6\ Salisbury. — On the Structure and Relationships of 
T. Shorensis , for in the latter anastomoses, if they existed, were probably 
of infrequent occurrence. 
2 . Trigonocarpus O liver i ; its systematic position. 
This seed was described and figured by Scott and Maslen in 1907, 1 
from a series of four sections through a single seed, and the diagnosis there 
given is as follows : 
‘ Length nearly 2 cm., diameter about 0-9 cm., characteristically coffin- 
shaped in vertical sections. Base flattened. Sclerotesta produced around 
the base of the seed in the form of a circular ridge enclosing the stalk of 
the seed. Longitudinal ridges of the sclerotesta acute-angled, not rounded 
as in Trigonocarpus P arkinsonil 
The number of longitudinal ridges which the seed bore is not explicitly 
stated, but, presumably from the description given, was assumed to have 
been six. 
The writer has carefully examined the preparations and employed for 
their interpretation the methods recently described. 2 The conclusions 
arrived at differ essentially from those of Scott and Maslen, who evidently 
did not fully recognize the marked effects of obliquity in this seed, which 
are so clearly brought out by the modelling method. 
As can be seen, the first section (S. 28, d) z is tangential to the surface of 
the seed and passes at unequal depths through two ribs, the divergent axes 
of which, towards the upper part, indicate that the plane was below the 
middle region, sloping away from the chalaza. The next section (S. 28, c) 
shows the more deeply cut of these ribs sectioned nearer the axis, and con- 
sequently represented by two angular projections, one at the apex and the 
other at the base. This section cuts the seed so far in that two lateral ribs, 
one on either side, are also encountered ; where these are cut near the apex 
they appear as angular projections, whilst near the base they exhibit 
a curious truncated outline, but are unequal in size, owing to the obliquity 
which was also manifested in the previous section. It is these two lateral 
ridges at the base which have been interpreted as a circular ridge, though 
the true character where sectioned near the apex was recognized. In order 
to explain the peculiar form of the lower projections, the assumption was 
made that they were incomplete. The objections against such an interpre- 
tation furnished by this section alone are, that if a chalazal ring were 
present, it is highly improbable that the central rib would extend below it 
and the lateral ribs be in no way represented ; whilst this difficulty cannot 
be overcome, since the interpretation of the median ridge as a part of the 
stalk is inadmissible, owing to the plane of section. 
1 The Structure of the Palaeozoic Seeds Trigonocarpus Parkinsoni and Trigonocarpus Oliveri. 
Ann. Bot., vol. xxi, No. lxxxi, 1907. 
2 Salisbury: Methods of Palaeobotanical Reconstruction. Ann. Bot., April, 1913. 
3 Loc. cit.,Pl. XIII. 
