1831. I assumed caloric to exist throughout the sublunary creation, as the luminiferous ether is assumed to be diffused throughout all space by the undulationists; the diffusion arising from the reciprocal repulsion of its particles being similar to that which had been supposed to cause the diffusion of caloric. There is the greatest analogy between this diffusion and that which is known to exist in the case of gases. The process is the same, whether the gas be dense like chlorine, or thirty-six times as rare, as in the instance of hydrogen, and in the luminiferous ether resembles the process by which hydrogen is rarified, or might be rendered more rare, were the pressure of the atmosphere removed.

1832. It is known that in any gas or gaseous mixture like that which we breathe, if a deficit of pressure be caused in any spot, the gaseous particles will quickly move toward it, in order to restore the equilibrium of pressure, and that if, on the other hand, any augmentation of pressure be produced at any spot, the gas will move outward to restore the equilibrium.

1833. The particles being symmetrically arranged in lines, a row of particles may be conceived to lie between every two remote points. If we suppose any number of points in the focal body, and a corresponding number in the surface of the mirror, it may be conceived that the intervening ethereal or calorific particles will move in rows one way or the other, as the pressure in the focal space may become greater or less. Thus an effect is brought about, equivalent to that which the Newtonian idea of radiation involves; lines of particles proceed from the hotter points to the colder ones.

1834. The arrangement of the particles of caloric, which was originally, in my view, confined to the sublunary creation, appears of necessity to belong to the luminiferous ether, required by the theory ascribing light to undulations, though the last-mentioned medium must be endowed with ubiquity as above stated, so as to abound in every part of space through which light reaches the eye.

1835. The undulatory hypothesis supposes that a wave-like motion being imparted to a row of particles, by a luminous point in the surface of the luminous body, is transmitted, like the sound producing waves in the air, to the other end of the row.

1836. This undulatory progression has been roughly illustrated by the transitory serpentine movements which may be made in a cord, stretched like a clothes-line between the tops of posts.

1837. In order to make this illustration elucidate the conception which I advance, we have only to suppose that the cord, instead of being attached to the post, should be drawn rapidly over pulleys, and, while thus actuated, be subjected to a cause of undulatory vibration. It may be conceived that, by this process, the ethereal particles, while performing all which the undulatory theory requires, might at the same time perform all required by that of emission and material calorific radiation. Directed upon a vaporizable liquid, the undulations might perform the part of sensible heat; the ethereal particles, successively combining, might furnish the latent heat requisite to the constitution of vapour.

1838. Agreeably to Newton, the seven colours of the spectrum are due to as many different kinds of radiant particles of various refrangibility, or susceptibility of being bent from the rectilinear path when passed through the same refracting medium.[40]

1839. According to the undulatory theory, the colours are caused by diversities in the undulations producing them. Retaining this feature, the last-mentioned hypothesis, as modified by myself, appears to be competent to explain the phenomena of light as well as those of vaporization, produced by calorific radiation, since not only is any vaporizing liquid subjected to the transient effect of the undulations, but also may combine with the ethereal particles as they come into contact with it.

1840. Thus modified, the rationale of the rainbow, or prismatic spectrum, would not be that the colours indicate as many varieties of original radiant particles, but that they are to be explained agreeably to the undulatory hypothesis, which ascribes them to as many varieties in the undulations, just as the notes in music are ascribed to diversities of vibration.