30 
Olive, Mitotic division of tlie nuclei of tlie Cyanopliyceae. 
count approximately twice tlie original mini her of cliromatin 
granules. Such a plane of lission, however, leaves tliis double 
tliread disadvantageously placed with reference to tlie plane of 
division of tlie cell which follows. Comparing with tlie process 
as seen in Oscillatoria and in tlie higher plants, we sliould ex- 
pect to find tlie subsequent plane of fission of the cell length- 
wise; whereas, in reality, we find it crosswise (figs. 73, 76). For 
a long time, tlie writer was at a loss to explain tliis curious 
discrepancy. but the finding of such nuclei as drawn in fig. 71 
furnished the clue to its solution. Judging from such appearanc- 
es, it becomes evident that tlie Separation of the chromosomes 
is acconiplislied simply by the pulling, or fiowing, of the two 
spireme threads in opposite directions, the one entering the one 
daughter cell, and the other being drawn into the otlier cell. 
In fig. 73, the daughter chromatin masses are completely sepa- 
rated, and we no longer see the elongated tliread arrangement; 
while in fig. 76, a transverse fission plane has cut in two the 
daughter cells, which liave not yet become rounded off at the 
cut end. 
We see, thus, in tliis species, two differences which separate 
Gloeocapsa widely from other Cyanopliyceae, — first, is the fact 
that the cell is cut in two -by simple constriction and not by 
a ring-formed wall; and, second, that liere we liave an excep- 
tional phenomenon in that the plane of division of the chronio- 
somes is abnormal. While, ordinarily, the plane of fission of the 
chromosomes is parallel to the subsequent plane of division of 
the cell, in Gloeocapsa , on the otlier hand, fission is at right 
angles to the resulting plane of division. 
Suminary of mitosis. 
It appears obvious to tlie writer that we liave at present a 
number of indisputable facts which point to the process of the 
division of tlie cyanophyceous nucleus as mitotic. And, further, 
that, although the process may be in some respects rather 
primitive, the essentials of nuclear division are, in the blue green 
algae, almost precisely similar to the well known karyokinetic 
processes seen in the higher plants. 
These facts are as follows: 
(1) Spin die. First, we liave in the dividing central body 
an achromatic figure, which consists of a central portion, situ- 
ated between tlie groups of separating chromosomes, and of a 
polar portion, corresponding in position to tlie niantle fibers, 
which lead from the chromosomes to the cross walls. The 
niantle fibers apparently liave to do witli the pulling apart of 
tlie divided chromatin granules. The complete achromatic figure 
evidently corresponds to the spindle, although it does not usually 
liave the common spindle shape. Instead, in tlie sliort celled 
filamentous species, it may liave tlie form of a more or less 
flattened, broad-poled disc. In the longer celled algae also, the 
central body assmnes somewhat the form of tlie cell in which 
