22 1 
Arber . — On the Past History of the Ferns. 
opposed to the (with rare exceptions) uniseriate annulus of the recent 
Ferns. In another group, the Botryopterideae, this type of annulus is 
correlated with other features which are not found in modern Lepto- 
sporangiatae. It is better, therefore, to regard these usually detached 
sporangia with some suspicion as to their exact affinities. 
And thirdly we have the evidence, from structure specimens, of the 
existence of a peculiar family, in Palaeozoic times, belonging to the Fern 
alliance — the Botryopterideae. The investigations of the late Bernard 
Renault 1 on these fossils, followed by those of Dr. Scott 2 , who was the 
first to point out clearly the affinities of this group, have shown that, 
although the Botryopterideae stand nearer to the modern Ferns than to 
any Palaeozoic group, yet they differ in certain important characters from 
any known Leptosporangiate Fern. It will not be necessary to enter here 
into the points of disagreement in detail. It may serve to mention the 
peculiar nature of the annulus and the fact that the wall of the sporangium 
is more than one layer in thickness, and to recall the unusual form and the 
size of the sporangia. 
Bearing in mind such differences between these Palaeozoic sporangia 
and those of modern Leptosporangiatae as have been pointed out above, it 
would appear to me to be more helpful in our present difficulties with regard 
to the life-line of that race, as well as more consistent, not to attempt 
to bring these ancient Fern-like plants into line forcibly with the modern 
Ferns, however closely we may regard them as allied. Neither the Botryo- 
pterideae nor any of these isolated sporangia can be included in any recent 
family of the Ferns — a conclusion which is not disputed. It would seem 
much more likely that these fossils formed the only fragments which are 
at present known to us of an ancient Palaeozoic race. Professor Lignier 3 , 
in a recent contribution, has also come to what I take to be a similar 
conclusion. 
That this race gave rise to the Leptosporangiatae — using that term 
in the sense applied to the modern Ferns, and here also to the Mesozoic 
representatives — seems more than probable. Such a suggestion is not by 
any means new. Dr. Scott 4 , in his ‘ Studies ’, has clearly pointed out the 
synthetic nature of the Botryopterideae, c in so far as they combine the 
characters of several of the existing families within this division,’ i. e. 
Leptosporangiatae ; ‘ and it is not improbable that they represent the 
stock from which some at least of the families were subsequently derived.’ 
Mr. Kidston 5 has also arrived at a similar conclusion. 
It has seemed to me that it would be convenient if this ancient race, 
of which, no doubt, the Botryopterideae were but one important family, 
1 Renault ( 75 ). 
3 Lignier (’ 03 ), p. 103, and footnote; and p. 133. 
5 Kidston (’ 05 2 ), p. 162. 
2 Scott (’ 00 ), chapter ix. 
4 Scott (’00), pp. 299, 506-9. 
