390 PROF. F. W. OLIVER ON THE STRUCTURE 
Aitheotesta subglobosa, Brongn., is very imperfectly known. Itis a globular seed, having 
a transv. diam. 12-15 mm. Brongniart sketches the general macroscopic characters of very 
imperfect specimens in figs. 16, 17, 18, pl. C., in his * Graines fossiles. The micropyle | 
is not shown, but the apical end of the seed was flattened and hairy, though there does 
not seem to have been the shallow, trough-like depression that we get in 5. caryoides. 
The basal end of the seed was peculiar, owing to the presence of an aril-like cushion 
around the point of insertion. The nucellus was much retracted and possessed a roomy 
pollen-chamber. Though differing in details, in its general characters this seed was not 
unlike S. caryoides. How near the resemblance goes can only be determined when 
the original specimens are compared *. 
Æ. elliptica, B. Renault.—Of this seed Renault gives considerable details T. Though 
larger and relatively longer than S. caryoides it possesses the same broad type of pollen- 
chamber as that seed. The vascular supply of the nucellus, which is derived in the 
usual way from the chalazal bundle, is nut a continuous mantle but consists of distinct 
strands. The pollen is interesting as the very large grains (320—400 „u X 270-310 nl are 
identified by Renault as being the spores of Dolerophyllum which have lost their extine. 
This author holds the view that the pollen of Dolerophyllum underwent dehiscence prior 
to pollination, and that the inner cell-complex alone penetrated to the pollen-chamber. 
The pollen-grains, except for their larger size, resemble those of S. akenzoides and like 
these became profusely perforated. | 
There was an outer fleshy (?) coat with chief development at the two ends of the seed. 
At the micropylar end this layer was much chambered, an arrangement which Renault 
interprets as a floating mechanism. "Though apparently very similar to Sfephanospermum, 
this seed, apart from the lacunate sarcotesta, differs in the grouping of the nucellar 
tracheides into bundles. 
Gnetopsis elliptica, R. Zeiller.—These are tiny seeds (2:5 mm.x1:2 mm.) bearing 
 feathery arms at the micropyle like the nut of an Abelia. They have been found by 
Renault surrounded to the number of 3 and 4 by a two-membered cupule j. The tracheal 
elements appear to be concentrated in four bundles, but whether they really belong to 
the nucellus seems open to question. The presence of the enclosing cupule here is an 
interesting fact, as it is the only instance as yet amongst the Radiosperms in which 
evidence has been obtained as to the kind of structure to which these seeds were attached. 
It does not appear, however, that the cupule exhibits any characteristic feature sufficient 
to warrant a reference to the plant upon which it was borne. 
So far, then, as this group is concerned, we have smooth seeds, circular in transverse 
section, one with a definite sarcotesta (_Ztheotesta), all with adaptations of one sort or 
another at the micropylar extremity. The Stephanosperma alone show a continuous 
tracheal mantle in the nucellus in contradistinetion to discrete bundles in the other 
* In the contingency of Stephanospermum caryoides eventually being recognised as identical with Ætheotesta 
subglobosa, Brongn., the discrepancy in respect of the pollen would still remain, as the pollen-grains of Ætheotesta 
elliptica, B. Renault, are stated to be of the S. akenioides type. 
T ‘Flore fossile d'Autun, &c. pp. 272-278. 
^t * Cours; iv, pl. 20. 
