Oecological Notes on the District of Manubie, Translcei. 
43 
The first two determinations were both made in typical savannah, and 
the mean of these, about 19 "5 per cent., may be taken as the basis of com- 
parison with the other two. 
The chief point of interest is that the soil of the Cyperaceous association 
actually contained less water than that of the typical savannah. 
It is to be regretted that no determinations were made after a consider- 
able period of drought, but it may be suggested that the proximity of a 
water level would rather determine the minimum water content of a soil 
than the rate at which the content would fall to the minimum after rain . 
Thus it may be that the texture of the soil, and the type of vegetation 
caused a higher rate of evaporation in the Sedge association, but that the 
actual minimum water content reached after a long drought would have 
been higher. 
The chief species in the Sedge association referred to are Fimbristylls 
complanata,lAuk., yii'iih. long, narrow, glabrous leaves, and Pycreus umbrosus, 
Nees. A species of Moraea (Iridaceae) of the same oecological type as the 
Cyperaceae also occurs, and a Mai vaceous shrub, Jfiftiscws diver sifolius, Jaccj., 
is present. 
By small sheltered streams an association of water-loving plants occurs. 
The chief species are Cyjperus jjzdcJier, Thunb., of which three forms occur, 
Cyperus immensns, C. B. Clarke, Scirpvs palndicola, Kunth., Juncus effusus. 
Linn., Juncus exsertns, Buchen., Ohirotiia jjaliistris, Burch. (Grentianaceae), 
.and Asclepias physocaiya, Schl. 
The grasses dominating the savannah formation show certain interesting 
features in their leaf anatomy, which may be briefly described for two of the 
species. 
Erianthus Sorghum. (PI. XII, fig. 1.) 
The leaf shows certain xerophilous features, e. g. the thickness of the 
cuticle, the small size and thick walls of the guard cells, the thick walls of 
the bundle sheaths, and so on. The most striking feature, however, is the 
presence on the adaxial surface of the leaf of marked furrows at the base of 
each of which is a longitudinal band of typical-looking motor cells. These 
are large, thin-walled, epidermal cells with nuclei, but without chloroplasts, 
and it might be expected that they would cause the leaf to roll up in 
drought. Nothing of the kind happens, however, as observations show 
which were made during a prolonged drought by Miss S. van Rooyen, to 
whom thanks are due for generous assistance in this and other ways. The 
leaves remain flat and extended in all conditions of weather. Moreover, 
stomata are found on both surfaces of the leaf. These longitudinal streaks 
of large clear cells, therefore, have no motor function, and must act simply 
