96 BOTANICAL GAZETTE [FEBRUARY 
the halves of the chromosomes already formed by splitting in the 
anaphases of the heterotypic division. The heterotypic is thus the 
reducing division. The shape of the chromosomes of course makes it 
impossible to determine whether the second splitting is longitudinal 
or transverse. 7 
Fig. 40 shows a telophase in which the reconstruction of the 
daughter nuclei has begun. Delicate nuclear walls are appearing 
and the chromosomes have assumed angular, lobed, and irregular 
forms, but very little nuclear sap has yet appeared. Materials staining 
like chromatin are scattered through the cytoplasm. Fig. 41 shows 
the same stage, in which bodies as large as chromosomes and taking 
the same stain are scattered through the cytoplasm (particularly in 
the region of the daughter nucleus). It has already been shown, 
I think, that these bodies cannot be fragments of the nucleolus. 
Neither are they likely to be fat bodies, since the cytoplasm in earlier 
stages contained abundant storage nutritive materials in the form of 
starch. It seems then most reasonable to conclude that these bodies 
are actually chromosomes, and fragments of chromosomes which for 
some reason have become separated from their fellows on their way 
to the poles of the spindle. It is difficult to understand, however, 
what forces carry them into the cytoplasm, for among hundreds of 
spindles studied no irregularities in their structure have been observed 
which would cause the chromosomes to be distributed irregularly. 
Indeed the heterotypic spindle is notably regular in form. The 
earlier stages of spindle formation have not been studied, however, 
and it may be that these would account for some of the chromo- 
somes not being on the spindle. It is doubtful if the heterochromo- 
somes are ever attached to the spindle. Of course the position of 
chromosomes near the nuclear membrane is easily accounted for by 
their having merely failed to pass to the poles of the spindle, but 
this will not explain the position of many which are found scattered 
in the cytoplasm throughout the cell and frequently near its periphery. 
Fig. 42 is a polar view of a daughter nucleus in a later stage of 
reconstruction. The nucleus has grown greatly in size, owing to 4 
very rapid increase in the nuclear sap. At the same time the chromo 
somes have become drawn out into irregular shapes, are much reduced 
in size and stain less deeply. Several nucleoli have appeared, but 
