) 
Genera Ceraria and Portulacaria. i i 1 7 
From what has been said it would appear that two distinct kinds of oil 
are found in the stem : 
1. The drops of oil occurring throughout the stem in the cell cavities. 
2. The oil permeating the corky walls of the bark. 
It is possible that the oil-drops have been converted from starch into 
oil during the resting period, and that these plants are of the same nature 
as those designated by A. F. W. Schimper ‘ fat trees \ 1 The fact that no 
starch but much oil is found in C. gariepina which was gathered during the 
resting period, and that C.namaqnensis which was obtained at the beginning 
of the resting period has a large quantity of starch and little oil, tends to 
support this hypothesis. 
Periderm. 
An interesting feature of Ceraria is the great development of leathery 
bark. This can easily be separated from the rest of the stem, and thus hollow 
cylinders of bark may be obtained in an unbroken condition (Text-fig. 3). 
In the bark of C. gariepina there is some inflammable substance which 
does not appear to be present in as large quantities in C. namaquensis , and 
which causes the living plant to burn easily. The results of a chemical 
investigation of the constituents of the bark are not yet available. This 
phenomenon suggests a comparison with Dictamnus albus , the Candle-plant, 2 
which secretes an extremely volatile oil. On calm hot days, if a match be 
applied the surrounding air takes fire and the plant itself burns. 
In Ceraria namaqaensis the apex of the stem is protected by a number 
of scale leaves, the base of which consists of mucilage cells similar to those 
of the stem (PI. XCVIII, Fig. 8). Immediately behind the scale leaves the 
epidermal cells appear to give rise to a cambium which persists throughout the 
life of the plants, cutting off cells on its outer side only. It cannot be stated 
with certainty that this is a cambium, although it seems probable. Only 
in three or four cases have cells been seen to have divided recently, but 
considering that the material was gathered at the end of the vegetative 
season, it would appear that the cambium had entered upon a period of 
rest. Support is lent to this supposition by an examination of two lots 
of material of Portulacaria afra obtained in the Cape Peninsula. Material 
(A) was collected in June at the commencement of vegetative activities, 
and this shows that a large number of cells of the phellogen had recently 
divided. Material (B) was gathered in January, during the resting period, and 
in this only two or three cells of the cambium had recently divided. There 
is no phelloderm formed and the number of cell rows separating the 
cambium from the large water-storing cells of the cortex is constant. As 
a rule only one row of these cells is present. The cambium first of all cuts 
1 Schimper (’03), pp. 436-7. 2 Willis (’04), p. 350 (9). 
