Division in Mnimn ho mum, .151 
chromosomes each as a dumbbell-shaped structure. As before described*, 
this appearance is produced by the two arms of the V seen in optical 
transverse section. No evidence for longitudinal fission in the chromosomes 
was discovered during their passage to the poles. At this stage one 
chromosome is often found lagging behind the rest near each pole of the 
spindle (Fig. 30). This, no doubt, is caused by the retarded division of 
one of the bivalent structures. 
No subsequent stages of the heterotype division were discovered, and 
the reconstitution of the daughter-nuclei must proceed very rapidly. It 
could not be determined whether these latter regain the completely resting 
condition, but, judging from analogy with other forms and from the short 
period allowed for these changes, this seems to be improbable. The 
telophase of the homotype division was found, and in this six chromo- 
somes are easily distinguishable at each pole (Fig. 31). The divisions of 
each of the two nuclei resulting from the heterotype mitosis take place in 
planes at right angles, and in the stage figured one division is seen from 
the side, while a polar view is obtained of the corresponding process in the 
other cell. A definite cell wall separates the two divisions, and although 
no such structure was found in the telophase of the heterotype division, its 
formation must take place before reconstitution of the daughter-nuclei. 
After the homotype mitosis walls are formed across the spindles of the 
two final divisions. These join up with the wall already present, and four 
daughter-cells are thus produced (Fig. 32). Spindle fibres can still be seen 
between the pairs of resting nuclei formed by the last mitosis. These 
nuclei possess a small lightly-staining nucleolus and a rather coarse 
reticulum ; at this stage numerous deeply-staining granules are found in 
the cytoplasm, particularly in the neighbourhood of the nuclei. 
Considerable growth takes place on the differentiation of the spores, 
and the spore-sac becomes completely filled with the developing tetrads*. 
Very soon the spores become closely packed, and, on account of movement 
of the cells due to pressure, the original arrangement is soon lost. At this 
stage the spores only possess one coat ; the nucleus is fairly large and 
possesses a distinct nucleolus, but the latter is not so large and evident as in 
the case of other cells of the plant. No details of the formation of the 
various walls during spore formation are discussed here, and this will form 
the subject of a separate communication. 
: ? I.'-; . : " • • r : V ■ / .,1 * l ' ■ ■ ;.\d f > 
General Discussion. 
At the present time the descriptions of the course of events during the 
early prophase of the somatic divisions given by various investigators show 
considerable agreement. The accounts of Strasburger ( 23 ), Farmer and 
Shove ( 9 ) , and Gregoire ( 15 ) agree in the essential facts. The import- 
ance of fixation in the elucidation of the early stages cannot be 
