PART IL. | THE MECHANISMS OF FLOWERS. 159 
between the upper and lower petals, the (two, rarely three) upper 
ones bearing “ path-finders” in the form of fine black lines, the 
_ (three, rarely two) lower ones being elongated to form convenient 
_ alighting-places. 
The flowers are distinctly proterandrous, so that cross-fertilisa- 
tion never fails if insect-visits take place in time: if they do not 
take place, then the flower is regularly self-fertilised. 
I have only noted the following visitors :— 
A. Hymenoptera—Apide: (1) Apis mellifica, L. §, s. and cp., ab. 
(Sprengel found hive-bees and humble-bees collecting pollen). 3B. Cole- 
optera—(2) Coccinella septempunctata, L., licking honey. 
The awkward way in which this beetle, which is not fitted to 
feed on flowers, obtained the honey, was too amusing to pass by 
without notice. While seated on a petal, it applied its mouth to 
one of the nectaries at the base thereof; while so doing, the petal 
_ usually broke off, and the beetle was left olinging to the next petal, 
or fell with its own to the ground. In the former case it kept on 
_ its way round the flower, and frequently pulled off: all five petals, 
one after another ; but when it fell it was always at once on its 
legs again, running to another stalk of the same plant to climb up 
anew. I saw one beetle fall four times to the ground without 
growing wiser by experience. 
_ __ Ludwig has published more comprehensive researches on the 
mechanism of fertilisation in Hrodium (433, 434, 438). 
Tribe Pelargonie, 
Some general remarks on the structure of the flower in 
| Tropewolum are given by Delpino (172); the structure in 7’ 
_ majus, L., was thoroughly described by Sprengel, who noted its 
_ proterandrous condition 
Tribe Oxalidee. 
_ Hildebrand has investigated all the species of Ozalis in the 
Botanic Gardens of Berlin and Bonn, in the Royal Herbarium of 
_ Berlin, the Munich Herbarium, and the collections of Treviranus, 
| A. Braun, and Selmeyer. He distinguishes twenty trimorphic 
) species, whose different forms were in many cases looked upon as 
distinct species previously; fifty-one species possess two forms, 
and thirty have one only. Our three native species, O. acetosella, ie 
0. stricta, L., and 0. corniculata, L, are shown to be mono- 
- morphic (348). 
