OF THE OYULE OF CHRISTISONIA. 581 
usually, however, be distinguished from a synergid by the fact 
that its nucleus lay always at the extreme anterior tip of the cell. 
As the young embryo thus grows out the synergidæ become 
eonsiderably deranged and eventually disintegrated (Pl. XXI. 
figs. 19, 20 a, 205, 21, 22). The process of fertilization of the 
ovum was not observed; and the only cases where the presence 
of a pollen-tube could be at all surmised are those given above. 
Figs. 19 and 205 afford instances where the development of the 
embryo may begin before fusion of the polar nuclei has taken place. 
As a result of the division of the definitive nucleus, the narrow 
embryo-sac becomes filled with a row of four elongated cells. 
Koch, in his description of the initial formation of the endosperm 
in Orobanche, states that at first several nuclei are formed, 
between which, after they are evenly distributed in the sac, cell- 
walls are intercalated. This will also probably be the course of 
events in Christisonia, although in this plant I have been unable 
to follow it. Subsequently, by longitudinal and transverse divi- 
sions in the cells, the embryo-sac becomes very much enlarged 
and swollen out, a large endosperm-tissue being produced. At 
each end, however, the sac remains as a narrow prolongation 
which is devoid of endosperm-tissue, enclosed by the layers of 
the nucellus. 
The further development of the young embryo was not 
clearly observed until the endosperm had undergone several 
divisions. It was as yet undivided, though greatly elongated, 
with an aggregation of dense protoplasm, containing a nucleus, 
at its anterior extremity (fig. 23). The subsequent development 
would seem to be exactly as Koch has described it, although I 
have not in these plants been able, as he was in Orobanche, to 
follow all the successive divisions which go to form the mature 
embryo. The reason for this is that, in the first place, the divi- 
sions which take place appear to be more irregular than they are 
in the latter plant; and, secondly, that in Orobanche the whole 
structure is muchclearer and better defined, owing to the absence 
of the purple coloration which in Christisonia is such a bane to 
the observer. 
The next stage in the development of the embryo is that in 
which a transverse wall divides it into two cells, an anterior 
swollen cell and the suspensor (fig. 24). Soon after the latter 
Joses its protoplasmic contents and becomes henceforth colourless. 
