CROSS-FERTILIZATION IN ROSCOEA PURPUREA. 205 
structure of Roscoea I may be better able to describe with the help 
of the diagram. The flower has one stamen ; 
the anther is seated on a stout filament, and is 
provided with two spurs which project over 
the tube of the flower The style is extremely 
slender and flexible every way; it comes up 
the tube, passes between the spurs, then runs 
between the anther-cells, and produces the 
stigma at the top of the anther. It must be 
observed that the anther swings on its filament 
when the spurs are touched. The result of this 
structure is, that an insect coming into the 
flower depresses these spurs, and this action 
necessarily brings the anther and stigma to 
the position I have here marked B in figure. 
or : E : Diagram flower of 
While in this position the stigma takes up Roscoea purpurea. A, 
pollen from the insect which has come from With anther and pistil 
: 8 in matural position; 
another flower; and the insect also receives and p, with dotted out- 
a further supply of pollen from the anther lineas bent onan insect 
over its back. I have referred to Salvia; and yaam 
should now point out that Roscoea differs in bringing down by its 
lever action, not only the anther, but also the stigma; while in 
Salvia there is no connexion between them, and the style either 
remains stationary, or is moved by another and distinct impulse. 
In Salvia there are two fertile stamens, and of each the anther- 
connectives unite to form one lever which swings between the two 
filaments; its structure, however, being well known, it is hardly 
within my province to recapitulate. The anthers differ greatly in 
structure; but there is the lever-process in Salvia which nearly 
corresponds with the lever-process of Zoscoea. 
The attention of the Fellows at the Meeting is directed to a 
drawing of Roscoea, from a specimen in the Cambridge Botanic 
Garden which flowered during the past summer; and a flower in 
spirit is also exhibited, which elucidates the structures now under 
consideration. 
There is one remark I should like to make with regard to this 
Salvia, which differs from most other species. If not typical 
S. Grahami,itis a slight variety. The flower is quite closed, and 
seems to form a trap. The only way in is by pressing back the 
anther-connectives, which form a valve over the entrance. Any 
insect getting behind is prevented from returning the same way. 
