Suppose, for example, we have some waves of wave-length DE (in Fig. 25), some of wave-length one-half {79} DE and some of one-third DE. Then in the direction DK we shall get plane waves of each of these wave-lengths setting out and being brought to a focus in the same place. This difficulty can be fairly simply surmounted where the measurement of wave-length alone is required, by placing in the path of the rays from the source of light, suitable absorbing screens, which will only allow a very small range of wave-lengths to pass through them. There will then be no overlapping and no confusion. Where the actual distribution of energy in the spectrum of any source of heat is to be determined the difficulty becomes more serious, and probably there is some error in the determinations, especially in the longest waves, which are masked almost completely by the overlapping shorter waves. +Rest-Strahlen or Residual Rays.+--A very beautiful method of isolating very long heat waves, and so freeing them from the masking effect of the shorter waves, was devised by Rubens and Nichols. It is found that when a substance very strongly absorbs any waves that pass through it, it also strongly reflects at its surface the same waves. For example, a sheet of glass used as a fire-screen will cut off most of the heat coming from the fire, although it is perfectly transparent to the light. If, now, it is placed so as to reflect the light and heat from the fire, it is found to reflect very little light but a very large proportion of the heat. Some substances have a well-defined absorption band, _i.e._ they absorb a particular wave-length very strongly, and these substances will therefore reflect {80} this same wave-length strongly. If instead of a single reflection a number of successive reflections be arranged, at each reflection the proportion of the strongly reflected wave-length is increased until ultimately there is practically only this one wave-length present. It can therefore be very easily measured. These waves resulting from a number of successive reflections, rest-strahlen or residual rays as they have been named, have been very largely used for investigating long waves. Quartz gives rest-strahlen of length .00085 centimetres and very feeble ones of .0020 centimetres long. Sylvite gives the longest rays yet isolated, the wave-length being .006 centimetres. +Range of the Waves.+--The lengths of the waves thus far measured are:-- Schumann waves . . . . . . . . .00001 to .00002 cms. Ultraviolet . . . . . . . . . .00002 to .00004 " Violet . . . . . . . . . . . . .00004 " Green . . . . . . . . . . . . .00005 " Red . . . . . . . . . . . . . .00006 to .000075 " Infra-red . . . . . . . . . . .000075 to about .0001 " Rest-strahlen from quartz . . .00085 and .0020 " Rest-strahlen from Sylvite . . .0060 " Thus the longest waves are six hundred times the length of the shortest.