304 THE FERTILISATION OF FLOWERS. [PART III, 
9 to 11 mm. long. The plant is andromonecious, and the herma- — 
phrodite flowers are markedly proterandrous (609). f 
Asperula azurea is adapted for diurnal Lepidoptera (590, 111). 
Asperula odorata, L.—In the structure of the flower and the — 
length of the tube this species closely resembles A. cynanchica. — 
The hive-bee is a frequent visitor. Additional visitors (four beetles, — 
four flies, one moth) are enumerated in No. 590, III. 
Sherardia arvensis, L., is gynodicecious (590, IIL, 605). 
Manettia, Mut., is very diligently visited by humming-birds in © 
South Brazil (360). 
Caffea arabica, L., according to Bernouilli,* produces in Guate- 
mala, at the bootinibie of the season, small pistillate fertile flowers. — 
Prof. Ernst was unable to discover these small flowers at Caracas” : 
224 cs 
. een Te species of Borreria, Hedyotis, and Manettia are dimor- 
phic (Fritz Miiller, 353); Hedyotes had already been il to be © 
dimorphic by Treviranus (742). F 
Mitchella, Knoxia, and Cinchona are dimorphic eae to 
Darwin (154, 167). | : 
Chasalia, Nertera, Ophiorrhiza, and Luculia are dimorphic * 
according to Kuhn (399). % 
Other heterostyled Rubiacez are. mentioned by Darwin in his 
Forms of Flowers, where he also discusses the passage from 
heterostyly to diecism in this order. 
Faramea, Aubl.—My brother Fritz Miiller (554) gives an account 
of a dimorphic species of this genus, which is of interest in several — 
respects. (1) It affords the only known instance of a striking dif- 
ference in the character of the surface of the two kinds of pollen- 
grains. The smaller pollen-grains, which are produced in the 
long-styled flowers (whose anthers are inclosed in the tube), are 
smooth ; the larger pollen-grains of the exserted anthers of the 
short-styled flowers are covered with short, rather close points, 
which prevent their being dispersed by a puff of wind. The pollen 
of the long-styled flowers, lying low down in the corolla, is in no 
danger of being blown away. (2) In the short-styled form the 
stamens twist round upon their axes, so as to turn their pollen- 
covered faces outwards. The short stamens (of the long-styled 
flowers) do not rotate but dehisce on their inner sides ; and in both — 
cases the pollen is thus placed in the best position for adhering to 
an insect’s proboscis when inserted in the flower. But it is a 
very remarkable fact that this faculty of rotation is by no means 
1 Bot, Zeitung, 1869, p. 17. 
