54 THE JOURNAL OF BOTANY 



normal occupants of the channel, visible running water is only occa- 

 sionally seen, upon the rare coming of flood. In such a river as the 

 Great Jb^ish liiver, at Seeheim, the river has cut itself a caiion-like 

 path through horizontally-bedded rocks. The gorge is wide, one 

 hundred yards and upwards; the floor is occupied by sand, with 

 occasional protrusion of rock ; the water, when it flows, may course 

 from side to side In lateral channel. Trees of considerable size grow 

 in the sand-bed out of reach of the ordinary flood- water. Such trees 

 take chances and may be overthrown in times of exceptional flood : in 

 the gorge they find water and shelter and the risk is incurred. Many of 

 these trees are Acacias, but there are others the writer could not identify. 

 One, a considerable shrub, showed an interesting feature. It had yellow 

 flowers, about an inch in diameter ; surrounding the superior ovary 

 was a gro^^^h of Avhite silk hair, enclosed within a membranous 

 bladder-like envelope.* After fertilization this envelope ruptures, the 

 hairs develop in size and strength, become brownish in colour, form- 

 ing regular rows upon the surface of the ovary, and serve to disperse 

 the wind-borne fruit. Hound about Seeheim, Gomphocarpits fruti- 

 cosus grew in profusion. This Asclepiad is a pestilent weed and 

 overruns many parts of Africa. 



The sand-river varies very much, according to the type of country 

 in which it occurs, and we find its fades varying with the slope. If 

 the declivity be small and the valley wide, the sand-river spreads 

 itself out over a wide area. Charged with water, the sand, in time 

 of flood, becomes mobile, be the fall ever so slight ; but upon a 

 considerable slope it does not, when dry, lose all mobility. In areas 

 occupied by ancient crystalline rocks, where the rainfall is a negligible 

 quantity, one may see a sand- river system very well. Small sand-' 

 streams course down the higher tributary valleys, join up with 

 others, and finally discharge themselves into the sand-river bed of the 

 main valley. This may be confined within definite banks of rock, 

 but, If the slope be slight, jdooIs and lakes of sand — if the terms be 

 admissible — are formed. A wide flat valley may be filled with such 

 a lake of sand, which at the margins may creep up the sides of the 

 enclosing hills like water up the sides of a burette. 



The sand-river system In a mountainous area is comparable to 

 a glacier-system. The sand-river is always there ; water may be but 

 rarely present, but some tree with deeply-striking roots may tell 

 of underground suppl3^ 



Many of the plants one sees are the desj^alr of the collector ; 

 they cannot be made to He decorously between sheets of botanical 

 drying-paper. The fleshy types have been so well protected by nature 

 against drying that it is diiiicult to preserve them In a way that will 

 give any Idea of the plant in its normal surroundings. Take Sarco- 

 caulon, two species of which were seen In flower ; it is one of the 

 hardiest of the desert plants. One species had pink flowers; the 

 other one seen had flowers of pale yellow. The fleshy stem bristles 

 with formidable spines. In full bloom it is a beautiful vision, and 

 only sketch or photograph could picture it. A flowering-plant in the 

 desert is a precious thing, for a plant speaks of life and comrade- 

 ship, as Mungo Park, the well-beloved, found in the long ago. 



