C. A. BARBER. 23 
numerous transitional cases occur in the body of the haustorium 
(Olax, 28 & 29). The vessels of the mother root are, as might 
be expected, much wider than those of the haustorium proper. 
In one place the vessels are peculiar, and it is doubtful 
if they have much to do with the transmission of fluids. This 
is at the top of the axis, where the two bands of vessels separ- 
ate outwards to curve round the loop. On the inner edge of 
the vascular loop at this point the vessels are broad and short, 
have in fact much the shape and size of the neighbouring cells 
of the nuclear ‘‘pith’’ (Plate VIII, figs. 2 and 3, sq.). 
Bast has not been met with in the haustorium proper of 
Cansjera. The sieve-tubes of the mother root are very small 
and difficult to locate, and little callus-formation has been met 
with in the sections, although the haustoria were collected at 
different times of the year. As far as can be seen, the sieve- 
tubes are found as far down as the end of the endodermis 
(Plate VII, fig. 2), but no trace of them has been met with 
lower down, even in cases of self-attachment. It will be 
remembered that, in Olax scandens, an abundant development 
of bast occurred throughout the haustorium in the latter case. 
In certain sections, especially in old haustoria (Plate IX, 
figs. 1 and 5, and Plate XI, fig. 3), a well-defined mass of tissue 
in regular layers is found outside the cambium. It is possible 
that a store of material is here laid up for the formation of 
further vessels in secondary thickening. On the other hand, it 
is just here that fewest vesselsareformed. The tissue has much 
the appearance of secondary cortex, the latter being usually 
crushed and dead in old haustoria, especially in the lower 
parts. <A protection would thus be afforded to the cambial 
cells within. 
16. Secondary thickening of the vascular system differs 
considerably in Cansjera haustoria from that in Olax and San- 
talum. In the latter (Sant. I/, Plate IX), wood is developed 
equally throughout the haustorium, as much in the sucker and 
axis as in the vascular loop. In Olax, however (Olax, Plate 
