268 Scott-Elliot. — Ornithophilous Flowers 
parallel to the peduncle. The stamens diverge considerably 
and curve upwards in this condition. Hence in this state the 
end of the leafless peduncle is surrounded by a sort of brush 
of stamens. 
The bird, as it hops along the peduncle and plunges its 
beak into the carina beneath the alae, has its breast well 
dusted with pollen ; but it cannot effect self-fertilisation, as 
(see Fig. 4) the terminal portion of the style is abruptly bent 
towards the vexillum. 
Subsequently both the stamens and styles, losing their up- 
ward curvature, bend still more downwards towards the 
peduncle until the style is so straightened that the stigma 
will touch the bird’s breast when it approaches along the 
peduncle. 
All the Nectariniae (and fide Mr. Scully Zosterops virens) 
are excessively fond of this flower, and apparently it wholly 
depends on them for pollination. Mr. E. S. Galpin tells me 
that ‘the bird puts its beak in the opening on top of the 
staminal column, and runs it along the latter, thus raising the 
stamens/ Unfortunately I have not been able to test this 
interesting observation. 
Erythrina indica, L. 
I was able to study the arrangement in this species in Mauritius. Fundamentally 
the structure is the same. A peculiarity, however, is the peculiar thickening of 
the upper borders of the carina, which renders entrance still more difficult. The 
nectary consists of ten small prominences enclosed by the staminal column. 
Though the stigma is bent over in the same way, I found, in the flowers I examined, 
that it lay below the level of the anthers. Not being in its natural habitat, 
I could not study its fertilisation, but in Mauritius it is certainly ornithophilous. 1 
SUTHERLANDIA FRUTESCENS, R. Br. (Figs. 6-8). 
This flower shows the same excessive length of vexillum 
(15-16 lines) and reduction of alae (5-6 lines) so noticeable in 
Erythrina. In the bud, the flower has the characteristic ap- 
1 In a species of Erythrina described by Belt, 1 Naturalist in Nicaragua,’ 
p. 130, the flower axis seems perfectly straight, though the flowers are ornitho- 
philous. Cf. also E. velutina and E. cristagalli. Hildebrand, Bot. Zeit. xxviii. 
p. 621. 
