In explanation of this experiment, M. Hachette says, “The air is pushed from the mouth A of the tube, towards the orifice E of the plate; it strikes the part of the disk opposed to this orifice, and the mean pressure on that part is greater than the pressure of the atmosphere. The blown air then takes place of that between the plate and the disk opposed to it; it moves in this interval with a velocity decreasing from the edges of the aperture: the elastic force of this air decreases at the same time, so that its mean pressure between the plate and the inner face of the disk becomes less than the atmospheric pressure; and as this last pressure is exerted on the whole external face of the disk H, I, this disk, subject at the same time to the two contrary pressures on its opposing faces, obeys the greater, and is pushed towards the plate C D.”
“It is not necessary that the disk, C D, should be near the orifice E, of the tube A E. Let Fig. 2 be an instrument composed of a hollow cylinder, C D F G, and a flat border of the dimensions C″ F, or G D″. Let a tube, A E, be fixed to the bottom of the cylinder, the orifice E having a diameter of about three millimeters (0.12 of inch). If air be blown in at A, against the disk, H I, in the neighbourhood of the flat border, the disk will be urged towards the orifice E. This instrument is also delineated on a scale of one half. The disk, with the attached weight, weighs about 12 grammes (184.87 grains), being 54 millimeters in diameter; the pressure of the atmosphere upon it equals 23 kilogrammes: from which it follows that, in this experiment, the pressure of the air blown upon the inner surface of the disk, and the atmospheric pressure exerted on the exterior of the same disk, only differs from each other by about one two-thousandth part of the latter.”—Annales de Chimie, xxxv. 34.
[34] See the last volume of this Journal, p. 473.
2. Considerations relative to Capillary Action, by M. Poisson.
Suppose that two different fluids, A, B, are contained in a vessel, and separated the one from the other by a vertical division; the heights being in an inverse ratio to the densities, so that the points, a and b, in the two faces of the division, and situated in the same horizontal plane, shall support equal and opposite pressures: suppose also that the division is pierced with one or more holes of small diameter, or, in other words, that it is traversed by several very narrow canals, as a, b, perpendicular to the two faces, and which may be regarded at first as filled with air, or any other fluid.
If the substance of the division exerts upon each of the two liquids an action superior to the half of that which the liquid has upon itself, each liquid will enter into the canal a, b, just as it would rise above its ordinary level in a capillary tube of the same size and substance. It would also be urged, by the excess of pressure which it would exert at the extremity of the canal, against the elasticity of the included air. When the two fluids have penetrated the interior of a, b, the air will be pushed on both sides in different directions by forces each of which is equal to the primitive pressure augmented by the corresponding capillary force, i. e. augmented by forces proportional, according to the known theory of M. Laplace, to double the action of the tube on the liquid, less the proper action of the liquid itself. It will only be in the case when the capillary force shall be the same on both sides, that the air, after being compressed to a certain degree, will remain at rest: for whenever this force preponderates at one end of the canal, the air will be driven out at the opposite end, and the liquid with the strongest capillary attraction will entirely fill the canal.
Suppose this liquid to be A, then let us consider the forces which will act on the portion a, b, of this liquid. At the extremity a, it will be submitted to the attraction of the exterior fluid A: at the extremity b, it will be attracted in the opposite direction by the liquid B. Now the two liquids being different, their attractions will be unequal, and we will suppose that that of B, on the matter [p196] of A, is greater than that of A for itself. As to the action of the canal on the portion a, b, that will be equal, and exerted in contrary directions at its two extremities; it will not, therefore, be either adverse or favourable to the movement of the fluid in the canal: and the same will be the case with respect to the pressures exerted at a and b, by the external liquids, as long as they are equal: nevertheless, the action of the canal, and the external pressures, will prevent the thread of fluid from being broken, so that it will move without interruption in the direction in which it is drawn by the greatest attraction, or from a to b. Hence will result an elevation of the level of B, and, consequently, an increase of pressure at the extremity b, of the passage, and this elevation will proceed until the difference of pressure in a and b shall be equal to that of the attractions exerted by the two fluids A and B, on the thread a b; this effect will be produced the more rapidly as the division is pierced with a greater number of passages similar to that which has been considered.
Now let us examine what would occur if the division were formed of two others different in their nature, and exactly superposed; exerting no action on one of the liquids, B for example, and one only acting on the other liquid. The liquid B will then retain its original position undisturbed; in consequence of the action it exerts upon itself it cannot penetrate the canal a b, just as mercury cannot escape by a capillary aperture made in a barometer-tube. It will be the same with A, when that face of the division which exerts no action upon the liquid is turned towards it; so that how numerous soever the apertures, the two liquids would, under such circumstances, remain separate and preserve their original level. But if the division be turned so that the face which acts upon A shall be in contact with that liquid, it will penetrate the canal a b by means of capillary attraction; and the velocity which the liquid urged by this force may acquire, may make it pass that point in the canal where the division changes its nature, and even make it reach the extremity in the liquid B, so that it is possible that the liquid A should entirely fill the canal a b, as in the case which has already been examined. Then if we always suppose the attraction of B for A to be superior to that which A has for itself, the thread a b will flow into B until the level of the latter is so far altered that the excess of pressure at b can balance the difference of attractions exerted by the two liquids at a and b.
M. Poisson then observes that, without pretending to assign a cause, exclusive of all others, for the phenomena of absorption by vegetable and animal membranes observed by M. Dutrochet, his object is to show that effects which have at least a great resemblance to these important phenomena, may be produced by capillary action conjoined with the difference of affinity existing between heterogeneous substances without the assistance of electricity, either moving or quiescent. It appears that M. Dutrochet afterwards [p197] found mineral substances, as a piece of slate, might be substituted for the organized tissues; this being the case, the opinion which refers such effects to a general cause, as capillary attraction, acquires more probability.