122 
Walker .—On Abnormal Cell- fusion in the 
diagonal as in Marchantia described by Ikeno ( 9 ), and Fegatella by 
Bolleter ( 5 ), consequently this easy method of distinguishing it is not avail- 
able. At this time the cell-walls are thick and swollen, and stain somewhat 
deeply with orange G (Figs. 18-20). The contents appear more or less 
rounded, and the spermatogenic tissue no longer presents the well-known 
tessellated appearance which characterizes it in sections of younger antheri- 
dia. Countings were also made of the cells after this division, and these 
were found to approximate those of antheridia containing spermatids with 
developing spermatozoids. As a result of the final division, two spermatids 
are formed within each thick-walled chamber (Fig. 23). 
The centrosome-like bodies, when first perceived, are situated upon 
opposite sides of the nucleus and appear as minute deeply stained particles 
of equal size, with a few delicate radiations extending from each (Figs. 13- 
15). According to Ikeno, these organs are present in the antheridium 
of Marchantia polymorpha during all the cell-divisions, and he therefore 
regards them as morphologically equivalent to centrosomes. During the 
following account this term will be used to designate these bodies. Their 
presence in the antheridium of Poly trichum at this stage is unmistakable, as 
they can be easily determined in all the cells. It was not found possible 
to discover their origin. No appearances were observed suggesting the 
extrusion of a parent particle from the nucleus. To determine this point, 
a differential stain is required to distinguish these structures, before they 
take up their position at the poles of the nucleus, from other deeply staining 
particles which are often observed both within the nucleus and in the 
cytoplasm. 
When the spindle is formed, the centrosomes occupy its poles and are 
now much less conspicuous than during the prophase. After the comple- 
tion of the division these bodies persist in the daughter-cells, one lying close 
to each nucleus (Fig. 23). As the time of the development of the spermato- 
zoid approaches, the persistent centrosome increases in size (Fig. 27) and 
becomes highly refractive, appearing as a very conspicuous object in the 
spermatid. This refractive body, or blepharoplast as it may now be called, 
agrees with the ‘stark lichtbrechender Plasmahocker ’ of Strasburger ( 17 ). 
In two antheridia belonging to the same rosette a nuclear division was 
observed, which, considering its rarity and the unusually large number 
of cells counted across the longitudinal section, was regarded as an extra 
division. In these cells the centrosome, which has already become refrac- 
tive, has probably in some degree lost its normal function. This is also 
suggested by the fact that its division does not precede the onset of mitosis, 
but was found to take place in one case at the time of the formation of the 
chromosomes (Fig. 24) ; in another, not until after the formation of the 
spindle. In some dividing nuclei of these special antheridia the two resultant 
refractive bodies are seen to occupy the poles of the spindle (Fig. 25) ; 
