114 THE FERTILISATION OF FLOWERS. [PART JIT. 
alone adapted for nocturnal insects, and Pringlea has become 
anemophilous in the absence of insects suitable for its fertilisation. 
On the whole, Crucifers are far behind Umbellifers in the number | 
and variety of their insect-visitors, both on account of their less — 
conspicuous flowers and their less accessible honey ; and not rarely 
plants remain altogether unvisited. The possibility of self-fertilisa-_ 
tion is useful if not necessary for the preservation of all the Crucifers— 
that we have considered; in many we find that self-fertilisation 
takes place to a very considerable extent; and in several we have — 
experimental evidence that it is productive of seed. Under these 
circumstances it would be better-for the plant to forego attaining — 
in its anthers the most favourable position possible for dusting — 
honey-seeking insects with pollen, if by doing so, while retaining ~ 
the chance of cross-fertilisation if insects did come, it could fully 
- insure self-fertilisation if they did not. 
It is impossible, therefore, to consider with Hildebrand that the | 
twisting of the stamens in Crucifers is “a remarkable expression — 
of repugnance to self-fertilisation” (351, p. 48); it rather refers — 
solely, as Axell rightly explains, to the dusting of insects with the 
pollen on their way to the honey (17, p. 19). But even Axell’s 
statement needed correction and completion, since he only acknow- 
ledged two distinct cases in the arrangement of the honey-glands 
and the twisting of the stamens, and referred all Cruciferz to 
one or other of these two types; and further, he did not take the 
position of the anthers into account as an aid to self-fertilisation. 
-The way in which conspicuousness is attained throughout the 
Umbelliferze by association of many flowers in one surface, and by 
asymmetrical development of florets for the common good, is 
exemplified only in isolated genera of Cruciferze (Teesdalia, R. Br., 
Iberis, L.); and, in TZeesdalia at least, it does not so far insure 
cross-fertilisation that self-fertilisation may be dispensed with. 
Orv. CAPPARIDE. 
Capparis, L., Cleome, L., and Polanisia, Rafin., are proterandrous, 
according to Delpino (172, 352). ) 
Orv, RESEDACEZ. 
42. RESEDA ODORATA, L.—The receptacle is raised up between 
the petals and stamens in the upper half of the flower into a 
perpendicular quadrangular plate, velvety on its anterior surface, 
