74 Salisbury. — On the Structure and Relationships of 
In Physostoma , the most archaic type of the Lagenostomales, the 
megaspore cavity projects into the free portion of the nucellus, the plinth 
being only slightly developed. 
In the more advanced Lagenostoma , the plinth exhibits intercalary 
growth, so that there is a greater extent of free nucellus in the older than in 
the younger phase, whilst in Conostoma the plinth reaches its greatest 
extent. 
These facts seem to indicate that the free apical portion of the nucellus 
was in this series a phylogenetically late development, resulting from a zone 
of intercalated growth. 
On various grounds Physostoma is regarded as the most primitive seed 
yet known , 1 therefore it is of greater significance as being the only 
member of the Lagenostomales exhibiting a free apex of the nucellus into 
which the megaspore cavity projects. We may therefore suppose that the 
potentiality for the development of a free nucellus was possessed by the 
ancestors of Physostoma ; but was only exhibited by the majority of 
the group in the intercalated growth of the region above the contained 
megaspore, resulting in the elaboration of the plinth. 
Probably, then, the common ancestors which gave rise to the Lageno- 
stomales on the one hand and the Trigonocarpeae on the other possessed 
a nucellus unenclosed at the apex, and round the base of which were fused 
a whorl of members with free apical portions, affording protection to the 
sporangium, in which the prothallus was now retained. With the inception 
of the seed habit came an increase in the size of the megaspore cavity, 
accompanying enlargement of the nutritive prothallus. Here is where the 
divergence in the two lines of descent arose. In the one case there was an 
upward extension of the megaspore cavity and the surrounding free pro- 
integumental lobes, resulting in the production of a free nucellus which may 
either have had its origin in an already vascularized sporangium , 2 or the 
nucellar system may have arisen in relation to the greater demands upon 
the water-supply now created. 
Along the line of the Lagenostomales the increased dimensions were 
brought about by a similar phylogenetic intercalation of growth, which took 
place, however, in the lower part where the whorl of protective members 
constituting the pro-integument was laterally fused to form a ring around 
the base of the sporangium. The close proximity of the integumental 
bundles would account either for the suppression or non-development of 
a special nucellar system. Such a view would not only explain the inter- 
mediate characters exhibited by the archaic Physostoma , but it further 
accords with the broad features of the testa as seen in the two groups. 
A marked characteristic throughout the Lagenostomales is that at the 
1 Oliver: Ann. Bot., 1909. 
2 Oliver: A Vascular Sporangium. New Phytologist, vol. i, 1902, p. 60. 
