LEAF IN THE VASCULAR CRYPTOGAMS AND GYMNOSPERMS. 
583 
phyllopodium terminates abruptly in a blunt spine. This I have never observed 
among the Leptosporangiate Ferns : there it grows on, and becomes attenuated and 
flattened, thus apparently returning to its more primitive condition. All these 
characters, together with the almost entire absence of winged developments upon it, 
point to the conclusion that the phyllopodium in Angiopteris has assumed a more 
independent position than in other Ferns, and appears rather as a structure fitted 
to bear flattened assimilating organs, than as a flattened assimilating organ itself: we 
may compare this gradual self-assertion of the phyllopodium as a supporting organ, 
with what has no doubt taken place in the differentiation of axis and leaf, and say 
that the phyllopodium in Angioyteris betrays its axial nature more clearly than is the 
case in the less highly organised Ferns. 
Cycadace^e. 
Cycas Seemanni (Adolph Braun). 
a . The Cotyledons. 
Observations on the development of the leaf in the Cycadacece were begun on seeds 
and seedlings of Cycas Seemanni from Fiji : the seeds showed on dissection a bulky 
embryo, embedded in a massive endosperm (Plate 38, fig. 20). The embryo itself 
before germination consists of two long, almost equal cotyledons, inserted on a very 
short axis, which bears at its apex a number of young leaves in various stages of 
development. The cotyledons are not exactly alike ; one is often larger than the 
other, while the margins of one usually overlap those of the other (Plate 38, fig. 21). 
It is not always the larger cotyledon which overlaps the other, though it may be 
presumed that it is the older one. Externally no trace of pinnae has been found at 
the apices of the cotyledons, such as is shown in Schacht’s drawing of Zamia spiralis 
(Sachs’s Text Book, 2nd Engl. Ed., p. 501). 
Transverse sections of the cotyledons show that there is sometimes a median 
bundle; in other cases no single bundle occupies a median position, but two equal 
bundles are disposed symmetrically near the centre of the cross section : intermediate 
modes of arrangement may be found between these two extremes. It might be 
assumed that the median bundle when present is merely the result of the coalescence 
of two equal bundles, which might be found to be distinct in the upper part of the 
cotyledon; but this is not the case, since the median bundle has been found to 
maintain its individuality in an upward direction. There is thus some irregularity in 
the arrangement of the vascular system in the cotyledons of Cycas Seemanni, though 
in the later formed leaves of this plant there is, as in the other Cycadacece, a 
constant absence of a single median bundle.* The development of the cotyledons has 
not been traced, since all the seeds at my disposal had mature embryos. 
* Compare the description by De Baby (Vergl. Anat. p. 246) of the vascular system in the coty¬ 
ledons of Phaseolus, &c.; also my own observations and those of Stkasburgek on vascular bundles in the 
leaves of different species of Gnetum (infra, p. 599). 
