parts as long as its temperature is uniform, the motion, if it exist, must be a vibratory or undulatory motion, or a motion of particles round their axes, or a motion of particles round each other.' Again, it seems possible to account for all the phenomena of 'heat, if it be supposed that in solids the particles are in a state of vibratory motion, the particles of the hottest moving with the greatest velocity, and through the greatest space. That in liquids and elastic fluids, besides the vibratory motion, which must be conceived greatest in the last, the particles have a motion round their own axes, with different velocities, the particles of elastic fluids moving with the greatest quickness; and that in ethereal substances the particles move round their own axes with different velocities, and separate from each other, penetrating in right lines through space. Temperature may be conceived to depend upon the velocities of the vibrations; increase of capacity on the motion being performed in greater space; and the diminution of temperature during the conversion of solids into fluids or gases, may be explained on the idea of the loss of vibratory motion in consequence of the revolution of particles round their axes, at the moment when the body becomes liquid or uniform; or from the loss of rapidity of vibration in consequence of the motion of the particles through greater space.'

CHAPTER X.

SECTION I.

GENERAL PROPERTIES OF LIGHT-PROPAGATION-INTENSITY

SHADOWS-VELOCITY.

189. The agent whose action on the eye originates the sensation of seeing, by which we are made conscious of the presence of external bodies, is at once admitted to differ most essentially from the ponderable matter of the globe. The cause or agent producing the effects which we refer to heat, was considered independent of ponderable matter, because the characteristic property of matter, or its weight, is never affected by it. Similarly, all the phenomena of seeing are referred to an imponderable agent, which is called light. Respecting the nature of this agent we shall at present venture no hypothesis, but proceed at once to the laws of the phenomena.

Light, whatever be its nature, is propagated in every direction; the simplest considerations will satisfy us of the truth of this law. The flame of a candle is visible from all parts of a room; the electric spark, or any phosphorescent, can be seen in all directions. This, which our experiments

can only shew on the small scale, is exhibited in all its grandeur in the expanse of the heavens; for the sun disperses in all directions through all space the same brilliancy, lighting up the earth, the planets, and the comets, and shining on all bodies, whatever be their position in the infinity of space.

All luminous bodies are composed of ponderable matter; we see that without ponderable matter there can be no light; it may be propagated or transmitted without ponderable matter, but it cannot be originated. Hence luminous bodies may be divided into ponderable fragments, more or less minute, and the smallest ultimate particles which we can conceive are called luminous points. Thus, as a body is a collection of molecules or atoms, a luminous body is a collection or combination of luminous atoms or points. Each luminous point propagates its light in all directions, as we have already seen; but from a luminous body, as from a red-hot ball, we only receive the light from the surface; the light from the internal particles is absorbed or stifled in some way by the matter of which the body is composed; the same is the case with the light which is propagated towards the interior from the points at the surface. Thus does light expand itself in every direction; but as we shall see hereafter it is not propagated under all circumstances with equal facility in every direction. Those bodies which are not in themselves luminous shine with borrowed or reflected light, the laws of which we shall presently consider.

190. Direction of Propagation. In a homogeneous medium, that is, in one which is absolutely or sensibly of the same density throughout, light is propagated in straight lines. The simplest experiments will shew the truth of this law. If any number of plates or flat pieces of any substance have a small hole pierced in them, a candle or any luminous point will be seen through them if their centres are all in the same straight line. If the centre of any hole be the least out of the line the luminous body will be obscured. Whenever light meets any surface, polished or unpolished, its direction is changed abruptly, as we shall see in speaking of reflexion and refraction, but after this abrupt change its new direction is rectilinear, provided the medium be homogeneous.

If the medium be not homogeneous, but heterogeneous,

that is, if it be of different densities at different points, the propagation will no longer take place in straight lines, its direction will be curvilinear. Thus the light of the sun or a star does not come to us in a straight line; the atmosphere, as we have seen (Art. 98), is a heterogeneous medium composed of successive strata, each differing in density from the preceding; hence it follows, that a heavenly body is never exactly in the place in which it appears to be. Our observations on all bodies at great distances are attended with the same illusions. When the distances are small, the deviation of the direction of propagation from a straight line is, on account of the smallness of the curvature, insensible; just as the deviation of the surface of still water from a plane is not sensible. But since the atmosphere is composed of layers of different densities, light must suffer this deviation in passing through. The change which takes place will be readily understood by considering the change which takes place when light passes from air to water, or from water to air; here it traverses two media of very different densities, and the deviation is very striking. Let any object, as a shilling, be placed in any vessel, and let the farthest point of the object be visible by an eye at E just over the edge of the vessel in the direction E C D. Then if water be poured into the vessel the whole shilling will come into view, though it is still really concealed by the vessel. Here the light is propagated in a straight line in the water, and in a straight line in the air, for each medium may be considered as homogeneous through this small extent, and it will be seen immediately that the light follows a broken line, as D F E. The same may be rendered at once apparent by placing a stick in a vessel of water, and looking down it; the stick will appear broken at the surface of the water, and the portion im

C

D

mersed will only be visible by placing the eye below the outer end of the stick. The light which comes from the heavenly bodies does not experience any abrupt breaks, as when the density of the media changes abruptly, but being constantly bent by insensible degrees has a curvilinear instead of a rectilinear path.

191. Ray and Pencil of Light.-Any line drawn from the luminous body along which light is propagated may be considered as a ray of light, and several such lines, or a collection of rays, constitute a pencil. When light is allowed to pass through a very small hole it appears as a single ray. It is, however, in reality a small pencil, since a single ray of light, like a mathematical line, has no actual existence; but, being by definition invested with certain properties, serves as the basis of our reasonings. If from any luminous point straight lines be conceived, drawn in all directions, these will represent the rays for all small distances. When light is propagated in a homogeneous medium round any point, and received on any surface, we say that any small portion of this surface is illuminated by a pencil of rays. This portion of the surface is considered as the base of a cone, whose summit is the luminous point, and the pencil of light is the light comprised in this cone; but when the medium is very heterogeneous, this cone can no longer be considered as having any existence. A pencil of light is naturally divergent, that is, its section increases as we recede from the luminous point; when, however, the luminous point is. at a great distance, the pencil may be considered as parallel, since all the sections are sensibly equal, or all the rays sensibly parallel. Thus, for example, the light from the sun constitutes parallel pencils, since whatever portion of surface we consider, lines drawn from it to the centre of the sun will be parallel. Pencils of light may also be convergent, that is, the rays may be so directed that they will meet in a point. This point of concourse for all