of Hippuris vulgaris. 261 



From the fact that the embryo-vesicle is developed at so long 

 a period before the bm'sting of the anther, little doubt can remain 

 as to its existing prior to the act of impregnation, and not being, 

 as supposed by Mirbel and Spach, a consequence of that act. 

 Still less can it be supposed to be the end of the pollen-tube, ac- 

 cording to the theory of Schleiden and his followers. 



We now proceed to notice the changes which the embryo- 

 vesicle undergoes subsequently to the act of impregnation. After 

 impregnation, the granular protoplasm, which has accumulated 

 at the larger extremity of the embryo-vesicle, becomes trans- 

 formed into a spheroidal cell. A septum is then observable at 

 the lower part, crossing it horizontally, by which it is divided 

 into two cells. Of these the inferior is developed downwards by 

 successive merismatic division, so as to form a oonfervoid fila- 

 ment, the suspensor. The upper assumes at the same time a 

 spheroidal form, and is distinguished from the rest by being 

 filled with granules, exactly as occurs in the Orchidacete. Soon 

 after it divides by a longitudinal septum, and subsequently by a 

 transverse. These are followed by successive divisions, and the 

 embryo with its suspensor is formed. While these changes are 

 taking place, the embryo-vesicle, which in the early stage is ad- 

 herent by one of its extremities to the micropyle-end of the em- 

 bryo-sac, becomes correspondingly enlarged and elongated. It 

 however never becomes completely filled with the cells of the 

 suspensor, or at least not until a very late period. It seems to 

 be narrowed at its apex, either by the absorption of its contents 

 by the developing embryo, or by the pressure of the contiguous 

 parts. Subsequently the round mass of cells described above, 

 to which the term embryo-globule has been applied, undergoes 

 further development, and the cotyledons and other parts being 

 gradually formed, the embryo assumes its characteristic appear- 

 ance. 



Thus we see in this plant — 1st. That the embryo- vesicle exists 

 at a period previous to the act of impregnation ; 2nd. That after 

 impregnation a number of cells are formed by an endogenous 

 process in its cavity which assume a confervoid arrangement ; 

 3rd. That of these one is selected to be developed into the em- 

 bryo; 4th. That the rest undergo no further development, but seem 

 to conduce to the nutrition of the embryo. These facts are in every 

 respect conformable to what is known of the embryogenic pro- 

 cess in the Oi^chidacecB, OnagracecBy Scrophulariacea, Crucifera, 

 and other natural orders. 



Since the above observations were made, I have had the oppor- 

 tunity of seeing the results of two very important series of re- 

 searches by Hofmeister of Leipzig and Tulasne. These researches 

 lead to the conclusion, that the mode of development above de^ 



