160 Digby . — On the A rchesporial and 
presynaptic prophases must represent the pairing of entire somatic 
chromosomes. This brings one to the bed-rock of the problem, and it will 
be discussed in detail. 
From evidence supplied by both schools of cytologists, it is clear that, 
as a general rule, 1 the somatic chromosomes in telophase apparently dis- 
organize by first dividing into longitudinal halves (threads), which gradually 
become dissipated throughout the nucleus. Prophase, therefore, entails a re- 
association of these halves for the reorganization of each individual chromo- 
some. If this dual nature of the chromosomes be borne in mind, it will be 
realized that in all prophases, whether somatic or meiotic, there must be a 
coming together of the halves (threads) dispersed during the preceding telo- 
phase involving a pairing of threads. Consequently, that parallel threads 
should be present in all nuclei, including those of parthenogenetic eggs (15) 
and gametophytes (9), is only a normal event, and it bears no relation 
whatsoever to the pairing of entire homologous chromosomes. 
Gr^goire (10) and his followers promote this view with regard to somatic 
prophases, but, on the other hand, they regard the paired threads of the 
heterotype prophases as the pairing of filaments representing entire homo- 
logous chromosomes. Such a theory postulates the complete rearrange- 
ment of the elements of the last premeiotic telophase, for which, however, 
there appears to be no supporting evidence, for its visible microscopical 
detail is in every respect identical to that of preceding telophases. On the 
contrary, when there is no interkinetal rest it is possible to trace the origin 
of the paired threads of the heterotype prophases back to the longitudinally 
split chromosomes of the preceding archesporial telophase. This is irre- 
futable evidence on behalf of the view that th ft paired threads of the hetero- 
type prophases are homologous to those of the somatic prophases ; that is to say, 
that each thread represents half a univalent chromosome of the preceding 
telophase. Having established this fact, the subsequent phases form 
a logical story. 
During synapsis the lateral association in pairs of the threads is con- 
summated. Whereas in somatic prophases the threads come together to 
form the univalent spireme filament which segments directly into univalent 
chromosomes, in the heterotype prophases this univalent spireme filament 
as it issues from synapsis is continuous, the univalent segments (filaments) 
being joined end to end, and their separation is postponed until a con- 
siderably later stage. 
It was not until the success in fixing was attained that all stages in the 
conjunction in pairs of entire univalent lengths of spireme (i.e. of filaments) 
were observed throughout the phases between post-synapsis and second 
contraction. In Osmunda these stages are not only difficult to fix suffi- 
ciently well to show this phenomenon, but they are also quickly passed 
1 See foot-note, p. 156. 
