428 On the Fertilization of Several Species of Lobelia. (July, 
in which the piston remains stationary while the barrel moves— 
pollen would be pumped out of the moving anther-tube by the 
stationary style with its ring of hairs. Moreover, when the 
staminate stage of the flower (lasting about three days in ZL. 
erinus) is ended, and’ the style protrudes from the orifice of the 
anthers, the hairs on the style become reflexed, thus apparently 
serving to prevent such pollen as is not infrequently left in the 
anther tube from falling out upon the papillose stigma when this 
is unfolded. 
In L. erinus, and twelve out of fourteen other species, indigen- 
ous and foreign, which I examined, I found essentially the rela- 
tions of corolla and stamens described by Prof. Todd; and in all 
cases, the lower lip of the corolla is provided at its base with two 
longitudinal ridges, including between them a guiding groove 
which leads into the tube of the corolla directly beneath the tip of 
the anthers (Fig. 1), thus forcing a bee to enter every flower in the 
same way. In the thirteen cases mentioned the two lobes of the 
upper lip of the corolla are nearly erect, as if to prevent an insect 
from creeping back on the tube of the corolla and obtaining nec- 
tar surreptitiously through the cleft on its upper side. The 
corolla-tube is also comparatively broad, and an open passage is 
left between its lower border and the filaments (Fig. 3), thus 
giving access to such insects as small bees, which creep bodily 
into the corolla in order to reach the nectar within its base. 
+ In L. cardinalis (Fig. 4) and the Hawaiian Z. macrostachya the 
lower lip Of the corolla is pendant, and there is but a narrow pas- 
sage between the corolla-tube and the stamens, which in these 
species are greatly elongated, with the tip of the anther-tube at a 
considerable distance from the corolla (Fig. 4). The two lobes of 
the upper lip of the latter are often horizontal or even deflexed, 
appearing to serve their purpose best when in these positions. 
The flowers go through the same dichogamy as the others, but 
they-are adapted to profit by the visits of large Lepidoptera or- 
even of humming birds, since a bee cannot enter’ the narrow and 
elongated tube of the corolla, while even if it could it would 
encounter the tip of the anthers or the protruded stigma only by | 
the merest accident. On the other hand a hawk-moth or hum- 
ming-bird could easily poise itself before the flower, and, while 
_ removing nectar by its long slender proboscis or bill, this would 
come’ in contact with the anthers or stigma, since the guiding- 
