Trigonocarpus Skorensis , sp. nov. 63 
sclerotic beak and secondary ridges opposite the bundles also point to 
a lesser degree of specialization. 
A further striking difference is the tapering insertion of the seed of 
T. Shorensis as compared with its abrupt insertion in T. Parkinsoni ; the 
former is probably a relatively older type than the latter, just as the 
horizontal departure of the leaf-trace characterizes the modern plant, in 
contradistinction to the oblique insertion, as found in the more ancient. 
Analogy with Physostoma elegans 1 would seem to suggest that the 
presence of a secretory system in the nucellar tissue is a more primitive 
feature than its absence. If this has any significance, and the close relation- 
ship between the Trigonocarpeae and Lagenostomales supports such an 
assumption, then in this respect T. Shorensis stands in the same relation to 
T. Parkinsoni as does Physostoma to the other known members of the 
latter group. 
We have shown good reason for believing that T. Shorensis was without 
a definite inner flesh, except in so far as we can apply that term to the 
lining of the micropylar canal. Such a tissue would, however, appear to 
have been present in T. Parkinsoni. Perhaps at first sight this might seem 
to be a pronounced distinction between the two, but the extreme peripheral 
situation of the sclerized layer in Pachytesta , surrounding an extensive inner 
flesh, coupled with the undoubtedly close relationship of the two genera, 
indicate that the position taken up by the region of sclerization was subject 
to considerable fluctuation in the group as a whole, and therefore of no 
great significance when exhibited in a smaller degree by members of the 
same genus. 
The well-defined nucellar epidermis, even more pronounced than that 
of T. Parkinsoni, indicates a stage less far removed from the condition in 
which the nucellus was a naked sporangium unprotected by a surrounding 
integument ; and probably the production of an inner flesh is likewise 
correlated with a phylogenetically more prolonged contact between the two 
surfaces, which functionally have become internal. 
As previously indicated, the vascular organization of the two seeds was 
essentially similar, perhaps the most outstanding distinction between them 
being the much more peripheral position occupied by the sarcotestal strands 
in Trigonocarpus Shorensis. On the whole, the internal vascular system of 
our seed probably shows a somewhat more pronounced tracheal investment 
at the base of the nucellus, whilst the separate strands into which this 
passed were broader, though this may well be an outcome of the larger 
nucellus they supplied. If, as may have been the case, the Trigonocarpeae 
were derived from fern-like plants having sporangia with a complete internal 
tracheal investment, then it would appear from Scott and Maslen’s descrip- 
tion that in this respect Trigonocarpus Parkinsoni was more primitive than 
1 Oliver : Ann. Bot., vol. xxiii, p. 73. 
