346 Bower . — The comparative examination of the 
Tr. reniforme towards the base of the leaf is for that 
species. 
From the above descriptions it would appear that in the 
Hymenophyllaceae the wings extend from the base of the 
frond upwards over its whole margin, that (with the two 
exceptions above noted) the marginal cells have the form 
of half of a narrow disc : that segments are cut off from these 
by successive transverse (anticlinal) walls, and the fact that 
this mode of segmentation is retained, whether the expansions 
consist of but one layer of cells, or of more, is of special in- 
terest as showing that the arrangement of cell-walls is not 
entirely dependent on bulk, or on the external conformation of 
the part, but that there is a conservation of the type of segmenta- 
tion in plants of natural affinity. Since the curvature of the 
segmental walls is very slight, while the walls by which the 
segments are further subdivided in Tr. reniforme and H. dila- 
tatum cut them at right angles, it is clear that the centre of 
construction lies within the marginal cells themselves, and is 
situated comparatively near to the extreme margin of the 
wing. 
Passing on now to the other leptosporangiate Ferns, in these, 
as in the Hymenophyllaceae, the leaf is habitually winged to 
its extreme base ; but the wings of the lower part (petiole) 
are of a reduced type, and frequently appear there as mere 
longitudinal, light-coloured streaks. It is in Ferns of this 
class that the development and segmentation of the marginal 
series of cells is best known, and chiefly owing to the work of 
Kny and Sadebeck. It is generally known that a marginal 
series of cells is present, and that from them segments are cut 
off successively to form the flattened wings. On examining 
Kny’s figure of Ceratopteris 1 it would appear that the seg- 
mentation of the marginal cells is similar to that in the 
Hymenophyllaceae, and this is made clear from a foot-note 2 , 
in which this is stated to be the case. As far as I am aware, 
this, and in an obscure manner Azolla 3 , and apparently also 
1 Kny, Parkeriaceae, Taf. XXIV, Fig. 7 . 2 1. c. p. 40. 
3 Strasbnrger, Ueber Azolla, p. 38, Plate I, Fig. 20. 
