AND AFFINITIES OF STEPHANOSPERMUM. 389 
COMPARISON WITH OTHER PALÆOZOIC SEEDS. 
Just as among recent plants, where all seeds seem much on the same plane and the 
pollen-tube is the dominant mechanism of fertilisation, though in archaic forms like the 
Cycads we get the end of the zoidiogamie epoch, so in Palzezoic times—or, more precisely, 
in the later periods of the Permo-carboniferous—fertilisation-mechanisms would seem to 
be also much on a plane. We have in the remarkable constellation of seeds from Grand' 
Croix a fair sample of seeds probably in the zenith of their zoidiogamie phase of evolution. 
That amongst them the beginnings of siphonogamy may here occur is possible, so that 
if Renault's surmise of a specialised operculum in Ætheotesta* prove correct, there 
would be distinct indications of a foreshadowing of a pollen- tube. That the Grang’ 
Croix seeds were without pollen-tubes, as we understand them, seems more probable, in 
view of the statement of so careful and indefatigable an investigator as Renault, who 
tells us in more than one passage (prior to the year 1901) that of the 300 odd seeds that 
have passed through his hands this structure has never come under his observation. 
It seems worth while to contrast Stephanospermum with the associated genera of 
seeds with a view to ascertaining in how far they present resemblances or differences in 
their general organisation, and to what extent they may be arranged under provisional 
groups. 
Brongniart and, since his time, all botanists who have touched the subject distinguish 
between the seeds of this period possessing a bilateral symmetry and those which, on the 
other hand, are radially symmetrical. 
The bilateral forms, which for convenient reference may be termed the PLATYSPERME.E, 
include the types known as Cardiocarpus, Rhabdocarpus, Diplotesta, Sarcotaxus, Lepto- 
caryon, Taxospermum, and Cycadinocarpus. Of these, many are probably Cordaitean in ` 
their affinity, whilst Cycadinocarpus may perhaps, as Renault supposes, approach the 
Cycads. They belong to one phase of evolution, and have in common more especially 
the complex chalazal plate from which the bundles both of testa and nucellus-wall take 
origin, the small pollen-chamber, and flattening more or less considerable. 
Brongniart’s other series, which we may term the RADIOSPERME#, include several 
groups based on the form of the transverse section :— 
(1) Stephanospermum, ZEtheolesta, Gnetopsis ................. nnn Circular. 
(2) Trigonocarpus, Tripterospermum, Pachytesta EE E eos ices 3-ridged. 
(3) Ertotesta, CORNOSPEIIUT o ee ee dé Ies o a Octagonal. 
(4) Polylebhospermum, Polypter ospermum, Hexaspermum, Ptychospermum. Hexagonal. 
Of these many are much complicated by the presence of ridges and other emergences 
from the testa. These peculiar specialised forms may be left on one side for the present, 
but attention may be usefully directed to the members of groups (1) and (2). 
Aitheotesta evidently, as a seed, very nearly resembles Stephanospermum. Two species 
have been described—_#. subglobosa, Brongn., Æ. elliptica, B. Renault, 
* Renault, “Sur quelques pollens fossiles,” Soc. d'Hist. nat. d'Autun: extr. du Procès-verbal de la Séance du 
13 juin, 1901, 
