British Coal Measures. 91 
the specimen figured by M. 1 ’Abbe Carpentier 1 as Carpolithes ? samar oides\ 
Carp., from the Westphalian of the Nord de France coalfield. The seeds 
agree in size, though the British specimens are somewhat broader as a rule. 
M. Carpentier regards his specimens as winged, which is certainly not the 
case in the English examples, and he does not mention the foramen, which, 
however, appears to be clearly seen at the broader end of his seed. In his 
figured example, which only shows one side of the seed, the median groove 
is much broader than in the specimens figured here. Judging chiefly by 
his figure, I am, however, inclined for the present to include the British 
examples in the French species. 
The present specimens differ from other seeds in certain important 
respects, chiefly in their unsymmetrical form. The foramen is, however, 
a feature met with in certain species of R adiospermum and Schizospermum y 
where, however, it is terminal. The fossils are solid bodies, and are certainly 
not foliar in nature. Neither do they appear to be sporangial. They may 
thus be regarded, at least provisionally, as seminal structures. 
Cordaicarpus , Geinitz. 
Next we have the genus Cordaicarpus , Geinitz, 1862, which in common 
with Geinitz, 2 Grand’ Eury, 3 and Zeiller 4 I am convinced belongs to a mem- 
ber of the Cordaitales . There are several species of impressions, of which 
Cordaicarpus Cordai , Gein. (PI. VII, Fig. 29), may be regarded as the central 
type. These are fairly small seeds, oval or orbicular in form, with a smooth 
testa. Two other well-marked species occur in Britain, C. areolatus (Boul.) 
(Text-fig. 7) and C. ovoideus , Berg. (PI. VII, Fig. 30). These are undoubtedly 
Platysperms. They commonly occur compressed quite flat, in which state 
there is usually a narrow rim bounding the seed (PI. VII, Fig. 29). They 
are, however, frequently also found in an almost unflattened condition, and 
are then elliptical in section (see PL VII, Pig. 30), resembling Diplotesta , which 
Bertrand 5 has shown to be a Cordaitean seed. 
Megafospermum, gen. nova. 
To this genus I ascribe a certain very large seed, without any indication 
of a wing or a nucule, which has been described by Dr. Kidston as Carpo - 
lithus Wildii , the testa of which is ornamented by numerous fine and close 
thread-like striae (PL VII, Fig. 28). There may be some doubt at present 
whether this seed is Radiospermic or Platyspermic, for only a single 
example is known from Britain. Others, believed to be generically similar, 
1 Carpentier (’ll), p. 7, PI. XIV, Fig. 3 ; Ibid. (’10), p. 599. 
2 Geinitz, H. B. (’61), p. 150. 3 Grand’ Eury (’77), p. 233, Pis. XXV, XXVI, Fig. 9. 
4 Zeiller (’86), pp. 637, 641. 6 Bertrand (’ll). 
