338 
IVEY F. LEWIS AND CONWAY ZIRKLE 
form of small elongated granules about the periphery of the chromatophore 
(figs. 27, 28). These granules then fuse end to end and a weir tangled 
spireme results (fig. 29). A striking thing about this spireme is the way 
various strands lie parallel to each other. One is greatly tempted to see 
in this a conjugation or perhaps a splitting of the spireme. However, the 
members of a pair do not necessarily go to different cells. The spireme 
breaks into pieces of varying lengths, and these segments frequently with- 
draw into two distinct masses before the cell has started to constrict. 
More often, however, the chromatin is constricted in two with the cell 
(figs. 30, 31), and it is nothing unusual to see strands* extending some dis- 
tance into each daughter cell when the cells are connected only by a narrow 
isthmus (fig. 32). Two granules of chromatin occur regularly at the poles 
of the dividing cell at the maximum distance from the plane of constriction 
(figs. 30, 31, 32). These granules occupy these definite positions too often 
for this arrangement to be due to a mere fortuitous placing of waste chro- 
matin, though what function is served is not at all clear. In the majority 
of cases the segments of the spireme in newly divided cells lie alongside the 
new wall formed by the constriction (fig. 33). If any spindle fibers were 
present in this division, the technic used caused them to be dissolved, 
as no traces of them were found. The evidence at present indicates that 
a typical resting stage is not necessary between successive cell divisions 
especially if the conditions are just right for rapid growth. 
The dividing cells studied came from an agar culture in Chodat- 
Grintzesco solution. For a culture to thrive it must not be in a liquid 
medium or kept in the dark. No organic energy-yielding compound was 
necessary for rapid growth. 
In searching for mitosis in a primitive or degenerate plant, the investi- 
gator is exposed to the danger which beset the late centrosome hunters, of 
mistaking a chance resemblance for a homologue. The eccentrically placed 
globule seems certainly to be chromatin, and whether we call it a nucleus 
or a nucleolus depends upon the relative flexibility with which we use these 
terms. In regard to its fragmentation it resembles the nucleolus, though 
if it is the nucleolus it contains all of the chromatin at this stage, which is 
not typical. The amount of chromatin apparently increases greatly as the 
fragmentation progresses, and this increase is too great to be explained by 
the increase in the precipitation of the stain on the increased surface exposed. 
Some of the stainable material may come from the fragmented pyrenoid. 
The chromatin is arranged in a typical spireme, which breaks up into 
segments of diverse sizes which may safely be considered as analogues of 
chromosomes. There is no lining up on an equatorial plane or any indica- 
tion of the segments splitting and having their halves drawn to opposite 
poles. This method of nuclear division may be recorded as mitotic, but 
the mitosis is quite primitive and of an exceptional kind. 
The bearing of this method of nuclear division upon the systematic posi- 
