68 MISSOURI BOTANICAL GARDEN. 
inate catkins then become from one to two inches long, 
generally curved outwards, and their scales spread just 
enough to expose the stamens and allow the very abundant 
and powdery yellow pollen to escape. The soft paren- 
chyma of the axis of inflorescence becomes torn in various 
directions as the catkins elongate, so that when they have 
reached their full development it is loosely fissured through- 
out. The fibrovascular bundles at the same time are poorly 
developed and almost unlignified, so that it is almost impos- 
sible to dissect a catkin without tearing it in pieces. The 
same loose texture exists in the basal part of the catkin 
scales, where they have increased in length during anthesis, 
and as the outer part is considerably longer than the inner, it 
assumes a series of characteristic transverse wrinkles below. 
This separation of the tissues in both axis and bract, gives 
rise to the curious appearance in longitudinal section which 
is shown in plate 32, fig. 4-5, for the cleft in each bract is 
decurrent down the axis to the point where the firmer fibro- 
vascular bundles emerge for the next flower. The staminate 
flowers, so far as I have examined them, are glabrous and 
quite destitute of a perianth or involucre of any description, 
and consist simply of a whorl of about ten short filaments a 
little dilated at base and surmounted by slightly versatile but 
nearly erect extrorse two-celled anthers dehiscing longitu- 
dinally. The pollen grains are nearly globose, smooth, 
slightly 3 to 4-grooved with underlying thickening of the 
intine, and fall from the dehiscent anther very readily, and 
there is no doubt that the species is wind pollinated.* 
The pistillate catkins possess the same loose lacunose 
structure as the staminate, though the axis is far less torn. 
When fully developed they are rarely over half an inch long, 
and it requires some little care to detect their presence on 
* The structure of the staminate flowers, aside from the lacunose 
character of the axis and bract, and the extrorse facing of the anthers, is 
well shown by Baillon, J. c. 240, f. 214; Oliver, J. c. f. 1 to 3; and Heim, 
i.c. pl. 10, f. 1 to 6. Baillon, J. c. 239, mentions bractlets as being some- 
times present. 
4 
