777
V O L U M E 2 1 , NO. 7, J U L Y 1 9 4 9 equalizing valve was opened, these fluctuatio;is all but vanished; the remaining noise was attributed to the position indicator. The utility of the analyzer for continuous measurements over extended periods was impaired by the persistence of drift, which made it necessary to recenter the drop from time to t,ime by momentarily opening the equalizing valve on the detector. The direction and magnitude of the observed drift were variable and may have been caused by slow changes in the source temperature, as compensation of the detector itself was apparently good. Drifts equivalent to 4 p.p.m. per minute were sometimes observed. Inasmuch as the analyzer was not compensated for fluctuations of the source, stabilization of the source deserved more attention than it received. Considering the simplicity of this analyzer, and its ability to detect changes of less than 3% of a 300 p.p.ni. concentration of carbon dioxide, this analyzer was considered successful. HOKever, improved performance might be expected from an analyzer with short dead cells a t the ends of the sample cell for insulation purposes, high quality reflecting surfaces, and a more stable source, or perhaps reduction of source noise by an independent feedback circuit.
ACKNOWLEDGMENT
The writers gratefully acknowledge indebtedness to Van Zandt Williams of the American Cyanamid Company for benefits derived from conversations which contributed materially to ,the progress of the work, and for checking the transmission of the mica used for windows. LITERATURE CITED
Barnes, R . B., U. S. Patent 2,431,019 ( S o v . 18, 1947). Fastie, \\-. G., and Pfund, 9. H . , J . Optical Soc. Am., 3 7 , 762 (1947). Kidder, 11. E., and Berry, J. W., U. S. Patent 2,443,427 (June 15, 1948). Luft, K. F., Z.tech. Physik, 24, 97 (1943). McAlister, E . D., Phys. Rev., 49, 704 (1936). Pfund, A . H . , Science, 90, 326 (1939). Pfund, d.H., U. S. Patent 2,212,211 (-lug. 20, 1940). Pfund, A . H., and Gemmill, C. L., BulI. Johns Hopkiiis Hospital, 67, 61 (1940). White, J. U., J . Optical SOC.A m . , 3 7 , 713 (1947). Williams, T’. Z., R eu. Sci. Instruments, 19, 138 (1948). Wright. N., and Herscher, L. W., J . Optical Soc. .4m., 36, 195 (1946). R E C E I V EJuly D 30, 1918.
Automatic Recording of Polarographic Data Obtained with Platinum Electrodes L. B. ROGERS, H. H. MILLER, R. B. GOODRICH2, ~
N D 4.
F. STEHNEYS
Oak Ridge ,Vational Laboratory, Oak Ridge, Tenn. Polarographic analyses appear to be feasible using a platinum microelectrode and the usual automatic recording technique. The effects on the half-wave potential and the diffusion current of using different rates of polarization, larger electrode areas, and stirring have been examined. Under controlled conditions, the precision of the measurements is somewhat lower than for the dropping mercurj- electrode.
P
L A T I S U l I electrodes have been used for obtaining polarographic information by Laitinen and Kolthoff (4)who determined the reproducibility of data obtained by a manual procedure, and by Miller (8) n h o suggested a modified procedure in which the deposit was stripped from the solid electrode after each point. Because the method of obtaining a curve point by point is very time-consuming, it appeared desirable to esamine the possibility of employing the usual polarographic procedure involving a continuously changing potential. Zlotowski (15) and Walen and Haissinsky (14) used automatic recording with solid electrodes to obtain electrochen.ica1 information about reactions, but they did not esaniine the polarographic aspects of the procedure. After the present investigation had been started, Muller (IO)reported po arographic studies in which solutions of electrolytes flowed past a stationary platinum electrode in a glass capillary. From Matheson ( 7 ) it was learned that polarographic studies with stationar?. electrodes had been carried out by H. =i. Robinson of the DOT\-Chemical Company using an oscillographic technique (8). Since the completion of the study reported here, other papers (1, 11, 19) have appeared 1 Present address, Department of Chemistry, Massachusetts Institute of Technology, Cambridge 39, Mass. 2 Present address, Electronics Division, S a t i o n a l Bureau of Standards, Washington, D . C. 3 Present address, Department of Chemistry, University of Chicago, Chicago, Ill.
on the general subject of polarographic techniques applic~ab!eto solid electrodes. I n examining the possibilities of recording polarographic data continuously, it was necessary t o study variables such as the effect of changing the size of the electrode, the effect of stirring the solution, and the effect of applying voltage continuously at different rates of change. The criteria used to check the po‘arographic re’iabilitg of the data were the reproducibility of the general shape of the curve, of half-wave potentials, and of diffusion currents. The chief consideration for quantitative analysis vias the linearity of the relationship between the concentration of reducible ion and the diffusion current. Silver was selected for this study because it has a simple redox reaction which is thermodynamically reversible, because knowledge of such a system is necessary if more complex systems are to be understood, and because the reaction illustrates an application of the platinum electrode t,o the analysis of an ion which is reduced in the region of potential that is positive in relation to the saturated calomel electrode (S.C.E.). EXPERIMENTAL DETAILS
Reagents and Solutions. All chemicals used in the investigation were reagent grade, and all solutions were prepared with distilled water. The concentrations of standard stock solutions were known to better than 1%. Jl in silThe standard polarographic solution was 5.00 X
778
ANALYTICAL CHEMISTRY
ver ion and 0.100 ill in putasslum nitIate. The pli, adjusted by the addition of about 2 drops of 6 I\i nitric acid, was 4.00 * 0.01. Tank nitrogen, used for removing oxygen from the polarographic cell, was freed from oxygen by passing it successively through three bottles of ammoniacal cuprous chloride followed by a bottle of 6 A’ sulfuric acid. CApparatus. The platinum electrodes were made by fusing platinum wire into a soft-glass tube and making electrical contact in the usual way by filling the tube with mercury. The rotating electrode used in this study was similar to the one described by Laitinen and Iiolthoff. The electrode was coated with ceresin wax several times. followine: which the nlatinum was scrannrd to PXpose
I I I I I I I I I
II
I I I I I I I I
IIi Ill
I I
IIILlLU
4 2 0
W l
[Tsing a 6-mm. wire electrode (0.10 sq. cm. area) in qiiiet solution I. Manual 11. 1.46 mv./sec. 111. 4.38 mv./sec.
An agar bridge saturated with potassium nitrate conriected the polarographic solution with a saturated calomel electrode. In order to be able to make a correction for I R drop, the resistance of the cell was measured before each run with a conductivity bridge, Model RC-lB, made by Industrial Instruments, Inc. Polarograph Models XI1 and XX, manufactured by E. H. Sargent and Company, were used for continuous recording. Although it was possible to use these instruments to obtain “manual” current-voltage curves, these curves were usually obtained with a setup consisting of a Leeds & Northrup potentiometer, Rubicon galvanometer, and an Ayrton shunt. A thermostat was felt to be unnecessary for most of this work. The experimenots were carried out a t room temperature, which was usually 26 * 2’ C. Each of the series of studies concerning the relationship between diffusion current and concentration of the reducible ion was completed in quick succession, during which time the temperature of the room (and the solution) did not change noticeably-i.e., *O,l O C. However, a bath maintained a t 25.0” * 0.1”C. was employed in studying the effect of the voncentration of reducible ion on the half-wave potential. For certain portions of this study it was necessary to stir the polarographic solution during a run. Mild stirring was obtained by bubbling nitrogen through the solution. For faster rates of atirring, a motor-drlven glass rod with a 60” bend 2.5 cm. (1 inch) above the end was used. Its speed was adjusted to 365 r.p.m. by means of a Variac. Procedures. CLEAXISG ELECTRODES. Preliminary tests Jhowed that the electrode could be cleaned by imposing upon it a sufficiently positive potential, or by dipping it into 6 nitric acid for 2 minutes or more, washing i t with distilled water, and then
drying with Kleenex before using. Because the acid treatment was faster, it was employed throughout this study. DEAERATION OF POLAROQRAPHIC SOLUTIOKS.Although deaeration was not necessary, in that silver ion was reduced before dissolved oxygen, the step was carried out in order to eliminate a possible variable. POLAROGRAPHY. The manual technique of Laitinen and Kolthoff was employed to establish standard curves. When curves were recorded automatically, the rate of change of potential was calculated in each case from the length of time required by the slide-wire to change the applied voltage by one or more volts. RESULTS
General Comparison of Results from Manual and Continuous Runs. As shown in Figure 1,continuously recorded curves usually had round maxima, whereas maxima were rarely observed in manual curves. Although the presence of maxima seemed to be the rule in continuous recording, some runs, which were duplicates of others for which maxima had heen found, showed no signs of a maximum. Maxima were expected frum experience with the Laitinen and Kolthoff technique and they indicate that the rate a t which the system reaches “diffusion equilibrium” is sloiyer than the rate at which the voltage changes (3). Evidence for such an explanation is found in the observation that these maxima are not of the same type as those encountered with the dropping mercury electrode, in that the presence of gelatin exerts no repressive effect on them. If the maxima do owe their existence to a change with time in the flux of silver ion a t the electrode surface, one would expect slower rates of polarization to produce curves with small maxima. However, the irregular occurrence of maxima prevented the authors from reaching a quantitative. conclusion concerning the effect of the rate of polarization on the height of the maximum. In Table I are listed the average values for the half-wave potential and the diffusion current obtained with three sizes of electrodes. Values listed in this table have been corrected for I R drop and residual current and represent the average of six or more runs. The instrumental limits of accuracy were considered to be about *0.01 volt and *0.1 microampere. However, the reproducibility limits of the experimental measurements were usually wider. Thus, although most of the values for half-wave potentials (in quiet solutions) fell within *0.01 volt of the average, some varied by 0.02 volt. Disagreement among the values for the diffusion currents was more marked, and the differences increased with the size of the average diffusion current. Table I shows that the half-wave potential is independent of the rate of polarization. The slightly positive shift observed for the 6-mm. electrode is within the limits of experimental error. The half-wave potential for a continuously recorded curve obtained by using the opposite direction of polarization also agreed within the limits of error with the values listed in Table I. Although faster rates of polarization did not affect the half-wave potential, they did produce larger diffusion currents in some instances. Effect of Electrode Size. The size of the electrode does not affect the half-wave potential. As one would expect, an increase in the electrode area appeared to produce no important change in the shape of the polarographic wave after it had been corrrctd for I R drop.
‘rahle I.
1. 2. 3.
Half-Wave Potentials a n d Diffusion C u r r e n t s of Silver i n Q u i e t Solutions R a t e of Applied Voltage Manual Ev2, id, volt Ma,
1.46 Mv./Sec.
Eva.
id, pa.
2.92 Mv./Se,c. Eira, t d , volt pa.
4.38 Mr./Se,c. El/%, Id, volt pa.
volt 1-mm. electrode 0.8 0.33 0.8 0.34 1 . 0 .. .. . (0.02sq. om.) 0 . 3 4 6-mm. electrode 3 . 0 0.35 3.9 0.35 4 . 1 (0.10sq.em.) 0 . 3 3 3 . 2 0 . 3 3 Foil electrode . . 0.33 32.7 (2.14sq. cm.) 0 . 3 4 2 9 . 4 0 . 3 4 3 2 . 7
.
.. . .
V O L U M E 21, NO. 7, J U L Y 1 9 4 9
779 Effect of Stirring. One would expect the results from stirred solutions to be similar to those obtained either with a rotating electrode or by flowing a solution past the stationary electrode, because in each case there is a movement of the solution past the surface of the electrode. This expectation was verified by the absence of maxima as well as the general results described beloa Stirring markedly decreased the reproducibility of both the half-wave potential and the diffusion current. In many cases, no half-nave potential could be calculated because of failure to obt ain a diffusion current plateau. T h e n half-wave potentials could be determined, the values were often spread over a range of 0 06 volt and the average value appeared to be somewhat lower than that for unstirred solutions. In general, these findings were i*onsistent with those of Muller (10) and TAtinen and Kolthoff (
.5),
Faster rates of stirring not only produced larger diffusion curents but also decrea>ed the probability of finding a diffusion current plateau. The probability for a particular electrode and a tixed rate of stirring could, hoRever, be increased by using a faster rate of polarization. These trends are consistent with studies of the rotating electrode ( 5 ) . Relation between Concentration of Reducible Ion and Diffusion Current. One set of results is shown in Figure 3, where each point represents the minimum diffusion current, xhich was usually constant for more than 0.2 volt, for a single determination ittempts to duplicate these curves were very succesbful. For iwrrents greater than 3 microamperes, differences betneen two values for a given solution were usually less than lo%, often 5%. For currents less than 3 microamperes, agreement between duplivates nearly always fell within the 0.1 microampere limit mentioned. Thus, the relationship between diffusion current and voncentration of silver was linear but the precision was not as high as that obtained with the usual dropping mercury electrode Relation between Concentration of Reducible Ion and HalfWave Potential. Deposition of an element on a solid electrode is analogous to the deposition of an element insoluble in mercury on the usual dropping mercury electrode. Therefore, a shift of the half-wave potential to more negative potentials with dilution of the reducible ion should be found ( 2 ) in accordance with the equation I
I'iyiirt-
2. Relation between Electrode Size and Diffukiiltl Current in Quiet Solution 0 Foil electrodes 0 Wire electrodes Rate of polarization. 1.46 rnv./ser.
W
520
a I
0
LT
/
/ O
I
Ell2 =
Figure 3.
Relation between Concentration of Silver Ion and Diffusion Current I. 1-mm. wire (0.02 sm. om.) 11. 6-mm. wire (0.10 s i . 0m.j 111. Foil (2.14 aq. em.) Rate of polarization. 1.46 mv./sec.
Difficulty was experienced in obtaining a good curve showing the known relationship between diffusion current and electrode mea, despite care in measuring and cleaning each electrode. Therefore, Figure 2 served only to substantiate the expected rrend. This figure illustrates the error that might be introduced in trying to predict the size of a diffusion current for an electrodr by comparing its plane area to the area of a second electrode. I t emphasizes the need for calibrating each electrode with a standard Jolution. Likewise, it is advisable to determine the residual current for each electrode individually. A comparison of the curves on Figure 3 emphasizes the desirability of selecting an electrode area according to the region of concentration that is being examined-smaller electrodes are more juitable for higher concentrations and larger electrodes for lower ,oncentrations of reducible ions.
E"
C8j8 pRT + ICT In -- + -In nP 2 nF
C,F,
where subscript s refers to the reducible ion, subscript z t o the Itomplexing ion, p to the number of coordinating groups, C to the molar concentration, and f to the activity coefficient of the ion. The activity of the deposit is assumed to be unity. The sign of the potential in Equation 1 is reversed from the convention in Latimer ( 6 ) , so that it agrees with the sign of the polarographic electrode. In polarography, where the concentration of the reducible ion is almost always less than 1% of the total ionic strength, fa is essentially unaffected by differences in the concentration of the reducible ion in two otherwise identical solutions. The third term need not be considered, providing the concentration of complexing agent is held constant. Similarly one can ignore corrections for liquid junction potentials, overvoltage, etc., by assuming that these factors will remain constant. A test of Equation 1 can, therefore, be made easily. Table I1 shows that the change in the half-wave potential agrees very well with the value of -0.59 volt predicted by the equation. Because of this shift, it is essential to specify the concentration of an ion whenever the half-wave potential is reported for a reaction in which one component is deposited. The average half-wave potential listed in Table I (0.33 volt) does not agree with the value (0.29 volt) calculated for a 5.0 X lo-* M solution from Table 11. It is felt that the latter is a better value because the analyses were obtained under more carefully controlled conditions.
ANALYTICAL CHEMISTRY
780 Table 11. Effect of Changes i n Concentration of Silver Ion on Half-Wave Potential Concentration, .U 1 , 0 0 0 x 10-2
( I n 0.1 .V potassium nitrate a t p H 4.00) Ed2 Trial 1 Trial 2 .Iverage 4-0.372 +0.373 4-0.374
1.000 X
t0.316
+0.322
f0.319
1 . 0 0 3 X 10-4
+0.275
+0.289
+9,232
r
18
I
I
1.6
14
(For ferric t o ferrous reduction in 0.1 .M hydrochloric acid) Concentration, El/? 21 Trial 1 Trial 2 Average A E112 1 . 0 0 x 10-2 to.514 +0.505 +0.510 +0.003 1 . 0 0 x 10-3 +0.520 +0.506 +0.513 +0.006 1.00 x lo-' +0.527 +0.511 +0.519
'2
-0.053
-0.057
I I I I IO 0.8 06 04 VOLTAGE VS. S C.E.
I 12
AE1
Table 111. Effect of Concentration on Half-Wave Potential
I
I
02
0.0
I I - 0 2 -0.'
OB
aero potential and going toward positive potentials until oxygen was evolved. Compositive curves were constructed for each electrode, and are shown in Figure 4 as curves I and 11. If, instead of starting a t zero potential and going negative, the start was made a t a potential where the oxygen was being evolved, the curve for the stationary electrode did not coincide n i t h the one described above. Throughout most of the region of positive potentials the residual current in the latter case was of the same order of magnitude but opposite in sign to that found when the potential was applied in the opposite direction. Curve I11 in Figure 4 is a representative example. The sign of residual current for the rotating electrode also depended upon the direction of polarization, but the differences were less conspicuous because the residual current was much smaller, on the order of 0.1 niicroanipere or less. Another comparison of the tTyo types of electrodes was made by recording curves for each electrode under conditions of an actual analysis. One set of curves for the reduction of silver ion is illustrated by Figure 5. (The curves have not been corrected for ZR drop.) In a similar study of the reduction of ferric ion (to ferrous) in 0.1 JP hydrochloric acid, the shapes and sizes of the curves corresponded closely, the only difference being a more nearlv symmetrical S-shape for curve 111.
The results indicate that although the particular stationary electrock used in this study gave a larger diffusion current than the rotating electrode, the curves from the rotatObtained with stationary platinum electrode and rotated platinum microelectrode i n 0.1 M HCI ing electrode were easier to interpret because of the smaller I. Rotated microelectrode, potentials started a t zero; 0.050 sq. c m . area, residual current, the sharper initial break in the deposition 600 r.pm. 11. Stationary electrode, 2.1 sq. c m . area, potentials started a t zero curve, and the usual absence of a maximum. The half111. Stationary electrode potential started a t approximately 1.3 volt wave potentials were within *0.01 volt of the theoretical reversible value for 5.0 X -11silver in a 0.1 JI solution of potassium nitrate. A shift in half-wave potential with dilution might not result if Maxima unexpectedly appeared in some of the curves of the a soluble ion were produced. I n the general equation rotating electrode. Many of the masima nere relatively large, and they were often immediatelv preceded by an inflection in the RT CO= + & + ET ln ,9) El,* = E" +-In nF Crad.flpd nF KIcd ?tF (Cf)*\-/ polarographic xave as shon-n in Figure 6. .ilthough this masiFigure 4.
Comparison of Residual Currents
~
KT
(sf)_"
the last two terms can usually be disregarded in the light of considerations discussed in connection viith Equation 1, and the remainder of the equation can be simplified by assuming that, a t El ), C,, equals Crtd andf,, equals.f,,d. The results given in Table I11 show a slight drift of the half-wave potential toward more positive values for the more dilute solutions of iron. The drift may be a real one resulting from the simplifying assumptions that were introduced, but poor reproducibility of the measurements limits the validity of such a conclusion. Unfortunately, the attempts to measure a 1.00 X 10-5 JP solution were unsuccessful, Comparison of Curves Obtained with Large Stationary Electrode and Rotated Microelectrode. Electrodes of larger area are more suitable than microelectrodes for analyses of dilute solutions Because a rotated electrode, compared to a stationary electrode, offers a similar increase in sensitivity, the following study was carried out to determine which approach was better. First a comparison was made of the residual currentsi.e., blank runs-for the two electrodes in 0.1 JI hydrochloric acid by starting a t zero potential and proceeding toward negative potentials until hydrogen was evolved. An analogous wave as obtained by starting a t
I
0.5
I 0.4
I 0.3
I
I
0.2 0.1 VOLTAGE VS. S.C.E.
1 0.0
I -0.1
Figure 5. Comparative Curves Obtained with stationary platinum electrode and rotated microelectrode for solution of 5.0 X 10-8 M AgNOa i n 0.1 .M KNOa and 0 . 0 1 7 ~gelatin a t pH 4.0 I. Stationary platinum electrode, 2.1 a q . c m . area 11. Blank (residnal current) with stationary electrode 111. Rotated p l a t i n u m microelectrode, 0.050 sq. c m . area, 600 r.p.m. IV. Blank (residul current) with rotated microelectrode
V O L U M E 21, NO. 7, J U L Y 1 9 4 9
781 dropping mercury electrode, has been esamined rTith a platinum electrode. These studies \Till be reported later. Solid electrodes simplify analyses in solutions containing hydrofluoric acid because it is possible to use a plastic cell to hold the solution and to coat, with wax, the glass in which the platinum electrode is embedded. It appears plausible to expect a solid electrode to be useful for studies in liquid ammonia and similar solvents. ACKYOWLEDGXIEYT
The authors are indebted to D. E. Ehrlinger for making some of the measurements reported in this paper and to Kathryn Odom for helping to prepare the manuscript. Work performed for Atomic Energy Commission under Contract K-35-058 eng. 71 r5ith Nonsanto Chemical Company and Contract 7405 eng. 26 viith Carbide and Carbon Chemicals Corporation. LITERATURE CITED
05
I -04
I
I
02
-0I VOLTAGE VS. S C E
-07
Figure 6. Polarographic Maximum Obtained with rotated platinum microelectrode i n solution containing 5.0 X 10‘6 M AgNOa i n 0.1 M KNOI and 0.01 %gelatin a t pH 4.0. Rate of polariaation, 4.65 mv./sec.
mum was undoubtedly connected with the discharge of silver, it was probably the result of a brief discharge of hydrogen. Curve I1 in Figure 4 shows a similar but less striking masimuni which was probably due to osygen evolution. Applications. The fact that polarographic procedures offer a rapid method for the determination of normal and formal (13) potentials has been known for some time ( 2 ) . If a solid platinum electrode is used instead of a dropping mercury electrode for studying deposition reactions, calculations can be simplified because there is no longer a need for considering displacement of the wave as a result of amalgam formation. The recording of curves for ions other than silver has been tested and found to be as satisfactory as for silver. Thus, it is possible to use the usual polarographic recording technique to make qualitative and semiquantitative analyses for ceric, permanganate, and dichromate ions whose analyses have heretofore been complicated by the spontaneity of their reactions wi h mercury. Similarly, oxidation of hydrazine and chromic ion, which cannot be performed because of the anodic dissolution of the
Harris, E.D., and Lindsey, -4. J., Sature, 162, 413 (1948). Kerlinger, H., in Kolthoff, I &I., and Lingane, J. J., “Polarography,” p. 155, New York, Interscience Publishers, 1941. Laitinen, H. .i., and Kolthoff, I. M., J . Am. Chem. Soc., 61, 3344
(1939). Laitinen, H. h.,and Kolthoff, I. &I., J . Phys. Chem., 45, 1061 (1941). Ibid., 45, 1079 (1941). Latimer, W.M.,“Oxidation Potentials,” X e w York, PrenticeHall, 1938. Matheson, L. A , , private communication. Matheson, L. A., and Nichols, K.,Trans. Am. Electrochem. Soc., 73, 193 (1938). *Miller, S. D., Trudy Vsesoyicz. Konferentsii Anal. Khim., 2, 551 (1943). -Muller, 0. H., 110th Meeting, AMERICAK CHEMICILSOCIETY, Chicago, Ill., Sept. 9 to 13, 1946; J . Am. Chem. SOC.,69, 2992 (1947). Randles, J. E. B., AnaZgst, 72, 310 (1947). Skobets, E. M., and Kacherora, S. A.. Zaaodskaya Lab., 13, 133 (1947). Swift, E. H., “System of Chemical Analysis,” p. 50, New York, Prentice-Hall, 1940. Walen, K.J., and Haissinsky, If., J . phys. radium (7), 10, 202 (1939). Zlotowski, I., Roczniki Chem., 14, 640, 651, 666 (1934). RECEIVED1Iay 17, 1948. Presented in part before t h e E a s t Tennessee Section of the AMERICAN CHEXICAL SOCIETI. Knoxville, Tenn., J l a y 24, 1947, and later in more detail before t h e Division of Analytical and Micro SOCIETY, Chemistry a t t h e 113th Meeting of the ERICAS AS CHEMICAL Chicago, Ill.
Chemicotoxicological Examination ’of Foods A Systematic Procedure N. I. GOLDSTONE, D e p a r t m e n t
of H e a l t h , C i t y of New York, A‘. Y .
A systematic procedure is presented for the isolation and detection of 26 common toxic substances, comprising the four groups: volatile, metallic, alkaloidal, and nonalkaloidal. Sensitive analytical methods and confirmatory tests fitted into a methodical sequence are described in detail.
T
HE branch of the science of toxicology that deals with the isolation and detection of poisons is divided into two major classifications: the forensic, involving a crime either premeditated or accidental, to which scientific principles first began t o be applied during the eighteenth century; and industrial poisoning, the object of aggressive social and scientific attack only for the past 50 years. A number of adequate tests (3-5, 8-10, 14-22, 25-27, 30) have been published in both categories, and these include precise chemical methods for the detection and estimation of a large number of toxic substances. I n the forensic field the emphasis is naturally placed on the examination of the cadaver a n d body organs, with a chapter incidentally devoted to a discus-
sion of the detection and identification of poisons in foods and related products. The latter subject is constantly receiving more attention in connection with the growing interest in the related field of trace elements in food ( 7 ) . Most textbooks present the subject in a tabular form incorporated within the framework of the particular system of classification preferred by its author, describing the various members of each group with their properties and physiological responses, in addition to analytical methods for detection and estimation. But nowhere in the literature is there to be found a simple, comprehensive, and systematic analytical procedure whereby the chemist without any considerable experience in this particular field is
