48 MISSOURI BOTANICAL GARDEN. 
number of stamens as usually 3. While at the base of an 
ament there may be sometimes only 3, higher up I find the 
number varying from 4 to 7, in one instance 5 to 12! 
Andersson says the stigmas are entire, and my previous 
observation was confirmatory in part. I now find, if 
examined when young, that the stigmas are, just as in 
nigra, almost always notched. The capsule of Wardi has 
been stated to be ‘* minutely granular’’ (Fl. Washington), 
or ‘*minutely glandular under a lens’’ (Garden and 
Forest). My examinations have failed to confirm these 
statements. In respect to the form of the capsule, the 
variation runs parallel with that of nigra; length of pedicel 
varies much, graduating into that of nigra. 
The distinction into broad-leaved and narrow-leaved 
forms is too indefinite or unstable to be of importance, the 
variation in this respect being coincident with the similar 
variation of nigra. And lastly, the statement made in pre- 
vious descriptions, of the prolonging of the flowering 
branchlet as a peculiarity of Wardi, is not distinctive. This 
character is generally present in the nigra of this region, 
but, owing to inferior vigor, is less developed. 
Now, having passed in review the above mentioned strik- 
ing differences, it seems most remarkable that there is 
almost precise agreement in their venation, a statement 
which would not be true, to my knowledge, respecting any 
other two species of willows. For the mode of venation 
should be regarded as usually one of the most constant and 
distinctive characters of the species of this genus. The 
foregoing observation has more especial reference to the 
extreme minuteness of the reticulation, a unique character- 
istic of the whole nigra group. The only difference noticed 
is the greater tendency of the primaries of nigra to form 
loops and the marginal line.* 
I have found evident hybrids between longipes ard nigra 
in all the localities visited. As might have been expected, 
* See my paper on Venation of Salix. Rept. Mo. Bot. Gard. 5: 52. 
