_ PART III. | THE MECHANISMS OF FLOWERS. 267 
opposite to that which the anther touches at the same time; and 
_ so cross-fertilisation may result if the insect has come from another 
flower. 
In absence of insects, the flowers mostly wither without self- 
fertilisation taking place, as I have observed in specimens in my 
room. I have found a few flowers in which the stigma, instead of 
being directed obliquely downwards between the two stamens, was 
in contact with one of them from the beginning. 
The only visitors that I have seen are small flies which behaved in the 
manner described. I have collected (a) Syrphidw: (1) Baccha elongata, F. ; 
(2) Ascia podagrica, F.; (3) Melanostoma mellina, L.; (b) Muscide : (4) 
Anthomyia sp., with yellow abdomen. Several other small Muscide and 
_ Syrphidee escaped me. 
or 
On a bunch of Circzea lutetiana which I kept in a glass of water in my room, 
I found a great number of house-flies: (5) Musca domestica, L., sucking honey 
_ and fertilising the flowers, 
Orv. LOASEZ. 
Cajophora lateritia is distinctly proterandrous, according to 
_ Delpino. In the first period the anthers, which lie in five bundles 
in the hollow and outspread petals, rise up one by one, bend 
_ inwards towards the middle of the flower, and pass back into their 
old position after giving up their pollen to insects. With these five 
_ bundles there alternate five groups each containing five metamor- 
phosed stamens; in each of these latter groups the three outermost 
__ staminodes cohere and form in their expanded base a honey-recep- 
tacle, while the two inner ones incline as stiff rods towards the 
_ middle of the flower, and their bases, expanded and fringed with 
hairs, give shelter to the honey. After the anthers have all withered, 
_ the pistil grows up and unfolds its stigma. 
In the first period the insects cling to the central tuft of 
stamens; in the second, to the stiff metamorphosed protective 
staminodes. The fertilisers seem to be bees, according to Delpino 
(177). This is in direct contradiction to Treviranus’ view, that the 
flowers of Cajophora are self-fertilising (742). 
Orv. PASSIFLOREZ. 
Passiflora cerulea, L., was thoroughly described by Sprengel, 
who recognised it as a proterandrous form in which the anthers in 
the first stage, and stigma in the second, came in contact with the 
