I46 M. Bensoii. 
of the sporangiophore in Pteropsida ? Is it possible that a body so 
characteristic of one half of the Pteridophyta should not have its 
homologue in the other ? I wish to lay stress on the resemblances 
between the synangium or sorus of the Pteropsida and the sporangio¬ 
phore or synangium of the Lycopsida. 
In the vast majority of both Fossil and Recent Ferns the 
sporangia are aggregated together in tufts or groups, or even in 
more compact bodies called synangia. Examples of such aggre¬ 
gations are found in the Palaeozoic ferns Diplolabis, Botryopteris, 
Zygopteris, where the constituent members have relatively little 
cohesion, and in the Marattiaceae where the cohesion may be 
complete. 
The range in the form and number of parts of the Fern 
sporangiophore is greater than in that of the Lycopsida, this being 
probably partly due to the prevalence of the strobiloid habit in the 
latter. 
In the Fern the sporangiophore (sorus) occurs on the margin 
or lower surface of the frond. In the latter case a peltate form 
would be a disadvantage, and we see a great contrast in the method 
of dehiscence of the sorus of Kaulfussia and that of Equisetum. 
When the pedicel is continued into a peltate shield as in Matonia 
we find it is caducous. The sporangiophores of Ferns are more 
comparable in structure with those of non-strobiloid types, such as 
Psilotum. The exceptional cases among the Ferns in which there 
is no aggregation of sporangia at first appear to present considerable 
difficulty, and I will refer to them in detail. 
I will select Senftenbergia and Ai'cluzopteris for discussion as 
they are both Paleozoic forms and hence a pnovi might be consi¬ 
dered primitive. Senftenbeygia is essentially of the Schizaeaceous 
type and bears single sporangia at the ends of veins on the under 
surface of dorsiventral pinnae, while Archceoptevis may bear its 
sporangia peripherally on a pinna. 
We do not know what were the antecedents of these Ferns, 
but I would suggest that the scattered condition of the sporangia 
may be due either to the great expansion and dorsiventral develop, 
ment of the sporangiophore, or to the reduction of a number of 
sporangiophores each respectively to one sporange. Both of these 
evolutionary processes are known to occur. In Acrostichum most 
morphologists would accept the view that the loss of the soral 
identity is due to expansion. In Diptevis bifnvcata 1 the sorus is still 
1 Armour. New Phyt., 1907, Vol. VI., p. 238. 
