Review - Pesticides - American Chemical Society

---

1

I

Pesticides

J. L. ST. JOHN State College o f Washington, Pullman, Wash.

M

ORE papers on analytical methods for pesticides hare appeared in the last 2 yearb than in the preceding 3

years (121). The major part of thew methodr are micromethods, primarily for spray residue in fresh processed fruits and vegetables, as well as in fresh products, and for toxicity study. This indicates increased emphasis placed on safety by industry and by state and federal governments. The following review of methods perhaps shows a greater increase in emphasis on herbicides than any other type of pesticide, in accord with the interest t h a t has been developing. Macromethods are available through the Methods Clearinghouse of the Association of American Pesticide Control Officials, A similar clearinghouse for micromethods should be established. Methods for the various pesticides are also included in “Condensed Data on Pesticides,” first called “Pesticopoeia,” the third edition of which will soon be published ( 4 ) . Methods are presented to the Pesticide Subdivision of the AMERICLK CHEMICAL SOCIETYa t each semiannual meeting. hIany such methods were presented a t the Milwaukee and Yew York meetings in 1961, and were scheduled to be published in a volume of the Advances

i n Chemistry Series. The tolerances published in the Federal Register for October 20, 1954, became official on December 20, 1954, unless challenged, in conformity with the requirements of the very recent law supplementing the requirement for tolerances in the 1938 law. Some chemicals will have a zero residue tolerance; others will require no tolerances: while specific tolerances are establiqhed for others. CHLORINiTED HYDROCARBORS

ilgazzi, Peters, and Brooks ( 1 ) discuss combustion techniques for the determination of residues of highly chlorinated pesticides, The chloride ion is determined by amperometric titration. Applied to the determination of aldrin and dieldrin, it was used at levels of 2 to 100 y and detected 3 to 400 p.p.m. of aldrin in fat. An accuracy within =k4 y of pesticide is indicated. Helmkamp, Gunther, Wolf, and Leonard (64) present a direct potentiometric method for the chloride ion applicable to residues of chlorinated insecticides in foodstuffs. They use a pH meter and a silver-silver chloride us. calomel electrode system. The system is reducible with 0.02 p.p.m. of the chloride ion and appears usable up to 10,000 p.p.m. Gordon (40) developed a colorimetric microdiffusion method for the determination of chloride in chlorinated insecticides-an improved and more sensitive modification of the Conrvay method. The chloride is oxidized by permanganate and decolorizes the dye, Fast Green. Residues of 1 to 10 p.p.m. of chlorinated insecticides in biological material can be determined. The limit of detection of chlorine is about 0.1 y or, with a Beckman spectrophotometer, about 0.02 y of chlorine. The chlorinated hydrocarbon may be partially identified by a calculation of reaction rates. From 10 to 100 p.p.m. may be determined in oils and fats Mitchell (93)developed a method for the separation and identification of chlorinated organic pesticides by paper chromatography. I n mixtures of the alpha, beta, gamma, and delta isomers of benzene hexachloride, as little as 10 y of each isomer may be clearly separated and identified. The same author ( 9 4 )developed a modification of this method for use with other halogen com-

654

pounds, including aldrin, isodi in, dieldrin, and endrin. This test appears positive for various halogen pesticides including chlordan, D D T isomers, heptachlor, methoxychlor, and toxaphene. Noynihan and O’Colla (100) described the application of paper chromatography to the analyses of chlorinated organic insecticides. They have studied the separation of the isomers of benzene hexachloride, (BHC, 1,2,3,4,5,6-hexachlorocyclohexane) and of D D T and are developing a quantitative method of determination, O’Colla (102) described a method for the analysis of chlorinated organic insecticides by partition chromatography on paper and on cellulose columns. He separates and determines benzene hexachloride in chlordan in commercial samples. BENZENE HEXACHLORIDE

Schechter and Hornstein (125’) utilized the dechlorination of benzene hexachloride in developing a colorimetric method for its microdetermination. The method can be used to determine as little as 5 -, of benzene hexachloride and may be used for spray residue and pharmacological investigations. I t has been used in the range up to 100 y. Recoveries up to 100% have been obtained with the gamma and alpha isomers, but smaller recoveries with the delta and some other isomers. Recoveries of the gamma isomer added to plant, ariimal, and vegetable fats and oils and soils have been 95 to 100%. Chlordan appears to interfere, although other common chlorinated hydrocarbons do not show interference. Hornstein and Sullivan (61) described a method for the determination of 0.1 y of lindane (BHC) per liter in air, using the Schechter-Hornstein method (12-9, and measuring the color photometrically. Five-tenths microgram of lindane can be determined, with a precision within about &2%. A minimum of equipment is required and the method is comparatively rapid. Hornstein (69) found two colorimetric methods (107,123)suitable for determination of benzene hexachloride residues, although they may not be useful for the determination of the gamma isomer in formulations and technical grades of benzene hexachloride. Vhile the infrared method may be used for technical grades, it is costly. The partition chromatographic procedure appears to be most widely used for determination of the gamma isomer. Klein ( 7 3 ) preferred the Schechter-Hornstein method for the determination of benzene hexachloride and its beta isomer. Klein (76) discussed the Schechter-Hornstein colorimetric method for the determination of benzene hexachloride in foods and suggested the need for additional study. Reith (113) discussed a modified SchechterHornstein method, sensitive to 1 p.p.m., for vegetable material. Phillips (107) described a method for the colorimetric detection of microgram quantities of benzene hexachloride residue in foods, including peanuts and various fruits and vegetables. The method, which is readily adapted to routine quality control, utilizes extraction with aniline and color measurement with an Evelyn colorimeter. It covers a range of 1 t o 140 y of benzene hexachloride and shows high recovery in the range of about 20 to 100 y. Harris (50), who recently replaced J. J. T. Graham, an AOAC referee for 33 years, recommended certain methods and revisions and a further study of other methods for determining pesticides. These resulted from a report on benzene hexachloride by Hornstein (60), parathion by Giang (%), pyrethrins by Kelsey (679, rotenone by Payfer (106), dieldrin by McDevitt (86), and

V O L U M E 27, NO. 4, A P R I L 1 9 5 5 rodenticides by LaClair (80). Graham ( 4 3 )reported on a selies of pesticide chemical.. Rosin and Radan (120) have developed a method for the determination of lindane in commercial benzene hexachloride. They revien briefly the different types of methods which have been presehted for this purpose, and call attention to difficulties with commercial samples. The method gives results within 5 5 % of the gamma isomer, compared with polarographic determinations a t levels of 10 to 30% of the gamma isomer. Willermsin (144)applied the Thorp solubility method to the determination of the gamma isomer in technical benzene hexachloride. Hornstein (58) presented a method for the determination of technical benzene hexachloride in peanuts and soils, designed to eliminate inteiferences previously encountered. In many instances analyses can be made directly o n the sample. A methylene chloride extract is used on unroasted peanuts, but on roasted peanuts a correction is required. This colorimetric method can, in general, he applied directly to soils. Streuli and Cooke (135) described a sensitive polarographic method for the determination of the gamma isomer of benzene hexachloride in the presence of other isomers and other chlorinated compounds. The method has been applied t o pure benzene hexachloride (lindane), natural isomeric mixtures, concentrates, dusts, and other products. The gamma isomer in the range of 10 to 40% was determined. Craig and Tryon ( 2 1 ) described a n isotope dilution method for the determination of the gamma isomer in technical grade benzene hexachloride, with a standard deviation of 50.2%. The method is specific. and has been used for samples containing 1 to 50% of gamma isomer. Riemschneider, Claus, and Schneider (116) have developed a method for the determination of heptachlorocyclohexane in crude benzene hexachloride, by reduction with zinc and measurement of the index of refraction. Riemschneider and Stuck (117) describe a polarimetric method. DDT

Mattson, Spillane, Baker, and Pearce (89) developed a method for the determination of D D T and the degradation product, DDE, in human fat. -2modification of the Schechter-Haller method was standardized t o make a differential determination of D D T and D D E in quantities totaling 5 -, and rough estimations of 2 y. These compounds ranged from 0 to 80 p.p.m. of DDE, representing 40 to 85% of the total in human fat. D D T was not found alone. Berck (11) adopted the Schechter-Haller method t o the microdetermination of D D T in river water and suspended solids. DDT as low as 0.003 p.p.m. in suspended solids was determined after solvent extraction and removal of interferences chromatographically. Nitchell(97) described the separation and identification by paper chromatography of the ortho-para and para-para isomers of DDT, rhothane [ 1,l-dichloro-2,2-bis(pchloropheny1)ethanel ( D D D and T D E ) , and methoxychlor. ALDRIZT-DIELDRIN

O’Donnell, Neal, Tf’eiss, Bann, DeCino, and Lau (103) describe methods for the determination of 1,2,3,4,10,10-hexachloro1,4,4a,5,8,8a-hexahydroendoexodimethanonaphthalene (aldrin) in crop materials and aldrin residues; 0.1 p.p.m. in crop materials is detectable. Aldrin is extracted, separated by adsorption chromatography, and determined by the photometric method of Danish and Lidov or by the chlorine determination procedure of ilgazzi, Peters, and Brooks (1). Accuracy on plant extracts is a few hundredths part per million on a crop material basis. A photometric method has a high degree of specificity for aldrin. Blanks may be in the range of 0.05 p,p.m. Interferences are eliminated, although mint, for instance, was not successfully analyzed a t the 0.1 p.p.m. level. Interferences of other pesticides were studied. Mitchell and Patterson (98) present a method for the separation and identification of as little as 20 y of aldrin and dieldrin by paper chromatography. Mitchell ( 9 6 ) describes a paper chromatographic method for the separation and identification of the four

655

components: aldrin, isodrin, dieldrin, and endrin. Beckman (8) presents a method for the determination of aldrin and dieldrinin forniulations in the presence of D D T and sulfur, utilizing partio of 0.02- t o 0.2tion chromatography. It has 98 t o 1 0 0 ~recovery gram quantities. Ewart, Gunther, Barkley, and Elmer ( 2 7 ) describe a micromethod for a determination of dieldrin residues. They utilized and improved a method of high precision and sensitivity, which would detect the presence of 30 y of organic chloride or 0.1 to 0.2 p.p.m. of dieldrin. DILAN

Mitchell ( 9 6 ) continued a series of articles on the separation and identification of organic pesticides by paper chromatography, and described the separation and identification of the components of dilan, including primarily prolan and bulan. Jones and Riddick (66) present a procedure for the separation of certain organic insecticides from plant and animal tissues. h procedure for the isolation of dilan from interfering biological materials was developed by the addition of extraction to the previous method, The method has been used t o separate microgram and milligram quantities of DDT, lindane, parathion, chlordan, and methoxychlor from plant and animal tissue, with an accuracy within 15%. Jones and Riddick ( 6 4 ) present a method which is specific for nitroalkanes and thus specific for dilan, which contains two nitroalkanes in a 2 to 1 ratio. The method has been used for residue on plant and animal samples and is accurate for quantities as low as 50 -(. METHOXYCHLOR

Kunze ( 7 1 ) reported on a simplified procedure for the determination of methoxychlor in fruits and vegetables, dairy products, and biological materials, which he believes is specific for methoxychlor and sensitive to 1 p.p.m. Claborn and Beckman (16) describe a micromethod for the determination of methoxychlor in milk and fatty materials, as well as nonfat feeds and foodstuffs. The method is specific for methoxychlor, uses nitration by fuming nitric acid and colorimetric determination using sodium methylate. D D T and TDE in Emall amounts do not cause significant interferences. With quantities from 25 to 100 y j recovery is high. Jennings and Edviards ( 6 3 ) present a method for the determination of methoxychlor on paperboard, utilizing ultraviolet spectrophotometry. The interfering substances, different from those found in foods, were eliminated. Determinations were made within a range of 0.04 to 0.14 gram per liter of cyclohexane extract, with a precision nithin i 5 % . Recoveries of about 93% were obtained vhen methoxychlor in the range of 50 to 100 mg. per square foot was added to paperboard. ORGAVIC PHOSPHORUS COMPOUKDS

Schonamsgruber (124) described a method for the potentiometric determination of parathion or methylparathion in formulations by titration and back-titration after saponification, using a glass electrode. Zeumer and Fischer ( 1 5 2 )determined parathion by reduction with zinc, diazotization, and reaction Tvith 1-naphthol. Giang (56) reported on methods for parathion and recommended that the potentiometric end point technique be studied further. Wilson, Baier, Genung, and hhllowney (145)proposed a modified semimicromethod for the determination of parathion in citrus oils. The end point was determined spectrophotometrically. Xorris, Vail, and Averell (101) presented a method for the colorimetric estimation of S-(lJ2-dicarbethoxyethyl) 0,O-dimethyl dithiophosphate (malathion) residues. Malathion is removed by extraction, decomposed by alkali, extracted, and determined colorimetrically. Amounts between 0.25 and 2.5 mg. in 100 ml. of extract may be readily determined. The method is being adapted to the technical product and t o commercial formulations and has been utilized with a large number of fruits and vegetables. It has given 60 t o 100% recovery of surface residues

ANALYTICAL CHEMISTRY

656 on fruits and vegetables and 60 to 00% recovery of residues on chemical products. lTetcalf (90) applied the cholinesterase method to colorimetric microestimation in poisoning by organic phosphate insecticides. The method is said t o be rapid and precise on finger puncture quantities of blood. Cook (18)reported on an enzymatic method for organic phosphate and for carbamate insecticides and ( 1 9 ) discussed enzymatic methods, presenting a method for denleton. Gardner and Heath ( 3 5 ) have presented a chromatographic method for the quantitative determination of isomers of 0,O-diethyl ethylmercaptoethyl thiophosphate in demeton. Giang ( 3 8 ) discussed systemic insecticides, including methods for schradan, demeton, and others, and recommended further study. Hartley, Heath, Hulme, Pound, and Khittaker (51) described a method for the determination of schradan. David, Hartley, Heath, and Pound (83) presented a method for the determination of the toxic residues of schradan, which is sensitive to 1mg. per kg. March, Metcalf, and Fukuto (88) discussed a paper chromatography method for use with systemic insecticides. Two separate techniques have been used in the separation and identification of the components of technical demeton, and a third method has been used for technical schradan in translocation studies with demeton and for residues in fruits, nuts, and vegetables. FLUORINE

Remmert, Parks, Lawerice, and McBurney (114) describe a method for the determination of fluorine at a level of 1 to 75 p.p.m. in plant materials, primarily forage samples. They present a procedure to overcome calcination and difficulties in previous methods and indicate that the fluorine found with the AOAC method may be related to the weight of sample. Silica may interfere. Killard and Horton (143) present a fluorometric method for the determination of traces of fluoride, which estimates don-n to 0.2 mg. of fluoride with good precision. Liddell (84)presents a note on the colorimetric determination of small amounts of fluorine. Venkateswarlu and Rao (139) estimated from 30 to 120 p.p.m. of fluorine in tea. AIacIntire, Jones, and Hardin ( 8 7 )made a critical study of fluorine losses in the calcination of analytical samples of fluoride-enriched soils. Harrigan (48)reported on a qualitative test for fluorine in foods, and later (49)recommended a qualitative test for fluorine which involves quenching of the oxine fluorescence. PYRETHRINS

Levy and Estrada (83) described a rapid colorimetric method for the determination concentrations of pyrethrins down to about 0.04 mg. per ml. in extracts of pyrethrum flowers. The method has also been applied to oil-based insecticides. It gives good reproducibility, unlike earlier methods. Cueto and Dale ( 2 2 ) presented a colorimetric method for the estimation of small amounts of pyrethrins, allethrin, and furethrin, which showed a sensitivity of about 200 y and an experimental error of about 4%. The method may be used for spray residue and for the analysis of concentrates and formulations, but cannot be used for one of these pesticides in the presence of one or both of the others. As a micromethod it was used in the range of 15 t o 40 mg. Schreiber and McClellan ( 1 2 5 ) described a micromethod for the qualitative estimation of traces of pyrethroids. llicrogram quantities were determined in impregnated paper or cloth bags. Skukis, Christi, and Wachs (130) described a method for the spectrophotometric evaluation of pyrethrum, which is used to determine total pyrethrins in both flowers and eytracts. It is rapid and reproducible (1.0%) in a range of 1 to 20%. Hogsett, Kacy, and Johnson ( 6 7 ) developed a more precise method for the evaluation of both large and small percentages of allethrin in commercial allethrin, which may also be applicable to formulations. Konecky ( 7 6 )reviewed methods for allethrin which are in use and under development and recommended further

study of the h) drogenation and the ethi-lenedianiine methods of determining technical allethrin. Feinstein (28) described a new reaction and color test for allethrin and prrethrins. Mitchell ( 9 9 ) presented a note on recent rlOdC changes in the method for pyrethrins. He helieves that the cancellation of recent changes was unfortunate and should be reconsidered, and presents evidence that those changes were an improvement in the method. Kelsey (66, 6 7 ) reported on pyrethrins and recommended that approval of the modification of the official procedure be rescinded. C4RB43l.IATES

Gard and Rudd ( 3 4 )presented a micromethod for isopropyl S(3-chloropheny1)carbamate (CIPC) in soils and crops. They separate by extraction, hydrolysis, and steam distillation, and measure photoelectrically. Average recovery is S9%, and the loner practical limit of sensitivity is 0.05 p.p.m, of the herbicide. This method has been used n i t h cotton seeds, vegetables, and other crops, with a precision a t the 0.05 p.p.m. level of &0.016. Lowen (85)presented a modified technique for the carbon disulfide evolution procedure for the determination of the active ingredient in the manzate fungicide. He also recomniended the use of the Clarke, Baum, Stanley, and Hester ( 1 7 ) method with minor modifications for the determination of residues of manzate fungicide. Levitsky and Lonen (82) published an alternative method for obtaining the desired dispersion, based on Lowen's method (85). These modifications reduce the time required for analysis and give improved precision. Fischer (29) described a colorimetric method for the determination of 0.5 to 10 mg. of thiurani conipounds after preparation of a copper salt. hliddledorf (51) described a method for the determination of thiocarbamates, thiosemicarbazones, and other carbamates. S h m ( 1 2 6 ) studied methods for the determination of isopropyl n-phenyl carbamate and related compounds. Smith, Vagner, and Patterson (133) revien-ed the literature on volumetric analytical methods for organic compounds and included a discussion of certain pesticides. QUATERNARY AMMONIL?vl COMPOUNDS

Wilson ( 1 4 7 ) recommended a method for determining the quaternary ammonium compounds in milk and in T\ ater solutions. Furlong and Elliker (33) described an improved method for the determination of the concentration of quaternary ammonium compounds in Tvater and in milk in the range of 1 to 100 p.p.m. Fogh, Rasmussen, and Skadhauge (31) presented a colorimetric method for the microdeterniination of quaternary ammonium compounds using bromocresol purple, and spectrophotometric determination of concentrations ranging from 0 to 25 y per liter. Kilson (148) reported on the determination of quaternary ammonium salts as reineckates and recommended a method for the detection of quaternary ammonium compounds in milk (146). He recommended a further study of the reineckate method for the determination and identification of quaternary ammonium compounds. ARSENIC

Bartlet, F o o d , and Chapman ( 7 ) modified the procedure of hlagnuson and K'atson for the determination of arsenic in fruits and vegetables. They eliminated interferences by tin, and applied the method to canned fruit and to fresh fruits and vegetables. In the range of 5 to 20 y, the standard error was 1 0 . 2 y of arsenic. Following tin extraction, the standard error was 10.4 y. The method was used in the range of 0.1 to 2.5 p.p.m. in fresh fruits and vegetables and 0.02 to 0.50 p.p.m. in canned fruit. Evans and Bandemer (26) described a method for the determination of arsenic in biological materials, and in eggs and other animal tissues. From 0.01 to 1.3 y was determined, with a recovery of approximately 90%. iilmond ( 2 ) developed a field method for the determination of traces of arsenic in eoils, using a

V O L U M E 27, NO. 4, A P R I L 1 9 5 5 modified Gutzeit apparatus (10 to 350 p.p.m. in the range of 1 to 40 y). Smales and Pate (231) developed a method for the determination of submicrogram quantities (0.01 p.p.m.) of arsenic by radioactivation. Young (151) discussed the use of ceric salts in the determination of arsenic and for other analvtical uses. MERCURY

Miller and Polley ( 9 2 ) developed a method for the determination of diphenyl mercury alone or in the presence of phenyl mercuric compounds, and presented a procedure for the estimation of diethyl mercury alone or in the presence of ethyl mercuric compounds. The error of the method m s not greater than 2 y. Polley and Miller (108) developed an adaptation of the diphenylthiocarbazone method for the estimation of mived phenyl and ethyl mercuric compounds. It was used to detect as low as 2 -1 of phenyl mercurj and to cover a range up to 4000 y. Klein (72)reported on a micromethod for mercury for use with foods and biological products, and later ( 7 4 ) recommended its adoption. HERBICIDES

Hogsett and Funk (56) developed a method for the determination of residue quantities of Crag herbicide 1 in agricultural crops. With 100 grams of fresh material the sensitivity limit was about 3 y. Residues a t levels of 0.1 to 5 p.p.m. were determined on various fruits and vegetables. Smith and Stone (132) described a colorimetric method for the determination of 2 y of the herbicide N-1-naphthylphthalamic acid residues in plant tissues. It gives 80 to 100% recovery in the 5-y range. It was used on soils as well as crops, but some soil types give excessive interference. Young and Gortner (150) described a method for the microdetermination of C R K in plant tissue and colorimetrical determination of the end point. They state the method is accurate and sensitive to concentrations of 0.01 p.p.m. on 300-gram samples of fruit tissue; with modification it should be applicable to other plant tissue and to soils. Bleidner ( 1 2 )presents a method for the determination of micro quantities of CMU by chromatography. Interferences, including those from plant material, were eliminated and the efficiency of the separation was 95 to 100% with microgram quantities. Bleidner, Baker, Levitsky, and Lowen ( 1 3 ) described a method for determining microgram quantities of 3-(p-chlorophenyl)-l,l-dimethylurea(CMU) residues in soils and plants. Samples were digested in concentrated alkali, hydrolyzed, separated, concentrated, extracted, and determined colorimetrically. -1 few parts per billion map be determined in cane sugar and a few parts per 10,000,000 in nearly all plant tissues and soils. Interferences are minimized. Freeman and Gardner ( 3 2 ) described a chromatographic separation method for the determination of four acids in formulations of the herbicide in chloromethylphenoxyacetic acid formulations (CP.1). TT’arshon-ski and Schantz (142) described a method for the determination of 2,4-D in soil using a countercurrent distribution method to determine quantities less than 1 mg. Stroud (136)discussed a method for the determination of 2,4-D. Gordon and Beroaa ( 4 1 ) described a spectrophotometric method for the determination of micro quantities of 2,4-D and 2,4,5-T utilizing partition chromatography. The method is used for small quantities of 2,4-D in liquid insecticide concentrates, insecticide dusts, and alfalfa hay extracts. Determinations are within 5% of added amounts. Mixtures of 0.5 to 4.0 mg. of the two acids Fvere separated completely, with recoveries of 96 to 100%. K i t h experience, results should be within 2 to 3% of the correct value. Heagv ( 5 3 ) after further study of several methods recommended approval of the total chlorine method for the determination of esters of 2,4-D and 2,4,5-T or mixtures of these in liquid herbicides, utilizing the Parr bomb boric acid procedure. Grabe ( 4 2 ) has determined 4-chloro-2-methylphenoxyacetic acid in the presence of chlorinated and unchlorinated acids by absorption in ultraviolet light. Sjoberg (189) has presented an infrared spectrophotometric method for the determination of 4-

657 chloro-2-methylphenoxyacetic acid. Sorensen (134)has described a method for 4-chloro-2-methylphenoxyacetic acid by isotope dilution analysis. Hill ( 6 5 ) presents a method for the determination of 2-methyl-4-chlorophenoxyacetic acid in the presence of related acids using an ultraviolet spectrophotometric method. Barrons and Hummer (6) studied derivatives of trichloroacetic acid and described a colorimetric method for its determination. Tibbits and Holm (138)present a colorimetric residue method for trichloroacetic acid in plant tissue. Quantities as low as 25 y per gram of fresh tissue may be measured. Hummer and Barrons have utilized the procedure with plant sap and on oil extracts. The error in direct measurement is less than 0.4%. R700d (149) presents a method for the colorimetric determination of maleic hydrazide residues in plant and animal tissues and some soils. It may be used for residue below 1 p.p.m. Interferences are eliminated. RODE\TICIDES

Ramsey (111)reported on metals, other elements, and residues in foods, recommending methods for 1080 and a further study of several other methods. Ramsey (112) reported on a qualitative test and a quantitative analytical method for 1080. He recommended revised qualitative and quantitative methods. Ramsey (110) has reported on 1080 and recommended further study. Lasco and Peri ( 8 1 ) developed a method for the determination of AXTU. Bellack and DeWitt (10) proposed a procedure for the analysis of zinc phosphide rodenticides. Coon, Richter, Hein, and Krieger (20) discuss problems encountered in the physical determination of warfarin. They describe a modified procedure for use with mixtures which have previously given difficulty. LaClair ( 7 9 ) studied methods for the determination of warfarin, comparing results by the Eble method with bioassay, which he used as a referee method. Further study of warfarin methods is suggested. BIO.ISS41

The B o p e Thompson Institute continued the study of bioassay methods for detecting spray residues in foods and for evaluating new pesticide chemical?. Burchfield, Hilchey, and Storrs (14) developed a photomigration method for bioassay using mosquito larvae, based on the tendency of larvae to move away from a strong light source. Information was earlier published as a news report under the title “Analysis by Paralysis.’’ The method has been used for the determination of spray residue, and should also be useful in studying synergism, antagonism, absorption rates, and various physical chemical processes in insect toxicology. Burchfield, Redder, Storrs, and Hilchey (15) developed standardized procedures for the rearing of larvae for use in insecticide bioassay, bringing together a complete description of the method partially presented in previous papers. The bioassay method secures results in an hour or less at concentrations of 0.1 to 1.0 p.p.m. Hartzell, Storrs, and Burchfield ( 6 2 ) compare chemical and bioassay methods for the determination of traces of chlordan and heptachlor in food crops. The highest average residue found vias 0.047 and the lowest was 0.003 p p.m.; most samples contained less than 0.05 p.p.m. Average results by the two methods agreed closely for chlordan, and the comparative results were fairly satisfactory, considering the very small amount of heptachlor found. Sun and Sun (137) applied a microbioassay method utilizing flies to the determination of dieldrin, aldrin. lindane, isodrin, endrin, and D D T in milk. The average LD,o for dieldrin was found to be 0.96 p.p.m. -1 sensitivity of 0.1 p.p.m. was obtained. Davidow and Sabatino ( 2 4 ) discussed development of a screening test for chlorinated pesticides. This is much needed to survey a large number of food samples for pesticide residues in order t o segregate contaminated samples at a minimum cost in time and money. Biological methods with flies and with mosquito larvae do not seem suitable for certain types of samples. They determined the approximate sensitivity of goldfish t o DDT, lindane,

ANALYTICAL CHEMISTRY

658 heptachlor, toxaphene, aldrin, dieldrin, chlordan, methoxychlor, and dilan and suggested the use of goldfish for the rapid sorting of foods contaminated with excessive amounts of DDT, lindane, toxaphene, methoxychlor, and dilan.

tion of the liberated halogen by the Volhard thiocyanate method. The mean percentage recovery was 99.5 f 1.3 to 1.6, lvhen added in quantities from 1 to 30 y . GENERAL

OIL EMULSIONS

Riehl, Gunther, and Beier (115) applied the photoelectric colorimeter to the determination of oil deposit on sprayed fruit. A specific weight of a red oil-soluble dye was added to the petroleum oil. Behrens and Griffin (9) have presented a method of evaluating emulsifiers for formulation purposes. They evaluated the factors affecting the efficiency of emulsifiers and developed a standardized method of checking efficiency. Griffin and Behrens (44) describe an apparatus for the examination of emulsions involving measurement of separation. Vonnegut and Neubauer (140) describe a procedure for the detection and measurement of aerosol particles, which may be used to give information regarding the concentration of particles and particle size. Gooden (59) reported on the physical properties of pesticide powders, and recommended collaborative studies on various methods. lCIISCELLAZTEOUS

Kutschinski and Luce ( 7 8 )present a colorimetric method based on the procedure of Gottlieb and Marsh, for the determination of p-chlorophenyl p-chlorobenzenesulfonate (ovotran, K-6451) in spray residues on fresh fruit. Certain interferences were eliminated. Quantities in the range of 0.05 to 0.30 mg. were determined with a recovery of 94 to 100%. The analytical procedure should be rigidly standardized. Grummitt, Marsh, and Stearns (45) studied the properties of [bis(p-chloropheny1)methylcarbinol] and present a method of analysis which may be used nith an error of 2% using 1-mg. samples. Kittleson (71) developed a colorimetric method for the determination of iVtrichloromethylthiotetrahydrophthalimide used as a fungicide, Quantities as low as 0.05 mg. may be determined with areasonable degree of accuracy. The method has been utilized to determine residues on fruits and vegetables. Payfer (104, 106) studied modifications of methods for the determination of rotenone. Rooney (119) recommended approval of the Elmore method for the determination of organic thiocyanate nitrogen in livestock and fly sprays (118). Fischer (29) has presented a procedure for the dilution of dinitrothiocyanobenzene, dinitrorhodanobenzene, trichlorodinitrobenzene, trinitrotrichlorobenzene, and tetranitrocarbazole in an acetone solution with the development of characteristic intense color. Kirchner, LIiller, and Rice (69) present a quantitative method for the determination of biphenyl in citrus fruits and processed fruit products using chromatostrips. Added biphenyl has been determined in citrus juices in concentrations as low as 0.1 p.p.m. and in citrus peel as high as 600 p.p.m. Peters, Rounds, and hgazzi (106) describe methods for the determination of sulfur and halogens using combustion apparatus. A precision of 0.03% has been obtained for both. Above 3% the accuracy may be nithin 1%. The method is also adaptable to the determination of chlorine-containing pesticide residues on food and forage. Dible, Truog, and Berger (25) present a simplified procedure for the determination of 0.5 to 50 p.p.m. of boron in soils and plants. Sinclair and Crandall (128) developed methods for the determination of ethylene chlorobromide and ethylene dibromide colorimetric. Results agree closely 1% ith the volumetric method. They are, however, better for the determination of very small concentrations. The methods have also been utilized for residue in orange peel and pulp. Recovery of milligram quantities is high, Sinclair and Crandall (127) determined ethylene dibromide in air, and attained a high recovery with 50 to 300 mg. of ethylene dibromide. Kennett (68) determined ethylene dibromide and ethylene chlorobromide by absorption in alcohol and determina-

Kirk and Duggan (70)emphasize the need for accuracy and precision in biochemical analyses and for adequate training of chemists and clinical analysts in this field. Too much emphasis has been placed on speed rather than accuracy and dependability of results. Biological variation has been overemphasized and errors in clinical laboratories have been far larger than those usually assumed, and too often such errors may invalidate the utility of the results. A referee system is suggested to improve the situation for both research and routine analytical results in this field. Archibald ( 3 )considers criteria for the use of analytical methods in clinical chemistry. He points out that in many cases the accuracy of results has been unsatisfactory. This discussion may also be of interest to research and regulatory laboratories. Emphasis may well be placed upon the widespread misuse and overuse of the word “test.” “Test” refers to a qualitative chemical procedure and not to a quantitative procedure (122). Authors of papers, particularly those appearing in AMERICANCHmncAL SocrmY journals, should avoid misusage and eliminate the n-ord “test” where it makes no contribution to the meaning. Similarly, the word “strong” is frequently misused when “concentratedJ’ is the desired meaning. Correctly used, concentrated and dilute refer to the relative quantity of a solute present, while strong and weak refer to the relative activity. The word “run” is rather generally misused. One does not “run” a quantitative analytical procedure; rather, he makes an analysis. “Assay” is also frequently misused in referring to quantitative methods of analysis. Gunther and Blinn (47) discuss the basic principles for the quantitative determination of pesticide residues. They consider the basic analytical approaches to be the evaluation by direct measurement (selective) and the isolation followed by measurement (nonselective). Even though every foodstuff containing pesticide residues must be individually investigated in selecting the final residue method, the authors feel that there is promise of systemization and standardization. They also have in press a book ( 4 6 ) on the analyses of insecticides and acaricides. 4CKNOWLEDGMENT

Thanks are expressed to C. U. Brewster, professor emeritus, for aid in searching out references. BfBLIOGRAPHY

(1) Agazzi, E. J., Peters, E. D., and Brooks, F. R., SUL.CHEX., 25, 237-40 (1953). (2) Almond, Hy, Ibid., 25, 1766-7 (1953). (3) Archibald, R. AI., Ibad., 22, 639-42 (1950). (4) Assoc. Ani. Pesticide Control Officials, “Condensed Data on Pesticides,” 3rd ed., in press. ( 5 ) Assoc. Offic. Agr. Chemists, “Official Nethods of Analysis,” 7th ed., 1950. (6) Barrons, K., and Hummer, R. W., A g r . Chemicals., 6 , No. 6, 48, 113 (1951). (7) Bartlet, J . C., TTood, Margaret, and Chapman, R . A., ANAL. CHEM.,24, 1821-4 (1952). (8) Beckman, H. F.,I b i d . , 26, 922-4 (19.54). (9) Behrens, R. W., and Griffin, W. C., J . A g r . Food Chem., 1,720-4 (1953). (10) Bellack, Ervin, and DeWitt, J. B., J . Assoc. O$ic. A g r . Chemists, 35,917-20 (1952). (11) Berck, Ben, A l s . 4 ~ . CHEM.,25, 1253-6 (1953). (12) Bleidner, W. E., J . Agr. Food Chem., 2, 682-4 (1954). (13) Bleidner, W.E., Baker, H. >I., Levitsky, Mchael, and Lowen, W. K., Ibad., 2, 47f3-9 (1954). (14) Burchfield, H. P., Hilchey, J. D., and Storrs, E. E., Contrib. Boyce Thompson I n s t . , 17 ( l ) , 57-86 (1952). (15) Burchfield, H. P., Redder, A. >I., Storrs, E. E., and Hilchey, J. D., Ibid., 17 ( 5 ) ,317-31 (1953).

659

V O L U M E 27, NO. 4, A P R I L 1 9 5 5 (16) Claborn, H . V.,and Beckman, H . F., A N A L . CHEM.,24, 220-2 (1952). (17) Clarke, D. G., Baum, Harry, Stanley, E. L., and Hester, W. F., Ibid., 23, 1842-6 (1951). (18) Cook, J. W., J . Assoc. Ofic. Agr. Chemists, 37, 561 (1954). (19) Ibid., pp. 561-4. and Krieger, C. H., J . (20) Coon, F. B., Richter, E . F., Hein, L. W., Agr. Food Chem., 2, 739-41 (1954). (21) Craig, J. T., and Tryon, P. F., ANAL.CHEM., 25, 1661-3 (1953). (22) Cueto, Cipriano, and Dale, W.E., Ihid., 25, 1367-9 (1953). (23) David, A., Hartley, G. S.,Heath, D. F., and Pound, D. T.,J . Sei. Food -40r.. 2. 310-14 (1951). (24) Davidow, Bernard, and Sabatino, F. J., J . Assoc. Ofic. A g r . Chemzsts, 37, 902-5 (1954). (25) Dible, W.T., Truog, Emil, and Berger, K. C., ANAL.CHEX., 26, 418-21 (1954). (26) Evans, R. J.. and Bandemer. S. L., Ibid., 26, 595-8 (1954). (27) Ewart, 1%’. H., Gunther, F. A , , Barkley, J. H., and Elmer, H. S.,J . Econ. Entomol., 45, 578-93 (1952). (28) Feinstein, L., Science, 115, 245-6 (1952). (29) Fischer, W., 2. anal. Chem., 131, 192-8 (1950). (30) I b i d . , 137, 90-8 (1952). (31) Fogh, Jorgen, Rasmussen, P. 0. H., and Skadhauge, Knud, AXAL.CHEM.,26, 392-5 (1954). (32) Freeman, F., and Gardner, K., Analyst, 78, 205-9 (1953). (33) Furlong, T. E., and Elliker, P. R., J . Dairy Sci., 36, 225-34 (1953). (34) Gard, L. X., and Rudd, K. G., J . A g r . Food Chem., 1, 630-2 (1953). (35) Gardner, Kenneth, and Heath. D. F., ANAL.CHEM.,25, 184953 (1953). (36) Giang, P. A , , J . Assoc. O f i c . Agr. Chemists, 36, 384-7 (1953). (37) Ibid., 37, 625-8 (1954). (38) Ihid., pp. 642-7. (39) Gooden, E . L., Ibid., 37, 639-42 (1984). (40) Gordon, H. T., A N ~ LCHEM., . 24, 857-62 (1952). (41) Gordon, Kathan, and Beroza, hlorton, Ibid., 24, 1968-71 (1952). (42) Grabe, E., Acta Chem. Scand., 4, 806-9 (1950). (43) Graham, J. J. T., J . Assoc. Ofiic. A g r . Chemists, 36, 365-7 (1953). (44) Griffin, IT. C., and Behrens, R. W., ANAL. CHEM.,24, 1076 (1952). (45) Grummitt, Oliver, hlarsh, Dean, and Steams, James, Ihid., 24, 702-8 (1952). (46) Gunther, F. A., and Blinn, R. C., “Analysis of Insecticides 1

,

and Acaricides,” Interscience, New York, in press. (47) Gunther, F. A., and Blinn, R. C., J . Agr. Food Chem., 1,326-30 (1953). (48) Harrigan, Af. C., J . Assoc. Ofic. Agr. Chemists, 36, 743-4 (1953). (49) Ibid., 37, 381-2 (1954). (50) Harris, T. H., Ihid., 37, 621-3 (1954). (51) Hartley, G. S., Heath, D. F., Kulme, J. M., Pound, D. W., and Whittaker, Mary, J . Sei. Food Agr., 2, 303-9 (1951). (52) Hartzell, Albert, Storrs, E. E., and Burchfield, W. P., Contrib. Boyce Thompson I n s t . , 17 (7), 383-96 (1954). (53) Heayy, A . B., J . Assoc. Ofic. Agr. Chemists, 36, 378-81 (1953). (54) Helmkamp, G. K., Gunther, F. A., STolf, J. P., 111,and Leonard. J. E., J . -4gr. Food Chem., 2, 836-9 (1954). (55) Hill, R., Analpst, 77, 67-70 (1952). (56) Hogsett, J . K.,Funk, G. L., A s a L . CHEM.,26 849-53 (1954). (57) Hogsett, J. S . , Kacy, H . W., and Johnson, J. B., Ibid., 25, 1207-11 (1953). (58) Hornstein, Irxin, Ibid.,24, 1036-7 (1952). (59) Hornstein, Irwin, J . Assoc. Ofic. Agr. Chemists, 36, 367-9 (1953). (60) I b i d . , 37, 623-5 (1954). (61) Hornstein, Irwin, and Sullivan, ST. S (1953). (62) Hummer, R. W,, .Yorth Central Weed Control Conf. Proc., 5, 103 (1950). (63) Jennings, E. C., Jr., and Edwards, D. G., AXAL.CHEM., 25, 1179-82 (1953). (64) Jones, L. R., and Riddick, J. A., Ihid., 23, 349-51 (1951). (65) Ihid., 24, 569-71 (1952). (66) Kelsey, David, J . Assoc. Ofic.Agr. Chemists, 36, 369-71 (1953). (67) Ibid., 37, 628-30 (1954). (68) Kennett, B. H., J . Agr. Food Chem., 2, 691-2 (1954). (69) Kirchner, J. G., hIiller, J. AI,, and Rice, R. G., Ibid., 2, 1031-3 (1954). (70) Kirk, P. L., and Duggan, E. L., An-n~.CHEM., 24, 124-31 (1 952). (71) Kittleson, A. R., Ihid., 24, 1173-5 (1952). (72) Klein, -1.K., J . Assoc. Ofic. Agr. Chemists, 33, 594-7 (1950).

Ihad., 36, 589-94 (1953). Ibtd., p. 596. Ibtd., 37, 576-8 (1954). Konecky, h l . S., Ibzd., 36,388-90 (1958). Kunze, F. hI., J . Assoc. Ofic. A g r . Chemists, 37, 57s-81 (1954). (78) Kutschinski, A . II., and Luce, E. S . ,ASAL.CHEM.,24,1188-90 (1952). (79) LaClair, J. B., J . Assoc Ofic.Agr. Chemists, 36, 373-7 (1953). (80) Zbid., 37, 634-8 (1954). (81) Lasco, G., and Peri, C. A, Chimica e industiia ( M ~ l a n )33, . 557-8 (1951). (82) Anr. . . Levitsky, AI., and Lowen, IT. K., J . Assoc. Oi9ic. .. - Chemists, 37, 555-7 (1954). (83) Levy, L. W..and Estrada, R. E., J . Agr. Food Chern., 2,629-32 (1954). (84) Liddell, H. F., *4nalyst, 78, 494-6 (1953). (85) Lowen, W.K., J . Assoc. O j i c . Agr. Chemists, 36, 451-92 (1953). (86) lIcDeritt, J. B., Jr., Ihid., 37, 633-4 (1954). (87) hladntire, W. H., Jones, L . S., and Hardin, L., I h i d . , 33, 653 63 (1950). (88) lIarch, R. B.. Metcalf, R. L., and Fukuto, T. R., J . Agr Food Chem., 2, 732-5 (1954). (89) Mattson, -1.hl., Spillane, J. T.. Baker, Curtis, and Pearce, G. IT., ANAL.CHEM.,25, 1065-70 (1953). (90) Metcalf. R. L.. J . Econ. Entomol., 44, 883-90 (1951). (91) lliddiedorf, K.,Arzneimittel-Forsch., 1, 311-13 (1951). (92) lIiller, T’. L., and Polley, Dorothy, 2 h . 4 L . CHEM., 26, 1333-5 (1954). (93) Alitchell. L. C., J . Assoc. Ofiic. A g r . Chemists, 35, 920-7 (1952). (94) I b i d . , p. 928. (95) Ibid.,36, 1183-6 (1953). (96) Ihid., 37, 216-17 (1954). (97) Ibid., pp. 530-3. (98) Mitchell, L. C., and Patterson, My.I., Ihid., 36, 553-8 (1953). (99) Mitchell, IT., J . Sci. Food Agr., 4, 246-8 (1953). (100) Noynihan, P., and O’Colla, P., Chemistry & Industru, 1951, 407. (101) Xorris, 31. V., Tail, W.d.,and drerell, P. R., J . Agr. Food Chem., 2, 570-3 (1954). (102) O’Colla, P., J . Sei. Food A g r . , 3, 130-5 (1952). (103) O’Donnell, A. E., Seal, 11. A I . , Keiss, F. T., Bann, J. XI., DeCino, T. J., and Lau, S. C., J . A g r . Food Chem., 2,573-80 (1954). (104) Payfer, Romeo, J . Assoc. Ofic. Agr. Chemists, 36, 371-3 (1953). (105) I h i d . , 37, 630-3 (1954). (106) Peters, E. D., Rounds, G. C.. and Agazzi. E. J., ASAL. CHEM., 24, 710-17 (1952). (107) Phillips, W. F., Ihid., 24, 1976-9 (1952). (108) Polley, Dorothy, and Miller, V. L.. I b i d . , 24, 1622-3 (1952). (109) Polley, Dorothy, and Miller, T’. L., J . Agr. Food Chern.. 2, 1030 (1954). (110) Ramsey, L. L., J . Assoc. Ofic.Agr. Chemists, 36, 597-8 (1953). (111) I b i d . , 37,574-5 (1954). (112) Ibid., pp. 581-6. (113) Reith, J. F., Chem. Weekblad., 49, 689 (1953). (114) Remmert, LelIar, F., Parks. T. D., Lawrence. A. M , , and AIcBurney, E. H . , ANAL.Cwmf., 25, 450-3 (1953). (115) Riehl, L. 1., Gunther, F. d.,and Beier. R. L., J . Econ. Entomol., 46,743-80 (1953). (116) Riemschneider, R., Claus, R., and Schneider. H., Anz. Schudl i n g s k u n d e , 25,89-90 (1952). (117) Riemschneider, R., and Stuck, IT., 2. anal. Chem., 136, 115-18 (1952). (118) Rooney, Herbert A., J . ilssoc. Ofic. Agr. Chemists, 34, 677 (1951). (119) Ibid., 36, 387-8 (1953). (120) Rosin, Jacob, and Radan, G. B., .~s.AL. CHEM.,25, 817-19 (1953). (121) St. John, J. L., As.\L. CHEX.,25, 42-7 (1953). (122) St. John, J. L., Science, 69, 474 (1929). (123) Schechter, 11. S., and Hornstein. I., AX.IL. CHEM.,24, 544-8 (1952). (124) Schonamsgruber. AI., 2. anal. Chew., 135, 23-36 (1952). (125) Schreiber, A. A., and AIcClellan, D. B., -4x.4~.CHESf., 26, 604-7 (19543. (126) Shaw, R. L., J . Assoc. Ofic.A g r . Chemists, 36, 381-4 (1953). (127) Sinelair, W.B., and Crandall, P. R., J . Econ. Entomol., 45, 80-2 (1952). (128) Ibid., pp. 882-7. (129) Sjoberg, B., Acta C’hem,. Scand., 4, 798-805 (1950). (130) Skukis, A. J., Christi, D., and Wachs, H., Soap Sanit. Chemicals, 27, 124-7 (1951). (131) Smales, A. ii., and Pate, H. D., ASAL. C m m , 24, 717-21 (1952). (132) Smith, A. E., and Stone, G. hl., Ibid., 25, 1397-9 (1953). (73) (74) (75) (76) (77)

ANALYTICAL CHEMISTRY

660 (133) Smith, 13'. T., Wagner, $3'. F., and Patterson, J. AI., Ibid., 26, 160 (1954). (134) Sorensen, P., Acta Chem. Scund., 5, 630-7 (1951). (135) Streuli. C. and Cooke, W.D., - 4 ~ 4CHEM.. ~. 26, 970-2 (1954). (136) Stroud, S. W., Analust, 77, 63-6 (1952). (137) Sun, J.-Y. T., and Sun, Y.-P., J . Econ. Entomol., 46, 927 30 (1953). (138) Tibbits, T. W., and Holm, L. G . , J . A g r . Food Chem., 1, 724-6 (1953). (139) T'enkateswarlu, P., and Rao, D. S . , Ax.4~.CHEM.,26, 766-7 (1954). (140) T'onnegut, Bernard, and Seubauer, Raymond, Ibid., 24,1000-5 (1952). (141) Walker, E. A., J . Assoc. Ofic.A g r . Chemists, 37, 647-8 (1954).

WarshonTsky, Benjamin, and Schantz, E. J., ANAL.C H E I I . , 22, 460-3 (1950). Willard, H . H., and Horton, C. A , , Ibid., 24, 862-5 (1952). Willermain, AI., Chintie & Industrie, 63, 09-71 (1952). Wilson, C. W., Baier, Rodger, Genung, Dale, and llullowney, James, AXIL. CHEM.,23, 1487-9 (1951). Wilson, J. B., J . Assoc. Ofic.Agr. Chemists, 36, 741-3 (1953). I bid., 37, 3T4-9 (1951). Thid.. . ..... , n n .. 27Q-81 - .. --. Wood, P. R . , . ~ - A L . CHEM.,25, 1879-83 (1953). Young, H. Y., and Gortner, IT. .%.,I b i d . , 25, 800-2 (1953). Young, Philena, I b i d . , 24, 157-62 (1952). Zeumer, H., and Fischer, W., Z . anal. Chem., 135,401-9 (1952). Paper s o . 1373, Washington hgricultllral Experiment Stations, Pullman, Wash

SCIENTrFrC

Solid and Gaseous Fuels HAYES T. DARBY The Pennsylvania State University, College of M i n e r a l Industries, University Park, Pa.

T

H I S fourth review of the analysis of solid and gaseous fuels continues former reviews (33, 34, 6 4 ) and includes a bibliography of references found from September 1951 through June 1954. T h e arrangement follow closely that of the other reviews.

SOLID FLELS This section discusses the sampling, testing, and analysis of such fuels as coal, coke, wood, charcoal, and briquets but does not include any reference to atomic fuels. SAMPLING

Liddell (99) has discussed the literature relevant t o the problem of coal sampling and the general principle on which a sampling specification should be based. T h e ideal specification should be flexible and should provide a self-contained estimate of its accuracy in any particular case. He is of the opinion that these requirements are not met in the present British standard method, nor do they appear likely to be met vihile the approach to the problem is along present lines. Sampling procedure should give an accurate sample and some measure of the variation in the lot of coal sampled. The solution of this problem consists of taking duplicate samples so that the difference between the duplicates can be used to indicate accuracy. Further, the use of subsamples will give an idea of the variation within a large quantity of coal. T h e sampling accuracy required TT ill depend on the purpose of the sampling. For much routine n-orlr the accuracy of each sample could be much reduced from the present accepted standards; the total weight of coal to be examined may not be substantially altered and the over-all gain vioultl be considerable. Emery (44) has presented a thorough analysis of the theory of small coal sampling. T h e principal factors affecting the number of evenly spaced increments to be taken from the moving consignment of coal are the mean ash content and the maximum size of the coal, but size distribution and the variation of ash content rvith size may affect the sampling results to a greater or less extent. It is generally conceded that, if the sampling is designed t o limit the variability of the ash determined to some predetermined value, other properties of the coal, such as volatilematter content and calorific value, will be obtained n i t h sufficient precision. Since the general level of ash content may vary from place to place in the coal consignment, a sampling procedure for commercial application should take this factor into consideration.

Orning (123) has shown that the economic sampling of a coal, which is low in value per unit TT eight but high in tonnage produced and contains variations in moisture and ash, requires attention to maximum piece, sample weight and applicability to estimates of ash content, method of mining as related to impurities, and methods of handling affecting mixing of varying components. The compositing of alternate increments into two gross samples, to be reduced and analyzed separately, provides one method of routine control of sampling. Bertholf (16) has stated that a minimum number of increments must be taken to obtain a sample of preassigned accuracy, regardless of increment weight. T h e principal variances include the the true variability of the coal from time to time, the random variation in increments such as number of particles, particle weight, and ash content, and variations of reduction and analysis. Bur ( 2 4 )has described a sampling device that consists of three lengths of S-inch pipe which are attached to the side of the car dumper and an angle 50" from the horizontal. TITO5-inch holes are cut into each pipe in the sections projecting beyond the edge of the dumper. As the car is dumped, six increments are automatically taken from the interior of the car contents, while coal from the top of the car is excluded. -1s the dumper returns to the upright position, the increments drop to a hopper and thence to a sample crusher. Comparative tests have shon-n satisfactory agreement with belt samples and a reduction of sampling labor time effected. 14sman and Mieernian (186) have reviened their method of sampling of coal and the statistical factors involved, such as deviations, and sampling sizing. K a r d ( 1 8 7 ) has shonn that the methods for sampling and determination of moisture in coke which have heen recommended by the Joint Coke Consultative Committee have a sound basis in experimental work and statistical theory. Frei ( 6 1 ) has discussed sampling and sample preparation of solid fuels. PROXIAIATE ANALYSIS

Rodriquez Pire (147) has described and discussed methods of sample preparation and analysis. T h e sample preparation and method for determination of moisture, ash, and volatile matter are British standard methods. The determinations of fixed carbon were made according to the ,4merican Society for Testing Rfaterials ( A S T l I ) and its British, German, and French counter-