UC-NRLF I GAL J ;HODS IN CONTROL AND RESEARCH LABORATORIES PUBLISHED liY ADAM HILGER LTD. 75.4 GAMDEN ROAD, LONDON, N.W.I ENGLAND , *->!, 5V v." OPTICAL METHODS IN CONTROL AND RESEARCH LABORATORIES PUBLISHED DY ADAM HILGER LTD. 75A GAMDEN ROAD, LONDON. N.W.I ENGLAND CONTENTS PAGE INTRODUCTION i THE SPECTROMETER AND SPECTROGRAPH FOR EMISSION SPECTRA - - 2 Qualitative Spectrum Analysis - - - - - 3 Quantitative Spe'ctrum Analysis - - 4 ABSORPTION SPECTRA AND SPECTROPHOTOMETRY - 8 Ultra-violet Absorption Spectra - 8 Spectrophotometry in the Ultra-violet - - 9 Spectrophotometry in the Visible Region, and Dye Analysis - 13 REFRACTOMETERS - 15 Historical Notes -- -15 The Abbe and Dipping Refractometers, and their Applications - - 16 Applications to Physiological and Clinical Analyses - 20 The Pulfrich Refractometer ----- - 20 The Rayleigh Gas and Liquid Refractometers - - 21 List of Substances in Wagner's Tables - 23 List of Biochemical Applications ----- - _ 24 Application of Jemgerature Corrections to Mixtures - 26 -,'r - - - - - - - - - 28 -\r' t - . - - 28 OPTICAL METHODS IN CONTROL AND RESEARCH LABORATORIES THE optical methods dealt with in this book are those employing spectro- scopes (or spectrographs), spectrophotometers, refractometers, and polari- meters. No detailed descriptions of these instruments or their techniques are included, but in each case references are given to sources of informa- tion on these points. We have already published an essay on the effect on light radiation of passage through matter. 1 There is a tendency to lose sight of the deeper significance of these physical properties when they are merely used for the identification of substances and such technical purposes. It seems to us that such a tendency, natural though it be, should be guarded against by both makers and users of these instruments. Recent physico-chemical investigations seem to promise that the study of these optical phenomena will within the next few decades result in a great extension of our knowledge concerning the intimate structure of the molecule, with a commensurately enhanced power of following and con- trolling the chemical transformations of matter. But the Works Chemist requires, one may almost say, the maximum of empiricism. To keep pace with his work he needs instruments which he can read at once, and which either have scales giving him the percentages he wants, or tables from which he may get the information he requires without calculation. Indeed, only if he can hasten and compress into a rela- tively short time the solution of the practical questions which urgently insist on being answered day by day, can he hope to have time to keep in touch with or to carry out research on the fundamental subjects. Such being the needs of the chief market for these types of scientific instru- ments, we have, in compiling this book, confined ourselves to practical applications alone. 1 Refractive Index, Absorption, Wavelength, and Rotatory Power in Relation to Molecular Structure. 449765 SECTION I THE SPECTROMETER AND SPECTROGRAPH FOR EMISSION SPECTRA THE spectroscope was, of course, a valuable tool in the laboratories of Kirschoff and Bunsen, and it has been in regular use by chemists ever since. The literature of Spectroscopy has during that time grown to great bulk and importance ; but whereas with the refractometer the practical applications at the present time overshadow the profound theoretical importance of refract! vity, with the spectroscope and spectro- graph the reverse is the case. This is not surprising. Those departments of physics which deal with radiation and with the structure of matter have, during the past twenty years, relied at every step on spectroscopy, both for the inspira- tion and for the verification of theory. The advances due to Spectroscopy in these fields have been so rich in sensation that the new industrial applications seem trivial in com- parison. Furthermore, until the introduction of recent instruments (commencing with our Constant Deviation \Yavelength Spectrometer in THE SPECTROMETER AND SPECTROGRAPH 3 1904, and continuing with our Quartz Spectrograph, sizes A [now known as size E 6] and C [now known as size E 2] in 1909), the table spectro- scope was tedious and difficult to use, inaccurate, and unsuited in design for any but academic laboratories ; while its results when obtained were not immediately utilisable. But as will be seen from the following notes on practical spectroscopy, there are indications that spectrum analysis is at length about to take its proper place as an instrument in constant use in industrial laboratories. Qualitative Spectrum Analysis for Chemists. With regard to qualitative spectrum analysis, which has developed from Kirschoff and Bunsen's early work, the literature is very extensive, and the main principles well known. Unfortunately, however, the literature deals chiefly with the use of the older instruments which were laborious to use, and which rendered their results in a form not con- venient to utilise. We give here a few references which will be found helpful to those undertaking spectroscopic work seriously for the first time. For methods of obtaining spectra with arc or spark discharges, see general text books on subject, viz. : BALY. Spectroscopy, 2nd edit. Longmans, Green. London. WATTS. Introduction to the Study of Spectrum Analysis (1904). Longmans, Green. London. HAGENBACH AND Translated by KING. Atlas of Emission Spectra of KONEN. most of the elements. Wesley. London. URBAIX. Spectrochimie. The spectroscopic notes given on p. 55 et seq. of Hagenbach and Konen give much information regarding the setting up of apparatus, exposure > etc. Urbain's book includes instruction for analysis of the rare earths, and for general spectrum analysis. The atlas of GISSING, Spark Spectra of Metals (1910), published by Balliere, Tindall and Cox, Henrietta Street, London, gives the prin- cipal lines of an element to be looked for, e.g. Ruthenium. Gissing also points out what is not always realised namely, that little practice is required to recognise immediately the characteristic spectra of most of the * elements without in each case having recourse to wavelength measurement. 4 OPTICAL METHODS Lines in the Arc Spectra of Elements, published by ourselves (price 173. 6d.), is a list of the principal lines in the arc spectra of most of the elements arranged in order of wavelength. A feature of this book that has been much appreciated is that opposite each line of an element the next prominent line of that element is given. Not all the rarer elements are included in the above atlases ; more complete works are : EDER AND VALENTA. Atlas of Spectra (1911). EXNER AND HASCHEK. Wellenlangen Tabellen (1911-12). F. Deuticke. Wien and Leipzig. 3 vols. KAYSER. Spektroskopie, vols. v. and vi. Finally, a brief but complete statement of the operations required in taking a spectrogram will be found in the leaflet The Spectrograph in Metallurgy, issued by ourselves (post free on application). Quantitative Spectrum Analysis. Hartley in 1882 first pointed out that " as the quantity of any element present in a substance is diminished, many lines due to that element cease to appear in the resulting spectrum." Those lines which are still present in the spectrum when the element is reduced to a very small quantity were called by Hartley the " persistent " lines. This work of Hartley (Phil. Trans. 1882 and 1884) was continued at Dublin by that of Pollok and others, and in France by the independent work of A. de Gramont. De Gramont distinguished by the name of Rate Ultime that line in the spectrum of an element which is the last to disappear. These Rates Ultimes common to the Spark, Arc, and Flame Spectra, include the rays which are spontaneously reversible. The above workers have based on their researches methods of quanti- tative spectrum analysis likely to become of great technical importance, and we therefore give here a bibliography of their papers on the subject, with notes of their contents. Author. Journal. POLLOK. Proc. Roy. Dublin Soc. (1907), 11, 184-216. The persistent lines of Mg, Zu, Cd, Al, In, II, Cu, Ag, Hg, Sn, Pb, Te, As, Sb, Bi, Be, Si. POLLOK AND Proc. Roy. Dublin Soc. (1907), 11, 217-228. LEONARD. . Quantitative spectra of Fe, Al, Cr, Si, Zu, Mg, Ni, Co. THE SPECTROMETER AND SPECTROGRAPH 5 Author. Journal. POLLOK AND Ibid. (1907), 11, 229-236. LEONARD. Quantitative spectra of Ba, Sr, Ca, Mg, K, Na. POLLOK AND Ibid. (1908), 11, 257-269. LEONARD. On the separation of cerium, lanthanum, and yttrium. LEONARD. Ibid. (1908), 11, 270-279. Quantitative spectra of Mo, \V, Th, Zr. POLLOK. Ibid. (1909), 11, 331-337. Quantitative spectra of Ti, U, V. Ibid. (1909), 338. The spectrographic analysis of a specimen of commercial thallium. DE GRAMONT. Compt. Rend. (1907), 144, p. 1101-1104. The " raies ultimes " of calcium, strontium, copper, silver, magnesium, zinc, aluminium, thallium, ger- manium, tin, lead, bismuth, chromium, and iron. DE GRAMONT. Compt. Rend. (1908), 146, 1260-1263. The " raies ultimes " of the metalloids, tellurium, phosphorus, arsenic, antimony, carbon, silicon, and boron. DE GRAMONT. Compt. Rend. (3 Jan. 1910), 150, 37. The distribution of certain persistent lines in the solar spectrum. An examination is made for the lines of K, Rb, Mg, Ca, Sr, Ba, Sc, Al, Ti, Pb, V, Cr, Mn, Fe, Ni, Co. DE GRAMONT. Compt. Rend. (January 1910), 150, 154. A similar examination of the distribution in stellar spectra. DE GRAMONT. Compt. Rend. (25 July, 1910), 151, 308-311. The place of the raies ultimes with regard to the spectral series, in particular the 2138 line of Zn. 2365 line of Cd. 2536 line of Hg. DE GRAMONT. Compt. Rend. (July, 1912), 155, 276-279. A further development in the technique, particulars of a new Hilger spectrograph, describes how certain causes of instrumental errors were overcome, and gives details of the " raies ultimes " for chromium, manganese, iron, nickel, and cobalt. 6 OPTICAL METHODS Author. Journal. DE GRAMONT. Compt. Rend. (1918), 166, 94. Researches on the line spectrum of titanium. An illus- trated description of his sparking apparatus is given. The chief lines of titanium are grouped after the manner employed by Hartley and Pollok (see below), and the application of the method to metallurgy is described. DE GRAMONT. Compt. Rend. (1918), 166, 365. The rates ultimes of colombium (niobium) and zirconium are investigated. Reference is made to a similar investigation by M. A. P. de Sampio Forjaz (Arquivos da Univ. de Lisbonia, vol. iii. 1916) on the lines of zirconium and uranium. DE GRAMONT. Compt. Rend. (1918), 166, 477, The persistent lines of boron : it is shown that the appearance of wavelength 3451.2 shows that the sub- stance is present at a concentration of 1/10,000 or over ; wavelength 2497.8 for a concentration of i part in 100,000 ; and wavelength 2496.8 for 5 parts in 100,000. " L'intensite notable de ces trois raies du bore a 67 dix milliemes dans le spectre d'etincelle de cet acier permet done de les indiquer comme susceptibles d'une utilisation pratique certaine pour les analyses metallurgiques." Later investigations go to prove that the ultimate lines vary with the conditions of excitation of the source. De Gramont, 1914, finds, for instance, that they shift towards longer wavelength as the tempera- ture diminishes : Pollok and Morrow, too, find that the ultimate lines with the vacuum tube are different from those with the spark. These ob- servations, which were not altogether unexpected, do not however detract from the importance of the work on quantitative spectrum analysis. Another and quite different method has been developed by Hill and Luckey (Trans. American Electrochemical Soc. (1917), 32, 335), depend- ing on a difference in volatility in the constituents of an alloy. They use the " arc " discharge, and measure the time required for a line to disappear w r hen a given weight of the unknown mixture is burned in the crater of the arc. This method is obviously limited in its applica- tion, but in the instance of lead in copper, a range of -004 - -216 per cent, can be estimated with all the precision required in refinery work. THE SPECTROMETER AND SPECTROGRAPH 7 At a discussion of the above paper, M. G. Lloyd of the Bureau of Standards, Washington, stated that the spectrographic method of quanti- tative analysis was in use at the Bureau for determining the impurities in tin boiler plugs, and that results are quite as reliable as chemical analysis when the impurity does not amount to more than o-i per cent. He adds : " Quantitative spectroscopic analyses of steel have been made, especially for the determination of chromium and titanium. For such Illustrating mode of obtaining comparison spectra by the Hartmann diaphragm supplied with Quartz Spectrographs K i, K 2, and K 6. elements as niobium and molybdenum the estimation of small quantities is more reliable than chemical analyses." The following further printed matter relative to the Spectrometer and Spectrograph may be obtained post free on application to Adam Hilger, Limited, 75 A Camden Rd., London, N.W. I : Catalogue Section D Catalogue Section E Catalogue Section F Wavelength Spectrometers. Spectrographs, including the Quartz Spectro- graphs, which give the ultra-violet region of the spectrum in which so many of the distinctive metal lines are found. Accessories for spectrometers and spectrographs. Catalogue Description of Outfit for producing spark discharge suitable for Spark Spectra. SECTION II ABSORPTION SPECTRA AND SPECTROPHOTOMETRY Ultra-Violet Absorption Spectra. The investigation of ultra-violet absorption spectra has thrown a good deal of light on the constitution of organic substances. The work of Hartley, Dobbie ; Baly, Stewart, Kayser, and others demonstrates the importance of this side of the subject. The utility of observations in the ultra-violet has been repeatedly confirmed, especially in the case of substances such as isatin which may exist in tautomeric forms. In such cases, the spectrograph is one of the readiest means of deciding between alternative formulae for the structure of a compound. (See Journal of the Chemical Society (1900), 77, 839, and Kayser's Hand- buck, especially vol. iii.) A number of similar cases are referred to in the British Association Reports for 1901 and following years. An interesting study in connection with the constitution of organic substances is that of the ultra-violet absorption spectra of alkaloids. Many communications on this subject appear in the pages of the Journal of the Chemical Society, some by Hartley, but mainly by Dobbie and his co-workers. One of the features which renders this method of in- vestigation so valuable is the fact that only very small quantities of the pure substances are required, and the manipulation is easy. Frequently one or two centigrams suffice (see Ley, Farbe und Konstitution bei Organische Verbindungen, Leipzig). So delicate is the method that such a minute quantity of pyridme as o-ooooi gram in 100 cc. of aqueous ammonia can be readily detected by the effect on the ultra-violet spec- trum. (See Hartley and Dobbie, Journ. of the Chem. Soc. (1900), 77, 318). As a method of detection for certain organic substances the ultra- violet absorption spectrum is unsurpassed, both as regards ease of manipulation and certainty. Thus, much less than 0-5 mg. of strychnine can be detected, and verified by its characteristic ultra-violet spectrum. 8 ABSORPTION SPECTRA 9 Similarly, as little as 3 mg. of cocaine can be identified by its charac- teristic band in the ultra-violet. It is interesting to note that this band is similar to that due to benzoic acid, 1 which fact furnishes additional evidence of the presence of a benzoic acid residue in the molecule of cocaine. Very minute quantities of benzoic acid itself can also be identified by this method, a matter of some importance having regard to the difficulty of detecting this acid. Again, as little as one milligram of phenol is shown by its absorption spectrum. Further, by photographing the ultra-violet absorption spectra of the vapours, extremely minute quantities of organic substances such as benzene, and also certain inorganic substances, can be identified gener- ally by the large number of sharp narrow bands which occur in their absorption spectra. For practical work in this field a small quartz spectrograph, such as the Hilger size A (now known as size E 6), is fre- quently better than a larger, such as the size C (now known as size E 2), since the wide bands in the neighbourhood of 200 JULJUL are sharper in the small instrument. Convenient sources of light are a spark between nickel and iron, the copper arc, and the cadmium spark or arc. The last source of light possesses the especial advantage of giving a con- tinuous background between the strong lines, and this furnishes an excellent means for the observation of sharp and narrow bands. Still better, it is stated, for certain work is a Tesla discharge between aluminium electrodes under water, though this arrangement has been found by us troublesome to keep in good condition. An important industrial application of the ultra-violet spectrophoto- meter is the testing of optical glass, goggle glasses, etc., for violet and ultra-violet absorption. Spectrophotometry. By far the greater part of the above work has been carried out without quantitative measurements of the absorption at each point of the spectrum. Jt does not seem yet generally realised that we have placed on the market apparatus whereby such work can be carried out even in the ultra-violet region (see our booklet Spectrophotometers). An apparatus for measuring the absorption enormously enhances the importance of absorption spectra, for not only does it enable the true 1 That is as far as an ordinary comparison of their spectra has revealed. No truly quantitative measurements (as by means of the Hilger Sector-photometer) have apparently been made at present. IO OPTICAL METHODS form of the absorption curves to be studied, but it makes possible quanti- tative measurements of the absorbing substance. The whole of the data for an absorption curve can be obtained on a single photographic plate, and such a curve uniquely characterises and defines the amount of the substance examined. SPECTROPHOTOMETRY ii Biochem ical A pplications. Work seeming to promise results of the highest importance has been conducted on the examination of the ultra-violet absorption of blood serum by Judd Lewis, 1 and recently by Tadokoro 2 and Nakayama. 3 The former observer claims to have established in this way the existence of specific changes in the blood serum in the early stages of certain dis- eases, and the latter have observed changes in the blood serum under various conditions of physiological interest. So little has been published on the subject of quantitative absorption spectra in the ultra-violet that we give an example below. Fig. I, obtained on a Hilger Quartz Spectrograph size C (now known as size E 2} with sector photometer, and Fig. 2, the curve drawn from it, show in the case of anthracene the method of drawing the absorption curve. If this be compared with the curves sketched by inspection of Hartley's 1 Judd Lewis, The Ultra-violet Absorption Spectra of Blood Sera. . z " Ultraspectroscopic Studies on Blood Serum," by T. Tadokoro. From Journ. of Infectious Diseases (January 1920), 26, Xo. i, 1-7. 3 " Ultraspectroscopic Studies on Blood Serum," by T. Tadokoro and Y. Naka- yama. From Journ. of Infectious Diseases (January 1920), 26, No. i, 8-15. 12 OPTICAL METHODS in / \ \ 510 \ \ \ I I 530 SPECTROPHOTOMETRY 13 method, it will at once be seen what an important new tool has not been placed in the hands of organic chemists. Spectro photometry in the Visible Region. Applications of visual spectrophotometry are naturally limited to those substances which have absorption in the visible region. Dyes, of course, are in this category, and among the most important industrial applications of the visual spectrophotometer are the identifica- tion, estimation, and analysis of dyes. This application is so far confined to a very limited number of progressive dye manufacturers and users ; but it is claimed by one of the latter that the saving to his firm in cost of dyes alone effected by the use of the spectrophotometer in the last five years to be reckoned in tens of thousands of pounds. Fig. 3 illustrates how a dye analysis would be conducted by means of the Hilger Wavelength Spectrometer with Nutting Photometer, this combination forming a convenient and efficient spectrophotometer. A solution of known strength is prepared from a reliable sample of the dye under consideration, the solution being of such strength as to give an accurately measurable absorption in the region of maximum absorp- tion. Such a solution will require to be only a very dilute one. This solution is then spectrophotometered namely, its extinction coefficient is measured for a series of wavelengths by means of the Nutting Photo- meter used in conjunction with the Hilger Wavelength Spectrometer. The results are plotted, and the curve kept as a permanent reference. When testing a sample of what purports to be the same dye at any future time, a solution of equal strength by weight is made up and similarly spectrophotometered. The densities are then plotted on the same sheet as the standard curve, and any non-coincidence of the two curves indicates, of course, a difference either in strength or nature of the dye. If the difference is in the strength only (through dilution by a colourless substance, e.g. dextrine), all the ordinates are reduced in like proportion. A particular case is illustrated in Fig. 3, showing curves of solutions of eosine and erythrosine, and a mixture of the two. It will readily be understood that the instrument can be used to measure the proportions in which two dyes are mixed. For instance, in the curves shown, let us suppose A and B to be our standard curves of known strengths of reliable eosine and erythrosine, C is a mixture of the two, but we will presume the proportions to be unknown, as also whether it is merelv A and B mixed, or whether a diluent has been added. 14 OPTICAL METHODS Let m be the proportion of A in the mixture before dilution, and let the mixture be diluted by some inactive substance so that its weight before dilution is to its weight when diluted in the proportion of i to n. Then, in Fig. 3, if a, b, and c represent the ordinates of any wavelength of curves A , B, and C respectively m(a - b) +bcn. Taking ordinates for two suitable wavelengths we get two equations from which m and n can at once be found. In the present case we find ^=46% n i viz., c is an undiluted mixture of 46% A +54%^. The following booklet on Absorption Spectra and Spectrophotometry can be obtained post free on application to Adam Hilgcr, Limited, 75.1 Camden Rd., London, N.W. i. Spectrophotometers : With bibliography on absorption spectra. v SECTION III THE REFRACTOMETER ALTHOUGH the work of Gladstone and Dale, and later of Briihl and Landolt, leading to the concepts of molecular and atomic refraction have been of great value in arriving at conclusions relative to constitution, it is undoubtedly true to-day that the chief uses of refractometry lie in purely empirical fields, such as the identification of substances, control of manufacture, estimation of mixtures, etc. It is these latter applica- tions with which this booklet deals. Modern refractometers of the Abbe and Dipping type require no knowledge of optics for their manipulation, but a few brief historical notes will not be out of place, even in a book primarily devoted to an account of practical applications. Historical Notes. Abbe and Dipping Refractometers both depend upon the observation of the critical angle of total reflection. The phenomenon of the critical XIIL: PROPOSITIO. Nullus radius, quiintra corpus CryHalli fuper unatn cjus fuperficicmplus42inclinatur,avcrticepoterdciliam fuperfi- ciem pcnetrare, S FIG. 4. angle was probably known very early in the history of Optics, but the first mention of it yet traced is in Kepler's Dioptrice, published in 1611. Here he states (Prop. xiii. p. 4) that a ray incident on the surface of a " crystal " cannot pass out of it, if the ray makes an angle greater than 42 with the normal to the surface. 15 16 OPTICAL METHODS A photographic reproduction from a copy of Kepler's Dioptrice in the British Museum is given above (Fig. 4). This original Latin edition of Kepler's book is very scarce, but a German translation has been made by Plehri (No. 144, Ostwald's Klassiker, Engelmann, Leipzig). The famous Sine Law of Snell was not discovered till about ten years afterwards. Although this was not published until the appearance of Descartes' Dioptrique (Leyden) in I63S, 1 even then it was left to Sir Isaac Newton to state the simple corollary that since sin i i - =-- for light emerging from denser bodv, sin r n sin i=- when r is a glancing angle of 90. The first to realise the possibilities of this property as a practical method of measuring the refractive index of a substance was Wollaston, who in 1802 (Phil. Trans. 92, 365) made the first critical angle refractometer, which embodied the principle of the right-angled prism later adopted by Pulfrich. With the difference that he used " sights " instead of a telescope with cross-lines, and a link motion giving the refractive index on a linear scale instead of using a graduated circle, Wollaston realised the most important principles of the modern refractometers, and succeeded in making an instrument giving results correct to the third place of decimals a remarkable achievement considering the limited resources for optical and mechanical manufacture in those days. Wollaston also applied his method for determining the indices of opaque bodies ; and, in the case of oil of cloves, actually used his refracto- meter to detect falsification. Wollaston must, then, be considered the originator of Critical Angle Refract ometry. Applications of the Refractometer. The introduction of the Abbe Refractometer and later of the Immer- sion or Dipping Refractometer facilitated greatly the introduction of refractometric methods of analysis, since the earlier Pulfrich accurate though it is is not very convenient for a number of consecutive deter- minations. The Pulfrich in its turn had been a great improvement on the spectrometer, which was employed by Gladstone and the earlier workers. l.Descartes, Dioptrique. Discourse second. THE REFRACTOMETER 17 The firm of Carl Zeiss of Jena showed great enterprise in attempts to popularise the Abbe and the Dipping Refractometers. Well-known workers in different industries were encouraged to use the instrument for purposes of control, and Wagner was employed to determine the strength of a large number of aqueous solutions with the Dipping Refrac- tometer, and to compare the results with specific gravity determinations, almost invariably to the advantage of the optical method. Wagner's tables are published, 1 and form up till now the most important con- tribution to the technical use of the refractometer for aqueous solutions which lie within the range of the Dipping Refractometer. In technical organic chemistry extensive collections of data had previously been made in their treatises on oils by Lewkowitsch and other writers on the fatty oils, and by Parry and Schimmel on essential oils. There is now available also the Refractive Indices of Essential Oils, which contains complete literature references. This volume is the first of a series of collected data of refractive indices. 2 In the sugar industry, Main introduced the use of the Abbe Refracto- meter for determining the sugar content or total solids, and his work has been followed by numerous publications, so that the refractometer is now well established as a routine instrument in the sugar industry and the necessary tables are contained in the handbooks on sugar analysis. In the determination of mixtures of methyl and ethyl alcohol the refractometer may be very conveniently employed, and this determina- tion is of particular interest as the specific gravities of the two alcohols diverge far less than their refractive indices. The same remark applies to mixtures of acetaldehyde and ethyl alcohol. Much work has also been done on the application of the Dipping Refractometer to milk examination. The interest in refractometrical analysis was well shown at a largely attended meeting of the Society of Chemical Industry on February 3, 1919. Mr. Main described the use of the instrument in sugar works. Dr. Annie Homer gave the method employed in the determination of the protein content of anti-toxin sera with the immersion instrument, and 1 Can be supplied from stock by Adam Hilger, Ltd. A list of substances included is given on p. 23. Our Dipping Refractometers are made exactly to correspond in scale reading with the German instrument, so that the numerous published tables (for instance, Wagner's Tables) can be used ; although it is hoped at some future time to publish conversions of all such tables into refractive indices, and to engrave the Dipping Refractometer in refractive indices, thus dispensing with the arbitrary scale with which the Dipping Refractometer originated. 2 Published by ourselves. The second volume, Refractive Indices of Oils> Fats, and Waxes, will appear shortly. i8 OPTICAL METHODS Mr. Berry spoke of the analysis of ethylene-chlorhydrin by the Abbe instrument. The papers were followed by an interesting discussion on refractometrical analysis. It was evident that the opinion of the meeting was that the instrument was capable of greatly extended use. Here we would enter a plea for assistance in our publication of lists of refractive indices. Many chemists must possess records of deter- minations which would be of great value if accessible to other workers. Once extensive tables are in existence, it is probable that the refracto- rneter will in many more cases displace with advantage the hydrometer and pyknometer. We shall be pleased to acknowledge any observations, and to publish them if possible in our series of tables of refractive indices. Among the substances for the determination of which tables exist nicotine and pyridine can now be included. The Abbe has also recently been used for the determination of paraffins in tar oils, and for the deter- mination of original gravity and the alcohol in beers. The preparation of tables for any specific purpose is, of course, a simple matter and well worth while undertaking by anyone requiring to make frequent determinations of the same kind. In a recent paper on the refractometer some of the industrial applica- tions of the refractometer have been well summarised as follows : 1. Acetic acid, carbolic acid, cresylic acids, and other organic acids. 2. Sulphate of ammonia, nitrate of potash, and numerous other inorganic salts in solvents. (A number of these are mentioned on p. 23.) 3. Ammonia. 4. Acetone, alcohol, and many other organic solvents. 5. Glycerine, formalin, nicotine, pyridine. 6. Essential oils, and the terpenes. (Turpentine and its substitutes.) 7. Fatty oils, fats, and waxes, and also fatty acids and other by-products concerned with the industrial application of these. 8. Mineral oils and waxes and products of distillation derived from them. To the above may be added sugar, resins, benzol, toluol, solvent naphtha, cresylic acid, and organic liquids generally. For creosotes the refractometer test is likely to become of consider- able importance. In the preservation of timber it is only the coal tar and coke oven creosotes which are of high value, the blast furnace and water gas tar creosotes being deficient in those constituents which are preservative. It has recently been shown by Messrs. Dean and Bateman (Circular 112 of the United States Forest Service) in a paper on the Refractive Index of Creosotes, that the refractometer forms a test for THE REFRACTOMETER 19 distinguishing coal tar and coke oven creosotes from blast furnace and water gas tar creosotes ; and the experiments and conclusions of Dean and Bateman have recently been verified independently in this country. It will be seen that there are few chemical industries in which the refractometer is not already well established as an approved means of test in control of manufacture or examination of raw materials or of products. There is one further industrial application of the refractometer recently described, 1 which we quote in extenso. Having referred to the main use of the refractometer in industry, the writer points out a further novel application to works control, which he has used for a considerable time in the works with which he is associated. He says, " It applies to mix- tures of liquids or solutions which form a viscous or semi-solid product, and in which it is difficult to ensure a uniform dissemination of all the constituents throughout the entire bulk of a batch of material. " Many manufactured products are of this nature, and it is usually very necessary to know if, and when, adequate mixture of the constituent materials has taken place. In some cases this is extremely difficult to determine with any degree of certainty, especially in the case of complex mixtures, when even chemical analysis would probably fail to yield conclusive results, and would, in any case, be too slow a method for works' control. In such instances the refractometer will prove of great assistance when used in the following manner : " Samples are taken from different portions of the batch of material, say, at the top, bottom and middle of the vessel, and examined in the refractometer. If the figure, whatever it may be, is identical in all the samples, there is a strong indication that a homogeneous product has been obtained, and it can be assumed that the mixing has been efficient. If, on the other hand, the samples differ, there is an indication of local excess of one or other of the constituents, proving that the mixing has been inadequate. The reading on the instrument may even be indistinct providing that it is equally so in each sample. Of course, the best control is obtained by this method when the refractive power of the different ingredients used varies to a considerable extent, and in in- dustrial products this will be found to be very frequently the case. "An example from the writer's own observations may be given. It is known that soap gives a fairly definite reading on the Abbe Refrac- tometer, the figure varying with the strength (fatty acid percentage) " The Use of the Refractometer in Chemical Industry," by Percival J. Fryer, F.I.C. The Chemical Age (March, 27 1920), 326. 20 OPTICAL METHODS of the soap. In mixing compound insecticides, of which nicotine is frequently an ingredient, it is often a difficult matter to know when such insecticides have been efficiently dissolved or distributed in the soap. If, however, samples are taken and examined as described above, the mixing is readily controlled in a perfectly satisfactory manner." The Refractometer in Physiological and Clinical Analyses. (See also Appendix II., p. 24). The refractometer has also been extensively used in physiological and clinical analyses. There are many references in Abderhalden's Handbuch der biochemischen Arbeitsmethoden, especially vol. i. pp. 568-583. The refractometer became from April i, 1898, an official instrument for examining butter in Germany. The ordinary Dipping Refractometer has a scale from n D 1-32539 to 1-36640, and reads by estimation to O'i of its scale division, corre- sponding to an uncertainty of 0-000037 in n D . Reiss employed this instrument for the determination of the protein in blood serum, in which it can be rapidly estimated on a few drops with an error of 0-2 per cent. Many other clinical uses were introduced by H. Strauss (stomach con- tents), by A. Strubell, J. A. Grober, and others (urine). In vol. viii. pp. 84-119 of Abderhalden's Handbuch further par- ticulars are given of the work on blood serum, fibrinogen determination, secretions and exudations, cerebrospinal liquid, stomach contents, milk, and for controlling the action of ferments and bacteria. Brailsford Robertson's method of serum analysis with the refractometer gives total protein, total globulin, insoluble globulin, and albumen. 1 Intending purchasers of a refractometer would do well, if in doubt, to mention the chief purpose for which it is intended to use it. In some cases the Abbe, and in others the Dipping (or immersion) refractometer is desirable. Full particulars of these instruments will be sent on applica- tion. In concluding this section a few words may be added concerning two other refractometers viz. the Pulfrich Refractometer, and the Rayleigh Interference Refractometer. The Pulfrich Refractometer. This instrument gives a higher accuracy than the Abbe, and also has a longer range (from 1-33 to 1-73), if both the interchangeable prisms 1 B.'^Robertson, Physical Chemistry of the Proteins. THE REFRACTOMETER 21 are supplied. It is for these reasons that this instrument is generally employed for determining the refractive indices of glass ; and for use with liquids where a greater accuracy than the Abbe is needed simul- taneously with a wider range than the Dipping. The Rayleigh Interference Refractometer again widens the possibilities of application of refractometry to chemical analyses. Many problems require a greater accuracy than can be obtained with the Abbe, the Dipping, or the Pulfrich Refractometers. In such cases the Rayleigh Refractometer for liquids, or the Rayleigh gas refract ometer, as the case may be, is indicated. One typical application of the former instrument is the determination of a sodium salt when present as an impurity in the corresponding potassium salt. By ordinary chemical methods, this is a laborious pro- cess, but with the Rayleigh Refractometer for liquids it is a matter of a few minutes only. A change from a 5 per cent, solution of potassium chloride to a 5 per cent, solution of sodium chloride gives, with careful working, 600 measurable units. Therefore, in a solution of 5 per cent, of the mixed salts an accuracy of 0-2 per cent, can be attained. The same type of instrument when arranged for gases (the Rayleigh gas refractometer) has of late years become of considerable importance in gas analysis. In this connection it is valuable in laboratories of chemical works manufacturing pure compressed gases, and in those requiring to control the composition of intermediate product gases, in Public Health offices and medical institutes for the systematic analysis of air, and in refrigeration works ; also in mines, in gas works, and to control the permeability of balloon and airship fabrics. Ranges and Sensibilities of the Different Types of Refractometer. RANGE, n D . SENSIBILITY. Abbe Refractometer 1-3000 to 1-7000 O'oooi Dipping Refractometer 1-32539 ,, 1-36640 0-000037 Pulfrich Refractometer - i'33OO ,, 1-7300 0-00005 (dispersions 10 0-00002) Rayleigh Liquid Refractometer 0-0063 0-0000021 (using a i cm. cell) Rayleigh Gas Refractometer- 0-000018 0-000000018 (using 100 cm. tubes) 22 OPTICAL METHODS Further printed information concerning refractometers may be obtained on application to Adam Hilger, Limited, 75a Camden Road > London, N.W.i, as follows : Catalogue description of the Pulfrich Refractometer, post free. Instructions for use of the Pulfrich Refractometer, containing description of the principles of construction, bibliography, etc. Supplied free with every instrument, or is. 8d. post free to any part of the world. Catalogue description of the Abbe Refractometer, post free. Instructions for use of the Abbe Refractometer, containing description of the principles of construction, use, bibliography, etc. Free with every instrument ; otherwise is. 8d. post free to any part of the world. Catalogue description of the Dipping or Immersion Refractometer, post free. Catalogue description of the Rayleigh Interference Refractometer, post free. Refractive Indices of Essential Oils, price i nett (already published). Refractive Indices of Oils, Fats, and Waxes, price i 55. nett (in the press). Refractive Index, Absorption, Wavelength, and Rotatory Power in. Relation to Molecular Structure ; an essay on the passage of electromagnetic radiation through matter. Price is. 8d. post free to any part of the world. LIST OF BOOKS DEALING WITH THE PRACTICAL APPLICATIONS OF THE REFRACTOMETER. LEACH. Food Inspection and Analysis, 1913. Chapman and Hall. LEWKOWITSCH. Oils, Fats, and Waxes. FRYER AND WESTON. Oils, Fats, and Waxes, vol. ii. ROTH UND EISENLOHR. Refraktometrische Hilfsbuch. E. J. PARRY. Chemistry of the Essential Oils. ALLEN. Commercial Organic Analysis. Vol. i., pp. 22, 126, 316. Vol. ii., p. 291. Vol. iv., pp. 243, 253. , THE REFRACTOMETER 23 SlMMONDS. Alcohol. ABDERHALDEX. Biochemische Arbeitsmethoden. Vol. i., p. 582 : Generalities. Vol. vi., p. 84-119 : Pathological Applications, etc. APPENDIX I LIST OF SUBSTANCES FOR SOLUTIONS OF WHICH WAGNER GIVES TABLES OF DIPPING REFRACTOMETER READINGS. Hydrochloric, nitric, sulphuric, phosphoric, boric, chromic, and perchloric acids. Potassium and sodium hydroxides. Ammonium, potassium, sodium, barium, calcium, strontium, ammonium, magnesium, ferric, mercuric, gold, and platinum chlorides. Potassium, sodium, and ammonium bromides. Potassium and sodium iodides. Potassium chlorate. Potassium, sodium, silver, and barium nitrates. Potassium, sodium, ammonium, magnesium, zinc, copper, ferrous, and aluminium sulphates. Potassium-aluminium, nickel, and manganese sulphates. vSodium sulphite. Sodium bisulphite. Sodium thiosulphate. Potassium and sodium carbonates. Potassium and sodium bicarbonates. Sodium phosphate. Sodium borate. Potassium chromate. Potassium bichromate. Ammonium sulphocyanate. Formic, acetic, oxalic, lactic, tartaric, and citric acids. Potassium, sodium, ammonium, and lead acetates. Potassium oxalate, potassium binoxalate. Methyl alcohol, ethyl alcohol. Glycerine, formaldehyde. 24 OPTICAL METHODS Cane sugar, dextrose and other sugars. Phenol, sodium salic^late, tannin. Furthermore, special methods of determination are given for calcium, magnesium, and phosphoric acid. APPENDIX II PRINCIPAL APPLICATIONS OF THE REFRACTOMETER TO BIOCHEMICAL METHODS REFERRED TO IN ABDERHALDEN, Biochemische Arbeits- methoden. Vol. i. p. 582. A. STRUBELL. " Uber eine neue Methode der Urin und Blutunter- suchung." Verhandlungen des XVIII Kongress fur innere Medizin zu Wiesbaden, 1900. Deutsches Archiv fur klinische Medizin. Vol. Ixix. p. 521 (1901). E. REISS. " Eine neue Methode der quantitativen Eiweissbestim- mung." Archiv fur experiment. Path. u. Pharmakol. (1904), 51, 18. " Der Brechungskoeffizient der Eiweisskorper des Blutserums." Hof- meisters Beitrage zur chemischen Physiologic und Pathologic (1904), 4, 150. " Klinische Eiweissbestimmungen mit dem Refraktometer." Verhandlungen der LXXVI Versammlung Deutscher Naturforscher und Arzte zu Breslau (1904), 2, 35. " Anwendungen der Refraktometrie in der Physiologic und Pathologic des Menschen." Bericht iiber die XV Hauptversammlung der Deutschen Bunsengesellschaft fiir angewandte physikalische Chemie (1908), 14, 613. // a solution of constant refractive index contains varying amounts of a second substance, this substance may be determined by the refractometer. The work of E. Reiss on blood serum is an instance of this problem. Blood serum from this point of view is a solution of varying quantities of protein in a liquid whose refractive index is constant or only subject to very small variations. If the refractive index of the protein-free liquid =n, that of the serum =n', then according to the table constructed by Reiss the protein content of the serum may be tead off from the difference n' -n, the average error is O'2 per cent, protein. Since the analysis requires only a few drops of blood, this refractometric method is of great clinical importance. H. STRAUSS. " Demonstration der refraktometrischen Blutunter- suchung." Deutsche med. Wochenschr. (1905), Nr. 2. APPENDIX 25 H. STRAUSS und B. CHAJES. " Refraktometrische Eiweissbestim- mungen an menschlichem Blutserum und ihre klinische Bedeutung." Zeitschr. f. klin. Med. (1904), 52, Heft 5 und 6. K. MARTINS. " Vergleichende Untersuchungen iiber den Wassergehalt des Gesamtblutes und des Blutserums." Folio haematologica (1906), Jg. 3, Nr. 3. H. STRAUSS. " Uber den Brechungsexponenten von Mageninhalten." Deutsche Arzte-Zeitung (1901), Heft 4. J. A. GROBER. " Quantitative Zuckerbestimmung mit dem Eintauch- refraktometer." Zentralbl. f. innere Med. (1900), 21, 201. E. RIEGLER " Die Refraktometrie des Harns im Dienste der Medizin." Atti del VI. Congresso internationale di chimica applicata (Rom, 26 April bis 3 Mai 1906), 5, p. 167 (Rom, 1907). Bericht in der Zeitschr. f. ange- wandte Chemie (1906), Jg. 19, 918. H. O. G. ELLIXGER. " Optische Bestimmung der Albuminmenge im Harn. Journ. f. praktische Chemie " (1891), 2 Reihe, 44, 256. H. STRAUSS und J. LEVA. " Uber eine neue Form der Motilitats- priifung des Magens." Deutsche med. Wochenschr (1907), Nr. 29. GERHARD SCHORER. " Uber refraktometrische Pepsinbestimmungen." Dissertation. Bern 1908. REFERENCES TO THE USE OF THE REFRACTOMETER IN PATHOLOGY AND FOOD ANALYSIS. " Methods of Refractometrical Investigation in Biology," by E. Reiss. Abderhalden's Arbeitsmethoden, vol. vi. pp. 84-119. Contents : 1. Abbe refractometer. 10. Stomach contents. 2. Milk fat refractometer. n. Milk. 3. Dipping refractometer. 12. Estimation of fat. 4. Blood serum. 13. Testing blue solution of milk. 5. Estimation of fibrinogen. 14. Testing milk serum. 6. Estimation of volume of blood 15. Estimation of milk sugar. corpuscles. 7. Excretions and secretions. 16. Testing the action of ferments and bacteria. 8. Cerebro-spinal liquid. 17. Further applications. 9. Urine. " Study of Protective Ferments with the Interferometer," by P. Hirsch, Abderhalden's Arbeitsmethoden, vol. vi. pp. 561-572. 26 OPTICAL METHODS APPENDIX III APPLICATION OF TEMPERATURE CORRECTIONS TO MIXTURES. Too little attention has been given in the past to the accurate deter- mination of temperature coefficients in refractometry, with the result that full advantage has not been taken of the accuracy of refractometers. Numerous attempts have been made to obtain temperature coefficients which can be applied indifferently to large classes of compounds. It must be remembered that any such rough and ready means of correction can only be applied where the working temperature is very near indeed to that taken as a standard. Every careful worker should ascertain the temperature coefficient of each substance in which he is interested. Furthermore, where there is a case of determining considerable amounts of one substance in mixture with another, it is necessary for him to obtain the temperature coefficient for the particular mixture. The trouble of preparing tables for the purpose is well repaid where the determination of a particular kind of mixture has to be frequently repeated. As an example of such a table, we give opposite, page 27, the re- fractive indices of various mixtures of petroleum spirit and benzol at different temperatures. This table has been prepared and is in use at a large chemical laboratory where work frequently requires to be done at high pressure. On this table the percentage by volume of petroleum in the mixture can be immediately obtained from the refractive index, as given on the Abbe Refractometer at any temperature from 10 to 30 C. This furnishes a ready means of sorting out samples. It is more expeditious than distillation or specific gravity determinations, and requires very little liquid for test. APPENDIX ^iO i LO.O,LO;O;LO,O i CN I CN -- r? i r? I Q I Q I o I co ! oo LO LO ' LO -LO LO ' LO 8O O O ! O : O o o ' o - o o rOOOnrOCNCNCNCN 8 I 8 I 8 8 I 8 v> >) O O ro -oo 'CO -CO '3 .3 8 8 S 8 ! 8 I R ' ' 8 b 8 l o 8 ,8-8 iS 8 8 CN CN CJ CN CN CS r\l CN I CM (N CS | fN fN:- CN CN i CN - c j SP SECTION IV THE POLARIMETER THE applications of the polarimeter in the sugar trade and in the essential oil industry are too well known to require more than the briefest mention. Starches can be determined by the polarimeter by Ewer's method and its later modifications ; thus the products of another im- portant industry may be controlled by this instrument. The relation between concentration and specific rotatory power has been determined for a number of substances with a very high degree of accuracy. We may particularly mention cane sugar, galatose, cocaine, lactose, maltose, glucose, camphor, nicotine. Synthetic camphor can be distinguished from natural camphor, and adulteration of inactive oils (such as almost all the fatty oils) with rosin or rosin oil is easily detected. Castor oil is the most important fatty oil possessed of optical activity. A review of the scientific applications of the polarimeter would require a treatise on stereochemistry, 1 but arising out of the purely scientific work is an interesting analytical method, an application of the fact of solvent influence to the analysis of mixtures. This method presents distinct advantages in certain cases in which chemical analysis is difficult to perform. Thus the quantitative estimation of benzene in cyclohexane is difficult and troublesome, but by taking advantage of the fact that benzene is almost without influence on the rotation of ethyl tartrate, whilst cyclohexane exerts a considerable depressing influence, the propor- tions of the two substances present in a mixture may be estimated within about 3 per cent, by the simple determination of the rotatory power of a mixture with a fixed proportion of the esters. (T. S. Patterson and A. Fleck, Journ. Chem. Soc. (1910), 97, 1772.) C. S. Annual Report (1910), p. 88. 1 See, however, Professor Frankland's Presidential Address to the Chemical Society, 1912. 28 THE POLARIMETER 29 The influence of the solvent is also clearly shown in an important paper by F. H. Carr and \V. C. Reynolds (J.C.S. (1910), 97, 1328) on the rotatory power of alkaloids. They found, inter alia, that hydrastine has a strong dextrorotation in 50 per cent, alcohol, while it is inactive in 95 per cent, alcohol and laevo-rotatory in absolute alcohol. These observations suggest that indirect determinations of inactive substances ^may often be possible, by measuring their influence on the rotation of an admixed active substance. It has been known for a long while that boric acid and also acetone and other substances have a powerful effect on the rotation of tartaric acid. For the two substances mentioned, there are excellent chemical methods available ; where this is not the case a polarimetric method might well be looked for. Some of the latest scientific applications of the polarimeter are highly specialised, and refer to rotatory dispersion, which is outside the scope of this book. Particulars of polarimeters for measuring rotatory dis- persion, and some of the chief modern applications can be forwarded on application. The special polarimeters made by us for rotatory dispersion are equally useful for ordinary polarimetric work, differing from the former only in the addition of apparatus for the production of a spectrum, Among the more recent applications of polarimetry in medicine may be mentioned, even if its value is disputed, the interesting study of the protective measures of the animal organism. " Die Diagnose der Schwan- gerschaft bei Mensch und Tier." Abderhalden, Bioch. Arbeitsmethoden, vol. vi. pp. 223-226. The following papers are important for bio-chemists : " The Develop- ments of micro-polarisation," by Emil Fischer, Abderhalden, Arbeits- methoden, vol. v. p. 572 ; also E. Fischer, " Synthese von Polypeptiden," Sitzungsber. der Berliner Akademie (1908), 552 ; cf. Chem. Zentralblatt (1908), 2, 315 ; Ferner iiber Mikropolarisation Berichte, (1911), 44, 129. (Fischer, by the employment of small tubes, was able to measure the polarisa- tion of volumes as small as 0*1 c.c.) Studying Biological Processes. The instrument has also been used for studying biological processes, in which case it has been useful in detecting changes which would other- wise escape observation. For instance, Abderhalden found that the serum of normal dogs retained its original rotation if preserved at 37 C. A change was only observed in cases of illness, particularly in the fevers due to infection. If a foreign protein or peptone was dissolved in a normal serum the original rotation of the mixture remains unaltered. 30 OPTICAL METHODS Quite different results, .however, were obtained with the serum of dogs which have been injected (subcutaneously, intravenously, etc.) with foreign protein or peptone. If such serum is mixed with protein, the mixture has no longer a constant rotation. Peptone formation was proved by dialysis, and thus the breakdown of protein by ferments con- tained in this serum or activated by the treatment. The detection of ferments not present or not active in the normal serum may be of great importance, although the numerical rotations measured may be of secondary interest. (Abderhalden, Arbeitsmethoden, vol. v. p. 575.) The applications of the polarimeter to the study of the action of ferments is also interesting. (See " Methods of Studying the Action of Ferments," by L. Michaelis, Abderhalden, vol. iii. pp. 32-35.) The above observations have been given as they may be the fore- runners of new applications of the polarimeter in medical research. Further printed information concerning polarimeters, etc., may be obtained post free on application to Adam Hilger, Ltd., 75 A Camden Road, London, N.W. i, as follows : Catalogue Section M : Polarimeters. Catalogue description of Saccharimeter. r.LASGOW : PRINTED AT THE UNIVERSITY PRESS BY ROBERT MACLEHOSE AND CO. LTD. UNIVERSITY OF CALIFORNIA LIBRARY OTHER PUBLICATIONS Refractive Index, Absorption, Wavelength and Rotatory Power in relation to Molecular Structure. By LUDWIK SlL'BERSTEIN. the passage of electromagnetic radiation through matter, pp. 17. Bound in stiff f>iU.-::r t 1'nce is. 6ci. ni:i.. I fixe is. 8d Tables of Refractive Indices, : K KANTHACK. : 'iited by J. N. GOLIJSMITH.. Ph.D., M.Sc., F.LC. Vol. I. ESSENTIAL Oll^S (contains over 1 500 measurements on over 500 oils), i nett. Vol, [I, M ; Press}. OILS, FATS, AND WAXES, ^3 5 s. nett. Vol. III. (in preparation). ^- l - TABLES. Elements of the Electromagnetic Theory of Light. r>> LUDWIK SILBERSTEIN. pp, 4.8 Crown 8vo. 43. nett Elements of Vector Algebra. By I.UWVJK STLBERSTEIN. ; 42. Crown