The curvature of the cotyledons of Phalaris towards a lateral light is therefore certainly influenced by the [page 461] degree to which they have been previously illuminated.  We shall presently see that the influence of light on their bending continues for a short time after the light has been extinguished.  These facts, as well as that of the curvature not increasing or decreasing in nearly the same ratio with that of the amount of light which they receive, as shown in the trials with the plants before the lamp, all indicate that light acts on them as a stimulus, in somewhat the same manner as on the nervous system of animals, and not in a direct manner on the cells or cell-walls which by their contraction or expansion cause the curvature.

It has already been incidentally shown how slowly the cotyledons of Phalaris bend towards a very dim light; but when they were placed before a bright paraffin lamp their tips were all curved rectangularly towards it in 2 h. 20 m.  The hypocotyls of Solanum lycopersicum had bent in the morning at right angles towards a north-east window.  At 1 P.M. (Oct. 21st) the pot was turned round, so that the seedlings now pointed from the light, but by 5 P.M. they had reversed their curvature and again pointed to the light.  They had thus passed through 180o in 4 h., having in the morning previously passed through about 90o.  But the reversal of the first half of the curvature will have been aided by apogeotropism.  Similar cases were observed with other seedlings, for instance, with those of Sinapis alba.

We attempted to ascertain in how short a time light acted on the cotyledons of Phalaris, but this was difficult on account of their rapid circumnutating movement; moreover, they differ much in sensibility, according to age; nevertheless, some of our observations are worth giving.  Pots with seedlings were [page 462] placed under a microscope provided with an eye-piece micrometer, of which each division equalled 1/500th of an inch (0.051 mm.); and they were at first illuminated by light from a paraffin lamp passing through a solution of bichromate of potassium, which does not induce heliotropism.  Thus the direction in which the cotyledons were circumnutating could be observed independently of any action from the light; and they could be made, by turning round the pots, to circumnutate transversely to the line in which the light would strike them, as soon as the solution was removed.  The fact that the direction of the circumnutating movement might change at any moment, and thus the plant might bend either towards or from the lamp independently of the action of the light, gave an element of uncertainty to the results.  After the solution had been removed, five seedlings which were circumnutating transversely to the line of light, began to move towards it, in 6, 4, 7 1/2, 6, and 9 minutes.  In one of these cases, the apex of the cotyledon crossed five of the divisions of the micrometer (i.e. 1/100th of an inch, or 0.254