WAYFARING NOTES FROM THE TRANSVAAL 835 
granite, affording a sneer: series of natural sections. By 
the stream there grow, among other water- -loving plants, several 
very fine tree-ferns (yaithen Dregei Kunze (1174) ), here well 
sheltered. 
Karly in November, 1902, I wrote, ‘The flowers of Vellozia 
have come and gone”; this was true of that detachment, but a 
e Gladiol, "which I thought had pa ee Sgr 
apts) the end of November. The bri . crimson on mall- 
bs de in evidence, e.g. the feathered achenes of Clematis Stanleyi, 
arranged in a mop-like ball; the glistening, silky plumes of variou s 
Asclepiads, which upon rm? windy days are very striking ; and 
the se ; of the various grasses. 
One is tempted to ask why some of the follicles of the Asclepiad- 
ace@ aul be inflated, others not. One may fairly assume that the 
air-chamber gpebo as a non-conductor and is protective. The 
follicle is very usually borne erect and freely exposed. It seems 
possible that the beat of midday may increase the se n within 
capsule from expansion of the contained air, and this, combined 
with the wind-factor, may determine the moment at whic e 
capsule bursts. The inflated capsule, which i is moreover sometimes 
winged, presents an extended surface to the wind. One notes that 
in various species the stalk of the follicle, sharply bent, lies close 
to the flattened axial side of the follicle, the outer side of the 
follicle being more bulging and gibbous, or it may form a bracket- 
like curve. The attachment of stalk to follicle is broad and firm 
If pressure be exercised by the wind, or, as may be imitated by the 
enger, upon the inner flattened Wee. of the follicle, so pushing it 
way from the vertical eng oy of the stalk, the broad base of 
ee a rene: ttached b ‘heir ir tips their expansion 
dry up, but as remain a y 
pe A : = hese cicks, balling between root and tip which lifts them 
