75 
Trigonocarpus Shorensis , sp. nov . 
apex of the seed the multiple nature of the integument exhibits itself with 
almost sudden clearness after it becomes free from the nucellus. In the 
seeds of the Trigonocarpeae, on the other hand, the component units of 
the testa are in general equally distinct at the base as at the apex, or even 
throughout their length. These two conditions seem only explicable on the 
hypothesis that in the one group lateral fusion was simultaneous throughout, 
representing in fact the elongation of the free parts, and in the other took 
place in the ‘ canopy ’ and body of the seed at phylogenetically distinct 
periods. 
Our investigation of the present seed has shown us that the external 
periphery of the integument was circular in outline throughout; in other 
words, the lobing is only shown superficially at the internal periphery, 
where it extends from near the base to the apex, though in the body of the 
seed the six lobes manifest near the chalaza, and reappearing at the micropyle, 
had become reduced to three. 
What is true of this species was, judging from impressions with sarco- 
testa preserved, true also for its congeners, and may well have been a group 
character. The Trigonocarpeae were then characterized by an internal 
lobing almost to the base of the seed, whereas in the Lagenostomales such 
internal lobing is only to be found in the apical region, so that superficially 
as well as internally the major part of the integument in the one is to be 
homologized with the distal extremity in the other. 
We have assumed that the ancestral fructification of the two groups 
possessed a whorl of surrounding members fused at the base between them- 
selves. In both Trigonocarpus Shorensis and Trigonocarpus Parkinsoni 
the sutures of the primary ribs disappear a little before the floor of the 
seed cavity is reached. 
The structural facts therefore justify the assumption that in both 
groups there was a phylogenetically earlier lateral fusion of the pro- 
integumental members at the base, followed later by their fusion at the 
apex. 
The congenital fusion of the integument with the contained nucellus 
was but a matter of time, naturally evinced first in the basal region or zone 
of earlier lateral fusion. This latter, in the Trigonocarpeae, is wellnigh 
vestigial, so that a free nucellus resulted, whilst in the Lagenostomales it 
represents the greater part of the seed body. 
Later on in the history of the Trigonocarpeae and their descendants, 
congenital fusion followed in the upper part, giving us the condition in 
modern Cycads ; though here, as in the more recent Lagenostomales, the 
apical region has remained permanently free in relation to its specialized 
structure and functions. 
