Page images

would all fall to the surface of the earth in time; but the great particles would fall most quickly, and so most straightly, and the other particles in proportion, as they were smaller, would fall more slowly, and be carried farther on. Thus the dust would fall, not into a single heap, but into a long band, so graduated that at every point the particles should be perfectly uniform in size. The thickness, then, of the dust at every point will indicate the quantity of dust of a certain definite fineness which existed in the quantity let fall. And if dust of any given fineness were altogether wanting, it would be indicated by a blank space, or bare cross-line in the band of dust we have supposed to settle on the ground.

In this experiment we have, then, fully analysed a mass of dust as regards the magnitude of its particles. It is indeed by a process exactly of this nature that our daily bread is prepared; for everyone knows that chaff is separated from corn by a stream of air. In the good old times the thresher used, we believe, to throw the corn and chaff into the air, the wind blowing freely through the two great doors which are seen in every aged barn. Then upon the barn-floor was witnessed the very separation of the light and heavy particles which we had just before taken as an imaginary experiment.

But the illustration represents the phenomena more closely than might be thought. On the wave theory of light, each simple ray of light consists of ethereal waves of some definite length, not less than the 60,000th of an inch, nor greater than the 35,000th of an inch, but of any length within these limits. On the wave length depends the colour of the light and other distinctive properties. Now the refracting medium or prism retards the passage of the ethereal waves in a degree depending on the wave length, and it follows, from the mechanical nature of the wave, that each ray is caused to bend in a degree proportioned to its retardation. Hence arises the spectrum, which we may define as the series of rays of every kind or colour, which a compound ray of light contains, laid out in the successive order of their wave lengths. Thus far we have used the spectrum as a general term; it is when we come to inquire what are the particular rays which any kind of light exhibits in its own peculiar spectrum, that the subject opens up before us in its infinite diversity. Any perfectly solid substance, indeed, raised to a white heat, gives every kind of ray perceptible to the eyes of men. But there are few or no substances always solid, and any substance becoming luminous in the gaseous state gives a spectrum peculiar to itself, containing a definite combination of rays, recognisable by their colour and angular position in the spectrum, such as no other substance could present. When a gaseous substance, then, gives out light of any colour, we may consider that all the rays of the continuous or perfect spectrum are omitted, except those which by their combination produce the peculiar colours of that light.

It may seem that the compound ray before prismatic decomposition should be a distinct mark of the gaseous substance from which it proceeds. And so it often is; for half a century back any chemist would have distinguished certain elements, such as sodium and strontium, by the yellow and red colour which they give to flame, and it would take a volume to detail all the particular tests for substances depending on their colourproperties, long since devised by chemists. But when several substances with their appropriate colours are mixed together, what are we to do? Any two colours, as every one knows, compound into a third distinct colour, and any such compound colour may be composed, in an infinite number of ways, of other simple colours. The compound colour does not spontaneously reveal its composition. Some substances, too, may give out a very powerful light, and others a mere glimmer of colour. To strain our eyes in looking at a coloured flame in hopes of discerning every trace of colour, would be just as profitable as looking for a needle in a bundle of hay. But place the prism before our eyes,-every different colour is now referred to a different position in the spectrum, and the weak light is independent of the strong. Every ray may be separately examined, and the needle will not be missed among the hay.

The compound beam of light, then, is like a letter directed to us by an unknown hand, the very receipt of which informs us of the writer's existence, but of little more. Analysing the light by the prison is like opening the letter; and as in the spectrum we see brilliant lines, the many-coloured characters of nature, so in the other have we lines written in pen and ink. Still we are none the wiser unless we know the language of the writing. Until the present time, chemists have scarcely bethought themselves even to open these letters of nature; if they did, the characters within seemed meaningless, or more puzzling than Egyptian hieroglyphics. It is the German Professors Kirchhoff and Bunsen who have studied the language of Light, with an acuteness and perseverance worthy of their greatest philological countrymen. They have found the key to the characters, and the first few pages deciphered contain information astonishing by its novelty and correctness.

Among numberless varieties of the spectrum, the solar spectrum is naturally of supreme importance, and we have a rough example of it in the transverse section of the rainbow. Newton first examined the spectrum obtained when the sun's light is ssed through a prism, and he stated it to consist of the seven colours, violet, indigo, blue, green, yellow, orange, and red, in unequal quantities, shading into each other. He employed, however, a thick beam of sunlight, so that in reality he had innumerable separate spectra overlapping each other and confusing all the details. Wollaston, in 1802, first used a very narrow band of light, so that each kind of light of definite colour would also form a very narrow line or band, like one veft or cross thread of a ribbon. The spectrum of compound light then would resemble a broad band or ribbon, continuous or broken, according as all the possible rays of light, or wefts, were present or were wanting. He now observed that the solar spectrum was not continuous, but crossed by narrow dark lines, indicating deficient rays. The excellent German observer Fraunhofer, however, first described the solar spectrum with fidelity, not only rediscovering the principal lines, to which he assigned the symbols B, C, D, E, F, G, H, but also actually detecting 590 minuter lines with which the spectrum is scored. Brewster carried up the estimate of dark fixed lines to 2000; and Kirchhoff has lately observed 70 lines between two of the principal lines, D and E, belonging, as we shall see, to one substance only, — namely, iron.

The constituent coloured bands of the solar spectrum, and the dark lines which separate them, increase in number with the improvement of the apparatus and the care of the observer, almost as the stars increase in number with the power of the telescope. Conceive, then, a piece of ribbon, of which the wefts or cross threads are almost infinitely fine and infinitely numerous, no two agreeing in colour or coming into contact, and we have an illustration of the solar spectrum. It will then be easily understood how, with these innumerable but invariable details of light and shade, the solar spectrum has become a kind of map, to which all coloured rays of light are constantly referred. Knowing that a ray of light corresponds, for instance, to that which forms the edge of Fraunhofer's line D, we know at once all its characteristics of colour, refrangibility, length of wave, &c. The mapping of the solar spectrum is, then, an operation almost comparable in importance and accuracy with the determination of the places of the fixed stars, or the trigonometrical survey of a kingdom.

Let us now consider the principal details of the process of spectrum-analysis.

The apparatus employed consists essentially of a prism and two small telescopes, inclined at an angle of about 122°, and so placed on either side of the prism that the first telescope throws the light under examination upon the prism, and the second collects and transmits it to the eye. The ocular lenses of the first telescope are replaced by a plate, in which a slit, made by two knife-edges, is arranged to coincide with the focus of the objectglass. The prism is a hollow vessel, of which two sides are of plate-glass, inclined at an angle of 60°; it is filled with the liquid bisulphide of carbon, which has a high refractive and dispersive power, and can moreover always be obtained in a state of uniform purity. When observations are to be made, a gas-lamp of a peculiar construction, in which the gas is mixed with air before combustion, and a colourless but very hot flame thus obtained, is placed just before the slit of the first telescope. The substance to be tested is fixed into a small hook of platinum-wire, or melted on to it, and then placed in the flame somewhat lower than the point at which the axis of the telescope-tube meets the mantle of the flame. According as the substance thus placed in the flame is more volatile, it is more quickly diffused in a gaseous state through the flame, and communicates to it all the characteristic colours of the elements it contains. Now, looking through the other telescope, and adjusting the intensity of the light and the magnifying powers at will, we see all these distinctive colours laid out as bright bands in the spectrum. We can see, indeed, only part of the spectrum at once, but then, by turning the prism or the telescope round, every part of the spectrum may be made to move over the field of view.

A very perfect form of the spectrum apparatus has already been made by the celebrated optician Steinheil, of Munich, under the direction of the Heidelberg professors.* The prism is fixed upon the top of a stout iron pillar, two arms from which carry the two telescopes. Of these, the second, or observation telescope, is capable of an angular movement; so that not only may different portions of the coloured rays be brought into the centre of the field, which is defined by a thin wire stretched across it, as in theodolites and astronomical instruments, but the angular breadth and interval of separation might easily be measured. A useful addition is a small prism placed over half of the slit at the end of the first telescope. Two kinds of light may now be admitted at once, and yet separately; so that their spectra are seen one above another, and may be easily compared.

To examine or detect the spectrum of any element, its chloride-salt is commonly preferred, as being most volatile, and thus giving the most powerful light. When several salts of different volatility are mixed, some of them are often driven off before the others begin to volatilise, and thus we have the cor

* We have seen some of these instruments, manufactured by Messrs. Ladd and Oertling of Bishopsgate Street.

responding spectra presented separately, like dissolving views. This, however, is an adventitious beauty of the process, and is not usually necessary for the detection of all the elements present, since the several spectra may be distinguished by a practised eye even when all presented at once, the lines seldom falling upon each other, and all being produced in perfect independence. This last point is of supreme importance, and therefore, before proceeding, we should notice that Kirchhoff and Bunsen have, by very numerous experiments, confirmed the law before stated by several observers, that the spectrum of each element is produced quite independently of any other substance. Thus it is, theoretically, quite a matter of indifference what compound of a substance, or what sort of colourless flame, we use: the spectrum of the substance will in every case be visible.

Bunsen and Kirchhoff have as yet published their results concerning six of the elements only, viz. the metals of the alkalies and alkaline earths. We will briefly describe them in the following order :

Calcium. This element, the metallic base of lime, can be examined only in those of its compounds which are volatile; others, like lime, are almost perfectly fixed. The lime spectrum is distinguished by a very characteristic bright green line, referred to as Ca. ß in the notation adopted by the professors. A second distinct feature is an intensely bright orange line, Ca. a, lying near the red region of the solar spectrum. There are other yellow, red, or orange rays, of inferior strength and importance. It has been found by certain sure experiments that a quantity of lime no larger than the 100,000,000th part of a grain, when present in the testing-flame, can be detected by that appearance of the two characteristic green and orange rays.

Barium gives a more complicated spectrum than the other alkaline and alkaloid metals. But among the many green and orange lines, the green ones Ba. a and Ba. ß are far the most distinct, and are a sufficient mark of the presence of this element. From the greater diffusion of the barium spectrum the test is much less delicate; still it was found possible to detect the 60,000th part of a grain of this substance.

Strontium may almost be known without the prism, by the splendid crimson colour which it gives to a flame; it is, in fact, the substance used with such effect in making red-fire for theatrical and pyrotechnic purposes. Its spectrum, as we should expect, has powerful red bands, six in number, with one orange band, Sr. a, and a very beautiful blue line, Sr. 8, separated, of course, by a considerable interval from the rest. The 1,000,000th part of a grain of strontia can be detected.

Potassium, the element found in pot-ashes, gives a widely

or corone of the solaright orange Tissors. A second

« PreviousContinue »