2& STUDIES IN ROOT-PARASITISM. 
The vessels in the axis and the sucker behave much as in 
Santalum and Olax, with the exception already noted that there 
is comparatively little secondary thickening. As the host’s 
tissue is approached, they form sepa-ate rows and apply them- 
selves to different parts of the woody tissue. As in the other 
haustoria, there is no special tendency for these rows to apply 
themselves to the vessels of the host and, in most cases, no 
definite connection can be noted between the vessels of host 
and pacasite. Occasionally, however, we find vessels formed 
in the sucker accurately opposite to the vessels of the host. 
Two such cases are figured on Plate X (figs. 1 and 4) and 
call to mind a similar example mentioned in our study of 
Sanialum (Sant. ITI, 29). 
A few lines, in conclusion, may be devoted to the yellow 
‘‘secretion’’ so oltenreferredto as occurring between the sucker 
and the host’s tissues, for some additional light has been thrown 
on its origin and structure by the examination of Cansjera 
sections (Compare Plate X, fig. 2, with Sant. I7, page 31, and 
Plate X, figs. 1—3, and Olax, Plate X, fig. 2). In the purely 
elandular cells and the lumen of the gland the yellow trans- 
parent substance is obviously the direct product of the proto- 
plasm accumulating beneath the cuticle and being discharged 
at intervals. A yellowish colour is not infrequently assumed 
by discharges of this kind (Marshall Ward, for instance, noted 
this colour in the secretion from Botrytis hyphe on Lilium 
candidum). In the more diffused action of the general 
haustorial surface, the yellow substance has not infrequently 
traces of structure, many instances of striation or lamination 
having been noted. This lamination appears to be caused by 
the breaking down of successive cells (refer also to paragraph 10 
where the yellow substance foimed by the breaking down of 
the cells of the collapsed layer are clearly continuous with 
the yellow secretion at the end of the sucker). In some cases 
these cells belong to the sucker, and it is natural to suppose 
that they have been dissolved by the action of the cells behind 
them, In othercases the yellow layer is evidently formed in part 
