HETEROSPORY AND THE SEED-HABIT 705 
type of the sporangia of other Lycopodiales, falls away, but the details 
of fertilisation and of embryogeny which follow are still unknown.! The 
nearest approach to a seed-like habit among the living Lycopods is seen 
in Selaginella apus and rupestris.2 Here fertilisation occurs while the spores 
are unshed, and the sporangia still attached to the strobilus: in S. rupestris 
the connection is maintained with the parent plant until the embryo has 
produced cotyledons and a root. Thus the Lycopodiales, both fossil and 
modern, show approaches to a seed-habit, though it is doubtful whether 
that habit was ever firmly established among them, or persists in the 
form of any of the Seed-Plants of the present day.2 The condition now 
so fully demonstrated for the Pteridosperms shows that a Seed-Habit was 
definitely acquired along another quite distinct phyletic linet These 
large-leaved types, bearing their large seeds of Cycad-like character dis- 
tributed on fronds effective also for assimilating purposes, probably sprang 
from the same stock as the Ferns, and it is especially with the Botryo- 
pterideae and the Osmundaceae that they show the nearest analogies. 
Thus the Seed-Habit appears to have been initiated certainly in two 
distinct phyla, and it is not improbable that it may have been repeatedly 
initiated within either or both of them. 
The establishment of a Seed-Habit does not necessarily bring immediate 
reduction of the supporting system in its train: but it has frequently 
happened that such reduction follows. The fact that the large seeds of 
Neuropteris heterophylla are borne on a rachis bearing characteristic 
vegetative pinnae shows that a correlative reduction is not obligatory. 
But on the other hand, a reduced state of the sporophylls does usually 
accompany the seed-habit: in Lyginodendron the female fructification is 
described as being borne on the rachis of fertile fronds which differed 
from the sterile foliage in the reduced leaf-area: and this applies also in 
some degree to the male sporophylls as well. From such minor degree 
of reduction of the megasporophyll to that condition seen in Cycas is 
no great step, and from this the sequence through the Cycads gives 
very convincing evidence of further reduction.® It seems not im- 
probable that in Cycadeoidea a still further step in reduction has been 
taken, so that while many of the sporophylls appear as minute sterile 
scales, those which are fertile exist merely as radio-symmetric pedicels, 
each bearing a single terminal ovule.© The microsporophylls show a 
series of reductions in less prominent degree, but without any strict 
parallelism with the megasporophylls: thus in Cycadeoidea where the 
1See Scott, Progressus Ret Bot., i., p. 171. 
2Miss F. Lyon, Bot. Gaz., vol. xxxii., pp. 182-3. 
3See Seward and Ford, ‘‘The Araucariaceae, Recent and Extinct,” Pd. Trans., 
Series B, vol. 198, p. 305, etc. 
4See Scott, Progressus Rei Bot., i., pp. 190-212, where the literature is quoted. 
> Engler and Prantl, Wat, Pflanzen., II. i., Fig. 7. 
6 This is the opinion of Wieland, American Fossil Cycads, p. 230, etc. 
2Y 
