598 ON THE USE OF POLLEN 
tions to this occur in Ericacee and Epacridacee, where the 
nature of the pollen points out the affinities. 
9. That the nature of the pollen is a character co-equal with 
the definite number of the ovules, among monopetalous 
dicotyledons. Exceptions, Asclepiadee, Ehretiacee, and 
perhaps Boraginee. 
10. The characters derived from the pollen are supericr to 
the nature of the exsertion, or presence of a perianth in 
Apetale and Achlamydee. Exceptions, Salicinee and Pro- 
teacee. 
11. The characters of the pollen correspond to the regularity 
of the perianth among epigynous monocotyledons. 
12. The characters of the pollen correspond with the exser- 
tion among perigynous monocotyledons. Exception, Alis- 
macee. 
13. The nature of the pollen affords characters of the 
highest importance, superior to exsertion, the presence of 
albumen, the kind of perianth, &c. among polypetalous 
dicotyledons. Exceptions, Ranunculacee, and perhaps 
Mimosee. 
4. The characters drawn from the pollen are superior to the 
number of cotyledons, or the position of the embryo 
as regards the albumen, among grasses and sedges. 
15. Transverse affinities or analogies exist ——— very re- 
mote families, as regards the pollen. Thus, Rosacee, 
Leguminose, Rhamnee, Lythracee, Myrtacee, Crassulacee, 
Saxifragee, Hypericacee, Aisculacee, and Tamaricacee, 
are allied to Solanee, Scrophularinee, and Ehretiacee among 
monopetalous dicotyledons ; and to Salicinee and Proteacee 
among apetalous dicotyledons, by having analogous pollen. 
Again, Alismacee has pollen similar to that most usually 
prevalent among dicotyledons. 
In treating a subject so immense, and hitherto so little inves- 
tigated, as that to which this Memoir is devoted, many false 
conclusions and erroneous observations most probably have 
