8 ARKIV FOR BOTANIK, BAND 10. w:o 15. 
suggest a Marattiaceous affinity, which is more conspicuous, at 
any rate, than in the case of Senftenbergia. It is, of course, im- 
possible to decide whether the sporangial wall is one or more 
cells thick nor in what manner the sporangia originated. It is 
worthy of note, however, that the sporangia not only remain 
free from each other but are sometimes even seen to be provided 
with a short stalk (pl. 2, figs 4, 5). The development of the 
broad apical zone of thick-walled cells is another character, 
which is without parallel in the recent Marattiacee. It appears 
therefore that the genus cannot be ranged within the limits of 
that family as defined by its existing representatives. Yet, 
there isevidently a closer relationship to the Marattiacee than 
to any one of the leptosporangiate families. 
The presence of a short stalk seems to point in the direction 
of the Osmundacew, to which family there is, moreover, some 
resemblance in the disposition and shape of the sporangia, and 
in the appearance of the fertile pinnules. But, in Cladotheca, 
the thick-walled cells form a continuous zone round the sporan- 
gium, which shows that there cannot well exist any close rela- 
tionship to the Osmundacee. 
If this zone of thick-walled cells may be at all compared with 
the annulus of the leptosporangiate ferns, it isevidently only 
among the Schizeacee that a somewhat analogous structure is 
found. The points of resemblance in this respect have been 
illustrated above by a comparison with Senftenbergia, which 
would form the connecting link with Klukia and the recent 
members of the family. In Cladotheca undans, there is, however, 
a more pronounced resemblance to the Marattiacew. It may 
be that this species forms like Senftenbergia a transitional stage 
connecting these two families, only coming nearer to the Ma- 
rattiacee than does that genus. 
For the present it seems to be the better plan to refrain 
from placing this plant in any definite family. It might, per- 
haps, be brought under the wide heading of Primofilices, as has 
been proposed for Senftenbergia. It is peculiar, however, that 
a fern which is evidently so little specialized and which cannot 
be included in any of the existing families, occurs in so new a 
formation, relatively, as the Lower Oolite. 
In regard to the characters of the frond, Cladotheca undans 
is very like the fern described as Pecopteris australis and referred 
by RENAULT to T'odea (see above). The resemblance is indeed 
