656 GENERAL COMPARISON OF THE FILICALES 
and pectinata, to the basipetal sorus of A/sophila: the addition of a basal 
indusium, together with increased vascular complexity, gives the full type 
of Cyathea. In relation with these genera, though on a minor scale of 
structure, is probably Ozoclea, and as possible last terms leading to the 
mixed sorus may be Woodsia and Hypoderris. But this line requires more 
full investigation before it can be accepted with assurance. 
A third line, which is however more certain, is the MMJatonia-Dipteris 
series. Again the sori are superficial, and the relation of Matonia is 
clearly with the Gleicheniaceae. There can be no doubt of the close 
relation of Dipieris to Matonia, as shown by external form, as well as 
by anatomy: but in place of the simple sorus of A/atonia and of Dipteris 
Lobbiana that of Dipterts conjugata shows the mixed condition. It seems 
clear that there has been a direct transition in this case from the simple 
to the mixed sorus, leading in fact to a form long included in the genus 
Polypodium ; but without the intermediate basipetal condition, though this 
has been found usual in other cases. 
It would thus appear that along several distinct phyletic lines the 
condition with mixed sori may have been attained: This is the most 
advanced state of the present day among the Filicales. It will remain 
for future workers, by anatomical and other enquiry, to disentangle more 
fully the confused phyletic lines of the Polypodiaceae: the present work 
will at least have served to show where the tangle actually lies, and 
some of the probable lines which have led up to it. 
Returning now to the base of the series of the Filicales, as represented 
in the scheme, it remains to consider what idea can be formed of a 
primary type for the group. As there is no clear evidence of the nature 
of the Fern-stock prior to the known types of Simplices, it is only on 
a basis of comparison of these with one another, and with other early 
Pteridophytes, that a cenception can be formed of the probable origin 
of the sporophyte in the Filicales. Comparison, checked by the Palaeonto- 
logical facts as stated at length above, has led to the recognition of the 
following characters as primitive for Ferns: the shoot consisted of a radial 
upright stock, showing occasional dichotomy, with protostelic structure, 
bearing radially disposed leaves, and supported by adventitious roots: 
the leaves were primitively flattened, of relatively simple outline, in which 
dichotomy was a prevalent, but perhaps not an exclusive feature: the 
leaf-trace consisted of a single strand: the meristems of axis, leaf, and 
root were not dominated by a single apical cell, but showed a group of 
initials: all the leaves were potentially sporophylls: the sori were probably 
definite, marginal or superficial, disposed in a single row on either side 
of the midrib: the sporangia of each sorus were simultaneous in origin, 
sessile, and of large size, with high individual output of spores. 
The characters thus summarised indicate an essentially strobiloid type 
not very much removed from some of those which Palaeophytology has 
brought to light. Among the Sphenophyllales and Equisetales are forms 
