The Origin and Development of the Apotheeium etc. 
389 
a swelling of tliis hypha. A third hypha at b (Fig. 23) seems also to 
be connected with the binncleate cell. I have noted that the hyphae 
anastomose very frecpiently, and it appears that such large cells as I 
have described may arise where the hyphae become connected. Tlu> 
spermatiophore shown in Figure 23 at c appears to come from the large 
cell at the base, but in tliis ease I am less certain that the two cells are 
connected. Tliere is need of much more evidence before we can accept 
it as a fact that these irregularly swollen cells are always concerned 
with the production of spermatia. I have merely noted here a few ob- 
servations, which strongly suggest such a relation. Further study of 
my preparations may make it possible for me to acld to these observa- 
tions at a later time. 
The spermatia have definite cell walls. Tliis is especially evident from 
the length of time that the c-ells retain their shape after the contents 
have passed into the trichogyne (Fig. 54, PI. XXXV; Fig. 26, PI. XXXII). 
Vhen the spermatia are first foimed eacli contains a single nucleus (Fig. 15 
at a, PI. XXXI), which, except for its minute size, appears not to differ 
from the resting nuclei in the vegetative hyphae. The nuclear membrane 
is quite distinct and the chromatin stains deeplv. If a nucleole is present, 
it is not possible to distinguisli it from the chromatin. The chromatin 
may be so evenly distributed on the nuclear membrane that the nucleus 
looks like a globule of c-lear material with only the outline stained. I 
have never seen a dividing nucleus in a spermatium, and yet these cells 
later appear to contain two or tliree globular bodies each of which at 
first siglit might be taken for a nucleus (Fig. 15 at d). After much study 
of my preparations, I am cpiite certain that tliere is but one nucleus. 
By referring to several figures the reasons for tliis conelusion may be 
made clear. The single nucleus is often somewhat elongated (Fig. 15 at a). 
Fig. 15 at b shows a spermatium with a single large pear-shaped nucleus 
and a smaller globule. At c in Fig. 15 and at c in Fig. 54, PI. XXXV, 
are cells with the containcd structures similar to tliose figured in Fig. 15, 
PI. XXXI at b, but seen in a somewhat different plane so that the larger 
body in each cell appears ovoid. Frequently the arrangement of the stain- 
able material in this larger body is such that it appears at first sight to be 
two bodies instead of one (Fig. 54, PI. XXXV at c and Fig. 22, PI. XXXI 
at a). If we turn to Fig. 15 at d we may explain the apparently three 
globales as really but one larger and one smaller one. Very frequently, 
however, these three globales do not appear to be so distinct. The stain- 
able material is quite evenly distributed on the margin of each as seen 
in optieal section and there seems to be but one nucleus which is either 
