928 KirKkwoop: POLLEN-FORMATION IN CUCURBITACEAE 
liquid surrounding the pollen-mother-cells, where they seem never 
to occur in Micrampelts. 
The history of these bodies in Micrampelis is against such an 
interpretation, inasmuch as it would be difficult to account for 
their collection in the form described in a circumscribed portion 
of the cytoplasm. Moreover the change of form which they un- 
dergo seems to indicate that they are something more than mere 
passive by-products. Strasburger maintains that the cytoplasmic 
nucleoli bear an intimate relation to the kinoplasm, and supports 
his contention by citing the behavior of such bodies in Larix and 
other plants, in which they appear in connection with the spindle 
and other parts of the spindle-fibers and disappear at the conclu- 
sion of the division. He believes, however, that they are derived 
from the nucleolus, inasmuch as they appear as the nucleolus dis- 
appears and vanish again with its reappearance in the daughter- 
nucleus. 
In this view Mottier’* concurs, and chiefly on the basis of 
their staining reaction states that “there is no doubt that these 
bodies represent nucleolar substance.” He suggests furthermore 
that the presence or absence of extranuclear nucleoli may depend 
upon the activity or condition of the cell, in view of the fact that 
they may be present in or absent from cells of the same tissue in 
the same stage of development. That the bodies here under con- 
sideration in Micrampelis are of the same nature as those described 
by Mottier in Lz/ium is difficult to say, though it seems a fair as- 
sumption that they are. 
In Micrampelis no relation between these bodies and the nu- 
cleolus could be established. They appear in the cytoplasm long 
before the disappearance of the nucleolus and the nuclear mem- 
brane (FIGURE 24). That nucleolar material in solution might dif- 
fuse out through the nuclear membrane and be precipitated again in 
the cytoplasm is possible, but it seems highly improbable, and if 
so it might reasonably be expected to diffuse equally in all direc- 
tions. In Micrampelis, however, the appearance of the darkly- 
staining granules is at first in a particular part of the cytoplasm 
and that the region occupied by the fibers above referred to. 
These cytoplasmic fibers seem similar to those observed by 
Duggar ® in Symplocarpus and by Lloyd™ in Crucianella. In 
