THE SPECTROGRAPHIC SENSE’TIVI‘N CF MANGANESE, COBALT AND MAGNESIUM ”Fixesis for the Degree 0? M. S‘ MECHIGAN 3"? TE COLLEG 1303122 Arihur Crane- E946 {'13 l 1 J. 0. c ’l aeti‘s‘gfcfijg 5.53;,1 } 1-451; 7;»; .16; En! ,4 , ”3;, . Fifi-"7%” (14“? ”Ml: x133 ‘\ s.’ 3? WIT“ "“ ‘* ”gt w’..'.‘:-‘('«"- "‘ 'l‘“ ll- " “I; a? " .\;~'~ $ . .v ‘ v.5) f: ‘ I ‘~ “ I ", ' 1F ‘03" ‘1 v i o ‘ l. ' . ‘ l “ ‘ ”IV, K/4 "'2‘ . Wk ' ‘ -‘ t " .1 . ‘ . 1F ' “If" ’ I 9‘ l ‘..“.- M, -| 'I l V. v ' ' ‘ P‘- 1 | . ':.\ , 1%.]; ) l . ‘ ‘ ‘3, N Y ' ‘ ‘ \‘ .. ‘ . g ll .‘ t .r l ' . u \ t ‘ .' . ‘ \ I This is to certifg that the thesis entitled Aid curd , amt». W pres med b1] ‘ _ j \3 MBA/K W (Lew has been accepted towards fulfillment of the requirements for _/;ai_wdegree in-“ < P L7 All Z M Major profesgr ( Date_ ( w& "4/! L? H (0 11-795 \ I I THE SPECTROGRAPHIC SENSITIVITY 0F MANGANESE, COBALT AND MAGNESIUM By John Arthur Crane 'A THESIS Submitted to the School of Graduate Studies of Michigan State College of Agricuflture and Applied Science in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Chemistry 1946 Ci tummy DEPT. ‘ {-1 ,' "x": ‘ ' Acknowledgement The writer wishes to express his appreciation to Dr. D. T. Ewing for his guidance and enthusiasm in the direction of this investigation. ,1». C3 {PIE nil Pg- \n. J t J c3 Introduction Spectrographic examinations of dilute solutions show that 30 milli-micrograms of manganese, 6O milli-micrOgrams of cobalt and 3 milli-micrograms of magnesium can be read- ily detected. This paper is largely a study of the various factors to be taken into consideration to show the presence of such small quantities of these three elements in various types of samples. Important factors necessary to consider in this investigation are purity of materials including electrodes, particular wave length of element under obser- vation, excitation voltages and exposure times. Two methods of excitation have been investigated. The preliminary work utilized.a direct durrent arc source in which the solution was placed in the cavity of a carbon electrode, the solution evaporated and an arc struck be- tween this and a second carbon rod. After the sensitivity of manganese was determined by this method, the condensed spark excitation was used. This consisted of evaporating the solution on pure ccpper electrodes and passing a high voltage spark between two cepper electrodes. This work has been concerned with the determination of the actual amounts of the elements present on the elect- rode and from these values, the concentration in the complete sample has been calculated. Apparatus The direct current arc source consists of a 220 volt direct current generator, a potential divider, and a series resistor by which the current through the circuit can be controlled. The circuit diagram is shown in Fig. l. The condensed spark apparatus is a little more com- plicated and is shown in Fig. 2. It is made up of a step- up transformer with 110 volt primary and a maximum second- ary potential of 25,000 volts, and a .02 microfarad conden- ser with a maximum of 20,000 volts. A self inductance coil is placed in series with the spark in order to elim- inate air lines from the spectrum. (1) The electrodes used in this investigation are shown in Fig. 3. National Carbon spectrographic carbons were used in the arc method, and special purity cOpper elect- rodes were used in all spark work. A Bausch and Lomb Littrow type spectrOgraph with a quartz Optical system was used in making all the photo- graphic plates in this investigation. The Optical system of the instrument is shown in Fig. 4. Light enters the slit (S) and passes through a right angle prism (R) by which it is rotated 90° to pass through a columnating lens (L) before it enters the Littrow prism (P). This is a 30° prism, the back surface of which is coated with aluminum so that it will reflect ultraviolet and visible radiation without absorption bands. The light is then reflected by the prism through the same lens (L) which now 220 D.C. Fig. l.- Direct current are circuit . O b IL jr Fig. 2.- Condensed spark circuit . acts as a focusing lens and serves to focus the energy on the photOgraphic plate which is maintained at the correct curvature to keep the spectrum in focus. (I) (2) Instrument was adjusted to position VII (focus 157.5; tilt 210) which covers the range from 2327 to 2973 Angstroms. Kodak Spectrum Analysis No. 1 plates were used in this investigation. They were processed in a constant temperature, 18°C., mechanically agitated deveIOping tank using chak deveIOper D-ll. After the plate was dried, the density of the various lines was determined by means of a Hilger misrephotometer. This instrument consists of a light source which is focused on the photo- graphic plate by means of a series of lenses. The light which passes through the plate is picked.up by a second series of lenses which focus the light on a photoelectric cell. The current which is generated in the photo cell is transmitted to a ballistic galvancmetsr. The galvano- meter deflection is a measure of the relative amount of light incident on the photo cell. (2) .---- - ”AL" ‘ - £5 mm 3 mm. J.—_::‘.Zl: 5° MELE mm Tm ’V L._J (a) (b) Fig. 3.- (a) carbon electrodes; (b) capper electrodes. “nigh l I I 1 I t I i t I i t l I I I I I dPL H 2,43? Fig. 4.- Optical diagram of the Littrow spectrOgraph. The are was used as a light source in the first part of this investigation. The anode electrodes were spectro- scepic grade carbons % inch in diameter and broken into approximately 2 inch lengths. A crater was bored in one end of the rod as shown in Fig. 5. The cathodes were of the same quality, but were 1/8 inch in diameter so as to prevent the are from wandering out of focus. The carbons were arced 30 seconds at 60 volts and 6 amperes in order to remove any superficial impurities. The anodes were placed in a 120°C oven for 30 minutes after which the craters were filled with the samples. The electrodes were than air dried for 30 minutes followed by placing then in the oven for 15 minutes. The arc can best be aligned by going through the follow- ing procedure. Remove the plate holder from the spectro- graph and adjust the focus and tilt of the instrument so that the visible portion of the spectrum is in focus. Open the shutter and remove the Hartman diaphram completely. This will allow the entire length of the slit to be illum- inated. Strike an arc between two carbon electrodes and by observing the spectrum with the aid of'a magnifying eye- piece, adjust the lateral positicn of the electrodes so the spectrum is visible. Next, the horizontal position must be aligned so that the spectral lines are of constant inten- sity over their full length. When the adjustments have been completed, the position of the electrode iamge is marked on a permanently located test screen so that the electrodes can.always be set at the same position. The base solution was made up of .1% copper as CuC12.2H2O in #% hydrochloric acid to which 25 grams of ammonium chloride was added per liter of solution. man- ganese was added to this solution as Mn012.4H20 (.3675 gram contains 100 milligrams of manganese) and the volume made up to one liter. By dilution of this stock solution with the base solution, the concentration of the various samples was varied to cover the range from .1 to 10 milli- grams per liter. By this method, it was possible to obtain solutions which contained definite concentrations of man- ganese, and each sample contained a constant amount of COpper which served as an internal standard for the quanp titative work. The ammonium chloride and hydrochloric acid were added to the solution in order to produce a more uniform arc. (3) The samples were placed in the crater by means of a calibrated pipette to be certain that .3ml. were used in all cases. It was necessary to determine the Optimum exposure time by observing the line density and background density of spectra produced by exposures covering the probable range of time. The results of this determination are shown in Table I. From the table, it will be seen that the maximum line density compared to the background density was obtained with an.exposure of 180 seconds. It should also be noted that the background intensity increased sharply with ex- posures of more than 120 seconds. Since it was desirable to use lines of low density, it was deemed better to use an exposure of 120 seconds in order to obtain the most dense lines and the least possible background. The re- lation of background density to exposure time is shown in Fig. 5. The data presented here is taken from plate #23. Table I Effect of Exposure Time on Phot0graphic Densities Clear plate deflection 2 30 cm. Exposure Time Deflection Density (LOg IO/I) Seconds Backgr. Cu Backgr. Cu Cu/Backgr. 30 27.6 12.8 .036 .370 .334 45 27.4 10.0 .039 .477 .436 60 27.3 9.4 .041 .504 .463 90 27.1 8.4 .044 .553 .509 120 27.0 7.7 .046 .591 .545 180 25.2 6.7 .076 .650 .584 240 23.9 6.4 .099, .671 .572 300 22.5 6.2 .125 .686 .561 360 21.9 6.0 Ii36 .700 .564 The develOping process was also investigated in order to increase the sensitivity of this method. The original development time was four minutes at 18°C. with continuous agitation of the develOpsr. By developing plates at various shorter periods, it was found that no appreciable change in the density of the line occured after two minutes in the develOper if the background density was taken into consideration. The final procedure decided upon which gave ' 350 300 250 200 100 150 50 .14 .12 .10 .08 .02 Fig. 5.- Relation of background density to exposure time. the best results was to develOp in D-ll at 18°C. for two minutes followed by fifteen seconds in a 5% acetic acid st0p bath. The plate was fixed in the acid fixing bath for two minutes after which it was washed in running water for fifteen minutes, rinsed in distilled water, sponged and dried on an.A.RtL.- Deitert plate drier. After the plate had been dried, the lines to be ex- amined were spotted on the plate and the densities deter- mined by means of a Hilger micrOphotometer. This instru- ment must be turned on ten minutes before any readings are taken in order to a110w the light source to attain equili- brium. The galvancmeter deflection for total darkness on the photocell must be noted and should remain at zero. The galvanometer deflection.was adjusted to a constant value for the background of each spectrum (Io), and the deflection (I) was measured for each line being used. The Density was calculated from these values; D = Log 10/1. (2) The various steps of the step sector spectrum were read in the same manner in order to obtain the plate calibration curve. In order to obtain any accurate results, an inten- city calibration must be made of each photOgraphic plate used. (4) This was done by removing the Hartman diaphram and using the motor driven rotating step sector, the seven steps of which are in a ratio of 1:1.5. The Density of the steps is plotted against the 1ogrithm of the relative exposure (LOg E) to obtain the calibration curve. The step sector was exposed by means of an iron are at 50 volts and 6 amperes for four seconds. The Hartman diaphram was used in making the exposures of the samples. The shutter was Opened before the arc was struck. The are was run.at 50 volts and 6 amperes for two minutes. The most sensitive arc line of manganese is that line at 4030.8 International Angstroms, but this is in the range of the carbon bands so it, of course, could not be used. The next most sensitive line is at 2576.1 Angstroms and is the one which was used. (5) The density values of the man- ganese line and of the 00pper line at 2618.4 Angstroms was determined and by refering these values to the calibration curve, the Log relative intensity of the two lines was de- termined. This value was then plotted against the LOg COD! centration to give the working curve. The OOpper line used for a reference line should be chosen such that its density lies on the straight portion of the calibration curve and should be a line of constant density. A typical set of data used for obtaining the working curve for the determination of manganese by the arc method is given in Table II. The data for this table have been obtained from plate #33. Figs. 6 and 7 show the calibra- tion curve and working curve respectively which were ob- tained from this data. It will be seen from the data that the extreme concentration of manganese which could be de- tected was .4 mg.1. Since the samples each contained .3 ml., the extreme amount of manganese which could be detected by this method was 120 milli-micrOgrams. J TABLE II , . CALIBRATION AND WORKING CURVE DATA FOR.MANGANESE DETERMINATION BY THE 120 METHOD Sector Step 0 l 2 3 4 5 6 7 Calv. Read. 25.0 24.8 24.5 23.7 22.0 19.3 15.6 12.3 Density .000 .003 .009 .022 .055 .113 .205 .308 Manganese Galv. Density . Lo Intensity Content Reading LOg 10/1 Mn 257 .1; Cu 2618.4 g. x 10‘9 Mn Cu Mn Cu Mn Cu Mn/Cu 3000 19.8 20.8 .101 .080 4.78 4.48 .30 2400 20.0 19.7 .097 .103 4.74 4.83 -.09 1800 20.8 20.1 .080 .095 4.84 4.70 -.22 1200 21.5 19.6 .066 .105 4.23 4.86 -.63 900 22.8 20.7 .040 .082 3.69 4.50 -.91 300 24.2 20.5 .014 .086 2.45 4.56 -2.11 240 24.2 19.5 .014 .108 2.45 4.90 «2.45 180 24.5 19.2 .009 .115 2.00 5.01 ~3.01 150 24.6 20.6 .007 .084 1.70 4.54 .2.84 120 24.6 19.7 .007 .103 1.70 4.83 -3.13 90 25.0 19.2 .000 .115 ---- 5.01 --- m st Eli 1| I M a 1 1 _ . [liq owulnt g ,_ M , _ i * 19r- a lllwll ill “ill I M m 1 a 1 1 m .1 _ u in a a (-IET : in 1 1 hit». a ll. 1 I» a h H _ 1 in _ a w l#-,l ll a II Tim. at _ ‘ _ _ _ a _ 3 On. I...“ hvflmcon Log E Fig. 6.- Plate calibration curve for D-ll develOpsr. lln 2 6.1 01'? . Log .4 /{ / // // / V 100 200 '500 500 1000 2000 4000 ‘ (Milli-micrograms manganese Fig. 7.- Working curve for the determination of manganese by the arc method. In the spark excitation method, the electrodes used were special purity cOpper rods & inch in diameter and about 3 inches long. The electrodes were machined level at first, but the drop tended to run off the sparking surface during the evaporation process. In order to correct this defect, various shapes were tried, and a 5° taper on each electrode was decided upon. The elect- rode on which the dr0p was placed was made concave and the other was made convex on the same taper in order to maintain a constant space between the electrode surfaces. The taper was held to a minimum so that the light would not be shielded from the spectrograph. The electrodes were machined before each exposure to remove the oxide coating formed by the spark. By removing .015 inch, a clean, fresh surface was obtained. Thus each pair of electrodes could be used for a large number of exposures. The standard solutions were prepared by dissolving a weighed amount of Mn012.4H20 in distilled water. The copper in the electrodes served as the internal standard. No ammonium chloride or hydrochloric acid was added to these samples. The electrodes were made ready for exposing by placing one drOp from a calibrated pipette on the face of the COD! cave electrode and evaporating the sample in an.0ven at 1200 C. for thirty minutes. The exposures were made immediately after removal of the electrodes from the oven. The spark gap was prOperly aligned according to the method previously used in this laboratory. (6) A spark is passed between two electrodes which have been set at approximately the correct position. The Hartman diaphram and step sector housing are removed from the Optical path; the focusing lens is adjusted horizontally so as to pro- duce a sharp image on the slit housing. The electrodes are shifted so that the image is vertically in line with the slit; then the lens is adjusted vertically so that the best possible illumination is obtained over the entire length of the slit. The alignment is checked by setting the spectrograph to the visible range and examining the spectrum with the aid of a magnifying eyepiece. A primary potential of 55 volts with a current of 7 amperes was used for the spark excitation. The Optimum exposure time was determined by making exposures of 15, 30, 45, 60, 75 and 90 seconds from which it was decided to use 45 seconds for exposing all the samples. The lines reached their maximum density in this time, and the back- ground density was held to a minimum. The step sector was exposed under the same excitation conditions using iron electrodes for thirty seconds. The plate develOpment process and the micrOphotometer procedures were the same as those described in the section dealing with the arc excitation methods. A typical set of data for the determination of man- ganese by the spark excitation method is given in Table III. The working curve from this data, which was taken from plates 59 and 60, is shown in Fig. 8. The curve has a tendency to level off at the lower concentrations, but this is to be expected due to the extremely low amount of manganese present in the spark gap. TABLE III WORKING CURVE DATA FOR MANGANESE DETERMINATION BY THE SPARK METHOD Manganese in sample g. x 10‘9 300 240 210 180 150 120 100 60 3O 24 GfilVe Reading Mn 22.9 21.0 20.5 20.3 20.9 23.0 23.1 24.2 24.3 25.0 Cu 24.2 22.8 22.5 21.6 21.5 22.6 21.6 22.2 21.8 23.2 Density Log Io/I Mn .038 .075 .086 .090 .078 .036 .034 .014 .012 Cu .014 .040 .048 .064 .066 .044 .064 .044 ~059 .033 Mn 257 .1; Cu 2482.3 Mn 2.64 4.39 4.57 4.63 4.44 3.56 3~53 2.03 2.35 Intensity Cu 2.45 3.69 3.90 4.21 4.25 3.81 4.22 3.81 4.12 3.47 “'Mn/Ou 1.19 .70 .67 .42 '3 In 2 601. Ci? Log 1;. )‘ 7 1 OH / t -1 e / -2 10 40 60 100 200 400 Milli-micrOgrams manganese Fig. 8.- Working curve for the determination of manganese by the spark method. _Since it appeared possible to detect smaller amounts of manganese by the spark method as compared to the arc procedure used, other trace elements were investigated by the spark method. Cobalt was investigated using the line at 2582.2 Angstroms since this is the most sensitive spark line. (7) The solutions were prepared by dissolving Co (NO3)2.6H20 in distilled water; .4938 grams of the salt contains 100 mg. of Cobalt. The same procedure as used in the spark determination of manganese was used in this case. The data obtained from making exposure Of various concen- trations Of Cobalt are shown in Table IV. The working curve covering this low concentration of Cobalt is shown in Fig. 9. TABLE IV WORKING CURVE DATA FOR COBALT DETERMINATION BY THE SPARK METHOD Cobalt in Galv. Density Lo Intensit sample Reading LOg IO/I Co 2582.2; Cu 2 5.1 g. x 10'”9 Co Cu Co Cu Co Cu CO/Cu 360 16.0 21.8 .194 .060 5.92 4.13 1.79 300 16.0 19.6 .194 .106 5.92 4.86 1.06 240 19.5 20.8 .108 .080 4.90 4.48 .42 180 19.4 19.8 .110 .101 4u93 5.02 .09 150 21.7 21.6 .061 .063 4.14 4.18 -.04 120 21.3 20.5 .069 .084 4.29 4.55 -.26 100 22.6 21.4 .044 .068 3.81 4.26 -.45 60 23.2 21 .7 .033 .062 3.47 4.17 -.70 00 2 82.2 9 . 55 Log -2 10 20 4o 0 80 100 200 400 Milli-micrograms Cobalt Fig. 9.- Working curve for the determination of Cobalt by the spark method. In detecting the slightest amount of magnesium, some difficulty was encountered due to the limitations of the electrodes. Magnesium 2852.1 line could not be used in this work because there was enough magnesium present as im- purity in the electrodes to bring this line into the spectrum when no sample was in the spark gap. The next most sensitive 'spark line, which occurs at 2795.5 Angstroms, was used. The effect of the electrodes on this line was noticeable, but not to the extent of the 2852.2 line. The standard solutions were prepared by dissolving Mg(N03)2.6HéO in distilled water; 1.0549 grams of the salt contains 100 mg. of magnesium. The procedure used in preparation of the eleCtrodes and making the exposure was the same as that used in determining the sensitivity of manganese. The date obtained from observa- tion of the photographic plates used are given in Table V. The working curve for magnesium is presented in Fig. 10. TABLE V WORKING CURVE DATA FOR MAGNESIUM DETERMINATION BY THE SPARK METHOD Magnesium Galv. Density Log Intensity in sample Reading Log IO/I Mg 2795.5; Cu 2493.1 g . x 10‘9 Mg Cu Mg Cu Mg Cu Mg/Gu 30.0 18.6 22.7 .128 .042 5.18 3.67 1.51 24.0 20.2 22.7 .093 .042 4.68 3.67 1.01 18.0 22.0 23.6 .056 .025 4.04 3.16 .88 15.0 22.2 23.3 .052 .030 3.97 3.35 .62 12.0 20.5 22.2 .086 .052 4.57 3.97 .60 10.0 21.1 21.8 .073 .059 4.36 4.12 .24 6.0 19.5 19.5 .108 .108 3.75 3.75 .00 3.0 22.4 21.8 .048 .059 3.91 4.12 -.21 -2 l 2 4 6 8 10 20 40 Milli-micrograms Magnesium Fig. 10.- Working curve for the determination of magnesium by the spark method. On the basis of the working curves obtained, a series of analyses was run on soils and organic matter. The re- sults of the investigation of these samples are summarized in Table VI. The soil samples were prepared for examination by digesting .1 gram of the sample in 10 ml. of concentrated hydrochloric acid for 24 hours at room temperature. After digestion, .03 ml. of the clear solution was evaporated on the cOpper electrodes, and the sample sparked under the standardized conditions. The organic matter wastreated according to the method suggested by Hess, Owens, and Reinhardt. (8) One-tenth gram samples were ashed after the addition of 1 m1. of con- centrated sulphuric acid. When charring was nearly complete, concentrated nitric acid was added drOpwise to the hot solu- tion until a clear liquid was obtained. The solution was then evaporated almost to dryness, cooled, and taken up in .5 ml. of 2 N nitric acid. A constant volumne, .03 ml., of the solution was then evaporated on the electrodes and sparked after evaporation. Sample Soil 1/0-4 1/6-12 8/0-7 8/8-16 9/0-7 9/8-14 Organic '3777' TABLE VI SUMMARY OF DETERMINATIONS Manganese .045 .020 .027 .023 .031 .015 OF TRACE ELEMENTS Cobalt Magnesium Others % 0 Found Fe, B, Zn, Ba. ..---- * Same Same Same Same Same Ni, Fe. * Magnesium was present in all samples in amounts larger than those covered by the working curves. The sensitivity of manganese as determined by the arc method is in good agreement with the work done in other laboratories. (8)(9) The minimum amount of man- ganese which could be detected by the arc method was 120 milli-micrOgrams. It was possible to detect smaller amounts of man- ganese by using the spark excitation, so most of the interest was centered in this method. Amounts of man- ganese as small as 30 milli-micrOgrams could be detected by this method. Magnesium appears to be more easily detected than either manganese or cobalt. The actual minimum of Magnesium was not determined due to the presence of magnesium in the cOpper electrodes, but a working curve was obtained covering the range down to 3.0 milli-micrOgrams. The extreme leveling off of the magnesium working curve is due to the impurity of the electrodes. The determination of cobalt was possible to approximately the same extent as manganese; the least observable amount of cobalt being 60 milli-micrOgrams. It was not possible to detect quantitatively the amount of magnesium in the presence of cobalt. The co- balt greatly increased the appearent concentration of magnesium. This was the only case of interference noted in this investigation. It was observed that high relative humidity while the plates were being exposed caused poor results in that the sensitivity of the elements was reduced. Summary 1. The spectrographic sensitivity of manganese under direct current arc and alternating spark excitation has been determined. 2. The sensitivity of magnesium and cobalt by the spark excitation method has been observed. 3. Working curves for low concentrations of the three elements have been derived. 4. Methods of handling various types of samples have been presented. fi References Cited Bausch & Lomb, “Instruments for Spectrographic Analy- sis," Bausch & Lomb Optical Company, Rochester, N.Y., 1932, p. 42. Brode, W.R., ”Chemical SpectroscOpy,' 2nd Ed., John Wiley 8.30m, Inc., New York, N.Y., 1943, p. 36, 115. Ewing, D.T., Wilson, M.F., and Hibbard, R.P., Ind. Eng. 9315., Anal. Ed., g, 11.10 (1937) Twyman, F., "The Practice of Spectrum.Analysis with Hilger Instruments," 6th Ed., Adam Hilger, Ltd., London, Eng., p. 32. Ryde, J.W., and Jenkins, H.G., ”Sensitive Arc Lines of 50 Elements," A. Hilger, Ltd., London, Eng., l9##. Brandt, S.M., "The SpectroscOpic Working Curves for Magnesium and Other Elements in.Aluminum Alloys,” M.S. Thesis, Michigan State College, East Lansing, Michigan, 1943, p. 7. Harrison, G.R., ”M.I.T. lave Length Tables," John Wiley & Sons, Inc., New York, N.Y., 1939, p. 382. Hess, T.M., Owens, J.S., and Reinhardt, L.G., Proceed- $252 3; the Seventh Bummer Conference on gpectroscOpy and its Applicatigns, 29 (1939) Melvin, E.H., and O'Connor, R.T., ibid. H2 (1939) . .. v, , . 1‘13335, If 12;... 2......wflfifr5, $.19... :1», 1 a ,. . .slhrx; 41:5 .. 111."; 7.171.. . .. .n . V ..I I...» «LLH detail} us: x! 1.5%. pun .9.sz UL? L 2,. 1 L. L . . ..wwt .h, . in J ..... 12.? vii...) Lass...‘ . L :2 _.~nw.3.t..vnv:w7tf... 3.131... ,. , , .r. ,..,., . ,, . . '2': ‘33 ‘41 WENISHY Liking-ah, o 1.. u I :I: i Ii"; \‘N‘ . . fF“ U. "1"”;é r" 1W I‘ ‘._ 5’; ‘I'Iq '7," .-:h17 ' I I; . . _'I ' I 0- .' - I .l ".‘ 5:3 I: I - _ t 'r. I.', . I '. -.‘~ .,‘ - it . , . . I I. | I I ' s". ‘ II . . . \‘ ¥ .l ‘40. I I ,_ .' . ’ I . I | I, - I ' a ‘ I . - ‘ ‘ \IO' ,‘ I ‘ t ‘ O . . ,i .' I '3‘. A 'I I I . ‘ ' {-5 . V ::"‘ Ill‘t . . - .' . "I" .- .I ' _ I. I _ .' I. l _ ‘I , } I § . .\.I, M{. . I ll 7' I. I' 0| . l .9. U ‘ I ' ‘ I" ~'§. I ... . :_ I '. .3 ..- , .J ' .. 2.? “. .' ‘ H .I ' ‘r ' ,1; J 'I. ' ! VI. . If" I . xv ‘ "'.. I. I - " ‘0. I I I - I o 9' ‘I.’ 1 I . '_ _ l I . F ‘ ' I -- ' -_ I . I l.‘ . 'I‘l‘ ' . ' I 7' l'I 'er .- vI ';.". P . . - & -. I I a _- .V ‘c ... ._. _ . _ 1’ ,‘u ' O I o l '. ' C‘ V I ' ‘ . .‘ - I 3' o. I I- . . I. | . ‘ -" . " - . u _ I :H’ a :r C: : ’ - 1’. ‘ ,‘ .V fHLMlsTRY DE"! ’. O I... 1 ' . . ‘ . V ‘; ‘ I . I‘ 1.; H '9 ‘1 I h I . . H 5&1 ' ’4' l [I ~ I "I . ‘ .I .1 , . ,. - I- I.. . - . . . . . . . . . . ”I 1 .. - -. « .. ., , 1 089 Crane ‘ 9b ‘. .' 1 ° .31 - .- I I .' 4| . f . p o I .1 I ‘, (“III I‘. I ‘ ‘ I “. ‘I'. . I ' ‘ ' I I I 0‘ ' ' t I ' I ‘ ! Ii: . t I 1 .' I l‘ I . .I. ., 'III. , ‘II' I .‘ . , I I , ’ I. ’ :r ‘1‘ . \ . 0 . '. - . \- . j I C II I .. I r . v o _.‘ I - I . I .1 III I' a i J‘ I l ' - I' I. l‘ i c I; I , . - O | . | I I |I| ’f‘ 1" .“I Y‘ ‘ I I '- ' W I 5' ’ I" I". ’5‘ ‘.. I ' ."‘.‘ . " | cl -‘ ' . O I."l I I 'n. i I II . -" . ’ I I 1.- I“. ‘ I I. O I ' 1. I 1 u ‘ I . I I I . , x“ , I II x " o I" .’ I ’ I . I. . u ,,v‘ ".9 ’ ‘ .7.“ I. ‘34”. ~ . . . ' ~ -. I; I I' 4“ v " o.‘ : I . I'.'\\‘ .1 ' . .. s' .. F. '.l. . - I ‘ O _ a . _ l . l ’ ' ‘fu ‘" I,‘ '\Ie‘ ‘ 1 'V s “I V " ' I1 . I I I i ; A \ ' . ). I ‘ '- I. II . b. '. "I . r " "Ii .’ l , i‘. V F 76.: III I. . I g “ ‘1' I. .l. I- ‘ I‘ I I I . ' K ‘ , 9 _ I . y 'e- . k ‘ l . I ,' ‘ I H . ' b 4 “ . I_ . 0 II ‘ |\ , 0' I. I: = I r. I’I‘I I I I y ' :\ ’ 4 1 e I I '.I ' -- ,h _ . ' '. . I ‘ I < ' J! I. ‘ I.“ I. Jr“ ','., ' J .« . .1 ' , I ' I I . O. -,: ' ' I 1 ‘ ‘ \ . . I - ‘ I- . ‘ I n; ‘ 1 ‘ I I V“ J II J p , . I —"‘Vl'¥. '. I. z ' ' I -. ' 'I u‘ ‘I . I 'II ._ g I ‘ ,' :‘ , ' u ‘v‘ a .’ V ‘.I 1‘.‘ § 4 I a; E r u. - .. .. ' \‘t ,I L; '. ‘ I I 7‘] ‘ I \I I I MIMIIII\IIIIIIIIILIIIIII 3 1293 02446 8 33