STEVENS: KINOPLASM AND NUCLEOLUS. SI 
not simply radially arranged cytoplasm, and that in this instance 
the kinoplasm has a nuclear origin. ! 
My own studies of the development of the pollinium of Asclepias 
cornuti have brought out some evidence in regard to the kinoplasm 
and nucleolus which will now be given. The pollen mother cells 
are formed very early in the development of the flower. Figure 1, 
plate V, is a young pollen mother cell from a cross-section 
of a pollinium from a bud which was about 2 mm. in diam- 
eter. The mother cells, which extend in a radial direction 
across the entire pollen sac, remain in a resting state until the 
bud has attained a diameter of about 4 mm. Then the mother 
cells divide twice, producing a radial row of four pollen grains. 
After the mother cells have been differentiated from the arche- 
sporial celis they pass through a long period of growth before 
they undergo their first division. Thus the mother cell grows in 
length from .o85 mm. to .180 mm. and the nucleus of the mother 
cell grows in diameter from .o1r to.o17 mm. At the same time 
the nucleolus of the mother cell increases from .006 to .oo8 mm. 
in diameter, or nearly 2.5 times in bulk. While the nucleoli are 
relatively very large the amount of the chromatin in the nucleus is 
very little, as shown in Fig. 2, plate V, which represents a portion 
of a mother cell shortly before division. At this time the plasma 
membrane is the only means of separation between the mother 
cells and the tapetal cells. The latter form a layer from two to 
three cells thick and are likewise richly supplhed with nucleolar 
material. At the time of the division of the pollen mother cells 
their relatively enormous nuclei become dissolved and no trace of 
them, as a rule, is to be seen, either in the spindle or in the cyto- 
plasm, and so far as my preparations show this takes place before 
the solution of the nuclear membrane. The chromatin, which 
before the division of the nucleus was, for the most part, apphed 
to the nuclear wall in the form of a loose net-work, breaks up into 
twelve small chromosomes, which are so small, indeed, that the 
manner of their division cannot be made out with certainty. It can 
be observed, however, that there is a reduced number of chromo- 
somes formed in this division, for it is easily seen in a pole view of 
the nuclear plate of the archesporial cells that the number of 
chomosomes there is twenty-four. The small nuclear spindle con- 
verges to a sharp point at the two poles, and there is no trace of 
centrospheres, centrosomes, or of a multipolar origin of the spindle 
(Figs. 3,4 and 5, plate V). My preparations leave me in doubt as 
to the origin of the spindle, but I am of the opinion that in this in- 
11. c: pp. 120-121. 
