228 Arber . — On the Past History of the Ferns. 
But a close inquiry will show that such evidence as has been brought 
forward to support this attribution is by no means convincing. The fossils, 
whose claims to be regarded as members of this group have been so 
confidently urged by Dawson 1 and Penhallow 2 in Canada, and by Reid 
and Macnair 3 in Scotland, are, curiously enough, among the oldest plants 
known to us, for they are derived from Devonian rocks. If the Hydro- 
pterideae really were, as has been insisted, a dominant race in that period, 
it is curious that we meet with no trace of them in the Lower or the Upper 
Carboniferous floras. But there are other objections, which are much more 
weighty. It has not, in my opinion, been demonstrated that the fossils, 
described as sporocarps by Dawson and others, have any real structural 
resemblance to those of the modern Water-Ferns. It may quite well be 
true that such bodies as Parka decipiens , Flem., are not typical sporangia. 
But they are not for that reason necessarily sporocarps. Also, in no case 
has any trustworthy evidence been found as to the nature of the plant 
which bore them. In some instances they have been attributed to certain 
stems occurring in the same beds, but it must be clearly understood that 
the evidence rests entirely on association alone, which in such cases is, by 
itself, of little or no value. There is also a total absence, among the 
fossils associated with these obscure fructifications, of plant remains, which 
could be regarded, by any stretch of the imagination, as in the least similar 
morphologically to members of the recent Salviniaceae or Marsiliaceae. 
In the typical Mesozoic floras there are very few cases in which one 
would be led to suspect that any of the Fern-like plants should be assigned 
to this group. Certainly, there are no grounds at present for believing that 
the Hydropterideae were a dominant race in Mesozoic times. It is not 
until we reach the Angiospermous floras that we find definite traces of 
these Ferns. In one case, however, there does appear to be fairly good 
evidence for the conclusion that we may be dealing with a Mesozoic Water- 
Fern. The genus Sagenopteris , a typical Rhaetic frond, and one which 
persisted until Wealden times, may have been of this nature. The form of 
the frond, composed of four leaflets springing from a common petiole, is 
not at all unlike that of Marsilia. In some species (e. g. 5. alata , Nath.), 
as Nathorst 4 has shown, the petiole is winged, and thus recalls a feature 
sometimes seen at the base of the stalk of the leaves of the recent genus. 
But a more important point is that impressions of fructifications, which 
certainly do resemble very strongly the sporocarps of the Water-Ferns, 
are sometimes found in association with the fronds of Sagenopteris . In 
the absence of any known instance in which a fructification of the Lepto- 
sporangiate type has been observed on the fronds of this genus, this associa- 
tion is distinctly suggestive. However, the affinities of this interesting 
1 Dawson (’86) and (’88), chapter iii. 2 Dawson and Penhallow (,’92).. 
3 Reid and Macnair (’99). 4 Nathorst (’78), p. 85, pi. I, fig. 17; ph XIX, fig. 4. 
