224 Sargent.—Fragments of the Flower Biology 
miles away from any congener, produced a few fruits from its wealth of 
blossom. 
The rosy calyces of Guichenotia Sarotes , Benth. x both in shape and 
colour are reminiscent of advanced inflorescence-buds of Bougainvillea 
Sanderiana of horticulture. The floral structure is essentially similar to 
that of Tho 7 nasia montana ; but nectar and staminodes are both wanting. 
The anthers form a cone round the pistil, which is an object of most 
exquisite beauty. This beauty, however, is only revealed by a lens. The 
pale green ovary is sparsely clothed with glistening starry hairs, amongst 
which are dotted plump ovoid trichomes filled with purple sap. The style 
is so densely and delicately clothed with starry hairs that it appears to be 
surrounded by a halo. Description, I fear, fails utterly to convey anything 
like an adequate idea of the beauty. The purple trichomes are, I presume, 
either gnawed or sucked by visiting insects. Either action would doubtless 
cause sufficient agitation of the anthers to shake out pollen. I have not 
seen the flowers visited by any insect. The close relationship of the 
species with TJioniasia induces me to make some mention of it here. 
I now come to one of the choicest gems of the Westralian flora— 
Leschenaultia biloba , Lindl. It seems almost sacrilege to attempt a descrip¬ 
tion, yet one, confessedly inadequate, must be offered. The flower suggests 
to me a bright blue ten-rayed satin star, crinkled and curved most elegantly. 
While celestial blue is the rule the shade of colour of the petals varies from 
almost pure white (very rare) to deepest ultramarine. Sometimes the 
bright blue blossom possesses a pure white £ eye ’, which at times occupies 
nearly one-third of the limb. The corolla is actually, though not obviously, 
two-lipped, and the ‘eye’ is always limited to the anterior lip of three 
petals. The anthers shed their pollen into the indusial pouch, then 
a gaping mouth, some time before anthesis. In the open flower the stamens 
are simply shrivelled shreds. The style is a thick column of tissue with 
a broadly expanded, slightly dorsiventrally flattened apex. The structure 
of the expanded part is remarkable. On its anterior edge there is a furry- 
looking whitish line, which indicates the closed mouth of the pollen- 
containing indusium. On the opposite side of the style top there is a broad 
slightly curved line of short fleshy hairs. The central space is occupied by 
a glistening, sticky, oval patch—the stigma. The edge of the indusium 
stands about one millimetre distant from the corolla tube. An insect 
entering the flower would first brush against the fringe of fleshy hairs and 
the stigma almost simultaneously; it would then brush against the edge of 
the indusium, and thus cause the pouch to open, and allow pollen to fall out, 
or at least expose it, so that some might be removed. I have spent much 
time in watching the plants ; but only twice have I seen flowers visited. 
Once a small indigenous ‘bee’ entered a flower just as I have described ; 
but the pouch was empty, so I did not see the complete operation. The 
