ON LIGHT AS A MEANS OF INVESTIGATION PROF. G, G, STOKES . a t REESE LIBRARY UNIVERSITY OF CALIFORNIA. Received \^/rf %s9 &*L^ /< Accessions NQ.&**. **-/ / Shelf No. BURNETT LECTURES. ON LIGHT. BURNETT LECTURES ON LIGHT. ON LIGHT AS A MEANS OF INVESTIGATION. DELIVERED AT ABERDEEN IN DECEMBER, 1884, d&j&W // ~ y \ {( TJNIVKi:- GEORGE GABRIEL TOKES, M.A., F.R.S., &c, FELLOW OF PEMBROKE COLLEGE, AND LUCASIAN PROFESSOR OF MATHEMATICS IN THE UNIVERSITY OF CAMBRIDGE. Honfcon: MACMILLAN AND CO. 1885 \The Right of Translation and Reproduction is reserved.} CTambrtoge : PRINTED BY C. J. CLAY, M.A. & SON, AT THE UNIVERSITY PRESS. If. CONTENTS. LECTURE I. PAGE Subjects of the present course Use of the mode of absorption of Light by substances as a discriminating character Examples Various effects produced by the action of Light incident on bodies Phosphorescence Epipolic dispersion of Light Fluorescence Its use as a means of discrimination Phos- phorescence produced by electric bombardment Delicate test of Yttria thus afforded . i LECTURE II. Rotation of the plane of polarization of polarized light produced by various liquids Its application to quantitative determinations, and to the study of molecular grouping Magnetic rotation of the plane of polarization Application to the discrimination between isomeric compounds Bright lines in the spectra of flames Application as chemical tests Discovery thereby of new elements Connexion between the powers of emission and absorption of the same substance for the same kind of Light Conditions as to temperature which determine whether a spectral line shall appear as bright or dark . . . 25 " H TJNIVEL LECTURES ON LIGHT. SECOND COURSE. ON LIGHT AS A MEANS OF INVESTIGATION. LECTURE I. Subjects of the present course Use of the mode of absorption of Light by substances as a discriminating character Examples Various effects produced by the action of Light incident on bodies Phosphorescence Epipolic dis- persion of Light Fluorescence Its use as a means of discrimination Phosphorescence produced by electric bombardment Delicate test of Yttria thus afforded. IN the course of lectures which I had the honour of delivering to you last year, I dwelt on the nature of light itself, and endeavoured to give you a fair idea of the evidence on which we accept that view of its nature which is now I may say almost uni- versally held by scientific men, namely, that light consists in undulations propagated in a highly elastic S. II. I 2 LIGHT AS A MEANS OF INVESTIGATION. medium called the ether, and that those vibrations are not, as we should previously have imagined, to and fro in the direction of propagation, as we know are those of air in the propagation of sound, but transverse to the direction of propagation. At the conclusion of the course I stated that according to an arrangement adopted in consultation with the Burnett Trustees the subject of the present year's course should be investigations carried on by the aid of light; or in other words on light as a means of investigation. In one sense of course this would include nearly every investigation that we can carry on ; for in nearly every case we make use of our eyes, and without light they would be of no avail. But it is obvious that it cannot be in this comprehensive sense that the title of the present course is meant to be taken. The investigations actually intended are those in which the properties of light in their re- lation to ponderable matter enable us to ascertain something regarding the nature or the condition of that ponderable matter. Even as thus restricted the subject remains a wide one, branching out into other departments of science, especially chemistry, miner- alogy and astronomy. The special subjects belonging to our general class which I have selected to bring before you il ORIGIN OF COLOUR. 3 are (i) absorption, and its application to the dis- crimination of bodies ; (2) the emission of light consequent on absorption, or produced by different means other than incandescence, and its application as a test of the presence of particular bodies, or of the condition of the bodies so emitting it; (3) the rotation of the plane of polarization of polarized light, and its connexion with the constitution of bodies ; (4) the emission of light by incandescent bodies in the state of vapour, and its application as a test of the presence of particular bodies; (5) the information thus afforded as to the constitution or condition of distant bodies ; (6) the influence of the motion of bodies on the refrangibility of the light emitted, absorbed, or reflected by them, and the information thence afforded as to the motion of distant bodies. That colours in all cases depend primarily on the heterogeneous nature of light, on its consisting of kinds differing from one another in refrangibility, and in the sensation of colour which they produce when separated one from the other, was long ago shown by Newton. In all cases the existence of colour depends either on the emission of a certain kind or certain kinds of light in excessive proportion in the source of light, as in the case of coloured I 2 4 LIGHT AS A MEANS OF INVESTIGATION. flames, or in a subsequent treatment of the light of such a nature that certain kinds are preserved for reception by the eye t* the exclusion of others, or preserved in larger proportion than others. The production of colour by fluorescence, of which more presently, forms no exception to this rule provided we regard, as we may regard, the fluorescent body as the source of light, albeit in this case a secondary source. By far the commonest exhibition of colour belongs to what is called absorption. This applies for in- stance to the verdure of the fields, the colours of flowers, those of dyed dresses, &c. If we take for example a coloured flower, and hold it in different parts in succession of a pure spectrum, it appears of the colour of the part of the spectrum in which it is placed, but its luminosity varies greatly from one part of the spectrum to another, the flower appearing comparatively bright in those colours which approximately agree with that under which it is seen in white light, and com- paratively dark, and sometimes almost black, in other parts of the spectrum. This shows, as Newton pointed out, that the proximate cause of the ex- hibition of colour in ordinary light is that the flower reflects copiously light of certain kinds, and reflects feebly or hardly at all light of other kinds. But what is the nature of the selection of certain COLOURS DUE TO ABSORPTION. 5 colours for reflection while others are not reflected ? Newton's attempt to explain this by reference to the colours of thin plates was not a happy one, and the inadequacy of such an explanation has long been recognised. To illustrate the true answer let us take one of the commonest of all colours, the green of vegetation. If a green leaf be put into alcohol a plant with an acid juice, like sorrel, should be avoided, as in that case special treatment is required to avoid decompo- sition the green colouring matter is dissolved, and we obtain a beautiful clear green solution. [This was shown.] There cannot be any doubt that the cause of the colour in the leaf and in the solution is the same, but in the solution it is quite plain that it is in the transmitted light that the colour is seen. If we examine the behaviour of the liquid with respect to any particular kind of light, we find that as the light travels onwards in the solution it continually grows weaker and weaker, and presently becomes too weak to be any longer perceived. It is found that in passing across a layer of the liquid of given thickness a given fraction of the light is lost, from whence it readily follows that as the light travels on in the solution its intensity decreases in geometric progres- sion as the distance traversed increases in arithmetic progression. But the rate of loss varies immensely from 6 LIGHT AS A MEANS OF INVESTIGATION. one kind of light to another. For one kind there may be hardly any loss at all : the liquid may behave almost like water; while for another kind the absorption may be very rapid, and the liquid may behave almost like ink. For other kinds again the rate of absorption may be intermediate, and the liquid behave like water with some ink mixed with it. Our clear solution requires to be looked through in order that the colour may be seen ; for if it be looked at, and the colour seen by means of light which has been reflected from the further side of the vessel, that comes to the same thing. Now suppose we mix some flour or powdered chalk with the liquid. We shall be able to see the colour by looking at the liquid mixture ; [This was shown.] but that is evidently because the white powder reflects light from its surfaces in an irregular manner, and the light so reflected has had to pass through the solution in the interstices between the reflecting surfaces before it reaches the eye. It is just the same thing in a natural leaf; the irregularities of the structure cause irregular reflections and re- fractions of light in the substance of the leaf, and the light in its progress is exposed to absorption on part of the colouring matter. The colour of solutions of metallic salts or of organic substances has naturally long been used as a ANALYSIS BY THE PRISM. 7 means of discrimination. But the eye can only very imperfectly judge of the composition of a mixed light by means of the colour. When the mixture is re- solved into its constituents by means of a prism, it can then be seen at a glance how each constituent kind of light has been affected. Now in many cases some two or more parts of the spectrum are specially attacked by the coloured substance, and when that is the case the character and position of the bands of absorption when the light has been strained by passing through a suitable thickness of the coloured substance, which is usually most conveniently examined in the condition of a solution, or of a clear glass, are. often highly dis- tinctive. Till of late years, chemists were unaware of the simple means of discrimination thus afforded, and occasionally made mistakes which a single glance at the absorption spectrum of the substance they had under examination would have prevented. Physicists indeed were familiar with these phenomena, but they usually studied them with other objects in view. Now, in consequence of the labours of Bunsen and Kirchoff, a spectroscope of some kind is considered an indispensable portion of the equipment of a chemical laboratory. I will mention two or three examples. Under certain circumstances red solutions are obtained by 8 LIGHT AS A MEANS OF INVESTIGATION. the action of acids or acid salts on certain oxides of manganese, which resemble a good deal in colour the permanganates, and*which like them are oxidising agents. It seemed not unnatural to infer that the colour of the former class of solutions was due to permanganic acid, and a paper was written in one of the scientific journals in support of this view. A single glance at the spectrum of the transmitted light would have shown this to be an error, for the permanganates are characterised by a remarkable and highly distinctive system of bands of absorption, of which there are five occupying the region of the brightest part of the spectrum. These are wholly wanting in the spectrum of the solutions of the class first mentioned. A long time ago the eminent chemist M. Fremy attempted to divide chlorophyll, as the colouring matter of green leaves is called, into a blue and a yellow substance, of which it was a mixture. He attempted to carry out this idea in two ways. In one he added to the alcoholic solution hydrate of alumina and a very little water, and filtered. The filtrate was of a yellow colour, and the precipitate on being treated with appropriate solvents for the colouring matter it contained yielded solutions which were not indeed blue, but of a decidedly bluer green than before. In another attempt he dissolved EXAMPLES OF THE APPLICATION. 9 chlorophyll in a mixture of hydrochloric acid and ether, when on separation of the two solvents an upper yellow stratum was obtained and an under one which was nearly blue. Fremy designated the colouring matters contained in the yellow solutions obtained in these two ways by the same name phyxo- xanthine ; whereas the prism shows that the first is a definite colouring matter characterised by two bands of absorption in the blue, but the yellow solution ob- tained by the ether process is a complex mixture. In the detection and separation of the rare earths of the Ceria and Yttria groups, the number of which has recently been largely increased, and which chemists are now actively engaged in investigating, great assistance is derived from the very peculiar spectra which the solutions of the salts of some of them present. Dr Gladstone was I believe the first to point out the peculiar spectrum of the salts of didymium, which show bands of absorption almost rivalling those of coloured gases in their narrowness. This brief sketch may serve to give some idea of the way in which the relations of light to ponderable matter may be turned to account in investigations belonging to a branch of science altogether different ; though instances far more striking will fall under our notice in a future lecture. I pass on now to a different though allied subject. IO LIGHT AS A MEANS OF INVESTIGATION. In the phenomenon of absorption, as the very name indicates, light is as it were swallowed up and disappears. But, as belonging to the incident light, a certain amount of energy is continually being brought to bear on the body in which the light is absorbed, and this energy cannot be annihilated ; there must be something to show for it. Now different effects are produced in different cases, or it may be are simultaneously produced in the same instance. The commonest effect of all, one indeed that would seem to be always present, whether alone, or mixed with others, is that of raising the temperature of the body on which the light falls. It is true that the rays which are the most luminous are by no means those by which this effect is most powerfully produced ; and that they are far surpassed in this respect by certain rays lying beyond the red, which though physically identical in their nature with rays of light, from which they differ only in the same way in which rays of one colour differ from those of another, do not neverthe- less affect the eye. Still, the effect is produced by rays of whatever refrangibility, though feebly it is true by those of high refrangibility. But this effect of absorption, though the commonest of all, is not of a nature to be made available as a means of discrimination. NTJVJ EFFECTS PRODUCED BY LIGHTRNlA II Another effect sometimes produced is that of effecting chemical changes. On this is founded the whole practice of photography. Although the seat of the chemical change varies from one part of the spectrum to another, with a variation in the chemical nature of the substance acted on, the variation can- not, except in rare cases, be conveniently used as a means of discrimination. Photography may indeed be most usefully employed for the purpose of de- tecting and recording what is going on in different parts of the spectrum, especially in the case of the ultra-violet and ultra-red regions, where eye obser- vations fail. But this does not properly come under the subject of the present course. Another effect sometimes produced as a result of the absorption of incident rays of light is that of phosphorescence or fluorescence. It has long been known that certain precious stones, especially the diamond, and the sulphides of the alcaline earths after exposure to light shine for some time in the dark. [A collection of phosphorescent sulphides was shown, which glowed with different colours after having been excited by the radiation from burning magnesium wire.] The laws of this phenomenon have been investi- gated by many physicists, especially by M. Edmond Becquerel, who has now given a collected account of his labours and those of his predecessors in 12 LIGHT AS A MEANS OF INVESTIGATION. his work entitled La Liimttre. This phenomenon forms no exception to the statement I have made, that in the phenomenon of absorption the incident light is swallowed up and disappears ; for though it is true that as a result of that absorption light is forth- coming, yet that light is not in any way of the character of the incident light, but of a different composition altogether. The phosphorescent body is rendered for a time self-luminous as a result of the action of the light upon it, but the incident light itself is spent in the process. This phosphorescence of long duration has not however been much used as a means of distinguishing one body from another, with one remarkable excep- tion, in the case of a similar effect produced in a dif- ferent way, to be mentioned presently. There is however a phenomenon which appears to differ from phosphorescence only in degree, in point of duration, which is more generally available for this purpose, at least in the case of organic bodies. In the course of some experiments on the action of light on vegetable colours, in which he had occasion to throw a more or less pure spectrum on the sub- stance to be examined, Sir John Herschel noticed a remarkable prolongation of the spectrum beyond the violet, which is usually regarded as the termi- nation of the visible spectrum, when the .spectrum EPIPOLIC DISPERSION. 13 was thrown on turmeric paper and a few other substances. This extension Sir John Herschel at- tributed to a peculiarity in the reflecting power of the paper : and as, in the case of turmeric paper, which showed the phenomenon best among the substances which he tried, the prolongation was coloured yellow, he even speculated on a possible repetition of the colours of the spectrum in order after passing beyond the violet. Some years later he observed a curious phe- nomenon ' in a solution of sulphate of quinine of moderate strength. This liquid when seen by trans- mitted light appears colourless and transparent like water, but when viewed otherwise exhibits in certain aspects a peculiar blue colour. This, Herschel found, proceeded from a narrow stratum of the liquid adjacent to the surface by which the light entered, though a small part came from greater depths ; and the blue light emanated from this stratum in all directions. But the most remarkable thing was that the light which had once produced this effect, though not apparently altered by transmission through the liquid, was deprived of the power of producing the effect; so that the light which had passed through a cell filled with the liquid, and then fell on another vessel containing a similar solution, no longer pro- duced a blue stratum near the surface by which it entered the second solution. Herschel called the 14 LIGHT AS A MEANS OF INVESTIGATION. phenomenon of the production of the blue stratum epipolic dispersion, and designated as epipolised the light which, by passage through such a solution, had been deprived of the power of producing the blue stratum again, though he did not explain wherein epipolised differed from common light. I have already alluded to the careful study made by M. Edmond Becquerel of the phenomenon of phosphorescence. While engaged in this research he noticed that when a pure spectrum was allowed to fall on several of the phosphori examined (especially sulphides of the metals of the alcaline earths), in certain regions of the spectrum, whether belonging to the visible part or to the invisible region of greater refrangibility, the phosphorus shone with a light of a different kind from that which fell upon it, but only so long as the light remained falling upon it, differing in this respect from the ordinary phenomena of phos- phorescence, in which the phosphorescent light lasts a very appreciable time. This phenomenon he rightly regarded as a phosphorescence of very brief duration ; but from connecting it too closely with ordinary phosphorescence he failed to perceive its full bearing ; and though he was actually working with an acid solution of quinine, the "dichroism" of which he expressly mentions, and had determined by pho- tography its great opacity for rays of high refrangi- FLUORESCENCE. 15 bility, he never thought of putting these things together, or perceived their intimate connexion with the phenomenon just mentioned of phosphorescence of short duration. In reflecting on the possible explanation of the epipolic analysis of light discovered by Sir John Herschel, I was led to believe that the rays which produced the blue light dispersed in a solution of sulphate of quinine were not the blue rays at all, but the rays of high refrangibility which are mostly invisible. Once this idea is suggested it is easily put to the test of experiment, and the result completely verified the anticipation. Perhaps the most striking feature in this phe- nomenon is the change of refrangibility of light which takes place in it, as a result of which visible light can be got out of invisible light, if such an expression may be allowed : that is, out of radiations which are of the same physical nature as light, but are of higher refrangibility than those that affect the eye ; and in the same way light of one kind can be got out of light of another, as in the case for instance of an alcoholic solution of the green colouring matter of leaves, which emits a blood red light under the influence of the indigo and other rays. Observation shows that this change is always in the direction of a lowering. 1 6 LIGHT AS A MEANS OF INVESTIGATION. But in speaking of a change of refrangibility I would guard against being misunderstood. All that is intended is that light m of one refrangibility being incident on the substance, light of a different re- frangibility is emitted so long as the first remains in action. It is not to be supposed, according to a view which has erroneously been attributed to me by more than one writer, but which I never for a moment entertained, much less published, that the refrangibility is changed in the act of reflection from the molecules. The view which I have all along maintained is that the incident vibrations caused an agitation among the ultimate molecules of the body, and that these acted as centres of disturbance to the surrounding ether, the disturbance lasting for a time which, whether it was long enough to be rendered sensible in observation or not, was at any rate very long compared with the time of a single luminous vibration. And now that M. E. Becquerel has shown experimentally by his beautiful phos- phoroscope the finiteness of duration of the emission of light in the case of solids in which it was so brief that its emission was described as "fluorescence," as in a solution of sulphate of quinine, there can no longer be any doubt as to the identity of nature of phosphorescence and fluorescence, even though the finite duration of the emission of light after the FLUORESCENCE AS A TEST. I/ incident rays have been cut off has not at present been experimentally demonstrated in the case of any liquid. I will not dwell further on the nature of this phenomenon, because the subject of the present course mainly confines me to the consideration of light as a means of research. In this point of view it is obvious that as the presence or absence of in- visible rays of high refrangibility in a pure spectrum is so easily shown by the exhibition or non-exhibition of fluorescence in a suitably chosen fluorescent body, the phenomenon renders these rays virtually visible, and enables us thereby at once to study the action of bodies upon them, such as the absorbing power of bodies for them. This however constitutes rather an extension of another optical means of discrimination than an independent method. But even as an in- dependent means of discrimination the phenomenon is by no means devoid of application. The pro- perty is sufficiently common to be pretty frequent- ly available, and yet sufficiently uncommon to be, at least when taken in all its features, highly dis- tinctive. Experience shows that at least in the case of a single fluorescent substance, as distinguished from a mixture of two or more such substances, the tint of the light emitted, with a given solvent if the active substance be in solution, is usually sensibly S. II. 2 1 8 LIGHT AS A MEANS OF INVESTIGATION. constant whatever may be the refrangibility of the active rays. The tint admits of being observed by suitable methods notwithstanding the admixture of various other bodies with the one in question, even though they should be coloured, provided they are not themselves fluorescent. In this respect it is somewhat superior to the colour due to absorption, the observation of which requires that the coloured body should be at least approximately isolated, so far at least as coloured substances are concerned. As an example of the use that may be made of the observation of the tint of the fluorescent light, I may refer to a powerfully fluorescent and easily obtained solution which may be made from the bark of the horse chestnut. A decoction of the bark is powerfully fluorescent, as may best be seen by pouring a little into water, but it is liable soon to become brown from tannin &c. contained in it. This may be removed by adding to the freshly made solution a suitable metallic salt, such as a salt of alumina or sesqui-oxide of iron, precipitating by ammonia, and filtering, when the filtrate shows the phenomenon to perfection. Long before the true nature of the phenomenon was known, a glucoside named aesculin had been obtained from the bark, to which the play of colour was attributed. But similar solutions of aesculin show a fluorescence of a jj FLUORESCENCE AS A TEST. 19 decidedly deeper blue than that of the solution obtained directly from the bark. This shows either that the aesculin is a product of decomposition a hypothesis which is negatived by the fact that the reagents employed do not affect the tint of the fluorescent light of the solution obtained from the bark or that in the latter the aesculin must be mixed with some other fluorescent body. In fact the bark of the horse chestnut contains a second glucoside analogous to aesculin, and yielding like it highly fluorescent solutions, which has been named fraxin from its having been first discovered in the bark of the ash. The slightly alkaline solutions of fraxin show a fluorescence which is intermediate between blue and green, and the presence of both bodies in the solution obtained from the bark explains the greater paleness of the blue of the light emitted from it by fluorescence than of that coming from a solution of pure aesculin. [The fluorescence of solutions of pure aesculin and fraxin obtained from the bark of the horse chestnut, and of a purified solution of the bark itself without any separation of the two bodies, were exhibited by holding burning magnesium wire over glasses containing the three solutions.] I have already mentioned the searching character of prismatic analysis when applied to the examination of absorption-colours, as compared with a mere ex- amination of the colours by the naked eye. Just 22 20 LIGHT AS A MEANS OF INVESTIGATION. in the same way the observation of fluorescent substances in a pure spectrum exhibits features by which they may be followed and detected in spite of the presence of other substances even in large quantity. In proceeding in the order of increasing refran- gibility, that is from the red to the violet and beyond, the fluorescence is found to commence at a part of the spectrum differing with the particular substance observed, and once commenced to continue from thence onwards. It is frequently however, indeed I might almost say generally, subject to fluctuations of intensity; in one place the fluorescence being copious, and the rays which excited it being soon spent, while in another place it is comparatively feeble, and the rays which excited it, unless they should happen to be absorbed by some other sub- stance present in an impure mixture, are able to penetrate to comparatively great distances in the solution before they are spent. Both these, the place of commencement and the copiousness of emission, form discriminating characters which may be usefully employed. The first is however very much bound up with the colour of the emitted light, and therefore hardly forms an independent feature, except in the case of a mixture of two or more fluorescent sub- stances, but in this case is often of great utility. PHOSPHORESCENCE BY BOMBARDMENT. 21 The second again is bound up with the character of the absorption due to the substance, so that it hardly forms an independent feature when the latter can be observed ; but it has this advantage over absorption as a discriminating character, that whereas the latter requires the substance to be approximately isolated from other coloured substances in order that it may be observed, the former can be observed independently in great measure of the presence of such impurities, and enables us in fact to predict the character which the absorption-spectrum of the substance will exhibit when isolated. I will not dwell further on these phenomena of fluorescence, which are too much of a speciality to be of very general interest. But before I leave the subject of phosphorescence there is one research which I must mention, to which allusion has already been made. In the instances of phosphorescence, including that phosphorescence of brief duration denominated fluorescence, which have hitherto been mentioned, the phosphorescence was excited by light, or by radiations of the same physical nature as light, though they might not be capable of affecting the eye. But there is another mode of exciting phos- phorescence which has been much studied by Mr Crookes and some others. Electric light in all its 22 LIGHT AS A MEANS OF INVESTIGATION. forms, abounding as it does in rays of high refran- gibility, is specially well suited to excite phospho- rescence in most phosphorescent substances. In particular, the beautiful discharge in pretty highly exhausted tubes is well adapted to excite it on bodies placed outside, or still better when that is practicable on bodies within the tube, or on the walls of the tube itself. So far, we have merely a particular case instance of the ordinary mode of exciting phospho- rescence. But when the exhaustion is very high a new mode of exciting it comes into action. We are familiar with the glow surrounding the negative electrode in tubes which are considerably exhausted. When the exhaustion is moderate the glow appears to invest closely the negative electrode. But as the exhaustion progresses it is seen to be separated from the electrode by a dark space, which becomes wider and wider as the residual gas becomes rarer and rarer, until at last it reaches the walls of the tube, and may even at extreme exhaustions fill the whole tube. Now the portion of the walls which lies within the dark space is seen to glow with a phosphorescent light resembling in colour the phos- phorescence produced by radiation from the gas glowing under the influence of the discharge with more moderate exhaustions, but usually a good deal brighter; and the same is the case with substances TEST FOR YTTRIA. 23 placed within the tube when the dark space extends over them. Experiments instituted with the direct object of determining whether the phosphorescence in this case is due to radiation or to an actual bombardment of the walls or enclosed substances by molecules projected with great velocity from the negative electrode indicate that the phosphorescence is to -be referred to bombardment, provided at least that we do not by the use of that term imply that the action is merely mechanical, but only assert that it is due to the actual transfer, or something accom- panying the actual transfer, of the molecules ; for it seems probable that the electric discharge, whatever the appropriate idea of that may be, has much to do with it in a direct way. If in this wider sense of the term the phosphorescence is really due to bombard- ment, that justifies us in speaking of it as a distinct mode of production from the ordinary one in which it is due to radiation. From the processes to which the tube has to be subjected in order to obtain exhaustions so nearly perfect as is requisite for this observation, the method is pretty well confined to the examination of in- organic substances. In studying the spectra of the transmitted light, Mr Crookes frequently came across a characteristic citron band, evidently indicative of some particular substance. Was it a new element, or 24 LIGHT AS A MEANS OF INVESTIGATION. rather compound of a new element, or was it a compound of one of the known elements, including therein a number of ne*v and rare earths recently discovered and as yet imperfectly known? The known elements were tried, and for a long time apparently with a negative result. Frequently the substance Mr Crookes was in search of seemed to be driven into a corner, and yet it managed to slip through the fingers. At last the loophole was discovered, and the substance proved to be the long known but rare earth yttria, in combination with sulphuric acid. Once the origin of the citron band was known, it furnished a very delicate test of the presence of yttria, and the application of this test showed that this rare earth is in reality very widely distributed, though in general in minute quantity. This illustrates what is I believe generally admitted by chemists, that chemical purity represents an unattainable ideal, to which we can only make a more or less near approach in actual experiment. LECTURE II. Rotation of the plane of polarization of polarized light pro- duced by various liquids Its application to quantitative determinations, and to the study of molecular grouping Magnetic rotation of the plane of polarization Appli- cation to the discrimination between isomeric compounds Bright lines in the spectra of flames Application as chemical tests Discovery thereby of new elements Connexion between the powers of emission and absorption of the same substance for the same kind of Light Conditions as to temperature which determine whether a spectral line shall appear as bright or dark. I HAVE mentioned the rotation of the plane of polarization of polarized light as another property of light which has been turned to account in in- vestigation. It was in the year 1815 that Biot, in pursuing his investigations of the colours developed in polarized light which is subsequently analysed, by the introduction of a crystalline plate which is variously inclined, being desirous of working with inclinations within the crystal greater than could be 26 LIGHT AS A MEANS OF INVESTIGATION. obtained when the plate is inclined in air, was led to try the effect of inclining the plate when immersed in oil of turpentine contained in a cell with parallel sides. The oil being of considerable refractive index, approaching that of plate glass, and accordingly that of ordinary minerals, would allow the passage of light within the crystal at an inclination higher than that attained at even a grazing incidence w r hen the plate is in air.- But in order to establish the legiti- macy of the process it was necessary to show that the oil did not itself act on polarized light. Contrary to what was to have been anticipated, inasmuch as the oil is alike in all directions, differing notably in that respect from a crystal, the oil did act on polarized light, and Biot was thus led to the discovery of the action of certain liquids on polarized light. This affords a striking example, of which there are so many, of the way in which the honest pursuit of scientific investigations will sometimes lead to a discovery unthought of by the person who made the experiment with a totally different train of ideas in his mind. The amount of rotation is found to be different for the different colours, being greater for the more refrangible ones. Accordingly the transmitted light is not wholly extinguished in any position of the analyser, but the different colours are extinguished in ROTATION OF PLANE OF POLARIZATION. 27 succession, so that on rotating the analyser there is a constant change of colour. For polarized light of a given refrangibility passing through an active liquid of given kind, the rotation is found to be proportional to the length of path of the light within the liquid, as might have been anticipated. When the liquid is not homo- geneous, but consists of a solution of an active substance in an inactive solvent, the rotation is found to be proportional to the strength of the solution. Hence the observed rotation divided by the length of path and by the strength of the solution, is a constant depending on the nature of the active substance, and may be called the specific rotation. The specific rotation, like any other physical constant belonging to the substance, may be used as one of the charac- ters by which it may be known, but its chief value arises in part from the comparative rarity of the rotatory property, so that in many cases there is but one such body liable to be present in the solution, which in that case may be determined quantitatively notwithstanding the presence of other substances ; in part from the delicacy of the molecular combinations of which it is able to take cognizance, and of the existence of which it sometimes constitutes the sole evidence. Let me mention an example of each kind of 28 LIGHT AS A MEANS OF INVESTIGATION. application. One of the commonest substances which in inactive solvents yield active solutions, and one which at the same tin^e possesses a high specific rotation, is sugar in its different forms. When sugar is the only active substance liable to be present, and the form in which it occurs is known, we can by simply observing the rotation of the plane of polar- ization produced by passage of the light through a known length of the solution determine the quantity of sugar present. Accordingly an instrument de- signed for effecting this measurement with accuracy and facility has been named a saccharimeter. A rather amusing application of the saccharimeter which has been proposed and I believe actually em- ployed in some foreign country is for excise purposes, in fixing the duty on beer, which according to a law there in force depends on its alcoholic strength. In distilled spirit we have practically only two substances liable to be present, namely, water and alcohol, which are of very different specific gravities, so that the specific gravity of the mixture, which can be easily and rapidly taken by the hydrometer, determines the strength. But in beer the specific gravity depends not only on the alcohol, which lowers it, but on the sugar, or rather dextrine, which raises it. (The dis- solved substances other than sugar may be neglected, as being present in no great quantity.) Hence the STUDY OF MOLECULAR GROUPING. 29 specific gravity gives only one relation between two unknown quantities. But the amount of sugar can be determined by measuring the rotation of the plane of polarization, and knowing this we can calculate what the specific gravity would become if the sugar were removed, and from thence deduce the alcoholic strength. As an instance of feeling, so to speak, a delicate molecular combination by the observation of the azimuth of the plane of polarization of polarized light, I will refer to the behaviour of a freshly made solution of grape sugar or glucose. This substance crystallizes by itself, i.e. with merely water of crystal- lization, and curiously enough forms also a definite crystalline compound with common salt. The crystals of either kind dissolve readily in water, and the solu- tions when once made remain apparently unchanged. Nevertheless when the rotation of the plane of polari- zation produced by the solution of either kind of crystals is immediately measured, it is found to be nearly double that of a solution of glucose of the same strength which has been made some time. The rotatory power of the fresh solution gradually diminishes, and in the course of seven hours or so at ordinary temperatures reaches a permanent value, a change which takes place at once on boiling. Had this change occurred only in the solution of the 30 LIGHT AS A MEANS OF INVESTIGATION. crystallized compound of glucose and common salt, we should naturally have inferred that the compound at first dissolved as such, but was of an unstable character, and this may possibly actually be the case ; only if it be so some analogous change must take place when we have nothing present but glucose and water, and what the nature of the change of molecular grouping in that case may be we do not know. It is remarkable that cane sugar, though so closely allied to glucose, shows no such phenomenon. As another example of the application of this method to the study of the mode of combination of bodies in mixed solutions, I may refer to an elaborate research by Dr Jellett, on the combinations formed when a vegetable alcaloid is present in a solution containing two acids, with either of which the alcaloid could combine. It was in the year 1845 that Faraday made the remarkable discovery that uncrystalline bodies, or at least those of high refractive power, when under the influence of a powerful magnetic force act on polarised light. The action consists in a rotation of the plane of polarization, agreeing so far with the natural action of liquids like sirup of sugar, from which however it differs in the circumstance that whereas in the latter the action is alike in all - * CA STUDY OF MOLECULAR GROUPING. 31 directions, in the former the amount of rotation varies with the direction of the light. It is greatest when the light travels in the direction of the lines of magnetic force, and vanishes in a direction perpen- dicular to those lines, and the direction of rotation is reversed on passing across the equatorial plane, or plane perpendicular to the lines of magnetic force. It has been found that for a given substance and kind of light the rotation is proportional to the length of path, to the magnetic force, and to the cosine of the inclination of the path to the lines of force, from whence by measuring the rotation under given cir- cumstances we can determine the specific rotation, or more readily the ratio of the specific rotations of two substances more readily because in this case we do not require the magnetic field to be sensibly uniform (or else calculable, so as to permit us to perform an integration) and the magnetic force known; it is sufficient that the magnetic field and the force in it be the same in successive ex- periments. The specific magnetic rotation, like the specific gravity, the refractive power, &c. is a physical con- stant giving one of the characters of each particular substance. An elaborate comparison of the specific rotations of several chemical groups has recently been made by Mr Perkin, with the result among 32 LIGHT AS A MEANS OF INVESTIGATION. others that the difference between isomeric com- pounds is in many cases clearly revealed by a difference in the valuts of this constant, whereas most other physical constants are nearly alike for the two. It appears therefore that the rotation of the plane of polarization produced by the action of magnetism on bodies across which light is proceeding, like the natural rotation belonging to such bodies as sirup of sugar, &c. is capable of laying hold of and revealing delicate differences of molecular grouping. It is less easily observed than the natural rotation, from which also it differs in being of general instead of ex- ceptional application. In all the phenomena which I have brought before you in my last lecture and in this, and indeed in all that I shall have occasion to mention in this year's course, there is a very intimate relation between molecular grouping and the optical features observed. We touch here on the boundaries of our present physical knowledge. That light consists in the vibrations of a subtile medium or ether, that self-luminous bodies, including phosphorescent bodies, which are for the time being self-luminous, are in a state of molecular agitation which they are capable of communicating to the ether, that consequently in the phenomenon of absorption molecular disturbance PODffTS ASCERTAINED OR CONJECTURAL* 33 r. i ". "i .-.--. it t!~--; -rX'^'.-.t "" tt-'-tr'i. ~.:ir^- tions all this is so weft established as to leave no reasonable room for doubt. But what may be die r-* ; : be die mode of connexion by winch die "f th t ~tr.tr ii'ititt t~~.t rr.:.t ; -Itr tr tr.t in their turn are able to agitate the i ~j t-.~- i i. - - t ii t.tt 1. 1 rr.ir. . ~.t t z. vt. i p i . t~.* 11 ir"iLi"it-in i r. r . " r L 1 1 '. i* rr. t i L. .".Lt rr.L*-* ~t t."_t mtut LnitL- cause of the difierence of the velocity of propagation "i r.^'-.t ir. i ..:t-^L~iti ::ri"-..2.r.y c;.Lr:iti _:r~t :r. rr. t z. i L ..--it rir^_iL ,i r-ifLr * r.-Z..""- i~ rr.Ln.rtTtti. i.v a r:tLt::- ::" tht ^int ::" c-iIirli.L:::- ::* polarized light, still more what may be the 1 1 *_*". "i i i* 1 1 r. "i ~ L ~TT~. "i *_ f rr. L r~_ r^ c i~i* ir *_". ^ i r 1 1 L- ^"^^""*" "* ~."^.T *-"" """ --~J"1 1. * " Z..^ ^ - ___ ~^_ "7~r *^r __"""1_- IL^ - - - .T* _,__,-- be We have seen how searching are the of light with respect to the bodies, although I have said i'f:r~iti:- they irfir e of crystals. The - * . *. f !____ . il.I.^I.. f f . .. ___ J _..-"- -.1.-. - - ''- I^lr.-V _-.t.t ^r . ___ --.-_ _ ~ :: ..^..t irt _i:rt.y :: i t: :: 34 LIGHT AS A MEANS OF INVESTIGATION. kind which could not exist at a high temperature. The subject which I propose next to bring before you relates almost exclusively to bodies in a state of incandescence, so that organic combinations are ex- cluded. If on the one hand the field of research is thereby limited, on the other the nature of the phenomena connects them with other branches of science, rendering the subject of more general interest. It has long been known that salts of particular metals, such as those of the alkalies and alkaline earths, when introduced into a feebly luminous flame, such as that of a spirit lamp, cause a coloration de- pending on the base of the salt. Thus salts of potash give a violet, of soda a yellow, of lithia a red, of baryta a green, of strontia a red, of lime a brick red. The colours thus produced have to a certain extent been used by mineralogists in discriminating between different minerals by the aid of the blow- pipe. But just as the prismatic analysis of the colours due to absorption reveals characters of the absorbing body which are often highly distinctive, but which would escape detection so long as we merely observed the absorption colour with the naked eye, so here a prismatic analysis of the coloured flames reveals immensely more than can be perceived merely by BRIGHT LINES IN SPECTRA OF FLAMES. 3$ looking at the colour. As long ago as 1834, the late Mr Fox Talbot showed that the red due to a salt of strontia and the red due to a salt of lithia can be at once distinguished by the prism, which in the case of lithia shows a narrow well-marked line in the red, not far from the line C, while the spectrum of a strontium flame wants this line, and is of a more complicated character. Several years before, both Sir John Herschel and Mr Talbot had drawn at- tention to the characteristic lines produced on intro- ducing the salts of certain bases into flames, and had pointed out how small a quantity of a substance suffices to produce the effect, though it is true that Talbot entered into some erroneous conjectures as to the origin of the bright line D. It is remarkable for how long chemists neglected the precious means of discrimination lying at their very hands in the use of the prism a striking example of how much may be lost by a too exclusive devotion to one branch of science to the neglect of others. Notwithstanding that W. A. Miller had published maps of the spectra of flames coloured by the alkalies and alkaline earths, it was not till Bunsen and Kirchhoff published their cele- brated researches that spectral analysis came into regular use with chemists. Bunsen and Kirchhoff engaged in a methodical 32 36 LIGHT AS A MEANS OF INVESTIGATION. chemico-optical examination of the spectra of salts introduced into an almost non-luminous flame, for which they used the flatne of a mixture of air and coal gas, burnt in a Bunsen's burner, taking the greatest care to purify the substances used, and examining separately the spectra given by the same base combined with a variety of acids. They found that the spectra depended on the metal of the salt, and not on the acid radical ; the acids which were sparingly volatile merely showing the spectra more feebly. The spectra showed bright lines or narrow bands characteristic of the metals ; the heavy metals as a rule showed no spectra when examined in this way. The order of importance of the various bands in the spectrum of the same metallic salt as tests of the presence of the metal was determined, and thus it became easy to detect even small quantities of these metals present in a mixture. These investigations were almost immediately rewarded by the discovery of two new elements, Rubidium and Caesium, which were traced by the appearance of certain bright lines in the spectrum of a Bunsen flame when a variety of specimens of substances from different localities, including waters from mineral springs, were introduced into it. They proved to be the metals of two oxides of the group of alkalies, and were named after the colour of their DISCOVERY OF NEW ELEMENTS. 37 most distinctive ' bands. Nor was this all. The facility of the test, which though indicated long- before had not been put in practice, enabled Bunsen and Kirchhoff to show that lithia, which previously had been regarded as a rare substance, was on the contrary very widely distributed, though usually present in small proportion, and led to the discovery of sources and means of separation of the alkali by which it can be obtained at a comparatively cheap rate. The method of spectrum analysis, carried out as above indicated, or modified by the employment of a succession of electric discharges instead of a Bunsen flame, has led in the hands of others to the discovery of three more new metals, namely Thallium indepen- dently by Crookes and Lamy, Indium by Reich and Richter, and Gallium by Le Coq de Boisbaudran. In their original paper, Bunsen and Kirchhoff contented themselves with establishing the fact that different salts of the same metal when intro- duced into a Bunsen flame gave the same spectrum, which could therefore be used as a test of the presence of the metal ; they did not commit them- selves to any theory as to what the particular vapour present in the flame might be which produced for each metal its characteristic spectrum. That it was a vapour of some kind, follows both from the cir- 38 LIGHT AS A MEANS OF INVESTIGATION. cumstances of the experiment, and from the con- sideration that it is only in the state of vapour that substances exhibit such -narrow absorption bands as are actually produced by the flames, or as would correspond to the bright lines in the spectrum of the light they emit. It might be supposed that as dif- ferent salts of the same metallic oxide yield solutions which as a rule exhibit similar modes of absorption, so different salts when volatilized in a flame might yield vapours consisting of the salts themselves, and yet having a spectrum in common. Or it might be supposed that the identity of the spectra was evidence that the salts were decomposed in the flame, and that the glowing vapour which yielded the spectrum common to them all was that of the metal itself. But a different conclusion resulted from the obser- vations of Alexander von Mitcherlich on the spectra of the chlorides, bromides, iodides and fluorides of the alkaline earths. When a bead of a chloride for instance is introduced into a Bunsen flame, in the manner followed by Bunsen and Kirchhoff, a spectrum with the bands of the salts in general of the same base is obtained. But as in this mode of observation a minute quantity of the volatilized chloride is present in an atmosphere in which there is plenty of oxygen and vapour of water, it may very well be that the chloride is decomposed, with the formation of an BRIGHT BANDS DUE TO COMPOUNDS. 39 oxide and hydrochloric acid. To prevent this von Mitcherlich used a solution of the chloride to which a comparatively large quantity of sal ammoniac, which itself gives no spectrum, was added. In this way the volatilized chloride was present in an atmosphere abounding in hydrochloric acid (from the temporary dissociation of the sal ammoniac) and was accordingly maintained as such, and now the spectrum showed bands indeed, agreeing so far with the spectrum obtained in the former manner, but the actual bands were quite different. Accordingly we must infer that in this case the glowing vapour was the chloride, but in the former method the oxide. The other haloid compounds behaved in a similar manner, showing spectra differing from each other, and from that of the oxide. Moreover there was a remarkable similarity of character between the spectra of the chloride, bromide and iodide of the same metal; a group for instance of bright lines in the chloride having corresponding to it in general arrangement, but differing a little in position, a group in the bromide, and the latter again having corresponding to it a group in the iodide. Moreover the order of the change corresponded to the chemical order, the bands of the bromide being intermediate between those of the chloride and those of the iodide, just as in its chemical relations bromine is inter- 40 LIGHT AS A MEANS OF INVESTIGATION. mediate between chlorine and iodine. We may infer that the vibrating molecular systems which disturbed the ether, and were thus