2g: 



F. BORGESEN 



On a shell a small tuft of a tiny Amphiroa was found which I am unable 

 to refer to any known species. I name it A. Yendoi in memory of my regretted 

 Japanese colleague, who has published so many valuable papers on the Corallinaceae. 



The plant was not much more than I cm high; its thallus is terete, rather 

 stiff, erect, much calcified, about 250 — 300 [j. in diameter. It is regularly 

 constricted, broader portions, about twice as wide as high, regularly alternating 

 with slight annular constrictions (comp. Fig. 33 b). Generally every third 

 constriction represents a node (Fig. 33 b), but sometimes the joint is composed 

 of more than three segments, 4— 5 or even more having been found. 



A longitudinal section of the thallus (Fig. 33 c) shows that the central 

 strand is rather regularly stratified, a sheet of long cells being followed by 

 another of short ones, this again by intermediate ones, then a second layer of 

 long cells and so forth. 



This central strand is densely covered by a thick cortical layer formed 

 by small cells. The nodes correspond to a layer of long cells (Fig. 33 c); the 

 cortical layer splits along the middle of this layer and the cells get a more 

 corneous consistence. Within the joints, the constrictions are, too, found just 

 above the row of long cells. The long cells measure c. 80 jjl, the medium 

 c. 35 [J. and the short c. 15 \>.. 



Regarding the ramification I can only state that nearly all the shoots were 

 simple. Only a few of the longest had divided subdichotomously. The plant 

 was quite sterile. 



Yendo and especially Mme. WEBER have shown that the anatomical 

 structure of the central strand is of great systematic importance in the Coral- 

 linaceae, the genus Amphiroa being characterised by 1, 2, 3 or more rows of 

 long cells alternating with a row of short ones. Further, the number of cell rows 

 in the node is of systematic value. 



As stated above the present species has one cell row only in the nodes 

 and according to the analytical key found in Mme. Weber's work on the 

 Corallinaceae of the Siboga, p. 99, this is the case only in one species, viz. 

 Amphiroa valonioides Yendo. This species is described by YENDO in his paper: 

 »Corallinae verae Japonicae» (Journal of the College of Science, vol. 16, Tokyo 

 1902). To judge from his description and figures this resembles the present 

 species in size and general appearance but seems to be more branched (as 

 mentioned above, the specimen of A. Yendoi is mostly unbranched), the regular 

 constrictions in the latter are wanting and, in the central strand, there are 

 according to fig. 3, pi. 1 of YENDO four rows of different length regularly 

 alternating, while only three rows of cells are found in our plant as described above. 



Area of distribution: Endemic. 



Jania Lamx. 

 J. tenella Kiitz. 



Kutzing, Tab. Phycol., p. 41, tab. 85, fig. II. Weber, A. and M. Foslie, The 

 Corallinaceae of the Siboga Expedition, p. 108. 



I have referred some small, 4—5 mm high repeatedly dichotomous plants 

 to this species. The joints are about 100 ;x broad and 2 — 3 times as long in 



