336 
DUNCAN S. JOHNSON 
the pollen grain. The cytoplasm about the smaller, more centrally 
placed, generative nucleus becomes denser and less vacuolated than 
the peripheral cytoplasm. Then a spherical mass of the latter, about 
half the diameter of the pollen grain, is cut off from the rest of the 
cytoplasm by a spherical cleavage surface or wall (fig. 50). The 
exact method of formation of this cleavage wall was not made out. 
Neither could a definite cellulose wall be demonstrated. In some 
spores, however, a considerable shrinkage space could be seen between 
the dense central and the vacuolated peripheral cytoplasm. This 
central, spherical mass may come into contact with the inner wall of 
the microspore, but never seems to become appreciably flattened against 
it, as often happens in the microspores of other angiosperms. This 
line of separation between the generative cell and the rest of the micro- 
spore is still visible in the ripe pollen grain (fig. 51). The nucleus of 
the generative cell has by this time become equal in size to the pollen 
tube nucleus. The brownish exine of the ripe pollen grain is greatly 
roughened by rounded wart-like protrusions, while the intine remains 
colorless, smooth and about three times the thickness of the exine 
(figs. 50, 51)- 
The germination of the pollen grain on the stigma and the fate of 
the two nuclei mentioned have not been observed. Structures have 
been seen in the interior of the style that stain differently than the 
surrounding tissue, and look like pollen tubes, but in no case could 
anything be made out of the contents of these tubular structures, and 
their identity as pollen tubes is not absolutely established. 
The above account of the development of the stamens and ripe 
microspores has been given in some detail for the sake of making it 
evident that there are no clearly primitive features in this phase of 
the development of this, in many respects, simplest of the Peperomias. 
Some details of the history of the chromatin of the spore nuclei have 
been given because the sequence of stages can be followed with such 
certainty in the successive flowers of the spike. It will also be im- 
portant for us to have these details in mind when discussing the origin 
of the embryo-sac nucleus, where the occurrence of a typical reduction 
division is of critical value in interpreting the nature of the mature 
embryosac. 
