22 THE JOURNAL OF BOTANY 
I was fortunate in getting one day (Feb. 19th) upon the. Maga- 
liesberg range, just to the east of Pretoria. The formation is mainly 
sandstone and quartzite, but the dip-slope faces the north, whereas 
in the Witwatersrand range the dip is to the south. The granite 
quartzite rocks, The shales of the Witwatersrand series carry mu 
iron, and when they lie close beneath the surface the grass is apt 
to be thin over them 
On the dip-slope of the Magaliesberg I found St a wines 
Combretum sp., conan ae Kraussiana, all common far north 1 
s har “ee 
species. 
Several weeks in June were spent in the Se! sabe tere 
about sixty miles to the south-east of Johannesburg. There 
large hills in this neighbourho a composed of mente basic lavil 
fiows ; an district probably carries a more varied flora in spring 
and summer than is seen Averary Siussiinen Where unburnt 
the ike was dry and brown. It was noticeable that two plants so 
et 
AC) 
is 
= 
-9 
3 
x, 
2 
< 
= 
bac 
ag 
i) 
Qu 
& 
— 
ie) 
09 . 
° 
i=] 
m 
Qu 
oO 
< 
ee 
oa 
oO 
na 
- 
ed 
are 
a 
Oo 
ee 
et 
ae) 
i) 
and so their dispersal bs oe tern: ee The idea may be to tinixis 
too broad a distribution, conceivably undesirable, as separating in- 
dividuals too widely to admit of ready cross-fertilization. Where the 
heap is caught up by a gust of wind, the individual achenes would 
be likely to be dropped in close ia ap 
Eucomis sp. (1828, 1329), is common n the open veld, and 
also in precipitous places upon the hills wound non No. 1828 i 
the veld form. No. 1829 the hill form, which has larger scape and 
withering paeitt e scape, in the case of the veld form, lies 
upon the veld alongside, to little Ba as it would seem. 
ae as in the hill form, the plan wing upon a — slope, 
foot of the hill. But mee are “lentefil at the higher levels, never- 
theless, and, as the seeds carry no sail, animal agency must be 
called in to get them up the hill again. 
Barleria macrostegia (7245). — The grouped capsules with their 
