Ontogeny. 9 
the plants which have practised such methods, though successful in 
a degree, have not become dominant and replenished the earth. 
The vast majority of Ferns have either a creeping rhizome, 
subterranean or superficial, or a short erect stem (“rootstock”), 
for the most part buried in the soil. Nearly all the members of 
the more primitive families of modern Ferns, such as Hymeno- 
phyllaceae, Gleicheniaceae, as well as others with a more advanced 
type of soral structure, but a relatively primitive vascular system 
(Dennstaedtineac, Lindsayeac) are characterised by a creeping 
rhizome, often with long internodes. It might, perhaps, be thought 
that the protostelic and solenostelic vascular system is an adaptation 
to the creeping rhizome, but this notion is negatived by the fact 
that the creeping stems found in many Davalliaceae and Polypo- 
diaceac are dictyostelic. Also Lachmann (’89) has shown that the 
creeping rhizomes of the primitive types have roots either distributed 
all over the surface of the stem or confined to the lower side, but 
not localised in relation to the leaves. Such a distribution we 
should expect in primitive forms. It is in the higher dictyostelic 
types that the roots are sometimes localised in relation to the leaf- 
insertions, clearly a more specialised arrangement. In all cases, 
however, fresh roots are formed close to the apex as growth in 
length proceeds and fresh leaves are developed, and in all cases there 
is a tendency, frequently very marked indeed, for the earlier formed 
parts of the stem to die off as the new assimilative and absorptive 
organs are developed from the apex. 
Owing to the fact that the fern-plant has a continuous uninter¬ 
rupted development from the egg, and no secondary thickening, the 
successive stages in the development of their adult shoots can 
frequently be traced very fully. The earlier ontogenetic stages, 
instead of being obliterated by the later ones, as in animals and to 
some extent in the flowering plants, remain recorded in the first 
formed portions of the shoot, and are only obliterated by its gradual 
decay. In accordance with the law of recapitulation we should 
expect this ontogenetic record to represent, in a general way, the 
history of the evolution of the race, and as a matter of fact these 
anticipations are strikingly realised by the close correspondence of 
the developmental evidence with that derived from a comparison of 
the adult structure of the different families. 
Let us confine our attention at first to the monocyclic forms. 
We always find a protostelic structure in the first formed stem, and in 
the protostelic types, so far as evidence is available, this structure 
