THE EMBRYO-SAO IN ANQIOSPEBMS. - 529 



fore part of the nucellus, i, e. the micropyle end, are becoming 

 loose and deliquescent, and at a later stage will be absorbed into 

 the embryo-sac. 



These cells are by no means so numerous as in Butomus; but 

 their relation to the axial row and periphery is similar: they 

 radiate outwards, especially in young ovules. 



The earlier divisions have occurred ; a large vacuole separates 

 two masses of protoplasm ; and the mass at the micropyle end has 

 formed four nuclei, one of which is commencing to wander into 

 the protoplasm leading to the other end ; the other three are be- 

 coming arranged into an " egg-apparatus.' ' At the chalazal end 

 of the nucellus are the " antipodal" cells commencing to form. 

 The first bipartition is complete ; and whether or not a second 

 occurs, as in Butomus, my figures do not show. 



That the two nuclei (one from above and one from below) wander 

 towards one another and fuse to form a central nucleus of the 

 embryo-sac, as in Butomus and others, is to be fairly inferred 

 from Plate XX. figs. 2 and 3, in both of which the egg-apparatus 

 also is seen in the micropylar end of the sac. In fig. 2 i3 also a 

 pollen-tube just entering the micropyle, and at the chalazal end a 

 cloudy mass of protoplasm appears to represent " antipodal cells." 

 The unsatisfactory nature of figs. 2 and 3 is due to the use 

 of alkali ; for though only the merest trace has been used to clear 

 the preparation (the cells of aquatics become tinged of a deep 

 brown in alcohol) ; it renders the outlines indistinct, and causes 

 these delicate nuclei to swell up. On the whole, however, I 

 regard the points referred to as satisfactorily established by the 

 sections. 



JEmlryo-sac $c. of Anemone japonica. (Plate XXI.) 



In the younger buds, before the peduncle begins to elongate, 

 longitudinal sections of the globoid receptacle &c. show the ovules 

 as papillae on the ventral base common to carpel and receptacle 

 (Plate XXI. fig. 1). As the carpel grows it carries up this base 

 (Plate XXI. fig. 2), and produces several more ovules, till we get 

 a series of three superposed. Only the lower one appears to at- 

 tain any prominence, however; and all my remarks apply to it 

 alone. 



At first this papilla consists of an axial row with two layers 

 around it (fig. 1), each one cell deep, just as in Butomus ; the 



2q2 



