634 Fraser . — The Behaviour of the Chromatin in 
Early Prophases. 
The pollen mother-cells, before meiosis begins, are packed with a dense, 
finely granular cytoplasm. The nuclei are large (PL XLIII, Fig. x) ; they 
generally contain a single nucleolus, lying in a more or less central position, 
and the reticulum is spread just inside the membrane, so that it forms 
a hollow sphere. It consists of an irregularly diamond-shaped mesh, the 
origin of which may be traced to the longitudinal fission and cross connexion 
of the chromosomes of the preceding telophase. The spaces in the network 
are thus alternatively due to the pulling apart of the sides of the chromosome 
where fission has taken place, and to cross attachment between neighbouring 
chromosomes which, It must be kept in mind, are viscous structures. 
The first evidence that reduction is about to take place is given by the 
partial separation of the reticulum from the nuclear membrane (Fig. 2). The 
enlargement of the membrane has been recorded at different times by a num- 
ber of authors, and has been specially emphasized by Lawson ( 5 ) in 1911. 
While it is taking place the chromatin mass in Vicia contracts (Figs. 2, 3), 
still retaining its reticulate character, till it closes in about the nucleolus 
and the synaptic stage is entered. The nucleolus is sometimes squeezed 
out (Fig. 4) and lies at the periphery of the nuclear vacuole, where it may 
be recognized during the succeeding stages. 
It is very difficult to gauge the amount of contraction occurring at this 
period, since nuclei at the same stage of development are by no means 
necessarily of the same size, but measurements of neighbouring nuclei show 
that whereas, in a particular case, the diameter of the presynaptic nucleus 
in which the chromatin lay in contact with the membrane was 12*5 [x, the 
membrane after synapsis had begun was 15 n in diameter, and the con- 
tracting reticulum 9 by io/x, while the chromatin mass in full synapsis, in 
a vacuole of similar dimensions, measured only 7-5 fi across. These differ- 
ences are too constant to be accounted for by the inadvertent selection of 
tangential sections, too great, in otherwise uncontracted material, to be due 
to artifact, and it seems impossible* to doubt that in Vicia Faba a definite 
first or synaptic contraction of the chromatin occurs, in addition to the 
enlargement of the nuclear membrane. 
During the earlier stages of synapsis fine threads may be observed 
bridging across the space between the nuclear membrane and the reticulum 
(PL XLIII, Fig. 2). As contraction increases these are broken and finally 
disappear, being no doubt absorbed either Into the chromatin network or into 
the cytoplasm. This connexion between the membrane and the reticulum 
accounts for the fact that the presynaptic chromatin is held around the 
outer limits of the nuclear vacuole, and it may be surmised that a similar 
provision exists in vegetative nuclei. 
The synaptic mass is dense and impenetrable when ordinarily stained, 
