Allen . — The Spermatogenesis of Poly trie hum juniperinum . 279 
observed it. It is sometimes approximately spherical (Figs. 25, PI. XV ; 34, 
45, PI. XVI) ; occasionally (Figs. 26, PI. XV ; 37, PI. XVI) it is slightly 
elongated in an axis perpendicular to the blepharoplast ; oftener it is more 
or less flattened against the blepharoplast (Figs. 28, 32, 40, &c.). Sometimes 
differences appear in the staining reactions of different parts of the apical 
body (Figs. 26, PI. XV ; 37, 38, 42, PI. XVI) ; but there is no uniformity in 
these appearances, and as a rule the whole body is homogeneously stained. 
The latest stage at which I have seen the apical body is that shown in Figs. 48 
and 49, at which the nucleus has attained approximately its final length. 
After this time the apical body seems to disappear very quickly and com- 
pletely. Occasionally a mature antherozoid seems to have a slightly thicker 
portion just behind its pointed anterior tip (Fig. 59), which at times I have 
thought might be due to the retention of part or all of the substance of the 
apical body ; although it may also be explained by the presence of the 
anterior portions of the cilia, lying close to, and indistinguishable from, the 
body of the antherozoid ; at any rate, the part (if any) taken by the apical 
body in the formation of the mature antherozoid is so small as usually 
(Figs. 56-8) to produce no perceptible effect. 
The limosphere, after the formation of the apical body, places itself, as 
already noted, in contact with the nucleus (Fig. 27). Sometimes (Fig. 28) 
the limosphere is flattened in the region of contact ; oftener (Figs. 32, 35, 
39) the nuclear outline is indented as though it had yielded where the 
limosphere pressed against it. When the nucleus begins to elongate, the 
limosphere is usually in contact with that portion of the nucleus which 
is least stretched out, and whose diameter therefore remains greatest 
(Figs. 28, 30, 35) ; and since the elongation of the nucleus is most marked 
at first in its anterior region, the limosphere thus lies nearer the posterior 
than the anterior end of the nucleus. Sometimes the limosphere lies at 
the very posterior end (Figs. 33, 37, 45), or close to this end (Figs. 36, 
39, 40), of the nucleus, and therefore farther back than the region of 
greatest nuclear diameter. However, in the rare cases observed in which 
there is both a posterior and an anterior elongation of the nucleus (Figs. 42, 
47), the limosphere is in contact with the thickest portion of the nucleus, and 
therefore at some distance from its posterior as well as from its anterior end. 
At the late stage shown in Figs. 48 and 49 the limosphere is still present, 
enclosed within the posterior part of the coiled nuclear body (Fig. 48) or at 
the very tip of the latter (Fig. 49). The same is true at still later stages 
(Figs. 50-3), an d the limosphere is similarly placed in antherozoids seen in 
sections of nearly or quite mature antheridia (Figs. 54, 55). In sections 
showing these very last stages, there is no longer a visible differentiation 
between the outer and inner portions of the limosphere. 
During that period in the metamorphosis of the androcyte which 
is represented by Figs. 30-5, a darkly staining body, distinct from both 
