350 BOTANICAL GAZETTE. 
The essential points in regard to fertilization are as follows: 
The development of the pollen is quite normal. Therea 
pear in each transverse section of the anther from six to eight 
mother-cells, which have somewhat thickened walls but do not 
float in liquid. The tapetal cells are unusually large and the 
three outer rows, particularly the inner, much smaller. 
The stigmatic surfaces of the style are covered with nipple- 
shaped papillz, the uppermost of which develop sooner than the 
ower. Those on any lobe interlock with those on the other two 
and thus prevent the premature separation of the lobes (fig. 4). 
As the growth of the outer surfaces of the style-lobes produces 
an increasing tension, the lobes finally separate. This separa- 
tion begins at the base and the tips hold till the stigma is com- 
pletely mature. 
The hairs which brush out the pollen from the anthers ({ig- 5) 
have their bases, 5, deeply sunk in the tissues of the style. They 
are single-celled. Their tips, ¢, become quite thick and exten- 
sively cutinized, and the inner thickening layer is highly refrac- 
tive. Just at the point of emergence the wall is thinner than at 
any other place. After these hairs, which are at first rigid, have 
served their purpose, they become softened and are in some way— 
I suppose by the contact of insects with the style—telescoped’, 
so that the tip is left slightly projecting at the entrance of the 
pit thus formed (jig. 6). The result is as though one grasped the 
tip of a glove-finger and thrust it in upon itself until only the 
tip appeared projecting from the hand of the glove. This intro- 
version of the hairs frees the pollen so that it is readily brushed 
off and carried away. Sometimes a pollen-spore drops into one 
of the pits and remains as a stopper for it. In the young hairs 
a beautiful streaming of the protoplasm is easily demonstrated’. 
The style has a small canal in the center, around which the 
conducting tissue is disposed in a variable number of strands of 
different sizes (fig. 7). Midway between the canal and the pe- 
riphery are (usually) six fibro-vascular bundles. The conducting 
tissue, et, is quite small in amount and sharply distinguished from 
the remaining tissue of the style (fig. 8). 
The ovules are small, flattened and anatropous, with a vey 
short funiculus, so that the micropyle is brought close against the 
placenta. The embryo-sac is straight and lies in the axis 0 the 
ovule (fig. 9). It is sharp-pointed at its micropylar end, some- 
what larger in its upper fourth than elsewhere, and ends at the 
3 Miller, Fertilizati " ula the haits 
ent 1 : = true for this spetion nor Elrapuneuoides, te Btrasburger, Bot. Pract 
: Tger’s account of similar hairs in C. rapunculoides, 1. ¢. 
“sn 
