Inter-relationships of the Phyla. 6 i 
the Marattiaceae, many Palaeozoic synangia probably belonged to 
Pteridosperms ; in this group, apparently so closely related to the 
ferns that their microsporangia when preserved isolated cannot be 
distinguished from fern sporangia, the balance of evidence seems 
to show that the synangium was a derivative condition (23). In 
recent ferns synangia are normally found only in the Marattiaceae, 
and though the cryptogamic nature of the numerous Palaeozoic 
synangia is doubtful, it would seem that if these fructifications 
belonged to Ferns they are more probably to be assigned to such 
complex forms as the Psaronieae, than to the more primitive types 
such as the Botryopterideae, though in this respect the synangial 
Corynepteris may prove to be exceptional. Thus the geological 
evidence is opposed to the primitiveness of the synangium of the 
Ferns; it is still more strongly opposed to the hypothesis that 
fern-leaves were primitively small or simple. The evidence on this 
point has been given in the earlier part of this article and needs no 
recapitulation. The view that the sorus is the homologue of the 
sporangiophore is not, however, bound up with the primitive micro- 
phylly of the Ferns. Miss Benson, for instance, regards the 
“ sorus ” of the ferns as the homologue of the sporangiophore of 
the Lycopsida, in which category of organ she includes, as ex¬ 
plained above, the stalked sporangium of the Lycopods. Yet she 
assumes the primitiveness of megaphylly, since she states that 
Psuedobornia seems to be a type of probably Calamitean affinity in 
which the microphylly of the Lycopods had not been evolved (2). 
As she regards the Lycopod “ sporangium ” as composed of two 
completely confluent sporangia she would presumably consider the 
synangial type of fructification as less primitive than the sorus of free 
sporangia and leading from the latter to the Lycopod type. If the 
sorus be regarded as the homologue of the sporangiophore such a 
view (leaving out the doubtful question of the synangial nature of 
the Lycopod sporangium already discussed in connection with the 
affinities of that phylum) would appear preferable to Professor 
Bower’s, since it is in accordance with the primitive megaphylly of 
the Spenophyllales and Ferns, and with the priority of the sorus of 
free sporangia, both of which are supported by the geological 
record. But on either view it is hard to account for the position of 
the sporangia ; in the ferns they are borne on the lower surface or 
on the margin of the frond or of the branch of the frond, while in 
all other Pteridophyta they are borne on or near the upper surface 
of the appendages. In this respect Cingularia , probably one of the 
