548 



S((n{J]iiils. — The vegetation of the sandhills to tlie east, 

 of Ooldea is a marked contrast to that of the Nullarbor Plain. 

 In place of the d^^arf grey or white Clienopodiaceous bushes 

 of the latter, the sandhills bear a relatively luxuriant covering 

 of trees and shrubs. These differ both in their much larger 

 size; and also very much in leaf form. While the plants on 

 tlie plain have more or less spreading, though small, leaves 

 which are succulent and covered with hairs, almost all the 

 plants on the sandhills have smooth leaves which are hard in 

 texture and j^laced with their edges to the light, being either 

 pendant or vertical. The leaves are either quite glabrous and 

 polished as in Kuccdyptus, spp., and Mj/ojmrum, or grey in 

 colour, due either to wax or to a covering of very small hairs 

 that do not spread from the surface, as in several species 

 of Acacia. Even those plants such as Casuarina^ liossiaea^ 

 and others which are almost or quite leafless have tlieir assim- 

 ilatory branches erect or pendant, not spreading. This leaf 

 character applies both to the larger bushes and trees, ^ and 

 also to the smaller undershrubs which here bear, for the most 

 part, small hard leaves placed more or less vertically. This 

 difference in leaf type makes the vegetation on the two parts 

 very distinct, even when seen from long distances (pis. xxxiii., 

 xxxiv., fig. 1, and xxxv.). In botii situations it may be 

 reinarked that all the plants are evergreen ; not a single 

 deciduous plant was found. 



For purposes of description the vegetation can be divided 

 into three j^ortions : (1) the sand ridges, (2) the hollows 

 between, and (3) the basin known as Ooldea Soak. Even with 

 this division the vegetation presents at first glance a rather 

 bew^ildering lack of uniformity; many plants are apparently 

 localized in their distribution, and' situations externally very 

 similar often bear different pl^ht populations. This variability^ 

 can to some extent be explained by a recognition of the fact 

 that the sand is not unifotinly stabilized. The plaiits in 

 different places vary in their' efficiency as sand retainers. It 

 may also be correlated with the frequent limitation of areas 

 drenched by rainstorms. These "patchy" rainfalls may make 

 certain places good seedbeds, while the surrounding ar^as are 

 too dry for a high percentage germination or even any at all 

 that season. The result is that while sorne sandhills are ablaife- 

 with the flowers of an abundant annual flora, others a few 

 miles away are without any appreciable annual growth at all. 

 it is well known to pastoralists that the season at which' "ia 

 soaking rain falls profoundly affects the type of annual flbra 

 that results. It seems to us quite legitimate to assume that 

 the germination of other plants is affected also, and hence that 

 different phases of an open flora may be shown under similar 



