Document 4aLXbDgJyrvOMNZ4p6bzJG1qQ

/ 7 3 s . 00 ./ INDUSTRIAL HYGIENE A ND TOXIC,OL0G Y / OCTOBEk, 1939 i 1 NUMBER 8 CHRONIC EXPOSURE TO BENZENE (BENZOL). I. TEE INDUSTRIAL ASPECTS* MANFREDBOWDITCAHND HERVEYB. EL~INS Division of Occupational Hygiene, Massachusetts Department of Labor and Zndwtrier, U Boston, Mass. FOREWORD In the three papers comprising this series, a common system of key lettering bas been adopted to designate the several industrial plants considered, and a common system of key numbering identifies theindividual workerswhose cases were investigated. The fact that the industrial and medical studies were not in all details co-extensive will explain the omission in one paper of a key letter or number which may appear in another. 0F THE 89 benzene workers The two principal processes involving discussed by Hunter in the benzene exposure were the mixing of a second paper of this series, 40 nitrocellulose dope, carried out in the were employed in plants where careful compounding room, and the coating of appraisal of the benzene exposure had fabric therewith, this being done in the been made, while 45 worked under coating room. conditions which had been inspected This plant was visited in June, 1936, but not subjected to detailed study. and the following December. Several It is therefore possible to estimate the determinations of atmospheric benzene average concentrations of benzene vapor concentrations were made at vapors to which these workers were both times. In the compounding exposed, although in the case of the room the average concentrationsfound latter group approximate data only in June and December were, respec- will be obtained. tively, 190 and 160 p.p.m., while in Plant A was an artificial leather the coatingmom, near the coating ends plant employing about 30 persons. of the machines, the corresponding * Received for publication July 15, 1939. values were 240 and 420 p.p.m. ' 321 ... ! 322 JOUJWAL OF INDUSTRIAL RI'GIEXE AND TOXICOLOGY [d2.1, 710. 8 It is believed that the average coating room concentrations were somewhat lower than those listed, as samples were taken only during actual coating processes. The compounding room values, however, are believed to be representative of those to which most of the workers in that room were exposed regularly. In addition, aorkers 1and 2 were partially engaged in cleaning out barrels and cans with a solvent containing 25 to SOY0benzene. During such n-ork they would inhale concentrations appreciably higher (possibly twice as high) than those indicated in the previous paragraph. The coating machines in this plant were enclosed and exhausted. The general ventilation in the compounding room was fairly good, but the room was of relatively small area with a resulting congestion of small mixers and other open or partly open containers, each a source of benzene vapor. The use of a benzene mixture for cleaning purposes and the presence of a large open container of this cleaning solvent were the only conditions not ordinarily present in plants of this type. Plant C was also an artificial leather manufacturing establishment, employing about 50 men. In this plant the basic solution of nitrocellulose was prepared in a sepaxate building termed the pyro room. The process of filling this dope into drums, which were then sent to the compounding room, was carried out by worker 34. This plant was visited in March, 1937, and air analyses mere made in three departments. A single benzene vapor determination in the pyro room showed a concentration of 270 p.p.m. during the actual dope filling process. This work was said to consume less than half the worker's time and his total average exposure was therefore estimated as about 100 p.p.m. As only a single determination was made, this cannot be definitely stated to be this worker's exposure. The average of nine vapor determinations in the compounding room of this plant was 200p.p.m. The average exposure of a worker while transferring dope from one container to another, adjusting the color of the dope in the open mixer, etc., was found to be 380 p.p.m. This is believed to represent the maximum benzene esposure of such a worker during his regular routine. Workers 79 and 88, who were found to have symptoms of benzene poisoning some time after t.heir exposure ceased, mere employed in this room. A wash solvent containing no benzene was used in this plant, but analysis showed that this solvent gradually picked up benzene from the dope remaining in the cans, so that cleaning processes involved a definite benzene exposure. In the coating room, an average benzene vapor concentration of 210 p.p.m. was found near the coating ends of the machines. The drying ovens of these machines mere well constructed and exhausted, but the general ventilation of the room was not especially good. Plant G was another artificial leather factory. Only compounding room workers were examined and this department alone will be discussed. Tests made in this room on a cold day in February, 1938,showed concentrations of benzene vapor in the vicinity of the mixers ranging from 200 to 350 p.p.m., the average cc being 265 p.p.m. Some im in the genera1 ventilation 1 were promptly made and : made in April showed that vapor concentration had 1 ally reduced. An averagi tion of 110 p.p.m. was f area previously studied , what lower values in otl the room. At the same sulfate ratios were deter workers in this room, va from 37 to 7oy0being obt In January, 1939, a f of tests were made in th; and benzene vapor c( averaging 100 p.p.m. wc Urine sulfate ratios ran: to 40y0 were found in most heavily exposed. changes in solvent corn other set of tests, made March, showed a slig benzene vapor concentra age value being 85 I sulfate ratios ranged be 52%The average exposure examined in this plant 3 to be just under 100 cold weather. The la room mas such that cor probably be greatly weather, as natural v( relied upon to keep the trations at a low level. tests ma& by the pla warm weather indica lower benzene expo& found in winter. A fourth artificialile: W~IS also studied. A )VC as a coater in this summer of 1937 fro1 c O c f . 19891 BEXZEXE: ISDUSTILIAL ASPECTS p.p.m., the average concentration benzene poisoning. Tests made in the being 265 p.p.m. Some improvements coating room of this plant indicated in the general ventilation of this room average benzene vapor concentrations were promptly made and further tests of 200 p.p.m. in March, 1936, 130 made in April showed that the benzene p.p.m. and 160 p.p.m. on two different vapor concentration had been materi- days in March, 193i, and 120 p.p.m. ally reduced. An average concentra- in July, 1937. tion of 110 p.p.m. was found in the From the statistical standpoint there area previously studied, with some- have been 8 known recent cases of Khat lower values in other parts of benzene poisoning terminating fatally the room. A t the same t h e urine in three artificial leather plants. In sulfate ratios were determined for 6 these plants there were, in all, about Korkers in this room, values ranging 60 workers exposed to benzene vapor from 37 to 70% bcing obtained. in concentrations ranging from 100 to In January, 1939, a further series over 200 p.p.m. K i t h the exception of tests were made in this same room of one worker who was a millwright and benzene vapor concentrations and whose exposure could not be averaging 100 p.p.m. were obtained. estimated, all the fatalities occurred Urine sulfate ratios ranging from 19 among men who had had these heavy to 40% mere found in the workers exposures. As tnble I indicates, the most heavily exposed. After minor compounding room is their chief changes in solvent cornposition, an- source. In the three plants of this other set of tests, made the following type where the mnsimum average March, showed a slightly reduced exposure was judged to be 100 p.p.m. benzene vapor concentratioo, the aver- or below, there hare been no fatalities. age value being 85 p.pm. Urine These plants employ at least 50 sulfate ratios ranged between 13 and workers exposed to benzene. 52y*. Plants D and E used benzene rubber The average exposure of &e workers cements in the manufacture of crepe examined in this plant was estimated rubber soles for shoes. No direct to be just under 100 p.pm. during measurements of benzene exposure cold weather. The layout of this were made, but tests in plant E on the room was such that conditions would operations performed by worker 78 probably be greatly affected by (which were the same os operations veather, as natural ventilation was performed in plant D by worker 35), relied upon to keep the vapor concen- after benzene had been replaced by trations at a low level. Urine sulfate naphtha, showed naphtha vapor con- tests made by the plant chemist in ccntrations of from 300 to 1300 p.p.m. warm weather indicated a much It might be pointcd out that the poison- lower benzene cxposure than was ing cases in these plants, both involving found in winter. females, occurred after relatively short A fourth artificial leather plant, I, exposures to apparently high concen- was also studied. A worker employed trations of benzene vapor. In the as a coater in this factory died in the artificial leather establishments, on I* summer of 1937 from the effects of the other hand, the exposures have in . ..I . .... - ... ' .- I . .. 324 JOURNAL OF INDUSTRIAL HYGIENE AND TOXICOLOGY [vol. 81, no. 8 all cases been longer and, as a rule, less severe. In another shoe factory where a crepe sole cement containing benzene was used, and where the cementing was done in or near an exhausted hood, urine samples from the two workers worker. When visited, the only benzene found in the shop was a special product used in small quantities, not over a pint a month. The rubber cement used most freely, about a gallon a month, was at that time a naphtha cement, but it was shown - TABLE 1 WOBXT - OCCVPATION 0 Coater 1 Compounding room 2 11 I 1 12 Coating, compounding and finishing 32 Mechanic 34 Pyro room-compound ing 79 Compounding room 87 Coater Executive 88 Compounding room 35 Crepe sole cementing 78 I 1 11 I 80 Rubber cementing -81 Paint removing Artificial leather 11 11 I6 II I I I1 I1 I1 I 1 11 `I I1 I 1 11 I 1 11 Rubber goods Shoe mfg. Shoe repairing Telephone exchangc Cone. P.P.m. 150 180+ 180+ Varied OrnOMIL Fatal I1 11 11 <IO0 (est.) lO0+ 200 Slight 200 500 (est.) 500 <25 <10 l1 I1 II `I 11 Recovering Fatal Recovering Fatal I1 TABLE 2 ?LAnT A Artificial leather.. .... . .. . . Compounding room A 11 .11 . ... . . . ... Coating room G 11 .I1 ...... . . . . Compounding room D Rubber goo& ............. . Rubber cementing 180f 250- 100500 (est.) --- 5 480 5 360 6 350 44 21 48 exposed were examined and sulfate ratios of 31 and 37% were found. The benzene exposure was almost certainly less in this case than in plants D and E. Cases 80 and 81 have also been investigated. Worker 80, a cobbler, waa found to work in a reasonably well ventilated shop employing one other conclusively that benzene cementa had formerly been used. A survey of 68 cobbler shops in greater Boston disclosed that the majority of cobblers did not use benzene cements in any quantity, but three establishments using 3 to 1 gallon of cement a week were selected and the urine sulfate ratios of the 4 ocl. 19s91 cementers exposed . All four samples we ratios of over 80` negligible exposure t Worker 81 appare quantity of a paint rc benzene for a few m The exposure to s( processes carried 01 phone exchanges w ceedingly slight. u1 additional contact 7 uncovered by our must be concluded tl absorbed much less quantity usually harmful. The serious cases ( ing on which we hm summarized in table on which we have d: cases of poisoning an From these data n to consider the thres allowable concentra vapor in workroom : this, however, it will the technic employ such concentrations. methods of evalua1 hazard. METHODS OF VAPOR Benzene vapor de made chiefly by the v developed by SI samples mere taken (dfritted glasspetri tut of nitrating acid 9nitric and concentra * A * large portio$ c work desoribed in psper waa carried 0 1 Project No. 465-14Progreee Adminiatrat ocl. 19391 BENZENE: INDUSTRIAL ASPECTS 325 cementers exposed -re determined. A rate of flow of 0.25 1.p.m. was All four samples were found to have ordinarily used. The exact technic ratios of over SOTl, indicating a has been described elsewhere (3). negligible exposure to &benzene. In an extensive series of experiments Worker 81 apparerdy used a small the colorimetric method developed by quantity of a paint remover containing Schrenk and others (4) was compared benzene for a few mimutes every day. with the Smyth method. For this The exposure to sobrent in similar purpose the sample was diluted to 50 processes carried out a t other tele- ml., and 45 ml. used for the Smyth phone exchanges w s obviously ex- determinations, portions of the re- ceedingly slight. Unkss there was an mainder of the colorimetric test. It additional contact with benzene not was found that fair agreement between uncovered by our investigation, it the methods was usually obtained and must be concluded that this individual that when the agreement was not good, absorbed much less benzene than the a further check on the colorimetric quantity usually believed to be harmful. The serious cases of benzene poisoning on which we have information are TABLE 3 COMPARISON OF SMYTHAND SCBRENK METHODS FOR BENZENE VAPOR DETEBMINATIONS summarized in table 1, and the plants on which we have data as to incipient cases of poisoning areshown in table 2. From these data we are in a position I-I-I-0riginalRewa.t Final to consider the thresbld or maximum allowable concentration for benzene vapor in workroom air. Before doing 2 this, however, it dlbe well to discuss the technic employed in measuring such concentrations, as well as other methods of evaluathg the benzene .h n n n r A LIal4LL1 u method usually gave a result closer to that obtained by the Smyth method. A summary of the checks obtained is shown in table 3. Our conclusions regarding these two METHODS OF VAPOR DETERUINATIOmNe*thods are as follows: Benzene vapor determinations were made chiefly by the volumetric method developed by Smyth. (2). The samples were taken in a special semifritted glass petri tube containing 5 ml. of nitrating acid (equal parts fuming nitric and concentrated sulfuric acid). 1. With all the apparatus and reagents available, the two methods are approximately equally laborious. 2. The colorimetric method is more accurate for low concentrations ( b e low 50 p.p.m.), the Smyth method for high concentrations. 3. There is no systematic error in A ' large portion of the experimental work described in the remainder of this paper was carried out as part of Official Project No. 465-14-3-386 of the Works Progress Administration. either method. Of 37 determinations, the average result was 104 p.p.m. by the Smyth method, 106.5 by the colorimetric method. , h ! ;' i i t ! 326 JOURNAL OF INDUSTRIAL HYGIENE AND TOXICOLOGY [uol. 32, rw. 8 4. The Smyth method is more reliable. If the colorimetric method is to be used, checks should be run on all samples. 5. The colorimetric method is invaluable when toluene is present. USE OF GRAB SAMPLE DEVICES In one plant where rubber cementing was being carried on, samples were taken both by the chemical continuous sampling method and by a sensitive combustible gas indicator. As the cement solvent consisted of a 40 to 60 mixture of benzene and naphtha, the data obtained by a method not specific for benzene could be approximate only. Readings of 100to 200 p.p.m., indicating 40 to 80 p.p.m. of benzene, were obtained with the indicator, while concentrations of 10 to 20 p.p.m. were found by the chemical method. Normal urine sulfate values confirmed the latter finding. While the use of grab sample devices should not be condemned on the baais of this single case, it may be said that it is difficult to obtain average values for exposure by such methods. There is always a tendency to obtain the maximum momentary exposure, rather than an integrated sample covering all phases of the worker's routine. Direct-reading instruments such as the combustible gas indicator and interferometer find their chief advantage in obtaining data for use in designing exhaust equipment, rather than in the evaluation of average exposure to solvent vapor. URINE SULFATE DETERMINATIONS Schrenk and others have shown that the ratio of inorganic to total sulfates in the urine is lowered if there is ab- sorption of benzene (5). Several authors have discussed the practical application of the test to benzene workers. In the Bureau of Mines' plant study, 15 vapor determinations were made and compared with 60 urine sulfate ratio tests (6). The correlation obtained was not especially good. Kammer and others found that workers treated with sodium salicylate gave low urine sulfate ratios (7). Jephcott and Bulmer suggest that the daily excretion of organic sulfate, wrather than the sulfate ratio, be taken URINE SULFATE RATIOS-- % 0 20 4 0 60 EO IqO 100-1'5 i n -75 100 PW 2Ian 40-75 Wx 80 0-40 NWz m 0 FIG.1. Correlation of urine sulfate ratios with atmospheric benzene concentrations. Each square represents one urine sulfate determination. as the criterion of benzene exposure (8). Both of these last papers report that the sulfate ratio of the sample does not change materially on stsnding, a findingwhich we have confirmed. Vigliani and Giannini made sulfate ratio determinations on 39 benzene and 22 benzine workers (9). They showed that even short exposures to moderate benzene vapor concentrations afected the sulfate ratio. We have made 36 urine sulfate ratio determinations where the exposure during the day was measured by air analysis, a total of 84 atmospheric Oct. 19391 samples being ta these tests are in These tests n. where relatively maintained, no &I the average be greatly exceeded whole, there is a between the two age ratios for tl groups are show; The benzene e obtained by ave all vapor detern given worker's : few cases correc- ~ QllOUP I I1 111 IvV T I-BINLI: EXPO3S. UO O <40 40-7 75-11 100-1 tracting 10to 20 that the job inc a greater or le2 measured. Thc benzene expost quite as definit cated in this t: given room un duplicated the : closely, even I samples varied These urine s: the end of the. least 6 hours off few tests werp tained in the& parison of mo+ results is showxi It is seen tli -I '1 11 Oct. 19S81 BESZESE: ISDUSTRIAL ASPECTS 327 samples being taken. The results of these tests axe indicated in figure 1. These tests were made in plants where relatively good control was maintained, no area being found where the average benzene concentration greatly exceeded 100 p.p.m. On the whole, there is a very good correlation between the two methods. The aver- age ratios for the daerent exposure groups are shown in table 4. The benzene exposure mas as a rule obtained by averaging the results of all vapor determinations made in the given worker's area. This was in a few cases correct,ed by adding or sub- which the sample is taken is very important. In every case except that of worker 3, the ratio of the afternoon sample was over 2OY0 lower than that of the forenoon sample. The exposure of worker 3 was highly irregular and might easily have been very high before the first sample was taken and much lower during the latter part of the day. The remarkable difference in fore- noon and afternoon samples caused us to question the effect of the previous day's exposure on the sulfate ratio. Accordingly one of us and an assistant spent nearly a full day in an artificial TABLE 4 I I I I I1 <4: 1I I 8 86 - I - I - I - I 6 81 111 40-75 11 61 IVv 100-12j 42 C 34 d 59 77 49 29 28 tracting 10to 20% where it was known that the job included operations with a greater or less exposure than that leather plant taking air samples, and measured. Though the degrees of urine samples were collected over a benzene exposure are probably not 24-hour period and the sulfate ratios quite as definitely known as is indi- determined. The results of this study cated in this table, repeat tests in a are shown in figure 2. given room under similar conditions It is seen that subject X, who had duplicated the arerage concentrations an average exposure of about 80 p.p.m. closely, even though individual air from 7:45 to 12:00, 60 p.p.m. from samples varied considcrably. 1:00 to 2:30 and 80 p.p.m. from 2:30 These urine samples were taken near to 4:00,showed sulfate ratios of 69% the end of the working day, after at at 2:30 and 55y0 at 7:OO p.m., with least 6 hours of benzene exposure. A 88% at 7:OO a.m. the following day. few tests were made on samples ob- Subject Y, whose exposure was tained in the late forenoon. A com- slightly less, showed ratios of 79% at parison of morning and afternoon test 7:OO a.m., 68% at 11:15, 40% at 4:OO results is shown in table 5. p.m., 46% at 6:30 p.m., 79% at 10:45 It is seen that the time of day at pm., and 79% at 8:C0 a.m. the next i .. . . .... .'-. 3!28 JOURNAL OF INDUSTRIAL HYGIENE AND TOXICOLOGY [vol. 91, no. 8 day. In other words, the d a t e ratio ately high benzene vapor concentra- returns to normal shortly after the tion showed a high sulfate ratio (97%). exposure ceases. This case is not included in figure 1. The area between the curve of SUIA-nother worker in the same plant fate ratio and the line representing the showed a very low ratio (40%), normal ratio is 200 units for X and 230 considering the exposure. Samples for Y, a fair check, although Y's expo- taken 4 months later showed that each sure was about 10% less than that of reacted normally (ratios of 71 and X. Y's ratio before and after exposure 82%, respectively), indicating that the I O 12 2?u. 4 68 IO 84M IO 12 zw. 4 6 8 IO 8AM. FIQ.2. Variation of urine sulfate ratio with time of day, relative to exposure period was slightly below the normal value, but it is not believed that this fact affects the results in any material way. From this evidence we are forced to conclude that the day of the week on which the sample is taken is of little importance, while the time of day, with relation to the benzene exposure, is of great significance. According to Schrenk (5), persons with liver damage do not react nor- -mally to benzene exposure. We found one case where a worker in a moder- previous physical condition of the worker was a temporary one. URINE SOLFATE RATIO V6. Am ANALYSIS Schrenk (5) and Sayers (1) emphasize the advantages of the urine sulfate test over the measurement of atmospheric benzene vapor concentrations, while Ka.mmer (7) goes so far as to say that the latter are practically worthless as a control measure. Our results indicate that the two methods, U c f . 29391 when proper1 results, and successfully hazard. An that the ai determine t process or of and thus in! where conti applied, whil indicates on1 the worker ( hours, cith of the actua In the 1 Bureau of b tions, urine detcrrninati( Massachuse medical ex2 tially negat ratios, 22 between 50 50% (only workers ml was evaluat 22 were e q and 35 to c From thi that the n cated no ha a slight h i serious ha; servations paper) and would be fc conclusion. It is prc posure w t was mitig poor busint that time, the act& during th EESZEXE: ISDl%'fRIAL ASPECTS when properly carried out, give similar results, nnd that either can be used successfully to evaluate the benzene hazard. An important ciifferencc is that the air analysis method will determine the esposure of a given process or of various steps in a process, and thus indicate directly the points where control measures should be applied, while the urine sulfate method indicates only the averege exposure of the worker during a period of several hours, cithout direct determination of the actual sources of exposure. In the 1933 plant study of the tion was increased, It should be said for the air analysis method that it could have been the means of predicting this situation, showing as it did high concentrations of benzene vapor for definite processes which would become increasingly hnzardous as they became more and more nearly continuous. A further limitation of the urine sulfate test is its inutility in measuring excessively high exposures. In figure 3 the organic sulfate ratio is plotted against benzene ,exposure in hours Massachusetts plants (C, G, I). The , '~~ 50% (only two below 40%). Of 57 workers whose exposure to benzene was evaluated by means of ah analysis, 22 exposed to le, than 50 P.D.m. and 35 to over 120 p.p.m. From this study we would conclude that the medical examinations indicated no hazard, the urine sulfate ratio a slight hazard and the air analyses a serious hazard. From Hunter's observations on plant C (see following paper) and our own on plant I, we would be forced to agree with the last It is probable that the actual exposure when the 1933 study was made was mitigated by the fact that, due to poor business, production was low at that time, and that, in spite of undoubted improvements in ventilation, the actual hazard increased materially during the years 1935-1937 aa produc- - UVSS WAICS. 5 FIG. Subjects and y. Normal ratio measured. 0 Plant workers. Nor- mal and Graitainonainssiu. mNeodr8m5a%l .ratAio From Vidiani assumed 89%. times concentration. It is obvious that, with a material increase in concentration, the organic sulfate percentage mould become practically 100 and still higher exposures would produce no change. It is also seen that a very low sulfate ratio, e.g. 10% represents an exposure only 50% greater than does a ratio of 30010, and only twice as great as does a 50% ratio. The urine sulfate method is, on the other hand, more convenient than that of vapor determinations, and will indicate absorption through the skin and variations in individual absorption due to variations in rate of respiration, etc. This information cannot be ob- 330 JOURNAL OF INDUSTRiAL HYGIENg AND TOXICOLOGY [uol. $1, tu.8 tained through air analysis. In general, more can be l e m e d from the use of both methods in combination than from either alone, and both should be employed in important cases of benzene exposure. Neither method shows benzene poisoning, which can be determined only by medical examinations. The latter, however, should not be relied upon as an index of benzene absorption or of the benzene hazard. THRESHOLD CONCENTRATIONS The commonly accepted maximum allowable concentration for benzene vapor is 100 p.p.m., although Greenburg (10) reported a definite hazard in even lower concentrations. On the other hand, 50% has been proposed as the limiting safe value for the urine sulfate ratio. It is obvious from table 4 that these two standards are not correlative. A 50% urine sulfate ratio, based on a single sample taken at the end of the working day, would probably correspond to an average benzene vapor concentration of slightly below 75 p.p.m. This value has been proposed as the maximum allowable concentration by the Massachusetts Dust and Fume Code Committee and, in the light of our experience, seems not unduly restrictive. While we have found that benzene vapor concentrations may readily be kept below 75 p.p.m. in the great majority of industrial processes involving its use, this is not true of all such operations. It is our belief that cases 80 and 81 were exposed to average concentrations well below 75 p.p.m., indicating the possibility of occasional cases of benzene poisoning with even the best of control. Close medical supervision of all benzene workers is thus 'most important and it is to be hoped that adequate substitute materials will be developed for the more hazardous industrial uses of this excellent solvent. BIBLIOGRAPHY 1. SCHILEXK,H. H., YANT, W. P., AND sulfate determination as a measure of SAYEXS, R. R.: A new procedure for benzene exposure. B i d . , 18, 349- the control of benzene exposures. 456 (1936). J. A. M. A., 107, 849-852 (1936). 7. KAY-B, A. G., ISE~'BEXON.,,' I N D 2. SYYTH,H. F., Ja.: The determination BEBG,M. E.:Medical supervision of of small amounts of benzene vapors benzene plant workers. J. A. M.A., in air. THISJ., 11, 338-348 (1929). 111, 1452-1455 (1938). 3. The determination of benzol vapor in 8. JEPHCOTT, C. M., AND BULXEX, F. the atmosphere. Air Hygiene Foun- M. R.: The urinary sulfatc test in # dation of America, Inc., Prev. Eng. the supervision of workers exposed Ser., Bull. no. 2,pt. l(1938). to benzene. THIS J., 31, 132-140 4. SCHXENKH,. H., PEAXCEJ,. S., AND (1939). YANT, W. P.: A microcolorimetric 9. VIGLIANI, E. C., AND GIANNINI: La method for the determination of prevenzione medical dell' intossica- benzene. U. 5. Bur. Mines, R.I. no. zione cronica da benzolo con partico- 3287 (1935). lare riguardo alla determinazione dei 5. SCHXENKH, . H., AND OTHEXS: Urine solfati orinari. Rass. Med. Indust., sulfate determinations as a measure 8, 376403 (1937). of benzene exposure. THIS J., 18, 10. GXEENBUXLG.,: Bcnzol poisoning as an 69-88 (1936). industrial hazard. Reprint no. 1096, 6. YANT, W. P., SCHXENKH,. H., AND U. S. Pub. Health Repte., 41, 1357- PATTY,F.A.: A plant study of urine 1375, 1410-1431, 1516-1539 (1926). ,