GERMINATION AND HISTOLOGY OF WELWITSCHIA MIRABILIS. 25 
separate, the bundle sheath makes its appearance. In the 
young root the latter is cuticularised on its radial walls only 
(fig. 21). It is developed from a layer of cells removed from 
the fibro-vascular tissue by a pericambium of two or three 
layers of cells in thickness; these are the direct outcome 
from the proto-meristem, and appear not to be formed by 
subsequent division. 
In a transverse section of a young root, at a point not far 
removed from that of transition, we find a structure as re- 
presented in fig. 21. On the outside will be seen a thick 
covering of tissue, whose cell walls are much swollen, and 
the cells often elongated and intertwined among one another. 
This is especially well seen on the outside of fig. 16. This 
covering may be found partial or entire. The older the 
root the more incomplete is the covering. In fig. 21 it 
may be seen that it is in the course of being thrown off 
from the underlying cortical tissue by a mucilaginous 
swelling of the cell walls. This tissue, thus thrown off, 
is the root cap. When the underlying cortical tissue is 
thus laid bare, the single cells of it grow out (fig. 21), 
and apparently form root hairs. I have, however, never 
succeeded in finding these in close attachment to the soil. 
The cell walls of the outer cells of the cortical tissue 
usually become cuticularised. 
As the root grows older the cuticularisation of the walls 
of the bundle sheath extends from the radial to the tangential 
walls (fig. 22). The change happens first opposite the 
phloem masses. The result of this is the cutting off of the 
cortical tissue from a physiological connection with the central 
cylinder. Accordingly it begins to break down and separate 
from the latter; and, as in fig. 22, the process is first seen 
opposite the phloem masses. The final result is that the 
central cylinder is completely laid bare. 
We may now turn to the consideration of the central 
cylinder. As before mentioned, the bundle sheath is sepa- 
rated from the vascular tissues by a pericambium, two or 
three layer of cells thick. Within this are arranged, 
according to the usual root type, the two xylem and two 
phloem masses. With respect to the constituents of these 
I have no remarks to offer, except that the phloem tissues are, 
from a very early stage, compressed, so that their cell walls 
are irregularly folded. There is at the centre a parenchy- 
matous pith, and I have never observed the connection of 
the two xylem masses centrally, so as to form a single plate.! 
Cf. “ Description of Bertrand,” who has already investigated the 
-ateral roots of Welwitschia, ‘ Ann. d. Sci. Nat,’ Série v—xx. 
