198 BOTANICAL GAZETTE [MARCH 
had become irregularly thickened, so as to have a pitted appear- 
ance in surface view. The outer layers of cells of this integument 
have all collapsed, as is best seen in fig. 20. In the outer integu- 
ment the development had been different, for while here too the 
outermost layer of cells had unthickened walls, all the remaining 
cells had the pitted wall (fig. 20), giving in section a rather con- 
spicuous seed coat. It is interesting to note that the outer integu- 
ment, which in the early stages of the development of the ovule 
seemed excessively developed (fig. 7), here extended but a very 
little distance beyond the tip of the nucellus (fig. 12). 
Discussion 
In handling a form about whose position taxonomically there 
have been so many differences of opinion, it seems worth while to 
attempt to summarize the forms which are similar to Leitneria in 
various particulars, in order to see whether it could be placed on 
the basis of its morphology. 
In the case of the stamen, there is no particular in which it differs 
strikingly from the stamen of other catkin-bearing forms. The 
microsporangia pass the winter in the spore mother cell stage, as 
do those of Salix glaucophylla (2), Alnus glutinosa (3), Corylus 
americana (3), and Ulmus americana (4). That such a character 
- should be given little weight taxonomically, however, becomes 
evident when one looks at such a group as Hamamelidaceae. 
SHOEMAKER (5) in his study of this family reports all variations 
in the stage in which the stamens of the different genera of spring- 
flowering forms pass the winter. Of Liquidambar Styraciflua he 
says ‘“‘stamens are only small protuberances which do not show 
any archesporium”’; of Fothergilla Gardeni, ‘they pass the winter 
in the pollen mother cell stage”’; while of Hamamelis arborea and 
Corylopsis pauciflora, the stamens “pass the winter containing 
nearly mature pollen grains with two free nuclei.” 
The developing megasporangium containing a single archespo- 
rial cell differs from most of the Amentiferae yet reported upon. 
However, this condition is found in Betula alba (6) and in Alnus 
glutinosa (6), and is usual in various species of Salix (2), and there- 
fore the character of a multicellular archesporium could hardly be 
