226 Sargent—Fragments of the Flower Biology 
below upwards. Not a single column flew over, though all were' in the 
irritable condition ! The back of this bee appeared to be pollen-dusted. 
I think the anthers or stigmas (according to the stage of flowering) usually 
strike the back of a visitor when the column reacts to gentle pressure. In 
some species the under-surface of the visitor may be struck. Anthers ripen 
first, and are pushed aside by the growing style tip just before the stigmas 
mature. The motion of the column is too rapid for the eye to 
follow. 
A pollination mechanism recalling that I have described for species of 
Proteaceae is found in Verticordia pennigera , Endl., and V. densiflora , 
LindL, and probably other species, but I have not specially observed them. 
The style is an erect column with an apical stigma. Just below the 
stigma is a ring of hairs; and on to this ring the anthers discharge their 
pollen-grains, floating in a sticky fluid just before the flower opens. The 
stamens are not functionless in the open blossom ; they partially close 
the floral orifice and so make the nectar less readily accessible. The 
flowers are visited by hymenopts, which are, I believe, efficient pollinators. 
The flowers of Actinodium Cunninghamii , Schau., which are borne in 
heads strongly suggestive of ‘ double J Beilis perennis capital a, follow a 
closely similar plan to that of Verticordia , and are, I believe, also pollinated 
by Hymenoptera. 
The mechanical structure of the flowers of Compositae scarcely differs 
essentially from the above-named. I have seen the flowers of Helichrysum 
Lawrencella^ F. Muell., visited by small hymenopts which seem to pollinate 
effectively. 
Marianthus Imeatus , F. Muell., depends chiefly for pollination upon 
a small dipteran, as I have shown at some length in ‘Journal N. H. & S. 
Society of W. A.’, vol. iii 9- No. i. The plant blooms during the hottest 
months of the year—January and February. Its elegant blossoms, cream- 
coloured with purple longitudinal stripes, are borne in great profusion. 
In the bud, neglecting the pistil, the flower, even to a fairly late stage, 
is actinomorphic: sepals, petals, and stamens are regularly spaced in their 
respective whorls, and in each whorl the units are not perceptibly dis¬ 
similar in any respect. When the blossom opens, though their claws are 
still almost equally spaced round the sporophylls, the limbs of the lateral 
and anterior petals are bent back, so that the complete corolla forms 
a half-bell only at the back of the flower. The stamens now differ in 
length and are so arranged that their anthers form a compact little cushion 
facing the petals. About four 'days after the flower opens the small stigma 
will be found occupying the place of the anther cushion, the stamens having 
spread widely away from their first position. Nectar is not secreted ; but 
the tiny teat-like trichomes, which thickly stud the petal bases, contain 
a sugary juice. My observations and experiments point to the conclusion 
