Tlwnningia.] cxxi. balanophore.t: (botting hemsley). 437 



times obsolete. Stamens o-5, or perliaps^ sometimes more, united in 

 a column ; anthers normally 12-celle(l, sometimes imperfect, or moie or 

 less embedded in the column, dehiscing longitudinally ; cells sometimes 

 irregular in number, length and height ; pollen subglobose. Receptacle 

 usually bracteolate. Female flowers very small and very numerous, 

 occupying the whole head or surrounded by one or more series of male 

 flowers, often exceeding 10,000 in a head. Perianth adherent, pro- 

 duced in a cylindrical tube above the ovary, 2-i3-lobed ; lobes unequal, 

 forming an oblique limb, or equal or almost obsolete. Ovary inferior, 

 1-celled, cylindrical ; style exserted, about the same length as the ovary, 

 fiUform or capillary ; ovule solitary. Fruit compound, fleshy, highly 

 coloured, strongly resembling a strawberry in external appearance, 

 sometimes as much as 2 in. in diam. Seed imperfectly known ; not 

 present in any of the specimens examined. — Herbs parasitic on the 

 roots of trees or shrubs, perennial, forming tubers at the points of 

 attachment to the host and emitting more or less branched cylindrical 

 usually hairy rhizomes, bearing sessile or more or less pedunculate 

 flower heads ; fleshy or coriaceous, destitute of chlorophyll, but brightly 

 or brilliantly coloured. Leaves none. Peduncles densely clothed with 

 imbricating rigid sharp-pointed scales. Heads solitary ; involucral 

 scales similar to and continuous with those on the peduncles. 



Specjes G, all eiideniic. 



Thonninrjia, excluding T. malagasica, Fawcett {Trans. Linn. Soc. ser. 2, 

 Bot. ii. 23y, f. 36) is restricted to tropical Africa. Tlie floral structure of the 

 Madagascar plant is so different that a revision of the whole order would i)robably 

 lead to giving it generic rank. In some respects the flowers are more like those of 

 some of the species placed in Bala?iophora itself ; the males having a regular perianth 

 enclosing the staminal column, and the females being destitute of a style which is a 

 conspicuous feature in Thuuninjia. 



Eichler, in DC. Frodr. xvii, l-J-l, describes the male flowers of Thonningia as 

 pedicellate; but as sonje of the lobes of the imperfect i)erianth are sometimes given 

 off near the base of tliis axis, or are easily detached to its base, it secuos more 

 consistent to treat it as a part of the staminal column. 



Hitherto botanists, with the exception of Prof. Lecomte, the author of T. sessilis, 

 have referred all the specimens of Thonningia WviXXvxse come under their observation 

 to the original T. sanguinea. Thus in Engler (FJl. Ost-Afr. C. 169) and Durand 

 {Sylloge FlorcB CongolancB,476) and in the herbaria consulted all the specimens were 

 under this name ; but excellent specimens in alcohol at Kew, collected by Monteiro 

 in Angola, and by Dawe in Uganda, at once suggested plurality of species, which was 

 confirmed by examination. Mrs. Talbot's excellent coloured drawings at the British 

 Museum, the only ones seen, were a great help, though unfortunately the specimens 

 drawn were attacked by insects before being put into spirit and the male flowers, 

 especially, almost entirely destroyed. Possibly segregation has been carried too far; 

 but it was considered better to err in this direction rather than confuse two species 

 under one name. Complete specimens of both sexes are much needed. The tubers 

 of most of tlie species are unknown, and no inlormation concerning the host-plant is 

 forthcoming. Kipe seed is wanting in all the specimens examined, except the 

 doubtful T. malagasica. 



We are indebted to Dr. A. E. Rendle for facilities for comparing the British 

 Museum material with that at Kew, 



