4 STUDIES IN ROOT-PARASITISM. 
3. Comparatively few young stages have been found in 
the numerous sections examined. In attempting to remedy 
this defect, a number of minute haustoria fixed to the roots of 
Abrus precatorius were imbedded in paraffin and cut with the 
microtome. But, unfortunately, some of the very smallest 
showed complete penetration, and most were fairly mature. 
The few cases of young haustoria met with show that the early 
stages have slight peculiarities which separate them both from 
Olax and Santalum, and it will, on that account, be necessary 
to describe the development of the haustorium. 
No haustorium has been cut in which the nucleus is not 
already indicated. The earliest is a microtome section, and 
the cells are not altogether suited for the study of their con- 
tents. Inthis haustorium (Plate I, fig. 8), there is a median 
area of small cells in the act of rapid division, and around 
this the cells are arranged in definite order. This is the first 
appearance of the nucleus and the curves of growth in the 
ground tissue around it. 
The haustorium is firmly attached to the root of the host 
(Abrus precatorius) and is fitted into all the unevennesses of 
its surface. This firm attachment of the haustorium to the 
objects it meets requires a moment’s attention, as it does not 
seem to be merely a pressure phenomenon. It appears probable 
that a gummy substance is secreted, as in Ampelopsis or the 
climbing roots of ivy and pepper, but as yet no definite proof 
of the existence of such a substance has been obtained. 
From a study of the tissues along the line of contact we 
learn several things. In the first place, a very distinct grow- 
ing edge is distinguishable (as in Santalum) on each side of the 
haustorium. The haustorium is thus seen to be growing tan- 
gentially to the surface of the host’s root. But, besides this 
growth at the edges, there is considerable increase in size in the 
middle part, resulting in a spreading movement of the tissues 
of the haustorium in contact with the root attacked. This is 
shown in the figure (8 6) by the doubling up of a piece of cork 
which is being rolled over in much the same manner as we 
