MEEK THE ABSORPTION SPECTRA ETC. 969 L X X X I I I . - T h e Absorption Spectra of some Polyhydroxy-an thr.aquinone Dges in Concentrated Sulphwic Acid Solution and in the S t a t e of Vupour. By DAVID B. MEEK. IN a previous communication (Meek and Watson T. 1916 109, 544 e t seq.) on the colour of the polyhydroxyanthraquinone dyes i t was shown that the wave-length of the maximum of an absorp-tion band of any of the substances depended orL whether the sub-stance was examined in alcoholic solution in potassium hydroxide solution or on cloth variously mordanted. Taking the absorp-tion as due to the resonance produced in a system capable of oscillating between the different tautomeric forms then this period of oscillation was regarded as modified by the nature of the radicle attached to the conjugate chain of double and single bonds.One of the conclusions drawn from that investigatdon was that “the more electropositive the nature of the radicle attached to a con-jugate chain the longer will be the wave-length of the maximum of the absorption band.” I n that work the spectrum of each of the dyes was examined (1) in alcoholic solution (2) in aqueous potassium hydroxide solution and on wool inordaiited with (3) tin, (4) alum and (5) chrome respectively. The present investigation was undertaken with a view t o obtain information regarding th 970 MEEK THE ABSORPTION SPECTRA OF SOME absorption spectra of the same polyhydroxyanthraquinone dyes in solution in concentrated sulphuric acid and in the state of vapour. Some of these dyes had given a number of fairly narrow sharp bands in alcoholic solution (Zoc.cit.) and i t seemed of importance to discover whether under other conditions the narrow bands which had been obtained could be broken up into absorption lines, for as a vapour a t atmospheric pressure iodine has very fine absorption lines whilst in solution i t gives absorption bands which are broader than those yielded by some of the polyhydroxyanthra-quinone dyes. The difficulty of obtaining the absorption spectra of many organic substances as vapour is that they frequently decompose before vaporising and hence the absorption spectra have to be observed when the dyes are vaporised at low pressures. With the same end in view namely of obtaining very narrow absorption bands one of the substances alizarin-cyanine which gave narrow bands in alcoholic solution was also examined in a number of organic solvents.It may be stated here that i t has not been found possible t o resolve any of the bands given by the polyhydroxyanthraquinone dyes into absorption lines but absorption curves havo been obtained which on resolution into elementary symmetrical bands have yielded further information regarding the effect of the number and position of auxochromes on the absorption and there-for0 on colour. E X P E R I M E N TAL. The absorption spectra of the dyes in concentrated sulphuric acid and of alizarin-cyanine in the various organic solvents were obtained in a manner similar to that described in the previous paper (Zoc. c i t . ) the apparatus being a Nutting photometer in combination with a large Hilger wave-length spectrometer.I n obtaining the positions of the absorption bands in the case of the vapours a slightly modified procedure was adopted. A brass tube 50 cm. in length and 1.5 cm. in diameter was fitted with air-tight caps taken from a polarimeter tube. The portion of these caps which is generally fixed to the glass polarimeter tube by means of hard wax was brazed t o the ends of the brass tube. They were then ground plane and by means of asbestos washers the caps could be made air-tight when screwed on firmly. Attached t o the side of the brass tube was a small brass tube through which the pressure inside the long brass tube could be reduced. To observe the absorption spectrum of the vapour of one of the dyes a small quantity of the substance was placed in a porce-lain boat and introduced into t h e brass tube.The polarimete POLYHYDROXYANTHRAQUINONE DYES ETC. 971 tube cap was then firmly screwed on and the whole tube placed in a resistance electrical furnace. The latter was larger than the brass tube and so the ends were well within the furnace. This prevented the deposition of the vaporised substance on the glass ends of the tube. The side-tube projected beyond the furnace and was attached to a Gaede pump by means of which the pressure inside the absorption tube could be reduced. The furnace was arranged so that a parallel beam of light could be passed through the absorption tube to one aperture of the photometer and another beam from the same source was brought by reflections to the other aperture.This method should have given measurements from which absorption curves could have been drawn but the difficulty of keeping the amount of vapour in the tube constant has not yet been overcome with the result that only the wave-lengths of the maxima of absorption have been determined up to the present. Although the absorption curves have not been obtained yet for the substances in the form of vapour an attempt has been made to classify the absorption bands according t o apparent intensity. The following table contains the results which have been obtained: TABLE I. Wavo-lengths of Comparative intensity Substance in the form of maxima of of these maxima of vapour mixed with air. absorption.absorption. (1) 5137 Very faint ( 2 ) 5040 Intense Intense ( 5 ) 4736 Very faint (6) 4635 Extremely faint Quinizarin or 1 4-dihydroxy-anthraquinone (fluorescent) Faint Purpurin for 1 2 4-trihy-droxyanthraquinone ( 1 ) 5256 Intense (2) 5045 Intense (3) Broad band of Intense general absorp-tion from violet upwards to about 4900 ( ( 1 ) 5135 ( 2 j 5050 Alizarin-Bordeaux or 1 2 5 8-tetrahydroxyanthraquinone [(5j 4693 Intense Intense Faint Very faint Extremely faint These absorption bands have been shown in Figs. 2 3 and 5 as straight lines drawn a t the wave-lengths of maximum absorption and of length varying according to the qualitative classification of intensity in the third column of Table I. It is interesting that the bands of quinizarin and Alizarin-Bordeaux vapours are closer P P 972 MEEK THE ABSORPTION SPECTRA OF SOME together than the bands of these substances in solution in alcohol, and it is also noteworthy that the differences between successive maxima of absorption of the substances as vapours are of the same order of magnitude as the differences between the successive maxima for alizarin-cyanine and anthracene-blue in alcoholic solution.The absorption clue t o the vapours of the other five FIG. 1. Wave-lengths. F I G . 2. 0 - -d4 * *m m m w (0 wdr * * m m m w w c o b 0 0 0 0 0 0 0 8 dr 0 0 0 d4 m o * 0 0 0 d4 0 0 0 d4 a30 * 0 0 0 2.0 1.8 1.6 1.4 1.2 1.0 0.8 0.6 5 0.4 .z 0.2 O a f c: 0 1.8 *s 1.6 2 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0 0 dr 0 0 0 * 0 0 0 * 0 0 0 * 0 0 0 dr 0 0 0 * 0 0 0 0 dr -0 u3 m o w w o * *o m u 3 0 w w o 0 t- 0 * 0 dr 0 co 0 m 0 W 0 m FIG.3. Wave-lengths. FIG. 4. _ _ _ _ - _ Elementary synzmetrical curves. polyhydroxyaiithraquinones considered here was not resolved into narrow bands. Alizarin vapour gave geii era1 absorption coming up from the violet end of the spectrum as the density of the vapour was increased. Anthragallol gave a similar result whilst anthracen+blue gave a broad indistinct weak band with its centre about h = 5000 POLYHYDROXYANTHRAQUINONE DYES ETC. 973 The Absorption Spectra of the PolyJ~ydroxyantl~ru~~~i?~ones iijh Conceiztrated Sulphuric A cid Solutions. Solutions of strength AT/ 1.04 in concentrated sulphuric acid were prepared and examined by the Nutting spectrophotometer in the manner previously described (Zoc.cit.). The absorption curves obtained are given in Figs. 1 t o 8. FIG. 5 . Wave-lengths. FIG. 6. 0 0 0 0 0 0 0 8 G - 4 -2 * O - 8 -6 5 .4 .f .2 % 8 P .+ - 8 . G 2 T *8 - 6 .2 IIvrrrlrrllllllllllrlII11IY1I 0.0 > * 0 0 0 dc 0 0 0 * 0 0 0 dc 0 0 0 bi 0 0 0 dr 00d ) * * g LC m 0 (0 (90 bi dco 10 m o (0 c D 0 > 0 I u3 CD t- 0 cD 0 u3 0 cr 0 FIG. 7. Wace-lengths. FIG. 8. (1) Alizarin.-In potassium hydroxide solution there are dis-tinctly three bands the middle one being the most intense. I n concentrated sulphuric acid solution there are indications that these bands still exist but the middle one and the one towards the red have diminished in relative intensity .and probably also P P" 974 MEEK THE ABSORPTION SPECTRA OF SOME have broadened.The one towards the violet end has becoine relatively more intense. (2) &fri)?imrin.-The results are given in Table 11. TABLE 11. Wave -lengths of maxima Solvent,. A. (3) 4926 (4j 4815 (5) 4736 (6) 4635 Vapour (1) 5150 (2) 5018 ( 3 ) 4780 Alcohol (1) 5470 (1) 5944 Potassium (2) 5526 f=l/X. A(l/h) 194G 38 1954 47 2031 46 2077 34 2111 47 2158 -1942 51 1993 99 1828 139 1967 126 1683 127 1810 128 2092 -2093 -1938 -Relative order of intensity. 3 2 1 1 2 3 1 almost 2} equal 3 Remarks. (4) - (1) = 131 ( 5 ) - (2) = 127 (6) - (3) = 127 The absorption spectra of quinizarin in alcohol and as a vapour have the absorption over the same area of the spectrum as regards both extent and position.I n the concentrated sulphuric acid solu-tion and in the potassium hydroxide solution the areas of absorp-tion are larger but again almost equal and they have been shifted towards the red end of the spectrum the shift being greater in the case of the alkali solution. (3) Yrrrpurin.-The results are given in Table 111. TABLE 111. Wave-lengths of maxima Relative order Solvent,. A. f= 1/A. A( 1 /A). of intensity. Remarks. (1) 5256 1903 79 (l)'\about -(2) 5045 1982 59? (2)j equal -(3) Broad 2041? - - -band of general ab-s o r p t i o n from violet up t o about 4900 (1) 5524 1811 110 4 (2) 5205 1921 123 1 (3) 4892 2044 48? 2 (4) 4780? 2092? - 3 (1) 5564 1798 141 2 hydroxide \(2) 5113 1956 - 1 1 2 I Vapour Alcohol Sulphuric acid { (2) 5155 1939 -Potassium j(1) 5450 1835 12 POLY HYDROXY ANTHRAQUINONE 1) Y E S ETC.975 Again in the case of purpurin the absorptions by the vapour and by the alcoholic solutions occupy almost identical parts of the spectrum but as in quinizarin the absorption maxima are closer together for the vapour t,han for the solution in alcohol. With concentrated sulphuric acid and potassium hydroxide solutions the absorption has been displaced towards the red and the maxima f o r sulphuric acid solution are nearer the red end of the spectrum than the maxima for the alkali solution. (4) L4 nthwrycrlZol.-Iii the sulphuric acid solution there are three bands giving maximum absorption at wave-lengths 5750 5240 and 4620 and as regards intensity these are in the order 3 2 1 respectively.These three bands are not apparent in the potassium hydroxide solution as distinct bands but their existence as broad bands may be inferred from the shape of the absorption curve of the potassium hydroxide solution. One effect of the sulphuric acid has been t o render the resolution better and t o increase the relative intensity of the absorption towards the violet^ end of the spectrum. (5) A Ziza?.in-Bor~ZeaziIL:.--The results are given ii-i Table IV. TABLE IV. Wave -lengths of maxima Relative order Solvent. A. j=l/x. A(l/h). of intensity. (1) 5135 1947 33 Intense (2) 5050 1980 62 Intense (3) 4896 2042 30 Faint (4) 4805 2081 50 Veryfaiiit (5) 4693 2131 - Extremely faint {);; mi? 1'309? '19? 2 2128 - 1 (1) 6400 1663 181 4 (2) 5732 1744 143 1 Sulphuric acid (3) 5300 1887 10'3 2 (4) 5010 1096 - 3 i Vapour Alcohol 2 1 i Potassium (( 1) 6045 1655 115 hydroxide ( ( 2 ) 5650 1770 -Comparing the absorption in the state of vapour tion in alcohol the maximum of absorption is nearer the violet for the alcohol solution than for the vapowr.I n the state of vapour the bands are narrower and the resolution much better. For sulphuric acid and potassium hydroxide solutions the absorp-tion has been displaced towards the red and in the two solutions it occupies the same part of the spectrum. The sulphuric acid solution gives better resolution on the blue side of the absorptio 976 MEEK THE ABSORPTION SPECTRA OF SOME curve whereas the pot'assium hydroxide solution gives better resolution on the red side.(6) Alizarin-cyanine.-Table V contains the results for alizarin-cyan me. TABLE V. X-for ( 1 ) 5630 (4) 5085 ( 5 ) 4978 I (6) 4776 Solvent. maximum. (2) 5473 (3) 5337 Alcohol Relative order f = 1 /A. A( 1 /A). of intensity. Remarks. 1776 51 3 1827 47 1 f - f = 1874 92 2 (2) (4) 130 1966 43 4 2009 80 5 2098 - 6 (1) 6332 1580 76 3 (2) 6040 1656 137 1 (3) 5575 1793 183 2 (4) 5060? 1976 - 4 Potassium hydroxide Gradually r i s i n g curve from violet to red. No maximum b e t w e e n 4000 and I 7000 From table V and also from the curves it is again obvious that the absorption has been displaced towards the red in the con-centrated sulphuric acid solution and still further towards the red in the potassium hydroxide solution.The band which has been numbered (3) in the alcoholic solution has been masked in the sulphuric acid solution by the intensity of band (a) so that the bands (l) (2) (3) and (4) in sulphuric acid solution correspond with the bands (l) (a) (4) and (6) in the alcoholic solution. The band (2) in both solutions is the most intense and i t is also prob-able that the position of the maximum of absorption is not very much removed from the centre of the absorption band so we may obtain the displacement towards the red in the concentrated sulphuric acid solution by comparing band number (2) in alcohol with band number (2) in sulphuric acid.These bands we may take a t h=5473 and 6040 giving a displacement of 567 Angstriim units towards red. I n frequencies the figures are f=l827 and 1656 with a displacement of 171. Anthracene-bZwc TV,R.-Table VI gives the results for anthra-cene-blue POLYHY DROXYANTIIRAQUINONE DYES ETC. 977 X .for (1) 5655 (2) 5487 Solvent. maximum. Alcohol (6) 4875? TABLE VI. Relative order f = 1/X. A( 1 /A). of intensity. Remarks. 1768 55 -1823 47 - f - f = 1870 05 - (4) (2) 142 1965 37 _- f - f = 2002 49 - (5) (3) 132 - 2051 -Gradually rising curve from violet to red with an inflect.ioi1 a t about 5450 which shows that the curve is not a single absorption band. Potassium hydroxide Again the effect of the sulphuric acid and of the potassium hydroxide is to displace the absorption towards the red and the displacement due to the sulphuric acid is not so great as that due t o the potassium hydroxide.The resolution is also much better in the solution in sulphuric acid. Alizarin-cyanin e in Various Solvents. The absorption spectra of alizarin-cyanine in various solvents have been examined and the curves are given in Figs. 9 and 10. The results are also contained in table VII. Alizarin-cyanine was chosen for the purpose on account of the large number of narrow bands i t gives in solution in many organic solvents and the ease with which corresponding bands can be identified in the various solutions. From Table VII and also from Figs. 9 etc. it will be seen that the absorption bands can be recognised quite easily when dis-placed by the various solvents.The bands have been given dis-tinguishing letters - 3 1 3 c' etc. The last column of Table VII contains the refractive indices of the various solvents for sodium light. According to Kundt's law the greater the refractive index of the solvent the greater should be the displacement towards the red. I n many cases there is dis-placement towards the red but this is not always accompanied by an increase in the refractive index of the solvent. Looking a t the graph of wave-lengths of maxima of absorption against refractive indices of solvents (Fig. 12) we see that if the solvents of acid and basic nature are omitted chloroform is the d y solvent of those considered which violates Kundt's law t o a This is however not the case TABLE VII.Alizarin-cyanine in Various Solvents. Wave-lengths and frequencies ,4. B. C. D. E. F. Q. - 5589 5460 5330 5214 5086 4970 - 1789 1832 1876 1918 1967 2012 - 43 44 42 49 45 Solvent. Ether ........................... Amy1 alcohol .................. Chloroform ..................... Acetone ........................ Methyl alcohol ................ - 5599 5483 5346 5217 5100 4992 - 1786 1824 1871 1916 1961 2003 - 38 47 45 45 42 5726 5576 5432 5311 5173 5058 4960 1746 1793 1841 1883 1933 1977 2016 37 48 42 50 44 39 - - 5458 5324 5195 5083 4962 - - 1832 1879 1925 1968 2015 47 46 43 47 - -5558 5441 5313 5178 5065 4961 - 1799 1838 1882 1931 1974 2016 - 39 44 49 43 4 TABLE VII. (continued). Alizarin-cyanine id Various Solvents.Wave-lengths and frequencies A . B . C . D . E . F . G. I 5568 5436 5308 5186 5058 4952 - 1796 1840 1884 1929 1977 2019 - 44 44 4.5 48 42 Solvent. Glacial acetic acid x Pyridine ........................ Amy1 ether ..................... ........................ ra Anisole Phenol ........................ 13 Nitrobenzene .................. - 5545 5422 5286 5159 5028 4913 - 41 48 47 50 47 1803 1844 1892 1939 1989 2036 -5708 5608 5478 5346 5223 5102 4986 32 43 45 44 47 45 1752 1784 1827 1872 1916 1963 2008 - 5643 5512 5362 5227 5113 5000 - 1772 1816 1865 1913 1956 2000 - 44 49 48 43 44 5483 5375 - 5115 5005 - -- - 1822 1860 - 1956 1998 - - 38 96 42 1722 1772 1815 1873 1913 - 2007 1722 1772 1815 1873 1913 - 2007 50 43 58 40 9 980 MEEK THE ABSORPTION SPECTRA OF SOME 1.4-1.3 1.2 1.1 1.0 0.9 i u Ei 0.8 * * g3 S 0.7 E -* u 0.6-v) rQ 9 0.5 0.4 0-3 0.2 0.1 large extent.Neglecting pyridine acetic acid and phenol there is a general tendency to displacement towxds the larger wave-length with increase of refractive index but the nature of the ------------O k FIG. 9. 45 47 ~-51 53 55 57 59 61 solvent as regards acidity or basicity certainly has an effect on the displacement of the absorption bands. The displacements cannot be due t o absorption bands in th POLYHYDROXYANTHRAQUINONE DYES ETC. 981 infra-red of the spectra of the solvents for the absorption con-sidered is that of the solution less that of the solvent,.So far as the effect of solvent on the colour of the polyhydroxy-anthraquinone dyes is considered the results are : FIG. 10. 41 43 Wave -1eiqths. -- Chloroform. - - - - - ._ - Glacial acetic acid. EthRr. - . - . - . -\ ' \ \ \ '. T (1) The absorption is displaced towards the red end of the spec-trum by solution in sulphuric acid and in potassium hydroxide as compared with absorption of the alcoholic solution. (2) The displacement is greater for potassium hydroxide solutio 982 MEEK THE ABSORPTION SPECTRA OF SOME than for sulphuric acid solution. The one exception is purpurin, which was shown t o be an exception in other respects (T. 1916, 109 561). (3) The resolution in sulphuric acid solution is generally FIG. 11. much Wave - lengt hs.- - Alizarin-cyanine in anisole. 9 ) )) pyridine. , ,) phenol. - - _ - - - -- . - . - . -greater than in potassium hydroxide solution and this is the case most frequently on the violet side of the group of absorption bands. (4) For the few neutral solvents experimented with Kundt’ POLYHYDROXYANTHRAQUINONE DYES ETC. 983 law is in a general way true but it is quite wrong when acid and basic solvents are included. T?he Chcrnge in t?Le Absorption S p e c t m Produced b y the TTnriation of tlir Nutnher nnd Posifioir of the rl7ixochroines. Previously (T. 1916 109 556) the gerieralisations formulated by Georgievics (Morintsh. 1911 32 329 et sep.) on ths influence of hydroxyl groups on the colour cf lakes were criticised and four FIG. 12.Wave-length of centre of same band i n various solvents. rules were formulated which seemed to be more in harmony with the facts. Pursuing this point further the absorption curves of (1) alizarin (1 2-dihydroxyanthraquinone) (2) quinizarine (1 4-dihydroxyanthraquinone) (3) purpurin (1 2 4-trihydroxyanthra-quinone) (4) anthragallol (I 2 3-trihydroxyanthraquinone) in sulphuric acid indicated by the full lines in Figs. 1 to 4 respectively have been resolved into symmetrical bands shown by the dotted lines in the figures. It has frequently been suggested by various experimenters that the absorption in the visible and i 984 MEEK THE ABSORPTION SPECTRA 03’ SOME the ultra-violet portions of the spectrum is due to electrons associated with masses of various magnitudes and calculations in the case of some substances showing selective reflection gives a mass of the order of the molecule.I f we take the elementary curves in Figs. 1 to 4 and find the ratio of the values of m / p e , where rn is the mass of the electron e the charge and p the number of electrons per molecule for each set of elementary curves we obtain the numbers in table VIII. TABLE VIII. Relative values of mfpe for elementary bands. Band No. 1. Band No. 2. Band No. 3. Substance. - Alizarin .................. 7 1 Quinizarin ............... 1 1 3 Purpurin ............... 2 1 Anthragallol ............ 4 2 1 -From Table V I I I we see that the value of m/pe for the alizarin bands increases with increase in tho wave-length of the maximum of absorption whilst the reverse is the case with quinizarin.To a less degree purpurin and anthragallol show the same effect as quinizarin and alizarin respectively. I n alizarin the hydroxyl groups are in the positions 1 2 whilst in quinizarin the groups are in the 1 :4-positions. Similarly purpurin is 1 2 4 - and anthragallol is 1 2 3-trihydroxyanthraquinone. Hence from the above and also from the curves in Figs. 1 to 4 it will be seen that the effect of the position of the auxochromes seems t o be as follows : the closer the hydroxyl groups are to each other in the benzene nucleus the broader and less intense become the bands towards the red side of the absorption group. A comparison of the curves given in the previous paper for these substances in other solvents, for example in potassium hydroxide bears out the same con-clusion.Also it seems to be true that the proximity of the auxo-chromes in the benzene nucleus determines the displacement towards the red but the mere displacement towards the red with closeness of the auxochromes becomes relatively unimportant when compared with the decrease in intensity of absorption and to the increase in the breadth of the bands on the red side of the absorp-tion group. So far as displacement toward the longer wave-length with proximity of auxochromes is concerned that could be ex-plained by an increase in the period of oscillation of the electron due to an increase in the capacity of the whole system produced by the closer proximity of the auxochromes. The broadening of the bands relative to the intensity would be due to the increas POLYHYDROXYANTHRAQUINONE DYES ETC.985 in the friction to which the absorbing vibrating electrons are sub-jected for if we take a simple absorption band produced by electrons moving according to the equation (12 d t then the term ti- represents the allowance made for friction and the greater the value of 7; the greater the breadth of the absorp-tion band relative t o its intensity that is the greater the friction the broader the absorption bands. The results then of bringing the auxochromes close together in the benzene nucleus are: (1) A displacement of the absorption bands towards longer wave-lengths and (2) A decrease of the intensity of the bands relative to their breadth. The latter result produces a greater change on the colour than the former.It is responsible for the brown colour of such dyes as antIragallo1 and rufigallol in certain solvents. The A BsorytioiL Spectrum of Aliznriiz-Cyn~iiite in Piperidine. When a solution of strength N/lO4 was prepared 1 cm. thick-ness gave the absorption curve (1) in Fig. 13 half an hour after preparation. Taking the wave-length where the absorption is a maximum as the centre of an absorption band then curve (1) in Fig. 13 gives the bands tabulated in table IX. TABLE IX. Band. - 4 . R . C . D . E . F . Q . Wave-length ...... 5990 5757 5544 5442 5295 5153 5064 Intensity ............ 0.48 0.526 0.64 0.613 0.50 0.434 0.354 tensity ......... 5 3 1 2 4 6 7 The absorption changes while the observations are being taken, so that time readings are necessary t o obtain the absorption curve a t a definite time after the preparation of the solution.Examina-tion of the solution immediately it has been prepared shows that the bands A B and E are absent or a t least very faint and that with time they begin to appear. When the solution is freshly prepared C is the most intense band and next in intensity comes D. The solution was allowed to remain in the dark for forty-seven hours and the absorption curve mas then (2) of Fig. 13. (The vessel remained sealed throughout the experiments from the Relative order of in 986 MEEK THE ABSORPTION SPECTRA O F SOME moment when the alizarin-cyanine was added to the piperidine, and therefore water vapour and carbon dioxide were excluded.) 47 49 M FIG.13. i 55 57 Wave-length. 61 63 (1) Alizarin-cyanine in piperidine 4 hour after preparation.. 9 9 9 , 47 hours ,) 1 7 9 71 7 7 ,, (2) ( 3) > t ? 9 9 1 95 f 7 9 9 7 9 ammonia. (4) ( 5 ) Y Y 9 3 The same bands are still present but E F and G are only shown by a change in the gradient of the curve. The intensity of al POLYHYDROXYANTHRAQUIONE DYES ETC. 987 the bands has increased and A and a have increased t o a greater degree than C and D with the result that B is now the band of maximum intensity. The wave-lengths corresponding with the maxima of A and B have increased. That need not be considered as due to any shift of the bands but merely as the natural dis-placement of the maxima caused by the change in the relative intensities of the d and B bands with respect t o the G and D bands.Curve (3) Fig. 13 gives the absorption of the same thick-ness of the same solution seventy-one hours after preparation. The intensity of absorption throughout has decreased. The bands A and E still remain distinguishable but C’ and D have become merged into one band. After ninety-five hours from preparation the same solution has only two bands remaining namely B and D with a slight suggestion of d. The intensity has decreased further and the fading continues. The wave-lengths of maxima of absorption are now 5902 and 5454 and comparing the absorp-tion curve (4) with the curve for alizarin-cyanine in concentrated ammonia solution ( 5 ) a strong resemblance is observed. Fading of the solution or what is the same thing general decrease in intensity of the absorption throughout must be ex-plained by a decrease per unit volume of the number of vibrating systems giving the various bands but the disappearance from the absorption curve of the individual bands .4 C’ 3 F and G can only be explained by a broadening of these bands.This on the ordinary mechanical theory would mean that the frictional element in the forced vibrations causing these bands was increased, and that therefore systems with periods differing from the true natural period by large amounts are made to resonate. It is not assumed that the individual bands A C‘ E F and G have dis-appeared entirely but they have broadened t o such a degree that the observed curve does not show them resolved as separate bands. The conclusion drawn previously (T. 1916 109 555) that “ t h e more electropositive the nature of the radicle attached to the conjugate chain the lopger will be the wave-length of the maxi-mum of the absorption band,” is supported by the comparison of the absorption curves of alizarin-cyanine in alcohol and in piperidine. I n this case the absorption curve of the solution in piperidine shows an intensification of the bands toward the longer wave-length and a diminution of the intensity of the bands towards the shorter wave-length. With the solution in potassium hydr-oxide and the dyed fabrics on chrome alum and tin mordants, the absorption curves were broad and did not show the separate bands as such. This would mean that the friction is much greater on these niordanted fabrics and also in the potassium hydroxid 988 RlITTER AND SEN ACTION OF PHENYLHYDRAZINE solution than in the alcoholic solution and the displacement of the centre of the broad resultant band would then be given by the increase of the intensity of the elementary bands towards the red end of the spectrum when the electropositive nature of the radicle is increased. DACCA COLLEGE, DACCA E. BENGAL, INDIA. [Received May 25tA 1917.