A critical study of the cytology of Crepis virens. 
119 
gression are to be found in the same loculns. This fact, together with 
the exclusion of the hoUow spireme stage, somewhat facilitates the study 
of the history of the univalent spirenie, froni the time that it can be distin- 
guished as it passes out of synapsis, to its fmal connection as a univalent 
Segment of the heterotype chroniosome. 
Second contraction. 
As the bivalent loops of spireme go into second contraction, their 
component univalent Segments may still show indications of Separation. 
The line of Separation may be clearly visible line in the substance of 
the bivalent spireme (fig. 71), or the two univalent segments may diverge 
widely (figs. 70, 72 and 73). Sometimes the independent univalent seg' 
ments are üvisted round one another (fig. 71). 
Gradually most of the bivalent loops contract into a mass, which 
is often so dense that little detail is apparent (fig. 74); those loops that 
protrude are thick and homogeneous and seldom reveal their duplex 
character (fig. 74), owing to the dose reassociation of the univalent Seg- 
ments. 
Lightly staining globules may be exuded from the contracted chro- 
matic mass (fig. 74). 
Coming out of second contraction. 
Eosexberg (46) has noted (p. 69) that the number of loops that 
emerge from second contraction is in excess of that of the chronio- 
somes. 
There is much variety in the character and appearance of these loops 
as they extricate themselves from the chromatic mass of the second con- 
traction. In some, the associated pairs of univalent segments appear as 
fine threads united at mtervals by large beads of chromatin (fig. 76); 
whilst in others the univalent segments are independent lengths of con- 
centrated spireme (figs. 75, 77 and 78). Again hi others, the univalent 
spiremes are widely separated, and are only united at their ends to form 
a loop (fig. 75), and even this pomt of juncture may break do\ra (fig. 78.) 
Somethnes two imivalent segments may run closely parallel and be joined 
together by fine cross Connections (fig. 75). 
At this stage it is occasionally possible to recognise fission in the 
substance of a imivalent segment (fig. 78), the fission that will eventuaUy 
clcave the univalent chroniosome into its daughter halves on the homo- 
type spindle. 
