116 
L. Digby 
chromatic presynaptic pliases, and did not observe those characterised 
by the presence of chromatic bodies. He describes the presynaptic stages 
of certain Compositae inclnding C. virens as “a ratlier coarse network with 
chromatin aggregates which are irregulär in size and form, and of no con- 
stant nnmber” (p. 707). And agatn, “so far as the presynaptic poUen 
mother-ceUs of the Compositae are concerned, I find definite prochromo- 
somes or gamosomes to be nonexistent” (p. 708). The occurrence of various 
types of prophases proceeding on different Hnes towards the same end has 
already been described by the Schreiners (50), by Gregoire (20) and 
by Mottier (35). It is proposed to defer the account of these fine pre- 
synaptic stages to the end of the description of the meiotic phase. 
Synapsis. 
Close synapsis has a finely granulär appearance. The nucleolus 
can be seen lying within, but apparently independent of, the synaptic 
knot (fig. 61). 
The density of the synaptic knot obscures the course of events there 
taking place, but it is noticeable that when the synaptic knot is about 
to loosen (figs. 62 and 63), the chromatic beads are larger and more definite 
than those of early synapsis (fig. 61). 
Düring close synapsis fine linin Strands, and occasionally thick loops 
of spircme (fig. 62), escape froni the laiot. This stage is of long duration 
as it is not uncommon to find all the pollen mother-nuclei of a bud m 
close s 5 Tiapsis. 
AMien the knot begins to imravel it may give off roundcd globules 
of a faintly staining substance (fig. 63). These droplets pass out of the 
nuclens into the surrounding cytoplasni where they remain throughout 
the second contraction pliases as very bright rcfractive chromatic staining 
bodies (figs. 64, 69, 70, 74 and 76). In Galtonia candicans (8) similar glo- 
bules of eu-nucleolar substance escape from the synaptic knot, and these 
subseqnently become chromatic. Lubbienko and Maige (28) have de- 
scribed the extrusion of nucleolar fragmehts into the cytoplasni. In Crepis 
virens this budding may lie contmued throughout the unravelling of the 
synaptic knot, and may still be proceeding during second contraction 
(figs. 74 and 75). Sometinies the ivhole s^maptic knot breaks up into 
chromatic portions, but such extreme fragmentation undoulitedly fore- 
shadows nuclear disintegration. From s^niapsis onwards, individual loculi, 
or even all the loculi of an anther, may show complete disorganisation 
of their pollen mother-cells. Rosenberg (46) has stated that sometinies 
the action of the. fixing fluid has been so violent “daß der Kern hier und 
