154 
Edith R. Saunders 
to the bottom flower on the axis. They may be detected perhaps 
up to the fifth or sixth, but I have never seen them attain develop¬ 
ment the whole way up the inflorescence 1 . Similar non-vascular 
bract structures are also frequently found in the Wallflower (Figs. 
1-2). Again in some cases, in addition to a subtending leafy bract 
an asymmetric sickle-shaped leaf structure may be present in the 
position of one member of the basal leaf pair (Fig. 6). This peculiar 
shape was at first attributed to injury, but on closer inspection it 
was noticed that the concave margin of these structures always 
differed from that of the convex side in having a transparent, and, 
in coloured forms, often a tinted membranous border (Fig. 7). The 
border and the upper surface of the concave half in hoary as well 
as in glabrous strains was devoid of hairs. This combination of 
characters showed at once that we are dealing here with a structure 
which is half leaf and half sepal 2 . The sepaline sector being capable 
of only a very limited longitudinal extension, the whole structure 
becomes curved. Sometimes the sepaline border may be developed 
for only quite a short distance and then cease; the sepal-bract then 
presents an appearance as though a piece had been bitten out on 
one side. In other cases the sepaline modification occurs along the 
whole of both margins, and we then get a symmetrical boat-shaped 
bract identical in form and appearance with the sepals proper 
(Figs. 8-9). When the complete transformation occurs, the whole 
young raceme may occasionally be found curved into a bow, for the 
sepaline bract is unable to keep pace with the growth in length of 
the axis, and the boat-shaped tip overarching one of the flower buds 
acts as a grapnel and holds the whole short apex down. Where, as 
is often the case with the upper branches there is but little elongation 
of the lateral axis and the basal leaf pair is followed at once by the 
flowers, both of these leaves may be thus modified. The normal 
floral arrangement is not as a rule affected by such transformations, 
the full complement of four sepals being present. If however the 
calyx follows upon a sepal-like bract without any appreciable inter¬ 
node, one of the true sepals may be lacking. This assumption of 
sepaline characters by the bracts was observed by Engler 3 in a 
monstrous form of Barbarea vulgaris, in which several flowers had 
1 In Iberis semperflorens, on the other hand, bracts may attain develop¬ 
ment in the upper part of the inflorescence and be undeveloped in the lower 
(see Eichler, loc. cit.). 
2 The leaf in a hoary stock is hairy on both surfaces, the sepal is glabrous 
on the upper (inner) face and hairy beneath. 
3 Flora, p. 450, 1872. 
