52 
EXPLANATION OF PLATE, 
Figs. 1 and 2 nat. size, the remainder highly magnified. 
Fig. 1. Root of Sweet Pea injured by the fungus. 
Root of Cypripedium injured by the fungus. 
Conidial (Milow?a) stage. 
Free conidia, some germinating. 
Resting-spore ( Torula) stage. 
Free resting-spores. 
Perithecium of ascigerous stage. 
Ascus containing eight spores. 
Free ascospores. 
s 
Oe eee PE Oe 
IV.—AFRICAN BUXEAE. 
J. HuTcHINsSoN. 
- Baillon, in his careful monograph of the Buxaceae, published 
in 1859, regarded them as a separate order and divided the 
genus Busus containing 13 species into two sections: (i.) Huburus 
and (ii.) Tricera. These sections besides occupying separate 
geographical areas are distinguished on account of differences in 
the form of the inflorescence. In both of them the anthers are 
supported by well-developed filaments and in the male flowers a 
large rudimentary ovary is present. At the time of Baillon’s 
monograph no species was known to occur on the African continent, 
but he recorded and described one from vip Pena (B, madagas- 
earica) without, however, having seen male 
Miiller in 1869 (DC. Prodr. xvi. i. 13-20) Tedacied Baillon’s 
arrangement, adding 5 other species. 
Baillon in 1873 (Adansonia xi. 268) described B. Hildebrandiii, 
and named it after the collector who found it in the mountains of 
Somaliland. ue pointed out its close affinity with the other Old 
World speci 
Oliver in 1882 (Hook. Ic. Pl. t. 1400) described and figured the 
gon Notobuxus, from Natal, which differs from Buzxus in having 
two additional stamens, sessile anthers, and by the absence of a 
rudimentary ovary from the male flowers. 
The same author four years later (l.c. t. 1518) figured and 
described Buxus Macowan’, from Rael pti and at the 
South Africa and the examination of male flowers, all toliseted 
by Mr. —— proves that both species have the same floral 
structure 
