150 Digby. — On the Archesporial and 
makes it quite clear that ‘ the reduction in the number of chromosomes has 
not yet taken place ’, whilst in his Fig. 16 the spireme threads 1 ‘ are segre- 
gating in pairs preparatory to a lateral union ’ (p. 625). 
When the majority of the spireme filaments have conjoined in pairs 
(Figs. 68 and 69), they necessarily occupy a more restricted area than when 
they were independent of one another (cf. Figs. 60, 62, 63, and 67 with 
Figs. 68 and 69). The nucleus at the same time decreases considerably in 
size. The conjoining spireme filaments stain uniformly and have a taut 
and strained appearance pricS to close conjunction (cf. Figs. 59, 63, and 67 
with Fig. 69). Close conjunction bears a striking resemblance to fission, 
the two approaching univalent spiremes (filaments) being identical even to 
their beading (Fig. 69). It is only by following every stage, and by con- 
stant comparison, that these pre-second contraction phases can be inter- 
preted and the fact substantiated that one is here dealing with conjunction 
and not with separation. It will be remembered that the associating halves 
of univalent spireme (threads) in presynaptic prophases exhibit precisely 
similar characters. 
The conjoining stages appear to have been missed by many cyto- 
logists, who have therefore concluded that these paired filaments going into 
second contraction are ^k-joining and not co 7 i-]oinmg, and hence that this 
separation is the opening out of the fission seen in the post-synaptic spireme. 
Gregoire ( 10 ) and Yamanouchi ( 19 ) have both come to this conclusion with 
regard to Osmunda. Farmer and Moore (8), on the other hand, described 
the coming together of the sides of the loops in the stages leading to second 
contraction, and they note that they approximate closely ‘and thus simulate 
an appearance of a longitudinal fission 5 (p. 520). It is proposed to refer to 
this point again in the ‘ General Considerations 
Gradually (Figs. 70 and 71) the greater part of the spireme filaments 
become collected together and lose their visible identity in the massing of 
second contraction, but isolated portions may remain independent and free 
(Fig. 71). In this nucleus (Fig. 71) the two outlying spireme filaments 
show a twisted conjunction, but in Fig. 72 some of the filaments, not involved 
in the aggregation of contents, are widely separated, whilst others are 
closely conjoined. It is suggested that close conjunction of two filaments 
(i. e. of two univalent spiremes) may, in some instances, never take place at 
all, for it is not unusual to find bivalent chromosome-like entities lying out- 
side the mass of second contraction ; these apparently concentrate and 
thicken, and thus take a short cut to becoming heterotype chromosomes. 
The main and final process of conjunction, involving the sorting out 
and conjoining in pairs of all the univalent segments (filaments), takes place 
during second contraction, whereas the association of the pair of threads which 
together make up the univalent spireme is consummated during synapsis. 
1 ‘Filaments according to the nomenclature adopted in this paper. 
