heat, but no luminous or actinic rays. As the temperature rises, we have along with a preponderance of dark rays, a few luminous ones of the less refrangible sort, such as the red. As the temperature still continues to rise we have, in addition to the dark rays, a similar proportion of the various rays of the visible spectrum, and a still smailer proportion of the chemical invisible rays which lie to the right.

299. In Fig. 99 we have a representation of the sun's visible spectrum, showing the comparative luminosity at different parts, while in Fig. 100 we have the same spectrum as given by a rock-salt prism, showing the heating effect of the various rays. We see from this that while there is most luminous effect, as far as the eye is concerned, about the yellow, yet there is greatest heating effect or true energy of radiation beyond the visible spectrum to the left, while the chemical rays represent in intensity but a very small fraction of the whole effect.

In fact, the eye is a very partial judge of the energy represented by a ray, for it is necessary that the rays of which it judges should be able to penetrate to the retina; but it is questionable whether some of the rays of small refrangibility are able to pierce the eye at all.

Another noteworthy point is, that the spectrum of carbon is a continuous one. Thus we get, at a sufficiently high temperature, a continuous band of light,--that is to say, the spectrum does not stop short and commence again, but goes

on without interruption; and in this respect carbon is a representative of liquid and solid bodies of which the spectra are generally continuous, like that of carbon.

300. The spectra of gases are very different from those of solid bodies, for they are discontinuous, consisting of one or more bright lines on a dark ground.

Thus the spectrum of ignited sodium vapour (see frontispiece) consists of two bright yellow rays very near one another in spectral position, forming what is called the double line D. In like manner the spectrum of thallium consists almost entirely of one intensely blue line. The light from burning sodium is as nearly as possible mono-chromatic, and all things seen by its light are either black or yellow, for a coloured body will not appear in its true colour unless those colours are present in the light by which it is viewed.

It forms a striking proof of this to put a little bit of metallic sodium into an iron spoon, and heat it over a spirit-flame in a dark room, when it will soon take fire and burn, and everything in the room will either appear black or of a ghastly yellow colour.

We are enabled by means of electricity to obtain even the most refrangible substances in a state of vapour: thus, for instance, when the electric spark passes from iron into the air, the flash seen consists of a few particles of highly-heated iron vapour; and the same holds for other metals. Now we can analyse these sparks by means of the spectroscope, and thus tell the nature of the light which they emit, and we find that they all give a discontinuous spectrum.

In like manner the spectra of the elementary gases are discontinuous, from which we see that bodies in a state of vapour are very different from solids and liquids, as respects the light which they give out.

LESSON XXXIII.-RADIATION AND ABSORPTION.

301. We have hitherto chiefly confined our remarks to the radiation of carbon at different temperatures; and taking that substance as the type of solids, and comparing its radiation with that from incandescent gases, we have found a very great and striking difference between the two classes of spectra, that of carbon being continuous, while those of gases are discontinuous. We shall now endeavour to connect the radiative properties of bodies with their absorptive properties.

Let us begin with the temperature of boiling water, or 100° Cent. Let us now, therefore, suppose that we have a large thermometer at this temperature hung up in a room having the temperature of melting ice. The thermometer will lose heat in two ways; by convection on account of the air which surrounds it, and which is continually carried off and renewed, and also by radiation. But in order to confine our thoughts to the latter process, let us suppose that the chamber is a vacuum. Now, in the first place, let the outside of the glass bulb of the thermometer be coated with a thin coating of polished silver, and let us ascertain how much heat it radiates in one minute. Next let the bulb be coated with lamp-black, the same experiment being repeated, that is to say, the thermometer at 100° C. being allowed to cool for one minute in a vacuum chamber of o°. It will be found that the bulb now radiates in a minute very much more heat than it did when coated with silver. Next, let the glass bulb be left uncovered, and the thermometer will still be found to radiate almost as much as when the bulb was covered with lamp-black. Finally, let it be covered with white paper, and its radiation will still be found to be almost equally great, We are thus entitled to say that at 100° C. a blackened surface, or one of glass or white paper, radiates much more than a surface of polished silver, and we may thus construct a table of the comparative radiating powers of bodies heated to Ico C., at the top of which we may put a lamp-black

surface, a surface of glass, and one of white paper, and much lower down one of silver, which is a very bad radiator. Our table of radiating substances for heat of low temperature will therefore stand thus :—

Good radiators

Bad radiator.

Lamp-black surface.
Glass

White paper

Polished silver ""

302. Suppose now that the thermometer is at 0°, and is carried into a vacuum chamber of the temperature of 100°, this being the reverse of the previous process; in the first place, let the bulb, as before, be coated on the outside with a coating of silver; it will absorb a certain quantity of heat in one minute; observe how much. Next, blacken the bulb with lamp-black, repeat the experiment, and measure the absorption which takes place in one minute as before; the absorbing power of the thermometer will now be considerably increased. Again, if the coating be entirely removed, and nothing left above the glass bulb, it will be found that the absorbing power of the glass bulb is almost as great as that of the blackened bulb, and the same result will be obtained if the bulb be covered with white paper. We may thus construct a table of the comparative absorbing powers of various bodies for heat of 100°, at the head of which we may place a lamp-black surface, a surface of glass, and one of white paper, and much further down one of polished silver.

Our table of bodies which absorb heat of 100° will therefore stand thus :

303. If these two tables, the one of radiators and the other of absorbents, be now compared together, they will be found to be identical; so that the blackened thermometer at 100° will, in the first case, cool much more rapidly than the silvered one when transferred to the chamber at o°, on account of its superior radiation, and will also, in the

second case, starting from o, become heated much more rapidly than the silvered one when transferred to a chamber of 100°. In fine, good radiators are also good absorbents, bad radiators bad absorbents. It is worthy of remark, before proceeding further, that surfaces behave very differently in their absorbing power for different rays. White paper and glass, as we have seen, are both very strong absorbents of low temperature heat, while both of them are manifestly non-absorbents of luminous rays.

Extending these considerations to visible rays proceeding from bodies of high temperature, they furnish us with some very interesting and instructive experiments, which will now be described.

Experiment I.-Take a porcelain plate of black and white pattern (the black of the pattern will of course be a strong absorbent of luminous rays, while the white will be a less powerful absorbent). Heat it to a good red or white heat in the fire, and when so heated, take it out and rapidly carry it to a dark place; the black will appear much more brilliant than the white, presenting a very curious reversal of the pattern.

of polished platinum foil Bring this foil to a red

Experiment II.-Take a piece and make an ink-mark upon it. heat with the flame of a Bunsen's burner in a dark room, and the ink-mark will shine out much more brightly than the polished platinum.

Experiment III.—Make a white mark on a black poker with a piece of chalk; when heated to a good red heat, examine it in the dark, and the chalk will shine out less brightly than the rest of the poker.

These experiments might be multiplied indefinitely, all tending to show that bodies which, when cold, are good absorbents, are, when hot, good radiators; and the observation may be extended to plates of various thicknesses as well as to mere surfaces. Thus a polished plate of rocksalt absorbs very little heat of low temperature (Art. 295), and when heated to 100° C. it is found also to give out very little heat. In like manner a piece of transparent colourless