374 Bower. — The comparative examination of the 
the scale. Accordingly, while we rightly regard aquatic 
Phanerogams, which in many respects show similarities in 
detail of structure to the filmy Ferns, as exhibiting degenera- 
tion in accordance with their aquatic habit, I think that con- 
sidering their position in the scale of Vascular Plants, we are 
justified in recognising the series of Ferns from the Hymeno- 
phyllaceae to the Marattiaceae, as an ascending series, and 
that this series illustrates the emergence of one phylum of the 
Vascular Cryptogams from the semi-aquatic to the aerial 
habit 1 : it is this which seems to me to give a special interest 
to the singular parallelism of characters of the meristems 
which I have demonstrated in this paper. 
But in our series of Ferns the progressive adaptation of 
structure to external circumstances is especially to be recog- 
nised in the characters of the sporangia : in the Hymenophyl- 
laceae these are aggregated in sori protected by a cup- like 
indusium from drying up while young, the youngest sporangia 
being nearest the base of the sorus, and accordingly 
most completely protected : in the Cyatheaceae the arrange- 
ment is similar, though the protection is somewhat less 
close and complete : in the Polypodiaceae the indusium 
serves as a protective cover during early stages : in all 
these Ferns the sporangia are stalked, and of small 
size, and each produces a limited number of spores, the 
sporangia being relatively numerous. In the Schizaeaceae 
the sporangia have shorter stalks and are of relatively larger 
size ; they are not aggregated in so large numbers as in the 
Polypodiaceae, but (to judge from Prantl’s figures 2 ) this is 
compensated for by the larger number of the spore-mother- 
cells, and spores produced by the single sporangium. These 
characters become still more pronounced in the Osmundaceae : 
1 Obviously the converse view is capable of defence : it might be held that the 
Hymenophyllaceae owe their simple construction to their adaptation to a semi- 
aquatic habit, and that the Marattiaceae are the original type, from which the 
simpler Ferns have degenerated: general considerations of comparison of both 
generations, and of their position in the whole system lead me to think this view 
improbable. 
2 Schizaeaceae, Plate VIII, Figs. 120, 121. 
