162 CANNON: STUDIES IN PLANT HyBrRIDS 
cells as well. The ratio between the two was an inconstant one ; 
in this regard the cotton is different from the G/adzolus hybrid, 
and agrees with the pigeon and Canna hybrids, as will be seen by 
referring to the summary given above. I may also say that my 
principal object in making the study was to learn whether the 
mitoses were different from those in pure races of analogous 
plants, rather than to make an exhaustive study of the divisions 
as such, and for this reason my material was not suitable for the 
study of the earlier stages, however important that may be. 
I shall turn now to the morphological part of the paper. The 
divisions in the normal cells will be spoken of first, and then I 
shall give some account of the various irregularities which may 
be seen in the abnormal mitoses of the hybrid. . 
The archesporial cells arise by the division of subepidermal 
cells and become pollen mother-cells directly, no formation of 
sporogonous tissue occurring. After the two divisions of its 
nucleus, the mother-cell undergoes two simultaneous divisions by 
which the tetrads are formed, and these probably become pollen 
grains in the usual manner, although the later history of the cells 
was not learned. The mitoses leading up to the formation of the 
tetrads are as usual heterotypic and homotypic, and therefore are 
the same as are found in like mitoses of analogous organisms. 
The cytoplasm of the archesporial cells is of a recticular or 
alveolar structure, and this structure is uniform throughout the 
cytoplasm of the cell. That is, the portion of the cytoplasm 
which is immediately around the nucleus is not differentiated from 
the portion which is in the more peripheral part, but has a similar 
appearance and structure. This primitive, or undifferentiated con- 
dition of the cytoplasm is retained in all cells in which direct nu- 
clear division occurs, but on the other hand it may not be observed 
in the normal cells. Thus the structure of the cytoplasm forms 
an interesting and reliable check to the observations of the older 
stages of development of the cells. 
When those of the archesporial cells which are to develop into 
normal pollen grains become spore mother-cells, the cytoplasm 
withdraws from the cell walls. This retreat of the cytoplasm may 
also be observed in certain of the abnormal cells, for instance, 
in those which do not undergo degeneration in the earlier stages, 
