See 
1922] BENSON—HETEROTHECA GRIEVII 135 
Lastly, a difficulty in the synangial theory of the seed has often 
occurred to the writer, but she has not heard it expressed by others. 
It is that in the compartments of the canopy of Lagenostoma a 
vascular bundle occupied a central position instead of, as one 
might have expected, a position in the plane of the lateral walls. 
As will be seen in text fig. 8, this difficulty is completely removed 
by the structure of Heterotheca and Sphaerostoma. Each compart- 
ment in the canopy of Lagenostoma is equivalent to a pair of loculi, 
which, although completely merged in Lagenostoma, can still be 
faintly traced in Sphaerostoma. 
We owe much to Professor OLIVER for the open expression of 
his views published in 1909. The conclusions he finally arrived 
at were opposed to those which seem necessarily drawn from our 
present knowledge. Both theories as to the origin of the seed 
were mere hypotheses in 1908. Unfortunately, the view that the 
canopy of Lagenostoma was the product of a cupular or indusial 
upgrowth, led to the further hypothesis that the Physostoma seg- 
mented integument was a relatively primitive form, and even con- 
tributed to a suggestion that Physostoma was perhaps the most 
- archaic type of seed known, a suggestion wholly contrary to what 
is known of its geological history. In the light of recent work, 
these latter hypotheses necessarily fall together, and it is to be 
regretted that a recent writer (14) should have referred to them as 
facts. OLIVER merely claimed that certain conditions should be 
proved to exist ‘in the ancestor,” but evidently the necessary 
conditions have persisted in’ Heterangium, so that they coexist 
with the formation of an ovule for the megasporangial apparatus. 
Thus a link is provided between an ovule and a microsporangium 
which is stronger than was demanded, and there can be no longer 
any question as to the seed of a Pteridosperm, a seed that may 
well have been the homologue of all Pteridosperm seeds, having 
been produced in the course of evolution as a transformation 
product of a synangium. In this investigation we thus stand 
upon the threshold of the origin of at least one group of the Sper- 
matophyta, and find more indications of it than were expected 
in the structure of a plant of the Lower Carboniferous. 
