452 
J. Osawa 
A. Taraxaeum platycarpum Dahlst. 
a) The development of the pollen-grain. 
At an earlier stage the nucleus of the pollen-mother-cell of T. platy- 
carpum has one very large nucleolus and fine linin-net-works, upon which 
many chromatin-grannles are scattered. Tlie latter gradnally increase 
in size and number; they might be perhaps homologizedto the prochromo- 
somes observed by Overton (1905) and some others (B'ig. 1). 
When synapsis approaches the linin-thread and chromatin-granules 
gradnally contract towards the nucleoli, away from the nuclear mem- 
brane as usual; this contraction proceeds more and more, until finally a 
dense ball of threads is produced, from which a few loops of very 
delicate Strands sometimes extend into the nuclear cavity. Düring this 
process many chromatin-granules, which stain deeply, are yet obser- 
vable, being distributed upon the delicate linin-fibres (Fig. 2). As sy- 
napsis progresses the thread beeomes thieker and thicker and more 
prominent tlian before. 
After a comparatively long duration of this stage, the ball loosens 
its structure and the threads begin to extend gradnally within the nuclear 
cavity (Fig. 3). Then the spirem appears to be uniform tliroughout 
its whole length and has the double structure, which is however in many 
cases difficult to be distinctly recognized (Fig. 4). The double structure 
is especially very clear in the places, where the halves of the spirems 
diverge for a short distance (Fig. 5); then the minute chromatin-sub- 
stances — chromomeres — are to be seen, inlbedded in the linin-threads. 
These double threads then contract gradnally and fuse eacli other to 
form a thick spirem, which now streches itself loosely tliroughout the 
entire nuclear cavity and forms a very beautiful thick homogeneous 
filament as in Fig. 6. 
After a short time this filament again separates into the double 
threads (Fig. 7), and soon afterwards there appears the indication of the 
cross-segmentation. An early stage of this process is shown in Figs. 8 — 9. 
where the thicker portion of the spirem presents the appearance of 
long variously bent or looped rods. Each rod presents the paired ap- 
pearance and between each rod we can observe the thin part, which 
seems to be the drawn out portion of the spirem. The spirem is now 
very angular or uneven in its outline, as shown in Fig. 9. 
When the contraction of the spirem proceeds further, these thin 
portions of the spirem disappear, thus gradnally producing a certain 
number of chromosomes. At first every double chromosome takes the 
