994 
Gates. — Somatic M itoses in Oenothera. 
occur. Thus, in over fifty cells of the nucellus the number was determined 
to be certainly 15, in three cases it was impossible to determine whether the 
number was 14 or 15 owing to the relative position occupied by certain 
chromosomes, in one cell only 12 chromosomes could be found though the 
cell was undisturbed by the knife, in two cells the number was certainly 16, 
and in one cell 20 or 21 chromosomes were found. 
These counts were all made with great care, nothing short of certainty 
with regard to the number being considered satisfactory. The counts were 
made chiefly in two stages of mitosis, (1) the late prophase just before the 
spindle begins to appear, and (2) the metaphase in polar view. In the 
former case only nuclei were considered which could be shown with certainty 
to be uncut. This stage is a peculiarly favourable one for counting the chromo- 
somes, because the latter are never in contact. They apparently repel each 
other, as they are invariably found at this time distributed just within the 
periphery of the nucleus (cf. PI. LXXXVI, Fig. 11). In the equatorial plate 
stage nuclei only were chosen for counting when the chromosome group was 
in a plane parallel to the plane of section and undisturbed (cf. Fig. 17). 
I mention these circumstances to show the care with which the counts 
were made. They show that the sporophyte number of chromosomes 
in this plant of O. lata was undoubtedly 1 5, although in nucellar tissue there 
are occasional variations to a lower or a much higher number. The signifi- 
cance of such variations is not at present clear. The smaller variations may 
be due to the temporary fusion or the fragmentation of certain chromo- 
somes, though it was impossible to obtain evidence for this from the size of 
the chromosomes themselves. Or they may be due, as E. B. Wilson has 
suggested, to the occasional failure of a chromosome to condense from the 
resting nucleus. In the single case in which about 21 chromosomes were 
to be seen in the equatorial plate (Fig. 21), the chromosome group appeared 
certainly larger than in normal nuclear plates, the cell also being larger than 
its neighbours. It seems probable that in this case the increase in the 
number of chromosomes occurred in some other way. Careful examination 
showed that all the chromosomes occupied approximately one plane, so 
that the increase in number could not have been due to precocious fission of 
certain chromosomes, unless, possibly, it had occurred in the prophase before 
the nuclear membrane disappeared and the equatorial plate was formed. We 
can, therefore, conclude that, whatever their significance or manner of origin, 
variations in the chromosome number in the nucellus do occasionally occur. 
The fundamental number of chromosomes, both in the somatic cells 
and the germ cells of this individual, is evidently 15 — one more than is 
found in O. Lamar ckiana. Miss Lutz (’ 09 ) has also reported 15 chromo- 
somes in the root tips of certain plants of O. lata as well as in several 
individuals of certain other mutants. As I pointed out (Gates, ’ 08 , p. 28), 
these deviations of one chromosome from the number (14) in O. Lamarckiana 
