My surprise was greatly excited, not only by the minuteness of the particles which caused movement, but how they could possibly act on the glands; for it must be remembered that they were laid with the greatest care on the convex surface of the secretion.  At first I thought—­but, as I now know, erroneously—­that particles of such low specific gravity as those of cork, thread, and paper, would never come into contact with the surfaces of the glands.  The particles cannot act simply by their weight being added to that of the secretion, for small drops of water, many times heavier than the particles, were repeatedly added, and never produced any effect.  Nor does the disturbance of the secretion produce any effect, for long threads were drawn out by a needle, and affixed to some adjoining object, and thus left for hours; but the tentacles remained motionless.

I also carefully removed the secretion from four glands with a sharply pointed piece of blotting-paper, so that they were exposed for a time naked to the air, but this caused no movement; yet these glands were [page 29] in an efficient state, for after 24 hrs. had elapsed, they were tried with bits of meat, and all became quickly inflected.  It then occurred to me that particles floating on the secretion would cast shadows on the glands, which might be sensitive to the interception of the light.  Although this seemed highly improbable, as minute and thin splinters of colourless glass acted powerfully, nevertheless, after it was dark, I put on, by the aid of a single tallow candle, as quickly as possible, particles of cork and glass on the glands of a dozen tentacles, as well as some of meat on other glands, and covered them up so that not a ray of light could enter; but by the next morning, after an interval of 13 hrs., all the particles were carried to the centres of the leaves.

These negative results led me to try many more experiments, by placing particles on the surface of the drops of secretion, observing, as carefully as I could, whether they penetrated it and touched the surface of the glands.  The secretion, from its weight, generally forms a thicker layer on the under than on the upper sides of the glands, whatever may be the position of the tentacles.  Minute bits of dry cork, thread, blotting paper, and coal cinders were tried, such as those previously employed; and I now observed that they absorbed much more of the secretion, in the course of a few minutes, than I should have thought possible; and as they had been laid on the upper surface of the secretion, where it is thinnest, they were often drawn down, after a time, into contact with at least some one point of the gland.  With respect to the minute splinters of glass and particles of hair, I observed that the secretion slowly spread itself a little over their surfaces, by which means they were likewise drawn downwards or sideways, and thus one end, or some minute [page 30] prominence, often came to touch, sooner or later, the gland.