Eriocaulon.'] cliv. eriocaule^ (brown). 259 



chestnut, about If lin. in diam., moncecious, with the outer flowers 

 female. Involucral-bracts | lin. long, J-J lin. broad, oblong, obtuse 

 or acute, entire, very thin and membranous, fuscous, glabrous. Flower- 

 ing bracts about j lin. lon^, lanceolate or spathulate-lanceolate, very 

 acute, entire, membranous, fuscous or dull olive-green, glabrous. Re- 

 ceptacle glabrous. Female flowers pedicellate. vSepala 2, free, about 

 J lin. long, linear-lanceolate, acute, membranous-, fuscous, somewhat 

 tessellated, glabrous. Petals none. Stipes between the sepals and 

 ovary none or excessively short. Ovary flattened, orbicular in outline, 

 glabrous ; style divided to about the middle into 2 filiform branches. 

 Male flowers pedicellate, glabrous. Sepals 2 or rarely ;^, free (always ?), 

 about ^ lin. long, linear-lanceolate, acute, membranous, fuscous. Stipes 

 between the sepals and petals variable, sometimes half as long as the 

 sepals. Petals reduced to 3 very minute ovate white segments about 

 TV~i ^^^- ^ong, bearing a black gland at their middle. Stamens -4 or 

 6 ; anthers yellowish. 



JtO'weT Crulnea* German South-west Africa : Ambolaiid ; on the margin of 

 a pool at Uashitenga, near Olukonda, Schinz, 859 ! 



I do not find that the bracts are lacerate tis stated in tlie original description, 

 and it is only the outermost or involucral bracts that nre sometimes obtuse ; both in 

 the type specimen (which Prof. Schinz has kindly allowed me to examine) and in 

 the example at Kew they are as described nbove. I find only 2 sepals present in the 

 female flowers, but the male flowers seem to be very variable in tlie number of their 

 parts, some having 2 sepals, 3 petals, and 4 stamens, others 2 sepals, 3 petals and 6 

 stamens, whilst a few have 3 sepals, 3 petals and 6 stamens. 



38. E. Stuhlmanni, N. E. Br. A very small plant. Leaves 

 almost capillary. Heads glabrous (probably pale brown). Involucral- 

 bracts about as long as the head. Female flower without petals. Sepals 

 unequal, narrowly linear. Male flower with the sepals connate into a 

 spathulate body, trifid at the apex. Anthers white. — E. sexangulare, 

 Ruhland in Engl. Jahrb. xxvii. 83, partly, not of Linn. 



Mozamb. Dlst. German East Africa : Eastern Uzinja, Stuhlmann, 3552. 



This plant is unhesitatingly referred by Ruhland to E. sexangulare, Linn., but 

 that species grows 12-16 in. high and the female flowers have very distinct and 

 rather peculinr petals. I have not seen the plant, but think it possible that 

 Ruhland made the comparison with E. sieboldianum, Sieb. & Zucc. {E. sexangulare, 

 Mart., not of Linn.) to which, from the description, it appears to be closely 

 related; but I doubt ils identity with that plant. The only African species witli 

 which it can be compared is E. amhoense, Schinz, from which it differs (according 

 to the characters given by Ruhland in his key to the species) by the connate sepals 

 of the male flowers. 



Imperfectly known species. 



39. E. quinquangulare, Linn. Sp. PL 87. Koernicke in Linna^a, 

 xxvii. 642, mentions that there is a specimen of this Indian species in 

 the Stockholm herbarium labelled as having been collected by Afzelius 

 in Sierra Leone, but that there has probably been some exchange of 

 labels. — Durand &l Schinz, Conspect. Fl. Afr. v. 5U3. 



