A zo l la filiculoides , Lam. 159 
time each sporocarp-rudiment grows by a three-sided apical 
cell (Fig. 8). Next a slight outgrowth is observed near the 
base of the young sorus, which forms a ring-shaped projection 
around it (Fig. 9 id .) ; this is the beginning of the indusium or 
sporocarp-wall, and corresponds exactly to that of Salvinia 1 . 
From this point the two sorts of sporocarps differ. In the 
macrosporic ones the apical cell forms at once the body of 
the single sporangium ; in the microsporic, it forms a colu- 
mella, from which latter the microsporangia arise as lateral 
outgrowths. My own observations in these earlier stages 
were confined mainly to the female sorus, but there was 
nothing to indicate any difference in the development of the 
male, except in the sporangia themselves. 
There has been more or less conjecture concerning the 
origin of the single macrosporangium, whether it was really 
the only one formed, or whether several were formed at first and 
one crowded out the others. A glance at a young sporocarp 
will show at once that only one sporangium is formed at first, 
and that this is formed directly from the apical cell of the 
sporocarp-rudiment. After a varying number of segments have 
been cut off, a periclinal wall is formed in the apical cell, 
which then proceeds at once to form the body of the sporangium. 
At this stage (Fig. 9) the sporangium has a very short pedicel, 
and the archesporium has the familiar tetrahedral form common 
to the other Leptosporangiatae. The next divisions follow 
closely those of the other Leptosporangiates, and offer nothing 
new. The tapetum (Fig. 10 1 \ is formed as usual, and, coincident 
with its formation, radial walls are formed in the outer cells of 
the sporangium, whose wall, as usual, remains but one cell 
thick. The tapetum also divides only by radial walls, so that 
it too consists of but one layer of cells (Figs. 12-15). These 
cells, as well as the central cell, contain more granular proto- 
plasm than do the cells of the sporangium-wall. 
The divisions in the central cell correspond with those in 
other Leptosporangiates. The first wall is nearly vertical, and 
this is followed by a transverse wall in each of the resulting 
1 Strasburger ; Ueber Azolla, p. 54. 
