CHAP. II. 
POLLEN. 
155 
the whole cavity of a cell of the anther, as in Asclepiadeas. 
But the most curious instances of the cohesion of the grains 
of pollen is to be found in Orchideae ; in which some genera 
have the pollen in its common pulverulent state, with no 
remains of the cellular substance in which it was developed ; 
others have the granules held together by some of the cellular 
substance in an elastic state, and forming a distinct appendage 
to the pollen called the caudicula; while others have the grains 
united either by threes or fours, or in wedge-shaped masses, 
or in a hard, dry, solid body. It appears from Mr. Fran- 
cis Bauer's observations, that the masses of pollen of both 
Asclepiadeae and Orchideae, in the most solid state are really 
cellular, the grains of pollen being contained in cavities, the 
walls of which are either separable from each other as in some 
Orchideae, or are ruptured without a separation of the cavities 
as in Asclepiadeae. (See the Observations on Orchidece and 
Asclepiadece before referred to.) And this is quite in con- 
formity with MirbeFs account of the origin of the pollen in 
the Gourd. 
The granules are generally discharged at once, upon the 
dehiscence of the anther, or at least are at that time wholly 
formed. But in some Aroideae, which emit their pollen by a 
hole in the apex of their anther, the formation or develope- 
ment of pollen must be going on for a considerable time after 
the first emission. A single anther continues to secrete and 
discharge pollen, till, as Brown remarks, the whole quan- 
tity produced greatly exceeds the size of the secreting organ. 
The surface of the pollen is commonly smooth. In some 
plants it is hispid, as in the Gourd and Ipomaea purpurea ; in 
others it is covered with strong points, as Hibiscus syriacus ; 
and in all cases, when there are asperites of the surface or 
angles in its outline, it is asserted by Guillemin to have a 
mucous surface, which was first observed in Proteaceae by 
Brown. 
The figure of the granules is various ; most frequently it is 
spherical or slightly oblong. Many other forms have, how- 
ever, been described. The cylindrical exists in Anethum 
segetum, and in a very remarkable degree in Tradescantia 
virginica, where the grains become curved. In Colutea arbor- 
