TWILIGHT, the name given to the light which remains after the sun has set, or which is seen immediately before it rises. The reason of this appearance is explained in Sus (coL 915), being the effect of the light which is reflected from the higher strata of the atmosphere, in a manner which will be understood from the diagram in the column cited.
In our latitudes, at the summer solstice, a portion of the twilight continues from the setting of the sun to its rising, circling, as the hours of night and morning proceed, from the western horizon through the north, to the east ; which is the cause of there being scarcely any true night at that period of the year.
J. H. Lambert endeavoured to distinguish, besides the primary twilight, a secondary and even a ternary twilight, both the latter being caused by the successive reflection, by the clouds and the air, of light already reflected from other regions of the atmosphere, their clouds, &c. In conformity with these views, Sir John F. W. Herschel has attributed to such a cause the phenomenon, seen in the clear atmo sphere of the Nubian deserts, which has been described by travellers under the name of the " after-glow." To a corresponding cause must be ascribed the rose-coloured illumination of the summits of high mountains after sunset (often witnessed of those of Mont Blanc and Monte Rosa), but in these cases the reflected light is coloured by its traversing tracts of the air of which the vapour is in an opalescent state, imparting to it various hues of red and orange. [VAPOUR.] The astronomer last named, always careful to explain the physical processes which are operative in the production of astronomical pheno mena, in addition to the mathematical principles on which they depend, has pointed out that a portion of the light of the sun and moon reaches us after they are set, by means of the atmosphere, "by reflection upon [and from] the vapours and minute solid particles which float in it, and, perhaps, also on the actual material atoms of the air itself." (See ' Outlines of Astronomy.' ed. 1858, par. 44, 45.) The observations which have been made during the last two or three total solar eclipses ere very instructive, in two respects, concerning the light ordinarily received by the earth, including that of twilight. The character of the darkness, while the totality continues, so much more intense than that of night caused by the mere aversion of the hemi sphere from the sun, evinces how intrinsically dependent upon that luminary for light the earth really is ; while the amount and pecu liarity of the illumination actually existing at the same time, shows in how great a degree we are indebted to the reflective and refractive powers of the atmosphere, and to the reflection by the floating particles alluded to, by aqueous vapour becoming visible but not yet cloud, by the clouds themselves, and by the earth, reciprocally and unitedly, for the light we enjoy after the commencement and during tho progress of that aversion.
But all this is, of course, affected by the degree of transparency of the atmosphere and its difference at different altitudes, the less trans parent, in the mass, so far as we know, being those nearest the earth ; while there are facts which indicate that air perfectly free from aqueous vapour, such as we must conclude it to be above a certain height, is less transparent, or more absorptive of light, than the mingled atmo sphere of air and aqueous vapour incumbent on the earth's surface, when the temperature is such as to sustain the latter in a perfectly gaseous condition. When the sky is thus free from visible vapour and cloud, however, the transparency of the atmosphere is almost invariable, as Professor Seidel has shown.
Another consideration affecting this subject relates to the probable nature of the highest regions of the atmosphere, on which the amount of reflected light causative of twilight must be greatly dependent. If as inferred by Graham and Poisson [SURFACE OF TI1E EARTH, COI. 932], the terminal stratum be solid—air-ice—the reflection from its inferior surface, and from the inferior surfaces of the comparatively dense strata immediately below it, must be more powerful than that from the rare, purely aeriform strata (of which alone the highest regions are commonly supposed to consist), and hence may exert a marked effect in prolonging the twilight, as well as in lessening the darkness during the entire time when the sun is .below the horizon. It may be that this reflection is the principal agent in producing the remarkable amount of light still remaining during a total eclipse of the sun, as noticed above. - What may be the bearing upon this subject of the observed polarising power of the sky during such eclipses has not been investigated, though the materials for such an inquiry, we believe, have been obtained, especially from the eclipse of July 18th, 1860. It seems probable that, the polariscope would furnish the means of determining whether such reflection does in reality take place, and also, if taken in conjunction with the atmospheric refraction of the heavenly bodies, of determining what the structure and constitution of the upper regions of the atmosphere really are.