1922] RANDOLPH—MAIZE 345 
plasmic streaming. Association may be but momentary, or gran- 
ules which have been seen to come together may remain at- 
tached for a considerable period of time. Thus when a separation 
is seen to occur, it is not possible to tell whether one is observing an 
actual division of a single granule, or a reseparation of two granules 
formerly distinct. The fact that a paired arrangement occurs 
much more frequently than do groups of three or groups of four 
might be considered as suggestive, but in view of the observed 
behavior described, this can scarcely be considered as convincing 
proof of division. This point is of interest in view of the undoubted 
division of similar bodies which have become larger in older cells. 
The marked variation in the size of these bodies and the fact that 
evidence for their division in these early stages is uncertain raise 
the question of their nature, a question which will be taken up in 
the discussion. 
Observations on the living cells afford no evidence for the 
existence of more than one kind of body in the class of cell elements 
under discussion. An appearance suggesting the presence of two 
kinds of bodies, lighter and darker ones, is seen when the cells are 
examined with lenses of high magnification. This, however, is due 
to their relative position in the cytoplasm. In the living cells 
these minute bodies are continually in motion, and as they pass 
into the plane of focus first appear faintly colored, but when sharply 
in focus they are much darker in color, then become lighter again 
as they pass out of focus. Thus a single granule may appear 
light or dark depending on its position in the cytoplasm. The 
use of intra-vitam stains, such as are commonly employed in studies 
of cytoplasmic organization, and the fixing and staining methods 
which have been shown by recent workers to preserve the cytoplasm 
faithfully with its various inclusions, have also failed to differentiate 
distinct kinds. Treatment with osmic acid does not appreciably 
alter the appearance of these bodies (fig. 52). When treated with 
Benpa’s fixation and stained with haematoxylin, the cytoplasmic 
inclusions are well preserved, and their appearance is similar to 
that of living cells (fig. 53). Since these bodies have been found 
to occur as a constant feature of the cytoplasm of meristematic 
cells in maize, and inasmuch as they have been found to be definitely 
