636 Davis.—Cytological Studies on Oenothera . //. 
mitosis the segments as fully developed chromosomes separate from one 
another. 
During the process of condensation and shortening of the chromosome 
segments in the spireme there is frequently found near the nucleolus a 
massing of the chromatic material (Figs. 17 and 18) which bears a 
superficial resemblance to the synaptic knot (Figs. 8-10). This is appa¬ 
rently the condition termed the * second contraction ’ and considered by 
some authors as of considerable significance, since it is supposed to bring 
the chromosomes into close association as pairs in preparation for the stage 
of diakinesis. The material, however, offered little support to such an 
explanation, because the chromosomes exhibit only slight tendencies to 
group in pairs, and it is doubtful if the stage itself is a regular occurrence. 
It is hard to see for the species under consideration any significance in 
this ‘ second contraction ’ further than an assemblage of the chromosome 
segments of a spireme so complexly looped that its parts are necessarily 
drawn close together during the condensation which transforms the four¬ 
teen segments into chromosomes. 
It is interesting to note the differences in the formation of the 
chromosomes between this material of Oenothera biennis and that of 
O. grandiflora. The spireme of grandijtora emerges from the synaptic 
knot as a much more complexly coiled thread (Davis ’ 09 , Figs. 16-18) 
than that of biennis , and the subsequent condensation and shortening of 
the spireme draw the chromatic material into a closely contracted mass, 
from which several thick loops extend into the nuclear cavity (Davis ’ 09 , 
Figs. 19-23). These latter conditions might be considered representative 
stages of a ‘ second contraction ’, but in grandiflora the changes of structure 
are so gradual and the variation of form is so great that it is very difficult 
to distinguish such a phase. Finally the thickened and contracted spireme 
of O. grandiflora breaks up into seven rings, some of the loops being 
transformed directly into these structures. The later history shows that 
each ring is bivalent in character, since it separates into semicircular halves 
(sporophytic chromosomes) during the heterotypic mitosis. There are 
then fourteen sporophytic chromosomes as in biennis , and these are like¬ 
wise segments of a spireme arranged end to end, but the involved coils 
of the spireme bring the segments together in loops in such a manner 
that the fourteen chromosomes become grouped in pairs to form seven 
rings (bivalent chromosomes). It would be hasty to conclude that the 
close association of the chromosomes in pairs, characteristic of the material 
of grandiflora studied by the writer, involved any greater affinity between 
the segments of the spireme than in biennis . The striking differences be¬ 
tween the two species in the final grouping of the chromosomes are perhaps 
simply due to the peculiarity of a more complicated looped spireme in the 
case of grandiflora. Furthermore, it will not be surprising if the study of 
