1895. ] Embryology of the Ranunculaceae. 299 
advanced for that purpose. The youngest stage observed 
showed three cells derived from the initial cell, the lower one 
becoming the mother cell. 
The mature embryo-sac of Hepatica is especially interest- 
ing. Before fertilization both egg-apparatus and antipodal 
cells, as well as the cavity of the sac itself, become unusually 
large. At the upper ends of the synergide there appears a 
number of protoplasmic strands converging to a point and thus 
forming a star-shaped cap (fig. 55). The egg-cell is broader 
at its upper end and rather deeply inserted. All three cells 
of the egg-apparatus are surrounded by very delicate mem- 
branes. The antipodal cells increase greatly in size during 
the subsequent growth of the embryo-sac, and at the time of 
fertilization each cell may contain as many as ten or a dozen 
nuclei (fig. 56), due to fragmentation. The process of frag- 
mentation may be readily observed here (fig. 57). Each 
nucleus contains two or more nuclei situated in spherical or 
oval vacuoles. 
Soon after fertilization, or when the embryo has reached the 
stage of development shown in fig. 58, the antipodal cells 
bear evidence of dissolution, the fusion of all the nuclei into 
a common mass being the preparatory ste ; 
On account of the large nuclei, the cells of the developing 
embryo and endosperm of Hepatica afford favorable objects 
for the study of karyokinesis. ‘ 
I may add here that by staining on the slide with aniline 
safranine and picric nigrosine I was able to see the centro- 
spheres or what seemed to be centrospheres (fig. 69). The 
centrosomes appeared as extremely small dark points sur- 
rounded by a colorless court, and the radiating kinoplasmic 
zone was also darker but difficult to see. 
N some instances I have observed what were taken to be 
centrospheres at the poles of nuclear spindles (fig. 4, upper 
end of spindle) in preparations that were not subjected to the 
Special process said to be necessary to demonstrate these 
Structures, 3 
Summary.—This work has been interesting especially as a 
Study in variation. That which first of all attracts our atten- 
tion is found in the presence of more than one initial cell 
and their development into normal embryo-sacs. In the 
genera under consideration the number of initial cells reaches 
its culmination in Caltha where five or more may be present 
