102 
F. W. Oliver. 
the other specimen (1487, not figured) a considerable strip of 
parenchyma remains in position all along one side of the seed and 
at places it is still in continuity with the stone. 
So far then as these fragmentary data go they tend to shew 
the essential agreement of the two genera of seeds described with 
Stephanospermum and the Trigonocarpons of the English Coal- 
Measures. The large, broad pollen-chamber, the nucellar tracheides, 
and the differentiated testa appear to have been group-characters 
shared by many of the Radiospermic seeds that were so abundant 
in Permo-carboniferous times. The most difficult point concerns 
the relations of nucellus to integument, whether the former stands 
up freely from the latter, or whether these regions are largely 
coalescent as in living Cycads. The indications on the whole 
appear to favour the former interpretation, but for quite conclusive 
evidence we must await the discovery of specimens in which this 
delicate portion of the seed is more perfectly preserved. 
However, leaving this last point open, there are quite adequate 
data for distinguishing between these forms, which may be provi¬ 
sionally termed the Stephanospermeae, and the Lagenostomas, 
which doubtless belong to the Lyginodendreae. 1 
As yet no seed of the Stephanospermeae group has been cer¬ 
tainly referred to the plant which bore it, though Mr. R. Kidson 
has recently found impressions of a Radiospermic seed of Rhabdo- 
carpus-type (in the sense of Goeppert and Berger rather than that 
of Brongniart) attached to a rachis still bearing pinnules of Neuro- 
pleris lieterophylla, one of the Medulloseae. 2 When it is remem¬ 
bered how deeply rooted the suspicion has become associating the 
Trigonocarpons with the vegetative remains of Medulloseae, 3 and 
further, that in the last few weeks M. Grand ’Eury from his long 
investigations at St. Etienne, has added the weight of his authority 
in favour of the same view, 4 the time when these surmises shall have 
passed into demonstrated facts should not long be postponed. 
In retrospect, the history of the views that have been held as 
to the origin af the Coal-measure seeds is not devoid of interest. 
Passing over the older attribution to Palms (which belongs to the 
1 Proc. Hoy. Soc. lxxi., p. 477. 
2 Phil. Trans. 13 . vol. 197, p. I. 
3 G. Wild, “O11 Trigonocarpon olivaefortne," Trans. Manchester 
Geol. Soc., vol. xvi., 1900; D. H. Scott. “ On the Origin of 
Seed-Bearing Plants.” Lecture delivered before the Royal 
Institution of Great Britain, May 15th, 1903, page n. 
« “ Sur les Rhizomes et les racines des Foug&res fossiles et des 
Cycadofilices,” Comptes rendus, tom. 138 (1904), p. 607. 
