7 * 
at the Linncan Society, 
Mr. E. A. Newell Arber, whose address was entitled “The 
Earlier Geological Record of the True Ferns,” said, that any 
consideration of the origin of Gymnosperms must also involve a con¬ 
sideration of the ancestry of modern ferns, since, as was evident 
from Professor Oliver’s remarks, the two groups had a common 
origin far back in the Palaeozoic epoch. His remarks would be 
confined to the Filicinean line of descent. Within the last three 
years so great a light had been thrown on the nature of the 
reproductive organs of the Pteridospermeae, that a correspondingly 
deep shadow had been cast over much of the available evidence 
for the existence of the Palaeozoic ferns. Anyone who had looked 
over museum collections of Carboniferous fossds or had collected 
on the tips or heaps at the mouths of coal-pits, would be aware 
of the fact that impressions of these ferns or fern-like plants were 
both numerous and varied. All such fossils had, so to speak, been 
collected into a great lucky-bag, from which we had, during the 
past few years, been helping ourselves with both hands. Every 
specimen hitherto drawn had proved, or was suspected to belong 
to the Pteridospermeae, so that now we were faced by the 
question, are there any true Palaeozoic Ferns ? Are they all at the 
bottom of the bag? Did they exist at all ? This situation had led 
to what Dr. Scott, in his abstract, had described as being of the 
nature of a “ scare.” Such a scare might perhaps be regarded as 
amusing, when it was remembered that after all there is ample 
evidence from the foliage and anatomy of the known Pteridosperms 
that they themselves must have been derived from Fern-like 
Ancestors. Where, then, were these Ferns ? It behoved us 
to search very carefully the older Mesozoic and the Palaeozoic 
records, and we should probably do well to distinguish, during 
our search, between the Leptosporangiate and the Eusporangiate 
types. There was ample evidence that in the Triassic, Jurassic 
and Cretaceous periods the Leptosporangiate Ferns had attained 
to the position of a dominant type, and that all, or practically 
all, of the modern Leptosporangiate families had been differentiated 
out during the Mesozoic epoch. But if we tried to trace these 
back to Palaeozoic times, we were met by difficulty. In fact 
he doubted very much whether Leptosporangiate Ferns, in the 
modern sense, existed at all in the Palaeozoic rocks. We had 
the Botryopterideae and also a fair number of fronds hitherto 
unassigned, but bearing annulate sporangia. The structure of these 
sporangia however, did not agree in details with those of the 
present-day Leptosporangiatae. The annulus was always bi or 
multi-seriate, in opposition to the uniseriate type of the modern 
forms. It was helpful in our present difficulties to regard the 
Palaeozoic Ferns as an ancient race combining the characters of both 
Lepto- and Eu-sporangiate forms. This was not a new idea, since 
it had been foreshadowed by Scott in his “Studies” (1900). For 
this ancient synthetic race, he (Mr. Arber) proposed the name of 
Primofilices. Such names as Palaeopterideae and Archaeopterideae, 
in many respects preferable, were unfortunately barred by the 
exitsence of the form-genera Palaeopteris and Archaeopteris. On 
one point the geological record was emphatic as regarded the 
Filicinean line of descent. Whether we regarded these Palaeozoic 
Ferns as members of an ancient race, the Primofilices, or as true 
Leptosporangiate forms, in any case they were not one of the 
