IN NATURAL CLASSIFICATION. 597 
that of the Leguminose, while this organ possesses in those 
genera whose fruitis in an eterio, an organization like that it 
has in the Geraniacee. 
I shall conclude this memoir with the following propo- 
sitions :— 
1. That the pollen is an organ of as great, if not greater, phy- 
siological importance as the em 
2. That the dehiscence, form, opacity, or transparency of the 
exine, and simplicity or compound structure of the pollen, 
are the characters of this organ of greatest taxonomic im- 
portance, in our present condition of knowledge. 
3. That inasmuch as these characters are found by experience 
to be of very inferior value, in most instances, to those 
derived from the embryo ; while it appears from the physio- 
logical utility of the pollen, that characters of at least an 
equal importance, are capable of being drawn from it, it 
follows that this organ must, as yet, have been studied in 
minor points of view. 
4. That inasmuch as each of the characters derivable from 
the pollen are subject to exceptions, the most perfect 
types may be obtained in this as in other organs, by a 
combination of characters rather than by any single one 
.. taken alone. 
5. That the taxonomic value of the pollen varies in different 
groups. 
6. That the species of the same genus have always pollen 
essentially alike. Apparent exceptions to this rule occur 
in Viola. 
7. That the genera of the same family usually have pollen es- 
sentially alike. Exceptions to this rule occur in Ranun- 
culacee, Fumariacee, and among Monopetalous Dicotyledons 
with numerous ovules. 
8. That characters derived from the pollen are inferior to 
those by which particular families are segregated among 
Monopetalous Dicotyledons with numerous ovules. Excep- 
