Davis, — Cy to logical Studies on Oenothera, I, 553 
one or more small nucleoli, and a variable number of deeply staining chro- 
matic bodies of different forms and sizes (PL XLI, Figs. 1 and 8 ). The 
chromatic bodies are distributed rather uniformly around the periphery, 
closely pressed against the nuclear membrane, and frequently traces 
of a very open, delicate, connecting network may be observed. 
The nucleoli vary in size and in numbers. Instead of one large structure 
there are occasionally two or three nucleoli of approximately equal size. 
The smaller nucleoli may be so nearly the size of the chromatic bodies 
as to be scarcely distinguishable from the latter except . by their more 
spherical form and generally more central position in the nucleus. Small 
nucleoli frequently lie beside the largest, and stages indicating that they are 
absorbed by the latter are not uncommon, as has been reported by Gates (’08) 
for O. rubrinervis. 
The chromatic bodies are of especial interest in view of Rosenberg’s (’04) 
conclusion that structures, apparently similar to these, in the resting nuclei 
of Capsella and some other forms, represent the chromosomes. It is frequently 
possible to make counts of these bodies that approximate the sporophytic 
number of chromosomes, which is fourteen, but one finds other nuclei 
in which the number of evident bodies is less, and also nuclei with more 
numerous and smaller granules. The investigations of Overton (’05, ’09), 
which appear to have established the presence in the nucleus of prochromo- 
somes representing the chromosomes of mitosis, make it probable that 
similar conditions are present in Oenothera, but I have not as yet been able 
to follow the history of the chromatic bodies with sufficient exactness to be 
certain that they correspond to prochromosomes. 
The first indication of approaching mitosis is the appearance of a net- 
work distributed throughout the interior of the nucleus (Fig. 2 ), but denser 
in the peripheral regions. As it develops, the chromatic bodies decrease in 
size, apparently contributing their substance to the reticulum, and finally 
they can no longer be differentiated. The smaller nucleoli are then more 
easily recognized, since there are no longer chromatic bodies with which 
they may be confused. 
Certain strands of the chromatic network gradually thicken and become 
sharply defined as a coiled and looped thread, portions of which in early 
stages of development are frequently united to form a very open network 
(Fig. 3 ). This thread becomes the spirem which, when fully developed, 
is clearly continuous for a great length, although portions of the thread in 
places may appear to fuse at points of crossing. It is, of course, possible 
that all points of union during the development of the spirem, as described 
above, are merely those of adherence through the medium of accessory 
substance, and that the true chromatic spirem is at all times a single 
thread. 
The spirem at an early stage of development is composed of a single 
