1120 Michell. — On the Comparative Anatomy of the 
6. No tangential bands of paren- 6. Tangential bands of parenchyma 
chyma in the xylem. in the xylem. 
7. Large quantities of oil present. 7. Very few oil-drops in the cortex. 
8. Scale leaves protecting the 8. No scale leaves present, 
growing regions during the resting 
period. 
Nodes. 
The leaves fall off when the plant is preserved in spirit, and therefore 
the exact position of the leaves at each node could not be ascertained. 
A section through a node showed a large number of scale leaves, each con- 
sisting of mucilage cells at the base and a dark-brown suberized mass at 
the apex. A growing point seems to be situated at each node, and in 
a large number of cases in the material worked upon this has given rise 
to an inflorescence. In order to investigate this point thoroughly it is 
necessary to have materials obtained at the different seasons, and this is 
impossible at present owing to the long journey which must be undertaken 
to reach the home of these plants. The following suggestion may be made 
as to the changes taking place. 
In the axil of each leaf a growing point arises protected by scale 
leaves. These fall off as soon as the need for protection ceases, and the 
bud gives rise to a short branch bearing a number of closely packed leaves. 
At the end of the vegetative season scale leaves again arise to protect the 
growing point, and, as before, these fall off and the short branch terminates 
in an inflorescence. 
In C. namaquensis , of which stem tips were available, scale leaves 
exactly like those of C. gariepina were found on all the nodes and also sur- 
rounding the stem apex. Portulacaria afra has no scale leaves, and the 
foliage leaves arise in a decussate manner at the nodes. Structurally there 
is no great difference between Ceraria and Portulacaria , and in comparing 
the two plants one is struck with the resemblances rather than with the 
dissimilarities. From the anatomy of these plants, therefore, one is led to 
conclude that a close relationship exists between these two genera. 
In conclusion I wish to express my thanks to Dr. H. H. W. Pearson 
for supplying the material, and for many helpful criticisms during this 
investigation. 
I am also indebted to the publishers of the ‘ Gardeners’ Chronicle ’ for 
the loan of the block from which Text-fig. 1 was taken. 
Summary. 
1. In Ceraria the nodes are arranged in four longitudinal rows, many 
leaves being borne at each node. Portulacaria has decussate leaves. 
2. Scale leaves occur in Ceraria at the nodes and at the stem apex, 
