347 
the Embryo-sac Mother-cell of Li Hum . 
lowing close upon synapsis is shown in Fig. 7. The spirem has not become 
completely distributed within the nuclear cavity. It is, moreover, shorter 
and thicker. A longitudinal fission cannot be made out, although here 
and there a few chromomeres are seen to be double. The developmental 
stage of the ovule (Fig. y a) shows that we have a stage in mitosis only 
a little later than Figs. 5 and 6, since the inner integument has just attained 
the height of the nucellus. Fig. 8 is a stage slightly older than Fig. 7. In 
a number of places the longitudinal fission of the chromomeres can be seen, 
and the whole thread is rapidly approaching the stage of the hollow spirem. 
The nucleolus, which lies in a neighbouring section, presents the same 
appearance as in the preceding figure. 
The somewhat closely entangled spirem of Fig. 8 soon passes over into 
the typical hollow spirem. At this stage there enters a phenomenon 
which has generally led to confusion in the interpretation of subsequent 
events. Generally speaking, the spirem in this stage in both micro- and 
megaspore mother-cells is a rather uniform cord more or less regularly 
and loosely disposed in the nuclear cavity. It may, or may not, reveal 
the presence of a longitudinal split. Occasionally this longitudinal fission 
is seen here and there in the cord, where the halves show a tendency to 
spread apart (Fig. 9), but more frequently no evidence of a fission is 
apparent. In Fig. 9 only a portion of the chromatin cord is shown. Com- 
paring the ovule containing this nucleus (Fig. ga) with that of the preceding 
figure (Fig. 80), it will be seen from the height of the integuments that the 
two nuclei represent very closely related stages in mitosis. It may be 
well to bear in mind, however, that the author does not regard the form 
of the ovule as the sole guide in determining the mitotic stage, but it 
has been found all along that certain steps in the growth of the ovule 
correspond closely to definite mitotic stages. While the rather uniform 
chromatin cord, with or without indications of a longitudinal fission, is 
the rule in the stage of the hollow spirem, yet there are found nuclei 
in which the spirem is more slender and the longitudinal halves more 
widely divergent than in Fig. 9, — the phenomenon referred to in the 
foregoing. A part of such a nucleus is shown in Fig. 10. Here the 
longitudinal segments diverge, becoming widely separated for relatively 
long stretches. Where short turns of the divergent halves are cut by 
the knife, the two pieces appear as two isolated portions lying somewhat 
parallel. Attention was called by the writer to a similar condition in 
the pollen mother-cells (Mottier, ’ 07 ), except in the latter this phenomenon 
seemed to appear more frequently at a later stage, — that of the looping 
or ‘ second contraction ’. Whether this divergence of the longitudinal 
halves occurs as frequently here as in the pollen mother-cell is not known, 
for in any study it is not possible to observe as many megaspore as 
pollen mother-cells. That Figs. 9 and 10 represent about the same, or 
