AND AFFINITIES OF STEPHANOSPERMUM. 391 
forms. In none of the forms other than the two seeds here deseribed has any observation 
been put on record as to a layer internal to the sclerotesta. But perhaps special attention 
has not been directed to this point. 
The second group of the Radiosperme:e, which includes Trigonocarpus, Tripterospermum, 
and Pachytesta, is an interesting one demanding some consideration. ` 
Triplerospermum, Brongn., seems to be as unspecialised a seed as any from the 
Paleozoic rocks. It is an oval seed with a beak-like micropyle and possessed a sarcotesta 
external to the sclerotesta. ` Longitudinally, there run three symmetrically placed 
ridges ; ultimately dehiscence probably occurred along these lines. "To this genus the 
Trigonocarpons from the Lancashire Coalfield described by Sir J. D. Hooker and Binney * 
probably belonged, as Brongniart has suggested T. 
The seed described by Williamson} under the name of Trigonocarpon oliveforme, 
though slightly larger than the Hooker and Binney specimens, unquestionably belongs to 
the same affinity. Consequently, for the purposes of the present discussion, both these 
Trigonocarpons may conveniently rest in the Brongniartian genus Tviplerospermum $. 
Particular attention was drawn to the Hooker and Binney specimens from a study of 
pl. 4. figs. 5, 9, and 10 of their Memoir. Here is shown a circular ridge surrounding the 
base of the micropylar beak which recalls the crown of a Stephanospermum. On consulting 
the original specimens || I find that this apparent ridge is in reality due to a corrosion or 
decomposition of the tissue in this region, so that the marked resemblance is illusory. 
The only other point about these seeds calling for mention is that the tracheal elements 
ensheathing the nucellus are well preserved, though it is diffieult to decide whether they 
formed a continuous mantle or separate bundles. The pollen-chamber is smaller 
relatively than in Stephanospermum. 
Pachytesta Y was the largest of the Palwozoie seeds, approaching a hen's egg in 
dimensions as in form. The testa was massive and complicated, whilst the chalaza was 
remarkable in the presence of a conspicuous peduncle upon which the nucellus stood. 
Apart from the complexity of the testa, which possessed an extensive endotesta, this 
seed is characterized in that the nucellus appears to have borne longitudinal ridges which 
engaged with corresponding grooves in the endotesta. In view of these and other 
peculiarities, Pachytesta may perhaps be detached from the near alliance with 7'/igono- 
carpus and Tripterospermum implied in Brongniart’s grouping of the genera. 
Trigonocarpus, Brongn., may be provisionally placed in this section. T. pusillus was 
— 
* * On the Structure of certain Limestone Nodules, &c.," Phil. Trans. vol. 145 (1855), p. 149. 
+ * Graines fossiles silicifices, p. 26, footnote. 
t “On the Organisation of the Fossil Plants of the Coal-Measures: Part VIIL," Phil. Trans, 1877, p. 248. 
$ I am in agreement with Williamson (loc. cit. p. 253) that a redundancy of genera has been created for the 
known Radiosperms. 
| It is owing to the kindness of Sir Joseph Hooker that I have been able to trace the originals to the Geologieal 
Department of the Natural History Museum. Here are to be found the specimens (all longitudinal sections) upon 
which are based figs. 6, 7, 8, and 9, pl. 4, of the Hooker and Binney Memoir. Sir Joseph Hooker's figures are drawn 
in natural size and are practically facsimile representations of the specimens. : 
«| For full description see Brongniart, loc. cit. p. 25, pl. B. figs. 4 & 5, and pls. 18-21; also Renault, ‘ Flore 
fossile d'Autun et d'Épinae, p. 389, and pls, 83 & 84. figs, 1 & 2. 
