PROFESSOR MACAIRE ON THE DIRECTION ASSUMED BY PLANTS, 
267 
upper surface by a little gum or a pin. In that case, all the marginal parts of the 
flat part of the leaf separate themselves from the screen, bend downwards, and the 
leaf takes a globular form round the point tied to the screen. It succeeds in that 
manner to expose a part at least of its upper surface to the diffused light below the 
shade of the screen. When the same experiment is repeated with a mirror placed so 
as to cast light upon the under surface of the screened leaf, the bending downwards 
of the marginal portions of the flat part is still more marked and rapid ; the mirror 
giving a stronger light to those parts of the upper surface thus brought to receive it. 
When the screen is large enough to allow of no light reaching the upper surface of 
the leaf, even after the bending of its flat part in a globular form, and no mirror is 
used to supply the deficiency, the leaf turns yellow and dies. 
Light, then, is the direct and indispensable agent in the turning over of inverted 
leaves, and causing them to assume their natural direction. Its influence is the more 
rapid, all other circumstances being alike, the greater the difference between the two 
surfaces of the leaf experimented upon. Thus the leaves of French beans, or of rasp- 
berry-bushes, where the two surfaces are of very different colours, turn over com- 
pletely when inverted, and frequently in less than two hours. In the lilac, where the 
two surfaces of the leaves are very similar to one another, the turning over is slow 
and does not seem complete ; the flat part of the inverted leaf assuming often a 
spiral form. 
Bonnet and Dutrochet agree in admitting that the turning over of leaves always 
takes place by a flexion or torsion of the footstalk. The last-named author has even 
said that a footstalk alone could raise itself erect by the action of light as well as if 
it were part of a whole leaf. My own observations have convinced me that the flat 
part of the leaf, or even a separate portion of it, can turn itself over. 
In some leaves, for instance in those of the lilac, the Polemonium coeruleum, it is 
always in the flat part alone of the leaf that the motion takes place, and the leaf turns 
itself into a spiral form to come back by degrees to a regular position. In others, on 
the contrary, the motion takes place in the footstalk. This is the case in the gera- 
nium, the French bean, the raspberry-bush, the horse-chestnut, the plane-tree, the 
judas-tree, &c. In this latter I have seen the turning over of the leaf take place 
without any change of position in the footstalk, but by a sort of rotation of the 
flat part on that portion of the footstalk where it is inserted, and where there is 
a natural swelling, which swelling was much increased by the change. The footstalk 
then formed an obtuse angle with the flat part of the leaf, instead of a very acute 
angle as usual. 
In this experiment, as in all the others, the branches had been inflected and tied in 
a convenient position for the light, but without touching the footstalks, in order to 
avoid the reaction of torsion. 
A vase of geranium, the leaves of which had all directed their upper surfaces towards 
a window, had been reversed so as to present to the window the under surface of its 
2 N 
MDCCCXLVIir. 
