ioo6 
Gates . — Somatic Mitoses in Oenothera. 
a marked change in the shape of the chromosomes of the root tip, often 
producing chromosome shapes closely resembling heterotype tetrads. 
The results, particularly of Nemec, in producing tetraploid and octo- 
ploid nuclei by subjecting growing root tips to chloral hydrate, are too well 
known to require discussion here. The question whether such tetraploid 
cells afterwards undergo reduction divisions which restore the normal number 
of chromosomes, or whether the disappearance of tetraploid cells from the 
root tips (which gradually takes place after subsequent growth) is wholly 
due to other causes, such as division into several smaller cells, fragmentation, 
and passing into permanent tissue, has been much disputed. Kemp found 
nothing in her studies which could be interpreted as true reduction figures, 
and she evidently, like Strasburger, doubts the existence of such. Nemec 
(TO, pp. 63-5) describes two types of reduction figures in chloralized 
roots : (1) indirect reduction figures which differ from the typical (a) in that 
the chromosomes are longitudinally split at the poles of the spindle, and 
(b) in their occurrence in abnormally large (tetraploid) cells ; and (2) direct 
reduction figures, which also occur in large (tetraploid) cells, and in which 
a segregation of the chromosomes takes place, half going to each pole 
of the spindle. 
If my interpretation of Fig. 37 of this paper is correct, the pairing 
of chromosomes in metaphase will be followed by a direct reduction 
division. Nemec (TO, p. 245) describes somewhat similar pairings with 
partial fusion of the chromosomes as occurring frequently in the root tips of 
Ricinus zanzibariensis (cf. his Figs. 92, PI. Ill, and 97, PI. IV). The 
possibility therefore remains that an actual reduction will not take place, 
though the large size of the cell and spindle suggests that something 
exceptional is occurring in the cases I have described. 
Summary. 
The results recorded in this paper may be briefly summarized as 
follows : 
The individual of O. tata from which the figures for this account of the 
somatic mitoses were drawn contained 15 chromosomes in its megaspore 
mother-cells and in the cells of the nucellus. In the nucellus, however, 
there were occasional cells containing a different number. Thus over 
50 cells were shown to have 15 chromosomes, one contained only 12, two 
contained 16, and one 20 to 21. These aberrant numbers were all, with 
one exception, observed in the metaphase equatorial plate, and it is possible 
that the higher numbers are to be accounted for by separation of certain of 
the daughter chromosomes immediately after their formation by the pro- 
phase split. If this is the real explanation of such cases, it does not follow 
that the increased chromosome number will be perpetuated, for the chromo- 
