194 Farmer and Digby.—On the Cy to logical Features exhibited by 
The details of the meiotic phase were carefully studied in the hope that 
they might serve to throw light on the processes that obtain during and 
after synapsis, but they proved too difficult for satisfactory analysis. Cases 
were often seen which might have been interpreted as due to a lateral 
approximation of two parallel filaments (PI. XVI, Fig. 2), but it was impos¬ 
sible to feel sure that such an explanation was not illusory. We have, 
however, seen examples of nuclei which at this stage very closely resembled 
the excellent figure of Janssens’ illustrating the corresponding stage in the 
spermatocyte of Alytes} But as we observed numerous other instances in 
which more than two filaments were apparently involved in the thickened 
thread, the evidence for approximation at this stage was less cogent than it 
would otherwise have been. It must also be remembered that during 
synapsis there is in any event a bunching together of a tangled thread- 
work, and that it would consequently be singular rather than otherwise 
if parallelism between some of the threads did not occur. This should 
be no less true for those filaments which extend from the central tangle 
to the nuclear periphery than for the rest. Indeed, on merely mechanical 
grounds, such parallelism might be expected to be found in these threads 
quite apart from the question of any significance arising such as has been 
attributed to them on theoretical grounds. We have not succeeded in 
convincing ourselves that the apparent reduction in the threadwork of the 
nucleus which certainly does occur at synapsis is to be explained otherwise 
than as the result of contraction and concomitant thickening of the leptotene 
filament. 
During the very earliest stages of synapsis the cytoplasm in the 
vicinity of the nucleus begins to exhibit that structural modification termed 
kinoplasm by Strasburger. It is probable that the extrusion of nuclear 
substance (PL XVI, Fig. i), in the form of granules or droplets of chromatin, 
is concerned with its appearance ; the kinoplasm surrounds the nucleus 
as a fine feltwork, and recalls the arrangement described and figured by 
Wilson Smith 1 2 for the pollen mother-cells of Osmunda. The spindle in 
P. aureum is very regular, and shortly after synapsis rapidly assumes 
a bipolar character, in sharp contrast to what happens in the other parent. 
We examined several other species of the genus, e. g. P. guatamalense , 
Dryopteris , &c., and found them to resemble P. aureum in this respect. 
When the spindle is formed (PI. XVI, Fig. 3), the granules in the cyto¬ 
plasm, just referred to, are fairly numerous. They are often to be 
found in the neighbourhood of the spindle poles, but do not seem to exert 
any appreciable influence on the distribution of the fibres, although some- 
1 Janssens and Willems, La spermatogenese dans VAlytes obstetricans. La Cellule, t. xxv, 
P1 - b Fig. 3. 
2 Wilson Smith, The achromatic spindle in the spore mother-cells of Osmunda regalis. Eot. 
Gaz., xxx. 
