468 Fraser and Welsford . — Further Contributions to the 
spindle (Figs. 28, 29) and, in the anaphase, eight travel to each pole (Fig. 
30). We were not able to find four chromosomes, as described by Maire, 
in definitely uncut nuclei at any stage during this division. 
On the spindle of the second or homotype division (Fig. 34), and 
also on the spindle of the third (Fig. 38), four chromosomes appear ; they 
divide so that, in both cases, four pass to each daughter-nucleus (Figs. 35, 
39, 40). Their formation is preceded by the withdrawal of the chromatin 
towards one side of the nucleus (Fig. 33) ; this may be regarded as 
analogous to one of the meiotic contractions and as indicating the moment 
of pairing of the eight chromosomes of the heterotype telophase. Since 
contraction takes place in both the second and the third prophase (Figs. 33, 
37) it would seem that the union of the paired chromosomes is very slight 
and breaks down between the end of the second and the beginning 
of the third division ; this conclusion is also borne out by Maire’s and 
Guillermond’s observations. 
Guillermond has described eight chromosomes throughout the second 
and third divisions, and Maire eight protochromosomes in both prophases. 
In our material the number in the anaphases and early telophases of these 
divisions is clearly four, but it is not impossible that, in some cases, the 
eight chromatin bodies found on the spindle may represent the ‘proto- 
chromosomes * of Maire rather than, as we have considered, the separated 
daughter-chromosomes. 
In either case, by the end of the third division a second reduction has 
been accomplished and the eight chromosomes of the heterotype telophase 
have been replaced by four. The process by which this is accomplished 
appears to correspond to that in H. rutilans and O. aurantia. 
Before the third division is complete spore formation begins. The 
spindle elongates, and during the telophases and the early stages of recon- 
struction of the daughter-nucleus a beak is formed (Figs. 43, 44) and 
appears to push actively towards the periphery. At the same time changes 
take place in the cytoplasm (Figs. 41-44) ; its staining capacity increases 
and the astral rays are bent backwards, giving the arrangement first 
described by Harper ( 15 ). The appearance of the rays suggests that cur- 
rents of altered cytoplasm are flowing back around the advancing nuclear 
beak. A little later the upper portion of the spore is seen to be defined 
by a limiting membrane (Fig. 45), while the lower part is irregularly 
outlined by the sides of neighbouring vacuoles, and becomes rounded off 
at a later stage. 
The arrangement of the vacuoles is in general very regular : a large 
one fills the lower part of the ascus and another occupies its apex. Seven 
others, more or less well-defined, are left by the first, second, and third 
mitoses ; and the main body of cytoplasm is thus broken up into eight 
masses within which spore formation takes place (Fig. 44). 
