DR. SPRUCE ON THE FECTINDATION OE GRASSES. 
7 
It is generiill}^ to be remarked of diclinous grasses, that either 
the male flowers are very numerous, as in Zea Mays, or the 
stamens are multiplied in each male flower, as in Pariana, Leersia^ 
Guadua, &c., or the stigmatic apparatus of the female flowers 
is enlarged, so as almost to ensure impregnation, as in Olyra and 
TripsaciLrn. 
In the Bambusese I have gathered, belonging to the genera 
Ouadua, Merostachys, and Gliusquea, the flowers are more or less 
polygamous, and the stamens of the male flowers often doubled. 
But there is scarcely a genus in the whole order which is not 
described as having some flowers, " by abortion," neuter or male, 
and especially those that have biflorous spikelets, such as the 
Panicese. Some grasses, of normally hermaphrodite genera, are 
not unfrequently truly unisexual, such as a certain species of 
Andropogon. I have occasionally seen panicles of Orthocladus 
rariflorus (Nees), a grass peculiar to the Amazon, quite destitute 
of stamens, and therefore purely female. 
To come home to our own country : — is all the pollen wasted 
that a touch or a breath sets free from the flowers of grasses in 
such abundance ? Watch a field of v/heat in bloom, the heads 
swayed by the wind lovingly kissing each other, and doubtless 
stealing and giving pollen. Consider, too, that throughout na- 
ture, heat or motion, or both, are essential to the emanation of 
of the impregnating influence. In all our Festucese, as well as 
in Cynodon, Leersia, and some other genera, the stigmas are 
protruded from the side or from the base of the flower at an 
early stage, often before the stamens of the same flower are ma- 
ture, thus, as it were, inviting cross-fertilization from the more 
precocious stamens of other plants which are already shedding 
their pollen. 
All who have gathered grasses will have remarked that some 
have yellow anthers, others pink, purple, or violet anthers, and 
that anthers of both types of colour may coexist on distinct indi- 
viduals of the same species. The same peculiarity is just as 
noticeable in tropical grasses ; and (without professing to give a 
complete physiological explanation of it) this is what I have ob- 
served respecting it. The walls of the anther-cells are usually 
of some shade of purple, but are so very thin and pellucid that 
when distended with mature pollen, the yellow colour of the 
latter is alone visible. When the pollen is discharged, the an- 
thers resume their original purple colour, shortly, however, to 
