234 BOTANICAL GAZETTE [APRIL 
varies, since one, two, or sometimes even four such bodies may be nor- 
mally present in a resting condition. In many cases where these 
structures do occur, they appear to be surrounded by a clear space, 
and some,show a filamentous connection with the chromatin (figs. 
22, 44). In other instances (figs. 47, 48), no such clear space is seen, 
but the nucleole appears instead to be closely surrounded by a mass of 
chromatin. 
In optical sections, the nuclei of E. sciarae (fig. 57) show darker 
granules which are apparently connected by more lightly staining por- 
tions, thus giving an appearance corresponding to the common concep- 
tion of the chromatin and linin in the resting condition. But careful 
focusing reveals rather a more or less homogeneous, much convoluted 
thread, or filamentous material. Since I cannot, in fact, see any appre- 
ciable differentiation into chromatic and achromatic portions, I am 
inclined to regard the chromatin in this instance as resting in the form 
of a spirem thread. 
I am hardly prepared, however, to accept for these nuclei the ideas 
of VAN WISSELINGH (’99) and of GREGOIRE and WyGAERTs (:03), 
who think that there is no distinction between linin and chromatin. 
For, though it is true that in the resting nuclei of Empusa the nuclear 
material appears to be homogenous, during mitosis, on the other hand, 
some parts retain the stain much more tenaciously than other parts. 
One may bleach out an iron haematoxylin preparation, for example, 
until only that portion of the dividing nucleus immediately around the 
centers remains dark. However, whether this difference brought out 
by staining is due to mere physical causes, I cannot say. 
- Resting nuclei take the stain readily and are thus sharply differen- 
tiated; whereas those which are in a state of division stain less deeply- 
Hence in searching for division-stages, one has but to find those nuclei 
which are lightly stained and from which the color has been more 
washed out. But this applies apparently only to those nuclei which 
are somewhat advanced in the process, for such differentiation is not 
readily noticeable in very young stages. The earliest stages of nuclear 
division in the two species of Empusa in which I have studied. 
the phenomenon, E. sciarae and E. aphidis, in fact are not altogether 
clear. It is to be hoped that other species will prove more favorable 
for the beginnings of the process. It is not quite clear, for example, 
