2 BOTANICAL GAZETTE [JANUARY 
much smaller branches are given off. These branches run in all 
directions over the surfaces of the adjacent scales, so that when 
they are separated the mesh of tubes sometimes appears like a fine 
spider web. Numerous nuclei were distributed throughout the 
tubes, sometimes singly and sometimes in groups of a half-dozen 
or more. Aside from the body cell they all look much alike, so 
that it was not possible to identify the stalk and tube nuclei. 
The tube branches appear to grow independently of the tube 
nucleus or of any other nuclei. Branches were sometimes found 
in which no nuclei could be found at all. This was true of those of 
considerable length, as well as those that were just beginning to 
orm. No cross-walls were observed in any part of the main tube 
or its branches. The main branch after entering the nucellus 
sometimes branches again (fig. 1). These branchings are very 
much less frequent than those outside. The available evidence 
indicates that only the one branch enters the nucellus. 
In the majority of the tubes examined the body cell remains 
in the extra-nucellar part of the tube until after its division. 
Sperms were often observed in the tubes at the extreme tip of the 
nucellus (fig. 5), while very few body cells were observed within 
the nucellus. The mitosis of the body cell nucleus was not observed. 
Fig. 2 shows what appears to be the spindle lying between the two 
nuclei of a body cell. The nuclei are both in the resting condition 
and the spindle is surrounded by a delicate membrane. This 
membrane and its continued persistence after the daughter nuclei 
have reformed may be urged against this structure being really 
a spindle. It strongly suggests the remains of the nuclear mem- 
brane and thus suggests an intranuclear spindle. I have observed 
no other structures with which it could be identified, unless there 
may be some connection with the blepharoplast-like body referred 
to in a later paragraph as occurring in the cytoplasm of the male 
cells and the two-celled proembryo. 
Aiter the division of the body cell nucleus the two daughter 
nuclei do not usually separate for some time. In some cases they 
even enter the archegonium still closely associated. The division 
of the cytoplasm of the body cell follows considerably later. In 
some cases no distinct division has occurrred when the two male 
