278 Allen . — The Spermatogenesis of Poly trichum juniperinum. 
nucleus ; at the latter point the blepharoplast seems to end, although it is 
highly probable that conditions are really substantially alike in the two 
androcytes of Figs. 28 and 29. Sometimes the blepharoplast can be 
followed for nearly or quite its full length at a much later period, as in 
Figs. 45 and 46 already referred to. So it seems necessary to conclude, in 
spite of the frequent impossibility of tracing the blepharoplast throughout 
its length, that it actually persists without material decrease in size until 
a comparatively late stage in the history of the metamorphosis of the 
androcyte. Eventually, however, the blepharoplast (except for its anterior 
end) seems completely to disappear ; and whether this final disappearance 
is due to its coalescence with the nuclear membrane or to an actual 
absorption or destruction of its substance, must be left unsettled. 
In the later stages of nuclear elongation (Figs. 38-44), the anterior 
end of the blepharoplast seems to lie outside the boundary of the remaining 
cytoplasm — whether or not still covered by the plasma membrane is not 
clear. Occasionally, as in Fig. 44, the apical body as well as the end of 
the blepharoplast seems to be partly outside the cytoplast. The projecting 
tip of the blepharoplast probably corresponds to the cytoplasmic Hocker 
described by Strasburger (1892), and also to the refractive bouton of 
Guignard (1889), although the latter author takes the swelling in question 
for the anterior end of the nucleus. The appearances in my preparations 
at this time suggest that the anterior end of the developing antherozoid is 
pushing out and thus beginning to free itself from the cytoplasm that is not 
to be used in its formation. Something like this is described by Ikeno 
(1903) as occurring in the spermatogenesis of Marchantia . But I do not 
find in Polytrichum a continuation of this process such as Ikeno’s figures 
suggest for Marchantia , namely, a further pushing outward of the anterior 
end so that the unused cytoplasm is as it were forced to the posterior end 
of the antherozoid. On the contrary, even after the cytoplasm has greatly 
diminished in amount (Figs. 51, 52), a portion of what is left remains 
attached to the anterior end of the body of the antherozoid ; and the final 
disappearance of this anterior mass of cytoplasm is one of the last events in 
the history of the androcyte. 
The same difficulty in the observation of cilia that was noted earlier is 
met with throughout this history, although portions of one or both cilia are 
frequently to be seen (Figs. 29, 34, 4 %, 43, &c.). In sections containing 
mature or nearly mature antherozoids (Figs. 5° - 2, 54 ), considerable portions 
of the cilia are visible ; but they can be seen in their full length and in their 
relation to the body only in preparations that contain unsectioned anthero- 
zoids (Figs. 56-9). 
The apical body remains for a long time in contact with the anterior 
end of the blepharoplast. No apparent material change takes place in its 
size, form, or density down to the latest period (Figs. 48, 49) at which I have 
