a giant Amphiroa must 

 be considered as a form of H. gracilis ; but in the collections to which I had access there 

 were a few plants which were distinctly intermediate in external form and habit between 

 H. laxa and H. gracilis. Furthermore in the original description of H. gracilis (Till. Alg. 

 Syst. 1. c.) the branches are said to be "longos simpliciusculos", which character is only present 

 to a sliaht deerree in Dr. Harvey's plants from Ceylon, but is very strongly marked in the 

 Funafuti specimens. There can thus be no doubt that H. laxa must now be regarded as an 

 extreme variation, but one which comes quite clearly within the limits of the species H. gracilis. 



The straggling habit of the Funafuti plants is perhaps to be connected with the depth 

 at which they were found, 35 and 37 fathoms. Thé label attached to the specimens does not 

 state whether they were actually growing at that depth; but this is not improbable, since other 

 plants of Halimeda were found living at Funafuti at a depth ot 45 fathoms. 



In Kützing's Tab. Phyc. vol. 7, tab. 21 fig. III. a plant is figured under the name of 

 H. cuneata and the locality is given on p. 8. as Kamtschatka. A reference is given to the 

 description of H. cuneata Hering in Kützing's Species Algarum p. 505, but that description applies to 

 the true H. cuneata of Hering. The figure of the "Kamtschatka" plant is however entirely unlike 

 H. cuneata Hering and has therefore been referred to by later writers on this genus as H. cuneata 

 Kütz. Professor Askenasy (1. c.) describes under this name plants from the Anchorite Islands and in 

 De Toni's Sylloge Algarum (1. c.) descriptions are given of H. cuneata Her. and H. cuneata Kütz. 



Through the kindness of Mad. me Weber — van Bosse, the possessor of the Kützing 

 herbarium, I have been allowed to examine the original plant from which the Kützing figure 

 was taken. The internal structure and external habit are those of H. gracilis, to which species 



