159 
Meiotic Miloses of Osmunda . 
in synapsis showing the ‘ filaments minces ’ 1 orientated towards one pole of 
the nucleus and associating in pairs. He regards this stage as the pairing 
of entire univalent chromosomes. With Strasburger, he considers the 
fission in the post-synaptic spireme as the separation of these same entire 
univalent chromosomes which had come together during synapsis and which 
will separate at the heterotype metaphase. 
In 1910 Yamanouchi ( 19 ) published a paper on O. cinnamomea , and 
his views coincide with those of Gregoire. He traces the duality of the 
threads, 2 from early prophase to their separation as daughter chromosomes 
on the heterotype spindle. He shows that ‘ the threads come out of the 
network as two independent threads from the start ’ (p. 5) ; that they come 
to lie parallel to one another, and that these ‘ double threads ’ shorten and 
thicken, and a pair of such shortened and thickened ‘ threads ’ become 
a bivalent chromosome (p. 6). 
It is evident that neither Strasburger, Gregoire, nor Yamanouchi 
noticed the conjunction of segments of entire univalent spireme (i.e. of 
filaments) in the post-synaptic stages, and consequently were unaware of the 
complications thus introduced, and Gregoire and Yamanouchi consequently 
assumed that the striking pairing of threads in presynapsis and synapsis was 
the pairing of entire univalent chromosomes. 
After years of hesitation the writer has come to the firm conclusion 
that Osmunda affords a striking piece of evidence in favour of the ‘ telo- 
synaptic ’ 3 theory of the origin of the heterotype chromosomes. It is 
possible to identify all the important stages illustrated by the previously 
mentioned workers, namely the parallel filaments (of Gregoire) in the pre- 
synaptic prophases, the association of these filaments in pairs during 
synapsis resulting in the emerging thick spireme, and the separation of the 
univalent segments as they come out of second contraction to organize the 
typical heterotype chromosomes. If this really represents the complete 
sequence of events, and if the fission in the post-synaptic spireme is that 
which eventually parts the univalent chromosomes on the heterotype 
spindle, how has the numerical reduction of the chromosomes come about ? 
The only possible answer is that the paired filaments (of Gregoire) of the 
1 Gregoire (1910, Les Cineses de maturation dans les deux regnes. La Cellule, xxvi) calls the 
‘ threads ’ (term used in the sense adopted in this paper) of heterotype prophases 1 filaments minces ’, 
and considers each ‘ filament mince ’ to be an entire somatic chromosome. He thus describes 
synapsis: ‘ C’est le stade de noyau leptotene (Winiwarter, 1900) ou leptonema (Gregoire, 1907), ou 
noyau a filaments minces . . . l’association deux a deux des filaments minces des noyaux leptotenes 
et un rapprochement graduel des deux filaments associes, donnant origine aux anses epaisses des 
noyaux pachytenes. . . . Nous avons (1907) propose, pour ce stade, le nom de noyaux zygotenes ’ 
( = noyaux amphitencs of Janssens) (p. 237 % 
2 Yamanouchi (19) retains the term ‘ threads ’ throughout the heterotype division. He considers 
the pairing of ‘ threads ’ in synapsis to be the coming together of two entire homologous somatic 
chromosomes which separate on the heterotype spindle. 
3 See foot-note on p. 156 . 
