( 389 ) 
he states that the embryo-sac shows a chalaza-ward exten- 
sion at about the time when the pollen-tube begins its disten- 
sion in the micropyle. He further states that when the ovary 
has reached a length of about fifty millimeters the endosperm 
has begun to show a cellular structure. Garreau” (1849) 
studied Momordica Elaterium and Cucumis sativus and shows 
that in the former case the endosperm had replaced about 
half the tissue of the nucellus while the embryo was still 
only a spherical, undifferentiated mass of cells. Although 
he misunderstood some of the structures which he saw, he 
correctly represented the elongated mass of endosperm of 
Cucumis narrowed at its lower end. In this case the endo- 
sperm has begun to show a cellular structure in its upper por- 
tion while the embryo was still a pyriform mass of cells. Hof- 
meister ** (1849) went further in his study of the endosperm 
of Ecballium (Momordica) Elatertum. He observed that the 
endosperm rapidly displaced the nucellus and that the em- 
bryo in turn displaced the endosperm. He further repre- 
sented the vacuolated granular character of the endosperm 
cells and expresses the opinion that the process of develop- 
ment in the ovule of Acballzum was repeated in those of 
other members of the Cucurbitaceae. 
The facts obtained in this study relative to the behavior of 
the endosperm have been recorded in each special case, and 
a comparison of the results will show that there is some 
degree of uniformity in the growth and development of the 
endosperm in the different species. It first expands into a 
sac-like body in the periphery of which are numerous free 
nuclei formed before the first division of the fertilized egg. 
The coenocytic condition is retained for some time and is 
gradually replaced by a cellular structure, first in the micro- 
pylar region. The expansion of the endosperm is rapid and 
always first in the direction of the chalaza. The most rapid 
expansion takes place usually in the coenocytic condition. 
The tissue of the nucellus is rapidly disorganized wherever 
the endosperm comes into contact with it. There seems to 
be no doubt that the nucellus is digested by ferments secreted 
