LECTURE III.

Historical Sketch.-Talbot, Herschel, Bunsen, and Kirchhoff.-Discovery of New Elements by means of Spectrum Analysis.-Cæsium, Rubidium, Thallium, Indium.-Their History and Properties.Spectra of the Heavy Metals.-Examination of the Light of the Electric Discharge.-Wheatstone.-Volatilization of Metals in the Electric Arc.-Kirchhoff, Angström, Thalén, and Huggins.-Maps of the Metallic Lines.

APPENDIX A.-Spectrum Reactions of the Rubidium and Cæsium Compounds.

APPENDIX B.-Contributions towards the History of Spectrum Analysis. By G. Kirchhoff.

APPENDIX C.--On the Spectra of some of the Chemical Elements. By Wm. Huggins. With Maps and Tables.

I PROPOSE to point out to you to-day the properties of the new elementary bodies which have been discovered by means of spectrum analysis, the principles of which we considered in the last lecture. Before passing on to consider this point, I wish to direct your attention, for a few moments, to the history of the subject.

The experiments of which I gave you an account in the last lecture, and the results derived from these experiments, have been carried out chiefly by a German chemist and a German physicist, whose important discoveries have made the names of Bunsen and Kirchhoff celebrated throughout the scientific world.

But although these philosophers are the real discoverers of this method, because they carried it out with all due scientific accuracy and placed it on the sure foundation upon which it now rests, yet we must not suppose that the ground was before their time absolutely untrodden. No great discovery is made all at once. There are always stepping-stones by which such a position is reached, and it is right to know what has been previously done, and to give such credit as is their due to the older observers.

So long ago as 1752, Thomas Melville, while experimenting on certain coloured flames, observed the yellow soda flame, although he was unacquainted with its cause. In 1822 Brewster introduced his monochromatic lamp, in which the soda light is used; the first idea, however, being due to Melville. A simple experiment will prove to you the nature of this monochromatic soda light. I have here the means of producing a very intense soda flame, and I will throw the light on to this screen with painted letters. You will observe that no colour is noticeable in these letters. They appear in various degrees of shade or intensity, but no difference of colour is visible, because the light falling upon them is of a pure yellow colour. Now, if I throw a small quantity of magnesium powder into the flame, you will at once notice how brightly the various colours come out. We have here white light containing rays of every degree of refrangibility; hence the different colours appear, each letter being able to reflect its own peculiar rays.

Sir John Herschel, in the year 1822, investigated the spectra of many coloured flames, especially of the strontium and copper chlorides, and of boracic acid, and he writes in 1827 about this as follows: "The colours thus

contributed by different objects to flame afford in many cases a ready and neat way of detecting extremely minute quantities of them."

Fox Talbot, whose name we know as being so intimately connected with the origin of the beautiful art of photography, makes the following suggestions respecting these spectra. Writing in 1826 he says: "The red fire of the theatres examined in the same way gave a most beautiful spectrum, with many light lines or maxima of light. In the red these lines were more numerous, and crowded with dark spaces between them" (these are the strontium lines which you see on the diagram), "besides an exterior ray greatly separated from the rest, and probably the effect of the nitre in the composition" (this is really the red potassium line caused by the nitre). "In the orange was one bright line, one in the yellow, three in the green, and several that were fainter." The blue line which he mentions is the blue strontium line which we saw so plainly. "The bright line in the yellow" (our friend sodium) "is caused without doubt by the combustion of sulphur." Talbot got wrong there, as did many of the early observers. They could not suppose that so minute a trace of sodium could produce that yellow light; and even Talbot says that the yellow line must be caused in certain cases by the presence of water. He continues: " If this opinion" (about the cause of formation of these lines) "should prove correct, and applicable to the other definite rays, a glance at the prismatic spectrum of a flame might show it to contain substances which it would otherwise require a laborious chemical analysis to detect." We cannot even now express the opinion entertained at the present moment more concisely than Talbot did in the year 1826. These early

66

H

observers did not, however, determine the exact nature of the substance producing the colour, inasmuch as the extreme sensitiveness of this sodium reaction put them off the scent they could not believe that sodium was present everywhere.

Both Herschel and Brewster found that the same yellow light was obtained by setting fire to spirits of wine diluted with water, and Talbot also mentions cases in which no soda was, as he thought, present, and yet this yellow line always made its appearance. Hence he says, 'The only matter which these substances have in common is water," and he throws out the suggestion that this yellow line is produced by the presence of water. In February 1834 Talbot writes: "Lithia and strontia are two bodies characterised by the fine red tint which they communicate to the flame. The former of these is very rare, and I was indebted to my friend Mr. Faraday for the specimen which I subjected to the prismatic analysis. Now it is very difficult to distinguish the lithia red from the strontia red with the naked eye, but the prism betrays between them the most marked distinction which can be imagined. The strontia flame. exhibits a great number of red rays well separated from each other by dark intervals, not to mention an orange and a very definite bright blue ray. The lithia exhibits one single red ray. Hence I hesitate not to say, that optical analysis can distinguish the minutest portions of these two substances from each other with as much certainty, if not more than, any known method." Still Talbot says further on, that "the mere presence of the substance, which suffers no diminution in consequence, causes the production of a red and green line to appear in the spectrum."

Professor William Allen Miller next made some interesting experiments in 1845 on the spectra of coloured flame produced by the metals of the alkaline earths, and came still nearer to the result which we now find Bunsen and Kirchhoff arrived at in 1861. Diagrams of these spectra accompany the memoir, but they are not characteristic enough to enable them to be used as distinctive tests for the metals, owing to the fact that a luminous flame was used. Hence the investigations of Miller in 1845 attracted less attention than they deserved.1 The first person who pointed out this characteristic property of sodium was Professor Swan, in 1857, and it is to him that we owe the examination and the determination of the very great sensitiveness of this sodium reaction. So much then for the history of the method as applied to the detection of the alkalies and the alkaline earths.

We will now pass on to the consideration of the new elements which have been discovered by spectrum analysis. And, in the first place, I would direct your attention to the new alkaline metals discovered by Professor Bunsen in 1860. Shortly after he made his first experiments on the subject of spectrum analysis, Bunsen happened to be examining the alkalies left from the evaporation of a large quantity of mineral water from Dürkheim in the Palatinate. Having separated out all other bodies, he took some of these alkalies, and found, on examining by the spectroscope the flame which this particular salt or mixture of salts gave off, that some bright lines were visible which he had never observed

1 See extract in Appendix B. from Kirchhoff's Contributions to the History of Spectrum Analysis, Phil. Mag. Fourth Series, vol. xxv. p. 250, 1863,