1882.] 
425 
[Trelease. 
The broadly tubular form of the flowers, and their copious nec- 
tar, are indications of adaptaion to birds. A native of the east- 
ern continent, the plant is removed from the possibility of fertil- 
ization by humming-birds. The latter are replaced by the sun 
birds or ISTectarinidae, many species of which are quite as active 
about flowers as the Trochilidae of our own continent ; and it is 
known that these tubular heaths are very attractive to them, a 
fact considered by Delpino in referring some of them to his micro- 
stomous type of flowers, adapted to pollination by birds (2, 
251 ). 
To discover the probable mode of pollination, I took a slender 
probe and inserted it in the open space below the stamens of a 
flower, bringing it in contact with the lower filament and so break- 
ing the connection of the anthers, when some of the pollen fell 
upon it. In other experiments the probe was pressed against the 
anther with a like result. On one occasion, hapjiening to use a 
leadpencil instead of the probe, I discovered what appears to be 
the usual action when a bird visits the flowers. When the broad 
surface of the pencil was pressed against the anthers, instead of the 
almost impassive breaking of their connection, and the resultant 
sprinkle of pollen, there was a sort of explosion, the separation 
beginning between the lower pair of anthers and extending 
upward, throwing a large quantity of pollen upon the pencil. 
Other experiments led to the conclusion that in visiting these 
flowers a bird will most commonly insert its bill in the position 
occupied by the probe, without pressing upon the stamens with 
sufficient force to cause their separation. Its bill being pushed 
in for its full length, its foreh.ead comes in contact with the lower 
anthers, the broad surface acting very much as the pencil did, 
exploding the anther tube, and causing a deposit of pollen among 
the small feathers of the front, whence it would be brushed by 
the viscid stigma of older flowers. There is here a case some- 
what similar to that of Posoqueria described by Fritz Muller and 
Charles Wright. The connate anthers forcibly bend and detain 
the elastic filaments until the contact of a visiting animal aids the 
latter to escape, when, the arch being broken, they rapidly sep- 
arate. In this case, however, the explosion is less violent ; and 
yet the specialization may be more perfect, for the shower of 
