168 BOTANICAL GAZETTE [MARCH 
Examples might be multiplied almost indefinitely, but these 
illustrate the general direction of the irregularities. _Monospo- 
rangiate and ambisporangiate flowers in Salix have been described 
before, but I can find no account of microsporangia borne inside 
the ovary, or of orthotropous ovules. 
The more minute anatomy deserves some attention. As a 
rule, the macrospores have a perfectly normal development. 
Most of the material showed the macrosporangia at the fertili- 
zation period, and the egg apparatus and primary endosperm 
nucleus could not be distinguished from those of normal plants, 
and in several cases, as in fig. 88, embryos were developing in 
the usual way. The stamens of monosporangiate flowers, as 
well as those of the ambisporangiate flowers, developed exactly 
like other stamens in every detail which I was able to observe, 
but the microsporangia which were borne within the ovary need 
separate mention. These sporangia were usually solitary, but 
sometimes in pairs, and the wall usually had no layer at all 
between the tapetum and endothecium, the former often being 
abnormally developed, as in fig. 7. It is not at all unusual to 
find cells of the tapetum with two, three, or even four large nuclei, 
as represented in this figure. This preparation also shows’ cells 
of the tapetum which have divided by periclines. The cells of 
the sporogenous tissue are irregular in shape and probably would 
not have developed spores. Another irregular case is shown in 
fig. 6, where the sporogenous cells, probably spore mother cells, 
have surrounded themselves with a thick wall. Instances like 
figs. 3 and 6 are common, where the sporangium development iS 
feeble and seems to have beenchecked. Many of the microspo- 
rangia, however, especially those which are more or less stalked, 
present a more normal development. A characteristic example 
of the microsporangia which continue their development is seen 
in fig. 7, the wall appearing much like that in fig. 8, which is 
drawn from a perfectly normal anther of S. cordata. The pollen : 
grains are somewhat vacuolated (as are the cells of the tapetum), 
and show the division into. fave nucleus and generative nucleus, © 
re which are slightly sm: s lly th case in S. es A 
