A zollci filiculoides , Lam . 183 
bipartition, Salvinia offers a strong contrast to Azolla , as of 
course it does in the absence of roots. This latter is, however, 
probably of secondary importance, as the roots are replaced 
by the peculiar submerged leaves. 
The origin and development of the sporocarp, as well as the 
structure of the spores, is too similar to be accounted for 
except by the assumption of a relationship. The massulae 
and the episporic appendages of the macrospore differ only in 
degree from those of Salvinia , where all the micros’pores are 
held together in one mass, and where the epispore of the 
macrospore is less developed. As indicated in the account of 
the macrosporangium of Azolla , there is every reason to believe 
that it is derived from some form with more than one macro- 
sporangium in the sorus. In regard to the embryo, that of 
Azolla in its younger stages corresponds very closely to those 
of Salvinia 1 . Indeed, so close is the correspondence that one 
is inclined to regard the embryo of the former as not really 
rootless, as the very first divisions in what corresponds to the 
root-quadrant of Azolla are very like. A further investigation 
of this point, by careful microtome-sections, is very much to be 
desired. 
The prothallium and sexual organs of Azolla are also 
extremely like those of Salvinia , the most noticeable differ- 
ence being the much greater size and regular meristern in the 
female prothallium of the latter. Whether the endosperm- 
nuclei are present in Salvinia remains to be seen. 
Compared with the Marsiliaceae there is much less resem- 
blance. The sporocarps are very different, being in the 
Salviniaceae single sori, while in the Marsiliaceae, each fruit is 
a whole leaf-segment with several sori. The prothallium is 
much less reduced, especially the female one, and the arche- 
gonia more nearly resemble those of the homosporous Ferns, 
both in the length of the neck and the division of the nucleus 
of the neck-canal-cell. A very noticeable difference, and one 
which must be regarded probably as an important one, is the 
presence of the endosperm-nuclei. 
1 Sadebeck in Schenk’s ‘ Handbucb,’ vol. i. p. 217. 
