88 BOTANICAL GAZETTE [auGcusT. 
of the nucleus farthest away from the cell wall (fig. 72). While 
the beginning of the weft resembles that described for the stage 
in Iris, it soon takes on a very different form as it increases in 
size. Instead of the meshes running more or less parallel to the 
membrane of the nucleus, they elongate more at right angles to 
it (figs. 72, 13). The fibrils composing the weft stain deep blue, 
while the peripheral cytoplasm stains slightly orange. If, how- 
ever, we follow the individual fibrils outward, they gradually 
lose their property of staining blue and stain slightly orange. 
They are, in fact, strictly continuous with the more granular 
threads of the cytoplasmic reticulum. They therefore seem to 
be nothing more than transformed threads of cytoplasm. 
As already stated, the growth of the weft is not uniform; it 
is much more conspicuous on one side of the nucleus than on 
the other. This irregularity is carried still farther as the 
weft increases in size. The meshes elongate much more, the 
individual fibrils which form them lengthen considerably, and 
their individuality becomes much more pronounced. As the 
weft now increases, it does so by several projections, which 
terminate in sharp points, so that we have distinct cones formed 
(fig. 74), in much the same manner as they are formed in Iris. 
These cones at first vary considerably in number and size, and 
are invariably much more numerous on the side of the nucleus 
away from the cell wall (fig. 74). From a careful study of the 
development of these cones as they project into the cytoplasm, 
there seems to be little doubt that the fibrils composing them 
are transformed directly out of the reticulum of the cytoplasm. 
This is particularly evident in the earlier stages, where there is 
no sharp differentiation between the ends of the fibrils of the 
weft and the surrounding threads of cytoplasm, the one passing 
into and apparently being continuous with the other. 
As the cones project outward, they terminate in very sharp 
points and extend over half way to the cell wall opposite (/igs. 
14,15). During the entire process of their development the 
nuclear membrane and the nucleolus persist. There is no evi- 
dence of the breaking down of these structures until the cones 
are fully developed. As soon as the cones cease growing out- 
IO ir 
devia iF 
