Davis.—Cytological Studies on Oenothera. II. 635 
angular and united, while Figs. 8-10 are from sections in which the cells 
had rounded off and separated. 
As synapsis proceeds it becomes very clear that the threads thicken 
(compare Figs. 9 and 10 with Figs. 7 and 8), and there is good evidence 
that the system becomes shorter. Thus the threads are very much more 
prominent in the later stages of synapsis (Figs. 8-10) than at the beginning 
of the contraction (Figs. 6 and j), and in the later stages the structure begins 
to take on the appearance of a much-looped and coiled spireme. This 
history of the differentiation of the threads during synapsis is the same as 
that for O. grandiflora (Davis ’ 09 ). 
It will be noted that the term synapsis is reserved for this period of 
the first contraction of the chromatic material. A type of contraction that 
frequently takes place much later, during the formation of the chromosomes, 
will be described in that connexion. 
The Formation of the Chromosomes. The period of synapsis, 
described above, is of relatively long duration, as is shown by the fact that 
in an inflorescence one to three buds may be found in which all, or almost all, 
of the nuclei in the pollen mother-cells are in this stage. The succeeding 
developments, leading to the development of the thick spireme from which 
the chromosomes are formed, take place more quickly. The chromatic 
material emerges from the contracted condition of the synaptic knot by a 
loosening of the coils, with the result that a thickened and very much 
shorter thread appears, which is evidently a spireme. Such a stage is shown 
in Fig. 11, which is of especial interest in comparison with Fig. 10, since the 
two nuclei were from opposite ends of the same pollen chamber. The 
spireme even in this thickened and much shortened condition (Figs. 11-13) 
is generally complexly looped and coiled, but the greater part of the thread 
system can be traced generally without great difficulty, which is. impossible 
in the stages of synapsis. 
Shortly after the emergence of the thickened spireme from the synaptic 
knot there appear indications of a process of segmentation that is to trans¬ 
form the spireme into a chain of fourteen chromosomes. Early stages of 
this process are shown in Figs. 12 and 13, where the segments are present 
as long, variously bent rods with thin regions between them, as though 
separating by constriction. The contraction which has been going on in 
the spireme continues in these segments until their long diameter is not more 
than two or three times their width, and by this process of condensation 
the spireme becomes so much shortened that the complexities of the looped 
arrangement largely disappear, and it is frequently possible to follow the 
chain of fourteen chromosome segments from end to end as a single seg¬ 
mented spireme (Fig. 14). Later the spireme may become broken into 
groups of chromosome segments (Figs. 15 and 16), each group showing the 
segments arranged end to end ; finally during the prophase of the heterotypic 
