346 Mottier . — On the Prophases of the Heterotypic Mitosis in 
tinuous thread; which is to appear later. In fact one sees now, in Fig. 3, 
the beginning of such a pre-synaptic thread, although there is not yet 
a continuous spirem. At this stage one large nucleolus is usually pre- 
sent, with sometimes one or more small ones. However, there may be 
considerable variation in regard to the number of smaller nucleoli. In 
preparing the ovary from which Fig. 3 was taken, not only was the outer 
wall of the loculus but a considerable slice of the chalazal end of the 
ovule cut away (Fig. 30), so that the killing fluid undoubtedly had a 
more direct access to the megaspore mother-cell than in the case of Fig. 2. 
In this ovule the outer integument had just become evident. Both cell 
and nucleus are elliptical in longitudinal section. Following close upon 
the stage of Fig. 3 is the beginning of the synaptic contraction (Figs. 
4, 5). Sections of the ovules (Figs. 4 a, 5 a) to which these nuclei belong 
show that the stage of mitosis is only slightly in advance of that in Fig. 3. 
The nucleolus has the form and position frequently seen in this stage 
of synapsis. That there is a single thread with a single row of chro- 
matin granules cannot be questioned, and that there is no union side 
by side of two spirems is equally evident. Figures 4 and 5 represent 
rather thick sections of the nuclei, but they do not include the whole of 
their respective nuclei. In Fig. 5 the contraction of the thread has pro- 
gressed further than in Fig. 4. The spirem seems thicker and more 
definite, whilst the chromomeres or granules are perceptibly larger. In 
the case of these two ovules, the wall of the ovary was cut away, giving 
the fixing fluid direct contact with the ovules from the instant the 
ovaries were thrown into the reagent. In Fig. 6 is shown the compactly 
balled-up synaptic mass, the stage so frequently observed, and the one 
almost invariably figured in the literature dealing with the heterotypic 
mitosis. Here the mass is so dense that little in detail can be made 
out. It seems, however, that the chromatin granules in this nucleus were 
much larger and less uniform in size than would result from a compact 
balling-up of the structure shown in Fig. 5. In preparing the ovary from 
which Fig. 6 was obtained, the outer wall of the loculus was not entirely 
cut away ; a layer of tissue five cells in thickness remained to be pene- 
trated by the fixing fluid before it reached the ovule. To what extent the 
fixing fluid was responsible for the denseness of the synaptic mass the 
writer is unable to say, but that the size and form of the chromatin 
masses observable in the contracted ball are due in some measure to the 
reagents is highly probable. The author called attention to precisely 
similar conditions observed in the pollen mother-cell of Lilium ( 1 . c., 
Figs. 19, 20, 21, 22 , and 23). Whether the chromatin thread presented 
by Fig. 5 normally contracts into a mass as dense as that in Fig. 6 cannot 
be stated with certainty, but subsequent events indicate the possibility 
that it may not always do so. The nature of the chromatin thread fol- 
