380 WoopgsurRN: DEVELOPMENT OF THE EMBRYO SAC 
nated. The fruits of the latter, however, continue to develop 
and the integuments of the ovule often enlarge considerably into 
flat thin seedlike bodies. The structures in the ovules are very 
hard to procure in proper sections for staining, due to the very 
hard nature of the integuments. Therefore great, care was exer- 
cised in arriving at conclusions in regard to the presence or absence 
and the nature of certain structures in the embryo sac. 
The sections for microscopical studies were all made from 
ovaries that would evidently have formed seedless fruits, as no 
indications of pollen tubes were evident, and preparations were 
examined which included the earlier stages in the development 
of the embryo sac as well as later ones showing the parts nearly or 
quite degenerate. The considerations then in the following paper 
_ must be taken to refer to ‘‘seedless persimmons,” as none of the 
material on hand contained evidence of embryonal development, 
and later observations showed that the trees from which the mater- 
ial was taken bore a very high percentage of seedless fruits. 
THE DEVELOPMENT OF THE EMBRYO SAC 
Only a small nucellus is formed, PLATE 16, FIG. 13 represent- 
ing its greatest development but showing no indications of a spore 
mother-cell or spores. The earliest stage observed in the develop- 
ment of the embryo sac is shown in FIG. 1. Only two nuclei are 
present and these are near the chalazal end. Whether this con- 
dition represents a normal stage or an incomplete development I am 
unable to say, for many preparations of incompletely developed 
embryo sacs were found. The growth of the embryo sac takes 
place well toward the micropylar end of the nucellus, all of this 
portion eventually breaking down, leaving a small but quite 
persistent remnant at the opposite extremity. Fic. 2 shows a 
condition frequently met with. The egg apparatus seems quite 
normally developed and the polar nuclei are closely associated, 
The two latter have not been observed in an actual state of fusion 
although some preparations show the contiguous sides to be 
somewhat flattened (F1G. 4). Other sections of the same ovule 
_ (FIG. 2) show one or two antipodal cells, but frequently these are 
not of definite form or constant in number. Fic. 3 shows three 
antipodal cells all present in the same section. Only two cells 
