INVESTIGATIONS OF INFRA-RED SPECTRA Part I — Infra-Red Absorption Spectra Part II — Ixfra-Red Emission Spectra BY WILLIAM W. COBLENTZ WASHINGTON, D. C. : Published by the Carnegie Institution of Washington October, 1905. CARNEGIE INSTITUTION OF WASHINGTON Publication No. 35 FROM THE PRESS OF THE HENRY E. WILKENS PRINTING CO. WASHINGTON, D. 0. CONTENTS Part L— Infra-red Absorption Spectra. Page Chapter I. — Introduction 3-4 Chapter II.— Historical 5-8 Other investigations °"^ ^ Solvents ^ ^ " ^2 Ultra-violet 12-13 Total absorption temperature, etc i3-i5 Chapter III.— Description of apparatus and methods 15-33 The spectrometers ^" The prisms used '..... 16-17 Adjustment and calibration of apparatus i7 Adjustment of large spectrometer 18-21 The radiometer 21-23 The absorption cells 23-25 Source of radiation 25-26 Methods of observation 26-27 (Sources of error 27-29 The chemicals used 29-30 Problems attacked 30-3i Chapter IV.— Investigation with a quartz prism 33-38 Chapter V.— Investigation vi^ith a rock-salt prism 39-ioi Class I— Methane derivatives. Absorption spectra of gases 39-54 The absorption cell 41-43 Gases studied 43-54 Absorption spectra of liquids and solids 54-99 Halogen substitution products 54-50 Water and alcohol 56-59 Compounds containing the methyl group 59-65 Compounds containing sulphur 65-67 The mustard oils and sulphocyanates 67-69 The fatty acids 69-72 Petroleum distillates 72-77 Methylene series 73 Paraffin series 73 Acetylene series 73 Class II — Carbocyclic compounds 76-99 Benzene 76-77 Methyl derivatives of benzene 77-8o Various other benzene derivatives 80-83 Amido and nitro derivatives of benzene 84-86 Nitro derivatives 86-87 Phenols 87-89 Aromatic acids 89-90 III t f n IV CONTENTS. Part I — Chapter V — Class II — Continued. Page Aldehydes go-gi Terpenes 91-96 Pyridine group 96-97 Other cyclic compounds 98-99 Chapter VI. — ^General discussion of the spectra 101-119 Effect of structure 101-102 Effect of molecular weight 102-106 Effect of temperature 106-107 The effect of certain characteristic groups of atoms 107-110 Total absorption iio-iii Grouping of the spectra lii Characteristics of the spectra of carbohydrates 111-112 Occurrence of harmonics 1 12-1 14 Summary 115-118 Appendix I. — ^Sources of radiation 119-121 Appendix II. — The emission spectrum of the Hefner lamp 122-123 Appendix III. — 'Electrification of radiometer vanes 124-125 Appendix IV. — Absorption of solids in solution 126-128 Appendix V. — ^Water of crystallization 129-131 Appendix VI. — Indices of refraction of rock salt 132-134 Appendix VII. — Correction of the work of Julius 135 Tables : Table I. — Line spectra of gases 136 Table II. — Line spectra of liquids and solids 136 Table III. — Observed transmission of benzaldehyde 137-139 Table IV. — Observed transmission, using a quartz prism at constant minimum deviation 140-141 Table V. — Maxima of infra-red absorption band — gases 142 Table VI. — ^Maxima of infra-red absorption bands — liquids and solids 143-146 Table VII. — Observed transmission through gases 147-148 Tables VIIIa to VIIIf. — Observed transmission through liquids and solids 149-165 Transmission curves 166-285 Part II. — Infra-red Emission Spectra. Chapter I. — Introduction 289-294 Historical 289-294 Infra-red spectrum 289-292 Visible spectrum 292-293 Temperature, dissociation, etc 293-294 Chapter II. — Scope of present investigation 295-301 The radiometer 295-298 Table I. — Sensitiveness of different instruments 298 Table II. — Sensitiveness of radiometer 299 Other experimental details 300-301 Object of the investigation 301 Chapter III. — Infra-red emission spectra of metals 303-322 Carbon arc 305-306 Sodium 306-307 Lithium 307 Potassium 307-308 CONTENTS. V Part II— Chapter III— Continued. p^ ^ Infra-red emission spectra of gases in vacuum-tubes 308-319 Cleveite gas _ ^ ^ Water vapor 311-31^ ^y^'"°°^" 312-313 Oxygen ^^^ Carbon dioxide •? n--? 14 Carbon monoxide ^i4-'^r'; Ethyl alcohol -j_ Nitrogen '. ^' '.'.'.'.'.'.'.'.'.'.'.''.'. 316-318 A'"'"0"i^ 318-319 Radiation from a vacuum-tube when heated externally 319-321 Radiation from a black body when heated to 100° C 321-322 Temperature of gas in the vacuum-tube 322 Chapter IV.— Theoretical -j^, ,on Summary ERRATA Page 8, first line.— For CaSog, read CaSOi. Page 17.— For equation Zi = arc sin (sin <^ x/;^^— sin^ 4— cos sin"Z), read Zi = arc sin (sin <^ s/n^—sin^ Zg— cos <^ sin z,) Line 25. — For "emerge from," read "enter". Page 59, the ninth line from the bottom.— For CH3 — O— N, read CH3-0-N = 0. Page 109, line 7.— For "caymol" read "cymol". Page 135, in the third column from the end of the table, the "cor- rected" values are shifted one line. Read i^.id = c).44 /x, instead of I4.id = g.'j fx, etc. PREFATORY NOTE Originally it was intended to issue the report on "Absorption Spectra" and that on "Emission Spectra" separately. The publication of the former was unavoidably delayed, and, in the meantime, the latter being ready, the two are now combined into one volume, as Parts I and II. The arrangement of the separate parts has not been altered, how- ever, and the appendices, which one would naturally find at the end of the volume, occur at the end of Part I, to which they belong. PART I INFRA-RED ABSORPTION SPECTRA CHAPTER I. INTRODUCTION. The free vibration periods of molecular aggregations may be studied by means of their emission or absorption spectra. Emission generally implies a high temperature and uncertainty of the composition of the radiator, due to dissociation; hence it is more advantageous to study a substance at a low temperature by means of its absorption spectrum, since we then have a more definite knowledge of its composition. This is especially true of the highly complex compounds which are so abun- dant in organic chemistry. Although generally assumed that in gases the molecules have the advantage of a greater freedom of vibration, it will be show^n in the present paper that, starting with a simple liquid or solid compound, the free vibration of a given molecule or radical is not always seriously modified as the molecular complexity of the compound is increased. In fact, it has been found that there may be two sets of particles (or "ions"), which vibrate side by side, yet still retain the free period that they had when present, alone, in a simpler compound. Further- more, their free vibration periods, manifested as absorption maxima, occasionally seem to repeat themselves harmonically throughout the spectrum investigated. This presence of groups of ions, each group having its own free period of vibration, is in accord with present ideas of absorption and anomalous dispersion. Previous investigations of infra-red absorption spectra have never been extended farther than about y i^ for alcohols and to about lOju. for several other compounds. Now, it so happens that with the limited dispersion at our disposal (if that be the true reason), all carbohydrates investigated show a large absorption band between the wave-lengths 3.0 /x and 3.5 /i, and then there are, as a general rule, no marked bands until we arrive at 6 jx. Beyond this point, to the limit of the working transparency of rock salt at 15 ^u,, there are numerous large, well-defined bands. These facts were noticed in some preliminary work, when it became evident that, in order to gain a better knowledge of infra-red absorption spectra than obtained for the region up to 7/1, a very extended^ and systematic investigation of the spectrum beyond this point would be necessary, in magnitude not unlike Kayser and Runge's work on the emission spectra of the metals, for the optical region. ^Subsequently it was found that this was the conclusion of all the preceding inves- tigators in this field, who, however, never made further contributions to the work. 3 4 INFRA-RED ABSORPTION SPECTRA. However, in infra-red work we are confronted with a difficulty not encountered in the optical region, viz., the permanent registration of the spectra under investigation. In the optical region we can photo- graph the spectra without much delay. In the infra-red, photography has not yet been possible beyond 1.2 jx, and, instead of being able to procure the entire spectrum simultaneously by projecting it upon a pho- tographic plate, it is necessary to begin at one end of it, and with suit- able apparatus explore point after point until we reach the other end. As a consequence, the process of mapping infra-red spectra, whether due to emission or absorption, is, at best, a slow and tedious one. How- ever, as a compensation for all this, with the limited dispersion at our disposal, the spectra, which are a series of curves, are quite simple, so that the process of " measuring up the plates " is of no importance. After this we are ready to study the curves individually and collectively. The present investigation was begun while the author was a graduate student, at Cornell University, during the first three months of 1903. During that time the absorption spectra of 38 compounds were explored to 14 /A. It then became evident that, in order to gain a better knowl- edge of infra-red spectra than obtained at that time, a very extended and systematic investigation would be necessary. This was made possible under an appointment as Research Assistant by the Carnegie Institution of Washington during the year following June, 1903. The successful completion of this investigation has placed me under deep obligations to numerous persons, to all of whom I am very grateful for services rendered. In particular would I mention Prof. E. L. Nichols, who placed every facility of the laboratory at my disposal and did for me many other deeds of kindness, and his colleague, Prof, E. Merritt, whose advice and suggestions were also asked. Professor Nichols has also read several chapters of this manuscript. In the De- partment of Chemistry I am indebted to Prof. L. M. Dennis for the facilities of the gas laboratories, to Prof. W. R. Orndorfif for the use of chemicals, and, last but not least, to Dr. J. R. Teeple, whose advice and suggestions on the chemical side of this subject were constantly sought. To Profs. A. C. Gill and H. Ries, of the Department of Geol- ogy, I am also indebted for material placed at my disposal. It is a further pleasure to acknowledge the generosity of Prof. C. F. Mabery, of Case School of Applied Science, who presented me with 25 selected samples of pure distillates of petroleum, which could not have been pro- cured elsewhere. When one considers that it has taken seven years to prepare them, and that his collection is the most complete in exist- ence, the value of the accession of these samples to the list of compounds investigated becomes apparent. CHAPTER II. HISTORICAL. In 1882 Abney and Festing/ by means of photograph}', investigated the absorption spectra of 52 compounds to i.2fi, which was the Hmit of the sensibiHty of their photographic plates. They found many inter- esting relations among the absorption bands, which are of the greatest importance in the present investigation, since their work is in the region which is difificult to explore on account of the small dispersion and the numerous small bands which can not be detected without other measur- ing devices than photography. Certain radicals were found to have distinctive absorption bands at about 0.7 fi and 0.9 fj.. The ethyl, CH3, series gave a line at 0.74 /x and a second band at about 0.92 fi. Hence they decided that " when we find a body having a band at 0.74 fx and another beginning at 0.907 /x and ending at 0.942 fx we may be pretty sure that we have an ethyl radical present. In the aromatic group (e. g., benzene) the critical line is at wave-length A =0.867/1. If that line be connected with a band we feel certain that some derivative of benzene is present." They found remarkable coincidences in ammo- nium hydroxide, benzene, aniline, diethyl aniline, etc. The presence of oxygen in a compound was observed to sharpen certain bands. They call attention to the " remarkable fact that the 0.866 /x band of the sun should be the basic lines of the benzene series," and " it would not be at all surprising to find that A = 0.76 ix was another nucleus of a hydro- carbon group." The recurrence of these lines is not to be overlooked in considering others far in the infra-red, which, like those of Abney and Festing, are broadened in some compounds and narrow in others. They found a marked linear spectrum for chloroform, CHCI3, but could not detect these lines in carbon tetrachloride, CCI4. Hence, they con- cluded that the C and CI seem to have nothing to do with the linear spectrum observed in chloroform and, since two lines of water are coin- cident with those in the spectrum HCl, HNO3, H2SO4, NH^OH, etc., that hydrogen must be the cause of the lines. Again, in some of the compounds containing oxygen, certain lines coincided with the iodides, which were free from oxygen, and they " were forced to the conclusion that there must be some connection lAbney & Festing : Phil. Trans., 172, p. 8S7, 1882. 6 INFRA-RED ABSORPTION SPECTRA. between the one and the other, since such an agreement could not be fortuitous." From 1889 to 1890 Angstrom,^ using a bolometer and rock-salt prism and lenses, found the absorption of CO and COo. He also found the absorption spectra of methane, ethylene, ether, benzene, and carbon disulphide, several of the latter in the vapor and also the liquid state. H is work extended to about 8 [x, and showed that the maxima of liquids and their vapors are coincident. He found the absorption band of CO at about 4.6 [x, and that of COo at about 4.3 fx, which would indicate that the location of the maximum does not depend upon molecular weight. About this time Julius- investigated the absorption spectra of some 20 organic compounds. He used a bolometer, and a prism and lenses of rock salt. His work extends to 10 /j., which by his straight-line extrapolation from 5 /j, was supposed to be 16 /a. He used a cell hol- lowed out of rock salt, about 0.2 mm. in thickness, and found that about one-third of the substances, mostly alcohols, became opaque at 7 /x. He found that all compounds containing the methyl, CH3, group had an absorption band at 3.45 /x. The conclusion reached from this investiga- tion, which had been the most extensive thus far, was that the absorp- tion of heat waves is due to intramolecular movements ; in other words, the internal structure, i. e., the grouping of the atoms in the chemical molecule, determines the character of the absorption. He found that the chemical atom lost its identity in a compound ; i. e., the effect is not " additive," so that one can not foretell the absorption spectrum of the compound from a knowledge of the spectra of the constituent elements. This conclusion that the constituent atom loses its absorbing power in a compound is further substantiated by the fact that of the six com- pounds containing chlorine investigated by Julius not one showed the CI band found by Angstrom^ at 4.28 /x. Donath,* in 1896, using a quartz prism and a bolometer, investigated half a dozen compounds (mostly the essential oils), and concluded that the absorption is intermolecular and not intramolecular, as found by Julius. His work extended to 2.7 [x, and covered the region of " prac- tical significance," as he expressed it. For the aromatic compounds and the fatty oils he found maxima common at 1.69 /x and 2.2 /x. To my mind there is not sufficient evidence to draw conclusions like the above. However, after subsequent investigations by other observers on pure ^Angstrom : Ofversigt Af Kongl. Vetenskaps-Akadem. Forhandl., Stockholm, S. 549, 1889; S. 549, 1889; S. 331, 1890. ^Julius : Verhandl. Konikl. Akad. Amsterdam, Deel I, No. i, 1892. ^Angstrom & Palmer: Ofversigt Kongl. Vet. Akad., No. 6, p. 389, 1893. *Donath : Ann. der Physik (3), 58, p. 669, 1896. HISTORICAL. 7 compounds, the work is valuable in showing that the spectra of the pure compounds were not seriously disturbed when they existed in the impure state in the essential oils. Aschkinass/ using a flint-glass and a fluorite prism, and a mirror spectrometer, investigated the absorption spectrum of water to 8 fi. He found characteristic maxima located at o.yy fi, i.o fx, 1.25 /a, 1.50^1, 1.94 ju, 2.05 fi, 3.06 /x, 4.7 fi, and 6.1 fi. Paschen^ found the 3 fi band at 2.916 fi, 2.975 /*» and 3.024 ix, depending upon the thickness of the film, while the second large maximum was found at 6.06 /t. This was fol- lowed by the study of six alcohols, by Ransohoflf,^ his object being to learn the effect of the OH-group. He found marked bands at 1.71 /x, 3.0 fi, and 3.43 fx, and tacitly concludes that since this 3.0 fi band agrees with the one found by Aschkinass, it is a characteristic of the hydroxyl- group. Although the alcohols were " chemically pure," that is a differ- ent question from the one of having them " water free," which he does not consider. This question will be considered later on. It is of interest to note that the 1.71 /x band is harmonic with the one at 3.43 ^, just as is true of the water bands at 3 /* and 6 fi. The deepest exploration ever made into the infra-red is that of Rubens and Aschkinass,^ who, using a sylvite prism, found the absorption of CO2 and water vapor to 20 /a. They found CO2 transparent except at 14.7 fi, while water vapor has a series of maxima throughout the whole region. In 1899 and 1900 Puccianti^ explored a number of benzene deriva- tives by means of a quartz prism, mirror spectrometer, and radiometer. He found that all compounds, in the molecules of which carbon is com- bined directly with hydrogen, presented a maximum at 1.71 fi, while all the benzene derivatives have two other maxima in common at 2.18 ^u, and 2.49 fi. He found his results in agreement with the " hypothesis that the absorption depends upon the groups of atoms which exist in the molecule." The isomeric xylenes exhibit absorption spectra almost but not completely identical. A series of double refracting crystals have been investigated by Konigsberger.^ He used a fluorite prism and bolometer. The observa- tions of most interest here are that impurities changed the absorption curve, but not the maxima, and that wafer of crystallisation, in corn- ^Aschkinass ; Ann. der Physik (3), 55, p. 406, 1895. ^Paschen : Ann. der Physik (3), 53, p. 336, 1894. ^Ransohoff : Inaug. Diss. Berlin, 1896. *Rubens & Aschkinass : Astrophys. Jour., 8, p. 181, i8g8. ^Puccianti : Nuovo Cimento, 11, p. 241, 1900. *K6nigsberger : Ann. der Physik (3), Gi, p. 687, 1897. 8 INFRA-RED ABSORPTION SPECTRA. pounds like gypsum, CaSo3 + 2H20, has the same maxima at 1.5 /x and 2.0 /x as those found in pure water. It is unfortunate that he did not compare the 2.95 /x band of muscovite (which has the H and O chemi- cally combined) with pure water, since Paschen and Aschkinass dis- agree on the exact location of this band. This disagreement will be noticed in the present work. The most recent work in this line is that of Ikle,^ who used a fluorite prism and a linear thermopile io'.5 wide. He investigated the relation of absorption and thickness of the liquid used. He found no relation between the refractive indices and dielectric constants, nor could he detect a shift in the maxima for increase in molecular weight. Unfor- tunately the thermopile used was wide, and for this reason the location of the maxima is not very exact. OTHER INVESTIGATIONS. Kundt- investigated a series of solutions showing anomalous disper- sion. He observed that, in different colorless solvents, the absorption band of the solute is shifted toward the longer wave-lengths with in- crease in the refractive and dispersive pozver of the solvent. H. W. Yogel^ found that this is not true, but that the shift occurs in both directions. A spectrophotometric study of indophenols by Camichel and Bayrac* also shows that Kundt's law of the influence of the solvent on the position of the maxima of absorption bands does not hold, for the shift was from red to violet in an ortho-phenol, while upon substi- tuting a meta- in an indophenol the displacement was from red to violet or vice versa. Since then numerous investigations^ have afforded verifi- cation of the statement that the absorption bands shift in both directions. The most thorough and at the same time the most important investi- gation of solutions, for the optical region, is that of G. Kruss,^ who examined 64 different compounds dissolved in CS2, CHCI3, and C0H3OH. He observed that by the introduction of a methyl (CH3), etiiyf (C2H5), oxymethyl (OCH3), or carboxyl (COOH) group, or bromine, etc., in the molecule of the solute, the maximum of the absorp- ilkle : Phys. Zeit., 5, p. 271, 1904. ^Kundt : Pogg. Ann., 1871-1872, and Weid. Ann., 4, p. 34, 1878. ^Vogel, H. W. : Bed. Monatsber., p. 409, 1878. *Camichel & Bayrac : Jour, de Phys., Ill, 11, p. 148, 1902. ^G. Kruss : Zeitt. f. Phys. Chem., 2, p. 372, 1888 ; ibid., 18, p. 559, 1895. Schiiltze : Ibid., 9, p. 109, 1892. Grebe : Ibid., 10, p. 673, 1892. G. Kruss : Ber. der Deutch. Chem. Gesell, 22, p. 2065, 1889. Kriiss & Oecomonides : Ibid., 16, p. 2051, 1883; 18, p. 1426, 1885. Bernthsen & Goske : Ibid., 20, p. 924, 1887. Liebermann & Kostanski : Ibid., 19, p. 2327, 1886. HISTORICAL. 9 tion band is shifted to the red, while by the introduction of a nitro (NO2) or an amide (NH2) group the band is shifted toward the violet. Subsequent writers on this subject always mention the shift toward the red zvith increase in molecular weight, but rarely mention the fact that there is also a shift in the opposite direction; so that, unless one is familiar with the original work, the quotation is misleading. In quoting such a complete investigation which records two well-defined series of phenomena, apparently opposed to each other when considering the ques- tion of molecular weight, it seems highly desirable to have the complete observation rather than the part which fits the particular problem under investigation. This is especially desirable in work like that of Ranso- hoff, who thought that a small sharp band found at 4.9 fx. in CH3OH was shifted to 5.2 /x in C2H5OH, " which would be an example like that of Kruss." He found no shifting for larger bands. The following is what Kriiss (loc. cit.) observed for indigo: I Indigo in CHCI3 "^^ max. at 0.6048 jn Shift to red. . •, Methyl indigo in CHCI3 ^- max. at 0.61917 [ Ethyl indigo in CHCI3 /" max. at 0.6526 „, .J ., I Indigo in CHCI3 / max. at 0.6048 Shift to violet -( ^,. . ,. . ^,,^, . Nitro-indigo in CHCI3 /- max. at 0.5858 Water solution. Alcohol solution. Fluorescein / max. 0.494 f-i 0.5048 0.5159 / max. 0.4808 // 0.5094 0.5251 Dibrom fluorescein Tetrabrom fluorescein . . This is a shift of 0.0055 P- P^^ atom of Br, which proportionality was found not to hold true. E. VogeP found that the occurrence of chlorine in the meta-position in the " Carboxylrest " in fluorescein shifts the absorption maximum more toward the red than when it is in the ortho-position. As with Kriiss, he found that the more hydrogen atoms that have been substi- tuted, the greater is the shift of the maximum of the absorption band, but that the shift is not proportional to the number of hydrogen atoms substituted. The fluorescence and absorption of dyestuffs in solution and in solid gelatin was investigated by Stenger.^ He concludes that the absorption of light depends primarily upon the size of the physical molecule, and it is only when a change in the aggregation conditions, or in solving ^E. Vogel ; Ann. der Physik (3), 43, p. 449, iJ ^Stenger : Ann. der Physik (3), 33, p. 578, 18J lO INFRA-RED ABSORPTION SPECTRA. process, is accompanied by a change in the physical molecule that a change occurs in the absorption spectrum ; and, vice versa, each change in the character of the absorption spectrum is connected with a change m the physical molecule. Hence, so long as the physical molecule has not changed, Kundt's law, of shifting toward the red, holds. In a solid state, i. e., at a low temperature, the physical molecule is more complex. As a test of this last question we have the observations of Wiedemann^ on iodine solutions. Iodine in CS2 is violet, similar to the gaseous condition, while in alcohol it is brown, like melted iodine. The latter is the more complicated. Hence, if this assumption be true, we would expect the violet carbondisulphide solution to turn brown on cooling. This is the case. He also observed that cold brown solutions of iodine in stearic and oleic acid became violet on heating to 80°. The absorption spectra of the alkaloids has been studied by Hartley^ and others, and a remarkable similarity has been observed in their absorption spectra. After examining 30 alkaloids they think that, " as a general rule, those which agree closely in structure give similar absorption spectra, while those which differ in essential points of struc- ture give dissimilar spectra. Most of these compounds have high molecular weight, and changes may be effected in their molecules with- out alteration of their spectra, which, in substances of lower molecular weight, would be attended by wide differences." Hence the alkaloids differ only in details of structure. This effect will be noticed, in the present work, for petroleum distillates. The effects of dilution, of temperature, of acids, and of different solv- ents upon the absorption spectra of solutions of didymium and erbium salts have been investigated by Liveing.^ He found that the spectra of the different salts of the same metal in a dilute condition are identical. Ostwald interprets this by saying that the spectrum, common to all the salts of the same metal, is due to the metallic ions. All the tests applied by Iviveing contradict this assumption. He explains his observations by assuming that, in solution, the molecules are ruptured by collisions, but immediately recombine. Increased temperature and concentration mean more frequent encounters amongst the molecules and more frequent ruptures, which are counterbalanced by the more frequent encounters of the parts. These effects will compensate each other and leave the average number of absorbing parts of molecules constant, under changes of temperature or concentration as observed. In other 'Wiedemann : Ann. der Physik (3), 41, p. 299, 1890. '^Hartley : Phil. Trans., Part II, 47, p. 691, 18S5 ; Hartley & Dobbie : Phil. Trans., 77, p. 846; Dobbie & Lawder : Chem. Soc. Jour., 83 and 84, pp. 605 and 626, 1903. ^Liveing : Trans. Cambridge Phil. Soc. xviii, p. 298, 1900. HISTORICAL. II words, the absorption depends upon the form of the internal energy of the vibrating mass, i. e., on its structure. LaubenthaP found for the absorption spectra of solutions of the chlorides of the alkali metals that the absorption bands shift toward the red with increasing atomic weights, and that the shifts are propor- tional, so that the ratio between the wave-lengths of the two absorption bands of each spectrum is constant for each group of metals. The absorption spectra are thus brought in line with the emission spectra, and also the densities and melting points, as having their origin in the same fundamental cause. In addition to these observations we have those of Stockl- on solutions of fuchsine, cyanine, and iodine, in which the maximum of the absorp- tion band depends upon the solvent. This corresponds to my own work on iodine in solution.^ To crown all this we have the " electromagnetic theory of selective absorption in isotropic nonconductors " of Planck,* in which Stockl's observations are applied to the question of intra-molecular resonance. This is of great interest, since it is the first theoretical recognition of the possible unification of the selective absorption of a solution and the selective absorption of the solvent. Whether there is a distinction between the two is to be discussed later on. It will be sufficient to notice that in my work on iodine solutions (loc. cit.) the selective absorption of iodine (solute) is lost at 7.3 /x in the infra-red, where solid iodine has a large absorption band, while the band continues in the visible spectrum, just as though there was a resonance of small particles in the optical region which is out of tune in the infra-red. ( See, how- ever. Appendix IV, on transparency of solutions in the infra-red.) SOLVENTS. In the visible spectrum the absorption of but few solvents has been investigated. Schonn^ investigated methyl, ethyl, and amyl alcohol for columns of liquid 1.6 to 3.7 meters, and found a shifting of the maxima to the red, as follows : Maxima. I. II. III. CH3OH C2H5OH CsHuOH 0.6430 •6515 .6591 0.6428 .6328 .6362 0.5591 .5627 'Laubenthal: Ann. der Physik, 7, p. 851, 1902. ^Stockl: Inaug. Diss., Tubingen, 1901. ^Coblentz : Phys. Rev., xvi, i, 1903. *Planck : Sitzungsber. d. Akad. d. Wiss., Berlin. Nos. 22 and 25, 1903. ^Schonn : Ann. der Physik (2), 6, p. 267, 1879. 12 INFRA-RED ABSORPTION SPECTRA. The amyl alcohol was yellow in color for the i. 6-meter cell, but color- less for a small thickness. ULTRA-VIOLET. Soret and Rillet/ using a fluorescent ocular, examined different thick- nesses of ethyl, butyl, and amyl nitrates, and found that, to completely absorb certain cadmium lines used, it required the least thickness of amyl nitrate, while greater thicknesses were required for the butyl and ethyl, which is interpreted that with increasing molecular weight the absorption increases toward the longer wave-lengths. The benzene derivatives have been investigated by Pauer,^ by photography. The benzene spectrum consists of four groups of bands. In the xylenes and aniline the original benzene spectrum seems crowded together and shifted toward the visible spectrum. For the liquid state the bands lie farther toward the visible spectrum than for the vapor phase, so that " if we consider each liquid body as a solution in itself, then Kundt's law holds for shift of the band with increase in molecular weight of the solvent." This seems to be giving the broadest possible interpretation of this law. In toluene the substitution of a CHg group has not merely destroyed the arrangement of the lines found for benzene, but the new spectrum consists of a series of double lines, not regularly distributed. The isomeric xylenes are of the most interest. The ortho has scarcely any bands, while the meta has three and the para has five bands, which show no regularity. The results show that the arrangement of the atoms has a great influence on the absorption spectra, just as previously found by Hartley.^ In the benzene spectrum Pauer observed several groups of lines in which the " constant difference " of the vibration numbers is 98. He observed that the absorption spectra of vapors show lines or groups of lines which become bands in the liquid condition. Traces of benzene in the air were sufficient to show the four strong groups of lines. He also found that for nitro- or amido-benzene the bands shift toward the longer wave-lengths (to the red), while Kriiss found exactly the oppo- site (to the violet). The question whether this is a real shift will be discussed in connection with the present investigation. Martens* examined the spectrum of a number of transparent non- conducting elements (e. g., C. P. S. CI. Se. Br. I.) and found that the wave-length corresponding to the principal absorption band in the ultra- *Soret & Rillet ; Compt. Rend., 89, p. 147, 1879. ^Pauer : Ann. der Physik (3), 61, p. 363, 1897. ^Hartley : Phil. Trans., 170, p. 270, 1879. *Martens : Ann. der Physik (4), 6, p. 603, 1901. HISTORICAL. 13 violet is approximately proportional to the square root of the atomic weight; and that this wave-length is independent of the state of aggre- gation of the element, and is unchanged in solution. The most recent investigation in this region is that of Magini,^ who examined some of the compounds already mentioned under the work of Pauer/ as well as 12 isomers, such as resorcin, hydroquinon, the oxybenzoic acids, and the pthalic acids. Like Pauer, he observed a shift toward the longer wave-lengths when an amido or carboxyl group is substituted for a hydroxyl group. For total absorption the arrange- ment of isomers is mcta, ortho, para. Para-compounds have greater effect in forming entirely different absorption spectra. He concludes that the double bond causes the absorption in the ultra-violet, w4iile it makes no dift'erence, in the infra-red, whether the bands are double or single. The double bond can be thought of as changing the elasticity and cohesion in the molecule in such a manner that the molecule will be resonant with ultra-violet light. He finds that the effect of absorption is greatest when the replaceable groups are joined to opposite carbon atoms of the benzene ring, and also when the molecule is symmetrical. Isomers having a double bond have dissimilar absorption spectra. As a whole, he thinks the absorption is caused chiefly by the molecular configuration on the one hand, and by the nature of the compound on the other. TOTAL ABSORPTION TEMPERATURE. ETC. The change in diathermancy of liquids with temperature was investi- gated by Dechant,^ who found that the transparency of mica did not change for a rise of 120°, while the diathermancy of water decreased. Konigsberger* found that, in solid selective absorbing media, a rise of temperature shifts the absorption curve toward the long wave-lengths, while for the metals the absorption is constant. The work of Hagen and Rubens^ on the metals, in which a film so thin that it transmits light but is uniformly opaque in the infra-red, can not be discussed here. Neither the work of Aschkinass and Schafer,® v/ho determined the dielectric constants of several compounds for electrical waves, and found that, with increase in molecular weight, the maximum of reso- nance shifted toward the shorter resonators, which, if properly inter- 'Magini : Phys. Zeit., 5, p. 69, and p. 145, 1904. ^Pauer, loc. cit. ^Dechant : Wien Ber., in, p. 264, 1902 ; Beiblatter, 27, i, 1903. *Konigsberger : Ann. der Physik (4), 4, p. 796, 1901. ^Hagen & Rubens : Ann. der Physik (4), 8, p. 432, 1902. ^Aschkinass & Schafer : Ann. der Physik (4), 5, p. 489, 1901. 14 INFRA-RED ABSORPTION SPECTRA. preted, indicates a shift toward the long wave-lengths, as demanded by Kiindt's law. Finally, we have to consider the total absorption of an extended series of carbon compounds by FriedeP and by Zsigmondy,^ using Tyndall's method, substituting a bolometer for a thermopile. They found that the transparency of a compound increases if, other conditions being equal, H, O, OH, or N are replaced by S, or halogens. Absorption does not depend upon the size of the molecule. The carbon atoms have little influence on absorption. As a whole the absorption of radiant heat depends upon the manner of the bonding of the atoms in the mole- cule, as well as upon the kind of compound in which the atoms of an element are united. If an amido group occurs in a carbohydrate it outweighs the hydroxyl, e. g., an NH group substituted for an H atom in benzene reduces the intensity to one-fifth its original value. This opacity of nitrogen compounds will be noticed in the present work, where it will be shown that not all nitrogens are highly opaque. ^Friedel : Ann. der Physik (3), 55, p. 453, 1895. ^Zsigmondy : Ann der Physik (3), 49, p. 531, 1893 ; 57, p. 639, 1896. CHAPTER III. DESCRIPTION OF APPARATUS AND METHODS. The work involves two distinct kinds of activity, viz, mapping the spectra, by means of a series of curves, and studying them. The spec- trum is produced by means of a rock-salt prism mounted on a mirror spectrometer. For a source of radiation the " heater " of a Nernst lamp is mounted before the collimator slit. The distribution of the energy in this spectrum forms a smooth, continuous curve. If we inter- pose a hollow cell of rock salt, containing some compound, between the " heater " and the collimator slit, the distribution of the incident energy in the resulting spectrum will no longer form a smooth, continuous curve. The intensity will now be found to rise and fall, forming a series of maxima and minima. These maxima and minima are char- acteristic of each compound interposed ; and it is with these curves that we are concerned in this work. To obtain these curves very narrow portions of the spectrum are successively projected upon a device (a Nichols radiometer), which is sensible to heat radiation. This is a slow process, but it is the most successful yet devised. The number of organic compounds is so large, while many are so mtimately related, that, after the preliminary work, it became evident that, in order to gain more definite information in regard to absorption spectra than had been obtained by previous investigations, a large number of compounds would have to be examined. This has been done. The absorption spectra of at least 130 compounds of hydrogen and carbon have been explored, the majority of them to 14 /x and 15 ix. About 30 compounds were exam.ined twice, while 19 were explored to 2.5 IX, using a quartz prism. They include solids, liquids, and gases, and belong to the principal groups of organic compounds. The field is large, while many of the compounds ordered could not be obtained in commerce; hence it was necessary to proceed in this manner. It really amounts to a preliminary survey of the whole field. Subsequent work must be a detailed study of individual compounds belonging to a particular group, now that we know which groups are the most promising of results. 15 1 6 INFRA-RED ABSORPTION SPECTRA. THE SPECTROMETERS. The general arrangement of the reflecting spectrometer is shown in fig. I. The rays enter the instrument through the bilateral slit F , are brought to parallelism by the concave mirror M^, traverse the prism P, and after being collected by the concave mirror M^, are brought to focus on the slit F^. The portion of the spectrum which passes through F^, which is in the vertical focus, falls upon the exposed radiometer vane. The spectrometer stood in an inner basement room with cement floor, and the radiometer deflections were read through a hole in the wall. As a general rule, however, the temperature of the rooms was fairly constant, so that no difficulty was experienced in making the observa- tions with the door open. The spectrometer and radiometer were mounted upon a large slab of slate, which lay upon a large table. Although the latter stood upon the cement floor, there was rarely any trouble with earth tremors. A 35 cm. focal length mirror spectrometer was used for the explora- tions to 15 /x.. With this instrument the image of the collimator slit was slightly curved. For the region up to 9.5 jx the spectrometer slits were 0.4 mm. in width, while beyond this point the collimator slit was gradu- ally widened to i mm. at 15 /x. A very considerable portion of the work was repeated to y jx, using a pair of i-meter focal length, 20 cm. aper- ture mirrors mounted on a large spectrometer. The figuring of these mirrors was excellent and the slit image was perfectly straight. These mirrors gave about twice the dispersion of the smaller spectrometer, and, although the concentration of the energy in the spectrum was greatly reduced, which resulted in small radiometer deflections, the measure- ments obtained with them were so uniform in their agreement that for the region from 3.43 to 6.86 /* the results are more trustworthy than those obtained with the smaller apparatus. With this larger apparatus the spectrometer slits were 2' of arc, while with the smaller they were 4' of arc on the spectrometer circle, so that for the large spectrometer the dispersion was comparable to that of fluorite. With it numerous bands were resolved from 6 to 7 /a, but only occasionally were small bands found in the transparent region from 4 to 5 /t, to be mentioned later, while the 3 to 3.5 /x region was sometimes found complex. THE PRISMS USED. For the region from 0.8 to 2.5 /la a 60 mm. quartz prism, having a refracting angle of 60° i' 32", and the small spectrometer were used. For the region beyond 2.5 jn to 15 ^u, a 70 mm. rock-salt prism was ^~~^ Fig. I . -Inner room containing radiometer R, spectrometer and prism P, exhaust pump G, lamps and scale L, and telescope T. C is the holder for the absorption cell. S is the shutter, H the Nernst lamp-heater. Fig. 1 A.— Large spectrometer with Nernst heater, /* , to the right, and radiometer, r, to the left. The gas-cell holde and glass cells are shown at 9; Ceissler pump in the rear. Photograph taken through doorway of inner room. DESCRIPTION OF APPARATUS AND METHODS. 1/ employed. Its refracting angle was 59° 57' 43" for the center of the faces. The prism was securely and accurately mounted upon the spec- trometer table, and a dish containing phosphorus pentoxide was placed over it. When not in use it was covered with a glass bell- jar. as a further protection against moisture. The whole apparatus was inclosed in a tin box to exclude air cur- rents, moisture, and stray radiation from the radiometer and prism (fig. I a). With the small spectrometer the prism was set at minimum deviation of the sodium lines by means of a Gauss eye-piece, which was placed in the focal plane of the second mirror. ADJUSTMENT AND CALIBRATION OF APPARATUS. Since the radiometer must remain stationary, the successive portions of the spectrum must be projected upon its vane by one of two methods, viz, by rotating the prism, or by rotating the collimating arm of the spectrometer. This involves the problem of keeping the prism at mini- mum deviation. With the small spectrometer it was not possible to have the different rays pass through the prism at minimum deviation by rotating the prism table as in the mirror-prism device described by Wads worth. ^ An automatic attachment, like that used by Paschen and others, continued to give trouble, so it was discarded and the colli- mating arm was moved, while the prism remained stationary after set- ting it for minimum deviation of the sodium lines. As a consequence, when the collimating arm is revolved about the prism, the different wave-lengths emerge from the prism at a variable angle, which is, of course, no longer the minimum deviation angle, except for the sodium lines. This makes the computation of the dispersion curve more com- plex, since we can use the minimum deviation formula i=sin(^±i) 2 \ 2 ^ n sm 2 V 2 only for the sodium lines. The appropriate formula as used by Asch- kinass^ is i\ ^arc sin (sin cf> \/n- — sin^ i^ — cos sin (2) w^here z\ is the angle of emergence, is the angle of the prism, n is the index of refraction, and /, is the angle of incidence for the new wave- length. The deviation, 8, from the sodium line is then, 8=^0 — ii, where the sodium line is at minimum deviation. This formula is very un- wieldy, and I preferred computing the calibration (dispersion) curve from the following simpler relations, from which the aforesaid equa- iWadsvvorth : Phil. Mag. (5), 38, p. 346, 1894. ^Aschkinass : Ann. der Physik (3), 55, p. 406, 1895. 1 8 INFRA-RED ABSORPTION SPECTRA. tion can be derived. It has been noticed that as the colHmator arm is rotated the different wave-lengths must emerge from the prism at a con- stant angle in order to be reflected into the radiometer slit. Hence we need to compute this quantity, sin i, but once, viz, for the sodium lines. Reverse the rays. Then from the relations, sin i{=n sin r^, 1'^=^ — r^, and sin i. = n sin ra, we obtain the value of ig very readily, while the deviation is 8 = j'2 — c, where c = -\- m — i, where (f> is the angle of the prism, m is the angle between the mirror and opposite face of the prism, i is the computed angle of incidence of sodium light, for mini- mum deviation, and p is the required rotation of the prism table. All we have to do, then, is to turn the prism-mirror table so that the beam from the slit returns on its own path and forms an image of the slit upon the slit. Then rotate the prism table through the angle p, and the beam will be reflected from the mirror and enter the prism at the proper angle of incidence for minimum deviation. A second method of minimum deviation adjustment used by Stewart^ was also tried, and the results obtained agreed with the preceding. Using these three methods, viz, the non-minimum deviation (con- stant emergence, as with the small spectrometer) method, and Wads- worth's mirror-prism device for constant mimimum deviation, which was adjusted by the two methods just mentioned, the results obtained were in excellent agreement for the large spectrometer. But when these results were compared with the work done with the small spec- trometer it was found that all the maxima, of a certain number of com- pounds, were shifted by a constant slight amount toward the longer wave-lengths, which would have been unimportant had it not been for the question of harmonics. Now, the work with the small spectrometer had also been found consistent. This was established by the repeated examination of certain sharp emission and absorption bands which were equivalent to the comparison spectrum in the optical region. After weeks of intercomparison only one explanation of this discrep- ancy could be found, viz, that on account of the slight curvature of the image of the slit found the small spectrometer, in projecting the sodium lines upon the radiometer slit, which served as a " zero reading " in testing the constancy of the adjustments and as a starting point in cali- brating the apparatus, the whole spectrum was thrown forward about 30". Consequently every spectrometer circle reading represented a wave- length situated 30" farther toward the infra-red than that indicated on the calibration curve. Applying this correction to about a dozen com- pounds, the location of the maxima, as found with the small apparatus, coincided exactly with that found v/ith the large spectrometer. In the discussion of the curvature of the slit, in Kayser's Spectroscopy it is shown that such a shifting toward the long wave-lengths, as noticed above, is likely to occur. ^Stewart : Phys. Rev., xin, p. 257, 1901. (The details of his method of adjust- ment were given the author in a private letter.) DESCRIPTION OF APPARATUS AND METHODS. 21 This is the largest spectrometer }et set up for emission and absorp- tion work. Greater sensitiveness of the radiometer would be necessary in order to be able to use a spectrometer of greater dimensions. THE RADIOMETER. The construction of the radiometer is shown in fig. 3, which is a section of the instrument at right angles to the one facing the spectro- meter. It was built according to the design of Nichols.^ The outer ^Topump ^^^^ ^^ ^ block of bronze, 5 by 5 by 10 cm., with an axial boring 3 cm. in diameter. The top was made from the neck of a large round bottle, into which was fitted a glass tube, which led to the exhaust pump. In series with the radiometer and pump was placed a tube containing gold foil on cotton to absorb the mercury vapor, which was found to electrify the vanes and thus cause one of them to adhere to the window. As a further precaution against electrification, the inner rock-salt window was partly cov- ered with tinfoil, which was in contact with the outer case. (See Appendix III.) There were two lateral borings in the outer case ; the one in front of the vanes was 2 cm. in diameter and admitted the energy to be measured upon the exposed vane ; the one in front of the mirror, not shown in the figure, was long and narrow and was used to read the deflections of the Th im ,^- R, r (i77Z suspension. This window was closed with plate glass. The w i n d o w s were placed upon ground surfaces of the casing, which had been covered with a mixture of beeswax and tallow. Around the edges of the win- dows was a layer of beeswax covered with shellac. After the latter had dried there was no appre- ciable leaking. Fig. 3. V=vanes-, M=miTTor; Ri,R2=rock-saIt windows 'Nichols, E. F. : Phys. Rev., iv, p. 297, 1897. 22 INFRA-RED ABSORPTION SPECTRA. The vanes were suspended from a cross-bar by means of a very fine quartz fiber, made by blowing it out in an oxyhydrogen flame. The vanes were of very thin mica, approximately 15 by 2.5 mm., and were held together by means of glass fibers, the distance between them being about 4 mm. On the line midway between the vanes was fastened a glass fiber, nm, which forms the axis of rotation, and carries near its lower end a bit of mirror, made by silvering a thin microscope-glass cover, which was then cut into areas of about 2 by 3 mm. The vanes were blackened by burning camphor gum. To cause the lampblack to adhere better to the mica a very thin coating of beeswax was first applied by means of a hot wire. When the vanes were held above the burning gum the wax softened and the soot was deposited in an even layer. The vanes, mirror, and fiber were fastened by shellac. The total weight of the suspension was about 10 mg. This was rather heavy, but since no other precautions were taken against earth tremors, it was more serviceable than a second one which was lighter. Gener- ally there was no difficulty in reading to tenth millimeters on a scale situated 1.4 meters from the radiometer. The sensitiveness of the instrument depends very much upon the nearness of the vanes to the inner rock-salt window. This distance was regulated by the tripod screw, which was on the rear side of the radiometer base. The pressure for maximum sensitiveness, measured with a McLeod gage, was from 0.05 to 0.08 mm., while the period was such that the maximum of the deflection was reached in from 30 to 45 seconds, so that it usually required about 1.5 minutes to make a reading. Such a slow period would militate against the use of a radiometer were it not for the fact that the readings are always trustworthy, so that nothing is gained by repeating them. The deflections also depend upon the dispersion, and the kind of radi- ator used. For the small spectrometer, using acetylene, the maximum deflections were about 40 cm. on a scale 1.4 meters distant, while with the " heater " of a Nernst lamp they were from 10 to 15 cm. But the latter gives out a stronger radiation from 4 to 10 fi, and is therefore the more satisfactory. The radiation from this heater, as well as from other clays, will be discussed in Appendix I. The advantages of the radiometer over the bolometer have been dis- cussed by Nichols.^ The chief difficulties to be experienced in using a radiometer is shifting of the zero, " drift," which was avoided by inclos- ing the radiometer case (see fig. i) in a tin box, packed with wool. This avoided all sudden changes of the " drift," which was then only in ^Nichols, loc. cit., p. 302. DESCRIPTION OF APPARATUS AND METHODS. 23 one direction as the temperature of the room increased, due to the pres- ence of intense radiation from the Nernst " heater." The drift lasted only a short time, in beginning a series of observa- tions, and after that the shifting for a single reading was rarely greater than 0.1 mm., while generally for a period of 10 to 12 minutes no shift- ing at all could be detected. As a consequence the deflections were always trustworthy, even at 14 fi, where they were only from 2 to 3 mm,, while in passing through absorption bands in this region they were often only from o.i to 0.2 mm. Since the rate of increase of a deflection with the time of exposure of the vane to heat follows a logarithmic curve, which increases very rap- idly at first, then more and more slowly, and finally being asymtotic, one can make a reading, in the region where the deflections are small, in a shorter time than the actual period of the instrumient, thus avoiding a possible drift. The radiometer slit was mounted upon the spectrometer arm, so that if any shifting occurred in the relative positions of the radiometer and spectrometer^ it would not affect the location of any point of the spec- trum with respect to the slit. This is of importance in measuring the distribution of the energy in the spectrum of the radiator, but in this work we are dealing with the ratio of the intensity of the radiation which has passed through an absorbing medium to the intensity of the direct radiation, in any region of the spectrum, so that we are not con- cerned with the change in the sensitiveness of the radiometer, which varies from day to day, and it was sufficient to know that the sensitive- ness did not change while determining this ratio. This is more likely to be afifected by variation in the intensity of the source of radiation, which will be noticed in discussing that subject. THE ABSORPTION CEI.LS, One of the chief difficulties to contend against in this work is to obtain pure chemicals, and it is of the greatest importance to prevent contami- nation while investigating them. To this end a suitable absorption cell had to be devised for containing the liquids. The cell walls were made by splitting the rock-salt crystal parallel to a cleavage plane. This gave thin plates that were quite plane, smooth, and of a finer polish than could be obtained by hand polishing. Furthermore, the surfaces are not so easily attacked by moisture. One form of cell, used in examining the low boiling-point liquids, consisted of a fine wire, from o.i to 0.3 mm. in diameter, which was covered with Le Page's glue, pressed between ^Stewart, loc. cit., experienced this difficulty. 24 INFRA-RED ABSORPTION SPECTRA. r- —J r r L lY ^' i_ _i LI Fig. 4a. =^ two plates of rock salt and permitted to dry. The substances investi- gated did not attack the glue. After filling the cell the top was covered with tinfoil, over which was melted a layer of beeswax. A second form of cell, used for liquids boiling above i io°, is shown in fig. 4A. The plates were about 4 by 2.5 by 0.4 cm., split from the natural crystal. Between the common cleavage plane of the two plates was placed a washer, iv, of tinfoil, o.oi mm. in thickness, while around the edge was placed a strip of pure tin, t, 0.1 mm. in thickness, to prevent evap- oration. This form of cell is much better than that used in previous in- vestigations, in that it can be thor- oughly cleaned, while the washer is discarded for each compound. An- other advantage in using this form of cell is that the tinfoil assumes any small irregularities in the surface of the plates, so that one knows the actual thickness of the film more accurately than in previous forms of rock-salt cells. The cell was filled by placing the washer upon one plate, placing several drops of the liquid upon it, then covering it with the other plate. The tin edge was then put on the out- side and pressed close to the plates. The cell, c, was then mounted in a constant position upon its holder (fig. 4b), which consisted of a heavy block of wood with a narrow opening cut through it. Below the cell was a clear piece of rock salt, r, which was used to eliminate the absorption of the cell, thus enabling one to obtain the transmission through the liquid di- rectly. Furthermore, by using this arrangement no radiation except that which passed through the cell, or through the rock-salt plate, could enter the spectrometer. In fig. 4A is shown the absorption cell, consisting of the rock-salt plates, r, the tinfoil w^asher, zv, and the tin shield, t, while fig. 4B shows the cell, c, mounted upon its holder, with the clear piece of rock salt, r, below it. 1 ' f^l » 1 * > \ i L.J » ' 4 1 1 ► 1 1 "^ 1 4 > 1 i » 1 1 Fig. 4b. DESCRIPTION OF APPARATUS AND METHODS. 25 In fig. 5 is shown the manner of mounting the cell-holder, C, in ver- tical guide F, upon the movable spectrometer arm, A. (See also fig. i.) A double-walled sheet-iron and asbestus shutter moves in the vertical guide, W. The heater of a Nernst lamp is at H, while the spectrometer slit is at F. The cell for the gases was of glass tubing, with rock-salt plates for windows. It is shown in fig. 10, while fig. 1 1 shows the manner in which it is mounted in its holder by means of the key, k. This holder could then be placed in the vertical guides shown in fig. 5. The shutter and the cell holder were operated from the observing telescope by means of cords and pulleys, so that it was necessary to enter the room only in making the spectrometer settings. SOURCE OF RADIATION. The question of the source of radiation, with several curves showing the distribution of the energy in several kinds of electrically heated Fig. 5. clay radiators is given in Appendix I, in which is shown the great adapt- ability of the " heater " of a Nernst lamp for this work, on account of its strong radiation beyond 4 ix, and also on account of the ease with which the constancy of the radiation can be maintained. Although two I lo-volt heaters were provided, only one was generally used. Current was obtained from a storage battery of 90 cells, which brought the " heaters " to a rich cherry red. The " heater " thus used would last for several months. But when used on a battery of 120 cells it lasted but a few weeks, when it was found that the platinum conducting wire, under the clay surface, had vaporized, and the surface of the heater was covered with beautiful microscopic crystals of platinum. 26 INFRA-RED ABSORPTION SPECTRA. This heater is not adapted to the region from 0.8 to 2 ;«, on account of the weakness of the radiation. Here the " glower," or an acetylene flame, is preferable. The constancy of the radiation and the cheapness of the " heater " make it a most useful source of infra-red radiation. Moreover, it has no products of combustion, such as water vapor, hence the room can be kept free from moisture, thus protecting the prism. METHOD 01^ OBSERVATION. The method of observation consisted in projecting successive por- tions of the spectrum upon the radiometer vane, noting its deflection when the absorption cell was before the slit Fi (fig, 5), and also the deflection when a clear piece of rock salt was substituted. The ratio of the deflection obtained for the radiation, from the " heater," which passed through the cell to the deflection for the radiation which passed through the clear plate of rock salt gave the percentage of transmission through the liquid directly and more accurately than by finding the absorption of the empty cell and deducting it. This also meant the reduction of the work by almost one-half. By plotting these ratios as ordinates, and the wave-lengths corresponding to the circle readings as abscissae, the " transmission curves " of the different substances are obtained. The general method of observation consisted in observing the zero reading through the telescope, then, with the cell before the slit, raising the iron shutter, by drawing a cord which extended to the observer in the outer room. The vane would then be deflected and come to rest in a new position in about 30 seconds. The carrier of the absorption cell was then raised, by drawing a second cord, until the clear piece of rock salt came before the slit. This height was regulated by a suitable stop on the ways, which permitted the cell to rise to a fixed height. The vane would then suffer a still greater deflection, and, when it came to rest, the shutter was dropped — when the deflection decreased to zero. If there was a shifting from the former zero reading, it was noted and the deflections were corrected. The spectrometer was then set for a new position in the spectrum and the operation repeated. If the zero shift was of any significance the reading was repeated, especially when going through an absorption band. This method of observation meant a still further saving of time, so that the time to make a single measurement was reduced to about 1.5 minutes, while the exploration of the entire spectrum required from 3.5 to 4 hours. In the mean time the observations were plotted upon DESCRIPTION OF APPARATUS AND METHODS. 2/ cross-section paper. In this manner many slight variations in the trans- mission curve could be verified by repeating the observation before setting for a different wave-length, and, in going through an absorp- tion band, the spectrometer settings were made more frequent, thus increasing the accuracy of its location. That this method has not been prejudicial against the results obtained is shown in the transmission curves of the gases in which it was neces- sary to find the transmission of the cell and the included gas, and then deduct the absorption of the empty cell, in order to obtain the transmis- sion through the gas itself. The manner of recording the observations is shown in Table III,^ which also shows the peculiar distribution of the energy in the spectrum of the " heater " used as a source of radiation. In the subsequent tables only the " transmission " of each substance examined is given. At the beginning and at the conclusion of the exploration of each compound the sodium lines were thrown upon the radiometer slit to determine whether the apparatus was still in adjustment. SOURCES OF ERROR. It is a difficult matter to determine the accuracy attained in this work, as well as the sources of greatest error. With the smaller spectrometer and rock-salt prism it was not difficult to find one's work in agreement with that of previous investigations on absorption spectra. But the emission band of COo, which Paschen^ and others found at 4.40 ij., by using a fluorite prism, which has about twice the dispersion of rock salt in this region, was sometimes found at 4.42 ju,, or a difference of about 10" in arc. This was found to be due to the curvature of the slit, as explained. With the large spectrometer, which gave about twice the dispersion of the small one, and accidentally equivalent to the fluorite dispersion of Paschen and of Ransohoff,^ the emission maximum came exactly at 4.40 /* for both Bunsen-gas and Bunsen-acetylene flames. This is the best defined, the most accurately known, and the most easily obtained comparison line in the infra-red. The estimation of the degree of accuracy of this band is based upon the number of observers who have located it. The next in order of accuracy of location is the 3.43 ^ absorption band found in compounds containing CHg-groups, by Julius,^ Ransohoff, and others. This band is not quite so sharp as the COo emission band. Using these as comparison bands the 6.86 /a band, found so frequently in the present work, has, in the writer's estimation, been ^Given at the end of the text. ^Ransohoff, loc. cit. Taschen, loc. cit. *Julius, loc. cit. 28 INFRA-RED ABSORPTION SPECTRA. located as accurately as the aforesaid lines, by using the large spec- trometer. This estimation is based upon the fact that the prism has been remounted and reset for minimum deviation so often, upon two different spectrometers, with the adjustments made in three different ways, that the consistency in the location of this band can not be attrib- uted to mere coincidence. If one were to set up similar apparatus, and repeat the work, the other absorption bands would be found accurate to about 0.02 fx. This is a liberal allowance for errors in computing, draw- ing, and reading wave-lengths from the calibration curve, in adjusting the apparatus, in drawing the curves, etc. As an example of the diffi- culty in keeping the apparatus in adjustment for a long period, the experiences of Donath^ may be cited, who found his apparatus, which was of the finest construction, out of adjustment by the time he had examined half a dozen compounds. In the present work the adjust- ment was tested, by means of the sodium lines, at the beginning and conclusion of each series of measurements. A source of error in determining the transmission is the possibility of the source of radiation varying in intensity while reading the deflec- tion of the vane when the absorption cell is before the slit, and the deflection for the direct radiation. This is likely to occur with a gas flame, but did not occur with the " heater " used on a storage-battery circuit. The rigidity of the mounting of the cell in its carrier, which moved to a fixed height in a plane parallel to that of the spectrometer slit, in accurately fitting wooden ways upon the spectrometer arm, eliminated errors due to variation in thickness of cell. A great deal of time was spent on the question of the use of a prism kept at minimum and at non-minimum deviation. As is well known the minimum of deviation is not sharp, well defined ; hence any vari- ation from the real minimum in the setting for the sodium lines will be magnified far out in the infra-red when using the non-minimum devi- ation method. This appears to be the chief objection in this work. Since in the Wadsworth minimum deviation method the prism-mirror table is turned through only half the required angle, thus doubling any error in making a spectrometer setting, the writer preferred the con- stant emergence method, which requires the collimating telescope to be turned through the whole angle. In this manner a smaller power microscope could be used in making the spectrometer setting, the circle readings could be made more quickly, while there was less likelihood of losing one's place in the spectrum, which easily results in this work, ^Donath, loc. cit. DESCRIPTION OF APPARATUS AND METHODS. 29 where one must shift back and forth in repeating any observations. Since the CO2 band at 4.40 /m has been determined so accurately by mini- mum deviation methods by Paschen^ and by Stewart/ the operation of setting for minimum deviation with the small spectrometer was repeated until this band came exactly at 4.40/*. With the large spectrometer this band was found exactly at 4.40 /x for the minimum and non-minimum deviation methods. The maximum is so sharp that it could be accu- rately located to two or three seconds of arc, with the large spectro- meter, and, since no difference could be detected in the position and in sharpness of the large absorption bands, it appears that it makes no difference which method is used.^ Ransohoff,^ assuming a lack of minimum deviation of 3' at y.y fx, com- puted a change in deviation of 11", or 0.0012 /x, which agrees with the theoretical work of Wadsworth.^ As an illustration of the difficulty in establishing the absolute value of any absorption band, my experiences with the large spectrometer may be cited. Thinking that a marked band of carbon tetrachloride, found by Paschen at 6.45 fi, would be an excellent check on the present work, it was examined and found to consist of two bands, of which his is the mean value. THE CHEMICALS USED. In the beginning of this investigation it was never suspected that it would grow to the present magnitude ; neither was the source of the chemicals considered of the greatest importance. But with the increase in my knowledge of the processes involved in manufacturing the chem- icals, it became evident that the label " C. P." was of little significance, and it is considered thus by chemists. Obviously it would have been impracticable to purify all the chemicals used. The alternative was to duplicate the work by examining the same compound manufactured by different firms. In the " transmission curves," for many of the compounds thus examined, the duplicate curves are given, and it will be noticed that, disregarding their general appearance, due to the dif- ference in the thickness of the films, which is of no consequence, rarely IS there a difference in the spectra of the two samples. This means that the two samples of a compound were equally pure, or what amounts to the same thing, that they contained the same impurity if any was 'Paschen, loc. cit. ^Stewart, loc. cit. ^Helmholtz : Physiol. Optok., p, 260, shows that for purity of color bands and sharpness of Frauenhofer lines it is not necessary that the rays are homocentric. *Ransoho£f, loc. cit. ^Wadsworth : Astrophysical Jour., 2, p. 264, 1894. 30 INFRA-RED ABSORPTION SPECTRA. present. As an illustration of the aforesaid, a series of '' petroleum distillates " may be mentioned, of which 25 samples were obtained from crude petroleum, while one, " hexane," was made synthetically. The samples obtained by distillation have the same number of large absorp- tion bands, while the synthetic product showed one new band, in addi- tion to those common to the distillates. For the earlier part of the work, the samples of the compounds were obtained from the m^useum of the department of chemistry. Many of these had been obtained from Merck, while some were prepared by students. Many of them were redistilled just before using. Out of a total number of 70 compounds, 43 were imported directly from Kahl- baum. Of this number, 23 were duplicates of the aforesaid compounds. They came in sealed glass-stoppered bottles, in quantities of 5 to 10 grams, and were insured to be the purest obtainable. He omitted to state the boiling points, however, as requested. With these small quan- tities it was difficult to determine the boiling points with precision, but of the samples tested every one had a boiling point so close to the theo- retical value that its purity could not be questioned. In addition to the above compounds, a series of 25 distillates was obtained from Prof. C. F. Mabery, of Case School of Applied Science (see p. 73). The gases were manufactured in the laboratory, as indi- cated elsewhere. It wall be noticed presently that water is the most opaque substance investigated; hence it is of the greatest importance to have the com- pounds free from water. The fact that ordinary alcohol blurred and glycerin etched the rock-salt plates of the cell, while all the other com- pounds did not, would indicate that the quantity of water in the latter must have been exceedingly small. Thus, for the petroleum distillates, of which a great number were examined in one cell, after completing the work the interior of the cell was found as clear and highly polished as when it was new. PROBLEMS ATTACKED. The problem before me was to determine the effect of molecular weight upon absorption spectra ; also the effect of chemical structure, i. e., the arrangement of the atoms in the molecule, and the effect pro- duced by the substitution of a CH3 or OH group of atoms. As a criterion for the effect of the substitution of a CHg-group, the conspicuous band occurring between the wave-lengths 3.0 fi and 3.5 fj. Avas critically examined. Julius^ found this band at 3.45 /x for com- ^Julius, loc. cit. DESCRIPTION OF APPARATUS AND METHODS. 3 1 pounds containing CHs-groups, and, hence, ascribed it to this group. As a standard for judging the effect of the OH radical in certain com- pounds, the water bands found by Aschkinass^ at 3 /a and 6 fx were selected. Ransohoft',^ in his study of several alcohols, had tacitly con- cluded that the band at 3 )u, was due to the OH radical. Such conclu- sions in regard to the CH3 and OH groups seemed contradictory to the work of Angstrom and Palmer/ who found that the chlorine band at 4.28 fx does not occur in the six chlorine compounds investigated by Julius. The latter had previously shown that the chemical atom lost its identity in a compound, so that one can not foretell the absorption spectrum of a compound from a knowledge of the spectra of the con- stituent elements. In the detailed study of each compound the chemical properties will constantly be noticed. In some cases it will be found that certain chem- icall}' related compounds, especially groups like the fatty acids, or the mustard oils, give similarly related absorption spectra, while others which are unusually similar in their physical properties, e. g., benzene and thiophene, show entirely different absorption spectra. In some compounds we shall find the evidence of the effect of absorption as being due to a definite group of atoms; in other compounds the evidence points just as strongly in favor of the manner of bonding of the atoms, while still other compounds show that both the groups of atoms and the manner of bonding with other atoms, as well as the kind of atom, are instrumental in causing absorption. ^Aschkinass, loc. cit. ^Ransohoff, loc. cit. ^Angstrom & Palmer, loc. cit. CHAPTER IV. INVESTIGATION WITH A QUARTZ PRISM. (Figs. 6, 7, and 8 ; Table IV'). As is well known, a quartz prism is the most useful in exploring the spectrum from o.8 to 2.8 (x, where it becomes opaque. In the introduction, the work of Puccianti, who explored a number of benzene derivatives up to 2.75 fi, has been noticed. His curves show that all compounds, the molecules of which contain carbon directly combined with hydrogen, present a maximum absorption at 1.71 /x, while all the benzene derivatives have two other maxima in common at 2.18 fi and 2.49 fi. The three alcohols examined have a band in com- mon at 2.05/1. He examined 16 compounds, viz, carbon-tetrachloride, carbon disulphide, methyl, ethyl, and allyl alcohol, ethyl ether, methyl and ethyl iodide, ethyl benzene, ortho-, meta-, and para-xylene, toluene, benzene, pyridine, and water. The results obtained by Puccianti seemed so unusual that a continu- ation of the work was deemed necessary. Accordingly, 18 new com- pounds were examined, also benzene to serve as a comparison with his work. The new compounds are 1-pinene, benzaldehyde, chloroform, thiophene, phenyl mustard oil, methyl salicylate, eucalyptol, caproic and oleic acid, ethyl carbonate, methyl acetate, ethyl succinate, glycerin, methyl cyanide, allyl sulphide, nitro-methane, eugenol, and safrol. Their transmission curves are given in figs. 6, 7, and 8. Of the marked ben- zene maxima those at 2.18 /x and 2.49/* agree with the values found by Puccianti. The band found by him at 1.71 /a was located at 1.68 )n to 1.69 /A, which agrees with Donath.\ As a whole, the maxima of the curves are in agreement. His curves show an oscillation of the maxi- mum about the value 1.71 /x, just as will be noticed in the present curves. The depression in the benzene curve at 1.02 fx is of interest, since the photographs obtained by Abney and Festing^ show lines in this region. . It will be unnecessary to go into details in describing the absorption spectra given here, and it will simplify matters to discuss the conspicu- ous maxima. The hand at 0.84 /x. — In this region the dispersion of quartz is only slightly less than at J.7 fx, while it is comparable with that of rock salt 'Table IV is given at the end of the text. ^Donath, loc. cit. ^Abney & Festing, loc. cit. 33 34 INFRA-RED ABSORPTION SPECTRA. at 12 fji, which would make it appear that the lack of large absorption bands in this region is due to the absence of ions whose free periods coincide with that of the heat waves transmitted. The photographs obtained by Abney and Festing usually show only fine lines, which can not be resolved with the radiometer. The characteristic band found by them at 0.867 fx is represented in the present investigation by a small band whose maximum oscillates between the values 0.83 fx and 0.86 (i. This band occurs in nearly every compound studied, and is the most marked in eucalyptol, at 0.83 /x, and in ethyl succinate, at 0.85 /x. It has been found so frequently that one can hardly doubt its existence. The hand at /./ fi. — ^Beyond 0.85 //. the bands are more variable in their occurrence, and in their depth, until we approach 1.7 fx. This band is found in all the carbohydrates studied, which agrees with Puccianti. Of course, it is slightly shifted at times, which does not always appear to be due to lack of precision in the apparatus. This is especially noticeable in thiophene, where the maximum occurs at 1.66 /a, which is just double the 0.83 fx band, and in ethyl succinate, where these two bands are shifted to 0.85 /x and 1.73 /x, respectively. In noticing the constant occurrence of bands in these two regions, one begins to inquire into their significance, and, after finding similar conditions at 3.43 /x, 6.86 /x, and 13.8 /<., this becomes still more urgent. The question of the importance to be attached to these facts must be reserved for a later discussion. The hand at 2.08 ix. — This shallow band in Puccianti's curve of the alcohols was found at 2.14 fx for a sample of de Haen's bidistilled glyc- erin. On adding a drop of water the band was broadened out, extend- ing from 2.08 IX to 2.1 fx, as in ethyl alcohol. The gradual increase in the general absorption beyond 2 fx is very marked in contrast to its almost complete transparency from 0.8 /t to 1.2 fx. The bands at 2.18 fx and 2.4Q fx. — In the present study of the simpler benzene derivatives, these two bands always occur, which substantiates the work of Puccianti. In the very complex derivatives, like pinene, eucalyptol, eugenol, and safrol, the bands are wanting or shifted to entirely new positions, which agrees with the observations farther in the infra-red. As a whole, the curves form an interesting study by themselves. In comparing them with the curves obtained with the rock-salt prism, for this region of the spectrum, we can see how useless it would have been to explore it thoroughly with the rock-salt prism and the small spectro- meter. Only in the chloroform curve do we find the marked bands resolved and in their proper places. With the larger spectrometer, however, the agreement in the location of the maxima with those obtained with the quartz prism is very close. INVESTIGATION WITH A QUARTZ PRISM. 35 Fig. 6. 36 INFRA-RED ABSORPTION SPECTRA. Fig. 7. INVESTIGATION WITH A QUARTZ PRISM. Fig. 8. 37 •=; =:=Z ^ t— > < L 2. — _ r*--- - 4 U- S o i t> , c« u < 'S > t» o > " ''g ^ o 00 R o ^ o 00 o o ^ ? ^ ^ -a» 1 1 k i> c — — — , . _ . . — _ r- sr" — ^. . . ^ -. J3 '^ y D ^ S jf \ o c z, ^ ^ o ^ O o ? o «G o O v5 ^ > — •: — i~- — ^ — •< S3 -X { *>^ u r r^ ^ N s \ ■o c X .a 7 u 5 a > < 3 \ jc X s V a X < \ i ny t< ,\ o o o CO o o o CHAPTER V. INVESTIGATION WITH A ROCK-SALT PRISM. CivAss I : Methane: Derivatives. This class comprises all carbohydrates having an open chain, the so- called " aliphatic compounds " as distinguished from the " cyclic " or ring compounds Absorption Spectra op Gases. (Figs. 9 to 25 ; Tables I, V, and VII. )i Our knowledge of the absorption spectra of gases is confined to the work of Angstrom,^ who studied CO, CO2, CH^, C2H^, and (C2Hg)20. In fact, he is the pioneer in this subject. In studying gases the chief difficulty lies in obtaining them in a pure state. The apparatus, including generator, purifiers, and gasometers, must be air-tight, so that after the air has once been removed, none can enter from outside. Fortunately, we have in liquid air a means for obtaining small quantities of other gases in a more tangible state. By combining fractional liquefaction and fractional distillation with the usual chemical methods of purification one can obtain gases in a very high state of purity, as will be noticed in the analyses of some of the gases studied. In the present work the gases were first purified by chemical methods. Several showed absorption bands in common, so the work was repeated, using the additional method of liquefaction. The way that the transmission curves have been changed and the impurity bands have disappeared as a consequence will be noticed in the several curves, especially ethane. The arrangement of the apparatus for generating and liquefying the gases is shown in fig. 9, in which a is the generating flask, h is the wash- ing tube, and c is the drying tube containing PoOs and cotton or glass wool — the latter to present a large surface. This tube was always thor- oughly cleaned before making a new gas. At d and e are the liquefac- tion tubes, and g is the mercury gasometer, of 200 cc. capacity. The operation consisted of fractionally liquefying the gas in d, and what passed from d into e was liquefied there at a lower temperature. The ^Tables I to VIII and figs. 12 to 132 are given at the end of the text. ^Angstrom : Ofversigt af Kon. Vetensk. Akad. Forhandhgar, Stockholm, Nr. 7, 1890. 39 40 INFRA-RED ABSORPTION SPECTRA. part that did not liquefy in e passed out at /. Petroleum thermometers were used to measure the temperature, which, of course, could be deter- mined only approximately, but as close as one can make the fractional liquefactions. The thermometer was placed in con- tact with the side of the bulb, and the whole suspended over the liquid air. By raising or lowering this combination, any desired temperature could be ob- tained. From 5 to 10 cc. of the lique- fied gas were collected in e, which was then closed at h, by means of a pinch-cock, and al- lowed to grow warmer. After the air had been thoroughly washed out at /, as shown by the lighted gas, the bulb was per- mitted to become still warmer, and the successive fractions were distilled into the mercury gasometers, g, where they were stored under mercury seals, awaiting analysis and examina- tion. This method of procedure, for the present work, was cer- tainly more desirable and effi- cient than the usual one of join- ing the generator, purifiers, and absorption cell with an exhaust pump. In that case one can not manipulate the gases after once generated. The desirability of handling the stored gases is shown in the purification of ethylene, which is al- most impossi- ble to obtain Fig. 9. INVESTIGATION WITH A ROCK-SALT PRISM. 41 in a state sufficiently pure for this work. Not being satisfied with the usual washing in bromine water, and liquefaction method, the gas- ometer was placed in series with a wash-pipette of fuming sulphuric acid, and the gas washed back and forth half an hour, after which only a suspicious trace of ethylene remained, as will be noticed at 10.5 ^ in the absorption curves. The Absorption Cell. The absorption cell for gases is shown in fig. 10, Two glass cells were used, in length 6.3 and 5.7 cm., diameters 2.4 and 2.2 cm., capacity 30.8 cc. and 21.5 cc, respectively. The rock-salt windows were split from the natural crystal, which gave smooth, plane surfaces which were not attacked by moisture, and lasted throughout the work of two months. The windows were attached with Le Page's glue, which became exceedingly hard on drying, and had such a low vapor pressure that no absorption bands could be detected, even after exhausting the cell to 0.02 mm. and per- mitting it to stand four days. One of the vapors to be expected from it would have been acetic acid. Before filling the cell with a new gas It was always washed out thoroughly with air. This was done by means of a water aspirator attached to the pump, the cell being exhausted from five to seven times, each time allowing it to fill with air. The final exhaustion was carried to 0.02 mm., and the absorption of the cell, thus exhausted, was found each time before filling it with a new gas. Only 42 INFRA-RED ABSORPTION SPECTRA. ^^h I 3 3 once was absorption found for residual gas; this was ethyl ether at 8.7 ii, the cell having been washed but three times with air. To fill the cell, previously exhausted to 0.02 mm., with gas from a gasometer, the capillary tube, c, was inserted in the rubber connection (A fig"- 9) ^^ the latter, which contained a drop of mercury. On inser- tion, this drop of mercury would displace the air, except the trace con- tained in the bore of the capillary tube. The capillary bore was 0.024 cm. in diameter, 3 cm. in length, and had a volume of 0.013 cc, which, compared to the total volume of the cell, would introduce one part of air in 2,500 parts of pure gas. This trace of air could also have been prevented from entering the cell by filling the capillary with mercury. Since the presence of this small quantity of air diminishes the total length of gas by only 0.003 cm., the effect on the total transmission can not be detected, as was found on trial, and in practice the capil- lary was not filled with mercury. On opening the pinch-cock at f (fig. 9) and the stop-cock of the cell, the gas en- tered the latter and was brought to atmos- pheric pressure, which, with the tempera- ture of the gas (room temperature), fur- nished data for computing the equivalent thickness for the liquid state. A helix of iron wire inserted in the glass cell served as a diaphragm to prevent reflection from the walls, and seemed more serviceable than a black paint, since it also prevented occlusion of gases. The cell was mounted in the usual man- ner, in a wooden carrier (fig. 11), which worked in vertical ways between the source of energy and the spectrometer slit. The transmission through the exhausted cell was found each time before filling with a gas. As mentioned before, this also served as a test for residual gas, which could have been detected, if present, by setting on one of the absorption bands. The ratio of the transmission of the cell filled with gas to that of the empty cell gave the transmission of the gas itself. The transmission of the empty cell changed but little in two months, and, as a typical exam- ple, transmitted 65 per cent at 1.7 /a, 75 per cent at 4^1, decreasing to 70 per cent at 6 \i, then gradually increasing to 74 per cent at 10 )U. and Fig. 1 1 . — Top view of gas-cell holder with cell c in place. «, top view, 6, side view. INVESTIGATION WITH A ROCK-SALT PRISM. 43 to 78 per cent at 12/1,. Excepting regions of selective absorption, the gases, for this length of cell, were usually perfectly transparent, so that windows made from plates of rock salt split from the natural crystal were more serviceable and cheaper than highly polished ones, which would have transmitted about 80 per cent of the energy when new, but would have soon become tarnished with moisture. Many of the gases show but few absorption bands for the length of cell used. Whether more exist, which are very weak and can be detected only in layers of great thickness, is an unanswered question. Rubens and Aschkinass,^ using a column of CO2 gas 65 cm. long, found the same number of bands as for one 5 cm. long. On the other hand, Langley's curves of the atmospheric absorption show a great com- plexity, which can not be attributed to a mere variety of gases. GASES STUDIED. Methane. CHi. (Cell 6.3 cm. ; barom., 75.0 cm. ; temp., 22° ; fig. 12. See end of text.) The methane used in this work was made by heating a mixture of sodium acetate and soda lime, and washing it thoroughly in pipettes of potassium hydrate and of fuming sulphuric acid, which removes the unsaturated hydrocarbons, like ethylene. The latter is also formed when methane is made by this method. Since methane boils at — 160° and ethylene at — 105°, it is easy to apply the method of fractional lique- faction and fractional distillation. This was done as an additional pre- caution, but no change in the intensity or position of the transmission maxima could be detected. The slight shift of 0.02 /x at 3.4 /* is due to resetting of the prism, and will be noticed in ether, which was also one of the first gases studied. For curve b the gas was purified by lique- faction. The spectrum of methane is distinguished by two large absorption bands at 3.31 fx. and y.y [i, and a smaller one at 2.35 fi. 'Methane was investigated by Angstrom, who found the same number of bands, but since his calibration is wrong beyond 5 /* no comparison can be made. The effect produced by substituting four CI atoms for the four H atoms, and thus forming carbon tetrachloride, is very striking. Here we have an entirely different spectrum, with no bands in the region of 3 /A to 3.5 ^, characteristic to carbohydrates. In fact, we find no bands until we arrive at 6.5 fi and its prominent harmonic at 13 fi. iRubens & Aschkinass : Ann. der Physik, 64, p. 584, il 44 INFRA-RED ABSORPTION SPECTRA. Acetylene. C2H:. (Cell, 6.3 cm. ; barom., 74.5 cm. ; temp., 22°. 5 ; fig. 13.) The acetylene used was generated from " selected calcium carbide, commercially pure, for bicycle lamps." In the preliminary examination the gas was washed in c. p. H2SO4 and dried in P2O5, curve a. For a subsequent examination the gas was washed by passing it through a solution of 20 grams of HC1+ 10 grams HgCla + So grams HgO to remove the phosphene (PH3) ; then through a solution of KOH to remove HjS ; then through concentrated H2SO4 to dry it, and finally through a U-tube cooled in CO2 snow. In the latter there was a slight deposit of vapor on the side where the gas entered. This was found to be water vapor. After all this treatment, as well as an additional one, to be mentioned presently, not a single band was changed appreciably in transmission, nor in the position of its transmission minima, curve h. That ordinary acetylene is quite pure has been found by Rands,^ who found a purity of 99 per cent. Acetylene is distinguished for three deep, narrow regions of absorp- tion at 3.08 II, 7.73 IX, and 13.63 jx, and two other bands, one at 3.7 fi, which is harmonic with the one at 7.4 /x. In order to be able to locate the band at 13.63 /u, the cell had to be exhausted to a pressure of 7 cm., curve c. Elsewhere the gas is perfectly transparent. At first it seemed that the band at 3.08 /x might be due to water, but washing it thoroughly in new cone. H2SO4 and leaving it stand in contact with it for three days, in a pipette, failed to alter the intensity or position of this trans- mission minimum, curve d, which would indicate that this band belongs to acetylene. As a further check, a thin film of water, curve h, between fluorite plates, was examined. The first band occurs at 3.02 p., while the second lies at 6.02 II. If this band at 3.08 /* in acetylene were due to water, one would expect a similar one at 6.0 /x, but in this region the gas is perfectly transparent. It will be interesting to note the shifting of this band toward the infra- red for an increase of two H atoms (keeping the number of C atoms constant) in each of the following series of compounds. Acetylene brought to a red heat will polymerize, three molecules uniting to form one molecule of benzene (CeHg), (Berthelot). This is one of the most striking transitions from the aliphatic to the aromatic series, and is at the same time a synthesis of the parent hydrocarbon of the aromatic substances. Acetylene is also chemically allied to ethyl alcohol, yet there is not the slightest trace of resemblance in the absorp- tion curves in either case. ^Rands : Phys. Rev., xiv. INVESTIGATION WITH A ROCK-SALT PRISM. 45 Ethylene. C2H4. (Cell, 5.7 cm. ; barom., 73.8 cm. ; temp., 22°. 5 ; fig. 14. J This gas was prepared from ethylenebromide, CaH^Brj (from Drs. Bender and Hobein, boiling at the theoretical temperature of 131°), by using a zinc-copper couple/ and was dried by passing it through a glass tube containing P2O5 and absorbent cotton, to present a large surface. In the preliminary examination the gas showed ether bands. The ethylene bromide was then redistilled, and a second examination was made, the gas having been passed through a tube placed in CO2 snow. After the liquid-air apparatus was set up, this gas was dried in P2O5, then fractionally liquefied in bulb (/(fig. 9), kept at — 97°, and the remainder liquefied in bulb e, kept at — 120°. These were the readings of the petroleum thermometers, which, although constant and in contact with the liquefaction bulbs, may not have been at the same temperature, since they presented a smaller surface to the evaporating air. In the first bulb there was a liquid having a strong odor of CoH^Brg. In the second tube were about 5 cc. of liquid ethylene, having a sweet odor much like that of C2H4Br2, somewhat sharper, like pure acetylene. It burned with a beautiful yellowish- white flame. The second frac- tional distillate was used in this examination. The transmission curves for these three samples of gas, differently prepared, showed but slight variation in the intensity and position of the transmission minima, excepting that the ether band found at 7.8 \x. in the first sample was absent in the others. That the gas was very pure is shown from the analysis by absorption, using bromine water. The quantity of unabsorbed gas was so small that it could not be measured accurately. Amount of C2H4 taken 50. 2 cc. Unabsorbed gas 0.6 cc. Total C8H4 49.6 cc. Purity 98.8 per cent. A combustion analysis, made by Mr. R. C. Snowden, of the first frac- tion distilled, showed considerable impurity. From general experience it appears that this is due to the method of analysis rather than to any real impurity. The transmission of ethylene is entirely different from acetylene, from which it differs in composition in having two more H atoms. It is also different from CH., in which there is one less C atom. ^A zinc-copper couple is made by placing pure granulated zinc in a dilute solution of CuSOi, when the zinc becomes covered with copper. When this couple is placed in the liquid it immediately decomposes it. 46 INFRA-RED ABSORPTION SPECTRA. The first large band is shifted to 3.28 /u,. The large transmission minima at 6.98 fi and 10.5 /x are of interest in connection with the study of CH, groups, and with piperidine. For the region beyond 9 fi the pressure had to be reduced to 20 cm. Ethane. C2H6. (Cell, 5.7 cm. ; barom., 74.5 cm.; temp., 23°; fig. 15.) This was the most difficult gas to be prepared. Unlike ethylene, the chemicals used are more complex, and the reaction can not be con- trolled, so that — besides ethane — ethylene, methane, etc., are formed. The ethane gas was generated according to the method first described by Frankland,^ who found a " volume purity of 97.88 per cent, using liquefaction in addition. Equal parts of CoHjI (Kahlbaum) and abso- lute C2H5OH were poured over a zinc-copper couple, when the gases began to evolve immediately. No heat was applied, since that causes greater impurity. The gas passed through bromine water in a U-tube, with liquid bromine at the bottom, to remove the ethylene, then through KOH to remove the Br, and then through a tube of P0O5 to remove the water. Since ethane boils at — 93° and ethylene at — 105°, fractional liquefaction and distillation is an uncertain method of purification. In this case the bromine water offers the most serviceable means of sepa- ration. Methane will not liquefy till cooled to — 160°, and hence will pass through the bulb without being condensed. The same is true of the free hydrogen. After fractional liquefaction (about 6 cc.) and fractional distillation, the first and second fractions of the distillate were again washed in fuming sulphuric acid and KOH, and dried in PaOg, it being thought preferable to run the risk of letting in air in doing so, rather than be uncertain about the presence of ethylene, carbon dioxide, etc. The analyses were not satisfactory. The second fraction was wasted in two attempts at combustion with oxygen, there being a violent explo- sion each time. The explosion-pipette method is less accurate on account of the small quantity (5.4 cc.) of gas that can be used. How- ever, three separate analyses of the method gave a fairly concordant purity of 96 per cent. This is the first gas noticed in which the absorp- tion bands decrease in intensity or disappear entirely, which is an excel- lent clue to the impurities present. The change wrought in the transmission curves is most conspicuous at 10.5 /A (fig. 15). That transmission minimum has entirely disap- peared, and the bands at 12 /x, which were obscured by it, now appear in their proper intensities. ^Frankland : Jour. Chem. Soc. London, 47, p. 236, li INVESTIGATION WITH A ROCK-SALT PRISM. 47 The position of the transmission minimum at 3.39 /x does not vary, however. The very deep band at 8.25 fj., which has almost disappeared in the second curve, is no doubt due to CMJ, since that substance has a maximum at this point. Ethane differs from ethylene in having the 2.31 fi and the 3.28 fi bands of the latter shifted to 2.36/* and 3.39/11, respectively, in the former. Some of the following gases show this same shifting for the region of 3 /J.. As noted before, for rock-salt dispersion, all carbohydrates studied show an enormous absorption band, varying in position from 3.1 /a to 3.7 fi. The same is true for the 1.69 yu, band, using quartz dispersion. Here it is difficult to establish a definite shifting, as is found for homol- ogous compounds at from 3.08/1 for acetylene (CoHg) to 3.39^1 for ethane (CgHe). This study of gases was undertaken to learn the behavior of the CHj group in the molecule, as found in ethane. After studying many com- pounds of CH3, simple and complex, having an absorption band at 3.43 fx, w^hich was first announced by Julius^ to be due to CH3, it was rather surprising to find that ethane (HC3 — CH3) has its maximum at 3.39 IX. In the simple compounds, like CH3I, this band occurs at 3.4 /u.. When we compare ethane (CaHg) with benzene (CeHg) we find the latter has its transmission minimum at 3.28/1, which again shows that the structure as well as molecular weight influences the absorption. The cell of ethane was partly exhausted, and one part of it was mixed with two parts of acetylene, and, as a result, the band due to acetylene was obliterated, except a slight break at 3 /i, while the band due to both is at 3.3 fi. Lack of time did not permit examination of other regions. The test was not quite a fair one, considering the one gas as an impurity in the other, yet it serves to show that, for the region of 3 /i, where the dispersion of rock salt is still small, reliance upon the occurrence of an absorption band of an impurity, as a means for detecting the impurity, is not permissible. However, at 4.4 /i, where the dispersion is greater, one can detect the presence of CO2 in CO. (See fig, 21.) In fig. 15 curve a is for the gas when purified by liquefaction, and curve b for the same gas after washing in fuming H2SO4. Butane. C4H10. (Cell, 5,7 cm. ; barom., 75.4 cm. ; temp., 22° ; fig. 15.) This gas was made^ from ethyl iodide (C2H3I) by pouring it over an amalgam made of sodium and mercury. The presence of the mercury retards the action of the metallic sodium upon the ethyl iodide. After ^Julius, loc. cit. *Loury : Jahrsber. in Forstchritte der Chemie, p. 397, i860. 48 INFRA-RED ABSORPTION SPECTRA. drying in P2O5 the gas was passed into a U-ttibe placed in a freezing mixture of common salt and ice. Butane liquefies at 1°, while its isomer liquefies at — 17°. The attempt to liquefy it was not very successful, and the gas collected may have contained the isomer as well as other impurities. The transmission minima of ethane (C^Hq) at 2.36/* and 3.39/* and 6.85 fi are shifted to 2.4 fx., 3.42 fi, and 6.89 fi, respectively. Beyond 7 fx there are a number of bands which occur as impurities, the most con- spicuous of which are the two at 7.8 ju, and 14 /*, due to ether, and the 10.5 ju, band in ethylene. The bands at 8.3 /* and 8.9 ju.(diff. = 0.60 /x) find their counterpart in methyl ether, where they are shifted to 8.58 /a and 9.16 /A (difif.=o.58/A), respectively, and in ethyl ether, where the bands are shifted to 8.75 ju, and 9.25 fi (diff,=o.50/u,), respectively. Methyl Ether. (CH3)20. (Cell, 5.7 cm. ; barom., 76 cm.; temp., 22° ; fig. 17.) Prepared^ by heating 1.3 parts of alcohol and 2 parts of sulphuric acid to 140° and washed in KOH to absorb the SOg, and then dried in A study of this gas is of interest in connection with ethyl alcohol, with which it is isomeric. A more striking illustration of the effect upon absorption of the arrangement of the atom in the molecule has not been found, except perhaps for the sulphocyanates and mustard oils. From lOjtt to ii/t alcohol (fig. 38) shows transparency, and beyond 13 ju, there is complete opacity, while, for methyl ether, the very oppo- site effect was found. Beyond 8 fi the pressure of the ether had to be reduced to 10 cm. of mercury. The only apparent impurity band to be noticed is that of butane at 5.7 /x. The bands at 2.5 ju, and 3.45 /x and 10.55 /* ^^^ shifted to longer wave-lengths, as compared with those of butane. The 6.9 n band is harmonic with the one at 3.45 fi. This compound, like all the gases, is conspicuous for the great depth of its absorption bands. Curve b is for a pressure of 10 cm. Ethyl Ether. (C2H5)20. (Cell, 5.7 cm. ; barom., 74.9 cm.; temp., 24"; fig. 18.) For studying this vapor some of the anhydrous liquid was placed in the absorption cell, which evaporated, thus giving a saturated vapor. Frequent opening of the stop-cock kept it at about atmospheric pressure. Like the methyl ether, it has very deep bands, which are shifted as compared with the two preceding gases. The 4.75 /x band of methyl ether occurs at 5.15 /a in ethyl ether. The 6.9 /x band of the former ^Erlenmeyer & Kirchbaumer: Berichte d. Deutsch. Chem. Gesellschaft, 7, 699, 1874. INVESTIGATION WITH A ROCK-SALT PRISM. 49 is double, and occurs at 7.0 /* and 7.3 fi. These two bands are to be noticed with this same region for ethyl alcohol (fig. 38), where there is but one flat band, which would, no doubt, be found complex, using a larger dispersion. The 14^ band is also to be noticed, since it occurs in compounds having CH3 and C2H5 groups, while the 12 /u, band is also found in many compounds. Ethyl ether has been studied by Angstrom,^ both as a liquid and as a gas. He found that the maximum at 3.45 /x. did not coincide for the two states, the vapor being shifted by 4' toward the long wave-lengths. He claims to have obtained this effect repeatedly, while no shifting was observed by him for other compounds examined in the liquid and the vapor state. In the present work the region of 345 fi for liquid ether was explored at the same time that it was for the vapor. No such shift- ing was observed, as will be noticed in fig. 18, where the curves (indi- cated thus : A A A and x x x ) for liquid and vapor are seen to coincide. The slight difference in the position of the maximum for this day and that for the time of making the complete curve is due to the resetting of the prism in the meantime. Ether belongs to the water type of compounds, i. e., the C2H5 takes the place of an H atom in H2O, but it is not apparent in the curves. The bands at 7.3 \i and 7.8 yu, almost coincide with those of acetylene. Beyond 7 /x three sets of bands, " triplets," have " constant differences " of their wave numbers, which agree so closely as to lead to the sus- picion that they belong to a spectral series. The two most marked sets are: 7.0 ,« -^ ^, 10. 14 /i -^ 7.3 >^i=6o 10.77 ^^^=58 7.8 >^2=9o „_g >r2=90 I1.LUMINATING Gas. (Fig. 19.) The sample used was taken directly from the pipes and examined. It contains bands due to CH^, C.H,, C.Hio, CO, CO2, SO,, and HoS. Ordinary analysis shows about 60 per cent of CH4, 6 per cent of CO, 2 per cent of CO2, and the remainder principally unsaturated hydro- carbons. The curve is of interest in showing the presence of CO and H2S. Oxygen. (Cell, 5.7 cm. ; barom., 74.6 cm. ; temp. 23° ; fig. 20.) Made by heating KCIO3 -f MnO,, purified in KOH, and dried by passing through a glass tube containing absorbent cotton covered with P.O.. 'Angstrom, loc. cit. 50 INFRA-RED ABSORPTION SPECTRA. This gas is of considerable interest, because it enters into so many compounds. It shows two large shallow bands at 3.2 ju, and 4.7 ju,, respectively. They are only 4 per cent in depth. It appeared that this might be due to impurities, but repeated attempts to remove them failed to change the intensity. After passing the gas back and forth through a tube of P2O5, from one gas pipette to another, the bands still existed. The oxygen was then placed for three days in an absorption pipette containing concentrated H2SO4, but, as will be noticed in curve b, the bands still remain, showing that their presence is not due to water vapor. One sample was not washed in KOH, and then showed the CI band at 4.3 fi. Angstrom^ found this band at 4.28 ju, for pure chlorine. It is interesting to note that Dewar, and also Olzewski, found several absorption bands of liquid oxygen in the visible spectrum. Hydrogen and Bromine. (Cell, 6.3 cm.) Generated from c. p. Zn -j- HCl, washed in HoSO^, and dried in P2O5. The hydrogen gas showed no absorption bands. Paschen^ examined hydrogen and nitrogen, but, after repeated measurements, failed to find any absorption bands. The vapor of bromine showed no bands. The cell was lined with paraffin to prevent the bromine from attacking the glue. Carbon Monoxide (CO). (Cell, 5.7 cm. ; barom., 73.2 cm.; temp., 23° ; fig. 21.) Made by heating oxalic acid (C2H4O2) and concentrated H2SO4. This forms CO + COo + H2O. The CO, was absorbed by passing the gas through KOH solution, after which the CO was dried in P2O5. As will be noticed in curve a (fig. 21), there was still some COo pres- ent, as shown by the maxima at 2.75 fx and 4.3 ft. After washing the gas back and forth for half an hour, in a burette of KOH, and drying in PjOg the CO2 was entirely removed, as shown by curve b. The maxima of CO occur at 2.4 /x and 4.59 /a. Angstrom^ found these maxima at 2.48 fx and 4.52 [i. The absorption spectrum of CO was examined to 14 ^u, but no max- ima were found. As shown by Angstrom, the fact that the CO band occurs at 4.59 ju, and the CO2 at 4.29 /x invalidates the assumption that molecular weight has a great influence on absorption. ^Angstrom & Palmer : Ofversigt Kongl. Vet. Akad., No. 6, p. 389, 1893. ^Paschen : Ann. der Physik, 53, p. 334, 1894. ^Angstrom, loc. cit. INVESTIGATION WITH A ROCK-SALT PRISM. 5 1 Carbon Dioxide. CO2. (Cell, 5.7 cm. ; barom., 74.0 cm. ; temp., 23° ; fig. 22.) Made from KXO3+H0SO, and dried in P0O5. This gas has been examined by nearly every person who has investi- gated absorption spectra. It is noted for its variation in the location of its maxima, in emission and absorption spectra. This is most evident at 4.4 fi and 14 fi. The emission^ band of a Bunsen burner occurs at 4.4 ju,, while the atmospheric absorption band occurs at 4.26 fj.. Rubens and Aschkinass,* using a sylvite prism, studied CO, to 20 /x, and found a large band at 14.7 /A for absorption and at 14.1 fi for emission. The column of CO2 in the absorption work was 65 cm. in length. It will thus be noticed that at 4.4 fjL the emission band shifts to the longer wave-lengths with rise of temperature, while at 14 fi the shift is toward the shorter wave- lengths. Angstrom found the CO2 bands at 2.6 [x and 4.32 fx., while in the pres- ent examination they occur at 2.75 /x and 4.29 fx. The fact that the atmospheric band of CO2 occurs at 4.26 /x leads me to think that the present value is more nearly correct. A slight trace of CO would tend to shift it toward that band, at 4.59 /a. The actual position of the atmos- pheric band of CO,, as located in a radiation curve, depends upon the temperature of the radiator. This is well illustrated in the curves of the heaters (fig. 126), where for the more intense radiation the absorp- tion band is shifted from 4.25 /x to 4.28 fx. The 14.66 fi band is in excel- lent agreement with the value, 14.7 fx, found by Rubens and Aschkinass, when we consider that rock salt is already quite opaque in this region. As a whole, the work agrees w^ell with that of other observers, consid- ering how precarious a problem it is to map out infra-red absorption spectra. Carbon dioxide is the only gas studied which has no absorption bands at 4.5 IX and 14 fx. From the fact that gases dissociate at high temperatures are we to conclude that as the temperature rises the COg dissociates into CO, thus shifting the emission maximum of COg from 4.26 ju, toward the absorp- tion band of CO at 4.59 fx, and forming the well-known band at 4.4 /a? But how are we to account for the great Y band in Langley's holo- graphs, which show a broad band extending from 4.3 fx to 4.65 fx? We can not say that it is due to the combined action of CO and COg, since the CO is not found in the atmosphere. Oxygen has slight bands at ^Paschen, loc. cit., vol. 53. ^Rubens & Aschkinass : Astrophys. Journal, 8, p. 191, iS 52 INFRA-RED ABSORPTION SPECTRA. 3.1 /t and 4.65 n, but the absence of bands in the holographs at 3.1 /* would seem to indicate that the broad Y band is not the composite of the CO2 and O bands. As mentioned by Rubens and Aschkinass (loc. cit.), the absence of CO2 bands in the curves shows that this gas can not have a great influ- ence upon the general transparency of the earth's atmosphere. SuivPHUR Dioxide. 'SO2. (Cell 6.3 cm. ; barom., 74.4 cm. ; temp., 23° ; fig. 23.) Generated by adding concentrated H2SO4 to sodium bisulphide (NaHSOs) and drying in P^O^. This same gas was fractionally liquefied, and then fractionally dis- tilled, but, as will be noticed, the curves coincide for the two samples, showing that nothing has been eliminated. This is a striking contrast to ethane, in which for successive purifications certain bands continued decreasing in intensity, showing that they were due to impurities. Compared with CO2 we have few examples which are more conspicu- ous in showing marked changes by substituting an S for a C atom. In the region where CO2 is transparent the SO2 has its greatest absorption bands. The one at 10.4 fi coincides with that of ammonia. Certain lines seem to belong to a spectral series, as determined by the law of " constant difference " of the wave-numbers. For example, take the following absorption maxima : 3-8>ro=6o5 i.to> ^0=610 Considering that the 8.7 fx. band is broad, and hence not well defined, the agreement is very close, but, unless we find more lines toward the visible spectrum, we can not determine the constants in the spectral series equations. The two curves given in fig. 23 are for ordinary SO2 and for the same gas when purified by fractional liquefaction and distillation. Hydrogen Sulphide. US. (Cell, 5.7 cm. ; barom., 74.0 cm. ; temp., 21° ; fig. 24.) Generated by adding HCl to ZnS, dried in PoOg. Oquefied in two fractions, one at — 15°, the other at — 60°. The sample liquefied at — 60° was distilled fractionally, and several fractions were examined, A sample which had been made from ZnS + HoSO^, but not liquefied, showed SO2 bands, which would naturally be expected. The changes wrought in the absorption spectrum by substituting hydrogen for oxygen are not so marked as that from CO2 to SO2. For HgS we have a greater number of lines, especially beyond 9 /*, but there are no deep bands like in SO2. It was noticed that the bands were not INVESTIGATION WITH A ROCK-SALT PRISM. 53 SO intense after the gas had stood a while in the mercury gasometer. Whether this is due to decomposition into HgS and H has not been determined, on account of lack of time. Certain samples tarnished the mercury, while others did not, after standing in the gasometer for some time. Sulphur has a band at 7.9 /x, while H^S has a band at 7.8 /t. Julius found the emission band of a sulphur flame at 7.85 /* (H2S?) instead of at 7.4 fi, which is the first large maximum in SOg. The second maxi- mum of SO2 occurs at 8.65 (i. It does not seem possible that in the sulphur flame he observed the mean of these two bands of SOg. Ammonia. NHs. (Cell, 6.3 cm., barom., 74.8 cm. ; temp., 22°. 5 ; fig. 25.) Made by heating NH4CI and solid KOH and drying over freshly heated CaO. Upon finding several bands in common with those of certain carbohydrates, and that NH2CH3 or NH2C2H5 might be present in the NH4CI, the latter was purified according to the method of Stas,^ by boiling with HNO3 for half a day. The gas was liquefied in order to remove the supposed water band at 2.95 fi. A sample was placed in a combustion pipette containing freshly heated CaO over mercury, for five days, when the band at 2.95 (i was found to be as intense as on pre- vious determinations. Moreover, the absorption band of water was found at the same time at 3.0 fi, which shows that the 2.95 fi band is characteristic of ammonia. It will be noticed elsewhere that this band is to be found in compounds containing amido (NH2) groups, as well as in certain ones containing nitrogen. Ammonia is one of the most interesting compounds studied, because of the numerous deep, narrow bands from 9/x to 13 /u,. At 5.7 ju. and 7.3 fj. there is evidence of existing bands, but the narrow dispersion of rock salt prevents their being resolved. This is the only compound studied having such a series of maxima, which are, in addition, so regu- larly distributed that it reminds one of the ammonia and the hydrogen spectrum in the optical region. Whether the " law of constant differ- ence " of the wave-numbers is true, or whether the coincidence in the values of certain wave-numbers is simply accidental, is difficiult to decide with the few examples at hand. We have the following examples : I II III ^;^>ro=ioo \t1s> '-'='' 11.18-^ rr "•43^ >r„=ixa ;i:^?>'--3' ii.98>^»-" ^Stas: Fresenius Zeit. f. Anal. Chemie, 6, p. 423. 54 INFRA-RED ABSORPTION SPECTRA. It thus appears that there are three subordinate series, but how are we to estabUsh the vaHdity of these observations ? More pairs of bands are necessary, but, as we approach the visible spectrum, the pairs of maxima must lie closer together, and, hence, can only be resolved with a larger dispersion. In addition to this, the regions at 2 /u, and 4 fi are transparent. This, however, can probably be remedied by using longer columns of the gas. In fig. 25 curve b is for a pressure of 24 cm. of mercury, while the curve thus, x x x x , is for the purified NH4CI. Absorption Spectra of Liquids and Souds. (Tables II, VI, and VIII a to f.) HAIvOGEN SUBSTITUTION PRODUCTS. iChIvOroform. CHCI3. (Fig. 26.) This is one of the first compounds investigated, and is of interest because it is one of the simplest combinations of carbon and hydrogen. Numerous small absorption bands are to be noticed throughout the whole curve, and one is reminded of the line spectrum found by Abney and Festing at 1.2/*. The wide band from 6.8 /a to 7.2 /a is complex. The 5.9 ju. band is very small here, but is to be noticed, since it occurs in so many other compounds having CH-groups. Since chloroform is one of the simplest compounds having a CH-group, one would expect to find absorption bands due to this group, provided it has the power to act alone. Turning to benzene (CgHe), which has a ring of CH-groups, we find that none of the conspicuous absorption bands are in common. The largest band of CHCI3 at 8.3 fi is shifted to 8.7 fi in benzene, while the 10.82 yu, band has disappeared entirely in the latter. If this disappearance of the 10.82 ju, band is to be attributed to the lack of CI in benzene, one would look for it in monochlorbenzene (CgHgCl). It occurs there (fig. 85) as a slight depression in the transmission curve. It also occurs in S and in benzaldehyde (CoHgCHO), which brings us back to the previous conclusion of the nonentity of the atom in the molecule. Iodoform. CHIs. (Fig. 27.) This is a crystalline solid, and hence difficult to obtain in a thin film. The film used was 0.05 mm, thickness, but, not being homogeneous, it increased the opacity by scattering the incident energy. As a result, the general transmission is reduced to about 40 per cent, and only two strong bands are resolved, at 8.6 [x and 9.3 fi, respectively. In chloro- form these bands are quite obliterated by others of greater depth. The 9.4 fi band is to be found in numerous other compounds. On melting, iodine was set free, thus coloring the film red, but this evaporated in a INVESTIGATION WITH A ROCK-SALT PRISM. 55 few hours. The iodine band at 7.3 fx. could not be detected, although special care was taken in the examination. This is due to the thinness of the film and the weakness of the band. The 3.3 ^i band is very weak, due to the opacity of the crystalline film. Julius^ found this same band very weak in bromoform (CHBrj). It is a little stronger and shifted in the simpler CH compounds. Carbon Tetrachloride. CCU. (Fig. 28.) This compound is noted for the absence of large absorption bands up to 6 n, as seems to be true of carbochlorates as distinguished from car- bohydrates. Two small bands, at 2.9 /a and 4.45 fi, do not appear in Julius's curves. The first deep maximum is at 6.5 /x. This band shows a break in the curve for the small spectrometer, and was found to be double, by using the large spectrometer, the maxima being at 6.45 /i, and 6.57 ft, respectively. The mean of these two is at 6.51 /i. Paschen^ found this band at 6.45 /i. Thinking that it would be a good check upon my work, the observations were repeated with the large spectro- meter (curve xxxx ), with the result that two bands were found. This is an excellent example of what may be expected if a large disper- sion can ever be applied to the region at 3 )u, for some of the other com- pounds. There are no further large absorption bands until we arrive at 13 /x. Here a broad, deep band exists which is harmonic with the double band at 6.5 II. Since the photographs of Abney and Festing^ show no lines for the region up to 1.2 fx, and since Puccianti* found no bands up to 2.5 11, it appears that the absence of bands up to 6/* is not due to the lack of resolution of the fine lines, but is more likely due to their entire absence. For this reason this compound ought to be an excellent solvent for studying selective absorbing solids in solution. (See Appendix IV.) The great change introduced by substituting four CI atoms for four H atoms is very evident by comparing this spectrum with that of methane (CH4), which has only two large bands, at 3.3 /^ and y.y fi, respectively. Tetrachlorethyeene. C2CU. (Fig. 29.) As in carbon tetrachloride, this compound has no absorption bands until we arrive at 5.5 /a. The large band of CCI4 at 6.5 /a is entirely wanting, while the 12.9 /x band is present, but is narrower. At 8/x are two bands common to CCl^. This compound has some of the properties ^Julius, loc. cit. ^Abney & Festing, loc. cit. "Paschen, loc. cit. *Puccianti, loc. cit. 56 INFRA-RED ABSORPTION SPECTRA. of ethylene (C2H4), but not a single band is in common, just as was found for CH^ and CCI4. The maximum at 6.4 /u, is closely harmonic with the one at 12,9 /x. The sample of C2CI4 used was presented me by Dr. W. C. Geer, who had prepared it with great care. Nevertheless the small bands at 8 /a seem to indicate the presence of CCI4, which may have been formed in the process of making of C2CI4. The latter is an intermediate product. However, the absence of the 6.5 fi band and the presence of the 12.9 /a band leaves the question of the presence of CCI4 somewhat in doubt. This, however, does not interfere with the comparison of C2CI4 with ethylene (C2H4). Ethylene Bromide. iCaHiBri. (Fig. 30.) (From Bender and Hobein ; boiling point, 131°.) The absorption spectrum of this compound is of special importance because the ethylene gas (C2H4) was made from it. The difference between the two absorption curves of these two compounds is found mostly in the general absorption. The C2H4 has a large triple band at 10.5 IX, where C2H4Br2 is quite transparent. The bands at 3.35 fi, 6.98 /i, 9.8 fi, 1 1.2 /A, and 13 ju, are found common to both compounds, although they vary in intensity. The C2H4Bro boiled at the theoretical tempera- ture, 131°, and must have been pure. The fact that so many bands are in common with ethylene indicates that the addition of two Br atoms has not had any great influence upon the original absorption bands, just as the substitution of a Br atom in benzene has not changed the benzene spectrum. Two pairs of bands at 8.0 //. and 8.44 /u,, and at ii.i4)U, and 12.0 |u, have a " common difference " of F= 66, which would indicate that they belong to a spectral series. WATER AND ALCOHOL. Water. H-O-H. (Fig. 31.) This substance has been extensively studied by Julius,^ Paschen,^ Aschkinass,^ and others, both as a vapor and as a liquid film — emission and absorption spectra. The investigators do not quite agree in the location of the maxima, although the work of each one, taken by itself, appears consistent. All have found an extreme opacity to infra-red radiation, so that the films of water had to be reduced to a few thou- sandths of a millimeter in order to be able to study it at all. Were it not for this property, water could be used as a solvent for studying sub- ^Julius : VerhandL Kon. Akad., Amsterdam, 1892. *Paschen : Ann. der Physik, 53, p. 334, 1894. ^Aschkinass ; Ann. der Physik, 55, p. 431, 1895. INVESTIGATION WITH A ROCK-SALT PRISM. 57 Stances in solution. The idea that water becomes more opaque when alum is dissolved in it is still to be found in some quarters, although Aschkinass has made a special study of it, which gave negative results. The work of E. F. Nichols^ also disproves this belief. The work of Aschkinass is of special interest, since he examined pure water, instead of a salt solution, as did Julius, and found no difference in the maxima for the two methods. The maxima differ somewhat for different kinds of water, depending, no doubt, upon its purity. A series of absorption bands at 3.06 /i, 4.7 /a, and 6.1 ix are of impor- tance in connection with the present work. First, the 6.1 ^l band is just double the 3.06 fi. Paschen found the first maximum at 2.916 fi, 2.97 fi, and 3.024 II, depending upon the thickness of the film, and the second maximum at 6.061 fx, which is a fair agreement, considering the diffi- culties in calibrating the apparatus. The 6.1 fx. band is as closely harmonic with the 3.06 /u, band as one can expect. A glance at the curves shows this. The bands are broad, and the exact location of the maximum can not be determined beyond the second decimal place. Curve c in fig. 31 shows the transmission through a gypsum crystal (CaSo, + 2H2O) 2.58 mm. thick. The bands at 1.5 ix and 2.0 IX coincide with those of pure water. This shows that the mole- cule of the water of crystallization is in a condition like that of pure water. This curve is due to Konigsberger,^ who made a study of the absorption of doubly refracting crystals. He found that muscovite mica (H^KaAlgSigOg), which contains water, or, rather, hydrogen and oxy- gen chemically combined, shows a band at 2.95 /x, but no band at 1.5 /x. Julius and Aschkinass found a layer of water 0.17 mm. in thickness to be almost opaque beyond 7 fx. The region from 7 /x to 14 /a has been re-explored by the writer, using a salt solution of water, about 0.03 mm. in thickness, between rock-salt plates. This film decreased gradually in transmission of 6 per cent at 7 /x to opacity at 12 /x. It thus becomes evident that no water solutions nor compounds containing water could be used in this work. Curve e is due to Rubens and Aschkinass,^ who examined CO2 and water vapor to 20 fx. No explanation is offered for its great transpar- ency at 1 1 /t as compared with liquid water. Water is the most opaque substance examined ; in fact, beyond 2 /x it is the most opaque substance known. Further work on the question of water of cr}-stallization is given in Appendix V. ^Nichols, E. F. : Phys. Rev., vol. i, p. i. ^Konigsberger : Ann. der Physik, 6x, p. 687, 1897. ^Rubens & Aschkinass, loc. cit. 58 INFRA-RED ABSORPTION SPECTRA. Since, in all infra-red work, the calibration consists essentially in setting for minimum deviation of some known line, like the D lines, in the visible region, and using that for a base, or " zero setting," and since a slight error in setting for minimum deviation or in the " zero setting " may introduce great errors in the infra-red, it is unfortunate that Konigsberger did not compare the 2.95 fx water of crystallization band with that of pure water, found by Paschen at from 2.916 /x to 3.024 /A and by Aschkinass at 3.06 jx. The same is true of Ransohofif,^ who considers the alcohol band at 3.0 /a in coincidence with the 3 /a band of water, as found by Aschkinass, though he made no direct comparison. In the present work, using the large spectrometer, the best value for the water band is 2.95 /n, which agrees with the value found for water of crystallization. (See Appendices V and VI.) Alcohols. (Figs. 32 and 38.) In the present work alcohols were but briefly studied, because of their great opacity beyond 7 /a, as well as on account of the difficulty of free- ing them from water. The higher alcohols, like glycerin, even if they could be freed from water, are so hygroscopic that they are difficult to investigate. Myricyl Alcohol. CmHcoOH. (Fig. 32.) Ethyl alcohol was studied with the small spectrometer, and has been discussed with the other simple CHg compounds. Amyl alcohol was examined at 3 )u, with the large spectrometer to compare with the work of Ransohoflf.^ The maxima were found at 2.95 ju, and 3.43 //.. Ranso- hoflf found these bands at 3.0 /x and 3.43 /a. The only other alcohol which has been studied thoroughly is myricyl, which is a solid from beeswax. A very thin solid film was obtained by melting between two plates of rock salt. The spectrum is very marked for several strong absorption bands which correspond to those in the petroleum distillates. The dispersion with the large spectro- meter at 3 ju,, curve h, is comparable with that of fluorite, and hence comparison can be made with the work of Ransohofif. In the present work the maxima occur at 1.71 /a, 2.95 ^i, 3.43 ii, and 5.8 jx. The water bands were found at 2.95 /x and 6.0 jm, so that the 2.95 /x band of myricyl alcohol coincides with the one for water, while the one at 5.8 //. does not. The discrepancy at 2.95 ix and 3 /* may be due to inaccuracies in the dispersion of rock salt as compared with that of fluorite. The rock- salt dispersion curve passes through a double curvature at this point, ^Ransohoff, loc. cit. INVESTIGATION WITH A ROCK-SALT PRISM. 59 hence difficult to determine, especially since data are almost entirely lacking at 2.9 /x, just where they are most needed. ( See Appendix VI.) Two examinations were made, the first with the myricyl alcohol just taken from the containing bottle. For the second examination several grams of the myricyl were heated to 110° for seven hours in a drying oven, which was sufficient to expel any water present. Immediately after this heating a thin solid film was examined, and the bands at 2.95 /* and 3.43 fjL coincided exactly with those found previously, which would indicate that the 2.95 fx band is not due to water. Whether it is due to the OH-group is a different question. If it is due to OH, then the 5.8 iM band should coincide with the water band found by Aschkinass^ at 6.0 IX. The spectrum of myricyl alcohol consists essentially of five large bands. The bands at 1.71 fi, 3.43^1, 6.86 /u., 10.2 ju, and 13.9^1 are closely harmonic. These large bands occur in ethyl alcohol, where they are very much broadened and obliterated in details, due to the greater opacity of the film investigated. Like water, the alcohols are very transparent up to i fi, beyond which point they suddenly become more opaque. In connection with these alcohols, phenol (fig. 99) and menthol (fig. 117) are also to be noticed, since they show the 2.95 /x band, which is characteristic of the alcohols. COMPOUNDS CONTAINING THE METHYL (CHg) GROUP. Under this heading have been collected a number of compounds that contain CHg-groups. This method of discussing these compounds is of great help in considering certain absorption bands which are prob- ably due to the methane radical. Thus the 3.45 fx band found in 8 compounds (mostly alcohols) containing CH3 led Julius- to believe that it was due to that group of atoms. In this present investigation, which covers nearly the whole field of organic chemistry, it remains to be seen how constantly this band occurs. NiTROMETHANE (Methye Nitrite). CH3NO2 ; CHg-O-N. (Fig. 33.) In this compound the introduction of the NOj-group has not dis- turbed the 3.41 fx band. Several new ones are to be observed from 4 to 6 /A, and are more prominent when examined with the large spec- trometer. The region from 6 to 7 /* is noticeable because of its great opacity, due to a number of large absorption bands. Beyond this there is the same general transparency of 80 per cent, as in the first part of the spectrum, interspersed with sharp absorption bands, especially at 9.1 IX. ^Aschkinass, loc. cit. ^Julius : Verhandl. Konikl. Akad., Amsterdam, Deel I, No. i, 1892. 60 INFRA-RED ABSORPTION SPECTRA. NiTROETHANE. C2H5NO2. (Fig. 33.) The 3.4 )u, band is still undisturbed. In the region from 6 fx. to 7.5 fi the addition of a CHa-group seems to have brought about a greater freedom of vibration, and, instead of a continuous band, we find several small sharp lines. The same is true of the 7.9 /x and 10 /x bands, which are also more prominent. There are two new bands located at 11.45 /a and 12.45 P-> as well as the 14 /x, band, which is of frequent occurrence in methyl and ethyl compounds. On viewing these curves with their profusion of absorption bands, one can see why nitrogen compounds are so opaque as compared with the sulphides. The thickness of the cell used was only o.oi mm., while in ethyl sulphide the thickness was ten times as great. In these two compounds the N atom is thought to be trivalent. The spectra are entirely different from nitrobenzene. Methyl Cyanide. CH3CN. (Fig. 34.) In the pyridine group mention was made of the great opacity of nitro- gen compounds. In those cases the N or NH replaced a CH-group in the ring compounds. In the present compound the N atom is joined to the C atom. The result is a more transparent compound like C2H5CN. The 3.4 fj. band appears complex, while a small depression appears at 3.0 /x. The 4.37 fi band occurs prominently here, as it does in a few of the succeeding compounds. No band occurs at 5.7 /x, but a small one is found at 6.2 /x, ending in a large band at 7.0 /x. The 8.2 /x band of CH3I is wanting, while the 10.5 /x and 14 /x bands, to be noticed fre- quently in other compounds, are very weak. As a whole, an entirely new series of lines has been produced, most of which are very sharp. The 7.08 /x band is to be noticed in connection with the ethyl cyanide. The band at 3.0 ja is to be found in many compounds containing N atoms. The curves are practically identical for the museum and Kahl- baum's samples. Ethyi, Cyanide. H2C5CN. (Fig. 35) The difference between methyl and ethyl cyanide is as marked as that of the respective iodides. This is especially true at 11 /x, where the large band of CH3CN is entirely wanting. The 9.6 /x band of CH3CN is found at 9.35 m. The transparent region extending from 11.5/x to 13.5 fi in CH3CN is shifted so as to extend from 10.5 /x to 12.5 /x in C2H5CN. The 3.4 fi band occurs in its proper place, while the y.y /x band is probably a new one. The unusually sharp bands at 6.98 /x, 9.3 IX, and 12.75 M are very marked, and they make it appear that oxygen is not the only element that sharpens the bands. The 6.98 /x band of C2H5CN is not the same as the 7.04 ai of the CH3CN. To make sure INVESTIGATION WITH A ROCK-SALT PRISM. 6l of this these two substances were examined on the same day, and the apparent shift is simply due to the fact that the CH3CN band is com- plex, with its second component at 7.25 fx. This is an excellent exam- ple of what might have been considered a shifting of the absorption band, with change in molecular weight, if a smaller dispersion had been used. In C2H5SCN the CN part is joined to the S atom. The result is a shifting or stretching apart of all the strong bands, beginning at 4.4 /* and extending to 10.4 ;u,, where a transparent region occurs which extends to 12,8 fi, just as in the CgHgCN. Methyi, Iodide. CHsI. (Fig. 36.) This compound gives a curve conspicuous for three regions of great absorption, viz, 3.4 /x, 7.2 /a, and 11.4/n, while there are two transparent regions, at 9.5 /* and 13 /x, where ethyl alcohol is opaque. At 10.5 /* there is gentle decline in transmission, with a possible band at 10.8 /*, and terminating in a large minimum at 1 1 .3 fi. Methyl iodide is quite transparent, 60 per cent, if considered in height of the transmission minima and thickness of absorbing cell, also if one were to consider the total transmission. It will be noticed that these great absorption bands lie in the region where the radiation from a black body is very weak, while the 3.4 /* band is shallow, so that the great transparency of CH3I observed by Friedel^ (using Tyndall's method of total absorption), as compared with nitrogen compounds, is due apparently to the lack of strong absorption bands in the region of strong radiation. That the transparency of a medium depends upon the radiation employed is shown in a striking manner in Drew's^ work on vacuum-tube radiation. He used CSg as an absorbing medium, and found it as opaque as water, due to the fact that the emission is concentrated in a strong band at 4.75 fi, which coincides with a strong absorption band of CS2 at 4.7 jit. No absorption band is to be found at 7.3 [x, where solid iodine has a large band. The 5.95 [x band is only slight. The 4.6/* band is to be noticed in connection with the several of the simpler CH3 compounds mentioned in the following pages. It is found in only a few compounds. Ethyl Iodide. C2H5I. (Fig. 37.) Like the preceding, this substance is quite transparent, so that a 0.16 mm. cell could be used except beyond 13 /x, where it is more opaque. The region from 6 to 12 ju, is conspicuous for its regions of great transparency and of opacity. The ii.4ju, band of CH3I is absent, and ipriedel ; Ann. der Physik (3), 55, p. 453. 1895. '^Drew : Phys. Rev., vol. xvii, p. 321, 1903. 62 INFRA-RED ABSORPTION SPECTRA. in its stead is a large band at 10.5 jx. It will thus be seen that the three large bands of CH3I occur in slightly shifted positions in CgHgl, and, in addition to this, they are also more prominent in the latter. The bands from 3-4/* to 6/* are more prominent in this compound, while beyond 12 fx several new ones occur. As a whole there are a greater number of transmission minima in C2H5I than in CH3I, which is to be expected, if absorption depends upon the groups of atoms in the molecule. The 14 /x band is overshadowed by the one at 13.6/1,. With the 0.05 mm. film, the band at 7.15 /a occurs at 7.06)11, the side's appearing unsymmet- rical. In photographic spectra such examples of the blurring of the band is of frequent occurrence. As with other compounds, no shifting of absorption bands could be detected. Ethyi^ Alcohol (Absolute). C2H5OH. (Fig. 38.) As already mentioned, a number of alcohols have been examined by Julius and found to become suddenly opaque at 7 fx, due to thickness of the cell of 0.2 mm. The curves (fig. 38) explain this, as will be noticed in the great opacity which extends from 7 to 10 /x, beyond which there is a transparent region at 10.5 ju., and again opacity beyond 13 /a. Puccianti^ has also examined this compound and several of the iodides up to 2.5 fx. He found that all compounds in which carbon is directly united with hydrogen showed absorption maxima at 1.71 /a. In the above curve there is a slight depression at 1.6 jx, which represents the band found by him, but, on account of the narrowness of the dispersion in this region, no attempt has been made to adjust this disagreement. This obliteration of the band is due entirely to the overlapping of the slit, on account of narrow dispersion. The 3.45 ix band is broad for the 0.18 mm. cell, but appears complex for the 0.04 mm. cell. No doubt the 7.3 IX band is likewise a composite of two or more bands, which would be resolved with greater dispersion. The 5.95 fx band is only a slight depression. The curve is of importance in showing the freedom from water, which the alcohols so easily absorb. Ordinary ethyl alco- hol and glycol were found to be opaque beyond 7 fx, using a thin film, like the one above. At its best, however, alcohol is very opaque to infra-red radiation, as seen in curve a, 0.18 mm., which is opaque at 6 IX, and continues thus. Curve h, 0.02 mm. thickness, was repeated, using a new film to be certain that the transparency at 8.5 /x and 10.5 /x is real. This sample had been treated with sodium and contained per- haps about 0.4 per cent of water. There is quite a contrast between this compound and the others considered in this group, all of which are much ipuccianti, loc. cit. INVESTIGATION WITH A ROCK-SALT PRISM. 6^ more transparent. Ransohoff/ using a larger dispersion, found the region a complex of two bands, at 3.0 /x and 3.43 /*, while the band at 5.2 /x is deeper and better defined. EXHYLHYDROSUtPHIDE. C2H5SH. (Fig. 39.) The low boiling-point, 36°, of this ill-smelling compound makes it a difficult subject for investigation, as will be noticed from the curves in fig- 39. which were obtained after giving the cell an additional coating of paraffin to prevent leaking. In depth and number of transmission minima the curve agrees exactly wdth that of Julius, and was used to reduce his values (given for his straight-line extrapolation from 5 fi) back to the true dispersion curve. The last band found by Julius was the one at 10.4 fi. No attempt was made to locate accurately the band found by him at 3.88 /i. The curve is quite simple. The deep double band at 7.0 fi and 8.0 fx is strikingly repeated in ethyl sulphide. The 5.9 /a band is not promi- nent, while the ones found in sulphur at 7.9 /t and 12 fi are replaced by a strong minimum a.t S fi and another at 11.6 fi. As a whole this com- pound is quite transparent. Ethyl Sulphide. (C2H5)2S. (Fig. 40.) As compared with ethylhydrosulphide this compound is very similar in general absorption, as well as in the location of absorption bands. The band at 6.86 /x is worth noticing, since it occurs in so many other compounds. The addition of a CaHg-group for an H atom has not seri- ousty disturbed the spectrum of the hydrosulphide. The latter repre- sents some of the earliest work, using a wider slit, and the evidently complex band at 7.3 fi is resolved in the present compound. In com- parison wath these sulphides it wall be noticed that the allyl is far more opaque. TriethylaminE. (CsHOsN. (Fig. 41.) In this compound we have an additional ethyl group, while the S has been replaced by an N atom. The result is an increase in the com- plexity of the region from 7 to 10 fi, thus making the compound more opaque as a whole. The greater transparency at 11 to 13 /x is quite a contrast to ethyl alcohol. There is not so marked an effect of the N atom at 3 /x as in the case of aniline and several other benzene deriva- tives. This compound has a strong ammonia odor, and is similar to it, yet with the large spectrometer no marked band of ammonia could be detected at 3 /x; in fact the curve has no resemblance to that of ammonia. ^Ransohoff, loc. cit. 64 INFRA-RED ABSORPTION SPECTRA. The 343 /i, 6.86 /i, 9.3 /a, and 13.6)11 bands are found in the petroleum distillates. The great difference between this curve and that of the iodides is significant of the effect produced by introducing such ele- ments as I, S, and N into the methyl and ethyl compounds. Allyl Sulphide. (C3Hs)2S. (Fig. 42.) H H2 The structural formula of this compound shows that H2C=C— C\^ ^hg CHg-groups of atoms predominate. There are also UiC=C—c/ two CH-groups. But when we compare the absorp- ^ ^2 tion curve with that of piperidine, which compound has a ring of CHg-groups, we find so similarity between them. The same is true when compared with benzene (CeHg), which has a ring of CH- groups, so that the compound seems to have a spectrum peculiar to itself, and the question of groups of atoms is of less significance than the manner of bonding of the groups with others in the molecule. In gen- eral, the sulphides have been found quite transparent. Allyl sulphide is an exception, which, according to Friedel,^ is to be accounted for by the preponderance of H atoms. Ethyi, Sulphate. (C2H6)2S04. (Fig. 43.) This compound decomposes when exposed to light. The sample used was distilled just before using. The distilling had to be done fractionally under reduced pressure. Since the best sample obtained had a slight odor of SO2, the exploration of its absorption spectrum was not continued beyond 7 ix, where there was complete opacity for the cell used. The 3.45 ix and 5.95 fi transmission minima are of interest in considering the action of the ethyl radical. Methyl Acetate. CH3OOCH3. (Fig. 45.) Methyl acetate belongs to the fatty acids and shows some similarity to that group. The region at 3.3 /x reminds one of phenyl acetate, while at 1 1 ju, the transparent region is similar to that of alcohol. This compound is very opaque, and its study has not aided materially in determining the effect of CHg-groups. The region of great opacity, from 6 to 9 yu,, is to be found in methyl carbonate, ethyl succinate, and several other compounds containing oxygen. Acetone. CHsCOCHs. (Fig. 46.) Acetone differs from methyl acetate in having a C atom substituted for an O atom in the latter. ^Friedel, loc. cit. INVESTIGATION WITH A ROCK-SALT PRISM. 65 This compound was studied in the vapor state, and also as a Hquid at 3.4 /i. No difference could be detected in the location of the 3.4 fx band for the liquid and vapor phase. The acetone curve has a great similar- ity to that of methyl acetate and methyl carbonate, especially at 4.75 /*, 5.75 [X, and 8 /i, but the great opacity and breadth of the absorption bands prevents a minute study of them. The large spectrometer failed to resolve the region at 3.3 jn, just as was found for the fatty acids at 7 jx, and as Ransohoff found for the alcohols at 6 /x.. Methyl Carbonate. CO 22ti.42- It will be noticed that some compounds appear a little more trans- parent than others in certain regions of the spectrum. This no doubt is due to absorption bands which are obliterated in some and do not exist at all in others. For example, dodecane (CigHje) has four small absorption bands between 4 and 6 \x., while there are none in octadecy- lene (CigHgg), but they are so small, i. e., narrow, that the large spec- trometer was necessary to detect them. The solids were more trans- 74 INFRA-RED ABSORPTION SPECTRA. parent because of the thinner films thus used. On account of evapora- tion they could not contain any quantity of the liquids boiling at ioo°, so that the absence of shifting can not be attributed to the presence of hydrocarbons of lower boiling point. The different hydrocarbons studied, with their boiling points, are given on page 73. The petroleum distillates are conspicuous for three regions of absorption, at 3.43 ^t, 6.86 /t, and 13.8 /i. The spectra are entirely different from those of the fatty acids, which have an addition of two oxygen atoms. In the methylene series it is not known whether the Ohio oils (C10H24 and C^Ji^^) have the same structure as that of the California oils. The Pennsylvania oils (C.Ji^^ to Ca.H^g) are probably the same as the Ohio oils. They occur with the solids of the paraffin series, and were separated from them at —10° by filtering under pressure. The impure paraffin series of distillates are a mixture of solid and liquid. The solid was separated in a crude state by cooling and filtering ; it was then purified by dissolving in ether and precipitating it from solution by means of alcohol. The chlorine compounds, C.Jio^ CI [b. p. I30°-I35° (12 mm.)] Ci.H^sCl [b. p. I45°-I50° (20 mm.)] and C^,U^, CI [b. p. i7oJ-i73j (15 mm.)], were obtained from chlorinated oils boiling at 223°-224°, at 240°, and at 189° (50 mm.), respectively. The band at 5.8 ju, seemed to vary in depth when examined at different dates. This was thought to be due to the absorption of moisture, but a subsequent examination, with the large spectrometer, of a specimen which had stood over P2O5 for several weeks, and the same specimen after it had been exposed to the moist air of the room, showed no differ- ence in the absorption band, which was small. Since the samples obtained were small, varying from a few drops to 2 cc, it was impossible to distill them, so they were placed in wide-mouthed bottles containing P2O5, and left there from several days to several weeks, depending upon their tendency to evaporate. This drying made no difference on the absorption spectra, and it was concluded that the oils were free from moisture. Hexane. CaHu. (Fig. 62.) This sample came from Kahlbaum, and is of importance here, since it belongs to the same group of compounds as the petroleum distillates. On account of its low boiling point, 69°, it had to be inclosed in a thicker cell than the oils, hence it appears more opaque. Except for a new band at 8.82 fi and a slight shift at 9.42 fi, the absorption spectrum is the same as for the oils. This is a synthetic product, which fact adds interest to the agreement of its absorption spectrum with the other com- INVESTIGATION WITH A ROCK-SALT PRISM. 75 pounds. The general absorption is somewhat different than the oils, which indicates more bands not resolved with the small spectrometer. DODECANE. CuH28. (Fig. 64.) In addition to the small spectrometer examination, the large one was used, with the result that four small bands were found between 4 and 6 fi. The other bands are much deeper because a film o.oi mm. was used instead of a cell of 0.15 mm. thickness, as for the small spectro- meter. This shows that the bands are sharp and narrow, and for this reason are blotted out when using the thicker cell and the smaller dis- persion. The bands at 3.43 fi and 6.86 /i are harmonic. OcTADEcvxENE. CisHse; Tetracosylene, CmHis. (Figs, yz and 74.) These two compounds were also re-examined with the large spec- trometer, but no new bands could be found between 4 and 5.5 {i. The bands at 3.44 and 6.87 fi are harmonic, and have no indications of being complex. A small band exists at 6.6 /n, where the curve is very asym- metrical for the small spectrometer. It has just been noticed that dodecane (CiaHge) has several bands in the region where these two compounds are transparent. It would have been interesting to determine whether this is due to the difference in structure of the two series of compounds. It seems quite probable that this is due to structure, but time did not permit an examination of more compounds, using the large spectrometer, in order to thoroughly estab- lish this fact. MoNocHLORTRiDECANE, CisHkCI ; Monochlorheptadecane, CmH»CL (Fig. 7$.) The two chlorine compounds studied do not happen to be substitu- tions in the distillates examined. However, since they have the same maxima as all the disillates, it shows that the introduction of a chlorine atom has little effect on those maxima. HYDROCARBO>f. CioHmS. (Fig. jS.) This product was obtained from the sludge of Ohio petroleum. Its properties are unknown. The transmission curve is essentially the same as that of the oils. It seems more transparent than the oils beyond 1 1 /n. ASPHALTUM. (Fig. 44.) Asphaltum is a mixture of different hydrocarbons, part of which are oxygenated. It is of variable composition. Two specimens were examined. The film of varnish was first examined, after it had been drying for several months. Thinking that some of the solvent might 76 INFRA-RED ABSORPTION SPECTRA. Still be present, a film was made by melting solid asphaltum on rock salt. The whole seemed of interest in connection with the absorption and anomalous dispersion of asphaltum investigated by Nichols^ in the optical region. The curves show several marked bands. As is to be expected, the bands are found in the petroleum distillates. Class II : Carbocycuc Compounds. In this class the carbon atoms are joined in a closed chain or ring. To it belong the methylene hydrocarbons of the petroleum distillates, already discussed, the pyridine group, thiophene and pyrrol, and, most important of all, benzene and its derivatives. As will be pointed out elsewhere, the benzene spectrum is so wholly unlike that of the petro- leum distillates that, if we had no knowledge of the latter, gained from organic chemistry, the evidence presented in their absorption spectra would be sufficient to show that we are dealing with two distinct classes of compounds. In a more restricted sense, pyridine, thiophene, and pyrrol belong to the so-called heterocyclic compounds of carbon. Benzene. CeHe. (Fig. 77.) Benzene is the parent hydrocarbon of a large number of compounds. The idea that the constitution of benzene is a closed chain or ring of carbon atoms was first pro- pounded by Kekule,^ in his " Benzoltheorie," in 1865. It is based on numerous facts, such as the power to form three isomeric biderivatives, which is not possible in an open chain like stearic acid, but which is possible if each C atom is joined to an H atom and the six CH-groups are joined together in a ring. This forms a fairly complicated molecule, not easily reduced to simpler com- pounds, like CO2. These facts should be remembered in considering the following curves, in which certain benzene bands persist even in very complicated derivatives. The relations of the benzene derivatives to benzene are very limited,^ although the derivatives are intimately connected by many reactions. This fact is not out of place in consid- ering the following curves. The great dissimilarity between them and the curve of benzene also shows that the relations are limited. Nothing has been found more marked than the change in the maxima by substi- tuting a CI or Br atom for an H atom in benzene. On the other hand, the curves of the CI and Br derivatives show many bands in common. ^Nichols, E. L. : Phys. Rev., xiv, p. 204, 1902. ^Kekule : Liebig's Annalen der Chemie, 137, p. 129 (1865.) ^Bernthsen : Organ. Chemie, p. 330; also, p. 327. INVESTIGATION WITH A ROCK-SALT PRISM. yj which makes it appear that the facts gained by chemical and physical analysis are intimately related. Benzene shows a great transparency throughout the whole spectrum to 14 /x, where it suddenly becomes opaque. The well-defined trans- mission minima are seven in number. The 7.3 /a band found by Julius^ occurs only as a slight depression, and the possibility of its being due to thiophene, which has a strong band in this region, is mentioned in the discussion of that substance. This sample was known to be free from thiophene. Certain regions of the spectrum were re-examined thor- oughly, but no new transmission minima could be found. The shorter curves show slight depressions sometimes, but no deep bands occur in them. The cell containing the benzene was dismounted and stood aside for two weeks, but on examination no change was detected in the maxima. The curve found was so similar to the first that it was quite evident that the benzene had not attacked the glue of the cell in the meantime. The complex band at 9.8 /x to 10.3 \i is to be noted, also the 1 1.8 /i, band, which is in common with other compounds ; also the 3.25 \i. and 6.75 /A bands, which occur even in very complex derivatives, show- ing that the vibration of the benzene ion is not destroyed. In benzene and its derivatives the 5.8 ju, band is not to be found except in those containing CHg-groups and in benzaldehyde. Whether this band is due to CH3 or some group of atoms vibrating like a CH3- group is a pertinent question. The slight band at 6.25 /i is to be noticed, since in certain derivatives it is very strong, just as though the vibrations were less damped. This is the only example found in the whole research. METHYL DERIVATIVES OE BENZENE. Toluene. CeH^— CH3. (Fig. 79.) p.p„ By studying a group of benzene derivatives, increasing HC^ ^CH in the number of CHg-groups, it was hoped that the effect HcL CH °^ those groups could be determined, as well as the shift- ^^^ ing of an absorption band, with increase in molecular weight. The introduction of a CHg-group has shifted the 3.25 fx band of ben- zene to 3.34 fi. The one at 6.25 /x has been strengthened, while the 6.75 /x is quite obliterated by the 6.86 fi maximum found in the petroleum distillates and other chain compounds. This is shown to better advan- tage in the curves obtained with the large spectrometer, which also show fine bands at 5/^, not detected with the smaller apparatus. Whether ^Julius, loc. cit. 1 ^8 INFRA-RED ABSORPTION SPECTRA. the shift at 3.3 fi is real will be discussed elsewhere. It will be suffi- cient to notice that in mesitylene the 6.2 /* band is double (6.1S /x and 6.3 fj,) , so that with a smaller dispersion it would appear as one band, with the maximum shifted. At 6.1 /x and 9.3 fi the bands are in common with thiophene. The sample was labeled "free from thiophene," so that this may be mere coincidence, rather than impurity. AnISOL. CeH^^-^CHa. (Fig. 78.) Here the benzene maximum is shifted from 3.25 ix to ^CH 3-32/*' With the large spectrometer numerous small I bands were detected in the transparent region from 4 to HC\^^^H ^ ^^ while the 6.8 /u, band was found double with maxima " at 6.65 /x and 6.86 /x. The whole spectrum is composed of deep, narrow bands, which reminds us of the observation of Abney and Festing that oxygen sharpens absorption bands. The strong band at 6.22 IX, which is only a slight depression in the benzene curve, is analo- gous to the work of Liveing and Dewar,^ who found that mixtures of metals (Mg and K) often showed lines not seen at all when these ele- ments are taken separately. It also strengthens the belief, mentioned elsewhere in this work, that many apparently new bands are not really new, but that they are due to a condition in the molecule such that the original vibrating ion is less damped by its neighbor. The spectrum is entirely rearranged as compared with the original benzene spectrum. Ortho, Meta, and Para Xylenes. C8H4(CH3)2. (Figs. 81 to 83.) \Ortho\ \Meta\ \\Paj'a\ HC^ ^CH HC^ ^C-CH3 HC. JcW The three isomeric xylenes afford an example for investigating the structure, as well as the question of bonding, in the molecule. As found by Pauer^ for the ultra-violet, the spectrum is channeled, i. e., the bands occur in groups, in the origins of 3.4, 6.7, 9.6, and 1.3 /a. The first maximum for the three compounds occurs at 3.38 /x. They have great similarity to 6 \x, where they begin to show the effect of structure. At 6.8 ix they show great differences in the appearance of their curves. Apparently the ortho is shifted the most, while the \>ara appears double. This is best shown in the curves obtained with the large spectrometer. ^Liveing & Dewar : Kayser's Spectroscopy, II, p. 251. ^Pauer, loo cit. INVESTIGATION WITH A ROCK-SALT PRISM. 79 The maxima occur in the following order of increasing wave-length : [Para, meta, ortho.] At 6.8 ,« At gn At 9.7 /i At 13// P (m),p P P (P) Apparently the ortho, in which the CHg-groups are the closest to- gether in the benzene ring, has the greatest shifting of the maxima. If we take the center of gravity of the groups of maxima found by Pauer^ in the ultra-violet, the shift of the long waves is just the reverse, viz, o, m, p. However, we can not place great reliance upon the location of the center of gravity determined in this manner. In cumene, in which the CHg-group is joined to the benzene ring by means of a CH- group, the band lies at 13.4 fi, while in orthoxylene it is at 13.6 /^. The probable significance of this must be reserved for later discussion. With the large spectrometer a trace of the 3.25 /a band is noticeable in the orthoxylene curve, while the 6.2 fi to 6.3 /x band is double, as is also the 6.81 fi band, which is the mean of 6.75 fx and 6.86 /x. This would indicate that the vibration of the benzene nucleus has not been destroyed. The para is still more complex at 6 /x, having an additional maximum at 6.55 fi. These three compounds furnish the best evidence in favor of a resonance efifect of small, electrically charged particles. In the orfJio, which has the two CHg-groups nearest together, the large absorp- tion bands lie farthest toward the long wave-lengths, while in the para, where the CHg-groups are separated the most, the large bands lie far- thest toward the short wave-lengths ; the meta has its bands in inter- mediate positions. This is right in line with the idea that absorption is due to a resonance of small, charged particles, whose capacity, and hence whose period, depends upon their proximity to similarly charged parti- cles ; the closer the particles, the greater the capacity ; hence the slower the period, and hence the farther is the absorption band shifted toward the longer wave-lengths. (See Appendix IV.) Mesitylene. C6H3(CH3)2. (Fig. 80.) By substituting three CHg-groups, the 3.25 fi band is entirely obliterated by the one at 3.4 fi. The region at 6.18 to 6.3 fi is similar to that of the other methyl deriv- atives. The curve as a whole is very similar to toluene, except at 5,5 fi and the new band at 1 1.95 /i. From the asymmetry of the curve there are indications of the 3.25 fi band of ben- zene. All these methyl derivatives are very transparent. The benzene HC HCv '^ ^C-CH3 H ^Pauer, loc. cit. 80 INFRA-RED ABSORPTION SPECTRA. cell was i6 times as thick, so that no comparison can be made for total absorption. According- to Magini/ the absorption of isomeric com- pounds in the ultra-violet increases in the order meta, ortlio, para. The present curves do not show this, except a possible indication at 4/*. Mesitylene is isomeric with cumene. Their absorption spectra are en- tirely different. VARIOUS OTHER BENZENE DERIVATIVES. MonochlorbEnzene. CeHsCl. (Fig. 85.) This compound is of about the same transparency as benzene. How- ever, the substitution of a CI atom for an H atom has wrought a marked change in the general outline of the benzene curve. The lower curve was made with a wide slit and a 0.19 mm. cell, hence serves only for a g-eneral comparison with benzene. The upper curves were made with a narrow slit, while the parts of curves were made with the large spectrometer. They show that the original bands of benzene at 3.25 ju, and 6.75 )u, still exist. There are new bands at 6.27 /* and at 6.94 /i. The 6.27 ju, band is a characteristic of benzene derivatives, and strength- ens the belief that it exists in benzene, since that curve has a slight depression in this region. It adds weight to the belief expressed fre- quently in this paper, that the substitution of a new element is not so much the cause of new bands as it is the cause, in some manner, of a greater freedom among the original benzene ions, whereby those bands are intensified. Comparing same thicknesses of cells, the benzene curve has been radically changed, which seems to show that the limited chemical rela- tion of benzene to its derivatives is also to be found in the absorption curves. MONOBROMBENZENE. CoHsBr. (Fig. 86.) In this substance we have a striking similarity to the preceding com- pound. In fact, the curves for the same thickness of cell seem almost superposable, excepting at 8 /a. The region of 5.7 /t shows great simi- larity, as does the 6.9 /a, ii.I/x, and 12.2 /x band. The general trans- mission seems to progress in two steps ; the first, of 75 per cent, extends to 5 /x; the second, of 40 per cent, extends to 13 /x, where the substance becomes opaque. The same facts are to be noted in CeHjjCl, and the discussion of that substance applies to this one. It is unfortunate that time did not permit a further examination of this compound, using a larger dispersion. As in C0H5CI, the 3.25 /x band of benzene still exists. ^Magini, loc. cit. INVESTIGATION WITH A ROCK-SALT PRISM. 8l Safrol. CioHioO.. (Fig. 87.) In safrol quite a number of points are to be ^C^H2-CH = CH2 noticed. It is far more opaque than benzene out ]| I to 12 IX, where it becomes more transparent. As H*^\ -i^'^^CH? in several other benzene derivatives, the absence C — O "^ , ..... of common transmission minima is conspicuous, except the one at 12.9 yu.. The 9.1 /x band is also to be found in other compounds not related to benzene, e. g., piperidine, which has a ring of CHa-groups. The curve, b, of safrol, at 3.42 ^ was obtained by using the large spectrometer. It shows the absence of the 3.25 /a band of ben- zene, while the 2.9 /a band of eugenol, to which it is related, is very small. In safrol the bands are sharp and deep, and are of importance in con- nection with the idea propounded by Abney and Festing^ that the O atom sharpens the absorption bands. The large spectrometer failed to detect new bands at 3 11, but several were found at 6 /x. This is an example where the benzene ring vibration is almost entirely overshadowed by that due to the preponderance of CHg-groups. CUMENE (ISOPROPYL BEN^ZNE). CeHr— CH=(CH3)2. (Fig. 84.) In this compound we have the original benzene ring complete, except for one H atom, which has been replaced by a CH-group, to which is attached two CHg-groups. The result is a struggle between the benzene ion and the CHg-groups, and there is a compromise for different parts of the spectrum. Thus the maximum at 3.43 ix is a little greater than that for mesitylene, in which three CHg-groups are joined directly to the benzene ring. At 6.2 /a we have the band common to the xylenes. At 6.65 /A there is a new band. At 6.86 /a we find the mesitylene band, while at 9.75 fx we have the band found in benzene, in orthoxylene, and in numerous other compounds not related to the benzenes. At 13.4 /* we have a deep maximum like the one in orthoxylene at 13.6/*. As a whole this compound behaves like the xylenes, anisol, and mesitylene, which strengthens the argument for CHg-groups, especially for their bonding with the benzene nucleus. iCymene. C10H14. H Cymene is obtained from caraway, and can also ^-./'^^^ _ be prepared from turpentine oil and thymol. On II 1 (^jj comparing cymene with the terpene group of ^^\.^<;#=^~^^"^CH3 curves, we see that the general transmission is in H two steps remarkably like the terpenes, while there are quite a number of transmission minima in common. Cymene is the more transparent of the two, but, like pinene, has three CH3- groups. ^Abney & Festing, loc. cit. 82 INFRA-RED ABSORPTION SPECTRA. The first maximum at 3.43 fi is to be noticed in considering the possi- biUty of its being due to CH3 ; also the 6.86 fx, which appears to be har- monic with it. The 5.8 jit band found prominent in so many compounds is shifted to 6.0 )u, as in the terpenes. As a whole the benzene is changed so that there is only a slight indication of the benzene band at 3.25 ti and at 6.25 /*. In this connection it is well to notice toluene and mesitylene, which have fewer CHg-groups. The upper curve was obtained with a narrow slit, 0.4 mm., and a film, o.oi mm. in thickness. It shows that the general transmission is on a line of about 80 per cent. It also brings out the fact that the greater opacity beyond 6 ^ is due to the numerous absorption bands, which, by overlapping, lower the whole curve when a thicker film is used. Of still greater interest is the fact that the 6.86 fi and 7.3 [j. bands occur in the petroleum distillates, which are chain compounds. The latter predominate in CH^-groups and gen- erally have two CHg-groups, but whether we are to attribute them to the former is difficult to decide. The similarity in the curves of cymene, pinene, and limonene is a strong argument in showing that the cyclic structure is not the only cause of the characteristic spectra of the terpenes. Cyanine. C29H35N2I. (Fig. 89.) The chemical constitution of cyanines is unknown. Nietzki^ says of it that "possibly the cyanines possess a structure analogous to that of the phenylmethane dyestufifs," e. g., fuchsine, C20H19N3HCI+4H2O. The absorption spectra of these two compounds had previously been investigated by me^ to 10 /a and found unusually similar. Hartley and Dobbie' found that alkaloids having similar structure give similar ab- sorption spectra, showing that it is only in the details that the structure is different. The same is true in the present work on the petroleum distillates. The band at 3.43 fi occurs in the usual place for compounds having CHg-groups. Whether the benzene nuclei have any effect is difficult to decide, although the 6.7 ix and 8.7 /a bands are close to those of benzene. The broad regions of opacity at 7 ju, and 13 /* are very marked. BenzonitrilE. CsHsCN. (Fig. 90.) The result of the substitution of a €N-group for an H atom in ben- zene has produced an absorption spectrum of an unusual character. Its general transmission is about 80 per cent throughout the spectrum. There are no deep absorption bands, except one of 80 per cent at ^Nietzki : Chemie der Organischen FarbstoEfe, p. 262, 1901. *Phys. Rev., vol. xvi, p. 119, 1902. ^Hartley & Dobbie, loc. cit. INVESTIGATION WITH A ROCK-SALT PRISM. 83 13.25 jLt. This is the only compound studied that has such a spectrum composed chiefly of fine lines. The vibration of the benzene nucleus at 3.25, 6.25, and 6.75 fx has not been obliterated. The 4.45 fj. band is to be found at 4.38 /x in CHoCN and C0H5CN, while it is shifted to 4.8/1 in phenyl mustard oil, CgHsNCS, which should be noticed in connection with this compound. But few compounds have been found which have so many absorption bands at from 3 to 6/t, The large spectrometer was very useful in determining them with certainty. The contrast between the curve for this compound and the one for CgH-Cl is very marked. In the latter the substitution of a CI atom has not affected the 3.25 fi (6.25 ix) and 6.75 /x bands of benzene, but beyond this point the absorption bands at 9.3 /x. are much deeper, while at 13.25 fx the band is broad and shallow. On the other hand the substitution of the CN- group has caused a narrowing of all bands except the one at 13.25 fx, which is unusually strong. The spectrum of this compound shows that the numerous shallow bands in other compounds are not always due to the small dispersion used, but that the actual width of the lines must also be considered. DiPHENYL. (C6H5)2. Diphenyl, as the structural formula shows, consists of two benzene rings. It is made by passing benzene vapor through red-hot tubes, the heat causing it to polymerize into this larger molecule. The sub- stance itself is a solid, of large lamellar crystals, which melt at 71°. It was of great interest to study this compound in connection with benzene. The results were somewhat disappointing on account of the difficulty, always experienced with crystalline solids, in obtaining a film that was smooth and continuous and at the same time thin enough to be trans- parent. This one shows great opacity, but this is probably due to scat- tering of the transmitted energy. The film was continuous and 0.12 mm. in thickness. No marked bands are to be found in this curve except at 7.3 fx and 9.3 fx. Only a few — the least conspicuous — bands of benzene are common with diphenyl, e. g., 10.2 /t. This compound was studied in solution in CCl^ (Appendix IV), where it was found that the 3.25 ix benzene band occurs in its usual place and its usual intensity. It is quite transparent, showing that in the solid film the opacity is due to the scattering of the energy. Naphthalene and azobenzene were also studied. They have the 3.25 IX benzene band, showing that the vibration of the benzene nucleus still exists. 84 INFRA-RED ABSORPTION SPECTRA. AMIDO AND NITRO DERIVATIVES 01^ BENZENE, Aniune. CeHsNHj. (Fig. 91.) This compound differs from toluene, CeH^CHg, in that it has an NHj- group instead of a CHg-group attached to the benzene nucleus. The selection was made purposely to compare the effect of an NH2 to a CHg group when substituted for an H atom in the benzene ring. The result is a marked change in the region of 3 fi. The NH^-group has the greatest effect ; at least it is able to manifest the 2.98 ju. band found in NH3, as distinguished from the benzene band at 3.25/*, while the CH3- group has only caused a shifting in the benzene maximum to 3.32 fi, which is the mean of the 3.25 fi band of benzene and the 3,43 n band found in CHg compounds. However, this may depend upon the actual width of the lines. In the latter case the CHg-group would appear to have the greater influence. At 6,1/1, the large spectrometer failed to resolve the line, so that it is difficult to say whether the 6.25 /* band of benzene is actually shifted or whether it is double, like the one at 3,2 /x. This apparent shift is toward the shorter wave-lengths, just as observed by Kriiss^ for nitrogen compounds. However, in the present case the band of the original benzene nucleus still remains, with a new band toward the shorter wave-lengths. It will be interesting to learn whether Kruss actually found a shifting toward the shorter wave-lengths or whether, in his nitro-indigo solutions, the original indigo band disap- pears in the process of diluting the solution. As a whole, the introduc- tion of the NHg-group has seriously disturbed the spectrum of the ben- zene nucleus, except at 3.25 yu,. When compared with its isomer, picoline, we find but few bands in common, except at 2.g2ix. The observations were made of freshly distilled aniline. Methyl Aniune. CeHoNH (CH3) . (Figs. 32 and 93.) It The introduction of the methyl group in aniline vc-^^CH ^^^ decreased the intensity of the 2.97 /x band, while II I ^u the 3.25 IX band of benzene is almost obliterated by \_<^~ ^CH3 the one at 3.4/u.. The compound is far more trans- H parent, so that two strong bands at 13.3 fi and 14.5 /x could be located. The aniline film of same thickness becomes opaque at 12 ft. Certain bands of aniline are in common with this compound. The large spectrometer shows the 6,i ju, band of aniline shifted to 6,2 /t, while a new band occurs at 6,95 ju., *Kruss, loc. cit. INVESTIGATION WITH A ROCK-SALT PRISM. 85 Dimethyl Aniline. C6HbN(CH3)2. (Figs. 92 and 93.) H As in the preceding compound, the 2.98 /x and yp^-^Pjj 3.25 //, bands are almost obliterated by the CH3 band 11 I CH3 ^t 3.43 /i. A new band is found at 10.58 fi, while ^^^ just as in ammonia, while in picoline we have the ammonia band at 2.92 fi. and a second band at 3.35 /x. It is to be noticed that in the xylenes and in pyridine the benzene band at 3.25 /i has not been entirely obliterated, just as though there were two resonating ions, benzene and CHg, vibrating side by side. This is more evident in xylidine and the mustard oils. In xylidine, C6H3(CH3)2NH2, which has an NHo-group and two CHg-groups, we have the respective bands found in ammonia, at 2.95 fx, and in compounds predominating in CHg-groups, at 3.43 /m. The struc- tural formula of aniline indicates that in the original benzene ring an H atom has been replaced by an NHg-group, while in picoline we have the double benzene ring containing an N atom and a CHg-group. The absorption spectra support this theory, for in the aniline spectrum we have the original benzene band at 3.25 fi and the NHg band found in ammonia, xylidine, etc., while in picoline we have the benzene band oblit- erated and the CHg band substituted. The latter band occurs at 3.35 fi, the mean of 3.25 and 3.43 fi, instead of 3.43 fx. Can we say, then, that there is a real shifting of the 3.25 fx band in the xylenes? It must be remembered that we are integrating through a complex band, which, with ordinary dispersion, can not be resolved with a bolometer or a radiometer. Hence, when we find the maximum shifted to 3.3 fx in anisol, and to 3.4 fx in mesitylene, and find the separate bands in aniline, etc., it is a difficult matter to decide whether we have a true shifting, or whether we have simply determined the center of gravity of the several bands. An excellent example of this type is thymol, which melts at 44°. The solid film gave a deep band at 3.2 fx. In the melted condition the film was more homogeneous, and two bands were found, at 2.92 fx and 3.42 /x, respectively, instead of the mean at 3.2 fi. Other examples have been observed when the layer of liquid under examination was too thick. In view of the fact that we have such a striking similarity between the phenomena recorded here and those observed by Kriiss, it appears highly desirable to make a spectrophotometric study of dilute solutions, say of indigo and of methyl and nitro-indigo, in chloroform, to see whether there is but one band, or whether there are two, viz, the orig- inal one due to the indigo ion, which disappears on dilution,^ and a second, due to the methyl or the nitro group of ions, just as in the ^Since writing this, through the kindness of the Badische Anilin- & Soda-Fabrik, Ludwigshafen, Germany, who sent me samples of methyl, brom, and dibrom indigo, I have made a cursory examination of Kriiss 's work, and have found but one band in the visible spectrum, which shifts as was found by him. A saturated solution showed but one band, with the possibility of a second in the extreme violet. 104 INFRA-RED ABSORPTION SPECTRA. present work on aniline we have the original benzene band, at 3.25 /x, and a second at 2.97/1. As already mentioned, it has not yet been shown that the selective absorption of a solid in solution and the intra- molecular absorption of the solvent are closely related, but the question can be better settled by a photometric study as just indicated. (See, further. Appendix IV.) There are other bands farther out in the infra-red, which shift back and forth, just as noted above, but here the original benzene bands are more numerous. The most noticeable ones are those of the methyl sulphocyanate at 7.06)11 and 7.61 /i, which occur at 6.91 /i and 7.27/1 in ethyl sulphocyanate. However, in all the benzene derivatives studied the occurrence of an apparently new band in the derivative does not always seem to be a new band, but simply that the derivative has brought about a condition within the molecule such that the original resonating ion has a greater freedom. In the gases there is a more definite shifting of the absorption band lying between 3 and 3.5 /i, as shown in the following table. Gases. Maxima. Maxima. 3.08 3-28 3-39 3-42 3-45 3-45 331 1^ 7.38 6.98 6.85 6.85 6.88 7.00 7.70 KfVianp CloFTfi Rntanp (^jPTia Mpthvl pther (CH-', one would not expect a band at 2.95 /a for the fatty acids. In the other compounds having an OH-group, e. g., eugenol, caymol, menthol, and phenol, strong bands are to be found shifting from 2.87 fi to 3.0 fx. They show no bands at 6.0 fi. Can we assume, then, that the bands at 2.9 /i to 3 /it are due to OH? The evidence is not very favorable. Considering the bands of ammonia and of the compounds containing NH2, or certain ones con- taining nitrogen, the coincidence appears to be somewhat accidental. Farther in the infra-red we have numerous cases of the coincidence of absorption bands. As a whole, the most definite conclusion we can draw at present is that the alcohols have a characteristic band at about 2.p^ fi, just as the band at 4.78 ii is characteristic of the mustard oils. The CH3-group of atoms is probably the most important to be con- sidered, but only a few cases can be noticed here. The most noticeable eflFect is in benzene derivatives. It was shown under the discussion of the eflFect of structure that the benzene group (CgHe), although it appears as a series (CnHgn-e)* is entirely diflferent from the chain com- pounds, like CnH,n-2. But a substitution of several CHa-groups com- pletely absorbs the 3.25 /n benzene band, and the 3.43 /a band, charac- teristic of all compounds containing CH3, takes its place. Whether the 3.25 fi band has actually disappeared is an open question. In mesi- tylene there is still a trace of the 6.75 /a band of benzene, showing that the benzene " ion " has not been destroyed by the substitution of three CHg-groups. In the xylenes the 6.75 /* band is least aflfected, while the 3.25 IX suflfers the most, and the whole strengthens the belief mentioned in the beginning, that certain vibrating ions always seem to be present, but that their eflFect in absorbing heat waves seems to depend upon their surroundings. Thus the eflFect of substituting an NH.-group for an H atom, forming aniline, has the least eflFect on the benzene, 3.25 ju,, band, while those from 6 to 7 /a have disappeared entirely. In benzal- dehyde (CgHsCHO) the 3.25 ju. band is not seriously inflenced by a more intense absorption band at 3.55 {i, while in benzonitrile (CeHgCN) and in monobrombenzene (CgHgBr) the 6.25 fx band suflFers no change. As a whole, the substitution of a CH3 or NH2 group has a great eflFect on the resulting absorption spectrum. In the benzene derivatives these groups form new bands, which occur beside the benzene bands. no INFRA-RED ABSORPTION SPECTRA. showing that the vibration of the original benzene ion has not been destroyed. Such examples as these would indicate that the new bands are due to the groups of atoms substituted. But how are we to estab- lish this with certainty, especially when in the myricyl alcohol the evi- dence is contradictory for the direct effect on the OH-group? TOTAL ABSORPTION. This is not so well illustrated here as in the work of Friedel and of Zsigmondy (loc. cit.), who used the undispersed radiation. The pres- ent work agrees with theirs in showing that compounds having sulphur and the halogens are more transparent than H, O, OH, or N, which they have replaced. But not all the nitrogen compounds are highly opaque, e. g., nitromethane. The present investigation illustrates best the question of the location of the regions of greatest absorption. Thus in pyridine and picoline a 0.16 mm. layer is almost opaque beyond 6 /x, while methyl cyanide is quite transparent. Methyl iodide is quite trans- parent, since its large absorption bands lie in the region where the radi- ation from a black body is very weak, while the 3.4 /x band is shallow, so that the great transparency of this substance, observed by Friedel, as compared with nitrogen compounds, is apparently due to the lack of absorption bands in the region of intense radiation. As a whole, the work agrees with that of Friedel and of Zsigmondy in showing that the absorption of radiant heat depends upon the manner of the bonding of the atoms in the molecule, as well as upon the kind of compound in which the atoms of an element are united. The curves of pyridine and picoline illustrate the meaninglessness here of the application of the law of variation of absorption with thick- ness, which law Angstrom^ found not to hold true in certain cases, and concluded that it is due to the presence of unresolved absorption bands. The pyridine curves show this, especially the one for the thicker film, where the overlapping of the bands has lowered and blotted out the deep depressions in the curves. Friedel and Zsigmondy (loc. cit.) found that total absorption does not depend upon the size of the molecule. In the present instance we have noticed that the number and intensity of the absorption bands does not depend upon molecular weight, e. g., in the petroleum distillates. ^Angstrom : Ofversigt af Kongl. Vetenskaps, Akad. Forhandl., S. 331-352, 1890 ; S, 549, 1889. He computed the absorption of a liquid or vapor as a function of the thickness, from the equation (A=Io(i — e'*') and obtained smaller values than those observed. He concludes that the observed bands are complex groups of lines which are unresolved on account of the small dispersion. GENERAL DISCUSSION OF THE SPECTRA. Ill In the present work on dextro-pinene and laevo-pinene the maxima of the bands coincide, showing that the shape of the molecule has no influ- ence on absorption spectra. GROUPING OF THE SPECTRA. Abney and Festing (loc. cit.) say their results indicate "without much doubt that the substances we have examined can be grouped according to their absorption spectra, and that such a grouping, as far as we have examined it, agrees, on the whole, with that adopted by chemists." In other words, certain great groups of compounds have characteristic absorption spectra. This is just what Hartley and Dobbie (loc. cit.) observed for the alkaloids, in which the ultra-violet absorption spectra vary only in minor details. Aschkinass^ found that the minerals con- taining Ca, e. g., fluorite, calcspar, marble, and gypsum, have a band of metallic reflection (absorption) in the region of 30 /a. These observations apply to the present work. For example, the terpene group of compounds has a series of bands which is common to all of the compounds belonging to this group. In the same manner the general trend of their absorption curves is similar. The petroleum distillates have all the principal bands in common. The spectra of the fatty acids are conspicuous for the lack, but great depth, of their absorption bands. A glance at their line spectra in Table III gives ample proof of the foregoing statements.^ Having observed that certain groups of spectra are similar, one would naturally search for certain characteristic absorption bands in them ; and, reasoning from this stand- point, we may possibly be able to locate the group of atoms which causes the band. Thus the 2.95 fx band and the OH-group of atoms are characteristic of alcohols, while the 4.78 fi band and the NCS radical are characteristic of the mustard oils. But this avails us little, for a great many other facts, besides the group of atoms, serve as character- istics of these groups of compounds. CHARACTERISTICS OF THE SPECTRA OF CARBOHYDRATES. The absorption spectra of carbohydrates are conspicuous for the recurrence of absorption bands in certain regions of the spectrum. The first of these regions was found by Abney and Festing (loc. cit.) at 0.74 fj. and 0.867 fi. Puccianti (loc. cit.) found that in all cases where the carbon atom is joined directh' to the h}drogen atom in the molecule the absorption spectrum shows a band at 1.71 ^u. This has been verified ^Aschkinass : Ann. der Physik (4), i, p. 42, 1900. ^The tables will be found at the end of this volume. 112 INFRA-RED ABSORPTION SPECTRA. in the present work, in which the band was found to oscillate between the values of 1.68 /u, for benzene to 1.74 /t for caproic acid, for ethyl suc- cinate, and for methyl acetate. From this point to 3 /x there are numer- ous small bands of minor importance. Somewhere between 3.1 to 3.43 fi a band is found for every carbohydrate studied, oscillating from 3.25 fi in benzene to 3,43 fi in the alcohols and compounds rich in CH3- groups. In the region extending from 4 to 5 /x there is great trans- parency and often no strong bands (except in the case of the mustard oils), and for the more complex compounds, e. g., the petroleum distil- lates, there are generally no lines that could be detected even with the large dispersion used. Beyond 5.5 fi the transmission curve decreases, often very abruptly, terminating in strong absorption bands varying from 6.75 yn in benzene, 6.86 ju, in aliphatic compounds, and 7 /* in ter- penes. Beyond this region the bands become stronger, better defined, while the transmission curve varies from great transparency to com- plete opacity. The region at 12 /<, is often lacking in absorption bands, and finally we come to a region of frequently great absorption, with bands occurring from 13.6 fi to 14.2 /a. Beyond this it is difficult to pen- etrate, but all observations made indicate conditions similar to those existing in the region investigated. In addition to the general characteristics of the spectra of carbo- hydrates, it will be noticed that the characteristic bands of benzene derivatives are at 3.25 /x, 6.25 ix, and 6.75 fi ; that of the aliphatic com- pounds, e. g., the petroleum distillates, at 3.43 /x, 6.86 /x, and 13.6 /x to 13.8 /x; that of carbon tetrachloride at 13 /x; that of compounds having N or NH2 at 2.95 fi and at 6.1 /x to 6.2 /x ; that of the fatty acids at 3.45 /x and 5.86 /ix; that of the alcohols at 2.95 /x and 3.43 /«,; and that of the mustard oils at 4.78 /x. The region of great transparency from 4 to 5 /x is also to be noticed, since the larger dispersion failed to show the pres- ence of strong lines. OCCURRENCE OF HARMONICS. In discussing the question of the presence of simple relations among the spectral lines in the optical region Cornu^ shows that it is useless to search for harmonic overtones, since the case is relatively rare. He adds that the law of vibration in whole numbers is applicable only to a particular form of sounding bodies, of which the type is a cylindrical column, whose length is great in comparison to the cross-section. In any other type except this special one the relations between the vibra- tion numbers of the successive tones is very complex. In Kayser's 'Cornu : Compt. Rend., 100, p. 1181, 1885. GENERAL DISCUSSION OF THE SPECTRA. II3 Spectroscopy it is remarked that such a search is delusive, and GriJn- wald's mathematical spectrum analysis^ is cited as an example in which all wave-lengths of the so-called compound spectrum of hydrogen can be converted into corresponding wave-lengths of the water-vapor spec- trum by multiplying by 0.5. Recent and more exact measurements on these lines show that no such relations exist. Schuster- speaks of the iron spectrum, which has two lines which are in the ratio of 2 to 3, while hydrogen has lines in the ratios of 20 : 27 : 37. He demonstrates that in accordance with the Theory of Probability a certain number of coincidences between lines of two spectra might be expected to occur, even if the spectra be quite unrelated. Furthermore, there appears to be a tendency for functions formed by two lines to cluster around harmonic ratios, and, " most probably, some law hitherto undiscovered exists which in special cases resolves itself into the law of harmonic ratios." Of course, as is well known now, the nearest approximation to such a law is Balmer's Law and the numerous other convergent series formulae used by Kayser and Runge and others. Nevertheless, in spite of these warnings, and fully realizing the danger from lack of dispersion, experimental errors, etc., I venture to call attention to certain marked absorption bands which occur so frequently in positions which so closely fulfill this relation that it is necessary to examine more fully into the probable significance. Abney and Festing (loc. cit.) found that compounds having CH3- groups have a band at 0.74 ix and another between 0.907 and 0.942 /*, while benzene and CH compounds have a band at 0.867 /*• Puccianti (loc. cit.) found a band at 1.71 ix for all compounds in which the C atom is joined directly to the H atom in the molecule. Aschkinass (loc. cit.) found the absorption bands of water at 1.5 1 11, 3.06 /i, and 6.1 /£, while Paschen (loc. cit.) found them at from 2.916 to 3.024 /x and at 6.06 fi, which values are closely harmonic, Ransohoff found closely harmonic bands for alcohols at 1.71 and 3.43 /a. In the present work, using a quartz prism, the first band occurs at from 0.83 to 0.86 fx, while the second one oscillates between the values 1.66 /i, for the thiophene and 1.73 /u. for ethyl succinate. The next dis- turbance is in the region of 3.4 ^, the maximum being at 3.25 /* for ben- zene and 3.43 fi for compounds rich in CH2 or CHg groups. The next region where there is a constant recurrence of bands is at 6.75 /A for benzene and 6.86 n for other compounds rich in CH2 or CHg 'Griinwald : Wien. Ber., 96 to loi, 1887 to 1892. 'Schuster : Proc. Roy. Soc, 31, p. 337, 1881. 114 INFRA-RED ABSORPTION SPECTRA. groups. In discussing the sources of errors it was shown that the 6.86 /t band is quite accurately known. Beyond 13.6 /x there is a band of frequent recurrence. In this region, however, it is difficult to locate the bands with great accuracy, because of the weakness of the radiation. As a whole, however, the bands at 1.71, 3.43, 6.86, and 13.6 to 13.8^ are closely harmonic, e. g., in the petroleum distillates, and, taken with the 0.867 ii band of Abney and Festing (0.83 to 0.86 |it in present work), would seem to indicate a vibration about a fixed point. Even if in the future this relation should ultimately be found false, the constant recurrence of these bands in so many compounds can not be without meaning. These bands are so sharp and symmetrical that it is difficult to conceive how, with greater dispersion, they can be resolved into lines which are very unsymmetric- ally placed about the present centers of gravity. In carbon tetrachloride and tetrachlor-ethylene it was shown that this group of compounds is conspicuous for its absence of absorption bands except at 6.5 and 13 ix (harmonics), where there are large bands, each of which is evidently complex. The complete table of wave-lengths of absorption bands contains still further illustrations. An explanation of the significance of these relations is not attempted, and it will be suf- ficient to add that any such harmonic relation would seem to indicate the resonance of a definite group of atoms, or " ions," to which these lines are solely due. But to attribute a given line to a certain group of chemical atoms is dangerous, for it has already been shown in the case of the mustard oils that the manner of grouping of the atoms is not the only characteristic of this group of compounds. Thus the physical properties of benzene and thiophene are so similar that these two com- pounds are readily confounded,^ yet their absorption spectra are entirely different. It might be added that in the gases, where one would natur- ally expect such harmonic relations, only acetylene at 3.7 and 7.4 /x, and methyl ether at 3.45 and 6.9 /x have bands satisfying this condition. Whether this relation will ultimately be proved absolutely true remains to be seen. To determine this question a very much larger dispersion will have to be employed than has yet been available. This means a far more sensitive recording apparatus than has yet been devised. In dismissing this question it will be sufficient to add that after a year's struggle with it, to prove or disprove it, the result has been a closer agreement in the values first obtained, especially for the bands at 3.43 /x and 6.86 fx. ^Smith's (Richter) Organic Chem., vol. 2, p. 47. GENERAL DISCUSSION OF THE SPECTRA. II5 SUMMARY. The infra-red absorption spectra of organic compounds have been studied, the majority to 15 /u, using a radiometer, two mirror spectro- meters, and a rock-salt prism. Out of a total number of at least 135 compounds examined with the rock-salt prism, 131 have been recorded in this paper. They include solids, liquids, and gases. In addition to this, 19 compounds were examined to 2.^ /x, using a quartz prism. The following are some of the results obtained : ( 1 ) A study of isomeric compounds shows that the arrangement, or bonding, of the atoms in the molecule, i. e., its structure, has a great influence upon the resulting absorption spectrum, which agrees with Julius (loc. cit.). This is of considerable significance, and is in marked contrast with stereomeric compounds, like dextro- and laevo-pinene, which were found to have identical spectra, showing that the spacial arrangement of the atoms, i. e., the configuration of the molecule, had no effect upon the resulting absorption spectrum. (2) No shifting of the maxima of absorption, with increase in molec- ular weight, " Kundt's Law," could be detected, except in the case of the band lying between 3.1 and 3.5 /u,, for gases. Instead of a shifting of the maximum in certain compounds, there occurs a new band beside the original one when a methyl or amido group is substituted for a hydrogen atom, the new band lying toward the longer wave-lengths when a methyl group is substituted, and toward the shorter wave- lengths when the hydrogen atom is replaced by an amido group. This disagrees with investigations of Kriiss in the optical region, where only the new hand was observed. (3) A rise in temperature of 20° had no eflfect upon the transparency of the compound, nor upon the position of its maxima. (4) The effect of replacing an H atom by certain groups of atoms, like NH, and CH3, is very marked, and usually shows new bands, e. g., 2.g6 and 3.43 /x, in the resulting absorption spectrum. In the spectra of certain benzene derivatives, however, the bands of the benzene spec- trum are usually present, showing that the vibration of the benzene nucleus has not been destroyed. However, the writer does not con- sider this sufficient evidence to consider the new bands to be due to the groups of chemical atoms substituted. (5) Total absorption is not influenced by the size of the molecule, while compounds having sulphur or halogens are more transparent than those having H, O, OH, or N, which they have replaced, just as found by Friedel and by Zsigmondy. Il6 INFRA-RED ABSORPTION SPECTRA. (6) The spectra of groups of compounds are similar, and are char- acteristic of the grouping adopted by chemists, as found by Abney and Festing. (7) Carbohydrates have a characteristic spectrum, with absorption bands at 0.83 to 0.86 /*, 1.67 to 1.72 n, 3.25 to 3.43 fi, 6.75 to 6.86^, and 13.6 to 14 /i. The first large absorption band in carbohydrates occurs in the region of 3.2 /a, which is in general followed by a transparent region from 4 to 5 /a. The work of Puccianti, in which he found that all carbohydrates have an absorption band at 1.71 /*, while benzene derivatives have two additional bands, at 2.18 and 2,49 /a, respectively, has been confirmed on 18 new compounds. The 1.7 /* band deserves especial notice. (8) In addition to the characteristic carbohydrate spectrum, certain bands in it occur in positions which are close harmonics, the maximum wave-length of each succeeding band being twice the preceding. The question whether this is merely a coincidence or whether it is an exact relation is not fully determined. To decide this question a larger dis- persion will be necessary, while the spectrum will have to be explored to 27.6 /x for the next harmonic. In the same manner more pairs of bands will have to be located in the ammonia spectrum, etc., in order to show that the constant difference of the wave numbers found is not merely a coincidence, hence, that there is a true spectral series present. (9) The three isomeric xylenes have banded, " channeled," spectra, in which the most important line in each group lies farthest toward the long wave-lengths, in the order ortho, meta, para. In other words, the ortho, in which the CHg-groups are the closest together in the ben- zene ring, has the " head " of each group of bands lying farther toward the infra-red than are the heads of the corresponding bands of the meta and para compounds. This seems to indicate a resonance of electric- ally charged particles (CHg), whose capacity is increased, whose period becomes slower, and, hence, whose maxima are shifted toward the longer wave-lengths, with a decrease in the distance between the particles. ( 10) In many compounds numerous bands are in coincidence, which would no doubt be found in different positions when using a larger dis- persion. Other bands, like the one at 3.25 fi in benzene, in benzalde- hyde, and in pyridine, or the 3.43 jx and 6.86 /x found in aliphatic com- pounds, seems to point to a specific group of atoms as their source, or to some " ion " or " nucleus " common to them. The most marked example of this type is phenyl mustard oil, in which the vibration characteristic of the mustard oils at 4.78 fi is superposed upon the vibration of the benzene " nucleus " or " ion," which has its maxima at 3.25 /a, 6.75 /a, etc. In some compounds there is evidence GENERAL DISCUSSION OF THE SPECTRA. 117 that certain bands, e. g., the 3.43 /x band, are due to a definite group of atoms, e. g., the CHj-groups in the chain compounds and terpenes ; in other compounds the evidence is just as strongly in favor of the manner of bonding of the atoms, e. g., the methylene hydrocarbons of the petro- leum distillates ; still other compounds, e. g., benzene and its derivatives, especially phenyl mustard oil, in which we have the characteristic vibra- tion of the mustard oils superposed upon the vibration of the benzene nucleus, show that both the groups of atoms and their manner of bond- ing with other atoms, as well as the kind of atom, have a great influence upon the absorption curve. In the present work the spectra have been discussed from the stand- point that since compounds of the same chemical composition (isomers) have different spectra the source of the disturbance is intramolecular. If we had assumed ignorance of the composition of the compounds, we would have expected, from our knowledge of the spectra of the ele- ments, that each compound ought to have a different spectrum. This has been found to hold true, except for certain lines in them, as shown in the following table : Group. Characteristic maxima. CHaorCHs NH2 3.43," 6.86/« 13.6 to 13.8 and 14/^ 2.96 6.1 to 6.15 3.25 6.75 8.68 9.8 II. 8 12.95 7.47? 9.08 2.95 4.78 CeHe NO2 OH NCS There would be no reason for deciding whether the cause is inter or intra molecular. The compounds might then be grouped according to the marked absorption bands which they have in common, e. g., those having a band in common at 3.25 ft, at 3.43 fi, or at 4.78 fi. One compound (phenyl mustard oil) would then be placed in the 3.25 fj. and in the 4.78 /u, group. Another (xylidine) would belong to the 2.95^1 group and to the 3.43 /a group. This would then suggest a disturbance common to both groups, and we are brought to the point arrived at by the other line of argument, viz, there is a something, call it "particle," "group of atoms," "ion," or "nucleus," in common with many of the compounds studied, which causes absorption bands that are character- istic of the great groups of organic compounds, but we do not know what that " something " is. The presence of these groups of ions, each group having its own free period of vibration, is in accord with present conceptions of absorption ;/^ .©" >^':.. Il8 INFRA-RED ABSORPTION SPECTRA. and anomalous dispersion. The most conspicuous regions where these ions manifest themselves is at 3.43 and 6.86 /x for CHg compounds, at 3.25, 6.25 and 6.75 fi. for benzenes, and at 4.78 fi for mustard oils. The latter is superposed upon the CHg or the benzene vibration depending upon the compound. Alcohols and compounds having NH2 have a band at 2.95 fi. To assign the cause of these bands to a particular group of chemical atoms rather than to a less definite, ultra-atomic source does not eluci- date matters very much, although it is true that these bands do not appear until these groups of atoms are introduced into the compound. APPENDIX I. SOURCES OF RADIATION. In previous investigations the radiator used was a zircon lamp, or a platinum strip covered with iron oxide and heated to redness by means of an electric current. In the foregoing work a light portable radiator was desired which could be mounted upon the spectrometer arm and be moved with it. For this purpose the Nernst lamp was found to be the most serviceable because it is light, compact, and has no products of combustion, such as CO2 and water vapor, which would contaminate the room (which was small), thus endangering the prism. The " glower " causes some trouble, and since the distribution of the energy in the spectrum of the " heater " is more uniform, the latter was used. This is a very satisfactory radiator, since it is not affected by air cur- rents,, while it can be maintained at a uniform temperature by using current from a battery of storage cells. As is well known, the "heater" consists of a hollow cylinder of clay wound with a fine platinum wire, over which is a thin coating of a more refractory clay. When the plati- num wire is heated to incandescence the clay gives out the desired radi- ation. By using such a heater requiring no volts on a 90- volt circuit it lasted for at least three months, when the platinum wire had volatil- ized sufficiently to cause it to break. This was easily repaired and cov- ered with clay. The volatilized platinum condenses upon the surface of the heater in small, flat, triangular and hexagonal crystals. Joly^ found that on covering a strip of platinum with topaz dust and heating it to redness microscopic crystals of platinum were formed on the par- tially decomposed topaz, the prevailing form being the octahedron. The extraordinary distribution of energy from these heaters is shown in fig. 125. The minima lie close to the well-known absorption bands of COo and water vapor. The maxima are not so easily explained. The first one, at 2.5 ju, is no doubt due to the hot platinum wire. The third maximum lies close to the absorption band of quartz found^ at 5.3 jx. The suppression of this apparently selective radiation at 5.2 /* is well illustrated in curve h, which was obtained from a " heater " cov- ^Joly : Nature, 43, p. 541, 1891. ^E. F. Nichols: Physical Review, 4, 1897. 119 I20 INFRA-RED ABSORPTION SPECTRA, ered with a film of F2O3. The film was obtained by dipping the heater in a FeS04 solution, which was then oxidized by heating.^ For the curves a and h the spectrometer slit was the same width. Curve c rep- resents the distribution of curve a for a wider slit. It was obtained while finding the absorption of a certain compound, and is of interest since, by its regularity, it shows the constancy of the radiation. The radiation from this heater was found when covered with borax, also when a strip of mica 0.3 mm in thickness was wound on it, but in neither case was the energy curve different from the original. The mica did not show emission minima at 8.4 /a and 9.4 /u. as computed by RosenthaP (fig. 126, curve h). That an emission band should exist at 5.2 /u, seemed doubtful. To test this some finely ground quartz (French flint), pure feldspar, and a number of kaolin clays were obtained from Prof. H. Ries, of the department of geology. The original clay was removed from the plati- num wire, and then the heater was covered with a thin paste of one of these clays. After drying, the heater was used as before. In the course of the investigation of absorption spectra as many curves (the "direct deflections ") could be obtained as was desired. The quartz powder would not adhere well to the heater, even when mixed with starch and applied as a paste. The result was that the surface never became so hot as it did when the finer-grained clays were employed. This, of course, decreased the intensity of the radiation, as will be noticed in curve d of fig. 125, where it will also be noticed that the sharp maximum has disappeared, showing that it is not a selective emission band of quartz, but that it is simply due to the adjacent absorption bands in the atmosphere. There are really two emission maxima, viz, the partially suppressed one at 2.5 ^i, due to the white-hot platinum wire^ and a second maximum in the region of 3.8 ;u, which may be due to the cherry-red (color) clay covering. The latter maximum is partly reversed by the strong atmospheric absorption band of COo at 4.25 /u,. The platinum con- ductor under the clay is very much hotter than it would be if exposed to the air. It does not seem possible, however, that it is sufficiently volatilized to give emission bands. It would seem as though the clay would be a sufficient covering to suppress them. In fig. 126 is shown the radiation curve c of a different heater, which shows the variation in the distribution of the emission of diflferent heaters, this one being made of a finer-grained clay than the preceding. This heater contained ^Paschen : Ann. der Physik, 56, p. 762, 1895. ^Rosenthal : Ann. der Physik, 68, p. 792, 1899. APPENDIX I. 121 less than 3 per cent of SiOa. The examination of the pure feldspar radiator was not successful, the platinum breaking soon after beginning the observations, and the examination was never repeated. Whether the platinum was acted upon by the feldspar is not known. The heater, when covered with ordinary blue clay, gave a curve quite similar to curve a, the only difference being slight depressions at 6 /i. Since some of these curves were obtained at different periods, several months inter- vening, the relative intensity of the radiation from the different heaters can not be compared, except curves a and b in the two figures. They are given for the purpose of showing the relative distribution of the energy in the different spectra. Curve d in fig. 126 shows the radiation from a new heater, using the large spectrometer, which had mirrors of i meter focal strength ; con- sequently the radiation from the heater had to traverse almost 3 meters of air before entering the radiometer. The result is a marked trans- formation of the whole curve. The water- vapor bands at 2.9, 4.8, 5.1, 5.8 to 6.0, and at 6.6 /a and the CO2 bands at 2.7 /x and 4.28 /* have been determined by Paschen (loc. cit.). The source of the 3.45 /u, band is not known, unless it be due to the presence of hydrocarbon vapors, which manifest themselves, just as Pauer (loc. cit.) found for the ultra- violet. As far as known to the writer this is the next to the largest spec- trometer ever constructed for such work (Langley's being the largest) , and it illustrates well the difficulties to be encountered in increasing the dispersion by lengthening the spectrometer arms. For example, the intensity of the radiation drops from 21 to 8 in the CO2 band at 4.25 fi, which shows that a very much more intense source of radiation will be necessary if a larger spectrometer is to be used. What is needed, then, is a device giving larger dispersion, with short spectrometer arms. For example, a fluorite prism on the small spectrometer would have given about the same dispersion as this large spectrometer and rock-salt prism. APPENDIX 11. THE EMISSION SPECTRUM OF THE HEFNER LAMP. Having had occasion to compare the radiation from a Hefner lamp (amyl-acetate flame) with an acetylene flame of the Bunsen and the cylindrical types, the results obtained seem of sufficient interest to record in this paper. The large spectrometer was used, hence the atmospheric absorption bands at 1.4, 1.8, and 2.6/1 are prominent. The products of combustion are solid carbon, water vapor, and COg. In fig. 127 curve c shows the radiation from the Bunsen acetylene flame, while curve a shows the radiation from the cylindrical acetylene flame. Curve b gives the radiation from the Hefner lamp. The Hefner lamp is weak in light-giving power as compared with the acetylene flame. The curves at 2 fi show that this is due to the diflPer- ence in temperature of the incandescent carbon particles. It would be interesting to learn whether this is due to the greater quantity of water vapor present in the amyl-acetate flame or to the greater completeness of the combustion ; it has twice as much water vapor as has the acety- lene flame for the same amount of carbon dioxide. The emission bands at 4.4 fi are of about equal intensity. The curves when corrected for slit width are shown in fig. 128. The maximum for the acetylene flame comes at i .05 fi, as found by Stewart,^ while the amyl-acetate maximum occurs at 1.50 /*. Now, the question of the precise relation between the temperature of an ideal " black " body, computed by means of the constant A (AmT = A), and the actual temperature of a flame is not settled. Stewart (loc. cit.) adopted the value A = 2282, which gave values for the temperature of the acetylene flame, the ordinary gas flame, and a candle flame, computed from the maxima of their respective energy curves, which agree with the temperatures obtained by Nichols^ by direct measurement, viz, 1900° C, 1780° C, and 1670° C, respectively. Using this value of the constant A, the computed temperature for the amyl-acetate flame is 1250° C. Using Paschen's constant, 2940, gives a value of 1690° C. Angstrom^ finds this temperature to be 1557° C, while Wanner found it to be 1162°. Paschen* found the shifting of the emission band of CO2 with rise of temperature, using a stream of ^Stewart, G. W. : Phys. Rev., xv, p. 311, 1902. *E. L. Nichols: Phys. Rev., x, p. 248, 1900. Angstrom: Phys. Rev., xvii, p. 302, 1903. ^Paschen : Ann. der Physik, 50, p. 409, 1893. 122 APPENDIX II. 123 CO2 heated to different stages from 300° to 1460°, which is the tem- perature ( ?) of the Bunsen flame. The latter has its maximum at 4.40 /I. It will be noticed in the curves that, for the amyl-acetate flame, the CO2 maximum is situated at 4.36 fi, while for the Bunsen acetylene and the Bunsen gas flame the maximum is sharp at 4.40 fx. This would appear to indicate a lower temperature than 1460°. Plotting the values of temperatures and wave-lengths, found by Paschen, it was found that for the amyl-acetate lamp the maximum at 4.36 /i indicates a tempera- ture of 1370° C. This discrepancy is no doubt due to the use of the large spectrometer, which magnifies the atmospheric absorption band at 1.37 /t (Langley's il/=i.4fi), and consequently lowers the emission curve of the amyl- acetate lamp, which happens to come just at this point. However, this still leaves a difference of 200° unaccounted for, which emphasizes the lack of our knowledge of the term " temperature " as applied to such a radiator. APPENDIX III. ELECTRIFICATION OF RADIOMETER VANES. One of the few defects of a radiometer is the electrification of its vanes, which is caused by air currents in exhausting the apparatus or by standing with the stop-cocks open to the exhaust pump. Since this effect has not been recorded except by Stewart (loc, cit.) and myself, and since no efficient remedies to prevent it have been given heretofore, it seems desirable to indicate the progress made in this direction. The electrification of the vanes by air currents occurs in exhausting the radiometer, especially when it is first assembled and the vanes are new. This does not occur until a critical pressure of about o.i mm. is attained, when the deflection is generally thrown entirely off the scale, and the vanes may not return for many hours. In the present work, for quite a while the joints of the apparatus leaked a slight amount, so that the stop-cocks were left open to the exhaust pump, in order that the change of pressure would not affect the sensibility of the radiometer for a day's work. It was then found that the vanes became electrified, without any apparent provocation, while making readings, and that it was more aggravated in the hotter summer months. As far as could be ascertained, the vane that was nearest to the rock-salt window would be attracted to it. The mercury vapor was at once sus- pected as the source of electrification. The remedy consisted in insert- ing a glass tube containing gold foil spread on absorbent cotton, to present a large surface. After that there was no further electrification either from air curents or the mercury vapor, until the gold had become amalgamated. As an example, the apparatus stood from October 20, 1902, to February 20, 1903 (four months) , without becoming electrified. The pump was then cleaned and a little more gold foil added, when the apparatus stood from March 10 to June 12, 1903, before it became elec- trified. The gold was then found to be entirely amalgamated. Of course, part of this could have been obviated by closing the stop-cocks, but this in time starts them to leaking, hence the present method was the more preferable. This was especially desirable after the joints had become so tight that there was practically no leaking when the radio- meter stood from June to September, 1903 (stop-cock closed) and main- tained a sufficient sensitiveness that a candle several meters from the slit threw the deflection clear off the scale. It is a peculiar fact that with gold foil in series old vanes do not become electrified in exhausting the apparatus. Just the opposite is true for newly blackened values, which on exhausting adhere to the 124 APPENDIX III. 125 window with great tenacity. After standing several days the vanes no longer become electrified so easily. Since the completion of this research the radiometer has been remod- eled, in order to gain a greater sensitiveness, and at the same time to shorten the period. It became evident that on account of the viscosity of the residual gas in the radiometer the period is more affected by the size of the vanes than by the moment of inertia of the system or the diameter of the fiber suspension. In order to decrease the area of the vanes they must be situated near the slit of the spectrometer. To this end the outer window, R^, was discarded, and only the inner one, R^, was used (fig. 3). It rested on a wide flange, and was made airtight by means of beeswax covered with shellac varnish. The slit F,, of the spectrometer (fig. i) was mounted directly upon the outside of this window, while the vanes were about 2 mm. from the inside surface. To prevent the spectrometer from slipping, with respect to the radio- meter (and the slit F), it was held in place by means of heavy weights. The vanes v/ere of very thin strips of mica, 12 mm. long and i mm. wide, secured to fine glass rods. The mirror was about the size of the preceding. Unfortunately, the weight the suspension was not deter- mined, but since the area of the vanes, instead of their mass, concerns us most, that is of minor importance. The half period, i. e., time for maximum of deflection, of this system was only 6 seconds, and at 0.05 mm. pressure its damping was so small that it was used ballistically. The sensitiveness was about 4, i. e., a paraffin candle at a distance of i meter gave a deflection of 4 cm. per square millimeter of exposed surface (slit o.i mm.) upon a scale situ- ated at a distance of i meter. This was a somewhat greater sensitive- ness than the preceding, and was equal to that used by Stewart (loc. cit.)j which had a period of 40 seconds. A still greater sensitiveness was attained by selecting a much finer fiber which was aperiodic in air. In comparison with the bolometer this does not seem a fair test of sensi- tiveness, for the large vanes of the Stewart radiometer, with a new fiber (period 20 seconds), had a sensitiveness of only i instead of 4. Nevertheless, the dispersed radiation from the hand was sufficient to give a deflection of from i to 2 mm. at 9 11. For work where a very narrow linear absorbing surface is required it seems quite probable that the sensitiveness of the radiometer can not attain that of the bolometer. Nevertheless for ordinary laboratory work it seems superior to it on account of its simplicity and its freedom from magnetic and thermal disturbances. The radiometer, as ordi- narily used, is from 4 to 8 times as sensitive as the Boys radiomicrom- eter, which was thought capable of detecting a rise in temperature of 0.000001° C. APPENDIX IV. ABSORPTION OF SOLIDS IN SOLUTION. In the ultra-violet Hartley and Dobbie (loc. cit.) have studied the absorption spectra of solids dissolved in water. They found that groups of chemically related compounds have similar absorption spectra. In the visible spectrum we also find absorption bands, especially when the solute is a colored substance. Iodine is an example of this type. But in the infra-red the writer found iodine transparent beyond i.i /i. As far as is known to the writer, only one other substance, sulphur in CS2, has been examined far out in the infra-red. This was done by Julius, who found that S had no appreciable eflPect upon the transparency of the CS2. That a solid in solution should be transparent to infra-red radiation seemed doubtful. It would indicate a resonance of small particles in the optical region, as distinguished from the intramolecular resonance of the solvent. After studying so many compounds it seemed imperative to consider this question more thoroughly. It was suspected that this transparency is simply due to the thin cell used and to the slight solubility of the solids. For example, assuming that 0.05 gram of iodine per cubic centimeter is dissolved in CS2, using a cell 0.3 mm. thick, this would be sufficient to form a solid film only 0.03 mm. It seemed that the proper method of answering this question would be to select a solid having a strong absorption band in a region of the spectrum where the solvent has no absorption bands. DiPHENYL. C12H10. The curves of carbon tetrachloride, CCI4 showed no marked bands up to 6.5 fi. Accordingly this liquid was selected for a solvent. The sample used with diphenyl showed the water band at 2.97 /x and a second band at 4.5 fi. The saturated diphenyl solution showed an additional large band at 3.25 fi, which is the characteristic band of the benzene nucleus. This is of considerable interest, since diphenyl, CgHg — CgHg, is a double benzene ring. The curve, fig. 129, shows that solids in solu- tion do absorb heat waves, and that the selective absorption of a solid in solution and that of the solvent are identical.-' It also shows that the *In my preliminary communication on Infra-red Absorption Spectra, Astrophys. Jour., XX, p. 215, 1904, it appeared to me that there might be a difference. 126 APPENDIX IV. 127 benzene vibration still exists, and that solids in solution can be studied in the infra-red, just as in the ultra-violet. Naphthalene. CioHj. Naphthalene was also examined at 3.25 fi. Its solubility in CCI4 is unusually great, e. g., the present solution contained 0.25 gram per cubic centimeter of CCI4. This is sufficient to make a homogeneous solid film 0.28 mm. thick (sp. gr. = 1.15 ; cell = 0.6 mm.). This compound is of considerable importance, for its molecule is formed by the condensation of two benzene nuclei. As a result, its chemical properties are quite different from that of benzene ; but its absorption spectrum shows the 3.25 fi band of benzene, with which it is also comparable in general transmission. As a whole, for the region investigated, this compound shows that the vibration of the benzene nucleus has not been disturbed. AZOBENZENE. CeHs — N=N — CeHs. This compound is also quite soluble in CCI4, so that a saturated solu- tion contained about 0.2 gram per cubic centimeter of the liquid. As with the preceding compounds, the rock-salt cell was 0.6 mm. thick. The curve shows the 3.25 fi benzene band, indicating that the presence of the N atoms does not disturb the vibration of the benzene nucleus. This solution is far more opaque than the preceding, showing the effect of the introduction of the N atoms, just as in other compounds contain- ing nitrogen. Moreover, this solution is of a reddish-brown color, showing an absorption band in the visible spectrum, as well as in the infra-red. This shows that both are due to an intramolecular disturbance, and dis- approves the idea of a resonance of small particles (of molecular dimen- sions, as for, example, Kosonogoff's butterfly scales and Wood's metal films) in the optical region, as distinguished from the intramolecular resonance in the infra-red. Of course, one might say that the infra- red band, e. g., the 3.25 ju, band, is also a resonance effect. Possibly it is, but the evidence of a resonance of small, electrically charged parti- cles whose capacity, and hence whose periods (like a condenser) , depend upon their proximity to similarly charged particles (the closer the par- ticles the greater the capacity, hence the slower the period, and hence the farther is the absorption band shifted toward the longer wave- lengths), is somewhat contradictory in the infra-red. The best evidence in favor of this idea of a resonance effect is in the xylenes, in which for the ortho, which has the two CHg-groups nearest together, the large absorption bands lie farthest toward the long wave- lengths, while in the para, in which the two CHg-groups are farthest 128 INFRA-RED ABSORPTION SPECTRA. apart, the large absorption bands lie farthest toward the short wave- lengths ; the meta has its bands in intermediate positions. The dielectric constants of a number of solvents and solutions have been determined by Schlundt^ and by Eggers.^ They found that the dielectric constants of C2N2, CH2(CN)2, and C2H4(CN)2 to be 2.52, 46.3, and 61.2, respectively, which would indicate that the dielectric constant is dependent upon the comparative freedom of the cyanogen radicals from each other. In general, however, they find that the deleCtric of a substance is affected not only by the elements entering into its decomposition, but also by the grouping of those elements in the molecule. The dielectric constant decreases with increase in molecular weight. Methyl, ethyl, amyl, and phenyl mustard oil have the respect- ive dielectric constants, 17.9, 22, 17.3, and 8.5, so that evidently we gain nothing in considering them in connection with the large band at 4.78 /*. The value of the phenyl mustard oil is just about one-half that of the others. These compounds were not examined for the 6.25 and 6.75 fx benzene bands, because CCI4 has a large band at 6.5 /u., and the cell was opaque. It is to be noticed that these solids in solution are practically trans- parent, except in the region of their absorption bands, which corresponds to the results found for iodine in solution, as already mentioned. In the present state of our knowledge of the subject it appears that a solid in solution is far more transparent than in its undissolved state in the infra-red, which seems a rather unusual condition that deserves further inquiry. 'Schlundt : Bull. Univ. of Wisconsin, 2, 353, 1901. '^Eggers : Jour. Phys. Chem., viii, p. 14, 1904. APPENDIX V. WATER OP CRYSTALUZATION. Brucite (Magnesium Hydrate). Mg(0H)2 or MgO'HzO. (Fig. 130.) This mineral was studied with reference to the question of the effect of the OH-group. It is also of interest because it does not contain carbon. The common massive foliated variety was used, from which a thin, clear lamina was removed. The transmission curve is conspicuous for a very large minimum at 2.5 [I, beyond which there are no bands until we arrive at 9 /t, beyond which point there is complete opacity. The characteristic band of carbohydrates, at 3.1 to 3.43 [x, is absent; in fact, there are no bands in the whole spectrum which are to be found in the hydrocarbons. This is the first compound studied which has a large absorption band situated near the visible spectrum.^ In fact, it is but the second com- pound discovered which has a large band in this region, the first min- eral being beryl, HgBeeAl^SioOg^, which Konigsberger^ found to have a large band at about 0.86 fx. The question of the condition of the water in these different com- pounds is of interest. By " water of constitution " is meant that the HoO is chemically combined with the other constituents of the molecule. On heating, water will be given off, but the residue will not take up water when placed in it. This is not unlike the combustion of a carbo- hydrate, which contains O atoms. Brucite, beryl, tourmaline, and mica are examples of this class of compounds. On the other hand, in crystals having " water of crystallization " the molecules of water and of the mineral are thought to exist in their entirety. Here water is also given off on applying heat, e. g., ordinary copper sulphate, but in these compounds water zvill again be taken up. Gypsum (CaS04 -f 2H2O), which has been discussed with water, belongs to the latter class of compounds. The curves of gypsum, which has " water of crystallization," and of brucite and of mica, which have " water of constitution," are radi- cally different. The former has the bands found in water, the latter ^Konigsberger: Ann. der Physik (4), 4, p. 796, 1901, calls attention to the fact that no substances are known which have large absorption bands near the visible spectrum. ^Konigsberger : Ann. der Physik, 61, 687, 1897. 129 130 INFRA-RED ABSORPTION SPECTRA. does not show the water bands. This would indicate that water of crystallization is not different^ from ordinary water, as mentioned by Konigsberger. The entire absence of the water band at 3.0 /a, which in this compound could not be obscured by the carbohydrate band at 3.1 to 3.43 /a, since it is absent, would make it appear that the 3.0 /a band in the alcohols is not due solely to the OH -group. Curves b and c of fig. 130 are from Konigsberger (loc. cit.). The former (muscovite, HoO-KzO-SAUOs-GSiOo), has a band at 2.g fi, but since it has no band at 1.5 /x, and since its whole aspect is unlike the water curve, we can not consider the 2.9 ja band due to water. Curve c is for biotite mica (reddish-brown color due to iron oxide), and has even less similarity to the water curve. As a whole the curves show that the H and O which exist in these compounds, and which tinite to form water, on applying heat, are in a different condition from the combined O and H, which exists as " water of crystallization." The mica curves are for polarized light. Aschkinass^ found that mica has bands of metallic reflection (absorption) at 8.32, 9.38, 18.40, and 21.25 fi. Selenite ( Crystalline Gypsum). CaSOi-f-2H20. (Fig. 131.) This is one of the most easily obtained minerals for studying water of crystallization. Moreover, it can be split into thin, highly polished folia, the thinness being necessary in order to obtain transparency at 3.0 fi. In fig. 131 the curve a is due to Konigsberger (loc. cit.), who used a fluorite prism. It shows the 1.5 /x water band. Curve b shows the water bands at 2.95, 4.55, and 6.0 /a, the latter being obliterated by the increasing opacity of the mineral, which has a large metallic absorp- tion (reflection) band at 8.69 /x. The comparison spectrum of water is given in curve c, which is for an exceedingly thin film of water pressed between fluorite plates, hence the 4.7 /x band is almost absent. The full curve for water is given in fig. 31, which shows curves for films 0.0 1 and 0.05 mm. in thickness, from which it will be seen that the 4.7 fj. band is considerably deeper. By computation it is found that the film of selenite, of 0.126 mm. thickness, contains a layer of water 0.059 mm. in thickness. This is indicated by the greater opacity of this curve as compared with the 0.05 mm. curve in fig. 31. Curve a con- tains a laver of water 0.12 mm. in thickness. The band of metallic ^In Graham (Otto's Lehrbuch der Chemie, up. 173) the statement is made that water loses its properties, just as other bodies do when they enter into chemical compounds. APPENDIX V. 131 reflection (equivalent to metallic absorption) at 8.69 /x, curve d, is due to Aschkinass.^ The method of selective reflection seems better for studying minerals far in the infra-red. In fact it seems the proper method for making an extensive study of minerals containing water of crystallization, since, with few exceptions, they can not be obtained in thin films, especially for minerals containing several molecules of water of crystallization. As already indicated, the unusual similarity of the curve of selenite and the great dissimilarity of the barite and the mica curves to that of water indicate that water of crystallization is not different from ordi- nary water, as mentioned by Konigsberger. The present investigation at 3.0 to 6 /A is additional evidence to this effect. The observations are of considerable significance, for,^ "in the pres- ent state of our knov/ledge it is impossible to say that there is an abso- lute difference between the so-called water of constitution and water of crystallization, and not merely one of degree," although there is some evidence that there is a difference. The above curves are so marked that it is difficult to conceive how they can be one and the same thing. I am indebted to Prof. A. C. Gill, of the department of mineralogy, for these two minerals. 'Aschkinass: Ann. der Physik, (4), i, p. 42, 1900. ^Encyc. Brit., 5, p. 505. See also Role of Water of Crystallization in Salts of Organic Acids, by T. Salzer, Zeit. f. Phys. Chera., 19, p. 441, 1896. APPENDIX VL INDICES OF REFRACTION OF ROCK SAIvT. (Fig. 132.) It will be noticed that in the present work, using a rock-salt prism, the maximum of the absorption band of water occurs at 2.95 fi, while Paschen (loc. cit.) found it at from 2.95 /^ to 3.0 /x, and Aschkinass (loc. cit.) found it at 3.0 /x, both using a fluorite prism. Furthermore, it will be noticed that in the present work the conspicuous alcohol bands occur at 2.95 ju and 3.43 fx, while Ransohoff (loc. cit.) found them at 3.0 IX and 3.43 fi. That this discrepancy can not be due entirely to errors in the observa- tions has been shown in the discussion of the " sources of errors." From the nature of the dispersion curve, as well as the magnitude of the dispersion of fluorite, which is about twice as great as that of rock salt for the present work, and, furthermore, from the fact that the rock-salt dispersion curve passes through a double curvature near this point, it would appear that this discrepancy is due to errors in the indices of refraction of rock salt in the region of 2.5 /*. Now, it so happens that we have no observational data for this region. The most recent and most reliable measurements of Rubens (loc. cit.) are for wave-lengths 1.71 /x, 2.35^11, and 3.34 /a (fig. 132). LangleyV values are all much larger, which can be interpreted, for the present, as a shifting of his zero of wave-lengths by about o.i /* toward the infra- red. The COo absorption band occurs at about 4.28/1,, while that of CO occurs at 4.58 fi. Unless it can be shown that the great atmospheric absorption band found by Langley at 4.4 fi is the composite of CO and CO2, it would appear that his greater value, 4.4 /*, instead of 4.28 fi, is due to the shift noted above. As indicated elsewhere, the great bench-mark in the infra-red is the emission band of CO2, at 4.40 fx, using a Bunsen burner. The accuracy of the location of this band has also been noticed; and it was shown that all observers agree in the location of this band for fluorite, includ- ing myself, using rock salt. Langley (loc. cit., p. 215) finds this band at 4.6/1, using a Welsbach mantle heated by the flame of a Kitson lamp, ^Langley : Annals of Astrophys. Obs., vol. i, p. 261. 132 APPENDIX VI. 133 which burns vaporized petroleum oil with a strong air draft. This is no doubt the CO2 emission band, as he himself thinks. Unfortunately, he does not give this band, using the lamp zvithout the mantle, so we do not know whether the shift from 4.4 to 4.6 /x is due entirely to rise of temperature, as found by Paschen (loc. cit.), or whether it is due, in part, to the difference in the dispersion curves, as found by him and by Rubens. Langley (loc. cit., p. 219) has also determined the dispersion of fluorite. " The method employed consisted simply in taking a number of holographs with a fluorite prism, comparing these with holographs taken with rock salt, picking out common absorption lines, measuring the position of these lines in the fluorite holographs, and, after reduc- tion of these measures, comparing their results with those obtained for the salt." The indices are all smaller than those recently found by Paschen,^ who compared the fluorite prism directly with a grating. Now, this is just what one would expect, viz, that Langley's indices of fluorite are too small, if the wave-lengths of his absorption lines are too large; and the wave-lengths of his absorption lines are too large beyond 2 p., if his indices of refraction of rock salt are too large, as will be noticed in fig. 132. For, reasoning from the fact that the index of refraction decreases as the wave-length increases, if we find the disper- sion of fluorite by simply "picking out common absorption bands" whose true wave-lengths are smaller than those found by Langley by means of the rock-salt prism, using his indices of refraction of rock salt, then it necessarily follows that, if the wave-lengths of the absorption lines are too large, the fluorite indices will be too small. • The whole depends upon the question of the indices of refraction of rock salt. From the general appearance of the dispersion curve of rock salt, using the indices found by Rubens, it appears that the value, n= 1.5255 for X = 2.35 ij. may possibly be too low. The dispersion curve may be a straighter line at 2.35 /x. than is indicated in fig. 132. If it be a straight line, then the index is n == 1.5256 for A = 2.35 /x, or the wave-length is A = 2.39 /x for 7i = 1.5255. It is not for me to say which of these two investigators' indices are the more nearly correct. It will be sufficient to add that it is quite evident that the Langley values do not harmonize matters, while by straightening the Rubens curve at 2.33 /x (using m= 1.5256) the dis- crepancy at 2.95 fx is entirely obliterated without aflfecting the wave- ^Paschen : Ann. der Physik, 4, p. 302, 1901. 134 INFRA-RED ABSORPTION SPECTRA. lengths at 3.43 fi, and everything ever done with the infra-red absorption spectra that is worth considering is in agreement as far as the absolute value of wave-lengths is concerned. In the same manner the absorption band at 13.8 to 13.9 ju., which is a close harmonic with the one at 3.43 ju and 6.86/* (true value= 13.72 /x), indicates the possibility of the observed indices^ of rock salt at 13 ju, being somewhat too low. The computed values are higher. Of course, it may ultimately be found that this apparent harmonic series is really a slowly converging " spectral series," in which case there ought to be this discrepancy between the computed value, 13.72 /x, and the observed values, viz, 13.8 to 13.9 {x — the latter values being more nearly correct. The whole shows that the absolute value of the wave-lengths of the absorption bands are known as accurately as is possible with the present knowledge of the dispersion of rock salt. iRubens & Trowbridge : Ann. der Physik, 60, p. 733, 1897 ; corrections in Ann. der Physii<, 61, p. 224, 1897. APPENDIX VII. CORRECTION TO TH15 WORK OP JULIUS.^ At the time when Julius found the absorption spectra of some 20 compounds the dispersion of rock salt beyond 5 fx was unknown. Lang- ley's (loc. cit.) dispersion curve extended to 5 /x, at which point it became practically a straight line, and, since he had penetrated the infra-red beyond this point, he naturally desired to know how far he had explored the spectrum, expressed in wave-lengths. To do this he tentatively extrapolated the dispersion curve in a straight line beyond 5 fi. Julius, with apparently less hesitation, has applied this extrapo- lation to his work, which appeared about the same time, and has given a table of wave-lengths of the absorption bands. As is well known now, the dispersion curve suddenly becomes curved beyond 5 fi, and conse- quently all the succeeding wave-lengths thus found are too large. All but seven of the compounds studied by him are recorded in this work. The remaining seven were not examined, and to make the com- parison complete the corrected values of the absorption bands of five of the most important ones are given in the following table. The cor- rection was made by drawing a straight line from 5 /*. The position of this line was determined from the dispersion curve by comparing the values of the maxima of compounds studied by both of us. For exam- ple, for CoHgSH the band at 7.1 /^ on the dispersion curve was found at a point above it (at 8.6 /a), while the 10.26 ju, band on the curve lies still farther from the curve (at 16.5 /a). The maxima of numerous com- pounds thus located lay quite close to the straight line drawn through them, so that with few exceptions the difference in the values of the maxima, as found by both observers, usually amounts to only from 0.02 to 0.05 fj-, depending upon the sharpness of the band. Compounds. PCI3... CHBrs. SiCU. . , SiKCls S2C2I. . . 10.08 6.426 6.1 b 6.0 6.45 o > o 7.840* 5.82 5-64 5-57 5-85 12.4 8.6 6.8 II. 15 9.656 8.82 7-1 6.1 8.27 7-63 14.1 10.05a II. I b 13.0 a II. 8 9.44 7.82 8.27 9.04 8.58 16.] 12.7a 13.06 14.96 12.9 O > V 10.13a 8.95 9.04 9-74 9.0 14.1 6 14.856 14.056 g.44 9-7 g.42 16.26 10.16 *Depth of bands is indicated by the letters a and 6, where a refers to the deepest band. ijulius: Verhandl. d. Kon. Akad. v. Wetensch. Amsterdam., Deel. I, Nr. i, 1892. Beiblatter, 17, p. 34. 135 136 INFRA-RED ABSORPTION SPECTRA. TABLES. Table I. — Line Spectra of Gases. [Height of line indicates depth of absorption band.] 3 D cr 0 3 > 3 3 0 ■I 0 TO 3 0 0 C/1 fC 3 p c • a. 0 •0 3- D. Z a. ,n 0 p 0 0 X 0* n X X __ - - - - - - - _ _ - - 0 — - - — oc - 0 - ■~~ - V=3l{ - ^^^ ■ - V=3/ / — - ^ . - - ■• — 7 _!__ ._.. ^4._^X4__ .^.^.iJl L ..-.a. ,j1-L1_lJ l_ M : SPECTRA OF VARIOUS COMPOUNDS. n^^ ___^J^__ 11,^ 1, 1.. LJ__j ^ . I I _ u.l -fftfa i#ffftt SPECTRA OF VARIOUS COMPOUNDS. _.J^_ .J. JJlJi ._ I L ^___L .-i-. _^J !-__ L 4.. I -^--l- I , L_l , __. , 1_L 1_. 4_ 1 +-^-+" X_L SPECTRA OF VARIOUS COMPOUNDS. S3SSpg||p L U A fc'fe^ : _L ML A _ 1 : X L J LI J: ±± Il_ 1 ^J_-L_._ ::ii:]^ii.;:Li:|£^.;.i^^|ii;[4.: ; SPECTRA OF VARIOUS COMPOUNDS. fsiiii ,^__i ^_t-^i L Lj lj_ L.. J J ., JNE SPECTRA OF VARIOUS COMPOUNDS. I>ac w P !zi s i Iz; '^ m O -< H ■s(ho}ShSo •nu3DXio o d ocoo • • OCO • :gS : ^ . ooo • • t- • \i : CO <33 -* • X S : o ■«hooooShd d ooTt<>n ■ coco • Igj: : in-* 1 • ooo 00 00 . o • • i^ : : o • ^ : o •»o^h8d in d oooeoio * I- o in 00 • i- t2 : o • o • o ■* IN IN IN O d CI CQ ^ iC CO 'CD CO CO CO CO CO • CO t- t- t- t~ i^ ■ i- ^ moo • coo t-^QO o E: • o • is : ig •S!5(«hSd) •apiqdins i^nv o d CD CO O t' 00 «•! O CO in i- CD CO 00 00 00 00 00 00 00 00 00 §8 : "^9 : • OoS ; ^05 oJoo 00 00 (N • (N • s ■ m • :g : • 03 CO •ND^HD d ■<)< O OO M CO CO CT (N (N ^ ^ lO ■* in 00 00 00 00 00 00 00 00 ■* 00 'mm • :£SS : • o is m • 00 • t-^ • 00 • 00 • :g : • o is is ■OSiH^'O 'lojditiBDna o d o -* 00 in o m i~ CD CO CD CO CO CD CD • in • cooo • • cod • m q CO • q • 00 •ho»h9odoo-£hd in o d ■* in c^ Tt< o CO CO -1^ m in in 1-' i-^ 00 00 00 00 OO 00 00 00 • o ^ oin • • 03 t-^ ■ •0000 ; eooin • 00 cooo • OOOOCO ; • m • ■* 88 • m :§8 • N is •SDN?H9D •jio pjBjsnui iXnaqj en d t^ CD r^ o (M 00 in irj CO t-^ T)! in 00 00 in in in in in in in : Sto • mcoin ; coSS • l-O • SS : q CO CO • • 00 • i 1^ : •S*H*0 -anaqdouix d incDoo • r-.-. CD IN in in t» i^ t- • 00 - -* (N • :^g2 : 00 i^-* - ■^."^ : 0000 ; CO OO ■* ; cDcoeo ^ .-^(NCO ■00 00 00 ■^lOHO -TOJOjojomo d eooocD • Tj* in in • minin • COININe* • in CD 00 oi • in in in in • ■OMCO 3i r-^ cj in CD CD ■*mco • ScD CO ; 00 t- j -^m • coco ; (N r* • (M"* INO CO CO • CO CO COCO •9IhOIo -aninij-^ p d ■Tj' in CD • 05 OSOO * MOOO ; 00O5OJOS ; ■J<(MCO O j t-; ; Oi • 05 ; 0O-* ■ co4i • 0105 ; 00O00-* t-- CO CO IN (NCO 03 05 05 05 GS • -* • 03 ■OH0SH90 apAijapiBzuaa en d C000©!t~CD-*!-O ; »— •O5O50CO5'?J'?-lCO • CD in in in in CO CD CO • ■*o CO CD CO CD O t- IM ; CO CO CO • OOOi ; CD CD • COCO ; • coo-* ■iii •9H9D anaznaa d j O j ; ; CO • '■ '• • 00 • • • r- • • • • m • t is :? i :g : : I' ; ; 03 • • ooq ; 0000 • • 00 ; 00 •qiaaai-aABAi S 8 is ^ s-i--ooooooooojcR05c:500c; j2S2SS5^?.??§§^^^5:S •Sni:| -13S aa^anioJiDads ^ t~ocoir (NCOCO c^ " o t-Oi (N CO CO ■» ■ d X CO r- • COOOO M • CO x' d X . Oifso o in inoo diN 00^ ■ OXOJ • CO ino . ooino • 1ft dx*» • 0600 t- T» . coooo _ 5 CD -^ in . i^ t- 1~ t- • coo , d-* ■ CO ■* O O O CD • •'^ CO in in d CO • XX xxxx CO o ^-x in coco ^co t^ c^ CD in in -rj^ . in CD Oi in . o t-coo ■ .intio--(Nint~x-*05««3-t^J-¥^f^l-?-xxSSooo.-'--'-cj»iiNNcocococo^-«;inininincDcD:^ *'■'-'---*-* -^c^^ ©^fl^^ iri ^ci c-i iN« IN iri ^J IN ffi N w WW NC4C4 i W IN « «■! Ci N IN W IN IN 1: eoxoxi M M fi fi M 2.35 3.08 2.31 2.36 2.4 2-5 2-5 4.24 2.98 2-35 2-75 3- 18 3-1 331 3 69 3-28 3-39 3-42 3-45 3-45 5.6 5-8 4-59 4.29 3-97 4-7 7-7 7-38 4-32 6.85 4-5 4-75 5-15 7.12 6.1 14.65 5.68 7-73 5-3 11.94 5-7 5-7 7.0 7.78 6.51 *i47 7-4 13-63 5.8 12.2 6.89 6.9 7-3 8.46 8.9 8.7 6. 98 12.5 7.8 8.58 7.8 9-65 9-3 10.37 lO.O 13-2 8.3 905 8.75 10.08 9.9 10.5 8.9 9.16 9-25 10.6 10.4 11.05 9.6 10.55 9.6 10.95 10.75 11.8 9-97 10.5 II. 2 12.0 13-3 14-5 10.9 10.2 10.75 11.99 14.03 II. 18 11-43 11.69 11.98 12.3 12.78 13-7 *Rubens and Aschkiuass. TABLE VI. 143 Table VI. — Maxima of Infra-red Absorption Bands — Liquids and Solids. [In iJ. . = 0.001 mm. Values from Puccianti are marked with a star (*).] sJtC tn W «o y CI Op 0 J? d«5 00 oo vXS 10 la XX n X X 0 .•25 0 30 0 Z 2z "1 aX "He jj'OO U 1 0 0 . 1 V la •S 6 0 fa 0 £-u So 0 S ■SO a •2 0 « 0 0 a ■* 0 1 tJffi >> >1 0 .0 01 0 a 0 0 >> .c u w 3 as W a V X V >. 0 W |5 < X X V 3.2 0.83 2.95 2.9 2.7 3.42 2.6 0.84 1.71 1.35 0.78 3.42 0.88 0.84 3.0 4.6 1.12 3.2 3.4 3.42 4.68 3.42 0.96 2.3 1.7 0.85 4.72 1.0 0.96 3.4 6.8 1.4 3.6 4.68 4.78 5.6 4.78 1.1 2.9 1.86 0.95 5.75 1.13 1.08 4.38 11.65 1.66 3.95 6.2 5.8 6.2 5.8 1.68 3.42 2.1 1.07 6.6 1.5 1.25 5.1 13.4 2.0 4.43 6.6 6.3 7.0 6.5 1.95 3.75 2.27 1.5 6.9 1.72 1.37 6.1 2.23 4.77 7.08 6.6 7.28 6.97 2.18 4.5 2.37 1.73 7.3 1.95 1.7 6.98 2.43 5.85 7.64 7.03 7.95 7.53 2.42 4.85 2.5 1.92 7.6 2.10 1.9 7.7 2.56 6.25 8.1 7.2 9.46 8.7 2.5 5.35 2.75 2.15 8.3 2.3 2.28 8.55 2.64 6.5 9.75 8.3 10.33 8.98 2.7 5.78 3.43 2.32 9.1 2.4 2.53 9.33 2.72 7.05 1 0.2 8.6 10.7 9.5 3.25 6.05 4.65 2.40 9.7 2., 53 3.0 10.8 3.22 7.85 1 1.3 9.28 12.3 10.1 3.58 6.55 5.8 2.55 10.4 3.0 3.35 12.75 4.3 8.32 1 2.6 10.05 12.98 10.66 3.9 6.65 6.3 3.45 10.9 3.3 4.37 13.7 4.62 8.85 10.45 11.3 4.82 7.05 6.95 3.75 11.3 3.7 6.1 5.6 9.3 11.3 12.6 6.25 7.54 8.1 4.35 12.65 4.8 7.04 6.29 9.5 12.6 14.1 6.75 7.75 9.2 5.85 13.1 5.8 7.25 6.5 9.95 7.3 10.4 7 to 8.6 7.1 8.3 7.19 10.3 7.97 11.0 9.1 8.2 9.3 8.07 11.47 8.7 12.65 9.8 9.6 9.65 9.28 12.8 9.39 13.2 10.5 10.5 10.15 9.7 10.08 13.95 11.05 11.2 10.6 11.1 10.8 14.3 11.65 12.0 10.88 11.98 11.1 12.6 12.7 12.0 1 13.4 13.55 12.6 14.0 13.4 14.5 13.8 15.2 0 Z d E p 0 . « DO n w K 0 •3 0 4: K 0 ■a •5 .2 0 0 u a 0.23 So 0 u •a >. 0. .X ^0 0. s in vX .So a^ >> a "3 X 5 0 a u 0 0 0 X V S u 0 0 ii v a 4^ P X 0 u V a 0 X X t 0 2 u 2 2 2 V S J3 < ? < -a 0 5 X W 0: 0 0.84 3.4 *1.71 *1.71 1.71 3.45 3.43 0.84 3.43 3.46 3.3 0.84 1.7 5.8 5.6 1.10 4.1 2.7 2.7 3.44 4.0 5.2 0.96 4.8 6.0 7.75 0.97 2.3 6.45 6.3 1.24 5.1 3.4 3.41 6.2 5.85 5.76 1.02 5.7 7.1 8.65 1.1 3.35 6.57 6.85 1.43 5.4 4.6 4.6 7-7.8 7.1 6.2 1.18 6.1 8.1 9.45 1.2 4.6 8.02 7.5 1.7 5.78 5.4 5.28 9.6 7.25 6.5 1.70 6.86 9.8 10.7 1.69 6.3 8.25 8.0 1.92 6.37 5.9 5.9 11.4 8.05 6.87 1.96 7.27 10.4 12.1 2.38 6.98 9.0 8.2 2.27 7.0 7.1 7.1 9.25 7.28 2.14 7.75 11.6 2.49 7.85 9.3 8.7 2.43 7.3 8.22 7 2 10.34 7.8 2.34 8.3 14.1 2.75 7.05 10.1 9.0 2.58 7.6 9.2 8!36 11.6 8.0 2.55 8.79 3.32 8.46 12.3 9.9 2.65 7.93 9.75 9.6 12.65 8.6 2.64 9.3 3.8 9.2 to 10.2 3.41 9.05 10.1 10.5 9.38 3.4 10.0 4.4 9.82 13.6 11.0 4.1 10.08 10.7 11.2 10.28 6.2 10.95 5.6 10.75 12.5 5.1 10.96 11.3 11.68 10.7 7.25 11.67 5.9 11.14 to 5.8 11.45 12.3 12.1 12.8 8.3 12.5 6.8 12.0 13.4 6.35 12.45 13.95 13.6 13.9 9.0 13.5 7.2 13.0 14.7 7.75 13.25 14.95 14.1 9.5 8.3 13.37 8.3 14.0 15.0 10.1 9.6 8.7 10.9 10.2 9.1 11.3 10.82 9.95 13.4 11.8 10.5 10.92 12.0 13.2 13.96 144 INFJiA-RED ABSORPTION SPECTRA. Tabi o'4 u 0 o a o a >a UK ao a 2 o a o u 1 1.1 a o N a V n X O X o O CO "^^ eg o 03U O 1 o o a J' bo '1 O V ">. u u . o (-5 •^ X u a a a >. o t-l a S 'E s a 2.97 2.7 3.0 3.4 2.95 3.26 3.25 2.65 0.85 0.84 3.45 1.7 3.43 3.44 3.25 3.25 3.25 5.0 3.42 4.3 5.6 3.25 1.0 0.96 3.9 3.43 5.2 3.7 3.75 3.43 3.43 6.26 4.0 5.4 6.28 3.75 1.36 1.5 5.88 6.05 6.25 4.7 4.25 4.0 4.2 6.86 4.3 6.27 6.93 4.3 1.5 1.7 6.6 6.25 6.7 6.05 4.6 4.7 5.2 7.4 5.3 6.77 7.85 4.45 1.64 2.15 7.1 6.6 7.55 7.15 5.15 5.3 6.19 8.18 5.7 6.94 8.13 4.95 1.71 2.3 8.0 6.86 8.3 7.5 5.5 6.17 6.7 8.63 6.13 7.50 8.6 5.27 1.88 2.. 54 9.9 7.30 8.7 8.1 6.1 6.7 7.48 8.9 6.68 8.0 9.42 5.62 2.02 2.60 10.7 7.72 9.5 8.55 6.68 6.95 8.2 9.43 6.85 8.7 9.85 6.0 2.17 2.72 11.2 8.2 9.9 9.23 7.6 7.. 58 8.65 10.2 7.25 9.0 10.2 6.25 2.30 2.88 11.8 8.6 10.4 9.9 7.88 7.96 9.4 10.6 7.78 9.28 11.15 6.75 2.42 3.4 12.3 9.1 11.7 10.1 8.62 8.65 9.7 10.85 8.18 9.86 11.8 6.90 2.47 4.0 13.5 9.5 12.5 10.52 9.0 9.38 10.07 11.3 8.7 10.6 12.2 7.2 2.58 5.1 13.9 9.7 13.2 11.4 9.4 9.8 10.58 11.8 9.18 11.1 12.9 7.9 2.90 6.26 10.6 14.1 11.8 9.7 10.15 10.85 12.18 9.6 12.2 8.6 3.42 6.75 11.4 12.7 10.03 10.5 11.45 13.75 9.84 13.23 9.37 5.78 7.0 12.25 13.3 10.73 11.48 12.2 10.2 9.78 6.07 8 to 9 13.95 14.05 11.32 12.3 12.7 13.35 14.48 12.7 13.35 14.5 10.85 11.5 12.28 13.6 10.00 10.3 10.85 13.25 6.82 7.43 8.15 8.4 9.12 9.62 10.06 10.85 11.67 12.36 12.9 13.9 9.75 10.1 10.7 11.0 11.78 12.2 12.5 13.7 TABLE VI. 145 Table VI. — Maxima of Infra-red Absorption Bands, etc. — Continued. to 1 0 1 HO V d K 0 u d 0 0 a S 0 •4( s to 0 i 0 0 w 0 ii a K u a 0 11 a X s 0 a 0 .a a Im 0 .3 V d d a as u 3 0 a g CI u d Id cS "p. a u V V u u 0 u 0 >. a 0 s V m i4 a V > aj u W V Q V 5 0 0 -a 0 Q u H V 0 • K W 0 V n u ■>. 0 V 3 u a ij % u V ■d 0 Q i w liO a V >. u V •a a V a 41 u -a CS X u W B .0 a ">. 0 u -a 0! a 0 a § 0 a u H ■£cj I1 0 S 0 0 a 0 DO 0 h S K .a i- d 1 0 2 H d 0 2' CI CS > u a 0 i 0 V is 'u 0 a a 0 u 2.55 2.48 2.43 2.48 2.43 2.35 2.35 2.43 2.43 2.40 2.92 2.93 1.70 1.70 3.44 4.2 3.44 4.2 3.43 4.2 3.44 4.2 3.43 4.2 3.44 3.75 3.43 3.75 3.44 4.2 3.44 4.2 3.43 4.2 3.42 4.0 3.42 4.0 *1.71 2.95 2.92 3.36 5.8 5.8 5.8 5.8 5.8 3.95 5.78 5.85 5.85 5.82 5.4 5.05 3.25 5.1 6.28 6.87 6.15 6.88 6.86 7.3: 6.2 6.88 6.3 6.86 4.45 5.95 6.3 6.6 6.88 7.35 6.88 7.34 6.28 6.86 6.15 6.3 5.37 5.85 5.2 5.8 6.27 6.75? 7.34 7.34 7.85 7.34 7.35 6.2 6.87 7.85 7.85 7.35 6.6 6.14 6.25 6.86 7.8 7 8 8.65 7.80 7.85 6.6 7.32 8.1 8.1 7.8 6.98 6.28 6.75 7.05 8;65 9.3 8.66 9.26 9.28 9.7 8.66 9.26 8.65 9.28 6.87 7.32 7.8 8.1 8.68 9.34 8.65 9.22 8.1 8.82 7.6 7.85 6.6 6.94 6.95 7.30 7.30 7.8 9.75 9.67 10.38 9 67 9.78 7.8 8.67 9.8 10.38 9.3 8.25 7.44 7.75 8.15 io!4 10.38 10.61 10.4 10.34 8.1 9.3 10.4 11.28 9.6 8.68 7.80 8.26 8.74 11 4 10.8 11.3 10.75 10.58 8.65 9.75 10.84 11.7 9.9 9.3 8.07 8.78 9.14 12! 35 13.2 13.85 11.4 12.3 13.3 13.85 12.32 13.2 13.78 11.4 12.28 13.1 13.65 11.3 12.3 13.2 13.78 9.3 9.9 10.42 11.4 10.38 10,8 11.8 12.35 11.2 11.7 12.34 13.2 12.35 13.2 13.83 10.35 10.73 11.3 12.32 9.98 10.58 11.63 12.33 8.58 9.08 9.5 10.08 9.35 9.73 10.08 10.53 9.58 10.07 10.54 11.3 12.35 13.2 13.8 13.3 13.6 10.65 10.85 11.85 13.2 13.83 13.8 14.4 11.55 11.28 12.4 13.83 12.3 13.1 13.55 11.7 11.9 12.4 13.3 13.25 14.3 146 INFRA-RED ABSORPTION SPECTRA. Table VI. — Maxima of Infra-red Absorption Bands^ etc. — Continued. 5 Z 0 0 0 0 0 0 0 0 0 0 w i X 1 w S .w K -dW !5 N 0 0 l-r il ,—. ^ » .9 2 il g. 0 V a 0 "3 0 0 V u I0 "3 > y "0 u 0. cs 0 p E. a 0 '0 ■6^ '0 a 0 'u cs D ^0 0 01 0 0 u il Co iiZ^ 3 * oB 'is + d a i) H 'p^ a a u n si V u ii a < 4t K 0 V a V to w 0 6 a us 0 0 d a a 3 M d V .a W a "3 Ml a V bo 2 t .2 "3 0 B a < .6 no 'd 'm 0 a 0 a a 0 •e ca U & 0 ■3 a 0 •e cd 0 0 ."2 '5 _0 •3 a d a u M 0 25 30 25 22 21 1.7 2.00 2.28 2.37 2.46 2.55 2.65 2.73 2.82 2.88 2.97 3.03 3.13 3.20 3.31 3.38 3.46 3.52 3.60 3.67 3.74 3.81 3.88 4.04 4.18 4.26 4.32 4.40 4.46 4.51 4.57 4.65 4.72 4.9 4.98 5.04 5.11 5.15 5.23 5.34 5.50 5.60 5.70 5.82 6.00 6.10 6.20 6.30 6.36 6.47 6.51 6.56 6.60 6.66 6.77 6.87 6.96 7.00 7.04 7.07 7.12 7.17 7.22 7.32 7.36 7.44 93.0 1 93.0 00.0 99.0 94.2 91.8 84.4 84.7 86.8 90.3 93.5 94.2 84.7 82.2 83.4 85.3 88.5 00.0 99.0 87.2 98.0 96.8 89.1 00.0 98.0 8.10 00.0 99.0 96.4 94.0 95.7 99.0 99.2 20 19 18 17 16 15 86.3 98.5 85.1 80.0 75.3 100. 0 97 3 99.0 91.4 96.4 95.7 96.8 90.5 83.3 77.8 93.0 99.0 94.0 87.0 82.8 98.5 98.4 97.5 95.4 92.8 83.8 73.0 93.3 88.6 92.2 88.7 94.1 93.7 93.5 93.7 88.7 80.6 74.0 72.6 77.2 79.2 85.6 93.0 "92 14' 82.7 97.2 86.3 92.8 91.7 87.8 81.8 78.2 77.3 81.2 85.7 83.3 14 13 12 11 10 9 8 7 6 5 4 3 2 25 00 58 57 56 55 54 53 52 51 50 47 46 45 44 43 93.4 88.7 86.0 78.2 88.0 85.2 91.2 94.8 75.0 77.3 63.0 70.2 56.2 69.0 55.0 71.3 58.3 74.4 62.8 84.1 83.4 94.4 93.0 94.6 70.0 77.3 91.0 gels' 95.2 92^8' 93.2 94.0 62.8 43.5 37.7 43.7 59.8 55.0 '31 is' 31.0 38.6 47.7 61.8 59.2 43.2 27.7 22.4 25.1 37.2 50.8 70.4 50.3 31.6 18.6 16.2 SJ.I 23.8 79.8 94.6 92.0 99.0 95.8 34.6 22.8 11.9 8.3 11.2 22.2 94.7 92.2 100.0 96.2 96.8 90.8 96.8 '98 !6' 93.3 93.2 92.2 97.4 93.6 81.3 80.8 48.3 95.7 [00.0 59.6 86.4 79.8 77.0 81.8 98.0 96.0 90.6 86.8 93.7 91.0 93.7 95.1 69.8 81.7 100.0 89.3 59.2 51.3 95.0 98.0 86.3 87.1 100.0 94.0 91.8 99.4 88.3 79.4 85.1 '76! 2 61.3 59.8 60.3 64.5 79.0 89.2 49.8 58.7 88.7 97.8 96.0 99.0 86.2 93.1 90.0 87.0 87.3 100.0 74.2 94.3 96.6 88.7 92.7 86.0 94.5 95.8 97.0 99.0 'm'.s 86.3 93.0 94.3 95.3 65.0 62.5 88.7 99.0 100.0 100.0 100.0 95.6 97.4 96.0 0 •8- J 85.8 95.0 '79.8 80.7 100.0 100.0 99.0 94.0 72.8 92.8 42 40 37 35 33 77.3 93.5 94.0 91.8 95.4 52.5 73.8 91.0 99.0 94.7 96.7 97.4 97.0 ioo.o 97.6 91.0 85.6 86.0 93.2 94.3 97.3 100.0 92.5 74.3 69.3 78.3 89.3 88.2 87.0 84.5 86.8 94.2 59.2 41.4 36.0 30.3 19.3 21.7 27.3 28.7 23.9 22.8 23.0 27.3 37.3 49.8 64.7 92.0 92.8 95.0 85.6 81.3 78.8 74.7 91.7 27 25 23 93.0 95.0 88.5 91.7 99.0 21 '^i'.s".'.'.'.'.'. '97 is' 'siis 56.5 57.7 "89!6' iooio' 97.3 20 18 17 16 15 14 12 85.4 18.6 34.1 33.6 50.0 97.8 97.3 43.7 21.8 19.6 18.9 12.7 11.9 '22! 4 20.8 18.8 19.4 "26!3 "28! 4 4.2 2.8 6.4 7.3 23.5 9.6 7.8 5.7 'gsio" 91.0 98.0 84.2 8 7 74.7 6 ' 89.0 '.'.'.'.'.'. 'geio '2i!6 '26;8 '2i!7 30.5 38.8 55.2 67.2 "e.b 8.4 6.7 3.3 2.2 87.8 78.1 44.4 5 4 3 '44!5 22.3 10.6 14.6 75.8 89.6 '83!6 90.6 13.8 3.1 •«• 50.0 72.6 50.8 24 00 58 57 89.5 81.6 68.4 74.8 7.6 . 6.0 .1 10.7 'si'.k".'.'.'.'. 05 5 Q« 8 55 v.oa 148 INFRA-RED ABSORPTION SPECTRA. Table VII. — Observed Transmission Through Gases — Continued. 1 'u u v i 5C m M a P S u V a a w o v a V o a it i o n OS W 6 o d a S P5 6 W u 6 "S vt/i a a V 1 a "3 0 a s .0 a 0 a a 0 .n u 0! 0 6 0 '53 .2 ■3 a 0 •e 0 0 .2 '■3 *-• a .a a "a d a' u be >^ 0 o / 53 ft 7.59 7.72 7.84 7.93 8.00 8.12 8.31 8.49 8.67 8.83 8.99 9.16 9.32 9.48 9.63 9.78 9.95 10.08 10.23 10.29 10.37 10.45 10.50 10.63 10.70 10.77 10.86 10.92 11.03 11.12 11.17 11.30 11.43 11.51 11.55 11.61 11.68 11.77 11.82 11.89 11.94 12.02 12.06 12.20 12.32 12.40 12.45 12.56 12.66 12.93 13.17 13.40 13.62 13.85 14.06 14.26 14.45 14.63 23.8 20.0 31.0 53.8 83.7 65.8 63.7 68.8 71.8 63.4 34.1 40.6 47.8 32.2 44.0 62.3 73.2 55.3 53.0 52.7 48.7 51.5 60.3 75.6 68.7 54.2 41.2 50.8 41.7 41.8 46.5 93.2 81.0 78.8 87.6 17.2 64.6 85.6 88.7 50 30.6 93.0 95.3 97.0 47 45 37.5 93.0 96.8 97.0 43 40 56.5 84.5 97.6 99.0 100.0 100.0 96.0 1 o . . 95.0 93.0 Pres. 20 cm. "97!o 89.0 'sols' 74.3 60.7 46.8 35.9 26.5 26.3 26.1 95.8 93.0 96.3 97.2 97.4 96.5 97.0 '95.'6' '95!6' 95.0 Pres . 10 cm. 75.0 53.7 7.1 9.1 39.3 18.4 12.9 39.9 86.0 95.0 96.8 99.0 89.7 75.2 61.7 51.8 6.2 0.0 0.0 "5.3 8.6 30.7 35.5 46.8 72.7 87.3 87.3 99.2 87.8 81.0 90.8 92.0 93.0 95.5 95.0 91.4 86.8 89. 6 94.6 90.6 95.3 89.0 82.3 66.7 76.8 25.7 1.8 0.0 4.2 10.7 30.8 69.0 82.2 88.6 92.3 94.8 94.7 77.0 67.8 56.8 61.6 70.3 92.3 35 23 30 25 20 Fres. S3 cm 62.4 60.0 55.2 40.0 44.3 50.0 58.7 62.7 64.8 68.7 42.8 17.1 31.4 44.3 33.7 16.0 17.1 40.8 82.3 78.8 70.8 55.8 81.0 62.9 74.3 80.0 61.1 35.7 65.0 73.5 68.4 42.3 41.4 58.1 67.2 45.8 67.2 75.0 62.0 53.0 55.0 67.4 15 10 5 23 00 55 50 45 40 35 33 30 25.3 18.1 5.7 19.8 94.0 57.0 54.3 95.0 99.0 27 25 '98.'6' 50.4 59.8 38.4 32.8 41.4 43.4 34.0 35.0 50.8 87.3 79.3 95.4 94.7 20 18 15 26.8 64.6 62.8 97.3 100.0 12 10 25.1 24.0 96.8 77.8 73.8 82.3 98.0 95.8 100.0 5 2 22 00 28.0 43.2 49.6 88.8 61.4 57.8 67.2 62.7 61.0 67.0 76.5 87.0 97.0 100.0 14.06=97.2 14.26=81.0 14.45=60.3 14.63=55.2 14.82=62.8 15.00=73.6 15.20=91.3 15.37=97.1 100.0 55 50 95.6 47 45 57.7 51.7 57.3 92.7 43 40 69.8 44.3 53.0 98.0 77.1 37 35 72.3 38.7 32 21 30 75.8 34.7 50.0 98.8 49.3 27 25 95.0 99.0 40.0 42.0 48.4 20 89.0 79.0 51.4 99.0 66.8 15 12 10 100.0 50.2 62.8 100.0 87.3 5 Pres. 7 cm. 21 00 96.0 78.4 66.7 55.7 42.2 61.8 70.3 75.0 81.3 84.5 54.8 80.8 87.3 100.0 67.7 73.3 73.3 75.5 84.3 77.6 71.8 77.4 84.0 97.0 100.0 86.0 50 40 87.0 76.2 63.8 55.0 40.0 64.3 74.3 75 0 30 20 71.4 10 20 00 83.0 50 40 19 30 99.0 TABLE VIII A. 149 Table VIII a. — Observed Transmission, Using Small Spectrometer, and Prism AT Variable Deviation, as Explained in the Text. [For the petroleum distillates a single cell was used, made by cementing a copper wire, 0.15 mm. in thickness, between two plates of rock salt.] Spectrometer s e t - tings ; angular ro- tation from the sodium lines. oi M a it V > a X v tJ R 0 X V K i u 1) n c u 0 0 ll' a a CI u 13 0 Q 0 6 a u V •a V g X 0 u' a C3 S u I CO w .0 a 0 £i u a u 0 u •0 >. W w n .0 n 0 .Q 0! 0 ;-■ 13 >, X 0 X 0 6 ii g X n V ">. u S a • i- = S ii £.- cc p 4 i n ^ g n w K W w o o a ^.o o u 0 o Xi ce o CS Cj rt u o -a o o o n! )-. ;-! >> o u •0 w r-i w a 0 0.15 0.03 0.15 0.15 0.15 Thickness in] mm. J 0.24 0.15 0.15 0.15 3 14 16 19 22 24 26 3 29 32 34 36 39 3 42 44 47 49 51 54 56 59 4 4 9 14 19 22 24 26 29 4 32 34 39 44 49 54 59 5 4 9 14 19 24 29 34 39 44 49 54 6 00 10 20 30 40 45 50 7 00 10 80 bO 40 8.20 8.27 8.35 8.46 8.56 8.63 8.70 8.80 8.90 8.97 9.07 9.13 9.19 9.29 9.35 9.45 9.60 9.76 9.92 10.06 10.15 10.21 10.26 10.35 10.42 10.48 10.61 10.75 10.90 11.00 11.14 11.25 11.41 11.53 11.65 11.80 11.92 12.03 12.18 12.30 12.42 12.54 12.66 12.93 13.17 13.40 13.62 13.72 13.85 14.05 14.26 14.45 14.63 14.82 8.5 11.3 16.3 17.5 20.0 19.4 18.8 19.4 16.8 18.5 25.5 27.5 28.2 22.8 15.4 12.5 15.0 16.4 16.0 20.2 26.8 25.2 25.8 27.0 28.4 21.2 15.0 10.0 10.4 11.0 13.8 36.3 25.0 21.2 20.2 22.5 23.8 15.7 15.0 5.9 5.4 25.0 38.0 50.0 39.4 40.8 44.8 53.0 54.0 49.0 45.4 48.2 53.7 55.8 59.4 59.0 50.0 55.4 60.3 61.0 61.0 60.8 59.4 55.2 53.0 49.2 45.1 46.4 48.0 56.3 55.0 55.0 55.0 55.2 57.6 66.0 70.0 71.2 69.5 65.0 71.4 71.4 66.8 57.2 52.0 48.0 35.0 25.8 46.8 64.0 72.0 76.0 71.0 Pe7- cent ti-ansmission. 15.3 19.8 22.9 30.6 38.5 38.5 36.0 37.8 39.4 45.0 44.0 40.8 40.2 42.4 46.7 46.8 46.7 46.4 41.3 34.5 35.6 30.0 26.0 28.5 33.0 34.0 35.0 36.4 41.5 54.0 63.0 60.0 53.8 53.0 56.4 55.6 50.0 48.0 30.1 19.0 10.0 3.0 22.0 42.0 71.0 80.0 12.5 15.3 19.4 24.3 25.0 19.6 18.1 18.6 20.0 26.3 27.7 28.6 25.0 23.8 22.5 25.4 30.5 30.0 29.2 29.7 31.0 26.5 23.0 22.0 24.2 26.5 25.5 26.5 25.8 25.7 26.6 29.5 33.4 39.0 39.7 39.0 37.5 35.5 43.0 48.0 30.1 20.5 23.2 7.5 7.2 15.0 30.0 50.0 56.0 64.0 84.6 89.0 87.3 84.9 84.2 80.0 77.3 78.5 80.0 80.4 78.8 78.5 79.0 74.0 76.1 79.6 80.8 81.0 80.0 79.9 82.0 78.8 76.0 80.0 80.0 81.0 79.5 77.0 63.8 50.0 26.5 38.4 57.5 72.5 74.0 27.0 29.8 35.4 40.1 42.8 37.1 37.4 38.0 41.0 48.3 56.0 45.0 41.8 42.8 43.8 42.7 42.0 44.7 45.4 46.3 42.0 37.8 38.0 41.6 46.3 46.8 47.3 47.3 47.2 44.7 42.0 44.7 48.2 47.4 36.0 30.0 32.1 44.5 45.4 35.0 31.8 30.0 20.5 0.0 15.5 27.0 55.6 71.0 25.9 41.8 39.8 37.6 52.3 43.0 43.7 42.8 41.0 41.7 44.0 '40.8 43.5 45.3 48.0 46.0 48.8 35.5 29.2 44.0 40.0 31.8 27.0 15.0 8.0 21.0 41.5 62.0 72.0 23.7 36.6 37.5 33.0 30.0 29.6 35.8 44.8 44.5 43.7 42.8 43.8 49.0 43.3 44.6 46.4 47.0 42.6 40.0 34.5 34.6 37.0 37.5 37.6 41.0 44.1 46.5 47.6 52.8 59V8 61.3 56.3 60.3 60.3 57.0 49.0 45.3 36.5 24.0 14.5 25.0 47.0 67.0 77.0 34.0 37.5 38.5 34.5 29.5 30.0 36.2 43.4 44.8 42.5 42.0 37.8 37.5 40.8 41.8 39.2 34.4 34.0 38.0 42.2 42.7 38.6 42.7 45.3 18.0 35.0 35.2 28.5 27.0 12.0 5.0 10.0 30.0 51.2 64.5 70.0 TABLE VIII A. 151 Table VIII a. — Observed Transmission, Using Small Spectrometer, and Prism AT Variable Deviation, as Explained in the Text — Continued. Spectrometer s e t - tings. Angular ro- tation from the sodium lues. in a > 1 V a V >. y V 0 Q ?i X a u u •0 a) V X .6 a. _s ■?. u V (S 0 0 K a ti s u IS ll t3 H 4 g| |u S "0 0 a 0 ■ U ll 0 1" i 1 a 6 0 ■£ (d u 0 ;-i -a >. K S a 0 la X tf 0 01 a < "9 V a u 0. 0 Thickness of \ cell in mm. J 0.15 0.15 0.15 0.15 0.15 0.15 0.12 0.01 0.15 0 ' 1 30 35 37 39 40 41 42 43 44 45 1.70 2.05 2.18 2.37 2.46 2.55 2.65 2.73 2.81 2.88 3.04 3.13 3.21 3.30 3.38 3.46 3.53 3.60 3.67 3.71 3.81 3.88 3.96 4.18 4.33 4.46 4.72 5.04 5.16 5.34 82.8 81.7 78.5 66.3 83.4 81.0 79.0 77.6 70.0 62.4 61.7 61.7 79.6 77.3 73.0 73.6 65.0 59.4 81.0 77.4 71.3 64.0 72.3 74.3 68.0 65.8 80.1 79.5 78.0 77.5 87.5 80.1 64.5 65.8 ""'86!5 64.6 67.0 69.7 61.0 57.0 63.3 65.5 68.4 68.3 62.4 61.8 60.0 53.6 66.0 76.5 77.4 67.0 55.4 59.0 50.8 59.7 57.2 47.4 62.8 53.8 62.3 54.7 73.4 69.7 47 48 49 1 50 51 52 53 54 55 56 57 58 59 2 2 4 6 10 15 17 20 75.0 66.8 32.6 25.2 16.0 12.2 12.9 19.1 25.6 38.0 38.1 37.1 32.3 38.0 41.0 53.9 57.7 55.2 20.0 15.4 15.3 17.5 19.0 14.3 12.9 14.8 21.9 18.9 14.6 12.7 19.1 15.0 14.7 19.0 25.0 20.6 15.4 14.4 27.2 25.0 28.1 36.2 43.0 44.6 51.3 56.3 ■■■■64'.4 20.2 23.5 66.3 77.5 32.6 51.3 39.6 30.8 38.1 74.3 61.5 61.8 65.8 67.0 52.8 43.7 48.3 55.8 57.9 47.8 75.1 75.7 75.3 85.4 81.5 68.2 57.4 59.6 80.8 79.0 74.0 77.5 62.8 64.1 62.3 63.5 69.8 69.8 75.7 69.3 69.1 76.3 78.0 65.6 8.40 77.0 63.2 65.6 63.5 69.8 84.7 24 27 28 2 30 5.55 5.70 5.79 5.84 5.89 5.95 6.05 6.15 6.30 6.41 6.56 6.70 6.80 6.91 6.96 7.00 7.04 7.08 7.17 7.32 7.37 7.47 7.49 7.56 7.68 8.80 68.2 48.8 33.5 30.1 32.0 35.5 47.7 50.6 49.3 48.6 37.1 17.0 7.5 4.7 8.0 77.3 71.3 65.6 60.0 62.2 60.7 56.2 50.7 48.9 49.4 41.7 57.3 39.4 23.3 19.0 20.4 20.0 35.2 44.6 44.3 44.0 37.8 23.6 12.3 2.9 3.3 59.3 46.5 36.9 32.9 32.0 31.0 35.4 35.4 36.0 38.5 37.5 20.4 9.3 2.6 59.8 49.7 42.0 41.4 40.0 41.0 43.6 45.4 46.4 44.9 39.1 23.0 10.4 3.9 4.1 60.8 51.3 43.9 40.0 41.2 40.8 42.3 44.7 46.2 46.4 66.3 54.5 47.2 45.8 47.2 51.4 53.2 54.0 54.7 56.8 50.7 32.2 16.8 6.6 6.8 69.8 69.4 67.4 66.2 64.0 52.2 61.0 " '72.8 32 34 36 39 41 46 47 49 2 51 52 53 54 55 57 59 65.1 55.0 34.4 22.6 57.1 40.0 12.4 5.8 16.1 21.6 76.2 73.8 57.6 34.2 21.8 25.7 25.8 10.0 3.7 39.8 8.3 3.1 55.7 51.6 4.5 6.0 9.6 12.6 16.0 15.5 14.1 13.5 13.8 19.3 22.9 21.2 19.3 10.0 5.8 7.1 8.9 8.3 5.6 4.9 9.0 10.8 12.7 9.2 8.9 11.2 13.0 8.3 6.0 6.4 8.6 11.7 10.0 1 7.2 6.9 11.3 14.3 9.6 6.4 13.6 17.4 19.9 15.8 14.7 15.6 25.5 27.8 23.8 1 18.5 42.3 29.5 15.6 11.1 8.4 9.3 63.7 65.4 14.0 9.2 5.0 5.2 8.6 15.5 3 1 59.8 49.3 37.5 23.7 12.9 44.0 2 4 6 9 12 55.8 20.4 21.8 21.7 12.7 11.7 10.9 13.9 14.9 60.3 1 47.7 152 INFRA-RED ABSORPTION SPECTRA. Table VIII a. — Observed Transmission, Using Small Spectrometer, and Prism AT Variable Deviation, as Explained in the Text — ^Continued. pectrometer s e t - tings. Angular ro- tation from the sodium lines. i i K en W 0 0 0 0 XI k' a a ^a ^a a 0! a ">■ CI
  • 0 .s a a M ¥ .Scj Co ta a X S 1 a. cc 0 11 a li 3 ^ 0 1 0) . m iilz; 2 II gen >. '2 'A a ^ a II 9 ^ 1 L i4 So a a t) 0 || a u 0 T u Si 5'^ 0 w 0 w s» 0 0 c 1 0 K 0 a u 3 0 H "5 Thickness of \ cell in mm. j 0.01 O.OI 0.01 0.01 0.01 0.01 O.OI 0.01 0.01 0.01 0.01 O.OI O.OI O.OI 1.70 2.05 2.46 2.65 2.73 2.88 3.04 3.21 3.30 3.38 3.46 3.53 3.60 3.67 3.71 3.88 3.96 4.18 4.33 4.46 4.65 4.85 4.98 5.10 5.28 5 45 Per cent transmission. 1 30 35 40 42 43 45 47 49 1 50 76.0 75.0 69.4 '51.0' 32.0 27.0 36.8 90.0 90.2 89.0 86.5 73.7 65.4 58.4 47.2 86.3 85.0 81.5 80.0 76.1 72.4 1 79.4 80.8 78.0 72.0 72.0 66.7 65.3 65.0 68.0 66.3 63.8 80.0 79.0 76.6 74.7 74.4 73.0 71.0 66.3 77.0 74.6 71.4 70.8 72.3 73.8 71.0 64.4 'golo' 89.0 53.0 55.0 53.4 77.0 80.0 79.3 79.2 79.6 79.6 85.3 83.6 82.5 "ssis 86.0 87.5 79.5 1 80.5 65.3 46.3 61.0 50.2 40.3 36.9 74.5 73.7 73.5 66.8 89.4 89.8 87.5 80.6 42.7 21.2 15.2 10.9 76.1 76.0 75.9 65.1 79.2 76.2 75.3 70.1 82.6 81.6 76.8 62.8 44.8 43.3 45.8 51.5 '72!5' ""87!2 84.8 72.3 51 52 53 54 49.7 'is.8 36.8 40.0 '58!2' 26.4 25.8 29.6 38.3 45.3 50.8 60.0 58.3 59.5 62.4 71.1 53.7 50.6 47.7 47.6 50.8 55.3 50.0 51.8 56.7 65.0 48.5 45.0 44.5 44.5 51.0 77.3 14.9 44.2 41.7 47.4 50.3 42.4 36.8 37.8 42.3 46.8 42.5 76.4 22.1 45.7 51.4 55 50.4 62.7 79.7 34.2 60.3 68.0 56 58 59 2 2 4 6 9 12 14 16 19 22 24 27 66.0 65.8 'm.7 'ei'.i' '65.2 79.2 80.6 84.8 85.0 83.0 80.5 81.6 80.5 70.7 74.8 '76!8' 'soio' 71.3 70.0 68.4 68.8 73.5 74.7 75.3 75.0 77.6 74.7 67.8 51.8 59.6 72.4 77.0 80.0 59.2 64.2 62.6 44.5 21.1 3.5 3.3 12.0 27.4 52.8 76.3 79.4 79.3 79.4 72.6 57.6 63.5 71.8 78.9 78.3 60.7 66.8 65.5 50.7 23.4 3.3 0.4 4.3 22.0 48.3 79.7 78.2 78.7 71.8 49.3 12.9 1.0 5.4 21.8 52.7 56 ! 3' 78.3 82.2 62.3 76.1 81.0 84.4 84.3 85.5 80.5 86.4 62.0 82.2 81.0 89.0 65.7 82.4 84.0 85.6 89.0 75.8 72.6 65.7 81.8 82.1 88.5 75.0 67.9 66.4 65.0 79.2 78.6 79.0 77.4 77.4 79.1 75.8 86.0 5.55 5.70 5.79 5.89 5.95 6.05 6.15 6.30 6.41 6.46 6.56 6.70 6.80 6.91 6.96 7.04 7.08 7.12 7.17 7.27 7.41 7.49 7.56 7.68 7.80 7.88 7.95 8.08 8.19 8.27 8.35 8.46 8.56 71.0 68.7 67.4 63.9 67.5 64.2 59.3 79.8 71.2 79.0 78.0 79.0 75.1 76.0 64.2 62.2 60.8 76.8 65.2 46.2 74.3 76.0 79.3 76.0 79.7 29 31 32 34 36 39 2 41 42 49.7 69.2 78.4 66.2 70.4 78.6 84.0 25.3 13.6 5.7 12.5 21.2 60.9 34.2 22.4 16.8 24.2 53.0 43.6 15.0 9.1 14.6 14.4 11.5 4.7 7.4 12.2 39.8 21.1 15.1 25.0 32.8 76.4 74.3 70.8 69.6 68.8 67.0 66.0 63.8 57.0 53.5 76.8 76.4 72.8 72.8 72.5 69.5 69.6 68.1 65.5 63.8 '73!4' 59.0 46.0 51.6 60.5 60.2 57.2 56.2 53.0 77.3 74.2 65.3 62.4 64.8 12.5 15.5 25.0 2:^.4 32.0 78.2 76.5 62.8 55.3 61.3 61.5 51.1 31.4 24.4 25.9 31.1 46.2 79.8 75.0 58.7 45.5 51.8 52.7 44 47 49 51 52 54 55 56 57 59 3 2 4 6 9 12 14 16 19 22 24 26 29 32 20.0 8.9 19.7 '43!7' 'bz'.-l' 14.2 7.1 14.9 22.6 21.5 10.0 11.9 '23 .'4' 33.3 61.0 62.3 56.9 39.9 34.7 19.4 17.6 44.8 59.3 34.3 24.6 'n.'6 13.0 72.5 69.7 55.3 35.0 27.4 25.4 "36 ! 6' 42.6 41.8 51.6 60.8 58.1 54.9 38.6 19.3 67.7 45.6 37.1 51.3 45.5 27.8 17.0 18.2 59.0 42.3 31.2 27.6 30.0 41.0 58.0 67.2 55.6 35.8 50.2 56.2 25.0 24.5 28.6 32.0 25.2 20.0 26.6 '31 ie' 28.8 18.7 13.4 9.2 26.3 44.3 54.4 72.8 27.3 37.7 82.4 "82!6' '56.0 39.1 22.0 n.Y 8.0 4.6 9.7 30.4 ■3.5;4 50.3 59.7 47.8 26.0 10.4 15.8 24.0 15.8 15.4 32.2 55.0 59.3 39.4 30.4 9.1 6.1 12.2 29.6 39.4 42.4 42.5 25.0 9.6 10.5 12.7 11.8 11.4 46.0 49.6 '56!i' '21.5' 11.8 19.5 31.2 42.7 25.5 42.9 59.2 59.3 47.8 34.6 58.3 66.8 11.0 15.0 '42.3 27.4 27.2 57.9 61.0 62.0 65.3 69.0 70.7 34.7 32.2 41.2 50.8 51.8 45.7 41.7 52.2 51.0 33.4 18.8 30.4 33.9 59.2 59.8 66.8 73.0 76.4 76.0 80.0 83.8 81.3 77.3 81.5 80.1 78.0 75.2 66.7 48.7 49.3 54.6 84.3 36.4 67.8 73.8 54.0 '57!4' 65.2 46.2 25.9 15.8 27.6 76.4 42.2 78.3 68.0 46.8 73.8 69..8 63.4 69.6 71.5 79.2 81.8 73.2 57.7 59.3 69.7 50.0 84.5 53.7 77.0 50.7 62.2 58.0 78.0 50.3 84.2 26.0 37.1 1 .58.1 1 51.7 76.8 47.5 78.1 51.5 81.6 51.3 74.0 62.5 84.0 80.8 154 INFRA-RED ABSORPTION SPECTRA. Table VIII b. — Observed Tkansmission Through Various Carbohydrates, Using the Smale Spectrometer^ and the Prism at Variable Deviation, I. E., " Angle of Constant Emergence/' as Explained in the Text. — Cont'd. Spectrometer set- ings. Angular rotations from the D lines. 0) be i 1 o V a < in V a S 'A w .u "0 !S S 1 Id 2 0 II X 0 £• II ^ H «v n II W 0. II •5 '^ 0 1 •M a w 2u a a V 0 g u 0 T w 0 V 1 0 6 K 0 W .0 "3 0 a 3 u w a u 3 0 M >> V Thickn cell in ess of } mm. i 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.0 1 0.01 0.01 0.01 0.01 0.01 0.01 8.63 8.70 8.80 8.88 8.97 9.07 9.13 9.19 9.29 9.35 9.45 9.54 9.60 9.66 9.76 9.86 9.92 9.97 10.06 10.15 10.21 10.35 10.42 10.48 10.52 10.61 10.70 10.75 10.83 10.90 11.00 11.14 11.25 11.41 11.53 11.65 11.80 11.92 12.03 12.18 12.30 12.42 12.54 12.66 12.80 12.93 13.00 13.17 13.40 13.51 13.62 13.72 13.85 19.96 14.06 14.26 14.45 14.63 14.82 Per cent transmission. O ' 3 34 36 19.3 24.6 9.3 10.0 13.9 41.4 38.0 47.2 76.4 44.7 78.3 48.3 71.4 52.6 62.8 64.7 71.6 26.0 32.8 49.7 61.6 60.8 54.6 51.8 50.8 54.8 59.0 49.6 33.3 41.3 70.0 73.8 74.4 78.3 81.8 81.6 39 41 34.2 29.6 77.0 52.5 78.3 48.0 70.7 48.7 44 47 33.7 47.2 31.6 63.0 77.0 55.6 78.4 30.0 33.6 40.0 42.3 39.0 30.0 23.0 79.4 45.5 63.7 75.6 71.8 68.7 56.0 44.5 51.6 64.7 60.3 53.0 81.4 49 51 33.0 61.2 52.8 50.8 45.4 50.0 76.5 46.7 77.0 80.8 43.0 61.0 81.1 54 56 31.2 '21.6" 36.6 20.0 31.8 40.6 30.0 22.3 29.3 34.2 76.3 31.0 74.6 68.7 23.0 69.0 81.0 59 4 2 52.2 46.3 42.2 75.0 42.0 46.2 64.6 7.5 66.4 77.0 4 6 24.6 69.5 64.6 51.1 58.7 33.2 80.7 20.5 52.3 54.0 9 12 23.3 'ss.'g' 24.4 24.0 28.8 32.2 33.8 60.5 59.6 66.5 55.5 52.5 39.0 55.0 59.8 51.7 36.0 44.5 60.0 55.0 72.8 56.3 84.0 24.0 34.4 50.6 45.2 61.4 68.2 51.0 14 16 45.0 55.2 54.6 73.7 59.8 75.0 30.8 52.5 47.8 47.0 52.7 68.5 19 22 48.7 28.3 53.6 70.0 59.8 69.2 34.2 59.8 74.3 75.5 24 29 4 32 51.5 61.6 44.5 62.6 53.7 57.8 17.0 22.5 65.0 67.3 56.0 25.0 35.0 48.0 61.8 65.6 71.0 75.6 37.2 44.0 64.7 70.0 58.4 66.6 74.2 78.0 76.3 82.0 34 36 35.0 65.6 34.3 64.5 55.0 68.0 53.3 66.4 51.8 73.3 69.0 79.0 85.8 39 42 33.8 71.2 22.2 62.5 67.8 72.0 68.6 10.6 24.7 56.3 75.7 71.0 80.0 86.0 44 47 26.7 71.2 55.8 56.2 68.7 75.8 21.9 4.7 58.6 75.0 68.8 81.5 80.0 49 54 28.7 25.8 14.2 3.5 4.0 20.0 21.8 72.0 72.5 71.8 57.3 33.4 29.5 55.0 70.0 75.8 77.5 74.0 47.0 57.1 80.2 77.5 73.8 53.8 56.0 52.4 40.7 36.5 34.5 36.4 38.0 27.5 69.0 74.3 83.0 55.0 65.8 65.8 14.8 32.9 53.8 58.8 "66.'6' 68.8 65.0 69.8 66.7 74.2 74.3 83.3 83.5 74.2 78.0 83.0 87.2 88.5 'ss.'s' 79.8 88.0 90.0 90.0 82.5 82.0 81.3 59 5 4 70.0 70.0 83.5 9 14 73.8 70.0 83.5 76.8 81.8 62.0 75.8 69.6 66.7 70.6 70.0 68.6 67.6 58.8 26.7 0.0 2.0 20.0 50.0 66.5 74.4 78.7 82.6 60.8 10.0 29.0 84.0 19 24 81.0 72.8 83.0 78.2 81.2 64.0 29 34 18.3 65.8 82.7 73.0 81.6 79.3 66.5 39 44 4.1 ' 0.0' 60.6 58.0 60.5 73.8 '73.0 6.5 0.0 10.0 73.5 74.3 75.1 73.5 62.1 58.0 54.3 64.7 "43!6 49 54 65.2 71.5 70.6 57.5 80.0 6 00 5 0.0 50.5 66.0 25.1 66.7 71.0 50.2 64.5 53.3 50.0 73.7 68.3 56.0 60.7 67.8 75.8 81.5 78.0 10 15 41.7 70.0 25.0 77.0 72.0 28.0 66.0 65.0 53.4 56.7 70.0 73.0 20 30 35 0.0 0.0 6.0 0.0 25.0 0.0 20.0 10.0 80.0 80.0 71.0 72.0 36.0 59.0 78.0 78.3 31.0 9.0 51.8 50.0 18.3 0.0 67.5 47.3 85.0 88.0 40 45 13.0 16.0 0.0 81.5 71.5 70.1 75.0 36.0 44.0 3.0 70.4 18.0 80.5 50 55 33.0 50.0 17.0 80.0 72.5 72.0 60.0 50.0 2.0 42.0 35.0 0.0 82.0 7 00 10 20 0.0 39.0 20.0 '15.0' 30.0 45.0 34.0 0.0 9.0 50.0 36.0 35.0 81.0 81.5 72.5 70.0 72.0 73.0 59.0 59.0 64.0 15.0 43.0 58.0 63.0 75.0 64.0 47.0 67.0 70.5 24.0 43.5 20.0 00.0 83.0 84.0 80.0 30 78.0 40 TABLE VIII B. DO Table VIII b. — Observed Transmission Through Various Carbohydrates, Using the Small Spectrometer, and the Prism at Variable Deviation, L E., " Angle of Constant Emergence," as Explained in the Text. — Cont'd. c s - o V a K 0 K tjrj u "o CI s d 0 (J "o >> d W 0 0 0 0 CS > « 0 i K >. V 2 "5 4-1 si ao 0 1 0 ."H'o' "a > s u 0 % 0 W to CS 0 s 0. 0 0 0 "0 ft C2 0 0 M w 0 ■6 '3 0 0 'v 0 C8 in 0 ;.< .So "u 1- w 2:l| ft Thickn cell in ess of \ mm. J 0.0 1 0.01 0.01 0.01 0.01 0.01 0.01 0.0 1 0.01 0.0 1 0.01 0.01 0.15 0.035 1^ 8.70 8.80 8.88 8.97 9.07 9.13 9.19 9.29 9.35 9.45 9.54 9.60 9.66 9.76 9.86 9.92 9.97 10.06 15.15 10.21 10.35 10.42 10.48 10.52 10.61 10.70 10.75 10.83 10.90 11.00 11.14 11.25 11.41 11.53 11.65 11.80 11.92 12.03 12.18 12.30 12.42 12.54 12.66 12.80 12.93 13.00 13.17 13.40 13.51 13.62 13.72 13.85 13.96 14.06 14.26 14.45 14.63 14.83 Per cent transmission. 8 36 52.6 52.3 72.5 73.3 75.0 75.0 '59! 6' 43.4 45.5 '64.6' 56.9 50.8 60.0 61.8 6.3 12.1 23.4 24.8 25.2 25.0 '27 .'6' 13.8 18.8 11.9 7.8 3.0 7.4 24.0 33.0 30.0 24.0 24.6 30.0 1.8 5.0 7.4 6.0 7.2 6.3 8.2 4.2 2.7 2.7 5.0 7.1 39 41 65.2 42.6 38.4 33.7 55.0 72.7 66.0 44 47 2.0 40.0 46.7 28.7 20.5 24.3 36.6 44.1 50.7 35.3 18.2 22.5 31.6 45.6 '53! 6' 22.2 13.7 18.4 35.1 47.7 52.2 56.4 56.7 56.9 59.8 59.7 59.6 76.3 76.5 77.0 70.3 49 51 0.0 31.3 61.1 1.0 54 56 6.1 '31.9 10.0 16.5 40.0 80.0 29.3 0.0 9.3 5.7 3.6 59 4 2 46.0 45.8 39.7 'si'.s 56.0 59.6 51.0 25.6 35.0 54.3 72 2 79! 7 77.5 75.8 78.0 48.0 20.0 20.4 29.4 44.6 '53 .'6' 35.4 31.8 38.0 43.6 51.3 65.4 69.8 64.7 67.5 30.0 '47!2' 62.0 81.5 31.4 4.3 4 6 53.7 65.2 58.4 57.8 58.3 54.0 51.3 51.0 53.4 48.4 47.2 65.6 79.8 45.8 19.0 11.5 9 12 59.8 58.4 55.0 52.3 58.0 35.6 61.8 53.0 60.5 56.4 66.0 77.0 48.8 20.0 17.0 14 16 32.7 25.4 15.2 66.7 59.0 60.4 50.3 63.2 72.9 47.4 14.0 20.8 19 22 64.3 65.2 65.4 35.0 74.3 54.9 41.8 61.6 69.6 31.4 15.3 19.4 24 29 4 32 62.5 61.0 49.2 31.0 16.7 21.0 38.0 57.8 69.0 72.0 75.0 81.0 76.0 66.5 55.0 48.0 C6.0 78.0 65.5 36.5 13.2 28.5 61.5 73.5 26.9 47.0 44.5 40.7 23.3 16.0 12.5 23.8 78.5 79.0 45.0 25.2 "ie.'g' 31.6 19.4 56.5 43.3 62.3 52.2 7.6 11.7 39.8 54.0 15.5 18.6 34 36 8.8 76.3 15.5 8.8 9.3 35.8 44.4 46.7 59.0 19.8 39 42 2.5 72.7 8.2 0.0 5.6 30.7 34.5 63.4 55.8 20.0 44 47 83.0 71.5 14.3 80.3 0.0 0.0 4.9 27.3 40.0 67.8 50.0 23.5 25.0 "28!6 25.0 22.8 19 8 49 54 59 5 4 84.0 86.5 83.0 78.4 82.6 84.5 82.0 82.0 81.0 84.0 80.6 75.5' 75.2 75.0 72.4 69.2 71.0 74.3 76.7 74.9 74.5 75.8 71.0 59.4 52.2 42.7 55.8 60.0 47.0 17.2 9.1 16.0 44.0 48.0 28.5 14.5 43.7 55.0 61.0 67.5 '35! 6" 0.0 10.0 21.0 81.0 78.0 76.0 77.2 80.0 "83.'6' 'ii'.o 7.7 12.5 24.6 30.0 30.7 37.0 40.0 39.4 39.1 0.0 6.6 19.5 7.9 10.0 21.7 33.0 40.2 45.3 39.0 39.2 42.8 54.0 65.0 69.0 69.0 74.8 79.0 72.0 76.0 78.0 'so.'s' 42.8 44.8 50.0 55.0 63.5 9 14 27.8 25.1 23.6 26.5 31.7 46.5 45.4 41.3 40.0 25.6 27.3 28.6 34.0 35.7 32.3 31.0 33.0 38.0 60.0 '66!6' 60.0 60.5 64.0 66.7 67.2 68.0 19 24 29 34 81.0 '86!4" 60.0 63.0 64.8 18.4 12.6 7.7 5 5 39 44 74.0 66.5 64.5 72.5 38.6 39.2 40.0 83.4 81.2 39.7 3.5 6 4 49 54 22.5 82.8 86.9 72.0 14.2 13.5 6 00 5 75.5 49.7 56.8 83.0 76.6 47.5 44.5 48.0 67.0 84.5 20.0 19.3 25.0 10 15 56.0 23.7 80.0 53.3 83.0 66.0 51.5 50.0 50.0 64.0 82.0 0.0 33.0 28.0 20 30 35 20.0 4.0 0.0 0.0 80.0 29.1 75.0 14.2 59.0 74.0 30.5 50.0 75.0 60.0 61.5 47.5 33.4 50.5 46.6 50.0 50.0 •61.0 59.2 87.5 75.1 25.0 50.0 48.5 13.0 29.5 35.0 40 45 30.0 22.0 54.0 10.0 70.0 33.0 29.4 51.0 50.0 50.5 51.0 '56.'6' '51 !6' 48.0 63.0 63.6 7.1 22.0 50 55 45.0 38.0 71.0 82.0 84.0 80.0 75.0 73.0 26.0 84.0 75.0 44.3 18.3 30.4 27.8 66.7 30.0 20.0 7 00 10 20 8.0 0.0 0.0 0.0 46.0 27.0 33.0 20.5 30.0 30.0 67.0 60.0 70.0 80.0 72.5 68.0 50.0 50.0 36.0 50.0 52.7 46.0 25.0 35.2 47.0 47.0 37.0 56.0 73.0 70.0 79.0 79.0 45.0 50.0 47.0 33.0 50.0 30 78.0 50.0 40 1 1 TABLE VIII C. 157 Table VIII c— Observed Transmission, Using Small Spectrometer, and Prism AT Variable Deviation, as Explained in the Text. meter s e t - Angular on from the ium lines. to M a si 0 2 w 0 1 0 u 1) 1/ — 1 0 0 a "3 0 li 2 1 > a a >> a '0. )> 0 0 2 5 !* ^ 2 <5 w CJ fH 04 0 0 H Thickne cell in ss of \ 777/5. ) 0.01 0.01 (a) O.OI 0.01 0.16 0.01 0.01 0.01 6 0.04 0.01 0.01 00. 1 0.01 0.01 1 30 35 40 42 44 46 48 1 50 51 52 53 54 55 56 57 58 2 00 3 5 10 13 15 17 2 20 23 25 28 2 30 32 33 35 37 40 42 43 45 48 2 50 52 53 55 56 58 3 00 3 5 7 10 13 15 17 3 20 23 25 27 30 Per cent trans7nis3ion. 1.70 2.05 2.46 2.65 2.81 2.96 3.13 3.30 90.0 87.5 84.4 8.75 88.3 84.7 1 89.0 87.0 75.4 94.0 91.0 89.9 86.7 90.5 88.9 85.1 81.2 83.6 85.7 81^0 79.6 .073 78.0 79.0 76.4 75.3 73.5 89.8 87.4 85.6 88.0 81.3 89.3 71.0 62.5 79.0 81.0 87.0 89.4 78.7 79.3 76.0 73.8 61.0 42.7 88.0 84.5 84.5 80.0 71.0 60.4 86.4 '75!8 62.7 60.8 55.4 38.4 26.7 82.0 78.0 70.0 62.4 iiiiii 81.0 64.7 39.6 42.4 56.3 80.0 78.7 72.1 60.0 '48!6' 49.1 'is.8 'so.o 77.3 73.2 66.2 47.6 42.8 41.3 43.7 50.3 46.8 37.7 31.4 32.2 29!8 31.2 44.9 83.3 81.2 62.7 44.4 39.0 42.2 50.0 67.7 88.0 82.8 67.8 50.3 '■:&.8 62.0 54.7 60.8 3.38 3.46 3.53 3.60 59!3 64.7 59.5 63.8 24.2 66.4 :::::"! 68.7 43.5 49.1 68.3 37.2 "73.5 35.6 77.4 74.3 "le.s 88.9 90.0 88.3 91.0 90.3 89.4 82.6 75.7 80.0 86.6 76.8 77.6 58.0 77.6 3.67 3.74 3.81 3.88 4.04 57.3 79.5 63.4 53.5 80.8 75.8 78.7 75.0 77.4 86.1 'ss'.s" 83.3 83.0 83.0 86.3 64.3 66.0 68.0 71.8 80.5 76.3 80.5 77.6 82.3 m.o 87.3 87.1 75.6 62.1 55.8 57.7 62.2 59.8 44.5 22.0 19.5 68.1 71.3 67.1 69.0 75.3 78.5 84.7 84.6 80.0 79.8 81.0 84.7 79.8 80-2 4.40 64.5 54.7 62.0 68.0 79.7 'sois" 80.4 81.2 80.1 75.6 85.4 80.7 80.4 4.51 79.7 76.8 74.4 74.8 75.7 85.0 84 6 4.72 4.92 5.04 5.16 5.34 5.50 5.60 5.75 5.84 76.4 84.8 85.4 87.4 '87!o" 73.0 62.4 38.2 14.1 3.5 1.3 85.3 83.8 84.0 85.6 79.8 73.8 81.4 78.9 76.6 78.3 79.1 79.3 79.7 82.5 80.0 77.3 '79.0 "78!9 80.6 72!3 '74.3 ■75!6' 74.0 70.0 87.2 "ssis 81.8 78.0 62.0 32.0 15.0 5.1 5.4 83.6 'si!? 'soia' 73.5 '■rlis' '70.1 77.0 69.6 65.6 69.0 72.8 82.7 83.8 86.8 84.9 80.5 79.7 78.8 72.7 '72!3" '65!2' 78.5 '74.2 72.0 69.3 71.5 71.5 69.4 "68.3 6.00 6.10 6.20 6.36 6.46 53.7 16.8 6.0 16.4 25.7 27.6 20.0 14.8 78.9 78.3 77.0 77.2 77.6 69.0 66.7 61.8 44.0 27.3 77.3 76.9 75.8 72.6 73.4 83.9 82.0 77.3 74.0 80.4 82.6 75.0 63.0 63.5 66.3 74.9 73.2 73.6 71.9 61.3 50.8 31.0 11.4 5.4 4.3 64.8 62.8 65.3 67.8 72.5 77.8 68.3 61.2 62.4 64.3 ■^:2' 43.4 48.5 46.7 72.3 64.3 63.5 60.7 46.3 52.0 28.0 13.5 9.2 3.1 6.60 6.75 6.86 6.96 '"i!6" 0.2 0.0 "6!o' 0.0 0.0 0.0 11.5 25.8 34.0 41.7 'wis' "■rsio' 7.8 13.8 11.4 6.3 3.6 "eii 4.0 2.8 11.6 17.3 '33!4' 27.9 25.0 26.3 42.3 11.8 8.5 7.3 6.3 5.2 5.9 "5.7 4.9 7.3 8.0 10.0 18.6 '32.2 14.4 8.5 9.7 74.3 58.4 42.3 20.8 19.8 34.0 22.5 49.8 61.2 77.3 76.3 74.9 75.5 26.4 27.8 23.9 19.7 42.3 34.8 26.8 26.9 34.1 41.4 47.2 46.3 42.8 53.3 61.1 66.4 64.8 66.8 71.8 '76!i 75.0 71.2 73.4 75.0 1.3 2.3 3.8 7.2 n'.'s 67.7 40.8 23.7 'nis" 24.4 47.7 21.2 8.2 20.0 25.9 15.3 16.2 26.9 27.1 23.6 22.3 22.0 1.6 10.4 25.4 7.00 7.08 70.0 76.8 74.0 17.1 17.9 23.4 31.6 48.6 54.3 54.4 51.8 47.4 49.5 50.2 48.6 44.3 41.8 43.3 45.5 71.7 'so'.s" 82.0 83.4 83.5 'si'.e '79.6 82.0 87.5 '82!6" 71.4 42.3 40.8 36.5 36.4 7.12 7.22 7.32 7.45 7.52 m.7 68.0 72.8 72.7 7l!8 61.2 37.7 24.7 16.0 22.5 40.8 36.3 24.6 11.4 6.4 3.4 "sis 1.7 2.0 6.3 7.6 8.7 6.2 3.3 1.4 1.4 5.8 22.1 12.6 20.5 26.3 33.7 37.3 37!6 '28.2" 22.9 25.5 32.3 30.0 49.7 53.7 '76!2 77.5 79.0 79.0 78.2 79.0 79.0 47.2 51.4 'n.8 45.3 51.4 15.7 70.5 '72!6 *72!i 63.7 59.0 58.8 59.0 65.0 68.0 67.8 54.0 50.3 55.7 71.3 71.8 71.7 66.3 66.8 67.0 69.3 'm.3 6.0 ""s'.o 7.59 7.72 7.84 7.91 79.0 81.0 'siio 25.0 "'44.4 8.00 8.12 79.0 50.0 8.23 8.31 79.5 77.0 66.2 69.8 57.8 8.50 '&4.3 ■7i:6" isis 79.5 77.8 73.6 43.3 a Examined as a vapor in the gas-cell 5.7 cm. in length, vapor only, divide each number by 70 per cent. 6 Solid film. To obtain the transmission through the iS8 INFRA-RED ABSORPTIOX SPECTRA. Table VIII c. — Obsertcd Transmission, Using Small Spectrometer, and Prism AT Variable Deviation, as Explained in the Text — Continued. Spectrometer set- tings. Angular rotation from the D sodium lines. 01 M 0 i si O F a o u 0 , u K a 0 N a m IS 0 (J u a u a >. 0 a CO 0 6 i "o il a g. u {-1 w 0 affi 0 u 0 0 Id V 0 'Z if N a s 2 Thickness of\ cell of mm. J 0.01 0.0 1 (a) 0.01 0.0 1 0.16 0.01 0.01 0.01 0,04 0.01 0.01 0.01 0:01 0.01 8.59 8.66 8.73 8.83 8.90 9.00 9.10 9.16 9.23 9.32 9.38 9.48 9.63 9.69 9.80 9.89 9.95 10.00 10.09 10.18 10.23 10.29 10.37 10.45 10.50 10.55 10.64 10.73 10.77 10.86 10.92 11.03 11.17 11.30 11.44 11.55 11.68 11.82 11.94 12.06 12.20 12.32 12.45 12.56 12.66 12.80 13.93 13.17 18.40 13.62 13.72 13.85 13.96 14.06 14.26 14.45 14.63 14.82 15.00 15.20 15.37 15.74 Per cent transmission. 3 33 58.7 57.0 'eiis' 'm.'i' 30.9 48.6 57.5 63.1 73.2 73.8 59!6' 43.8 33.7 25.8 20.0 22.5 33.3 72.6 "si. 3 72.1 72.3 74.6 71.5 4.3 3.3 3.6 7.2 15.8 24.0 37.0 40.6 45.6 34.0 19.6 15.8 27.0 15.4 8.0 5.0 8.6 io!?' 30.0 33.0 34.8 35 37 75.3 72.8 71.6 3 40 42 47.5 35.1 43.3 73.4 39.0 69.0 60.5 63.0 77.8 79.5 74.0 '7i!3' 6.95 55.7 40.0 45.5 42.7 21.8 45 48 18.2 20.0 28.8 60.4 54.6 47.1 49.4 57.5 69.5 26.8 88.0 60.0 58.3 58.0 50 52 5.8 23.3 27.3 70.0 37.6 38.7 19.7 8.8 5.8 2.0 8.2 77.3 57.7 51.0 12.8 55 57 28.0 61.0 32.3 70.0 63.8 65.6 64.5 76.6 84.2 70.8 64.0 76.0 80.0 63.6 46!7 53.6 43.4 'ssie" 58.3 81.8 67.7 6.10 56.0 71.4 65.6 4 00 5 50.0 68.2 75.0 76.0 45.7 47.8 70.0 73.6 70.3 71.0 70.4 68.2 61.2 46.3 50.6 30.3 (d) 76.8 63.44 39.4 42.8 57.0 58.7 61.0 66.7 67.0 10 13 83.8 73.0 50.8 68.9 67.7 53.4 10.0 4.2 0.0 59.8 37.5 38.0 38.2 42.0 47.2 47.2 45.6 51.8 44.5 15 17 82.8 63.1 52.7 70.7 60.0 47.0 46.8 55.6 66.3 61.0 4 20 23 85.3 '87!8' 37.8 52.5 67.8 50.0 79.5 56.5 56.5 60.0 42.2 22.7 82.5 76.7 44.2 55.0 71.4 68.8 25 27 47.9 85.7 33.5 47.0 48.8 56.0 83.0 76.5 47.5 43.3 70.0 75.4 30 33 85.0 84.5 44.0 86.6 65.0 39.3 38.0 51.5 22.8 13.1) 33.7/ 10.0 84.6 75.5 47.8 45.8 31.4 28.4 74.5 76.0 67.4 35 37 80.0 85.5 43.3 84.3 73.3 63.8 86.8 74.0 48.2 77.8 79.5 68.0 55.5 44.3 39.0 54.6 4 40 43 80.0 82.3 39.3 78.8 76.0 (c) 67.2 84.6 72.3 50.7 29.4 27.8 19.4 79.4 80.5 45 48 66.0 76.5 29.2 77.0 67.3 63.5 75.5 56.0 80.5 82.4 50 55 39.5 63.0 84.5 "ss.'e" 86.0' 58.0 60.5 69.0 35.7 14.0 25.0 52.0 21.5 24.1 25.4 26.0 32.6 77.3 56.0 50.0 67.0 78.0 76.5 2.5 0.0 4.5 66.7 61.7 57.0 45.8 29.0 33.0 40.1 40.6 40.0 40.0 39.7 64.0 80.0 83.0 84^0 77.4 '77!2' 77.3 77.0 64.0 68.0 69.0 63.2 54.5 50.0 45.0 47.0 37.5 25.6 16.8 12.0 10.0 9.0 10.0 6.5 18.3 33.4 44.6 59.4 82.0 83.3 73.8 76.4 77.2 73.0 65.0 35.0 5.5 14.3 43.5 60.5 64.7 56.3 35.3 1.4 30.3 64.5 65.7 68.7 66.5 63.7 5 00 5 83.0 83.0 77.5 60.0 41.8 70.8 80.0 83.3 83 !6" 84!6* 10 15 69.6 18.8 34.0 41.8 20 25 61.2 68.8 71.5 52.0 17.0 25.0 70.4 66.7 83.5 77.5 66.7 51.8 38.0 58.0 66.7 60.0 40.0 44.7 55.5 53.0 5i.6 60.0 61.0 71.8 84.0 30 35 80.0 68.4 53.7 "XQ.'O 44.3 37.5 26.8 13.1 3.0 84.5 64.0 83.5 84.0 78.5 40 45 80.0 57.3 71.8 84.6 20.0 17.0 38.0 82.5 82!5" 82.0 70.1 isio 50 55 77.0 45.0 66.0 74.4 0.0 40.0 85.0 6 00 5 75.0 13.0 50.0 62.7 0.0 0.0 0.0 0.0 "k'.h' 40.0 61.0 27.0 78.0 62.0 82.3 50.0 5.2 29.0 55.1 73.0 80.0 10 20 30 40 45 73.0 71.5 69.8 67.0 84.0 77.5 80.0 85.0 54.1 51.3 60.5 70.0 42.0 40.0 33.0 50.0 68.0' 0.0 0.0 "6!o" 21.0 0.0 3.0 0.0 57.0 0.5 12.0 50.0 77.0 81.0 79.0 11.6 7.0 11.0 15.0 70.1 47.0 10.0 00.0 10.0 50.5 2i!4" 48.0 59.0 50 55 63.0 64.4 70.0 75.0 75.1 59.0 68.5 83.0 85.0 73.0 0.0 80.0 61.0 26.0 66.0 56.0 58.0 67.0 62.6 83.0 42.70 20.0 4.0 0.0 0.0 0.0 7.0 20.0 50.0 60.0 65.0 7 00 10 20 77.0 75.0 77.0 40.2 55.0 66.0 65.0 63.0 55.0 60.0 76.0 0.0 66.0 15.0 00.0 62.5 73.0 30.0 34.0 50.5 50.0 67.0 72.0 77.5 64.0 66.7 66.7 ss.'s 30 40 50 8 00 10 30 c Thinner cell for the following observations. dNew film. TABLE VIII D. 159 Table VIII d. — Obser\'Ed Traxsmissiox of Various Carbohydrates, Using the Large Spectrometer, and the Prisji at Variable Deviation, as Explained IN THE Text. [Thickness of cell o.oi mm., except for myricyl alcohol, which was a solid film between rock-salt plates, thickness less than o.oi mm..] Spectrometer s e t - tings. Angular de- viation from Na lines. a > 5 'J 3 a < w .a a %, u u •a . s u 2 I 1 u 5 0 u 0 0 a a V 3 "o a 11 = 1 'S 1 V '■J W .0 a« it's >.o M 6 u 0 cSl w 'C a 0 N a a 0 5 «/ a a < 0 K 0 W 0 0 a S 3 0 0 « >> M 1.46 1.50 1.55 1.60 1.65 1.71 1.75 1.85 1.90 1 96 Per cent trammission. 1 25 26 27 84.4 86.5 87.0 89.0 87.8 85.9 87.6 88.0 28 29 64.2 63 3 30 61 2 31 63 7 32 67 0 33 69 8 34 an 3 ■ 68.0 35 2.05 90.0 2.12 ; 87.8 2 18 ^^ ^ 69 3 36 78.0 77.8 75.2 74.3 77.6 80.2 81.0 81.3 89.2 72.3 37 67.2 38 2.28 2.37 2.46 2.55 2.64 2.73 2.81 2.88 2.96 3.04 3.13 3.21 3.30 3.38 3.46 3.53 3.60 3.67 3 74 79.4 80.7 83.5 85.3 87.6 60.7 39 59.3 1 40 61.2 41 67.5 42 84.5 81.4 76.0 43.8 29.3 49.4 63.6 62.5 41.2 24.3 32.8 63.7 78.5 75.9 75.0 75.5 75.0 74.8 73.5 71.0 67.5 47.6 26.0 14.8 21.2 31.2 42.7 60.8 70.0 43 69.8 54.7 16.4 4.7 14.0 33.1 27.3 32.7 54.5 75.3 '79!o' 67.3 40.0 28.5 19.6 18.1 42.7 68.0 66.3 44 85.0 82.7 82.8 81.3 75.4 59.0 53.3 43.3 47.5 56.0 64.3 67.3 71 fl 48.0 45 18.0 46 81.3 70.4 66.8 57.8 43.4 16.4 4.1 19.3 40.1 61.8 5.7 47 6.80 59.6 26.8 19.7 21.5 40.3 64.7 70.3 64.3 45.4 34.1 22.4 12.4 9.3 27.1 42.0 90.0 76.0 47.3 26.5 17.0 42.0 77.3 76.0 49.0 25.9 15.2 39.0 71.0 85.4 80.0 68.0 72.5 79.8 85.7 87.7 89.4 88.6 88.6 90.6 92.7 6.7 48 49 1 50 51 52 53 54 48.0 21.1 10.0 36.2 62.3 81.4 54.5 22.2 9.2 36.8 61.0 65.7 70.0 75.3 78.2 79.1 80.5 78.5 79.0 79.3 80.0 86.3 82.2 80.6 80.6 80.6 80.6 "hh'.z 40.4 12 7 11.6 43.6 58.4 20.8 28.6 20.7 10.0 4.9 13.6 33.8 55 46.7 56 48.1 49.5 58.0 68.8 69.8 85.3 'ssis" 90.5 88.0 90.0 90.5 89.0 85.3 86.8 87.8 'seio' 78.8 78.8 81.5 52.0 57 3.81 77.3 3.88 80.5 3.96 82.8 4.04 82.8 4.12 80.0 J. IS 87 Q 60.3 58 77.0 70.2 59 2 00 80.8 81.3 80.0 78.8 80.0 82.8 78.9 80.0 81.6 71.8 1 89.4 89.0 84.3 83.7 70.0 54.0 72.0 89.3 92.2 78.0 2 75.0 78.0 3 ( d •)« ^8 3 76.0 4 4.33 i -Vi 7 78.7 80.4 74.3 5 4.40 4.46 14.4 5.0 0.5 2.0 1.5 0.0 0.0 0.0 2.0 4.8 12.5 32.8 76.0 6 79.0 78.8 81.0 77.4 77.4 78.7 78.6 75.3 80.3 8 i 4 ?;7 81.0 9 4.65 4.72 4.79 4.85 4.92 4.98 5.04 5.10 5.16 5.23 5.29 5.34 5.39 5.45 5.50 5.55 79.0 79.5 2 10 81.3 11 12 81.0 91.8 81.3 81.4 13 14 87.5 69.3 71.3 58.3 50.7 61.8 69.3 66.8 63.2 61.8 59.2 64.3 15 88.0 86.7 84.2 82.2 80.5 16 81.2 83.4 81.9 79.4 79.0 83.5 80.1 "ik'.i' 74.5 73.3 75 4 17 80.2 18 59.7 19 2 20 66.4 81.1 84.7 79.8 21 22 75.2 88.3 85.0 81.0 80.5 23 24 84.8 '89.'6' 81.6 76.7 81.3 i6o INFRA-RED ABSORPTION SPECTRA. Table VIII d. — Observed Transmission of Various Carbohydrates, Using the Large Spectrometer, and the Prism at Variable Deviation, as Explained IN THE Text — Continued. Spectrometer s e t - tings. Angular de- viation from Na lines. s > 1 < i. S CI a u oi > CS V K o u' a CS 0 Q a a u a i Si's i o o p V O 6 X. u O o Ji'K a ts u Cli u W 0 a U X u .2 a a u V a '•5 < a .2 a 0 a 0 a u 0 a a s 0 a 0 So" ci •c >• 5.60 5.65 5.70 5.75 5.79 5.84 5.89 5.95 6.00 6.05 6.10 6.15 6.20 6.25 6.30 6.36 6.41 6.46 6.51 6.56 6.60 6.65 6.70 6.75 6.80 Per cent transmission. 2 25 79.0 70.5 62.8 43.3 15.0 3.3 2.0 12.5 40.0 66.0 76.8 65.5 63.3 72.2 64.7 55.6 34.2 14.3 1.0 0.6 0.0 14.0 20.0 37.0 26 78.3 63.0 46.0 45.2 56.8 60.0 62.5 55.4 51.3 59.8 73.8 75.5 75.8 75.7 74.8 73.7 72.0 66.0 61.0 64.5 66.0 68.7 59.5 44.8 30.3 23.6 16.7 9.1 0.0 9.4 14.3 87.6 76.0 73.6 79 8 27 C'.'.'.'.'. 1 1 1 75.0 69.4 68.3 73.0 72.0 75.1 77 8 28 85.5 84.9 72.5 74.6 29 73 5 2 30 86.8 87.3 71.0 57.3 43.3 16.0 9.0 0.0 2.0 7.0 '23.0' 41'. '7' 40.0 25.0 4.0 0.0 0.0 0.0 2.0 5.0 7.0 8.9 11.0 71 0 31 40.8 33.0 18.0 10.0 0.0 4.5 6.2 10.5 3i'.'5' 53.2 38.5 18.0 1.0 1.0 10.0 21.0 28.6 74 0 32 80.0 "78. '3' "72.0' 65.5 58.5 66.0 78.0 SS.V 'ei'.b' "56.0' 66.0 62.0 64.0 73.0 78.0 7'! 8 33 86.0 88.8 79.7 66.4 65.3 66.7 74.8 76.4 80.3 80.4 80.6 70.0 58.8 25.0 20.8 20.0 19.2 18.0 22.7 28.0 87.5 78.3 72.0 73.7 76.4 81.0 81.2 72.8 71.0 66.7 52.0 21.5 25.1 37.4 42.0 40.0 34.0 25.0 25.0 26.0 74.8 74 X 34 63.3 50.0 34.3 32.6 43.5 47.4 53.7 63.8 50.0 22.0 22.8 35.4 39.5 44.8 35 36 74.0 70.0 37 38 73.4 17.0 16.0 20.0 16.5 22.6 48.3 62.0 68.5 69.0 39 2 40 72.5 64.9 61.8 41 53.5 35.0 14.0 11.8 20.8 33.3 51.4 76.3 76.3 73.0 70.0 63.0 62.5 51.8 42.0 16.7 10.0 17.0 25.5 45.8 42 23.3 60.0 54 0 43 44 10.5 55.0 53 0 45 46 2.5 4.0 4.0 68.5 61.8 54.0 33.3 20.0 21.1 35.5 35.5 30. '5' 25.0 23'.'8' 32.5' 56". '2' 47 50.0 26.2 12.3 10.0 18.0 27.9 48.4 57.4 65.8 42.3 22.0 19.5 30.8 42.5 39 0 48 20 0 49 00 0 2 50 6.86 17.0 00 0 51 6.91 6.96 7.00 7.04 7.08 7.12 7.17 7.22 7.27 7.32 7.37 7.41 7.45 7.49 7.52 7.56 7.59 7.63 7.68 7.72 7.76 7.80 7.84 7.88 7.91 7.95 8.04 8.12 8.20 12 5 52 22.0 21 6 53 27 8 54 39.6 26.4 38.4 58.5 36.0 27.0 27.0 28.0 26 0 55 56 60.0 74.3 66.8 48.8 46.5 58.5 77.5 47.5 30.6 57 58 39.7 46.5 31.7 28.4 36.7 51.8 60.0 77.5 65.0 28.0 20 9 59 3 00 38.6 60.0 20 0 1 29.0 33 0 2 28.5 19.2 4.0 0.0 72.6 55.0 3 4 83.3 23.7 5 6 7.5 7 8 8.0 8.9 9 3 10 20.0 27.0 11 12 21.8 34.0 13 14 32.0 15 16 18 20 3 22 " " "1 1 1 TABLE VIII D. l6l Table VIII d. — Observed Transmission oe Various Carbohydrates, Using the Large Spectrometer, and the Prism at Variable Deviation, as Explained IN THE Text — Continued. [Thickness of cello.oi mm., except for myricyl alcohol, which was a solid film between rock-salt plates thickness less than o.oi mm.] Spectrometer set- tings. Angular de- viation from Na lines. a i) > 6 W .Si 2 "3 n O i -a •s 2 CS P4 B w o 'D C6 O "u O O S w 'o CS u o u V a 6 o U "2 o t a d i o 0 a a o O jj a V a d X o U o a g 3 w d w 0 0 0 11 a 0 OS 0 u 0 a «s >, 0 oO "1 _gffi a % a V W 0 V a 11 .0 a 0 .11 X 1*1 X **< 0 0 t Ph w 0 V a v a 3 u ^1- 1.46 1.50 1.55 1.60 1.65 1.71 1.75 1.83 1.90 1.96 2.05 2.12 2.18 2.28 2.37 2.46 2.55 2.64 2.73 2.81 2.88 2.96 3.04 3.13 3.21 3.30 3.38 3.46 3.53 3.60 3.67 3.74 3.81 3.88 3.96 4.04 4.12 4.18 4.26 4.33 4.40 4.46 4.51 4.57 4.65 4.72 4.79 4.85 4.92 4.98 5.04 5.10 5.16 5.23 5.29 5.34 5.39 5.45 5.50 5.55 Per cent transmission. 0 1 1 25 26 27 28 29 83.3 80.8 78.0 81.4 82.5 82.8 82.9 84.3 83.0 80.0 75.3 71.5 74.9 74.0 77.0 73.0 68.5 55.7 37.0 20.7 7.9 3.8 2.0 0.0 0.0 0.7 6.7 9.3 10.0 10.7 16.7 24.9 35.6 41.2 46.7 50.0 52.5 54.7 80.5 30 31 32 80.0 81.8 33 34 35 36 89.7 89.8 89.0 90.7 90.0 37 38 39 1 40 41 42 43 44 45 46 47 48 49 1 50 51 52 53 54 55 56 57 58 59 2 00 1 S 3 4 5 6 7 8 9 2 10 11 12 13 14 15 16 17 18 19 2 20 21 22 23 24 89.0 88.4 64.3 45.4 35.1 22.4 12.4 9.3 27.1 42.0 87.0 62.0 28.2 8.1 5.2 14.5 34.0 43.2 31.7 17.3 25.6 58.7 70.6 83 4 70.0 36.9 13.6 9.0 25.9 47.7 55.7 39.8 14.0 6.2 28.3 58.9 69.8 71.4 88.6 88.0 67.7 30.0 14.6 31.2 63.8 66.4 73.6' 77.8 'ss'.'s 52.0 20.0 8.2 11.5 30.9 48.3 41.5 29.1 11.2 23.3 52.6 64.0 87.7 69.0 54.0 55.4 65.0 75.5 54.4 24.6 7.3 21.5 62.8 74.0 80.3 85.4 90.0 80.0 68.5 69.4 74.0 80.3 75.0 59.8 17.8 17.8 51.7 78.3 88. '3' 88.3 88.4 "84.5 76.0 51.2 33.7 47.2 76.8 81.3 80.0 73.6 71.3 56.0 48.7 51.2 66.7 81.3 78.7 78.0 55.6 53.0 63.8 58.0 40.0 37.5 57.0 75.8 90.0 79.0 76.0 63.7 38.0 15.8 22.4 41.2 56.3 70.0 89.0 90.0 80.0 66.7 64.7 76.7 81.0 79.8 77.6 82.7 79.0 70.8 40.8 31.6 44.0 65.0 70.0 80.5 '82 .'7 59.5 39.5 18.4 20.2 57.0 80.7 48.0 49.5 58.0 68.8 69.8 78.6 67.6 89.0 89.0 80.7 88.4 93.3 76.1 90.0 88.0 86.4 83.5 88.0 91.8 "78'. 0 75.8 61.3 42.3 45.8 70.8 81.0 69.8 70.4 77.5 82.8 89.4 93.4 78.6 79.8 "79. 6 75.0 86.0 96.5 62.3 25.7 6.0 0.0 84.3 80.0 78.7 78.7 87.0 96.0 78.8 95.0 79.8 77.3 75.9 59. 2 89.0 96.0 79.2 '96.6 "ei'.'-r" 96.0 ■95.0' 78.4 78.6 78.8 91.0 92.3 0.0 74.3 85.5 80.0 77.0 78.0 78.5 82.6 87.5 64.5 91.0 95.3 0.0 81.0 78.6 89.4 88.7 89.0 90.4 89.0 91.4 91.8 89.8 89.8 92.3 94.7 0.0 9.0 20.2 85.8 95.2 78.0 94.3 62.7 58.3 54.7 50.8 83.5 90.3 90.8 90.5 90.1 94.2 78.8 91.8 86.2 90.0 88.3 91.8 71.3 53.6 35.6 '94.2 l62 INFRA-RED ABSORPTION SPECTRA. Table VIII D. — Observed Transmission of Various Carbohydrates, Using the Large Spectrometer, and the Prism at Variable Deviation, as Explained IN THE Text — ^Continued. Spectiometer s e t - tings. Augular de- viation from Na lines. a V > a d 2 "3 a n 0 i >. V 2 (S ;-■ 1 (J 2' 'u a CS 0 d u 0 g S w a v a ii a •r4 d w 0 0 a. >. "3 u s W d W 0 0 i) _n ft i u "?. V oo a n 1) 0 V a 0 w 0 0 0 a u S X 0 K 0 0 a V .a 0 So 0 0 13 < 0 a >> X u a 0 0 i 6 a !2 u 0. a'S" gu .Q u 0 S tj 0 a 0 0 M 0 0 0 a U bo 3 W 0 6 a v e s 0 K 0 u 1 OS 05 pu V a u W ■CO ou u nl a 0 .Q u d U 0 V a V a n •12 iHT) a g 0 0 K u V a V 3 "o H « w 0 0 w u 0 tn '5 < Thickness of \ cell in mm. J 0.01 0.02 0.001 0.01 0.01 0.0 1 0.01 0.01 0.0 1 0.01 0.01 0.16 2.73 2.81 2.88 2.96 3.04 3.13 3.21 3.30 3.38 3.46 3.53 3.60 3.67 3.74 3.81 3.88 3.96 4.04 5.55 5.60 5.65 5.70 5.75 5.79 5.84 5.89 5.95 600 6.05 6.10 6.15 6.20 6.25 6.30 6.36 6.41 6.46 6.51 6.56 6.60 6.65 6.70 6.75 6.80 6.86 6.91 6.96 7.00 7.04 7.08 7.12 7.17 7.22 7.27 7.32 7.37 7.41 Per cent transmission. 1 43 90.0 75.0 67.2 61.0 69.0 81.6 84.0 77.4 65.4 74.3 92.7 97.6 84.3 79.6 79.5 80.0 76.0 46.0 39.0 51.0 80.3 90.0 79.0 59.8 43.7 44.3 .54.4 63.5 53.3 38.9 39.1 62.2 80.7 71.0 27.2 13.7 30 0 54.5 63.2 51.0 31.0 26.3 34.5 59.5 72.0 87.3 83.3 69.5 45.6 16.4 12.3 46.8 75.3 83.5 81.8 80.2 80.0 80 0 77.8 51.7 37.4 25.0 31.7 46.0 64.3 76.4 44 53.5 20.4 9.4 8.2 19.0 38.3 42.3 30.8 11.7 17.7 33.0 22.8 7.6 3.3 5.0 20.0 29.3 41.7 58.0 66.7 69.8 71.4 87.4 75.0 53.2 50.0 65.5 69.8 52.2 24.3 8.0 7.6 18.2 47.0 63.7 89.3 85.0 68.0 65.8 77.2 87.5 45 34.0 17.5 6.7 5.5 3.0 0.0 0.0 0.0 ' '4.0" 46 92.0 73.5 73.0 62.3 29.3 22.3 40.7 58.7 73.6 47 87.0 64.7 27.3 17.5 19.4 44 8 74.8 82.8 55.0 29.8 14.0 38.1 68.3 90 0 48 49 1 50 51 52 53 54 90.8 87.0 76.5 71.4 80.3 89.0 80.0 61.0 40.0 34.0 47.0 66.0 55 56 45.3 71.5 78.7 5.5 7.6 15.0 57 58 63.8 90.0 59 2 00 27.6 ■ ■■ 24 25 70.8 71.0 14.7 7.0 0.0 26 70.0 67.5 27 71.2 69.0 70.0 70.0 69.0 70.0 28 87.0 61.6 62.3 68.3 73.4 29 0.0 30 88.0 86.8 31 2.5 32 66.8 72.8 84.0 73.3 81.0 77.4 70.8 60.0 3.5.1 33 16.0 34 64.5 68.0 70.7 67.8 50.0 22.8 9.3 14.0 25.6 52 3 73.0 63.0 83.4 76.8 35 81.6 38.5 43.7 36 63.6 63.5 65.0 43.7 21.2 20.0 36.6 55.2 59.4 31.6 10.0 00.0 11.0 20.0 84.3 'm.o 77.5 77.8 37 58.2 19.0 20.0 31.3 38 77.8 58.0 39 61.0 44.8 32.0 28.7 32.0 20.0 22.0 38.0 65.0 2 40 79.0 67.0 41 42 56.7 79.4 85.3 82.8 73.2 50.0 28.0 lO.O 11.0 16.5 76.0 80.0 43 44 55.6 5(j.'0 ■39.7 20.0 13.0 9.0 12.0 15.0 '22.0 66.0 50.0 40.0 36.0 36.0 40.8 46.0 86.7 56.7 77.2 54.0 33.0 52.7 61.2 51.5 24.2 16.6 15.0 25.5 45 39.0 34.3 15.8 11.3 17.3 42.0 46 78.0 43.5 26.0 29.0 41.0 25.0 12.5 25.0 36.0 53.8 31.0 12.0 5.0 0.0 5.0 10.0 54.0 20.0 10.0 12.0 28.3 24.2 25.0 47 48 88.3 73.3 68.3 .58.1 40.5 47.0 18.0 49 2 50 3.0 ' 0.0 1 51 52 64.0 53 54 65.0 18.5 38.3 76.0 47.7 0.0 55 61.0 56 58.3 47.0 76.0 82.7 57.3 0.0 57 26.3 85.0 58 72.0 58.8 57.8 56.0 65.8 73.0 46.0 32.8 36.0 50.0 57.4 10.0 59 3 00 75.0 19.0 1 2 79.0 1 TABLE VIII F. 165 Table VIII f.— Obser\t:d Transmission at Certain Regions or the Infra-red Spectrum, Using the Large Spectrometer, and the Prism at Variable Devi- ation, AS Explained in the Text— Continued. J.0 Thickness of \ cell in mm. ] 0.01 0.02 ^•w 0.01 0.01 0.0 1 0.01 us 0.01 0.01 0.01 0.01 0.01 00 0.16 7.80 7.84 7.88 7.91 7.95 8.00 8.04 8.08 8.12 8.16 8.19 8.23 8.27 8.35 8.41 68.3 'ei.V 50.0. 56.0 69.0 78.0 Per cent transmission. 81.0 S.4 50.0 55.0 56.8 58.5 40.0 20.6 26.0 38.5 64.0 80.0 78.0 -ffi J 66 INFRA-RED ABSORPTION SPECTRA. Fig, 12. TRANSMISSION CURVES. Fig. 13. 167 1 68 INFRA-RED ABSORPTION SPECTRA. Fig. 14. TRANSMISSION CURVES. Fig. 15. 169 170 INFRA-RED ABSORPTION SPECTRA. Fig. 1 6. TRANSMISSION CURVES. Fig. 17. 171 172 INFRA-RED ABSORPTION SPECTRA. Fig. 18. TRANSMISSION CURVES. Fig. 19. 173 174 INFRA-RED ABSORPTION SPECTRA. Fig. 20. Q^ -o Fi'g. 20. Oxygen 2l o o 00 O O - o o o TRANSMISSION CURVES. Fig. 21. vo -IvO o o o Fig. 21.. Carbon Monoxide CO o f^ M 176 INFRA-RED ABSORPTION SPECTRA. Fig. 22. >^ o o o o o Ov CO t^ so o TRANSMISSION CURVES. Fig. 23. 177 \^ 0 0 0 0 0 00 t^ 0 INFRA-RED ABSORPTION SPECTRA. Fig. 24. -o Fig. 24. Hydrogen Sulphide. H,S o o o o TRANSMISSION CURVES. Fig. 25. 179 i8o INFRA-RED ABSORPTION SPECTRA. Fig. 26. 0 0 0 0 0 0 0 0 0 CO l^ >o U-l ^i- re C) TRANSMISSION CURVES. Fig. 2^. I8l .182 INFRA-RED ABSORPTION SPECTRA. Fig. 28. TRANSMISSION CURVES. Fig. 29. 183 1 84 INFRA-RED ABSORPTION SPECTRA. Fig. 30. transmission curves. Fig. 31. 18: \^ 0 0 0 0 Q c> 00 t^ NO 0 (^O N i86 INFRA-RED ABSORPTION SPECTRA. Fig. 32, TRANSMISSION CURVES. Fig. 33. 187 i88 INFRA-RED ABSORPTION SPECTRA. Fig. 34. TRANSMISSION CURVES. Fig. 35. 189 190 INFRA-RED ABSORPTION SPECTRA. Fig. 36. TRANSMISSION CURVES. Fig. 37. 191 192 INFRA-RED ABSORPTION SPECTRA. Fig. 38. TRANSMISSION CURVES. Fig. 39. 193 IQ4 INFRA-RED ABSORPTIOX SPECTRA. Fig. 40. TRANSMISSION CURVES. Fig. 41. 195 196 INFRA-RED ABSORPTION SPECTRA. Fig. 42. TRANSMISSION CURVES. Fig. 43. 197 198 INFRA-RED ABSORPTION SPECTRA. Fig. Ad. 0 0 0 0 0 0 0 0 00 t^ vO VT) Th ro TRANSMISSION CURVES. Fig. 45- 199 200 INFRA-RED ABSORPTION SPECTRA. Fig. a6. TRANSMISSION CURVES. Fig. 47 201 202 INFRA-RED ABSORPTION SPECTRA. Fig. 48. TRANSMISSION CURVES. Fig. 49. 20.^ 204 INFRA-RED ABSORPTION SPECTRA. Fig. 50. TRANSMISSION CURVES. Fig. 51. 205 206 INFRA-RED ABSORPTION SPECTRA. Fig. S2. TRANSMISSION CURVES. Fig. 53. 207 208 infra-red absorptiox spectra. Fig. 54. TRANSMISSION CURVES. Fig. 55. 209 210 infra-red absorption spectra. Fig. 56. TRAXSMISSIOX CURVES. 211 Fig. 57. 212 INFRA-RED ABSORPTION SPECTRA. Fig. 58. TRANSMISSION CURVES. 2LS Fig. 59. 214 INFRA-RED ABSORPTION SPECTRA. Fig. 6o. TRANSMISSION CURVES. Fig. 6i, 215 2l6 INFRA-RED ABSORPTION SPECTRA. Fig. 62. TRAXSMISSIOX CURVES. Fig. 62,. 21' 2l8 INFRA-RED ABSORPTION SPECTRA. Fig, 64, TRANSMISSION CURVES. Fig. 65. 219 220 INFRA-RED ABSORPTION SPECTRA. Fig. 66. TRANSMISSION CURVES. 221 Fig. 6y. 222 INFRA-RED ABSORPTION SPECTRA. Fig. 68. ^=r5 o 0 o O O"' CO r^ O ui Tf- ro N TRANSMISSION CURVES. Fig. 69. 224 INFRA-RED ABSORPTION SPECTRA. Fig. 70. TRANSMISSION CURVES. 225 Fig. 71. 226 INFRA-RED ABSORPTION SPECTRA. Fig. y2. TRANSMISSION CURVES. Fig. jz. 227 228 INFRA-RED ABSORPTION SPECTRA. Fig. 74. TRANSMISSIOX CURVES. Fig. 75. 229 230 INFRA-RED ABSORPTION SPECTRA. Fig. y6. TRANSMISSION CURVES. 231 Fig, yy. 2^2 INFRA-RED ABSORPTION SPECTRA. Fig. 78. TRANSMISSION CURVES. ^?>?> Fig. 79. ^34 INFRA-RED ABSORPTION SPECTRA. Fig. 8o. TRANSMISSION CURVES. Fig. 8i. 235 236 INFRA-RED ABSORPTION SPECTRA. Fig. 82. s -1 V 0. m _ -^> — -< —^ ? 0 ( ;:r: r> '8 / > — =• o o o o o o TRANSMISSION CURVES. Fig. 83. 237 238 INFRA-RED ABSORPTION SPECTRA. Fig. 84. TRANSMISSION CURVES. Fig. 85. 239 240 INFRA-RED ABSORPTION SPECTRA. Fig. 86. o o o o o o O c> oo t^ vO «^ rh fO TRANSMISSION CURVES. Fig. 87. 241 0 0 0 0 0 0 0 0 0 CO r^ vO ^^ -"i- ro 242 INFRA-RED ABSORPTION SPECTRA. Fig. 88. w^ 0 0 p 0 0 0 0 0 0 00 rx so 4^ Ti- P^ 0 TRANSMISSION CURVES. Fig. 89. 243 244 INFRA-RED ABSORPTION SPECTRA. Fig. 90. TRANSMISSION CURVES. Fig. 91. 245 246 INFRA-RED ABSORPTION SPECTRA. Fig. 92. TRANSMISSION CURVES. Fig. 93. 247 248 infra-red absorption spectra. Fig. 94. TRANSMISSION CURVES. 249 Fig. 95. ^^ 0 0 0 0 0 c^ 00 t^ vO 0 250 INFRA-RED ABSORPTION SPECTRA. Fig. 96. TRANSMISSION CURVES. Fig. 97. 251 0000 252 INFRA-RED ABSORPTION SPECTRA. Fig. 98. TRANSMISSION CURVES. Fig. 99. 253 s 0 -— ■ D _: 0 0 i: c u t=> ^ 1 1 1 Fig. 99. Phe Large spe t t \ \ 1 » 1 ! •^ - 1 --^ — ' 1 -^ c c c 254 INFRA-RED ABSORPTION SPECTRA. Fig. ioo. TRANSMISSION CURVES. Fig. ioi. 255 2^6 INFRA-RED ABSORPTION SPECTRA. Fig. I02. TRANSMISSION CURVES. Fig. 103. 257 2S8 INFRA-RED ABSORPTION SPECTRA. Fig. 104. TRANSMISSION CURVES. Fig. 105. 259 26o INFRA-RED ABSORPTION SPECTRA. Fig. io6. TRAXSMISSION CURVES. 261 Fig. 107. 262 INFRA-RED ABSORPTION SPECTRA, Fig. io8. TRANSMISSION CURVES. 263 Fig. 109. 264 INFRA-RED ABSORPTION SPECTRA. P'iG. 1 10. _; 0 0 ■ 1 Cumi :ct. t = ( ^ > - 6 S. ^ z r- ^ 1 s _-»- - _ 3 \ v_ ~- ■ — - — .^ y ^ ^"^ — ■ — ^ f ■ 50 ^ o o 00 o o o TRANSMISSION CURVES. Fig, III, 265 266 INFRA-RED ABSORPTION SPECTRA. Fig. 112. O o p o o o tv. >o »-o Ti- o o ro CM TRANSMISSION CURVES. Fig. 113. 26^ 268 INFRA-RED ABSORPTION SPECTRA. Fig. 114. TRANSMISSION CURVES. Fig. 115. 269 2/0 INFRA-RED ABSORPTION SPECTRA. Fig. 1 1 6. TRANSMISSION CURVES. Fig. 117. 271 X - a * - u p. bfl - -J - bi 1 V. ■ — —«_ ^ . fl -1^ f .1 Y 1 1 \ \ \ — ^^ c «^j 1 ^ ^-^ \ f i ) C N CC I ^ - i > C C r C 1 c C - iy>>- . 2/2 INFRA-RED ABSORPTION SPECTRA. Fig. 1 1 8. TRANSMISSION CURVES. Fig. 119. ^72> 274 INFRA-RED ABSORPTION SPECTRA. Fig. I20. ^ o o o o c» 00 t^ so o TRANSMISSION CURVES. Fig. 121. 275 276 INFRA-RED ABSORPTION SPECTRA. Fig. 122. "^ 0 0 0 0 0 0 0 Q 0 CO t^ so *~n ^ ro 0 TRANSMISSION CURVES. Fig. 123. 277 278 INFRA-RED ABSORPTION SPECTRA. Fig. 124. TRANSMISSION CURVES. Fig. 125. 279 280 INFRA-RED ABSORPTION SPECTRA. Fig. 126. TRANSMISSION CURVES. Fig. 127. 281 Fig. 128. F ig. 128. c< Acetyl jrrected ene and for slit- Hefner width flame \ ■ ^ V \ \ \ \ \ b N \ tk , r -»^ \J ^^ I \ I f ■^•..^ ^^ ^ \ — , J 5/* 282 INFRA-RED ABSORPTION SPECTRA. Fig. 129. TRANSMISSION CURVES. 283 Fig. 130, 284 INFRA-RED ABSORPTION SPECTRA. Fig. 131. TRANSMISSION CURVES. Fig. 132. 285 /.^/8 /.5t9 1.510 1.5 Zl I.5ZI 1.513 I.SZ^ 1.5Z5 1.5 ZG /.5Z7 Fig. 132 Dispersion curves of Fluorite and of Rock salt Langley xxjtxx Rubens . 00000 Paschen .... (Fluorite) / 11 1 I jj n // / /I ; ; / / / 1 ll 1 1 f 1 / i < 1! If /• 4 •k rf \ 1 1 u /; f 1, 1 1 1 i i 1 f - / / 1 1 h A / /;' i , , 7 ? ^ 4^ PART II INFRA-RED EMISSION SPECTRA CHAPTER I. INTRODUCTION. The following experimental investigation was performed in the Physical Laboratory of Cornell University during the academic year 1904-5. It forms the second part of an investigation of infra-red radia- tion, rendered possible by a grant from the Carnegie Institution of Wash- ington. The first grant was for "Investigating infra-red emission and absorption spectra." Finding it impossible to complete the work in the time allotted, the Institution very generously renewed the grant, and the writer takes this opportunity to express his gratitude for the assist- ance rendered. In the Physical Laboratory he is under deep obligations to Profs. E. L. Nichols and E. Merritt for advice and criticisms as well as for the numerous facilities placed at his disposal. His dealings with the two institutions have been so agreeable that it is with regret that additional phases of the work could not be continued with them. HISTORICAL. i Infra-red Spectrum. In the visible and ultra-violet regions of the spectrum the most suc- cessful method of mapping the position of emission lines is by means of photography ; and so long as we do not desire a measure of the en- ergy in these lines this method is very satisfactory. On the other hand, in the infra-red, the photographic plate has never been made sensitive to rays of w'ave-length greater than about 1.2 /x. Other methods have been resorted to, one of the earliest being a phosphorescent plate. This method consists in placing the phosphorescent material (which is made into paint) upon an even surface and exposing it to the spectrum. The infra-red lines have the property of extinguishing the phosphorescence, thus leaving dark lines upon a bright background. The plate thus ex- posed is now placed in contact with a photographic plate which will be acted upon by the phosphorescing parts, but not by the darker lines. The photographic plate when developed is the equivalent of an ordi- nary positive plate. Becquerer was one of the first to apply this method with success. He photographed quite a number of lines of the alkali metals in the carbon arc to 1.2 fi. Recently this method was taken up anew by Lehmann," who was able to extend his investigations to i.y fi. Phosphorescent substances sensitive to radiations beyond I.//* are ^E. Becquerel : Compt. Rend., 99, 374, 1884. ^Lehmann: Phys. Zeit., 5, 823, 1904. See also Sci. Abstracts, 57, 425, and (1903) ; 750 (1905) ; Ann. der Physik (4), 5, 633, 1901. 289 290 INFRA-RED EMISSION SPECTRA. unknown. Since the effect upon the phosphorescent plate is cumulative, as is true of the photographic plate, this method is highly desirable for detecting weak lines, which could not be detected with a bolometer or thermopile. For measuring the energy of the different lines, the bolo- meter, radiometer, or the thermopile is the most useful. Julius,' using a bolometer, applied this method to explore the emission spectrum of a Bunsen flame, when different vapors, e. g., CS2 and Br, were burned in it. His results are of considerable interest, especially for CS2, for which he found emission bands whose maxima coincide with those of the absorption bands, and will be noticed in the present work on the radiation from gases in the vacuum-tubes. The infra-red spectra of the alkali metals were investigated by Snow," using a glass prism and a bolometer. He used the chlorides of the metals in an arc between hollow carbon electrodes. His observa- tions extend to about 2 /x, and show no strong lines beyond 1.4 ju,. The metals Rb and Cs show small lines up to i.y /x where Lehmann (loc.cit.) found small lines for these same metals. Whether this absence of lines beyond 1.5 /a was due to the opacity of the prism was not determined by Snow. Lewis,' using a radiomicrometer and a large grating, examined several alkali metals in the carbon arc. For Xa he found a doublet at A = 0.81837 and A = 0.8194 /x. The emission bands of CO, and H^O (vapor) have been investigated by Paschen.^ He found a strong emission band for COo which shifts toward the long wave-lengths with rise in temperature, the maximum being at 4.27/1, at 17° (absorption), at 4.3/^ for a temperature of 600°, at 4.388 fji for 1000° and 4.40 /* for the Bunsen flame, which has a tempera- ture of at least 1800°. His method of observation consisted in heating the CO, by passing it through an electrically heated platinum spiral, and finding its emission just as it leaves the spiral. Rubens and Aschkinass' have also investigated CO2 and H^O (vapor) to 20 /x, using a sylvite prism and a linear thermopile. They found a small emission band of CO2 at 14. i /x. The gas was heated by passing it through a metal cylinder, open at the ends and heated by means of Bunsen flames. Other emission bands of HoO and COo will be referred to in the text, and it will be sufficient to add that where one shifts toward the long wave-lengths another shifts toward the short wave-lengths. ^ Julius : Licht. u. Warmestrahluag verbannter Gase. Berlin, 1890. ^Snow: Phys. Rev., i, p. 35, 1893. ^Lewis : Astrophys. Jour., 2, p. i, 1895. *Pashen : Ann. der Physik (3), 53, p. 324, 1894. ^Rubens & Aschkinass: Ann. der Physik (3), 64, p. 584, 1898. HISTORICAL. 291 The dispersed radiation from gases in vacuum-tubes has been less extensively investigated. Runge and Paschen' investigated the infra- red emission spectrum of hehum (cleveite gas) to 7 fi, using for the pur- pose a bolometer and a fluorite prism. They found strong emission bands at 0.729 /x, 1.134 yu,, and 2.057 /x. The bands shift slightly for the different vacuum-tubes used, and also vary in intensity in a manner that could not be explained. No lines were found beyond the region of 2 fji. The three strong lines belong to those of the visible spectrum as predicted by the spectral series formulae. Angstrom^ measured the total radiation of gases in vacuum-tubes and remarked that to do so is a difficult matter, so that after dispersing the radiation the task is well-nigh impossible. In fact, the bolometer used by him was not sensitive enough to measure the intensity of the indi- vidual lines, hence only the total radiation was measured. For the positive column, at constant pressure. Angstrom found that the total radiation as well as the luminous radiation is proportional to the cur- rent. For hydrogen the total radiation is a maximum at a pressure of 1.02 mm., while the luminous radiation decreases with increase in pres- sure. For nitrogen the total radiation is a minimum at 3.6 mm. pres- sure. In the work of Drew (to be mentioned presently) the curves of total radiation of air also show a minimum, which, however, shifts with increase in size of the tube, the minimum being at a pressure of about 1.7 mm. for tube 0.9 cm. in diameter, and at a pressure of 0.5 mm. for a tube 1.8 cm. in diameter. From the fact that the total radiation increases, while the luminous radiation decreases, with increase in pressure of the gas. Angstrom concluded that there are two kinds of radiation present during the electrical discharge, viz, "regular" and "irregular" {i. e., lumines- cence). With decrease in pressure the former decreases while the irregular radiation increases in proportion as the motions are less ob- structed by the mass of the gas. At constant pressure a certain pro- portion of the energy in each molecule is converted into radiation; as the strength of the current increases the number of active molecules, and hence the radiation, increases in the same proportion. The number of the active molecules being relatively small, the damp- ening effect of the rest may be taken as constant, and the composition of the radiation remains practically unaltered as the current increases. On increasing the pressure the dampening effect changes, and the radia- tion becomes richer in infra-red rays. A greater proportion of the ' Runge & Paschen : Astrophys. Jour., 3, p. 4, 1896. ^ Angstrom ; Ann. der Physik (3), 48, p. 493, 1893. 292 INFRA-RED EMISSION SPECTRA. energy supplied is spent in heating, and, for the same current work, the total radiation increases with increasing pressure. After noticing the present results it will become quite apparent that Angstrom's predictions are in remarkably close agreement with observed facts. On the other hand, in his review of the recent work on electrical gas spectra, the predictions of Berndf are not so close in agreement with the present work. For example, in speaking of Ang- strom's observations, which show that total radiation increases while the visible radiation decreases with increased pressure of the gas, he remarks that this would show that the spectral distribution of the energy must shift with change in pressure, which has been observed. But, to say that the distribution of intensity changes in such a manner that, with decreasing pressure, the center of gravity of the distribution of the spectral energy shifts toward the shorter wave-lengths, does not seem admissible, since slight traces of CO, may have been present. In his efficiency investigation of vacuum-tube radiation Drew' ex- amined the emission spectrum of air, using for the purpose the writer's radiometer, which at that time was not very sensitive. He found several weak lines in the region of i /^, while the greater part of the energy is concentrated in a strong band at 4.75 /x. The other maxima worthy of notice are those at 0.66, 0.74, 0.89, and 1.03 )w. For air the 0.89 /A band is more intense than the one at 1.03 fx, while in the present work on nitrogen the reverse has been found. The infra-red emission spectrum of the mercury arc, in the form of the Arons lamp, has also been explored.' Three strong lines were found just beyond the red at 0.97, 1.045 and 1.285 /*> beyond which no lines were found, except in the region of 4.2 to 5.8 /x where the radiation from the hot cell interferes with the work. The fact that no line was observed at 4.75 IX is of interest in connection with the present work. Visible Spectrum. Several investigations in the visible spectrum are of interest in the present work, and they will now be mentioned. Certain prominent lines of hydrogen and nitrogen have been meas- ured photometrically by Ferry.' The tube was excited by means of a battery of accumulators consisting of 1,200 cells, the same as Ang- strom used for his infra-red work. He found that at constant pressure the intensity is proportional to the current, and with constant current the intensity increases with decrease in pressure, in a regular manner, ^Berndt : Jahrb. der Radioaktivitat u. Elektronik, i, p. 247, 1904. ^Drew: Phys. Rev., 17, p. 321, 1903. ^Coblentz & Geer : Phys. Rev., 16, p. 279, 1903. * Ferry : Phys. Rev., 7, p. i, 1898. HISTORICAL. 293 and differently for different lines. Langenbach' obtained similar re- sults for the visible spectrum, except that with constant current the intensity passes through a maximum for decrease in pressure of the gas. He found the three prominent hydrogen lines most intense for a pressure of about 2.5 mm. to 3.5 mm., and that they are very similar in their behavior with change in current, which was obtained from an induction coil. With increase in pressure the intensity of the red line increases the most; the energy maximum shifts toward the long wave- lengths, /. c, hydrogen behaves like a black body. This work was repeated by Berndt," who verified the work of Ferry, and added a new result, viz, that a rise in temperature of 200° has no influence upon the individual lines. For hydrogen he found that the green line grows more rapidly in intensity than does the red one, while for nitrogen the reverse is true of the red and green bands, which dis- proves the displacement law for gases. This work was continued by Waetzman,^ who verified the work of Ferry and Berndt, and showed that for a slight impurity the gas be- haves as the pure, but this impurity decreases the intensity of the long wave-lengths the most. Temperature, Dissociation, etc. The explanation of the luminous radiation from a vacuum-tube which is itself quite cool, has received considerable attention. Wiedemann,* using a calorimeter, shows that the average temperature of air in a vacuum-tube, at a pressure less than 3 mm., is less than 100° C, and that it depends upon the pressure. At a low pressure the temperature of the cathode is higher than that of the anode, while the reverse is true at a pressure greater than 2.6 mm. i\lthough the average temperature is low, few molecules may possess an amount of kinetic energy much greater than that indicated by the bolometer. Naccari and Gugliemo" found that at a pressure of 5 mm. the cathode is the hotter, reaching a maximum at 0.3 mm., then decreasing until at very low pressure, where the anode is the hotter. They found this to be due to a secondary action coming from the cathode rays. Wiedemann (loc. cit.) does not consider the radiation from the gas in a vacuum-tube to be due to a pure thermal effect, but of a phosphor- escent nature. He applies the term luminescence to all those luminous ^ Langenbach : Ann. der Physik (4), 10, p. 789, 1903. ^Berndt : Ann. der Physik (4), 12, p. iioi, 1903. ^Waetzman: Ann der Physik (4), 14, p. 772, 1904. * Wiedemann ; Ann der Physik (3), 6, p. 298, 1879 ; 7, p. 500, 1879 ; 9, p. 150, 1880 ; 10, p. 202, 1880. * Naccari & Guglielmo : Nuovo Cimento (3), 15, p. 272, 1884. ' 294 INFRA-RED EMISSION SPECTRA. processes in which the radiation is more intense than that correspond- ing to the sensible temperature of the tube. The luminosity of a gas is due to vibrations within the molecule rather than translatory motions of the molecules as a whole, which determine temperature. These intermolecular vibrations result from chemical or electrical disturbances. The theory of the low temperature of the vacuum-tube has been worked out by Warburg.' He computes the temperature upon the as- sumption that the energy, computed from the product of potential gra- dient and current, is changed into heat. He also shows that for a gas at a low pressure the heat conductivity is very rapid, so that in a small fraction of a second the gas assumes a constant temperature. In the same manner, on stopping the discharge the temperature will return to its original value ; hence there will be a rise and fall in temperature and a corresponding rise and fall in pressure of gas. For nitrogen he com.putes the following temperature for a current of 0.0012 ampere: Parts of tube. Temperatiires. Pressure 2 mm. Pressure 3 mm. Pressure 5 nim. Pressure 8 mm. 0 0.19 ao.36 21.6 039.0 II. 0 020.0 0 0 0.38 0 0.55 01.29 59-2 0132.6 30.4 069.6 29.4 41-3 14.9 21. 1 Temperature of inner wall=Wi.... Temperature of axis of tube^Wo' Mean temperature of gas, W o Computed for a current of 0.0032 ampere. This shows that while the inner wall is not above i°.5 the axis is at a temperature of 21° to I32°C. The interior of the vacuum-tube was explored by Wood," who used for the purpose a bolometer of very thin (o.ooi mm.) platinum-iridium wire. The highest temperature recorded was 43° C. In all cases the computed and observed temperatures were in very close agreement. Since the question of dissociation must be considered in this work it is of interest to note that Townsend' shows that the dissociation of the molecule by collision with an electron in a vacuum-tube is very small. All observers agree in thinking that when radiation is emitted by a gas, in one case by heating it and in another case by sending a current through it, the mechanism which is brought into play must differ in some important respects in the two cases. ^Warburg: Ann. der Physik (3), 54, p. 265 *Wood : Ann. der Physik (3), 59, 238, 1896 'Townsend: Phil. Mag. (6), i, 226, 1901. 265, 1895. CHAPTER II. SCOPE OF PRESENT INVESTIGATION. The present work divides itself into two parts: (i) emission spectra of the arc between metalHc electrodes and the emission spectra of the chloride of the metals in the carbon arc; (2) emission spectra of gases in vacuum-tubes. The contrast between these two forms of radiators is worthy of notice. The arc is noted for its enormous heat radiation in proportion to its light radiation. On the other hand, the vacuum-tube radiates but little heat. Consequently, in the study of these two kinds of radiators, the form of the device for exploring their spectra must differ. If a radiometer is used its period must be short for examining the arc in order to avoid heating of the window and the consequent shift- ing of the zero reading. This, however, is of less importance than the variation in intensity of the radiation from the arc, which requires a recording instrument having a short period. For the vacuum-tube a much greater sensitiveness must be used, which means a longer period for a linear radiometer vane. Fortunately the radiation from the vacuum-tube is uniform, which permits the use of a slow-period instru- ment. THE radiome;te;r. In the present investigation a Nichols' radiometer, a 7 cm. rock-salt prism, and a 35 cm. focal length mirror spectrometer were used.^ As a device for exploring infra-red spectra the radiometer has some inherent advantages as well as disadvantages over the bolometer. The most prominent advantages are its freedom from magnetic disturbances and its ease of construction and use at a high degree of sensitiveness. Its lack of mobility is not objectionable in the present work. It is also of interest to notice that in all my infra-red work this radiometer has stood upon a table, which itself stood upon a cement floor of a base- ment room, without any protection against earth tremors. Neverthe- less, only on rare occasions were the observations interrupted by earth ^E. F. Nichols: Astrophys. Jour., 13, loi, 1901. ^Coblentz : Phys. Rev., 16, p. 35, 77, 1903 ; also in "Absorption Spectra." (Wash- ington, 1905). 295 296 INFRA-RED EMISSION SPECTRA. tremors, even with the most sensitive arrangement used in the vacuum- tube radiation. Those v^ho have used sensitive galvanometers with their bolometers have been obliged to provide elaborate suspensions to protect the instrument against earth tremors and confine their observa- tions to the quiet hours of night. The chief disadvantages in a radio- meter of great sensitiveness, having linear vanes, is its long period, due to the viscosity of the residual gas. In fact, viscosity is the chief prob- lem to contend with in constructing a very sensitive radiometer having a short period. In a bolometer the spectrum is projected directly upon the bolometer strip, which can be made very narrow, while the period of the accompanying galvanometer is not to such a great extent deter- mined by viscosity of the air. In the radiometer, since the vanes re- volve about an axis, the spectrum must be projected upon a stationary slit, back of which one of the vanes is exposed. Now, the width of the vane that must be used depends entirely upon its distance from the slit. The slit is in the vertical focus of the spectral lines, which wiil diverge again after passing through it into the radiometer. In the original Nichols radiometer' the slit was at least 3 cm. from the vanes, which, as a consequence, had to be about 2.5 mm. wide and 1.5 cm. long. For his work on the radiation from stars^ no slit was required and the size of the circular vane was made equal that of the star image, viz, 2 mm. diameter. This eliminated viscosity to a great extent, and he succeeded in obtaining a period of 11 seconds. For line spectra the omission of the slit would hardly answer the purpose where great accuracy in the position of the wave-length is concerned. In the latter case the prob- lem is to place the vane as near the window as is possible. This can be done by using but one window,' as shown in figure 133. The use of but one window introduces a new complication in that any slight change in its temperature will immediately affect the radio- meter vane, i. e., will cause a shift in the zero reading. This was par- ticularly noticeable in the arc-spectra work, where the radiation is very intense. Consequently a suspension was used which gave a maximum deflection in 8 seconds. This was less sensitive than for the vacuum- tube work, but the arc lines are very intense and it answered the pur- pose, since exposing the vane for a longer time would warm the win- dow and cause a large shift of the zero reading. On the other hand, for the vacuum-tube radiation, where the radiation is very weak, the radiometer must be of extraordinary sensitiveness, to obtain which a ^Nichols: Phys. Rev., 11, p. 891, 1897. ^Nichols: Astrophys. Jour., 13, p. loi, 1901. THE RADIOMETER. 297 very fine quartz fiber must be selected — the present one was quite in- visible except by dififraction in sunlight. Here one must sacrifice per- iod for sensitiveness. In the present work the maximum of the deflec- tion was reached in 45 to 50 seconds. Since the radiation is weak, the shifting of the zero was not very great while making observations. But as soon as the making of observations in the region of 4.75 /x ceased, the deflection moved off the scale, and had to be brought back by turning the torsion-head. By using a double window of rock salt (zv zv, fig. 133) some of the shifting of the zero due to changes in tem- perature of the outer window was avoided. In addition to this the pressure in the radiometer was kept at o.oi mm., since at the point of maximum sensitiveness, 0.05 mm., heat conduction and convection cur- rents were much greater. For the emission spectra of the metals, the radiometer vanes had an area of about 2 by 15 mm. each, and the short period was due to the heavy fiber suspension. The vacuum-tube radiation being very weak required great sensitiveness, which was obtained in part by reducing the size of the vanes (of mica) to i by 10 mm. and selecting a fine fiber. The behavior of such a vane is entirely different from a heavier one. At low pressures the vane before the window was suddenly re- pelled from it, due apparently to the radiation through the window. This repulsion occurred when both vanes were black and when the unexposed one was not covered with lampblack. It was not due to electrification, and throughout the vacuum-tube work it was necessary to use a torsion-head (t, fig. 133) to keep the deflection on the scale. All greased joints were covered with beeswax and painted several times with dilute shellac, which dries quickly and forms the most satis- factory protection against leaking yet found. The rock-salt window was secured in the same manner. The radiometer was packed in wool. The window and spectrometer were inclosed in one continuous box, and no radiation could enter except at the slit. As described elsewhere,^ the apparatus stood in an inner room with an intervening door, which could be closed at will. In the region of 4.75 /x, where the deflections were large, the spectrum was explored in the daytime; but in the visible spectrum, and just beyond it, the observations were usually made at night, when the temperature could be more easily controlled. It was then necessary to enter the room and stay a while until the heat from the body established a new equilibrium. It was then possible, at times, to make readings for a quarter of an hour without detecting a shift of ^ Phys. Rev., 16, p. 35, 1903. 298 INFRA-RED EMISSION SPECTRA. the zero reading. Generally, however, the "drift" was from i to 2 mm, per 100 seconds, and since it was always in one direction it was easily eliminated form the total deflection. From this prolix description it will be seen that a radiometer of such extraordinary sensitiveness as the present one is useful only for work where the radiation is weak. In all other work a less sensitive, quicker- period instrument is more serviceable. From Table I the reader will obtain some idea of the sensitiveness of previous instruments used in radiation work. The paraffin candle is not a very satisfactory comparison source, but it is the only one for which data are available. The "sensitiveness" is expressed in deflections in centimeters per square millimeter of vane exposed, for a scale and a candle each placed at a meter's distance. Table I.— Sensitiveness of Different Instruments. LSensitveness is defined as the deflection in centimeters per square millimeter of exposed surface, for a candle and a scale at a distance of i meter.] Obser^'er. Boys, radiomicrometer — Nichols, first radiometer... Lewis, radiomicrometer... Snow, bolometer Stewart, radiometer Another form, esti- mated Nichols, radiometer for star work Drew, radiometer Coblentz : Radiometer for vacu- um-tube work Radiometer for absorp- tion-spectra work Deflec- tion. cm. 6 6 2 15 10.3 I . I 30 26 Distance of scale. cm. (?) 1-3 (?) 300 63 183 100 140 Distance from can- dle to in- strument. cm,. 152 600 300 100 300 811 200 140 Area of surface. sq. m,m. 4 2X1.5? 1.4 •35 30 3.14 7 i{=.095X9-8) Period. sees. 10 12 20 14 40 45 50 45 Sensi- tiveness. cm,. 0.9 7(?) 1-3 (?) 14.7 4.9 12.5 17. 1 52-3 8(?) The test of sensitiveness of the radiometer used in the vacuum-tube radiation is given in Table II. Since it was impossible to place the candle at a sufficient distance to keep the radiometer deflection upon the scale for the slit width, 0.7 mm, used in the regular work, the slit was decreased to 0.095 X 9-8 mm. (as found upon subsequent measurement under the microscope) and readings made for the candle at different distances. Several commercial paraffin candles having a diameter of 2 cm. were tried. The height of the flame was 5 cm. By keeping the wick w^ell trimmed there was no appreciable change in the deflections. THE RADIOMETER. 299 Numerous shields were provided to prevent stray radiation from enter- ing the radiometer. The data below the line in the center of the table show the constancy of the inverse square law which would indicate that within experimental errors the deflections are due to radiation from the candle onh'. Table II. — Sensitiveness of Radiometer. [Deflections are in centimeters per square millimeter of exposed surface (viz, i sq. mm.) for a candle and a scale each at a distance of one meter.] Deflections. Remarks. cm. c6.^ Pressure o.oi mm. ; candle at i . 5 meters. Trimmed candle ; flame 5 cm. in height. Exhausted to a pressure of 0.005 rnm. Pressure 0.02 mm. 52.3] CI . 2 V 52.3) 54-4 \ 53-3 f 68. 67.8 1 ca. 1, Pressure kept constant at o.oi mm. Another candle, at a distance of 1.5 m. from vane. Candle at 1.4 meters. Candle at 1.5 meters. Candle at i . 3 meters. Shutter raised ; radiation from a black wall when candle is extinguished. J '-> 50.8 1 49.0 ( CO. ^ 49-81 47.6 1 50.5 49-8 J 0.1 to 0. 2 In figure 134 is given the energy curve of acetylene for the visible spectrum, and that of a Nernst heater for the region of 14 ^u, for the very sensitive radiometer used under similar conditions in the vacuum- tube work, curve a, and, for the old radiometer used in the absorption- spectra work, curve Z?— ordinates are deflections in centimeters, radio- meter slit 0.7 mm., spectrometer slit i mm. The sudden drop from 30 cm. deflection at 13 /^ to 2 cm. at i6/x is due no doubt to the opacity of the rock-salt prism, which is very opaque in this region. Radiometer construction is still in its infancy. Other improvements suggest themselves. For example, there is no special reason for having two vanes, for symmetry, when using a torsion-head. B}' using a metal counterpoise instead of one of the vanes, the viscosity effects would be reduced by almost one-half, hence the period shortened. Another form, which is a combination of a radiomicrometer, which 300 INFRA-RED EMISSION SPECTRA. also serves as the vane of a radiometer, suggests itself. Even the finest quartz fibers are quite strong, so that they would support the extra weight of the loop of conducting wire which passes between the poles of the magnet. The extra weight would increase the period, while a coarser fiber would decrease it ; but whether the selection of the proper combination would compensate for the new difificulties involved remains to be determined. OTHER EXPERIMENTAL DETAILS. For the adjustment and calibration of the apparatus, the reader is referred to the memoir on "Absorption Spectra" (loc. cit.), and only certain details in the manipulation of the experiment will be mentioned. Since the arc varies with such great rapidity it was necessary to have a radiometer of quick period, as already mentioned, while special pro- vision had to be made to manipulate the arc and observe the deflections. The length of the arc was regulated by securing the electrodes in holders which could be separated or closed by means of a rack and pinion. This was mounted upon a holder which was in an asbestus- lined box, having a heavy sheet-iron shield through which the energy from the arc passed into the radiometer slit. The slit in the shield was I cm. high, while the length of the arc, which was directly back of OBJECT OF THE INVESTIGATION. 3OI it, was about 2 cm. in length. In this manner no radiation from the electrodes could enter the spectrometer slit. An image of the arc could be seen in the prism, and any slight trace of the radiation from the electrodes, when visible in the prism, was sufficient to cause large deflections, or even to throw the deflection entirely off the scale. The spectrometer arm was rotated, and the arc was adjusted before the slit by hand. The image in the rock-salt prism showed whether the adjust- ment was correct in the horizontal direction and whether any part of the electrodes was visible. By raising or lowering the shield or arc only the radiation from the vapors could enter the spectrometer slit. By placing a mirror above the scale an image of the latter, in the radiometer mirror, was reflected back into the viewing telescope, which was situated near the spectrometer. In this manner the observer could adjust the arc, raise the shutter, and make a reading before the arc had changed in intensity, which was very annoying when using metallic electrodes. With the salts of the metals in the carbon arc there were no serious fluctuations in intensity after starting the arc. OBJECT OF THE INVESTIGATION. Angstrom (loc. cit.) and others have found the absorption band of CO2 at 4.28 fx. and CO at 4.59 fx ; Paschen (loc. cit.) found that the COj emission band shifts toward the CO band with rise in temperature, being at 4.4 fx. in the Bunsen flame. The emission spectra work, par- ticularly that of gases in a vacuum-tube, was undertaken with the hope of gaining information on the subject of the dissociation of CO^. As already mentioned. Snow (loc. cit.) mapped the infra-red emission lines of the alkali metals to about 2 fi, and found numerous lines just at the end of the red, but no strong lines were located beyond 1.5 /a. The present investigation deals with the question of the distribution of emission lines (bands) in the infra-red, especially with the question of presence of lines beyond 1.5 /u.. All the infra-red lines predicted by our spectral series formulae end in the short wave-length just beyond the red. Any information as to the presence of lines beyond this point will aid in establishing these formulae upon a firmer, less empirical basis than they have at present. From our knowledge of the radiation from the "black body," which is most intense in the region of 1.2 to 2.5 ju, at high temperatures, one would expect the emission bands at 2 fi, if there be any, to be just as intense as those found by Snow at i fx. Other points of interest which developed as the work progressed will be noted in their proper places. CHAPTER III. INFRA-RED EMISSION SPECTRA OF METALS. This work was begun by examining the spark spectra of such metals as Zn, Al, and Cu. An induction coil and condenser were used. No emission Hnes could be detected; instead of lines a weak continuous radiation was detected in the region of 2 /a to 3 yu,, which appeared to be due to the hot particles from the electrodes. The arc between metallic electrodes of Fe, Zn, and Cu was then tried ; but no lines could be detected in the region of i /x, beyond which point the incandescent oxides gave such an intense, continuous, "black body" spectrum that the emission lines would have been obliterated by the radiation from the oxides. The vapors from the copper arc had but little "black body" radiation. No emission lines were detected, however, although several have been predicted in the region of 2.5 fi. V a 3 2 1 0 3 [ yv 1^- ■< 1 ~>- — x^ - . J A. 1 _' ■*^ a 1 ^ ""~"~~~ ^ • J 1 — *- 3 -^ Fig. 135. 7^ Copper. In figure 135 two sets of curve a are given for the vapors of the Cu arc, while curve h shows the radiation from the hot electrodes. In this figure it will be noticed that the terminals give a more intense radiation than the vapors, in which the density of the oxides is not so great as in the iron arc. 303 304 INFRA-RED EMISSION SPECTRA. Iron. In figure 136 are given the curves ot the arc between iron terminals. Curve b is for the electrodes, while curve a (scale is one-half of b) represents the distribution of the radiation from the vapors. Here the vapors have a more intense radiation than the solid terminals. This is no doubt due to the fact that we do not get the radiation from the hottest part of the electrodes, while in the arc, which is very rich in oxides, there is a better chance for the hottest region to emit radiation. In the Zn arc the oxides are formed so rapidly that it is almost le r\ \ f* \ - /2 — \ \a ^ JO .5 \ y ---> ^2> \ Q e \ / / / > ^^ ■^x 4- 1 / / / \ 1 ^^ ^v Z ]■ / -..^/^ '-~- -. f— < P 1 2 3 1- ? 5 ^ Fig. 136. impossible to work with this metal. The problem then is to separate the black-body radiation of the oxides from that of the vapors, which is very different from the work in the visible spectrum. Chlorides of Metals. The chlorides of Na, Li, and K were then examined in the carbon arc, using for the purpose hollow carbon electrodes filled with the salt. The carbons used varied from 6 to 9 mm. in diameter, the holes being from 1.4 to 2.5 mm. A direct current of 15 amperes from a 104-volt circuit was used. The radiometer slit was reduced to o.i mm. in width, nevertheless the radiation at 2 /a, which at most gave a deflection of only a few millimeters, was not resolved into individual lines. This is in marked contrast with the strong emission lines at i \x., and as will be noticed later on, the continuous radiation at 2 /a would blot out any weak emission lines, as far as a radiometer or a bolometer is concerned. Here a photographic process would be better, since the effect upon the METALS. 305 plate is cumulative, and one would have dark lines superposed upon a dark background, just as Abney and Festing' found for their absorp- tion spectra at i fi. Carbon Arc. In figure 137 is given the emission spectrum of the carbon electrodes, curve a, and that of the violet vapor of the arc, curve h. It will be noticed that there is but little radiation from the vapor except a slight amount from 2 to 3 /x. On the other hand the deflection was thrown off the scale for the radiation from the electrodes, just beyond the red. Snow (loc. cit.) found the radiation from the arc vapors con- 10 a 7 ■ y^ ' > / V j^ ^p^ r / /. ^ .66A ■^ -'^ e^.s'. 6^u. 3 -.013 =.02/ ".030 Fig. 150. 6 Amp. in primary =.033 Amp. in secondary RADIATION FROM A VACUUM-TUBE. 319 (fig*. 151) some of the latter vapor must have been present. The second sample, made with greater precautions, did not show any radiation beyond 2.5 /x. In the region of 4.75 /*, curve b, the observations for ' at collision, in any case, will be that acquired bv the charged particle in moving through its path under the action of electrical forces. When the electron collides with a molecule several effects will be produced. The first is to increase the kinetic energy of the molecule as a whole, which in turn, by colliding with other mole- cules, will cause a rise in the thermal temperature of the gas. As a result the electrons in the molecules wall be thrown out of their posi- tions of equilibrium, and will execute a series of vibrations. In so doing they will emit radiation in the form of heat or light, depend- ing upon the intensity of the excitation, which in turn depends upon the temperature. Since the mean temperature of the molecules is only about 300° abs. (Warburg, loc. cit.), it will be well to discuss presently the 4.75 fi COo band in this connection. When an electron collides with a molecule the second efifect is an acceleration of the former, which, when properly interpreted, means a wave motion, hence periodicity (Stark, loc. cit.). This period will be determined by the time of impact, independently of the chemical nature of the body. Since all possible times of impact are possible, an infinite number of dififerent electro-magnetic waves will be emitted, and we have a continuous spectrum. This is for the free electrons. The elec- trons that remain bound in the molecule will also be set into vibration by the impact of the collision. Since they are thus bound, there will be a relative motion among them, and the period of any one or group of these electrons will be characteristic of the kind of atom. The line and band spectra of the elements are, from this standpoint, due to these electrons within the atom.' In a more recent paper on this subject by Nutting^ the recombination of these dissociated aggregates is empha- sized as being the source of line spectra. 1 Stark : Ann. der Physik (4), 14, p. 506, 1904. ^Nutting: Astrophys. Jour., 21, p. 400, 1905. 326 INFRA-RED EMISSION SPECTRA. For this second effect the energy increases with the potential gradient, since the radiation from the gas is caused by the colHsions between molecules and electrons, which latter are moving with high speeds. As already mentioned in the high temperature radiation, where the collision is between molecules, the energy increases with the temperature. The electrical temperature distribution will be different from the thermal, and Kirchoff's law for the relation between emission and absorption will not hold for radiation from a gas in a vacuum-tube. The spectral distribution of intensity will be a function of the velocity distribution of the electrons. The greater the number of electrons with high speed in a volume element, the more will the intensity maximum be shifted towards the short wave-lengths. For this reason the cathode glow is blue, since the cathode fall is about 300 volts. On the other hand, in the position column, where the fall is only about 30 volts, the light emitted is red (nitrogen). Accord- ing to Stark's (loc. cit.) computations, in which the kinetic energy is equated to the temperature, this would indicate an electrical tempera- ture of some 6,000° for the cathode glow. The ionic energy^ (the minimum kinetic energy) necessary to disrupt an atom is in the order, metals Hg (8 volts), H, N (27 volts), and O. From this it would follow that if we mix H with N the fall of potential and the electrical temperature of the positive column will be changed. The temperature will be increased if the gas to be added has a higher ionic energy, £?. ^., N to Hg., or N to H. (Heuse,' Herz). This will be of interest in comparing the relative intensities of the 4.75 /x band of COo when it occurs as an impurity in N (very intense) and in O or NH3, where it is weak. These views will now be briefly considered in connection with the results obtained in the present research. Prior to this investigation on vacuum-tube radiation only one type of selective emission of gases in the infra-red had to be accounted for, viz, emission bands of water vapor and CO,. They were thought to be due to the thermal temperature of the gas. However, the data bearing upon this subject are so scarce that writers, in referring to them, generally expressed their opinions rather cautiously. Some have vaguely inti- mated that it might be something similar to luminescence in the visible spectrum — call it thermalescence. From the present research on the intensity of the infra-red emission bands of N and CO,, for constant current and var}ing pressures, and vice versa, it becomes evident that we have to deal with two distinct ^Herz: Ann. der Physik (3), 54, p. 244, 1893, ^Heuse ; Verb. d. d. phys. Ges. , i, 269, 1899. THEORETICAL. 327 types of radiation, the one being represented by the 4.75 (i band of CO and COo, the other being represented by the lines of N at 0.90 /a and 1.06 [x. The 4.75 fi band of CO and CO, behave in an entirely different manner from all the rest. Its intensity increases with increasing pres- sure (for constant current) of the gas, but never reaches a maximum, becoming asymptotic at 5 to 6 mm. pressure. On the other hand, the other bands increase in intensity with in- crease in pressure (for constant current), become a maximum at about 2 mm. pressure, and then decrease in intensity with a further increase in pressure, which agrees with observations in the visible spectrum. All lines increase in intensity with increase in current, as found in the visible spectrum. Condensers in parallel with the vacuum-tube caused a slight increase in the intensity of the lines, due to an increase in the current through the tube. This is due to the well-known fact that on account of the high self-induction of the coil, the discharge of the condenser takes the easier path through the vacuum-tube. The whole shows that the bands of N (He), and H, near the visible spectrum, are related to the visible bands, while the 4.75 /x band is of an entirely different type. Returning to the theory, it is interesting to recall Angstrom's (loc. cit.) predictions in regard to the mechanism which produces these radiations. As noticed elsewhere, he found that the total radiation increases, while the luminous radiation decreases with the increase in pressure of the gas, and concluded that there is a "regular" and an "irregular" radiation present during the electrical discharge. This would tend to change the efficiency of a vacuum-tube, as found by Angstrom and by Drew. In the present work, the decrease in infra-red radiation (4-75 /* band) and the simultaneous increase in the visible radiation, with decrease in pressure, explains very clearly the rise in efficiency of vacuum-tubes. It also explains why the total radiation passes through a minimum as observed by Angstrom, and by Drew (loc. cit.). In connection with the theoretical work just mentioned the behavior of these two types of radiation may be explained in the following man- ner. Consider the lines in and near the visible spectrum. At high pressures the electrons will not attain a high speed on account of the numerous neutral molecules, and their freedom of motion will be limited. At a lower pressure their freedom of motion will be greater, the number of collisions will be more frequent, the ionization will in- crease, and the electrical temperature,' which is proportional to the mean square of the ionic speeds, will attain a maximum. At a still lower 328 INFRA-RED EMISSION SPECTRA. pressure, on account of the scarcity of the molecules, there will be fewer collisions in a given time, the ionization will decrease, and the "electri- cal temperature" will decrease. On the other hand, this explanation will not account for the behavior of the 4.75 IX band, which appears to be due to a thermal radiation, excited by the collision of electrons with the neutral gas molecules. The gas molecule as a whole will suffer an increase in its kinetic energy, and, in colliding with other molecules, will cause a rise in the thermal temperature of the gas. With increase in pressure, i. e., in the number of gas molecules, the number of collisions will increase and the intensity of the thermal radiation will increase, but will not pass through a maxi- m.um, as is true of the other bands, because a stage will be arrived at where there will be a decrease in the ionization and in the collisions of the molecules. At still higher pressures the gas would cease to con- duct the current. Aside from these theoretical considerations, there is some experi- mental evidence for believing that the 4.75 jx band is of thermal origin. First, the gas must be hotter than the tube, for during the passage of the current the radiation tangential to the axis of the tube is probably different from the longitudinal, and the cell-walls assume the mean temperature of the gas only after the current has passed for some time. It has already been shown under the discussion of the temperature of the gas in the vacuum-tube (p. 322) that the mean thermal tempera- ture is from 300° to 400° C, depending upon the current and the press- ure. It was also shown there that the black body at these temperatures did not emit a perceptible radiation at i fx, while the emission lines in this region are very intense, indicating a temperature of perhaps 4,000" abs. On the other hand, the "black body" at 4.75 /a radiated almost as intensely as the vacuum-tube. This would indicate two distinct tem- peratures, which is hardly the case, and the term "electrical tempera- ture" seems appropriate for the intense lines at i /x. Second, the distribution of the heat in the vacuum-tube is very dif- ferent for different pressures. At a high pressure the constricted por- tion of the vacuum-tube is the hotter, while at low pressure it is quite cool, and the region surrounding the electrodes is the hotter. Third, the emission increases with the pressure (equivalent to an increase in the thickness of the emitting layer) and approaches a limit- ing value. Paschen (loc. cit.) has found for CO2 at atmospheric pressure, in a brass tube heated by a Bunsen burner, that a column 7 cm. long emitted and absorbed energy just as strong as a column 33 cm. long. In the present case the 15 cm. column of CO, at 5 mm. pressure is equivalent to a column i mm. long at 760 mm. pressure. Stark, Elektricitat in Gasen, from consideration of the kinetic energy of the elec- tron computes an electrical temperature of some 6,000°. THEORETICAL. 329 This is not a very long column of the gas as compared with Paschen's, nevertheless, judging from the behavior of the small traces of CO, in air, it does seem impossible for it to emit radiation as intense as that observed at 4.75 jj.. This would require a very high temperature. If the shifting of the COo band toward the long wave-lengths continues with rise in temperature for the region beyond 4.4//., just as Paschen (loc. cit.) found for the region preceding 4.4/*, then the 4.75 ju. band would indicate a temperature of some 6,500° to 7,000° (found by extrapolating from Paschen's values). This is close to Stark's (loc. cit.) "electrical temperature" of 6,000° for the cathode glow. Return- ing to the strong emission lines just at the end of the red, if we consider the maximum of the envelope (the curve) drawn through the highest points on these emission lines, which maximum lies just beyond the red, then, from the "displacement law," A^a^T=const, the thermal tempera- ture appears to be about 4,000° abs. From this line of reasoning it would appear that we can consider the 4.75 jj. band and the bands at the end of the red to be due to a high thermal condition in the vacuum-tube, without having recourse to an "electrical temperature." The continuous spectrum of alcohol vapor would indicate a higher temperature than that found by bolometric measurements. But even here the evidence is contradictory when compared with the emission of water vapor which showed no emission spectrum at 2.8 /i, where the Bunsen flame has emission lines. Evidently further investigation is needed to elucidate this subject — and such an investigation is in progress. Since submitting this paper for publication a similar investigation of CO, and N, by Drew (Phys. Rev., 21, p. 122, 1905) has appeared. He studied the emission band at 4.69/* (the 4-75/^ band in the present work) but did not succeed in eliminating it from N, which was made from sodium nitrate, as in the present work. Unfortunately he did not examine CO, and he reached a less definite conclusion as to the source of this line. He drew straight lines through the observed points, which gave an isosceles triangle, and from this he thinks that an observed shift of 0.02 /x for a change in pressure of the gas is real. His deflec- tions were much smaller than in the present work, and the observed points do not always lie close enough upon the lines to convince one of the reality of the shift. Such a shift of the maximum from 4.67 fi for a pressure of 3 mm. to 4.695 /x for a pressure of 0.6 mm. is to be ex- pected if CO2 dissociates into CO with decrease in pressure. In the present work the deflections were very much larger, but even here the variations in the readings are too great to be certain of a shift of the order of 0.02 /*. 330 INFRA-RED EMISSION SPECTRA. SUMMARY. The present investigation of infra-red emission spectra had for its aim the study of the region of the spectrum lying beyond 2 {x, which heretofore had never been examined. The question of the presence of emission Hues beyond this point is chiefly of theoretical interest. Two classes of radiation have been investigated, viz, the arc between metallic electrodes and the chlorides of the alkali metals in the carbon arc, and the discharge through a vacuum-tube using different vapors and gases. It was found, for the arc between metal electrodes, that the oxides emitted a black-body spectrum of sufficient intensity to obliterate any emission lines, if any were present. Using the chlorides of the alkali metals, the strong emission lines mapped by Snow were verified, but beyond 2 fx no emission lines could be found. The emission spectra of the following vapors and gases were exam- ined in a vacuum-tube; H,0, C^H^OH, H, N, NH3, O, CO, and CO,. Of this number the CoHgOH, CO2, and CO have a very strong emission band at 4.75 /a. Nitrogen is the only gas studied which has strong emission lines in the infra-red. The maxima are at 0.75 /x, 0.90 /a, and 1.06 ft. The behavior of these lines is entirely different from the 4.75 ju, band found in CO2 and CO. At a constant current the intensity of the 4.75 ix band increases with the pressure, but never reaches a maximum, becoming asymptotic at 5 to 6 mm. pressure. On the other hand, the nitrogen bands increase in pressure, become a maximum at about 2 mm. pressure, then decrease in intensity with a further increase in pressure, which agrees with observations in the visible spectrum. At a constant pressure all lines increase in intensity with increase in current, as found in the visible spectrum. Condenser in parallel increased the intensity slightly, due to an increase in the current through the tube. The aim in using a vacuum-tube was to avoid oxides. No lines, how- ever, were found beyond 2 /a, except the 4.75 yu. band, which seems to be due to the warming of the gas. Since the intensity of the vacuum- tube radiation is only from -j-^g to gr/oo as great as that of a black body, if there be weak emission lines beyond 6 |U. it would be almost impossi- ble to detect them with our present measuring instruments. The emission spectrum of C0H5OH shows that a vapor in a vacuum- tube can emit a continuous spectrum. INDEX OF COMPOUNDS. Page. Acetone 64 Acetic acid 69 Acetylene 44 Acetyl-eugenol 89 AUyl mustard oil 68 sulphide 64 Ammonia 53, 318 Arayl alcohol 58 Aniline 84 Anisol 78 Asphaltum 75 Azobenzene 127 Benzaldehyde 90 Benzene 76 Benzonitrile 82 Bromine 50 Brucite 129 Butane 47 Caproic acid 70 Carbon dioxide. . . .51, 313 disulphide ... 66 monoxide. .50, 314 tetrachloride. 55 Carvacrol 88 Cerotic acid 71 Chloroform 54 Cleveite gas 311 Copper 303 Cumene 81 Cuminol 91 Cyanine 82 Cymene 81 Decane 73 Decylene 73 Dextro-limonene 92 pinene 93 Dimethyl aniline 85 Diphenyl 83, 126 Dodecane 75 Dodecylene 73 Ethane 46 Ethyl alcohol 62, 315 cyanide 60 ether 48 hydrosulphide . 63 iodide 61 succinate 65 sulphate 64 Page. Ethyl sulphide 63 sulphocyanate . 68 mustard oil. ... 68 Ethylene 45 Ethylene bromide. ... 56 Eucalyptol 94 Eugenol 88 Glycerin 34 Hexadecane 73 Hexadecylene 73 Hexane 74 Hydrocarbons : C19H36 73 C22H42 73 C10H20S 73, 75 Hydrogen 50, 312 Hydrogen sulphide. . . 52 Illuminating gas 49 Iodoform 54 Iron 304 Isocaproic acid 71 Le Page's glue 70 Limonene 92 Lithium 307 Laevo pinene 93 Menthol 95 Mesitylene 79 Meta-xylene 78 Methane 43 Methyl acetate 64 aniline 84 carbonate .... 65 cyanide 60 ether 48 iodide 61 isosulphocyan- ate or mus- tard oil ... . 67 salicylate 8g sulphocyanate 67 Monobrombenzene. .. 80 Monochlorbenzene. . . 80 Monoheptadecane. ... 75 Monotridecane 75 Myricyl alcohol 58 Naphthalene 127 Nitrobenzene 86 Nitroethane 60 Nitrogen 316 Page. Nitromethane 59 Nitrotoluene 86 Octadecylene 75 Octane 73 Octadecane 73 Oleic acid 71 Ortho-nitrotoluene. . . 86 toluidine 86 xylene 78 Oxygen 49, 313 Paraldehyde 91 Para-nitrotoluene. . . . 86 nitrosodimethyl aniline 85 xylene 78 Pentadecylene 73 Petroleum distillates. 72 Phenol 87 Phenyl acetate 89 mustard oil. . . 68 Picoline 96 Pinene 93 Piperidine 97 Potassium 307 Pyridine 96 Pyrrol 99 Quinoline 98 Resin 93 Safrol 81 Selenite 130 Sodium 306 Stearic acid 71 Sulphur 66 Sulphur dioxide 52 Terpineol 95 Tetrachlorethylene ... 55 Tetracosane 73 Tetracosylene 75 Tetradecane 73 Thiophene 98 Thymol 88 Toluene 77 Triethylamine 63 Valeric acid 70 Venice turpentine. ... 93 Water 56, 311 Xylenes 78 Xylidine 85 331 INVESTIGATIONS OF INFRA-RED SPECTRA Part I — Infra-Red Absorption Spectra Part II — Infra-Red Emission Spectra BY WILLIAM W. COBLENTZ WASHINGTON, D. C. : Published by the Carnegie Institution of Washington October, 1905.