36 STUDIES IN ROOT-PARASITISM. 
the sucker. It is probable that something of this kind also 
takes place in Santalum, but at present it has only been noted 
that a few of the cortical ceils near the sucker are dragged in 
with the penetrating nucleus. In Ola# a similar subdivision 
into starchy and protoplasmic cells takes place, but these 
cells belong to the cortex and nucleus respectively. 
An endodermis is very clearly seen in most haustoria. It 
accompanies the branch of bundles from the mother root and 
is found round each of its subdivisions in the transitional 
region, and ceases at the interrupted zone. The latter zone 
is especially clear in Cansjera and is evidently centrifugal in its 
development. 
The activity of the cambium varies in different parts of 
the haustorium. In the axis it soon ceases to form wood, but 
for a long time forms layers of cells outwards which may be 
regarded as secondary cortex. In the vascular loop a great 
development of xylem takes place, with the result that the 
vascular strands assume a characteristic and peculiar shape. 
The older haustoria are chiefly composed of xylem, and it is 
chiefly due to its development that the form of such old haus- 
toria is turbinate. In Santalum, although there is a similar 
development of secondary cortex, the formation of wood 
takes place equally in the axis and the loop. 
The sucker is narrow in Cansjera and little lobed, resem- 
bling the cortical folds in this limited growth. An abundance 
of yellow “gummy ’’ substance is found in different parts of the 
haustorium from the decomposition of its cells. This extends 
to the collapsed layers whose decay is seen to be centripetal 
and sometimes complete. Such decayed tissues are frequent- 
ly surrounded by meristematic divisions in the neighbouring 
cells. A brilliant cap of clear yellow substance is found over 
the end of each of the host’s wings. This formation of yellow 
substance and internal meristems is much rarer in Santalum. 
BoTANY SCHOOL, CAMBRIDGE; 
30th December, 1907. 
