Allen— The Spermatogenesis of Poly trichum juniperinum. 271 
uniform thickness, excepting that it is slenderer toward the ends, one or both 
of which may be pointed. The blepharoplast is stained deep black in iron- 
alum-haematoxylin, and deep blue in the triple stain. Some authors — for 
example, Ikeno (1903) in his study of Marchantia , Arens (1907) and the 
van Leeuwen-Reij nvaans (1908) in their investigations of Poly trichum — -have 
described the blepharoplast as growing only into a relatively short body to 
whose posterior end is attached another thread-like or band-like structure 
formed by a gradual differentiation within the cytoplasm. Something 
similar is reported by Walker (1913) for Polytrichum. I cannot confirm 
these descriptions.. All the evidence furnished by my preparations indicates 
that the long, darkly stained rod (shown for example in Figs. 11-14) is 
a homogeneous body formed by the growth of the originally small blepharo- 
plast. In this respect my conclusions agree with those reached by Yama- 
nouchi (1908) and Miss Allen (1911) in studying the spermatogenesis of Ferns, 
by Woodburn (1911, 1913) and Miss Black (1913) working with various 
Hepaticae, and by Woodburn (1915) in his study of Mnium. 
While the blepharoplast is still comparatively short, another definite 
cytoplasmic body appears (Figs. 5, 9, 10). This is the large limosphere , to 
use a term suggested by M. Wilson (1911). This body seems to correspond 
to the one which, in the androcytes of Marchantia , was called by Ikeno 
(1903) the ‘ chromatoider N ebenkorper \ It is certain that in Poly trie hum, at 
any rate, the body in question is quite dissimilar in behaviour and history to 
the structure appearing in the spermatids of certain animals to which the 
name chromatoider Nebenkorper was first applied ; and for this reason 
the term introduced by Wilson will be used in the present paper. The 
limosphere is at first rather faintly stained (commonly taking the orange of 
the triple stain) ; it seems to arise by a concentration of a portion of the 
cytoplasm. Indeed, the bodies shown in Figs. 9 and 10 are so slightly 
differentiated from the rest of the cytoplasm by their tint and their greater 
degree of homogeneity that one can hardly be certain either of their exact 
outline or of their identity with the body which shortly afterwards is so 
conspicuous. However, a comparison of these figures with Figs. 5 and 12, 
in which the limosphere is very distinct, although orange in colour, and with 
Fig. 13, in whose cells also the limosphere is slightly stained, leaves little 
doubt that it is the same body which is seen in each case. Except for 
a brief period at the beginning of its existence, the limosphere is deeply 
stained by iron-alum-haematoxylin and by the safranin of the triple stain. 
As a rule, the outer portion of the limosphere stains more deeply than most 
of its internal substance, so that it looks much like a hollow sphere (or, as 
Figs. 11, 12, &c., show, a ring in cross section). Very commonly there is 
also some deeply-stained material in the interior (Figs. 11, 12) ; but some- 
times (Fig. 21) the stain is confined entirely to the outer shell. At still 
other times the limosphere seems to be uniformly stained throughout. This 
