740 Digby.—The Somatic , Premeiotic , and 
only between the four lobes of a single anther, but also in each individual 
lobe itself. 
During synapsis the ‘ chromatic bodies ’ as described by von Derschau 
( 2 ) are thrown off in abundance (Figs. 42 and 43). There is at this time 
a great increase in size of the anthers. The contained pollen mother-cells 
begin to separate from one another. Details concerning the extrusion of 
the ‘ chromatic bodies ’ and the rounding off of the cells have already been 
given ( 3 ). 
Hollow Spireme. 
Prior to coming out of synapsis the knot loosens. Shavings of a knot 
at this stage show it to consist for the most part of irregular masses of 
chromatin. These masses may occasionally be broken up into smaller 
granular portions. The knot unravels, and its substance emerges in the 
form of loops or strands (Fig. 44). Some of these may be more or less 
ribbon-like (Fig. 44), whilst others may be formed of large beads strung 
together (Fig. 45); some may show marked longitudinal fission (Fig. 45 ), 
whilst others may be homogeneous (Fig. 46). Lengths of spireme may be 
thick for a certain distance, and then their sides may divaricate widely 
(Fig. 48) and join other strands; others may be very delicate and anasto¬ 
mose freely and show no order which could be co-ordinated into a scheme. 
This great irregularity has been emphasized by Marechal and Saedeleer ( 17 ). 
In Galtonia it is evident from subsequent events that the spireme at 
this stage is univalent in nature, that is to say that it represents lengths of 
single somatic chromosomes as described by Farmer and Moore ( 4 ), and 
not lengths of paired somatic chromosomes as described by Gregoire (10) 
(1907); consequently that the fission in the threads is homologous with 
the parallelism found in the presynaptic stages and in the somatic pro¬ 
phases ; further, that this fission will ultimately divide each univalent 
chromosome into two daughter chromosomes at the second meiotic division, 
and will not separate two univalent chromosomes at the first meiotic division. 
As will be shown, in Galtonia the homologous lengths of spireme do 
eventually fuse into thick bivalent strands (Figs. 53 and 54) which split 
apart, after second contraction, into their two component chromosomes 
(PI. LXII, Figs. 58 A and 58 B, &c.). Although it is only at a later stage 
that the univalent strands pair, yet when the spireme emerges from synapsis 
there are already indications of the joining together of lengths of homolo¬ 
gous spireme. The thick loops (PI. LXI, Fig. 44) are striking examples of 
a telosynaptic union of univalent chromosomes. At the end of the loop 
there is often a swelling to be seen as if the junction of the two homologous 
lengths of spireme, each of which will be a univalent chromosome, caused 
some physiological disturbance. The sides of the loops are sometimes 
beaded (Figs. 45 and 46), sometimes they are more homogeneous (Fig. 44), 
