Nov., 1907.] 
Life History of Cornus Florida. 
201 
pollen. The grains when shed are very much like the two-celled 
pollen grain of April 30th (Fig. 13), except they are more elon¬ 
gated, being sub-spindle shaped. In 1907, although the season 
was somewhat late, the shedding period was about over on 
May 27th. Fertilization was not observed. 
Those flowers which w r ere fertilized, however, could easily be 
distinguished externally by their increase in size over the others. 
Of the fertilized flowers there were usually from one to three 
upon a head. Those receptacles which contained no fertilized 
flowers were soon shed by the formation of cleavage-planes in 
the peduncles. Great numbers of the heads thus cut off could be 
gathered under a single tree. 
ENDOSPERM AND EMBRYO. 
The first material after the last eight-celled embryo-sac 
(May 21st) was collected June 12th. By this time there had 
been a rapid development of the endosperm. This filled the 
upper third of the embryo-sac, extending up to where the svner- 
gids lay. The contorted, double pointed cone of the synergids 
at this time still retains its characteristic shape. The tissue of 
the integument is beginning to break down. 
On June 20th the endosperm had pushed its way in the upper 
end of the sac practically to the apex of the cone. In the lower 
half were a large number of loose endosperm cells. By June 
29th, the endosperm had reached a very characteristic arrange¬ 
ment. The upper part forms a cap-like structure, the cells of 
which are arranged concentrically. The integument is nearly 
destroyed, there being merely strands of disintegrating cell walls. 
In the material of June 12th, 20th and 29th, appeared a short 
chain of irregular cells whose walls were much thicker than those 
of the cells of the surrounding endosperm. Whether or not these 
cells were the young embryo rvas not determined. The first 
material which, without question, contained a young embryo was 
taken July 9th. At this time the embryo is a spherical mass of 
cells suspended by a suspensor from the cap of endosperm cells 
(Fig. 21). The endosperm cells do not as yet completely fill the 
embryo-sac and fragments of the integument cell walls remain. 
The embryo has developed rapidly by July 21st. On this 
date the two cotyledons are present (Fig. 22). Some of the 
cells are becoming differentiated to form the \ r ascular cylinder. 
The endosperm completely fills the sac except for the small 
ca\dty in which the embryo lies. 
The last section was made from material collected July 28th. 
In this the embryo had become quite large (Fig. 23). The 
cotyledon, hypocotvl, root tip, root cap, and embryonic tissues 
were well differentiated. The endosperm, as on July 21st, 
completely filled the sac. 
