meristems of Ferns as a Phylogenetic Study . 357 
naturally follows that the £ filmy ’ character is not a safe index 
of affinity, but that it is rather to be regarded as the result 
of adaptation. The general discussion of the filmy character 
in Ferns will be taken up later (page 378). 
Passing on to Angiopteris evecta as a type of the Marat- 
tiaceae, the thick coriaceous wings may be traced in the 
mature leaf as extending, with varied development, over the 
upper part of the phyllopodium, pinnae, and pinnules.: passing 
down the phyllopodium the wings become gradually reduced, 
till above the ‘ stipules 5 they are lost, and the base of the 
phyllopodium is almost perfectly cylindrical : at the base are 
the massive stipules, joined transversely by the fleshy com- 
missure, and containing a well-developed system of vascular 
bundles. The development of these basal parts has already 
been described elsewhere 1 , their first appearance being as a 
thick and massive weal, in which the component cells undergo 
repeated T divisions, and on comparative grounds it has been 
concluded that notwithstanding their size, massive character, 
and the presence of vascular bundles in them, they are to be 
regarded as peculiarly metamorphosed wings, and as compar- 
able to the somewhat similar, though smaller growths in Todea 
and Osmunda 2 . It is well known that these stipules as they 
develop run out to a thin edge, which is almost transparent, 
with uneven margin, and consists at the extreme edge of but 
a single layer of cells ; this we may I think rightly compare 
with the thinning out of the wings of the pinnae in Todea 
hymenophylloides and superha , and in this sense even the 
massive stipules of Angiopteris may be said to have £ filmy 3 
margins. 
The development of the wings of the pinnules was traced with 
1 Phil. Trans. 1884, Part II, p. 582 ; Plate 37 , Figs. 10, 11. 
2 The presence of vascular bundles in the stipules of Angiopteris , while the 
corresponding outgrowths in Todea have none, need be no obstacle, since wings in 
different plants are very variable in this respect, e. g. bundles are absent in the 
wings of the pinnae of Cycas , while they are present in those of Stangeria. Again, 
the occurrence of vascular bundles is very irregular in the wings of the leaf of 
Angiopteris itself : it would appear rather that the presence of vascular bundles, 
here as in other parts of the shoot, depends roughly upon the area and bulk of the 
wings. In the stipular growths of Stangeria no vascular bundles are present. 
B b % 
