Gales —Somatic Mitoses in Oenothera. 
997 
At the stage represented by Fig. 7 the last threads of the reticulum 
have been withdrawn, and the nuclear ‘cavity is now occupied only by 
karyolymph in which the chromosomes and nucleoli float. They are 
peripherally arranged and undergo progressive condensation as showm in 
Figs. 7-10. This is the best stage of all for counting. 
In a previous paper (Gates, T1 b } p. 931) I reached the conclusion, 
after a preliminary examination of somatic prophases in O. gigas , that a 
continuous spireme is produced, which segments to form the chromosomes. 
A study of the earlier prophase stages shows that this conception requires 
modification. The chromosomes are coiled spirally round and round the 
nuclear cavity in diakinesis in such a way as to give frequently the appear- 
ance of an end-to-end arrangement. But they do not originate in an end-to- 
end position, as shown by Figs. 4-7. They originate, as already described, 
by a progressive condensation of the threads in certain portions of the 
reticulum, in a manner similar in some respects to that described by Stomps 
(TO) in the pollen mother-cells of Spinacia . When first differentiated 
from the reticulum they are neither end to end nor side by side, but are 
irregularly arranged with regard to each other, a single chromosome in the 
earliest stage of its appearance sometimes extending across the whole 
diameter of the nucleus (Fig. 4). They are also at first much coiled and 
very irregular in shape, often forming a loop, though usually of nearly 
uniform thickness throughout. They thus correspond to Boveri’s (’ 09 ) 
conception of the origin of the chromosomes from the resting reticulum 
as shown in the blastomeres of Ascaris. While we may consider that each 
portion of the resting reticulum represents in general an alveolated chromo- 
some, yet anastomoses between adjacent parts, as well as perhaps move- 
ments between parts of the nucleus, must cause more or less intermixture of 
portions of the reticulum derived from separate chromosomes. This may 
perhaps account for the manner in which the chromosomes are intermingled 
when they first appear, but it of course assumes a more strict individuality 
of the resting chromosomes than can be visibly demonstrated. 
I was unable to obtain any satisfactory evidence that the chromosomes 
are paired at all in arrangement when they are first differentiated from the 
reticulum. Later, in Figs. 8-1 1 , indications of a side-by-side pairing begin 
to appear. In these figures each nucleus is drawn in two parts, representing 
the chromosomes seen in upper and lower focus. In each case the full 
number of 15 chromosomes is present. In Figs. 8 a and 8 b the chromo- 
somes are still considerably looped and twisted. In Figs. 9 and 10 they 
have condensed to a shorter and thicker, more rod-like shape. Delicate 
(linin ?) connexions are still to be seen attached to some of the chromosomes 
in Fig. 9. 
In a former paper (Gates, 11 b) I mentioned that the chromosomes 
in somatic prophases usually undergo a longitudinal split. A single 
