16 STUDIES IN ROOT-PARASITISM. 
earliest case examined (Plate III, fig. 1) a peripheral layer of 
irregularly arranged cells surrounds the actively growing part of 
the haustorium, and in this layer the cells are clear of starch. 
In the radiating cells of the growing part, on the other hand, 
starch is abundant, but it is denser in the outer cells, those nearer 
the host's root being clearer. This general arrangement has 
been met with in all subsequent stages of development, the densest 
layer being, however, formed further in in successive stages. 
It is fair to assume that the outermost clear cells have yielded 
up such starch as they had because they have ceased to take part 
in the life of the haustorium and (from the accumulation of 
calcium oxalate crystals) become moribund, merely functioning 
as a protective layer to the tissues within. Growth in the haus- 
torium being centrifugal, it is natural to suppose that the com- 
parative clearness of the inner layersis partly due to the cells 
being in a more meristematic condition. They are actively divid- 
ing and have protoplasmic contents. The deposition of masses 
of starch will presumably indicate that the meristematic condition 
has been passed or is in abeyance. The absence of dense masses 
of starch in the inner zone is thus readily explained, as also is 
the constant inward movement of the region where the masses 
of starch are densest, the outer cells becoming in turn permanent 
or moribund and yielding their starch to those within them as the 
latter lose their early meristematic condition. This region of 
greatest density is seen in thick sections as a dark shadow and 
may be followed by comparing the figures on Plates III—VI. 
But while the comparative clearness of the inner layers 
presents no great difficulty in our study, when the gland appears 
we are confronted by a very puzzling phenomenon. The line of 
demarcation between dense and clear regions, at first very indefin- 
ite and difficult to locate, becomes sharp and the transition 
sudden, when the gland becomes mature. A_ perfectly clear 
inner zone with watery contents is differentiated, and this zone is 
bounded outwards by cells containing the densest masses of 
starch, the line of demarcation being thus intensified to the last 
degree. We have, if possible, to account for this curious fact, 
