17 



discoid joints of H. Tana and the peripheral cells were only slightly connected with one 

 another; but the filaments of the central strand shew the close adhesion at the apex of each 

 joint, characteristic of H. cuncata. This plant constitutes the only link I have yet seen between 

 H. Tuna and H. cuncata, but on the evidence of a single specimen it does not seem right 

 to unite the two species and I therefore keep them separate for the present. It is interesting 

 to note that up to the present H. Tuna is the only species recorded north of the tropical 

 zone, while H. cuncata is the only one recorded south of it. 



The form of the joint varies in H. cuncata from being broadly cuneate or circular, to 

 long, straight and fingerlike, as many as 9 of these straight-sided joints arising like a fringe from 

 the upper margin of one cuneate joint. This form I have called forma digitata (fig. 9). Dr. 

 Harvey (Phyc. Austr. vol. V. PI. CCLXVII) describes and figures a plant from Australia 

 under the name of H. niacroloba, and this (Alg. Austr. Exsicc. n° 562) has been taken by 

 Professor J. G. Agardh for the type of his species H. versatilis. Dr. Harvey's figure resembles 

 II. macroloba, but the specimen under n° 562 of the Harvey Australian algae in the British 

 Museum shews the internal structure of H. cuncata. In the description of H. macroloba (Phyc. 

 Austr. 1. c.) Dr. Harvey says most of the joints are broadly cuneate in Australian specimens, 

 "some of the uppermost only verging to roundish or reniform. Specimens from Singapore are 

 not dissimilar except that some of the medial articulations are cither repand or somewhat 3 — 5 

 lobed or obscurely fmgered". This "fingering" is not figured, but the expression probably 

 refers to the elongate narrow joints which are found scattered singly among the much larger 

 joints. This occurs in plants from Australia and S. Africa, even in those which in all points 

 resemble Hering's typical form. There is therefore no external character in H. versatilis to 

 distinguish it as a special form of H . cuncata and it must therefore be included under forma typica. 



In certain plants, generally those found on S. African coasts, some of the joints are 

 stalked. The central filaments, which compose the stalk and are thus freely exposed, have 

 thickened walls and are often subtorulose in shape. In these joints the point of fusion of the 

 central filaments is surrounded by a small cushion of tissue raised above the upper margin of the joint. 



Some of the specimens of H. cuneata from the Siboga Expedition have joints of which 

 the margin is very undulating, thick, recurved and grooved. These I have called forma a un- 

 dulata" (tig. 10). Up to the present I have not seen them from any other locality. 



H. cuncata is only recorded from the Indian and Pacific Oceans. It advances further 

 south than any other species of the genus, its most southerly recorded limit being Port Elizabeth 

 on the S. E. coast of Cape Colony. lts presence there is probably to be connected with the 

 warm current which flows south from the Indian Ocean to Cape Agulhas. 



3. Halimeda macrophysa Ask. 



Halimeda macrophysa Askenasy, Forschungsreise S. M. S. " Gazelle". Th. IV. Bot. Algen. 1888. 



p. 14. tab. IV. figs 1 — 4. 

 Halimeda macrophysa De Toni, Syll. Alg. vol. I. 1889. p. 520. 



Hab. Indic. — Siboga-Expedition. N. E. point of Timor ! — Paternoster Islands ! — Muaras reef! 



Pacific. — Matuku, * Gazelle" \ — Tongatabu, Graeffe\ 



SIBOGA-EXPEDITIE LX. 3 



