Trelease.] 
416 
[March 15, 
Proteaceae. 
Hakea nodosa (Australia). The small white flowers are arranged 
in axillary clusters of two or three each. In the bud about to 
expand (fig. 5) the style, which is a little longer than the sepals, 
but is held at its tip by them, protudes from the upper side of the 
perianth, its enclosed end being enlarged into an oblique head, 
about which the four sepals, closely united at their end, form a 
closed pocket. Each sepal bears on the inside of its spoon-shaped 
tip an anther, which is already dehiscent, the pollen forming 
little contiguous mounds on the convex end of the style. Cf. 
fig. 10. 
Soon the segments of the perianth expand, the elasticity 
of the bent style playing an important part in their separation. 
The style, freed from the restraint to which it has been sub- 
jected, becomes slightly straighter, while the sepals curve outward 
irregularly, fig. 6. Very little pollen now remains in the anthers. 
The greater part forms a mound on the end of the style, to 
which it adheres from a slight viscidity of the latter, the mass 
holding together through the irregularity of the individual grains. 
Cf. figs. 11, 12 and 16. At first sight the flowers appear to possess 
every facility for self-fertilization ; yet, as in Goodeniaceae, Lobel- 
iaceae, Compositae, Campanulaceae, etc., where self-fertilization 
appears similarly provided for, crossing is the rule. A further 
examination shows that the pollen-loaded end of the style, which 
on a superficial examination would naturally be taken for the 
stigma, is in reality a complex structure. The greater part (cf. 
d, fig. 15) is not stigmatic but is a surface covered with epidermis 
and serving merely as a temporary place of deposit for the pollen. 
It has received the convenient name of stigmatic disk. The true 
stigma (st) is at this time immature, forming a minute nipple in 
the center of disk. Cf. fig. 13. Some days later the stigma comes 
to maturity, appearing now as a papillose cup. Cf. figs. 14-15. 
The flowers therefore present the opposite kind of dichogamy to 
that found in Lemna, being strongly protandrous. 
Nectar is secreted by a large gland at the base of the ovary. 
Fig. 7, n. This will probably be found to attract insects, which 
judging from the size, form, and clustering of the flowers, will 
most likely be small bees of about the size of our Osmias. In 
