220 
Arber. — On the Past History of the Ferns . 
giate Ferns in the primary rocks, still the evidences of their occurrence are 
at least exceedingly rare, and the question whether they existed at all in 
those early times is not even yet placed beyond the possibility of doubt by 
observation of microscopic sections.’ These conclusions, at that time } 
were probably regarded as somewhat extreme, at least by those who were 
more especially concerned with the fossil evidence. At any rate, at the 
present time we know from more than one source, that annulate sporangia, 
which, in certain respects at least, may be somewhat closely compared with 
those of recent Leptosporangiatae, do occur in the Palaeozoic rocks. How 
far we are warranted, in the absence of any developmental evidence in the 
fossil state, in referring them to the Leptosporangiate Ferns, is a point 
which may now be considered in some detail. 
The sources of evidence are threefold. In the first place certain im- 
pressions of Sphenopterid fronds have been described, sufficiently well 
preserved to permit of a detailed study of the sporangia which they 
occasionally exhibit. One of the best known of these, Hymenophyllites 
quadridcictyl ites (Gutbier), the fertile fronds of which were described by 
Zeiller 1 , is generally held to be at least a Leptosporangiate Fern, and very 
possibly a member of the Hymenophyllaceae. Similarly the fructifications 
known as Oligocarpia and Senftenbergia , which are borne on fronds belong- 
ing to the form-genera Sphenopteris and Pecopteris respectively, are among 
others which appear to be, so far as structureless specimens will permit us to 
judge, of the nature of Leptosporangiate sporangia. But even if we regard 
some, at least, of these sporangia-bearing fronds as really Leptosporangiate 
Ferns, a conclusion which I am inclined to think is probably correct, it must 
at the same time be admitted that, in point of numbers, such plants were 
but few in Upper Palaeozoic times, and did not then form one of the 
dominant groups. I think it possible that, in some of these, we may 
recognize the beginnings of the thin, subsidiary phase of the life-line of 
Leptosporangiatae. 
In the second place we have more satisfactory specimens, in which the 
structure is preserved. In both the Lower Coal Measures of England and 
the Permian of France, the calcareous or silicious nodules are found to 
contain, here and there, isolated sporangia or sori, for the most part not in 
obvious continuity with any other organ. Dr. Scott 2 has recently described 
a striking case in which such sporangia contained spores which had begun 
to germinate within the sporangium^ In this and other examples there is 
a well-marked annulus 3 . If, however, we closely compare these organs 
with those of a recent Leptosporangiate Fern, we shall find that they do not 
exactly agree in certain respects. For instance, as Dr. Scott has pointed 
out, we invariably find a bi- or multiseriate annulus in these fossils, as 
1 Zeiller (’88). 2 Scott (’04). 
3 Exannulate sporangia also occur, vide Scott (’05 2 ). 
