Archegonium ; and on Spermatogenesis in Poly trichum. 121 
figured these extruded particles but misinterpreted them as centrosomes, 
which, after occupying the poles of the spindle, have come round to the 
inner side of each chromosome group, and ultimately pass in amongst them 
and become included within the daughter nuclei. 
In view of J. and W. van Leeuwen-Reijnvaan’s statement that a 
reduction in the number of chromosomes takes place during the last division 
in the antheridium of Poly tri-chum formosum , countings were made of 
the chromosomes in this species during the anaphase, and the full number, 
six, was repeatedly found to be present (Fig. 20). 
The Reconstitution of the Daughter Nuclei. 
During the telophase the chromosomes become united in such a 
manner as to constitute a figure resembling the spireme of the prophase 
(Fig. 22). Many nuclei exhibit an end-to-end union of the daughter 
chromosomes. Fig. 21 shows an example of this where five chromosomes 
have become united ; while in the cell on the left in Fig. 20 is a sixth 
chromosome which shows indications of becoming laterally united with 
adjoining chromosomes. Connecting threads next appear between some of 
the chromosomes, probably as a result of contact at certain places, followed 
by a pulling apart and the drawing out of viscid threads, the whole present- 
ing the appearance of a simple reticulum (Fig. 20, cell on the right). There 
is no definite approximation of the chromosomes such as that observed by 
Fraser and Snell (7) in Vicia Faba> nor could any alveolization of the 
chromosomes be made out at any stage in the division of the nucleus 
as described by Gregoire and Wygaerts (8) and others in various plants. 
The establishment of the reticulum is accompanied by a change 
in the distribution of the chromatin, which is now seen to leave the network 
and collect into small drops of various sizes (Figs. 10 and 11). These 
deeply staining drops gradually run together and form the large rounded 
nucleolus. Sometimes a single chromatin droplet remains for a time 
distinct from the main mass (Fig. 12), but ultimately this is usually taken 
in, thus leaving the reticulum with very little chromatin distributed upon its 
threads. It is probable that the last outstanding particle of chromatin 
is identical with the body which J. and W. van Leeuwen-Reijnvaan state is 
cut off from the nucleolus and extruded from the nucleus to form, by 
division, the centrosomes. 
The Occurrence of Centrosomes. 
During the early prophase of the last division of the spermatogenic 
cells, the centrosome-like bodies, stated by J. and W. van Leeuwen- 
Reijnvaan to be present at all divisions, are observed for the first time 
in the development of the antheridium. The determination of the last 
division gave considerable difficulty. Except in a few cases, this is not 
