the I’oots, all indicate an affinity witli Cycadophyta. Broadly speaking, these results
hold good for the Medulloseae generally, though in the more elaborate Permian
species the differentiation of the steles among themselves, and certain peculiarities
of the secondary growth may render the likeness to Cycadean stems more stiiking.
On these anatomical grounds the Medulloseae were included by Potonié in
1897, in his happily named intermediate grouj) Cycadofilices.
The point on which the (piestion of the systematic position of the family
turned, was clearly that of the nature of the fructification. It had been suggested, some
years ago, on grounds of association and of common structural features, that seeds
of the genus Tr igo no ca r pon might probably prove to belong to species of Al e thopt
e r i s or Ne u r o p t
e r i s (W il d 1900). But
this remained nothing
more than a probable
suggestion, and it was
not till the close of 1903
that we received, from
my friend Mr. K id s t o n ,
tlie first definite proof
that a member of the
Nenropterideae was a
seedbearing plant.
Ml'. K id s t o n showed
that certain large seeds,
I'esembling a hazel-nut in
size and form, from the
Coal-Measures of Dudley
in England, ivere borne
on portions of a rachis
to which the characteristic
pinnules of the w'ell known
species, Ne u ro p t e r i s
Fig. 4. N e u r o p t e r i s h e t e r o p h y l l a . Part of vegetative frond,
slightly enlarged.
h e t e r ophyl l a , were still attached (Figs. 4 and 5). The seeds have a fibrous envelope,
and are placed by their discoverer in the genus Rh a b d o c a rp u s of G o e p p e r t and
B e r g e r (K id s t o n , 1904). Tliey ai'e, however, of the radiospermic type, and must
not be confounded with the platyspermic seeds, which, following Brongniart, we more
usually associate witli the generic name Rhabdocarpus. Mr. K i d s t o n ’s discovery
thus proved that Neu ropt e r i s h e t e r ophy l l a was a seed-bearing plant, and that the
seeds were borne on a frond, differing but little, so far as one can judge by the
leaflets, from the ordinary vegetative foliage. This one case, so clearly established, is
sufficient to afford the strongest presumption that the family Nenropterideae or Medulloseae
as a whole, ivere Spermophyta, a conclusion which has received much indirect
support from Mr. G r a n d ’ E u r y ’s extensive observations on the constant asso-
elation in the Ujtper and Middle Coal-Measures of France, between definite species of
seeds, and definite types of Neurojiteridean fronds. (G r a n d ’ E u r y , 1904). Thus
this observer is led to refer vaiions forms of , Pachy t e s t a (including the largest
fossil seeds, reaching the size of a hen’s egg) to the Stejdianian species of Al e thop
t e r i s , and certain seeds of the Trigonocarpon type to the Westphalian species
of the same frond-genns. Other species of seeds are constantly associated with the
fi'onds of Cal l ip t e r i d i u m . of O d o n t o p t e r i s ,
of Lin op t e r is ( = Dictyopteris) and so on. The
evidence is as yet direct and conclusive only in
the case of Mr. K i d s t o n ’s Neu ropt e r i s , l)ut
considering that in no frond of this group is
there any evidence for a Eilicinean form of fructification,
all indications jioint to the Spermo-
l)liytic nature of the entire family Neiiroptei'ideae,
and thus completely justify the opinion originally
expi'essed by S t u r ').
There are, however, other extensive groiqis
of Palaeozoic Fern-like plants in which the evidence
tor Spermophytic affinities is equally strong.
This ajiplies in the first place to a ])ortion of
the great genus Spheno p te r i s . As long ago
as 1877 Stur described a form of fructification
in the species S. Stangei ' i , from the Lower
Carboniferous beds of Bohemia (S t u r 1877).
Owing to the kindness of the Director of the
Geologische Reichsanstalt of Vienna, I was enabled
to examine the oi'iginal specimen, and am
thus in a position to confirm S t u r ’s interpi'eta-
tion in every particular. There can be no doubt
Fig. 5. Seed of the same plant,
attached to a branch of the rachis hearing
that the fertile pinnae — Calymma t o t h e c a
two characteristic pinnules, x 2 .
S tang e ri , S t u r — belong to the' same ])lant
Figs. 4 and 5. From Journal of
iioyal Microscojiical Society, after
as the vegetative foliage of the S jihenop t e r i s
])hotogra])hs by Mr. Kidston.
with which they are associated. The fructification
is I'epresented by calyx-hke bodies, each with about (5 lobes, terminating the branches
of the fertile rachis. S t u r regarded the lobes as tlie valves of an opened indnsium.
Subsequent writers have intei-preted the lobes as sporangia, Imt examination of the
specimens shows that they are foliaceons in character, and thus confirms the original
interpretation of the discoverer. We are still without direct evidence as to the organs
contained wdthin the cal_yx-like indusia, or cuimles as we may term them; this question.
J) The pinnule of A l e t h o p t e r i s S e r l i figured by AL Ze ii.lkr (Zeii.ler, 1900, p. 96)
as apparently bearing globular sporangia is, I understand, no longer regarded by that author as
affording any evidence of fructification. An examination of tlie specimen convinced me that the
irregularities of its surface had nothing to do with tlie presence of siiorangia