474 BOTANICAL GAZETTE [DECEMBER 
are of uniform size and full of protoplasm. Fig. 2, representing 
a much enlarged view of pollen from the same region, shows clearly 
that there are no defective grains, but that all are equally well 
developed. Fig. 3 illustrates similarly perfectly developed micro- 
spores from the anther of a “fireweed”’ collected near Pocono 
Plateau, Pennsylvania, and the uniform structure of the pollen 
av s oe ae i ria. : 
oe ie hee ye te 
tee Mee a ees, NS a yi é 
y~ eee icy 
ele = ae °” 
is tate =d 7 ‘te Ne ps 
: Fe 
a be 4 i 
Iehewt & ond Xt L 
Le 
i viurrn @ han aa 
astes 2 ae Bas x 
shat : Sie Ee GS oa ae 
BR dn. | a A or Sain aC ce a 
a) Peg in a Py 
: “er + N 
° : f. it b A? 
. Fae. eae rS 
Fic. 1.—Relative distribution of Epilobium angustifolium L. and E. latifolium 
L.; solid black indicates region where the two species are coexistent; hatching indi- 
ae territory where E. angustifolium extends beyond range of E., latifolium. 
from a West Virginia specimen is pictured in fig. 4. The two 
abortive grains among the perfectly formed ones in fig. 4, without 
doubt, have resulted from physiological conditions. In contrast 
to this perfect development of the pollen cells in these more south- 
ern representatives, fig. 5 represents the formation characteristic 
of the more northern regions where it grows within the range of 
E. latifolium. The pollen for this and the following illustration 
was chosen from a plant collected on Mount Albert in Gaspé 
