io8 Oliver. — On Physostoma elegans , Williamson , 
in Physostoma arises from the possibility that this seed presents us with the 
primitive structure of which the ordinary ‘ tent-poles ’ ( Ginkgo and Cordaites ) 
are surviving representatives — that the apical prolongation of the mega- 
spore-chamber is the primordial tent-pole. In the accompanying Text- 
fig. 9, the three cases referred to are represented, as far as the relations at 
the apex of the nucellus are concerned. 
If the conjecture prove well-founded that the extension of the mega- 
spore-cavity right up to the apex was an archaic trait, derived from a very 
remote past, it would be intelligible that this part of the megaspore should 
show relative arrest or atrophy in its capacity for expansion ; for the 
necessity for the pollen-chamber to become effective at an early period in 
the development of the ovule is quite obvious. 1 And if the pollen-chamber 
matured at an early stage in ontogeny, the stage being accompanied by 
cuticularization of the pollen-chamber wall, the immediate result would 
be an arrest of expansion at the apex, such as we find in Physostoma. The 
cuticularization, no doubt, may have been related to the imperfect pro- 
tection afforded by the integument in the seeds of the Lagenostoma-group 2 
If it be conceded that functional necessity determined an arrest of the 
nucellar apex, the final disappearance or abortion of the megaspore papilla 
(‘ tent-pole ’) would follow in its train, in those cases in which its persistence 
would be without significance. For in the Lagenostoma-group the wall 
of the pollen-chamber was probably sufficiently rigid to resist collapse 
without assistance from a ‘ tent-pole ’, and it is not very likely that the 
portion of the female gametophyte that lay within it was concerned in the 
production of archegonia. If this were the case, the disappearance of the 
apical process and its replacement by nucellar tissue, as we find it in 
L. Lomaxii and L. oroides, is hardly surprising. 
Closely bound up with the features discussed in the last section, is the 
question of the relation of the pollen-chamber to the pristine mode of 
dehiscence of an ancestral megasporangium. In the present state of our 
knowledge no data are available that throw any light upon this very 
important stage in seed-evolution, and under the circumstances discussion 
would be unprofitable. 
3. The Systematic Position of Physostoma. 
Throughout the preceding account of Physostoma a close relation to 
the seeds of the Lagenostoma-group has been assumed. Little remains 
to be said under this head beyond bringing together the scattered references 
in the text which emphasize the affinity. It should be stated at the outset 
1 The small-sized seeds of Lagenostoma Lomaxii possess fully developed pollen-chambers. Cf. 
Oliver and Scott, loc. cit., p. 212. 
2 The tip of the pollen-chamber protrudes beyond the integument in L . Lomaxii. Cf. Oliver 
and Scott, loc. cit., PI. IX, Pigs. 21 and 24. 
