254 MR. A. C. SEWARD AND MISS S. O. FORD ON 
Several Tertiary plants have been referred to the Osmundace:se, partly on evidence 
afforded by anatomical structure and partly on the grounds of resemblance in the form 
of the fronds. A well-preserved stem was described by Carruthers* in 1870 as 
Osmundites Dowkeri; the specimen, which was probably from Lower Eocene beds, was 
found on the shore of Herne Bay. The structure of the stem and leaf appears to be 
very similar to that of recent Osmundacere, but we have not enough data before us to 
enter into a more minute anatomical comparison. Carruthers describes the adventitious 
roots as arising from the interpetiolar spaces, and not from the base of each petiole. 
It is clear that the fossil stem is that of an Osmundaceous fern constructed on the same 
general plan as the existing forms, Another Tertiary stem very like Osmundites 
Dowkeri was described by Unger in 18547 from a Tertiary locality in Hungary as 
Osmundites schemnicensis, this generic name being substituted for Asterochlena, which 
had been applied to the same plant by Pettko 7. 
A petrified Osmundaceous stem in the British Museum (Botanieal Department) from 
New Zealand exhibits the same type of structure as that of Todea; the stele contains 8-9 
xylem groups, some of which are small and oval, while others form long sinuous bands. 
Unfortunately the geological horizon of the specimen is not recorded. 
The fronds of Tertiary age referred to the Osmundacee need not be dealt with, 
as they are few in number and usually founded on sterile and imperfect specimens. 
Osmunda regalis is recorded from the Norwich forest-bed ||, and its frequent occurrence 
in newer submerged forests affords an indication of the wide distribution of this species 
in Europe in post-Tertiary days. 
THEORETICAL CONSIDERATIONS AND CoNCLUSIONS. 
Reference has already been made to the divergent interpretations of the Osmundaceous 
type of stem-anatomy. In instituting a comparison of the structure of filmy Ferns with 
such species as Todea barbara and the representatives of Osmunda, it must be 
remembered that we are comparing plants which have acquired different characters 
in response to dissimilar external conditions—a fact -that increases the difficulty of 
discriminating between features of primary phylogenetic importance and such as may 
be described as merely the expression of adaptation to environment. It is indeed 
impossible to draw a definite line of separation between characters that mark evolutionary 
progress, and structural peculiarities that reflect the plastieity of a plant in its response 
to the stimulus of external conditions. 
In Osmunda Claytoniana T the stem is characterized by the large number of xylem 
strands—in well-nourished plants there are usually about forty; in O. cinnamomea the 
medullary rays, and therefore the xylem strands, are much fewer; O. regalis possesses 
an intermediate number of strands—usually about fifteen. In Zodea barbara the xylem 
groups do not exceed eight, at least in the stems we have examined; in T. hymeno- 
phylloides the number is approximately the same; while in T. superba the xylem 
* Carruthers (1370). T Unger (1854). t Pettko (1850). $ See Gardner (1879), p. 49, pl. 4. 
k Caruthers (1870) ; Reid (1899), PP. 91, 96; Solms-Laubach (1891), p. 172, *| Faull (1901) 
iid VEN 
