192 
The Ohio Naturalist. 
[Vol. XIV, No. 1 , 
Gates (4) in his paper on Oenothera rubrinervis states that 
the spirein constricts into fourteen chromosomes which break 
apart in pairs and then form the bivalent chromosomes by a 
folding together and fusion of the parts of each pair. Davis 
says there are ring-shaped chromosomes, some of which are 
linked together in O. grandiflora (1). He says these are present 
as soon as the sporocyte passes out of the svnizetic stage. In 
O. biennis (2) he finds a chain of fourteen chromosomes breaking 
into seven pairs from which seven chromosomes are formed by 
fusion. This method of chromosome formation of course is 
essentially the same as that of loop formation, but I have found 
the loops defiinitely formed and just as definitely contracting 
until there are seven chromosomes formed from the seven loops. 
These results are the same as were found in Fuchsia (8). The 
loops frequently form quite definite rings as is seen in figure 1G. 
In figure 19, the chromosomes still show something of their 
ring and loop character and there are two nucleoli shown. The 
next figure (Fig. 20) shows a certain amount of loose material in 
the nucleus which may be derived from the nucleolus although 
there is no direct evidence for this conclusion. The next two 
figures (Figs. 21 and 22) show the chromosomes broken apart and 
the cytoplasm flowing into the nuclear space. The nuclear wall 
has entirely disappeared. In the cytoplasm are seen great num¬ 
bers of prominent granules. These remain in the cytoplasm 
throughout the reduction process. Whether these are starch or 
not was not definitely determined. Figure 23 shows the beginning 
of the formation of the spindle with the chromosomes being 
drawn into the equitorial plane. Figure 24 is the mother star 
stage at the time when the chromosomes begin to be segregated 
into the univalents. The next two figures (Figs. 25 and 26) do 
not show the full quota of chromosomes but show the beginning 
of the true reduction in those that can be seen. The next two 
drawings (Figs. 27 and 28) represent metakinesis stages with the 
chromosomes half way to the poles. Figures 29 and 30 are 
daughter star stages. The lower pole of figure 30 shows a slight 
beginning of the nuclear wall. The seven univalent chromosomes 
are about half the size of those appearing on the mother star. The 
number can be easily counted at this stage. 
Following this stage the nuclear membrane develops rapidly 
and the daughter nuclei swell to a much larger size. The chromo¬ 
somes remain as distinct bodies although there is some distribu¬ 
tion of the chromatin material (Fig. 31). Even in the resting 
condition the chromosomes in the two daughter nuclei remain as 
seven distinct bodies and there is no real reticulum developed 
(Fig. 32). At this stage all traces of the spindle have disappeared. 
Soon after, the second division begins (Fig. 33) and the chro¬ 
mosomes in the mother star are again distinctly visible as small 
