842 REPRODUCTION OF THE ANGIOSPERMIA. 
elongate by growth at its apex, and passes down through the 
conducting tissue of the style (figs. 577, tp, and 1169, ¢) when 
this exists, or directly into the ovary when the style is absent. 
This growth of the pollen-tube is occasioned by the nourishing 
influence of the viscid secretion which it meets with in its passage 
through the stigma and conducting tissue of the style. 
These tubes vary in length, but are frequently many inches, 
and are extremely thin. ‘They are commonly unicellular, and 
have therefore but one cavity ; but, according to Martin Duncan, 
in Tigridia, and all other monocotyledons with long styles which 
he examined, they are composed of several elongated cells, and 
hence have as many cavities as cells (see page 264). The time 
required for the development of these tubes also varies in dif- 
ferent pollen-cells ; thus, sometimes they are developed almost 
immediately the pollen comes into contact with the stigma ; while. 
in other cases many hours are required for the purpose. The 
pollen-tubes also occupy a varying time in traversing the 
canal of the style—that is, from a few hours to some weeks or 
even months. When the pollen-tubes have penetrated the 
stigmatic tissue, the secretion of the latter ceases, and the stigma 
dries up. The upper part of the pollen-tubes also withers above, 
as growth takes place below. 
The pollen-tubes having reached the ovary are distributed to 
the placenta or placentas, and then come ultimately in contact 
with the ovule or ovules. One (or sometimes two) of these 
pollen-tubes enters into the micropyle of each of the ovules ( figs. 
1169, ec, 1170, ¢, and 1171, t), and thus reaches the nucellus and 
embryo-sac. When it arrives at the latter it is generally some- 
what enlarged (fig. 1171, t), and adheres firmly to it at or near 
its apex. The embryo-sac is frequently introverted to a slight 
extent at the point of contact with the pollen-tube (jigs. 1170 
and 1171), and it is stated by Hofmeister to perforate it in Canna: 
but if such a perforation occurs in this case, it is altogether an 
exception to what is generally observed. As soon as the contact 
of the pollen-tube with the embryo-sac is effected, a kind of 
osmotic action between the contents of the two takes place, the 
result of which is the development of one (or rarely two, as in 
Orchis and Citrus) or more of the germinal vesicles into embryos. 
The germinal vesicle (oosphere), in its development into an 
embryo, becomes surrounded by a membrane, and is then the 
oosperm. This generally divides at first in a transverse manner 
into two cells (fig. 1171, e); and then by further division forms 
the pro-embryo or suspensor (fig. 1170, s). The apical cell assumes 
commonly a somewhat globular form (fig. 1170, 7), and ulti- 
mately by cell-division forms the embryo, whether mono- or di- 
cotyledonous. The suspensor is not present in all cases, while in 
others, where it is found, it varies in length. It is evidently not 
essential in all instances, as it always shrivels up during the de- 
velopment of the apical cell into the embryo. The latter, there- 
