■J'ln: FUOACEiE OK JAPAN. ll.J 



leaf-scar^. Leaves are, as a I'lile, narrowly lanceolate, acuminated 

 above, and gradually attenuated towards the base into a delicate 

 filiform petiole. The margins are almost always entire l)ut some- 

 times coarsely dentated in the upper portion of the leaf. A few 

 obvious, elevated cryptostomata occur on the leaves. 



Vesicles are elliptical, apiculated above, and acuminated be- 

 low into a loni>- filiform stipe. They are disposed racemosely on 

 the lateral branches on which not a single leaf, except the 

 fiilcraiit leaf, i.s usually to be found. And on the ultimate 

 ramulets the fulcrant leaf itself is always transformed into a 

 vesicle. Hence when a plant is sterile there is danger of its 

 being mistaken for a Cysiophyllum. 



Receptacles are cylindrical, unarmed, tapering above, with 

 a short filiform stipe; borne solitarily on each stipe of a vesicle 

 or terminal on the ramulets. 



Remark on the affinity. The plant has a close resemblance 

 to Sargassum confiisum on the one side, and to Sargasswn 

 Thunhergii on the other. One is liable to refer the young 

 shoots, densely covered as they are with small leaves, to either 

 of these two species, oi- to Scü^gassum Kjellmanianum. A fertile 

 branch, however, readily distinguishes the present species from 

 the others. 



The disposition of the receptacles and vesicles of this species 

 is somewhat alien-ant among the Bactroj)hycus. It may be taken 

 as an extreme form of the Racemoscc^ standing near Sargassum 

 nipponicum, with the receptacular ramulets reduced to single 

 rece2)tacles only. But I am inclined to think it an ally of Sar- 

 gassum confusum. The essential difference between them lies in 

 the fact that in the former the bractlets on the ultimate fertile 

 ramulets are invariably transformed into vesicles. Such a modi- 



