848 Fraser and Snell .— The Vegetative 
Gradually it becomes erected till it stands out at right angles to the fibres 
(PL LXII, Figs. 19, 20), and the ends of the two daughter chromosomes, which 
at first lay close together, begin to slide away from one another along the 
spindle (Fig. 20) till the double rod becomes transformed into a wide V or 
U, at the apex of which the ends remote from the spindle lie pressed 
together or just crossing one another (Figs. 21, 22). The ends upon the 
spindle travel further and further apart till the daughter chromosomes are 
no longer in contact and lie flat or almost flat along the fibres (Fig. 23). 
They have the form of straight or somewhat sinuous rods ; frequently the 
end towards the pole of the spindle is bent over, forming a hook; whether 
this takes place depends on the position of the chromosome during the 
early stages of its attachment to the spindle. Indications that such hooks 
will be produced may be recognized in Figs. 20 and 21. 
The daughter chromosomes have now reached the stage represented in 
Fig. 1, and they show no recognizable indication of the longitudinal split 
which is about to occur ; indeed, we may perhaps suppose that, for the 
short period between the metaphase and the telophase, the chromosome is 
a single structure. 
Mitosis in the Gametophyte. 
Various authors have called attention to the double nature of the 
chromatin elements in somatic cells, and it has been interpreted by some of 
them (Overton ( 12 ), Sykes ( 14 ), Takara ( 15 )) as due to the lateral approxi¬ 
mation of structures derived respectively from the male and female gametes, 
while to other investigators (Hof ( 7 ), Digby ( 1 )) it has appeared rather 
that the duplication is brought about by an early longitudinal fission, which 
in the telophase of one division already prepares for the next. 
In the Broad Bean a study of these stages in the cells of the sporophyte 
seemed to us to point pretty clearly to the latter interpretation ; in order 
to test this conclusion we proceeded to a study of the vegetative divisions 
of the gametophyte, and examined the nuclei of the pollen-grains in various 
stages of development. 
Here the chromosome number is seven instead of fourteen as in the 
diploid cells, and there is no question of the association of paternal and 
maternal structures since only a single set of chromosomes is present. But 
in the pollen-grain, exactly as in the cells of the root or flower, a double 
reticulum was observed (PI. LXII, Fig. 24; PI. LXIII, Figs. 31, 32), and 
here also the first evidence of duplication appears in the late telophase 
(Figs. 30, 31). 
The formation of the first ‘ resting 5 gametophyte nucleus is preceded 
by the homotype division. It is proposed to discuss the meiotic stages in 
a forthcoming paper, and it suffices here to say that in the homotype telo¬ 
phase, as in all others studied, longitudinal fission can be seen. 
