THE VEaETABLE CELL. 135 



brane of the embryo-sac for a short space. In extremely rare 

 cases (in Ganna^ according to Hofineister), the pollen-tube breaks 

 through the membrane of the embryo-sac, and thns comes imme- 

 diately in contact with the ger}ninal vesicles. In the great jnajority 

 of cases, however, as already observed, the pollen-tube is separated 

 from the germinal vesicles by the membrane of the embryo-sac, 

 and frequently even, the point at which the end of the pollen-tube 

 is in contact with the embryo-sac, does not correspond exactly to 

 the point at which a germinal vesicle lies in the inside of the em- 

 bryo-sac (pi. 1, fig. 15). Therefore the only way in which a mate- 

 rial effect can be produced by the pollen-tube upon the germinal 

 vesicle, is by the fluid part of the fovilla transuding through the 

 membranes of the pollen-tube,' the embryo-sac, and the germinal 

 vesicle. It cannot be demonstrated that such a transudation does 

 take place, but it is in the highest degree probable, since it is in- 

 comprehensible how the impregnation of the germinal vesicle could 

 take place without it. 



The pollen-tube begins to decay more or less rapidly after it 

 has reached the embryo-sac. Its growth is arrested, as before 

 noticed, and the fovilla contained in it undergoes a visible change 

 in its characters, acquiring a granular, half coagulated aspect ; the 

 pollen-tube itself is by this time evidently dead, and disappears 

 sooner or later (sometimes, however, not until the seed is ripe), 

 apparently through absorption. 



Shortly after the meeting of the pollen-tube with the embryo- 

 sac, but only when this has occurred, the farther development of 

 the germinal vesicle begins, this exhibiting a rapid growth, and 

 usually displacing the two other germinal vesicles which ordi- 

 narily accompany it (pi. 1, fig. 15); it is only in rare cases that 

 two or more of these vesicles simultaneously undergo enlargement. 

 The form which the growing germinal vesicle assumes is very un- 

 like in different plants ; in many it grows but moderately in 

 the longitudinal direction, and thus becomes ovate; in others, 

 particularly in the Scrophularinese and Cruciferse, it grows into a 

 long cylinder, winch frequently does not much exceed the pollen- 

 tube in diameter, and exhibits a clavate expansion at its lower 

 extremity. During this enlargement, the protoplasm, which ori- 

 ginally filled up the germinal vesicle pretty uniformly, becomes 

 principally collected at the lower end, after which cell-formation 

 by division commences (pL 1, fig. 15, 16). In this conversion of 

 the germinal vesicle into a cellular body, to which Hofmeister 

 applies the name o{ pro-embryo (vorkeim), abundance of modifica- 

 tions present themselves in different plants. In all cases the 

 vesicle first divides by a transverse wall into two cells, one above 

 the other (pi. 1, fig. 16, a, 6.) ; the lower of these may at once be- 

 come converted into a parenchymatous body (the embryo) by suc- 

 cessive subdivisions, as occurs in Monotropa, or, as is ordinariljr 

 the case, the formation of the embryo does not commence until 



