1906} MERRIMAN—NUCLEAR DIVISION IN ZYGNEMA 49 
pletely cohering, while a few appear lying in delicate strands about 
them (jig. 31). In some cases all the chromosomes may cohere to 
form the central body. The nuclear membrane now emerges from 
the condensation of cytoplasm (fig. 32). As the chromosomes are 
now shut off from the influence of currents in the cytoplasm they 
generally remain unchanged in position, fusing either to form one 
mass (fig. 33), or three or more smaller masses (fig. 34), or rarely 
(jig. 35) all the chromatic material may be diffused in the nuclear 
plasm, forming numerous more or less tetrahedral granules. 
It is to be noted that not until the chromatic rings have separated 
and have approached the chromatophores do the pyrenoids ordinarily 
show any evidence of division. Tviis observation was easily con- 
firmed from the study of living cells. Fig. 29 represents the only 
one seen, out of many filaments examined, in which the pyrenoids 
divided before the formation of the nuclear membrane. As the 
newly divided nuclei approach their respective chromatophores, 
one or both plastids begins to show a constriction. This deepens 
until when the nuclei come to lie directly over, only a narrow band 
of less dense substance resembling linin connects the two daughter 
pyrenoids (figs. 37, 38). This becomes gradually reduced until it 
appears only as a thread (fig. 39). Later the nucleus sinks down 
and the separation is complete. The constriction of the plastids 
forming the center of the two pyrenoids takes place synchronously, 
as is the case with the stages in the daughter nuclei. One instance 
only was observed in which one plastid suffered division when other 
plastids had just begun to elongate (jig. 32). 
Although division of the pyrenoid may be influenced by division 
of the nucleus, that it is not wholly dependent upon it was demon- 
strated by leaving actively dividing filaments of Zygnema for one 
hour in a watch crystal containing 10°° of water to which two drops 
of chloroform were added. There were but few visible signs of plas- 
molysis in filaments killed and stained as in other material, but while 
a majority of the nuclei had ceased to divide, a majority of the pyre- 
noids were dividing as in normal filaments. That this division was 
not merely fragmentation was shown by sequence of stages and the 
presence of the band connecting the plastids. Fragmentation of 
the pyrenoids took place in filaments in stronger solutions of water 
Mo. Got.Garacn 
1906 
