XKXH PREFACE. 
that it is difficult to trace its existence. In the 
second modification its existence in the seed is al- 
ways distinct, ex Composita, and Boraginee. 
The last stage in the development and the ‘only 
one dependent upon fecundation, is the production 
of the embryo. This invariably consists in the ap- 
pearance of a cell attached by a slight cellular 
pedicel to the apex of the embryonary sac, it is deve- 
loped by the production of other additional cellules, 
which always contain much active molecular matter 
but less than the albumen. It soon becomes densely 
cellular in dicotyledons, the cotyledonary division is 
soon indicated, but the radicle is necessarily, from what 
. I have just mentioned, the first part formed. With 
regard to the direction of the cotyledons, I have not 
as yet ascertained any law, they are often opposite to 
the true faces of the seed and often alternate. Em- 
bryos are generally axile when albumen is present, in 
those cases in which they are subsequently peripherial, 
this is owing to an equal deposition or formation of 
the albumen. i 
The albumen is of a comparatively early develop- 
ment, arid when it exists is always partly formed before 
ihe embryo begins to appear, I find that with the excep- 
tion of those cases in which the embryo is lodged 
within a vitellus that it it is never formed within the 
nucleus. 
In its earlier stages, it contains myriads of active 
molecules of a regular size, these molecules form in 
fact the milk of the albumen. 
The presence of albumen is the test of an imperfect 
