The Groivth-Forms of Natal Plants. 
625 
approach of winter, leaving dormant shoot apices just below the surface of 
the soil. The difficulty of separating this class from the preceding one has 
already been sufficiently emphasised. 
The total number of Hemicryptophytes is cir. 540. It is noteworthy 
that out of this number about 200 belong to the Compositae (including a 
large number of rosette and half-rosette forms), 100 to the G-ramineae, and 
about 40 to the Cyperaceae. With regard to the latter, it should be noted 
that in the wetter parts of the veld we have grasses and sedges, also 
characteristic of the vleis, and it is not easy to determine which belong to 
the Hemicryptophytes and which to the Helophytes. 
The Hemicryptophyte class, as a whole, is much more homogeneous than 
the last one, and there is therefore less need of subdivision. The rosette or 
half -rosette form is common. 
In colder countries, where the Phanerophytes are mostly deciduous, the 
Hemicryptophyte class includes a large number of woodland plants. These 
are not abundant in Natal. There is not very much undergrowth, except around 
the edge of the bush, and what there is consists mostly of Nanophanerophytes 
and Chamaephytes, especially the Acanthaceous type of the former. 
The Natal Hemicryptophytes are mostly veld plants, l)ut, as pointed out 
in former papers (1, 2), the veld of Natal has several ecological features 
similar to deciduous woods — i.e. as regards the associated plants. The fall 
of the leaves from the trees in the one case, and the burning of the grass in 
the other, exposes the ground during winter. Herbaceous plants, in both 
cases, lose their aerial parts. In spring they grow quickly, and a great 
many of them get their flowering over before the leaves appear on the trees 
or the grass grows tall enough to shade them. The next class, Greophytes, 
are similar in this respect, and further reference will l)e made to the cjuestion 
when these are dealt with. 
Another question dealt with at considerable length in former papers by 
the writer (1, 2) is that of changes in the veld due to the influence of man, 
both direct and indirect, through the agency of fire. A careful study of the 
growth -forms of the grasses, with reference to the principle underlying 
Raunkiaer's system of classification, throws a good deal of light on the 
subject. Aristida junciformis and other coarse, wiry species oust the more 
valuable fodder species, such as Aiithistiria imherhis, with the result that, 
from the farmer's standpoint, the veld deteriorates. That the change 
is due to the burning of the grass is shown by the fact that it is greatest 
along the railway, where the grass is always burned early in the year. 
The resulting " secondary " associations of Aristida, etc., are perfectly 
stable. Ayithistiria is a Chamaephyte, Aristida a Hemicryptophyte. 
Anthistiria has the innovation shoots intravagiual, protected by the sheathing 
bases of the leaves. When the plant is burned, the fire may not only 
consume the withered leaves, but may also injure the innovation buds. If 
