A large number of measurements of the absorption of radi ation by metals has been made by Hagen and Rubens. The absorption coefficient, defined by them as the reciprocal of the thickness which reduces the incident radiation to one-tenth of its initial intensity, varies somewhat with the wave-length, but has a value of the order of pp' for gold and silver, for instance, in the visible and near infra-red.
While metals show the same opacity to the visual region as to the long waves, the behaviour of non-conducting or badly con ducting substances in this region is widely different. All such sub stances show regions of marked selective absorption either in the visual region or in the neighbouring ultra-violet or infra-red, the theory of dispersion resting upon the presence of absorption bands in this range. A few of the most important examples will be given. Glass, which transmits the visible radiations so well, is opaque for the infra-red and the ultra-violet except in regions immediately adjoining the visible. The ordinary glass fire screen offers a familiar proof of the opacity for infra-red, and this opacity is also the cause of the high temperature prevailing in glass-houses. The visible radiations enter the glass-house and heat the soil, but the very long heat rays emitted by the soil cannot escape. Quartz is transparent in thicknesses of some centimetres to the infra-red beyond mu , but has a strong absorption from 4.5,u to somewhere about this value. It is also transparent in the ultra-violet down to I,800 A.U. Substances which transmit the infra-red well are sylvin, rock-salt, and fluorite down to about 23, 15 and 91z respectively. The substances transparent to the visible are familiar to every body, as are coloured glasses with selective absorption. The ultra violet is often divided into groups by the absorption of various substances: 3,800-3,400 A.U., glass ultra-violet, transmitted by ordinary glass: 3,400-3,000 A.U., Jena glass ultra-violet, trans mitted by Jena and other glasses specially prepared for the pur pose: 3,000-2,200 A.U., quartz-glass ultra-violet : 2,200-1,800 A.U., quartz crystal ultra-violet : 1,800-.1,200 A.U., fluorspar ultra violet. The absorption of glass both for the ultra-violet and the infra-red is a matter that is, to a certain extent, under the control of the glass maker. Crookes prepared glasses very opaque to the
far red and near infra-red for the use of furnace-men, while there are many special glasses designed both to stop the ultra-violet as near the visible as possible and to transmit it as far into the short wave region as possible. The position of the absorption bands of solids in the visible and the adjacent parts of the spectrum is not determined atomic properties, but molecular properties.
Of particular interest is the absorption of the atmosphere for radiation, as it determines the limits of the sun's radiations which reach us. At the long wave-length end of the spectrum the in tensity of the radiations has been carefully measured by Abbot up to 5.3A where the energy is practically zero, probably owing to a strong absorption band of water vapour, which extends to 6.5,u. There are signs of a feeble transmission of infra-red between ro and 13/1, but it is established that there are no measurable radia tions between wave-length i5ih and 300,u . This is in any case not entirely due to atmospheric absorption, for if the sun be treated as a black body of temperature 6,000° absolute it can be calcu lated that the intensity of radiations of wave-length greater than 19.4 would only be about 3/io,000ths of the total energy. In the infra-red of wave-length less than 6.5/2 there are marked ab sorption bands due to carbon dioxide and water vapour.
At the other end of the spectrum the atmosphere transmits the ultra-violet down to about 3,00o A.U. The limit varies, of course, with the thickness of the atmosphere traversed, and hence i; different in summer and in winter. Measurements made in Switzer land by Dorno show that according to the season and the time of day the shortest wave-length of the sun's spectrum present in measurable intensity varies between 2,976 and 3,197 A.U. It may be noted that it is quite a narrow range of wave-lengths from about 2,800 to 3,100 A.U., with a sharp maximum at about 3,000m that can produce sunburn (light erythema), which accounts for the great variation in the burn produced by strong sunlight at different times of year, for the greatest action lies just in the region about which the limit of wave-length transmitted by the atmosphere oscillates with the seasons.