Humboldt endeavours to account for the increase of sounds during thenight, from observing that the presence of the sun affects the propagationand intensity of sound by the obstacles opposed to its transmission by cur-rents of air of different densities and partial undulation—the result of theunequal heating of various parts of the earth’s surface. In air at rest, whe-ther it be dry, or mixed with elastic vapours equally distributed through it,the sonorous undulation is propagated without difficulty. But when thisair is crossed in every direction by small currents of a warmer tempera-ture, the sonorous undulation divides into two waves, at the spot wherethere is the most sudden change in the density of the medium; thus pro-ducing partial echoes, which weaken the body of sound, because one of thesonorous waves is reflected back upon itself. The theory of these partitionsof sonorous waves has been explained by M. Poisson.Ann. de Chimie, t. 7. It is not, therefore, the motion of the passage of the particles of air from below upwards, northe small oblique currents of this fluid that we consider as opposing, by im-pulse, the propagation of the sonorous waves. A stroke or impulse im-pressed on the surface of the liquid will form circles around the impingingcentre, even when the liquid is in agitation. Several kinds of waves maycross in air, as well as in water, without interfering with each other; butthe true cause of the less intensity of sound in the day-time appears to bethe want of homogeneity in the elastic medium. There is at this time asudden change of density throughout, produced by small currents of air, ofa high temperature, rising from portions of the earth’s surface that are un-equally heated. The sonorous waves are then divided in the same manneras luminous rays are refracted, and form a mirage of sound whereverstrata of air of unequal density are contiguous. A distinction must be keptbetween the intensity of sound or of light, and the direction of the sonorousor luminous wave. When these waves are propelled across strata of differ-ent densities two simultaneous effects will be produced—there will be achange in the direction of the wave, and extinction of light or sound. Thereflection that accompanies each refraction weakens the intensity of light;the separation of the sonorous wave causes partial echoes, and that portionwhich returns on itself becomes insensible to our ear, in weak noises, at thespot where the density of the medium suddenly changes. In the mirage with double images, that which has undergone refraction contiguous to theearth is always weaker than the direct image. Strata of fluids, of very dif-ferent density, may so alternate, that the primitive direction of the lumi-nous or the sonorous ray will remain the same, but the intensity of the raywill be not the less weakened on that account. During the night the sur-face of the earth cools; the parts covered with grass, or with sand take thesame temperature: the atmosphere is no longer crossed by currents of hotair, rising obliquely or vertically in every direction. The medium beingnow become more homogeneous, the sonorous wave passes with less difficul-ty, and the intensity of sound increases, as the separations of the sonorouswaves and echoes become less frequent.