366 PROF. F. W. OLIVER ON THE STRUCTURE 
thin-walled palisade, or perhaps of a succulent cushion. On the whole a delicate tissue 
seems indicated, for the filaments are rarely preserved although they occupy the most 
secluded position that the whole surface affords. The projecting rim may have been 
continued to the lip of the micropyle as a thin shell overlying a spongy tissue such as 
we know to have existed in Ætheotesta elliptica*. If so, the lid must have been 
caducous, as no indications of disintegration or splintering are found on the sharp edge 
of the crown. In these circumstances our seed must have presented a very close 
resemblance to Ætheotesta. 
In this connection it is interesting to recall the fact that Renault has conjectured, 
from analogy with Gnetopsis (another seed from the same horizon), that the crown of 
Stephanospermum was the base of attachment of a plumed organ concerned in the dispersal 
of the seed t. : 
Whatever may have been the function of this micropylar organ, the fact is not without 
interest that the micropyle of this seed was more complicated than had previously been 
shown. Apart from theapex of the seed, no certain trace of a soft tissue external to the 
shell has been found. But, as it sometimes happens that even the hard sclerotesta of this 
seed shows evidence of superficial wear, it would be rash to conclude that a sarcotesta 
had not existed. From analogy with Ætheotesta $, Trigonocarpus$, and Polylopho- 
spermum ||, members of the Stephanospermum-group from the same horizon, the existence 
of such a layer is not wholly improbable. 
As regards the lining of the integument, this layer of flattened cells with dark contents 
may be traced from the chalaza to the commencement of the micropylar canal. Here it 
seems to lose its characteristic appearance (cf. Pl. 48. fig. 29, and Pl. 41. fig. 4, e£.), and is 
continued as an inconspicuous lining to the canal traceable with difficulty to the apex. 
That it was a sclerotic layer seems open to question. 
The Chalaza §.—-The base of the seed where nucellus and testa are confluent is not 
without interest. The sclerotesia in this region is much broader than elsewhere, due to 
the fact that the palisade-cells become about three times as deep here as elsewhere. At 
the centre, around the entering bundle with which these cells are parallel, they are 
undivided and have a length of ‘8 mm. (see Pl. 43. fig. 21, /p.), whilst further away from 
the bundle they are segmented into two or three storeys. Beyond, as we come level with 
the separation of sclerotesta and nucellus, the palisade-zone has become reduced to its 
normal proportions. Curved rows of short oblong cells of the same histological characters 
pass backwards from the chalaza and constitute the inner zone of the sclerotesta (see fig. 21, 
st.). Lying above the palisade-elements, and occupying the continuation of the plane of 
the gap between sclerotesta and nucellus, is a plate of not very thick-walled polygonal 
elements (sc/.) stretching from the angle of confluence of the lining of the sclerotesta and 
* Renault, * Flore fossiie d'Autun et d’Epinac,’ pt. 2, p. 272. 
t Cours de bot. fossile, vol. iv. 1885, pp. 184 & 186; also ‘Etudes sur le terrain houiller de Commentry, Flore 
fossile, pt. 2, 1890, pp. 669—670. 
t Renault, * Flore fossile d'Autun et d’Epinac,’ pt. 2, p. 272, fig. 49. 
$ Renault, loc. cit. p. 397. || Brongniart, * Les Graines fossiles, p. 28, and pl. C. figs. 6 & 7. 
‘| Cf. Renault, ‘Cours de bot. fossile, vol. iv. pl. 22. fig. 1, & pp. 185, 186. 
