McAllister—Cytology and Embryology. 633 
in time to the “second synapsis” first described by Miss Sargent 
for Lilium in 1896 and 1897, and since then reported by a num¬ 
ber of investigators among whom are Farmer and Moore, Mot- 
tier, Schaffner, Gregoire, Allen and Miyake. 
Figure 19 shows the beginning of this central aggregation 
drawn from a median section. Here it will be seen, the uni¬ 
formly distributed condition has been converted into a system 
of irregular coils consisting of broad loops, no two of which lie 
in the same vertical or horizontal plane. It will be seen that a 
part of each coil extends to the periphery of the nucleus, where 
for a considerable portion of its length it lies near or in contact 
with the nuclear membrane. It is the crossing and recrossing 
of the inner parts of these coils in the central region of the nu¬ 
clear cavity that gives the appearance of a central massing of 
the spirem which is characteristic of this period. 
The spirem continues to shorten and thicken and as a result 
the inner portions of the coils are drawn nearer and nearer to 
the center of the nucleus thus increasing the density and com¬ 
pactness of the central tangle of threads. The peripheral parts 
of the coils do not change greatly in their configuration but re¬ 
main open and for some distance in close proximity with the 
nuclear membrane, suggesting strongly that they are here at¬ 
tached to the membrane. (Fig. 21). 
This attachment of one side of these coils to the periphery of 
the nucleus, the opposite side of the coil remaining unattached 
and free in the nuclear cavity further suggests a possible ex¬ 
planation of the central tangle of spirem strands,—for as the 
spirem shortens and thickens, as it clearly does at this period,, 
the inner parts of the coils would be drawn toward the center 
of the nucleus and since they are being drawn in that direction 
from all sides they would necessarily cross and recross in such 
a way to produce a central tangle. This explanation of the 
phenomenon of “second synapsis” has been suggested by Jans¬ 
sens in his idea of a “tension nucleaire” which causes the seg¬ 
mentation of the spirem into chromosomes in Batrachoseps and 
in the Tritons. According to Janssens the loops of the bouquet 
stage following synapsis are whole chromosomes. They extend 
