Cc. A. BARBER. £5 
packed, and pass by easy transitions into the rest of the nucleus 
(fig. 2a). The clear space, both above and below, is seen to be 
traversed by irregularly jointed, dark parallel lines, and as these 
can without difficulty be traced into the rows of cells at the sides 
composing the wall of the gland, it is fair to assume that they 
consist of the contents of broken down glandular cells. These 
are shown in Plate IV, figs. 2a and 2b, which illustrate the lowe 
and upper halves of the gland lumen respectively. 
The next change which takes place in the gland is the great 
development of its lower half. This rapidly increases in breadth 
and becomes lobed. At the same time its walls become less 
distinct and the separation of the upper and lower halves becomes 
more pronounced (Plate IV, fig. 3). This stage does not, how- 
ever, last long. While the upper half retains its cap, the lower 
increases in width until little if any of the nucleus is left at the 
sides. At the same time the gland tissues in this region become 
completely disorganised, and a large lumen is formed filled with 
a glairy substance and occasional indications of what were 
once rows of secreting cells (see Plate V, fig. e, where such 
cell remains are drawn). Meantime the shadow spoken of in des- 
cribing figure 2 has become mucn more definite and, when care- 
fully examined, proves to be a well-defined channel through the 
dense, compressed cells of the lowest part of the nucleus to the 
tissues beyond, whether of haustorium or host. This is the duct 
of the gland and is constantly present (see also Plate V, fig. d). 
We see then that, while in Santalwm the gland is formed by 
the separation of two sets of cells by a subcuticular secretion 
formed layer by layer until the lumen is filled (Santalum, Part I, 
Plate V), in Olax long lines of cells are decomposed, the contents 
becoming more and more attenuated until they disappear entirely 
in a glairy matrix. The gland in Santalum is more or less schi- 
zogenous, while that of Olas is essentially lysigenous. 
10. In the stages thus far examined we have again and 
again had occasion to draw attention to the peculiar way in whic 
the starchy and clear areas of the cortex are arranged. In the 
2 
