162 THE EXTREME INFRA-RED RADIATIONS. 



nickel, blackened, and introduced as a resistance to form part of an 

 electric circuit; but sometimes it is a very minute thermopile caused to 

 affect a highly sensitive galvanometer. 



The production of extreme infra-red rays is very simple, as all bodies 

 whatsoever emit them constantly. If, however, an opaque body is at 

 the same temperature as the receiver, the radiations which it sends out 

 are imperceptible, because the two bodies exchange equal quantities of 

 energy. 



The hotter a body, the more it emits radiations of all kinds. But 

 the difficulty is to isolate separate bundles of these rajs from their 

 neighbors, so as to deal with rays of approximately the same wave 

 length. Prisms of most substances are useless for these long wave 

 lengths, as they generally absorb the rays totally, just as lampblack 

 does in the visible spectrum, while the grating, as is well known, super- 

 poses many orders of spectra having wave lengths which are multiples 

 of each other. 



A very ingenious device originated by the American physicist, jS'ich- 

 ols, has been of very great service here. In studying the nature of the 

 radiations reflected by quartz, Nichols found that the fraction reflected 

 was very slight through the greater part of the spectrum, but increased 

 very greatly at certain regions in the infra-red, where, indeed, this sub- 

 stance seemed to have a really metallic reflection. These regions of 

 nearly total reflection were tolerably narrow and well marked. At each 

 successive reflection from quartz surfaces all the radiations become 

 more and more enfeebled by absorption in the quartz, with the excep- 

 tion of those whose wave lengths correspond to such regions of metallic 

 reflection. These bands thus become relatively stronger and stronger 

 at each reflection till, after four or five reflections, they alone remain 

 perceptible. 



By interposing a grating in the path of the beam the wave lengths 

 under investigation may be determined. The apparatus for this pur- 

 pose is, however, quite different from that employed in the visible spec- 

 trum. For example, the grating is composed of metallic wires, each one 

 or two tenths of a millimeter in diameter, and the slit through which 

 the radiations pass may be as much as a centimeter wide without making 

 the spectrum too impure. It is not necessary that the reflecting sur- 

 faces shall be polished with that degree of accuracy and finish required 

 in the visible spectrum. For the proportion of radiations diffusely 

 reflected by a surface depends not only on the size of the grains of the 

 surface, but on the wave lengths of the radiations reflected. Infra-red 

 measurements now frequently deal with radiations of a wave length a 

 hundred times as great as that of green light, and thus the optical sur- 

 faces may be much inferior to those which would be required for work 

 in the visible spectrum. 



Quartz is not alone in possessing this property of selective reflection 

 of which we have been speaking. A great number of other crystalline 



