June, 1953

Health workers who may be experienced in fields other thantoxicology will find useful background material-especiallyin the introductory and summary sections of this paper-upon which to base their opinions in matters of air pollution.With our rapidly expanding industrialization more and moreof us will become embroiled in arguments over this worseninghazard to urban health.

Toxicologic Perspective in PlanningAir Pollution Studies*

H. E. STOKINGER, PH.D.Chief Toxicologist, Occupational Health Field Headquarters, Public

Health Service, Cincinnati, Ohio

AIR pollution is so much with us andthe public clamor against it so loud

that investigators are being pushed intoexpensive and time-consuming studies ofthe problem. There are obvious reasonsfor the clamor. The public daily notespoor visibility, and is often irritated byunpleasant odors and by burning eyes;it is, moreover, keenly aware that acutedisasters from smog can occur. It istherefore natural that concern is mount-ing over the possibility of long-termchronic effects on health, especially nowthat all the known chemical compoundsare being expelled into the air by awidely diversified industry. Logically,the public reasons, if acute episodes havehappened, may not the lives of citydwellers be injured or shortened fromliving continuously in the presence ofchemical agents the same as or similarto those that brought on acute disaster?Investigators meanwhile are confrontedwith vexing questions. Is health actuallyinjured or even endangered from contin-uous low-grade exposures to chemicalsubstances in the air? If so, how cansuch substances be recognized and their

*Presented before the Section on Medical Sciences,at the Annual Meeting of the American Associationfor the Advancement of Science in St. Louis, Mo.,December 29, 1952.

effects differentiated from those of otheragents of disease? How can evidencebest be obtained? No precedent or blue-print exists in medical annals of themanner of proceeding to obtain answersto the above questions. Indeed, accord-ing to some who have given the problemserious thought, there is no assurancethat such an investigation would lead toresults of value. On the other hand,there are some, believing the solutionpossible, who have begun a small-scalestudy of the problem.The present discussion is concerned

with what toxicology can offer in shap-ing the direction and course of studiesof effects of air pollution on urbanhealth, so that efforts may not be mis-directed or wasted. A critical evaluationwill be attempted of past studies ofacute and chronic air pollution incidentsand of current investigations, and perti-nent toxicologic information will be pre-sented. Already, some air pollutionstudies have reached a point whererealistic planning for a medical study ofurban health is opportune.

LESSONS FROM PAST ACUTE AND

CHRONIC EPISODESAcute Disasters-Five such incidents

have now been recorded: two Belgium[742]

VAIR POLLUTION STUDIES

Meuse Valley incidents of 1915 and1930; two Donora episodes of 1945and 1948. and one at Poza Rica, Mexico,in 1950. However, detailed reports,sufficient for critical evaluation, areavailable on but three: the 1930 MeuseValley incident,' the second Donoraepisode in 1948,2 and the Poza Ricastudy.3

Review of the scientific reports on thecausative agents for the Meuse Valleysmog disaster reveals several facts im-portant to present-day air pollutionprogram perspectives. Despite occur-rence of the incident in December of1930, scientific controversy was stillraging on the identity of the causativeagents as late as October, 1933! Be-cause certain articles were still attrib-uting the toxicity of the fog to fluorinecompounds, chiefly hydrogen fluoride,the last report on the subject attemptedto explain why the "theory," of the roleof fluorine could not be maintained.The final adjudicators, who had servedas technical experts during the coroner'sinquest, excluded fluorine on the basisof insufficient source, and inclined to theview that sulfur compounds had beenreleased in far greater quantities. Forour purpose the important fact to noteis that the identity of the agents re-sponsible for the fatality and injury wasnever conclusively determined. Therewere, however, plausible grounds foradmitting the role of irritant gases inthe disaster.

It seems highly significant to theauthor that a remarkably similar con-clusion was reached in the Donora studyof 1948. The Donora episode was farmore extensively investigated than theMeuse incident and was approachedwith more modern technics; moreover,the investigators had the benefit ofknowledge of previous incidents, never-theless, the identity of the offendingagents could not be expressed in termsmore precise than "no single substancebut rather the combined action of two

or more agents were responsible" for theinjuries sustained. Emphasis again wasgiven to the irritant gas sulfur dioxideas a probable injurious agent. The in-vestigators of the Donora study were,furthermore, plagued by the vexing factthat they were unable to demonstrateby direct analysis of the air or by aposteriori extrapolation of concentra-tions of the toxic agents that any werein excess of normally allowable limits,to say nothing of injurious or lethalconcentrations.* Similarly low, thoughsomewhat more elevated, concentrationsof contaminants were calculated to bepresent during the Meuse incident.tOne of the more important questions

to explain, therefore, is how toxic agentsin the form of irritant gases can producewidespread injury and mortality at smogconcentrations below levels known toproduce such effects normally. Assum-ing that the extrapolated or calculatedvalues of irritant gases are approxi-mately correct, the following explanationis presented as suggestive of a partialsolution.

It is now well known that the greatestnumber of fatalities and the most severeillnesses at Donora occurred among theolder age group and among those withheart disease, bronchial asthma orchronic bronchitis, or pulmonaryemphysema. These groups of indi-viduals belong to a class especially sus-ceptible to changes in air composition.The death of persons so afflicted may be

* The calculated air concentrations of sulfur dioxideextrapolated from determined values at Donora duringthe time the fog was lifting was of the order of 0.5ppm; the MAC value is 10 ppm; lethal values approx-imate 100 ppm for short exposures.

t The calculated air concentrations of hydrogenfluoride in the Meuse Valley during the days of thefog was 0.5 ppm. The maximal allowable concentra-tion of hydrogen fluoride is 3 ppm; irritating butrespirable concentrations are 15 to 20 ppm; lethalconcentrations are of the order of 100 or more ppmfor exposures of brief duration. The estimated airconcentration of sulfur dioxide in the Valley was 8ppm and possibly may have arisen as high as 40 ppm,although of course, direct evidence on this and otherpoints is lacking, unfortunately here as in all studiesof acute episodes.

Vol. 43 743

AMERICAN JOURNAL OF PUBLIC HEALTH

hastened by minimal exposure to nomore specific agents than irritant gases.It has been postulated that death inthese persons is caused under normalconditions by an acidosis produced byinability to eliminate carbon dioxide.Under conditions of exposure to irritantgases there is an even greater inability toeliminate carbon dioxide, due to theconstricting effect of the gases on thebronchioles, and therefore an evengreater tendency to develop acidosis.Because all irritant gases irrespective ofchemical nature act uniformally in thismanner, it is difficult to identify specifi-cally any one air pollutant as thecausative agent.

Contributory explanations may alsocenter on other natural factors occurringin smog conditions. Among these arethe effects of combinations of irritantgases plus other agents (asphyxiants ortoxic metals), or the absorption of suchgases on particulate material (catalyticeffect). Such combinations could wellpotentiate the effect of any one indi-vidual substance. One such experi-mental piece of evidence has alreadybeen reported 4; hydrogen fluoride en-hanced the lethal action of berylliumwhen inhaled by animals at concentra-tions at which neither was lethal alone.Another report 5 has shown styrene inthe presence of bromine to be more in-jurious than either alone. Such findingsagain indicate the difficulty of detectingprecisely the offending toxic agents inthe air.*

Chronic Episode-One instance oflong-term effects on community healthby a chemical air pollutant has recentlybeen brought to light.6 In a small in-dustrial city, a beryllium manufacturingplant had since 1943 been discharginginto the community atmosphere certain

* The recent Poza Rica incident of November, 1950,is excluded as not being germane to the present dis-cussion because of its peculiarly local character. Herehydrogen sulfide, an asphyxiant gas, from natural gasoperations was fairly conclusively demonstrated to bethe cause of localized fatalities and illnesses.

beryllium particulates that by 1951 gaverise to 16 cases of chronic berylliumpoisoning. Five of the persons affectedhave died to date. Because of the radialdistribution of the cases the source ofthe poisoning was unmistakably tracedto the beryllium plant. What is im-portant to the present discussion is thefact that an extremely minute concen-tration of beryllium in the air broughton the disease. According to best esti-mates derived from air analyses andfrom calculated stack output and fall-out, beryllium levels were of the orderof 0.01-0.1 ug/m3 of air. For a metalupon inhalation to produce such devas-tating effects, this is an astoundinglysmall concentration. No other chemicalelement with the exception of the radio-elements, has yet been shown to act withsuch fatal effects at so low a concentra-tion. If it is granted that other metallicelements, which currently exist in theatmosphere of industrial cities, may alsobe injurious at low concentrations, thenthere may be cause for concern of health.The true importance of this singlechronic occurrence is not so much thatberyllium can behave in this mannerbut that a demonstration has been madethat air pollutants can have chroniceffects on health.

LESSONS FROM CURRENT STUDIESCharacter of Chemical Air Pollutants

-Three tables have been drawn up topresent a realistic picture of the atmos-pheric pollutants in some of our cities.These show both the character and theamount of (1) total chemical air pollu-tants so far as these have been dis-covered to date, (2) 20 of the mostprevalent metallic air pollutants fromthree cities' samples and analyzed in asemiquantitative manner, and (3) eightof the most prominent metallic airpollutants according to area of two largeindustrial cities.

Table 1 shows that organic mattercomprises the major air contaminant,

744 June, 1953

AIR POLLUTION STUDIES

TABLE 1

Major Group Constituents in Industrial AtmospheresData Chiefly from the City of Los Angeles

ApproximateConcentration Approximate

Constituents Range * Per cent ofmg./rm Total

Solid Particulates 0.2 - 1 0.5- 2 .5"Carbon" 0.04- 0.4Metals 0.01- 0.1

Tars and Oils 5 -20 10 -50Gasoline 5 -10Oil 0.5

Ketones and Aldehydes 1 -10

Gases and Vapors 10 -15 25 -40Organic chlorides 10 -15Volatile organics 1 - 5

Aqueous Droplets 0.5 - 2 1 - 4SO2 0. 5 (ppm)S03 0.2Nitric acid 0.5NHs 0.1H2SO4 0.04

Total 15 - 40

* Includes only those constituents thus far identified

the bulk of which is now considered tobe gasoline and organic chlorides;ketones, aldehydes, and acids appear toadd significantly to the total organicpollution. Solid particulates and "aque-ous droplets" comprise far smallerproportions. According to more recentinformation (Los Angeles), the valuesfor oxides of nitrogen might be con-siderably higher than shown. Of thesolids, the metals represent a relativelysmall proportion of the total pollutants,the remainder being chiefly carbon andthe various anionic combinations of themetals. The inorganic gases, mostlyacidic, have been identified as oxides ofsulfur and nitrogen which probably existin the atmosphere as fine aqueousdroplets; ammonia constitutes a rela-tively negligible quantity. The totalpollution load is thus shown to be ashigh as 40 mg/m3, at least in the onecity (Los Angeles) in which such a studyhas been made. Whether such a highproportion of organic constituents is

representative of all large cities remainsto be determined.

Metallic Elements-Table 2 shows ananalysis of the metallic constituents ofthe atmosphere of three cities, Detroit,Mich., Windsor, Ont., and the vicinity ofCharleston, W. Va. In all instances,the same method of filter paper samplingand spectrographic analysis has beenused. The various determinations weremade possible through the recent de-velopment by the Division of Occupa-tional Health of a rapid, spectrographicscanning method of analysis that per-mits the determination of more than 20elements simultaneously with semiquan-titative accuracy.7 It is seen that certainelements, namely, iron, calcium, mag-nesium, and silicon, comprise the bulk(80-90 per cent) of these constitaents.The elements below manganese in Table2 occur in amounts generally less than0.1 tg/m3. In comparison with the pre-dominant elements, those of lower con-centration are commonly considered the

745Vol. 43

AMERICAN JOURNAL OF PUBLIC HEALTH

TABLE 2

Concentration of Metallic Elements in Urbant Atmospheres

Average Values in Micrograms Per Cubic Meter of Air

Charlestots, W. Va.tDetroit, Mich.* Windsor, Ont.`i and Vicinity

SiliconAluminumIronCalciumMagnesiumLeadZincManganeseCopperTitaniumTinMolybdenumBariumCadmiumChromiumNickelAntimonyVanadiumCobaltBerylliumTotal Concentration of

all pollutantsMetallic elements,

per centElements not found **

3.53.23.02.50.60.40.40.20.050.05

0.040.030.010.010.0080.0060.0050.0020.0010.001

233

6As, Hg, Bi, W

6.43.02.87.80.90.70.30.30.40.10.060.050.070.0060.020.03abs.0.009tr.0.002

210

11As, Sb, Bi

* Averaged results of 1,035 filter paper samples during 6-week periodt Averaged results of 82 filter paper samples during 6-week period$ Averaged results of 9 filter samples

** Certain of these elements (As, Hg) may fail to be detected owing to their smalland nature of analytic method

more toxic. Neglecting an occasionalvariation, one is struck by the uni-formity of the results among the threecities. Such uniformity, if substantiatedby further study, is of great importance,for it indicates that any industrial areaof sufficient size and diversity maypossess an atmosphere similar in com-position to that of any other city, thus.reducing investigational work appre-*ciably. One may also note in Table 3that the total metallic pollution load in-our present state of industrializationlies between limits of 0.2-0.5 mg/m3.An interesting comparison of the dis-

tribution of certain elements in the.atmosphere of two cities according todistricts is presented in Table 3. It isseen, as expected, that as one progressesfrom the residential through the semi-industrial to the industrial area, the total

8.72 .30.81.40.40.2abs.0.130.110.010.010.010.030.002tr.tr.abs.tr.abs.0.001

291

5

As, Co, Hg, Bi,Zn, Sb, W

quantity in the mixture

weight of air-borne particles increases.Although this relationship may also holdin a given city for certain elements, suchas iron, lead, and aluminum, it may nothold for certain other elements, such ascopper, tin, and zinc. Caution shouldbe exercised, therefore, in interpretingaveraged city-wide pollution data withrespect to the degree of inhabitant ex-posure to a particular substance in agiven area, and to the consequent effecton health, plant life, and property.

Distinct seasonal variations were

noted in the Cincinnati study, both inregard to total collection and individualelements. Again as might be antici-pated, the total concentration was

greater in the cold months than that inthe summer by a factor of almost two,and particle counts, by a factor of more

than four. The fact that many of the

746 June, 1953

Element

1.-2.3.4.S.6.7.8.9.

10.11.12.13.14.15.16.17.18.19.20.

AIR POLLUTION STUDIES

TABLE 3

Concentration of Certain Elemental Constituents in the Atmosphereof Cities According to District

Average Values in Micrograms per Cubic Meter of Air

Residential Semi-industrial IndustrialElement Detroit * and Cincinnati Detroit * and Cincinnati Detroit * and Cincinnati t

Iron 3.0 6.0 5. 4 7.7 7.4Aluminum 3.8 2 .2 4.6 4.0 5.0Silicon 2.6 .. 3.7 .. 4.1Calcium 4.3 .. 4.3 .. 3.6Lead 0.6 1.0 1.0 2.0 0.9Zinc 0.2 1.6 0.4 1.3 0.7Manganese 0. 18 0.2 0.23 0.4 0.3Copper 0.11 2.0 0.13 0.8 0.2Tin 0.03 0.1 0.04 0.1 0.04

Total concentra-tion of allpollutants 184 191 279 344 381

* Values are the average of 28 samples taken at one representative district for a 6-week period.t Values are the average of 2, 6, and 7 districts, respectively, sampled during 3-year period.Detroit values adapted from Keenan and Byers. Arch. Ind. Hyg. & Occup. Med. 6, 1952.Cincinnati values adapted from Cholak, et al. Ibid. 2, 1950.

elemental constituents changed onlyslightly during this time would indicatethat coal ash contributed the majorportion of the increase during thewinter months.

Of considerable value in distinguish-ing sources of inorganic pollution wouldbe data on the contribution from the soiland from painted surfaces. Some in-vestigators are exploring the possibilitythat certain soil constituents, becomingair-borne, may give rise to pulmonarydisease. One recent report by Cholak 19of a five-year air pollution study ofCincinnati included some information on

the composition of rural air. Althoughamounts of most metallic elements wereless than those found in industrial areas,some were as great as, or even exceeded,those found in the city (Be, Cu, Mg,Ag, Sn, Ti, V, Zn). In no case was an

element absent in the country whenpresent in city air. Such findings cer-

tainly indicate a careful re-evaluationof past data from industrial and urbanresidential districts.

Before the effects of metallic pollu-tants on health may be postulated, in-formation on both their particle size andthe nature of their chemical combina-tions should be at hand. Particles

larger than 5 ,u in average diameter tendto settle out of the air too rapidly to bebreathed in sufficient concentration, andthose larger than 2 f,u if inhaled, do notpenetrate into the lung in appreciableamount nor long remain there.8Cholak 19 found the majority of theparticles to be less than 2 A,u very few as

large as 5 ,u. More important to a

health study, the nature of metalliccombinations has a decided bearing on

the toxicity of an inhaled metal. For ex-

ample, certain forms of beryllium, suchas the silicate, found in ore, or derivedfrom cinders of certain types of coal,have never been implicated in poisoningfrom this element; whereas the oxide,fluoride, and sulfate of this element are

distinctly injurious.Identified Organic Constituents

Considerable evidence has been adducedby the Los Angeles group of air pollu-tion workers led by Haagen-Smit 9, 10that the smog constituents responsiblefor eye irritation as well as for injury tofoliage originate from the interaction ofozone or nitrogen oxides on unsaturatedhydrocarbons to form organic peroxidesor similar oxidized compounds of highirritant potency. Such a discovery rep-

resents a great forward stride in air

12.75.317.75.23.72.00.41.20.1

472

747Vol. 43

AMERICAN JOURNAL OF PUBLIC HEALTH

pollution research, for it is the firsttime that a definite group of organicair-borne substances have been directlylinked with injury to man.

Moreover, through the recent effortsof the biochemist Dixon," an explana-tion for the manner in which suchlachrymators and eye irritants act hasbeen advanced. All lachrymators act onchemical groups common to a numberof important enzymes, resulting in in-hibition of the activity of these biologiccatalysts. This has been shown to bethe SH group common to manyenzymes. These enzymes in turn, arefound at nerve endings. Lachrymationstarts at once on exposure to the vaporand continues during exposure with anintensity depending upon the concentra-tion. It recommences upon renewed ex-posure, but after sufficiently prolongedexposure, it ceases and may no longerbe elicited, at which point signs of con-junctivitis make their appearance.From a basic scientific view, however,

an even greater importance may be at-tached to the findings of the Los Angelesgroup. These atmospherically formedtissue oxidants resemble in structure andactivity the chemical substances now be-lieved responsible for the injury derivedfrom ionizing radiation.'2 These radia-tion effects are believed to be broughtabout chemically by the strong oxidizingaction of peroxides, HO2 and OH radi-cals acting on organic constituents ofbody cells.

Should a relation be established be-tween the effects of air pollutants andradiation, a mutually beneficial advancemay be expected in both areas. For ex-

ample, means have been found of pro-

tecting against radiation injury causedby intracellular oxidants. Use is beingmade of antireactants in the form ofreducing agents (hydrogen donors), suchas alcohols and similar substances. Apossible mechanism for this reaction hasrecently been disclosed.'3 The actionof these hydrogen donors stimulates

catalase, a cellular enzyme, to increasedactivity, thereby resulting in a morerapid rate of decomposition of the in-jurious tissue oxidizing agents.

Another organic air-borne constituentwhich is possibly hazardous is benz-pyrene which was recently found in theair of eight English cities.14 The factthat this well known cancerigenic agentwas derived largely from domestic heat-ing fires directs attention away fromwholly industrial sources as the origin ofall air pollution ills. Evidence directlyalong this line is accumulating here inthe United States. A recent survey ofcontributory sources of air pollution inthe Los Angeles area shows that house-hold burning of trash constitutes thebiggest single source of organic air pol-lution.'5 The long-term effects of thesesubstances on health in the concentra-tions encountered, however, are largelyconjecture at present. In the followingsection more definitive statements willbe made of these effects for certain sub-stances based on well founded toxi-cologic information.

TOXICOLOGIC AID TO AIR POLLUTIONSTUDY PERSPECTIVES

Toxicologic information should beutilized in setting air pollution studyperspectives. The substances among thelarge number now known to be viewedas suspect in regard to chronic effectson health, may be brought into sharperfocus through a consideration of theirknown toxicity. Because of the multi-tude of materials involved, only a fewexamples are offered to indicate howtoxicologic information may be used.

Chronic toxicity or injury can mani-fest itself in two ways: (1) by a repe-tition of acute or subacute effects or (2)by an insidious, gradual development ofchronic disease. The following are briefillustrations of the manner in whichtoxicologic information may be appliedto shape air pollution perspectives.

Sulfur dioxide and related substances

748 June, 1953

AIR POLLUTION STUDIES

have repeatedly been implicated in acuteepisodes, and much experimental evi-dence in animals 16,17 show that adanger may be expected from short-termexposures. On the other hand, there hasyet been no evidence and no physiologicreason to expect that injury may becaused by low-grade chronic exposuresto sulfur oxides, such as currently existin all large industrial cities (muchless than 0.1 ppm). For example, aboveaverage tolerance to an appreciableconcentration of sulfur dioxide is wellrecognized in workers around oreroasters.Hydrogen fluoride and fluorine, like

sulfur oxides, are irritant gases pro-ductive of acute injury, as shown fromconsiderable experimental study.18 Un-like sulfur oxides, however, fluorides arecumulative poisons producing irreversi-ble skeletal changes after prolongedexposures; at higher exposures renalchanges may develop. Currently, thereare certain localized areas in the UnitedStates where long-term exposures tofluorine are creating a problem thatshould be given serious attention.Hydrogen sulfide, although an excep-

tionally effective and highly toxicasphyxiant at elevated concentrations(0.07-0.1 per cent), because of its para-lytic effect on respiratory nerve endings,would anticipatedly be a source ofdanger in isolated areas as in the PozaRica incident. However, concentrationsof hydrogen sulfide in all of the largeindustrial cities are exceptionally low.Should concentrations of this substancerise in the coming years, a hazard maybe anticipated, possibly from accumula-tion of effects.

Nitrogen dioxide, like phosgene, iswell known to produce edematouschanges in the lung that can end fatallyif exposure is to relatively high concen-trations (10-20 ppm). It is not believed,however, that such concentrations wouldnormally be attained, chiefly because ofthe reactivity of nitrogen dioxide with

organic air constituents. This assump-tion is borne out by current surveyswhich show generally low air concentra-tion. No chronic effects from nitrogenoxides are anticipated, primarily becausethere are no indications that such gaseshave cumulative action. Small amountsare readily metabolized.

Arsenic and arsine, on the basis ofpresent information, would not be ex-pected to produce acute injury in acommunity, except in circumscribedareas about plants. There have beenfrequent instances of serious exposuresto arsine in and about metal processingplants, but there have been no reportsof widespread acute injury from otherarsenic compounds. Arsenic is a cumu-lative, chronic poison with a capacity toengender tolerance, certainly in the con-centrations (0.4-0.14 jug As/m3) foundin most cities (London, Cincinnati). Thefact that individuals may developtolerance to arsenic may affect the spec-ulation that arsenic from air sources iscancerigenic. Moreover, arsenic fromother sources, such as cigarettes, exceedsthe amount inhaled from city air.Examples could profitably be ex-

tended. In all cases, however, it is thebalance between the degree of exposureto a particular substance and the stateof the physiologic function involved thatdetermines the final outcome on health.Prime Desiderata for Medical Study

-Thus far, both from evidence basedon a study of actual events and fortheoretical reasons, the improbabilityhas been pointed out of identifying anyone particular agent in a generalizedsmog atmosphere. As the number andcomplexity of substances in the air con-tinue to increase and the opportunityfor interaction becomes greater, theprobability of doing more than identi-fying compound types seems remote.*

* A partial but a slowly evolving solution is pos-sible only from a study of "one industry" areaswhere relatively few (less than half dozen) pollutantsexist in sufficient concentrations among sizable popula-tions.

749Vol. 43

750 AMERICAN JOURNAL OF PUBLIC HEALTH June, 1953

The question then to be considered is,what is the best method of studyingeffects on health in such a complex ofconstantly changing mixtures on urbanpopulations that are in turn equallvcomplex as to physiologic response?

Enumeration of the factors influenc-ing the response of individuals to smogconstituents shows at least 5 inde-pendent variables and at least 16 de--pendent variables excluding the vari-ables in smog itself. Factors affectingresponse are (1) group susceptibility,-(2) population density and movement,(3) habits of living, (4) climaticfactors, and (5) topography. One can-readily list many factors under each ofthese major headings that influence in-dividual response. Most important ofthese from a health study standpointare the effects of other agents of disease.bacterial and parasitic, as well as per-sonal hygiene and nutritional factors.

Such a large number of variableswould have to be reduced to an absolute-minimum to assure a health study of anyreasonable chance of success. Practi-cally, this can be done by limiting thestudy to population groups that are-homogeneous in respect to the most im-portant variables of group suscepti-bility, racial constitution, populationdensity, housing, and nutritional status.Similar care must likewise be taken in-the selection of controls. Smog areas.their intensity and character, will haveto be defined within narrower limitsthan is now the custom in most surveys;averaged data from large areas will beuseless for the present purpose. As itwould be obviously desirable to studyseveral such "homogeneous" segmentssimultaneously so as to include all ofthe factors among the total populationof a city, a vast number of samplingstations would have to be set up and alarge volume of analyses and statisticalwork would be entailed; on the medicalside, a large staff of well trained epi-

demiologists would be required. Thepopulation covered by such a studywould be restricted because of limitedknowledge at present. It is not reallyknown whether or not the lives of urbandwellers are affected by smog. Thereis also room for speculation thatphysiologic adaptability may overridepotential insults to health. For thisreason, some have suggested concentrat-ing a health study on groups whoseadaptability has been limited by somepartial disability such as chronic heartor pulmoniary disease.

Before answers can be found to thesequestions, engineering controls may haveprogressed to such a point as to removethe cause for alarm or at least to reduceair pollution to such an extent that itwill no longer be a problem.

REFERENCESla. Mage, J., et Batta, G. Resultats de lexpertise

judiciare sur la cause des accidentes survenue dansla vallee' de la Meuse pendants les brouillards dedecembre 1930. Chimie et Ind. 27:961, 1932.

l b. . Le role de l'acide fluorhydrique dansla nocivite' du brouillard de la Meuse en 1930.Ibid. 30:787, 1933.

2. Schrenk, H. H., Heimann, H., Clayton, G. D.,Gafafer, W. M., Wexler, H., et al. Air Pollutionin Donora, Pa. Pub. Health Bull., No. 306.Washington, D. C.: Division of Industrial Hy-giene, Public Health Service. 1949.

3. McCabe, L. C., and Clayton, G. D. Air Pol.lution by Hydrogen Sulfide in Poza Rica, Mexico.Arch. Ind. Hyg. & Occup. Med. 6:199, 1952.

4. Stokinger, H. E., Ashenburg, N. J., et al. AcuteInhalation Toxicity of Beryllium. II. The En-hancing Effect of Hydrogen Fluoride Vapor onBeryllium Sulfate Poisoning in Animals. Ibid.1:398, 1950.

5. Adams, E. M., and Schneider, E. J. Eye IrritantsFormed by the Interaction of Styrene and Halo-gens in the Atmosphere. Midland, Mich.: Bio-chemical Research Department, Dow ChemicalCompany. April, 1952.

6. Eisenbud, M., Wanta, R. C., Dustan, C., Stead-man, L. T., Harris, W. B., and Wolf, B. S. Non-occupational Berylliosis. J. Indust. Hyg. &Toxicol. 31:282, 1949.

7. Keenan, R. G., and Byers, D. H. Rapid AnalyticMethod for Air-Pollution Surveys. Arch. Ind. Hyg.& Occup. Med. 6:226, 1952.

S. Stokinger, H. E., Steadman, L. T., Wilson, H. B.,Sylvester, G. E., Dziuba, S., and La Belle, C. W.Lobar Deposition and Retention of Inhaled In-soluble Particulates. Ibid. 4:346, 1951.

9. Haagen-Smit, A. J. Chemistry and Physiologyof Los Angeles Smog. Ind. & Eng. Chem. 44,1 342, 1952.

10. Haagen-Smit, A. J., Darley, E. F., et al. In-vestigations on Injury to Plants from Air Pollu.tion in the Los Angeles Area. Plant Pkysiol. 27:1S, 1952.

Vol. 43 AIR POLLUTION STUDIES 751

11. Dixon, M. Reactions of Lachrymators withEnzymes and Proteins. Biockem. Soc. Symp. No.2, p. 39, 1948.

12. Tobias, C. A. Mechanisms of Biologic Effects ofPenetrating Radiations. Fed. Proc. 10:595, 1941.

13. Chance, B. The State of Catalase in the Respir-ing Bacterial Cell. Science 116:202, 1952.

14. Waller, R. E. The Benzopyrene Content of TownAir. Brit. J. Cancer 6:8, 1952.

15. Magill, P., and Benoliel, R. W. Air Pollution inLos Angeles County. Contribution of CombustionProducts. Ind. & Eng. Chem. 44, 1:347, 1952.

16. Treon, J. F., Dutra, F. R., Cappel, J., Signmon,J., and Younker, W. Toxicity of Sulfuric Acid

Mist. Arch. Ind. Hyg. & Occup. Med. 2:716,1950.

17. Amdur, M. O., Schulz, R. Z., Drinker, P. Tox-icity of Sulfuric Acid Mist to Guinea Pigs. Ibid.5:318, 1952.

18. Stokinger, H. E. The Inhalation Toxicity ofFluorine and Hydrogen Fluoride, in the Pharma-cology and Toxicology of Uranium Compounds.(C. Voegtlin and H. C. Hodge, Editors.) NewYork: McGraw-Hill, 1949.

19. Cholak, J., Schafer, L. J., and Hoffer, R. F.Results of a Five-Year Investigation of AirPollution in Cincinnati. Arch. Ind. Hyg. &Occup. Med. 6:314, 1952.

The Journal 25 Years AgoSAFE WATER BY THE ROADSIDE

As the tourist season opens the following has some timely interest:"Michigan began its annual roadside water survey earlier this year than usuaI

so that tourists will be safeguarded against polluted and contaminated water andthe safe water supplies will be posted before the tourist season is in full swing.Two men are sent out to collect samples of water which are sent to the statelaboratory for analysis. A third man is sent out a couple of weeks later to postthe supplies found safe upon analysis with a 'Safe Water' sign." (A.J.P.H. 18,8:1002 (Aug.), 1928.)