124 Walker. — On Abnormal Cell-fusion in the 
extruded chromatin and consequently does not appear in the three figures 
representing these changes. The nucleus now stains very feebly, and 
its limits are often very difficult to distinguish (Figs. 29-31). In an 
occasional spermatid the extruded substance is collected into one drop, and 
there can be little doubt that this single mass of chromatin, and conse- 
quently the two masses normally present, corresponds to the ‘ chromatoiden 
Nebenkorper ’ observed in M archantia by Ikeno ( 9 ), the origin of which he 
was unable to determine. It agrees also with the ‘ limosphere ’ described by 
Wilson ( 22 ) in Mnium hornum and in Atrichnm undulatum. 
In Mnium Wilson states that the ‘limosphere’ originates through the 
coalescence of rod-like structures derived from the nucleolus, and in 
Atrichnm it is derived directly from the nucleolus. This observer believes 
that J. and W. van Leeuwen-Reijnvaan misinterpreted these rod-like bodies 
as three chromosomes, and upon this founded their statement regarding 
reduction. As these rod-like structures do not occur in Polytrichum , it 
therefore follows that Wilson’s explanation regarding the observations 
of J. and W. van Leeuwen-Reijnvaan does not apply. 
In those spermatids where the blepharoplast is embedded in the 
extruded chromatin, it soon becomes free and is next seen to lie at 
varying distances from this substance (Figs. 32 and 33). In Fig. 32 
the blepharoplast is the larger particle underlying one end of the nucleus. 
Meanwhile, a prominent organ, which when mature takes the form of 
a curved band, has now commenced its development at the periphery of the 
spermatid. This structure is first organized in the vicinity of the rounded 
masses of extruded chromatin, which at this stage appear hollow, the 
substance in the interior of each mass staining very faintly. The resemblance 
between these deeply staining bodies and the single ‘ limosphere ’ described 
by Wilson is now apparent, while the difference in number may probably 
be accounted for by the much larger amount of chromatin which is passed 
out of the nucleus in Polytrichum. With regard to the function of these 
bodies, their intimate association with the band-like structure and the 
similarity in their capacity for stain suggest that they contribute some 
of their substance to its formation. At a later stage, when this band is 
completely formed, only one deeply stained vesicle remains (Figs. 34-6) ; 
the other probably owes its disappearance in some degree to the transference 
of its substance to the developing band. This interpretation of its disap- 
pearance is supported by the spermatid shown in Fig. 32, where that part of 
the band in close proximity to the diminishing vesicle is seen to develop first. 
The spermatid, hitherto a compact protoplasmic body, now develops 
numerous large vacuoles and increases considerably in size. The nucleus 
stains more deeply and contains a central nucleolus-like structure (Fig. 34). 
At the same time the developing band has become extended towards, and 
finally reaches, the blepharoplast. This strongly arched band now connects 
