646 Journal of Agricultural Research voi. xii, no. 10 



mass. Continued contraction of the bivalent spireme results in seg- 

 mentation. 



Segmentation. — In order to determine definitely whether a telosyn- 

 aptic or a parasynaptic arrangement of the univalent chromosomes 

 prevails during the synaptic and postsynaptic stages, it would seem nec- 

 essary to determine the exact procedure from the bivalent condition just 

 previous to segmentation, through segmentation to the paired condition 

 in diakenesis. If it can be shown that the bivalent threads appearing 

 during the pre and post segmentation stages are identical in Fragaria 

 spp., it will be a strong argument in favor of the parasynaptic arrange- 

 ment of the chromosomes. As Fragaria spp. is not complicated by a 

 second contraction and, as the segmentation stages are rapid, all stages 

 from that shown in Plate B, figure 7, to diakenesis being found within 

 a single loculus; and as the stages during this period are unusually 

 distinct, such a determination is not difficult. 



Digby {11) has recently presented the results of a very detailed study 

 of the cytology of Crepis virens in which the conclusion is reached that 

 there is an end-to-end arrangement of the chromosomes during the 

 synaptic stages. The details from the loosening of the synaptic mass to 

 segmentation are very similar to those in Fragaria spp., although the 

 chromatin in C. virens is apparently much more viscous and gives less 

 clear-cut images than does Fragaria spp. Her figures 76, 78, and 79 

 may be considered as in the same stages as my figure 7 of Plate B and to 

 present the same condition — that is to say, a split or a double spireme the 

 pairs of which are somewhat twisted about one another. She considers 

 that these figures do not show two univalent threads lying side by side, 

 but that the bivalent loops are due to the folding back upon each other 

 of univalent segments during second contraction. At this stage the 

 chromatin mass became so viscous that — 



it is generally impossible to individualize the future three bivalent chromosomes. 

 The chromosomes are in fact evolved out of what appears to be a tangle of viscous 

 nuclear contents. 



It appears hardly logical to conclude from this evidence that there is 

 an end to end rather than a side to side pairing of the univalent 

 chromosomes. 



The nimiber of loops present it segmentation in species of Fragaria 

 is always less than the number of chromosome pairs which appear at 

 diakenesis. It has been mentioned that these loops extend for long 

 distances on the periphery of the nucleus, forming more or less regular 

 equilateral triangles. When segmentation takes place, it is usually at 

 the outer angles of these loops and at or near the nucleolus, which gen- 

 erally forms more or less of a center from which the loops radiate. Thus 

 the bivalent loops are often divided into three pairs of bivalent chromo- 

 somes (PI. B, 9). The pairs continue to contract, those attached to 



