6o Arber . — On a New Pteridosperm possessing the 
reflexed. On the whole, so far as one may arrive at a provisional con- 
clusion, it seems probable that these seeds were enclosed in cupules, not 
unlike those of Lagenostoma Lomctxi in some respects, but perhaps having 
a closer resemblance to certain other fossils, which may now be reviewed in 
this connexion. 
Since the discovery of the seed of Lyginodendron it has been generally 
agreed that certain organs borne on Palaeozoic fern-like fronds, such as 
Calymmatotheca Stangeri , Stur \ and possibly Zeilleria delicatula (Sternb.) 2 , 
whose nature had hitherto remained obscure, are in all probability persistent 
cupules from which the seeds have fallen. The fertile fronds of the latter 
species are very closely similar to that described here. The cupules are 
even smaller, and, at the same time, appear to have been more deeply cleft 
than in the present instance. But this latter feature no doubt depends to 
some extent on the state of maturity of the frond at the time when 
fossilization took place. 
Again in Calymmatotheca Frenzli , Stur 3 , we have in all probability 
a fertile frond with cupules still in continuity, agreeing fairly closely in 
habit with the present specimen, and with cupules of about the same size, 
cleft at the apex into several lobes, as is apparently the case here. 
The remarkably small size of the seeds may raise the question 
whether these organs are really of that nature. The smallest Lagenostoma 
(L. ovoides , Will.) known in the petrified state has a length of 4-5 mm. In 
the case of L. Sinclairi ) Kidst. MS., a species founded on an impression, 
the length is 4-5-5 mm. 4 Most of the other seeds, suspected of being of 
Pteridospermous affinity, are much larger. On the other hand the seed, 
Gnetopsis elliptica , Ren. and Zeill., has a length of 2 mm. There does not 
appear to me to be any valid objection to the present conclusion, merely 
on the score of size, for a large number of species of the frond genus 
Sphenopteris are known in which the segments of the pinnules are, as in 
the present instance, very small indeed, and one would naturally expect to 
find, in such cases as may prove to belong to the Pteridosperms, some 
correspondence between the size of the seeds and the pinnules which bore 
them. Again these fructifications have all the characters of a seed on 
a small scale (cf. PI. VI, Figs. 2 and 3), and do not in the least resemble any 
known Palaeozoic sporangia or synangia. Further, the possibility, by no 
means slight, that the seeds are invested in cupules, strengthens this con- 
clusion. 
1 Stur, ibid., see also Oliver (’05), p. 412, Text-fig. 6. 
3 Kidston (’84), p. 590, PI. XXV. It seems to me conceivable that the specimen figured by 
Kidston, PI. XXV, Fig. 2, and described (p. 598) as showing ‘ closed globular involucres/ representing 
* the fruit in an early state of development ’ may, judging by the figure, be a case in which the seeds are 
still enclosed in their cupules. The general similarity of the specimen to that figured here is striking. 
3 Stur (’85), p. 268, PI. XXXVIII, Fig. 3, PI. XXXVII, Figs. 2-3, and Text-fig. 42 on p. 239. 
4 Arber (’05 1 ), p. 252. 
