158 Campbell . — On the Development of 
position (Fig. 5, III. Z.); and the first division-wall in it divides 
it into two nearly equal cells, which develop into the two lobes 
found in the full-grown leaf. No trace of an apical cell can 
be found even in the youngest leaves, and in this respect, as 
well as in the secondary divisions of the segments of the apical 
cell, Azolla differs from Salvinia , its nearest relative. Each 
leaf-lobe is next divided into an inner small cell, and an outer 
larger one, and the latter then is divided into two equal 
cells by a radial wall. This formation of alternate tangential 
and radial walls is repeated with great regularity in the 
ordinary leaf-lobes, and in their young stages can be traced 
for a long time. 
The sporocarps or sori always arise, as Strasburger showed 1 , 
from the ventral lobe of the first leaf of a branch. He was not 
able to get the earliest stages owing to lack of material, and 
based his conclusions upon a study of the later stages. 
According to his statement they arise from a part only of the 
ventral lobe, the rest giving rise to the enveloping involucre. 
From a careful study of the very young stages I have been led 
to a somewhat different conclusion. I find that the whole of 
the ventral lobe goes to form the sori, and that the involucre 
is derived from the whole of the dorsal lobe of the leaf. The 
latter, instead of developing as in the sterile leaves, remains 
but one cell thick, and forms a sort of hood arching over the 
sori. The latter are always found in pairs in A. filiculoides. 
These are sometimes of the same kind, or one may be male 
and the other female. 
The leaf-lobe which is to develop into the sporocarps is 
distinguishable at an extremely early period. Its first divi- 
sions are like those in the sterile lobes, and like them it is 
divided into two very nearly equal parts. Each half now 
developes at once into a sporocarp. As soon as the first 
median wall is formed, each of the resulting cells becomes the 
initial cell of the future sporocarp. In it walls are formed 
that cut off three segments from its base, and these are 
followed by others following the same order, so that for some 
1 loc. cit. p. 52. 
