928 Gates.—Pollen Formation in Oenothera gigas. 
under a high power. The chromosomes in these figures appear considerably 
larger than in the figures previously described. This is not due, however, to 
their containing more chromatin, but to the fact that many of the chromo¬ 
somes are less dense and compact, and therefore stain less deeply. But 
although not alveolated, they occupy a larger space, and therefore of course, 
although they stain less deeply, appear to contain more chromatin. On 
account of this difference in compactness, the chromosomes differ greatly in 
apparent size, and in PI. LXIX, Figs. 59-61, PI. LXX, Fig. 62, while most 
of the chromosomes are in evident pairs, it is not possible in every case to 
say which are pairs and which single chromosomes. 
Figs. 63 and 64 are anaphases of the heterotypic mitosis, showing the 
polyhedral character of the cells. Figs. 65-67 are later stages, showing the 
heterotypic telophase. All are from the same flower. In Figs. 65 and 66 
the heterotypic spindle is in different stages of its disappearance, and 
a definite cell-plate is laid down. In Fig. 65 the interkinetic chromosomes 
are very compact and show no indication of bivalence, while in Fig. 66, 
which is a later telophase, the daughter nuclei have grown to larger size, 
the chromosomes are less compact, and most of them show clearly a bivalent 
structure. 
In this telophase the chromosomes frequently undergo a sort of swelling 
or distension without vacuolation. They thus occupy a larger area, as 
already mentioned, but are less deep staining and less compact, though 
without any open spaces. Then stringing out at the corners occurs, and 
then the vacuolation process may begin. 
Even the smaller nuclei in Fig. 66 afford evidence that their size 
depends upon the amount of chromatin they contain. In all the Figs. 65-67 
there is evidence of irregularities in the chromosome distribution. It will 
be seen also that these mother-cells are polygonal, and still in contact with 
other cells on all sides. 
One of the most striking features of these sterile anthers is the enormous 
variation in the size of the mother-cells. In normal development the pollen 
mother-cells, even when still in the archesporial condition, are approximately 
equal in size. In these sterile anthers, one mother-cell may have many 
times the volume of an adjacent mother-cell of the same anther. 
Fig. 68 shows a mother-cell in the prophase of the homotypic mitosis. 
There was an incomplete attempt at segmentation (?) of the cytoplasm after 
the first division. These prophase chromosomes nearly all show their 
bivalent structure. Fig. 69 represents a homotypic telophase. The daughter 
nucleus in the higher focus on the left, being uncut, shows fourteen chromo¬ 
somes. Here also there has been an incomplete attempt at a segmentation of 
the cytoplasm. Fig. 70 represents a condition sometimes occurring, in which 
a complete cell-wall is formed following the heterotypic mitosis. The two 
daughter cells are each in the telophase of another (the homotypic) mitosis, 
