PART III. | THE MECHANISMS OF FLOWERS. ; wa 
41. RAPHANUS RAPHANISTRUM, L., has four honey-glands 
| arranged as in S. arvensis, but all the anthers remain turned 
towards the stigma, the shorter ones standing on a level with it, so 
that the chance of self-fertilisation is increased. 
‘I have seen the honey-bee, which usually confines itself to 
_ one species of plant, visit flowers of R. raphanistrum and S. arvensis 
_ indiscriminately. 
| Subularia aquatica, L.—When the water is unusually high, the 
| flowers remain closed and submerged, and fertilise themselves (17, 
i; REVIEW OF THE CRUCIFER&. 
With great uniformity in the general structure of the flower, 
_ Crucifers show such variability in the number and position of 
_ the honey-glands and in the situation of the anthers in relation 
_ to these and to the stigma, that scarcely two of the species that we 
have discussed agree fully i in these points. 
_ The number of honey-glands may be two, four, or six; they 
are placed on or between, the bases of the stamens: and in the 
priier case, either on their inner side, their outer side, or round 
about them; intermediate positions are sometimes assumed by 
: Bstortive and furictionless glands. The honey when secreted either 
remains in drops upon the glands or fills spaces between the 
_ stamens and pistil or accumulates in sepaline pouches. 
____ All or certain of the anthers are placed so that honey-seeking 
insects must touch them with one side and the stigma with the 
other; but their situation is not always the most favourable for 
this Fedult; and the more they diverge from the position most 
| favourable to cross-fertilisation, the more chance do they give of 
' self-fertilisation. 
| This is explained by the following circumstances. 
Tn all the Crucifersee which we have considered, the position of 
_ the anthers in relation to the honey-glands and the stigma renders 
_ cross-fertilisation likely, but not by any means inevitable, on the 
visit of a honey-seeking insect; but it does not in any way favour 
_ cross-fertilisation more than belffortilisation if the insect comes to 
feed on or gather pollen. The abundance as well as the diversity 
_ of insect-visitors is not great in the species we have considered ; we 
find chiefly ‘flies (especially Syrphide), bees (especially the less 
Specialised forms), and in less degree other Hymenoptera (espe- 
| cially Sphegide), Coleoptera, and Lepidoptera. Hesperis tristis is 
I 
