PREFACE. xxxi 
additional distinct cellules from the surface of the 
original simple one, To the existence of the first 
modification, I find the presence of albumen invaria- 
bly attached, so much so that in describing ovula, T 
would say with confidence whether the seeds would be 
albuminous or not. Even in those cases in which no 
opportunities have occurred of examining it aborgine, 
inspection of the margin of the sac will always settle 
the point, even although the whole of its inner surface 
may be occupied by cells, the rudiments of the future 
albumen. 
The only instances in which albumen exists with- 
out this membranous sac is in those few cases in 
which that substance is deposited in the body of 
the nucleus, and the embryo is developed within a 
distinct sac (vitellus) which is always of the exal- 
bumenary or membrano-cellular type. 
In the exalbumenary form which in my descrip- 
tions. I term membrano-cellular in contradistinction 
to membranous, by which I characterize the albu- 
minary form, the surface is even at a very early 
period distinctly cellular. This form- may be seen 
to perfection in Composite. 
Whatever its nature may be, as it bicis has 
important relations. with the embryo, and as this is 
invariably formed within it, I propose calling it the 
embryonary sac, 
Another important feature of this sac is its indes- 
tructibility, it always forms portion of the seed, al- 
though in the albumenary form, it occasionally be- 
comes so identified with the albumen of the periphery, 
