73 
Trigonocarpus S /wrens is, sp . nov. 
and double system of integumental strands in Pachytesta evinces a tendency 
that might well have subsequently developed, and its origin in consequence 
have gradually become obscured. 
With regard to the second argument in relation to Lagenostoma , recent 
work has shown that, in spite of the free nucellus exhibited by the Trigono- 
carpeae, they agree both in the structure of their fructifications and stem 
anatomy much more closely with the modern group than do the Lageno- 
stomales . 1 
When we turn to Trigonocarpus Shorensis we find that the evidence 
for a homogeneous origin for the whole integument is abundantly clear. 
Such inner flesh as is present is but the unsclerized internal lining of the 
hard shell, and this latter, though broadly composed of an outer and inner 
part, is formed from one and the same tissue by the different course which 
its elements pursue. 
The sarcotesta on its inner periphery shows a complete transition from 
the outer sclerotestal cells, so there is no zone to which we can point as 
possibly representing the fusion plane of two morphological units. And if 
further evidence were necessary it is furnished by the course of the secretory 
elements which pass from sclerotesta to sarcotesta irrespective of the 
differences of texture which, in the mature fruit, these regions present. 
IX. The Origin of the Integument in the Trigonocarpeae 
AND THE LAGENOSTOMALES. 
The addition of yet another Trigonocarpean seed, showing a well- 
defined nucellar epidermis, adds to the certainty that the group was 
characterized by the possession of an extensive free region of the nucellus, 
to which part the megaspore cavity was almost entirely confined. This 
would seem to be a fundamental difference separating them off from the 
other members of the Pteridospermeae. 
The considerable resemblances which the Trigonocarpeae bear to the 
Lagenostomales, both as regards the general organization of their seeds 
and the broad features of anatomy, in the few stems which have been 
allocated to fructifications, indicate that the two must have had a closely 
allied ancestry, either in some semi-Pteridospermic group which possessed 
a generalized type of fructification, giving rise to the characters of both ; 
or in two parallel developments arising independently from the Pteridophytic 
stock. 
The theory of intercalated growth put forward by F. W. Oliver in 
a similar connexion 2 to homologize modern with fossil forms, applied in the 
manner indicated below, seems to offer the best explanation of both the 
resemblances and differences which these two groups exhibit. 
1 Kershaw, loc. cit. ; and de Fraine : On Sutcliffia insignis. Ann. Bot., vol. xxvi, 1912. 
2 The Ovules of the Older Gymnosperms. Ann. Bot., vol. xvii, 1903. 
