118 
L. Digby 
univalent Segments being temporarily concentrated. True fission reappears 
in the univalent segments as they come out of second contraction (fig. 78), 
but it is not easily seen until alter the lieterotype metaphase, when it 
occurs as tlie cleavage so ölten visible in the daiighter chromosomes as 
they retreat to the two poles (PI. X, lig. 84). The l)ivalent spireme, as it 
conies out ol synapsis, displays all degrees in the Separation ol its uni- 
valent Segments. These may be either 1. approximated so closely side by 
side showing no apparent space betwecn the two; or 2. partially separated; 
or 3. completely separated throughout their length but united at their 
ends to lorm a loop. The junction ol two univalent lengths ol spireme 
is always emphasized by a chroniatic swelling. 
1. Some ol the thick bivalent loops and Strands 2nay be homogeneous, 
and hence display no sign ol their duplex nature, though in character, 
and in thickness, they resemble others in which traces ol Separation into 
their univalent segments can be seen. These homogeneous bivalent 
portions ol spireme point to an intimate side to side association ol uni- 
valent lengths ol spireme arranged diiring synapsis (PI. IX, ligs. 64 and 65). 
2. All degrees ol partial Separation ol univalent Strands can be lound. 
The line ol luture Separation may be inerely indicated as a clear space in 
the substance ol the bivalent spireme (ligs. 66 and 67); or the two uni- 
valent segments may have recently beconie dissociated, their edges being 
still somewhat ragged and beaded (lig. 67); or the univalent segments 
may be entirely parted lor an appreciable length, and be rounded in 
outhne (ligs. 65 and 68). 
3. Paired univalent segments ol spireme, although separate through- 
out their lengths, may be united at their ends to lorm a loop (ligs. 69 
and 73). It is evident that such univalent lengths ol spireme have 
separated and are not in the act ol approximatmg lor the two lengths 
are identical, and may constantly be joined by Ime thread-like cross 
Connections (ligs. 69 and 73). These univalent scgmeiits demonstrate 
an end to end association. 
Gradually the loops and Strands extend mto the nuclear cavity, 
and a lew may reach the peripheiy, but as a rule they remain congre- 
gated more or less in the centre ol the nucleus (lig. 70). In C. virens 
there is no ‘iiollow spii'eme’ stage, lor the spireme, as it conies out ol synap- 
sis, passes ahnost dnectly into second contraction. In only one series 
ol sections was the spneme lound to be evenly distributed throughout 
the nucleus. 
It is evident, that the phases, Irom spiapsis onwards to diakinesis, 
are quickly passed through, as nuclei in very (üllerent stages ol pro- 
