
378 BOTANICAL GAZETTE [NOVEMBER 
the micropylar end of a four-nucleate sac are shown. The entire 
figures being found in two adjacent sections, both are drawn 
separately, as the number of chromosomes is so large that it is 
difficult to reconstruct the two figures and make one accurate 
drawing. In one nucleus the spindle is almost parallel with the 
section. In this spindle at one end 11 chromosomes appear, and 
the same end in the next section shows 1 3, giving 24 chromosomes 
at one end of this spindle. It is not possible to make an accurate 
count at the other end, but there seems to be about the same num- - 
ber. The other nucleus is almost perpendicular to the section and 
at one end the 24 chromosomes are unusually distinct. A meta- 
phase from the chalazal end of another four-nucleate sac is shown 
in fig. 7. It also shows the 24 chromosomes. A similar one might 
be given from the micropylar end of another sac. A somatic cell 
taken from the walls of the ovule is given for comparison (fig. 8); 
it shows 24 chromosomes. Fig. 9 is a telophase of this same 
division from the micropylar end of the sac. The old spindle of the 
previous division is still quite distinct, but less so than the spindle 
of this division. This spindle shows wall beginnings in the one 
parallel with the sections. Only one end of the other spindle 
appears in this section, as it is perpendicular to it. But t 
nucleus, which already has the nuclear membrane developed about 
it, shows approximately 24 chromosomes. : 
The 8 nuclei arrange themselves in the usual fashion, ees 
apparatus, antipodals, and two polars that fuse, each group being 
very distinct and almost diagrammatically organized. The filiform 
apparatus is well developed, but not so strikingly as in Parnassta 
(20a). The indentations so noticeable in the Parnassia synergids 
are not often seen at all in Atamosco (fig. 10). It was present in a 
few instances, but was never a deep notch. The only noticeable 
feature of the sac is the large nuclei containing little other stainable 
material than the large nucleolus. There are often two nucleoli 
the egg and occasionally in the synergid nuclei also. They both 
stain red in the triple stain, or if stained in the iron hematoxylon, 
they are apparently alike. The antipodal end of the sac has the 
usual organization (fig. 11). The three antipodal cells are definitely 
organized, but are without cell walls. The polars are in contact nm 

