18 
and restitution, being much the same as in Arbutus, I may illustrate 
by reference to figures of that genus. Fig. i illustrates a mature 
stamen of Arbutus Menziesii, The anther is erect, introrse both as to 
position and pores, attached very near its upper extremity, and fur- 
nished with a pair of curved horns projecting backward, the concavity 
of the curve looking directly upward. These horns seem to have their 
origin at, or very near, the upper edge of the pore, and to be adnate 
to an extent which varies in different genera and species, and in the 
same species at different periods. If now, we imagine this anther 
tipped over backward until it is exactly upside down, we shall have 
a part of what we find in Fig. 2, an illustration of a stamen from a 
young bud of the same species. That is, the anther would be reverted, 
and extrorse as to both position and pores, and the filament folded up- 
on itself. But the horns would assume a different position, for they 
would project inward, the concavity of the curve looking directly 
downward. Such, however, is not the case in the bud. The filament 
is bordered by thin, but strong, transparent wings, and to this the horns 
Fig. 5. Fig. 6. Fig. 7- 
■ 
are firmly attached,apparently imbedded in its substance. The filament 
IS at this time very short. From these observations, the design of the 
arrangement, and the use of the horns is very apparent. As the fila- 
ment lengthens in its narrow quarters it becomes variously curved 
and cramped, so that at the first intimation of freedom, from the 
spreadmg of the corolla, which it probably hastens, it begins to 
straighten by pressing the anther in the only direction possible, 
namely, the apex (lower end) outward. It is evident that powerful 
assistance must be rendered by the elasticity of the growing horns, 
pressed, as they are, far out of their normal line. (See also Figs. 3 
and 4.) 
In Arclosfaphylos the principle is the same, with slight differences 
in the details. It must be remarked that in this genus especially, the 
development and growth of the horns is very rapid. In the youngest 
buds that can be readily examined, the horns are scarcely discernible- 
The rapidity of their subsequent growth is well seen by comparing 
their length in the five precocious, with that in the other five anthers 
(best seen in Arctostaphylos Uva-ursi), for in all of these changes, five 
