Drinking Water and Health,: Volume 1 (1977)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

lv Solid Particles in Suspension INTRODUCTION In addition to dissolved substances, drinking water typically contains small amounts of very finely divided solid particles of several kinds. These particles, ranging in size from colloidal dimensions to about 100 ,um, are composed of inorganic and organic materials that are derived from soils and rocks and from the debris of human activity with which the raw water has come in contact. They include clays, acicular or fibrous particles of asbestos minerals, and organic particles resulting from the decomposition of plant and animal debris in the soil. Little is known about the effects that these suspended solids may have on the health of those who drink water that contains them. However, there is widespread concern over the biological effects of the asbestos mineral fibers that occur in water, since similar fibers are known to be carcinogenic when air heavily laden with them is inhaled for many years. In view of this concern that such fibers as occur in water may be injurious to health, their occurrence, characterization, analysis, and biological ejects are reviewed in some detail. No evidence has yet been discovered that either of the other classes of common particulate contaminants of drinking water-clays and organic colloids has any direct effect on health. Nevertheless, it is possible that both may indirectly affect the quality of drinking water because they can adsorb a variety of toxic substances, bacteria, and viruses from solution or suspension and bind them more or less strongly. By such means these 135

136 DRINKING WATER AND HEALTH materials may serve to concentrate and transport some water pollutants and protect them from removal by water treatment. For this reason, the properties of clays and organic particulates are also discussed, together with the tendency of chemicals, bacteria, and viruses to become concentrated at the surfaces of such particles. Removal of suspended particles from water is briefly reviewed, together with the significance of measurements of turbidity as an index of water quality. CLAY PARTICLES AND THEIR INTERACTIONS Clay is usually defined on a particle-size basis, the upper limit being 2,um diameter. Soils and sediments in nature will therefore have vaporing proportions of clay material containing clay-mineral components (usual- ly the phyllosilicates), as well as nonclay-mineral material that may include a variety of substances such as iron and aluminum oxides and hydroxides, quartz, amorphous silica, carbonates, and feldspar. The clay minerals themselves are classified in Table IV- 1 (Grim, 1968~. Clays are ubiquitous in soils and sediments derived from soils. They may be formed in soils during soil development through the weathering of various minerals, or they can be inherited essentially without change from the parent material upon which the soils are formed. Parent material, climate, topography, and vegetation determine the kinds of clays that are found. Hydrothermal activity may also lead to clay formation. As erosion acts on the landscape, clays may be suspended in water and carried until they are deposited by sedimentation. Most sedimentary rocks contain more or less clay as, for example, shales (almost exclusively clay), limestones, and sandstones. A number of scientific techniques are useful for studying clays, but the most useful for identification and indication of relative abundance is X- ray diffraction. The diffraction properties of the various clay minerals, as well as the methods of treatment and sample preparation, can be found in publications of Grim (1968), Brown (1961), and Whittig (1965~. Infrared spectroscopy is a valuable adjunct to X-ray diffraction in characterizing clays, and this subject has recently been reviewed by Farmer (1975~. Infrared spectroscopy is the most powerful method for study of organic- clay interactions (Mortland, 1970; Theng, 1975~. Other techniques useful in characterizing clays are electron microscopy (Gard, 1971), thermal methods (Mackenzie, 1957), and chemical analysis (Weaver and Pollard, 1973~. The layer-lattice clay minerals, in themselves, do not appear to have

Solid Particles in Suspension 137 TABLE IV-1 Classification of the Clay Minerals (Grim, 1968) I. Amorphous Allophane group II. Crystalline A. Two-layer type (sheet structures composed of units of one layer of silica tetra- hedrons and one layer of alumina octahedrons) 1. Equidimensional Kaolinite group Kaolinite, nacr~te 2. Elongate Halloysite group B. Three-layer types (sheet structures composed of two layers of silica tetrahedrons and one central dioctahedral or tr~octahedral layer) 1. Expanding lattice a. Equidimensional Montmor~llonite group (smectite) Vermiculite Montmor~llonite, sauconite b. Elongate Montmor~llonite group Nontronite, saponite, hector~te 2. Nonexpanding lattice Illite group Regular mixed-layer types (ordered stacking of alternate layers of different types) Chlorite group D. Chain-structure types (hornblende-like chains of silica tetrahedrons linked to- "ether by octahedral groups of oxygens and hydroxyls containing Al and Mg atoms) Attapulgite Sepiolite Pal ygors kite deleterious effects when ingested by humans. Some of them are, in fact, constituents of pharmaceuticals such as kaopectate (kaolinite). Other indications (some from folklore) suggest beneficial results from ingestion of clays. The effect of ingestion of fibrous clay minerals of the chain- structure types (e.g., attapulgite, palygorskite, sepiolite, sometimes called "asbestos"), is still open to question and is the subject of extensive study at the present time. If layer-lattice clay minerals have deleterious e~ects on human health, they are probably indirect, through adsorption, transport, and release of inorganic and organic toxicants, bacteria, and viruses. Several reports have shown that concentrations of many pollutants are much higher in sediments of streams and lakes than in the waters with  which they are associated. Clays and organic particulates are the

138 DRINKING WATER AND HEALTH materials chiefly responsible for such concentrations. Since clays are ubiquitous in many waters used as sources for human consumption, it is to be expected they will appear as particulate matter in some drinking waters and thus it is of interest to consider the kinds of interactions they have with dissolved materials. Considerable knowledge exists regarding the surface chemistry and adsorptive properties of clays, and thus, with information on the nature of a solute, it is possible to have some idea of their interaction. Clays are very adsorptive substances. The possibility exists that clays could act as vehicles for transport of toxic compounds through adsorption in one environment, followed by release of the toxic material when the clay entered a different environment. It has been well established that some pesticides applied to watersheds can be adsorbed by soil components and subsequently removed into water by erosional processes (Bailey et al., 1974; Nicholson, 1969; Nicholson and Hill, 19701. Inorganic Pollutants This classification of pollutants would include metal cations and some anions. Among the metal cations that have been found to be polluting some water and soils are Pb, Cr. Cu. Zn, Co, Mn, Ni, Hg, and Cd, while radioactive isotopes of Pu, Cs, and Sr, among others, over potential threats as pollutants. On the other hand, anionic species such as phosphates, arsenate, borate, and nitrates are considered pollutants in some situations. The interactions of metal cations with clays include adsorption by ion exchange, precipitation as hydroxides or hydrous oxides on clay surfaces, and adsorption as complex species. ObviouslY. OH and Eh are critical 1 J ' AL factors In determining the nature of the interactions between clays and some transition and heavy metal ions. Hodgson's review (1963) includes some reference to earlier work on clay interactions with some of the transition ions and heavy metals. Jenne (1968) has electively described the various factors controlling the concentrations of transition cations in waters and soils, while Jenne and Wahlberg (1968) and Tamura (1962), among others, have considered the interaction of radionuclides with clays. Holdridge (1966) has reported adsorption studies of heavy metal cations on ball clay. With regard to phosphate, it is likely that its interactions with calcium ion and amorphous hydroxides of Fe3+ and A13+ and with allophane are more important than adsorption by clay minerals in affecting its concentration in natural waters. In addition to adsorption by simple ion exchange, much work indicates the retention of transition and heavy metals at clay mineral surfaces via

Solid Particles in Suspension 139 precipitation of insoluble compounds, notably hydroxy and oxidehy- droxy polymers. The incorporation of A13+, Mg2+, and Fe3+ hydroxy polymers within the interlamellar space of swelling clays to form chlorite- like species is a well-known pedogenic process. It has also been shown that these brucite-gibbsite-like materials may often be withdrawn if the mineral is subjected to a different environment, usually one involving a change in pH. Gupta and Malik (1969) have reported the incorporation of Ni2+ in smectite to form a nickel-chlorite, while Blatter (1973) found similar reactions of smectite with Hg2+ . Thus it seems that many of these kinds of metal cations have the ability to form interlayer complexes in swelling clays. It would seem likely that in natural systems where the polluting species might be present in very low concentrations compared with other interlayer-forming species such as A13+, they might be incorporated within the gibbsite-like layer as it forms, in essentially isomorphous substitution for A13+, although reports of this phenomenon were not found. It would also appear that incorporation of a polluting metal cation within the intergrade clay is no guarantee that it might not again be released to the natural system when the clay, through erosion and deposition, is placed in a different environment where the interlayer material may be removed. This phenomenon has been shown for vermiculite-chlorite intergrades that, upon erosion from an acidic soil, are deposited in a calcareous freshwater lake or floodplain to form discrete vermiculite, within relatively short periods of time (Frink, 1969; Lietzke and Mortland, 1973~. The clay mineral vermiculite has a special affinity for K+ ion, the ion is initially adsorbed in the interlamellar regions of the mineral and then trapped by collapse of the layer structure. The ion is thus removed from direct interaction with the surrounding solution. This process is called potassium fixation, but will occur with other ions of similar diameter to more or less extent. Cations that might be considered pollutants that undergo this reaction with vermiculite are Ba2+ and radioactive Cs+ . in which The hydroxides and hydrous oxides of iron, manganese, and aluminum are often components of the clay fraction of sediments and have important e~ects on pollutant concentrations in natural waters. They often exist as coatings on the surfaces of other minerals and thus may exert chemical activity far out of proportion to their total concentrations. Jenne (1968) suggests that they furnish the principal control on the concentrations of heavy metals such as Co, Ni, Cu. and Zn through adsorption processes. The principal factors a~ecting adsorption and desorption of heavy metals from these kinds of particulates are pH, Eh, concentration of the metal in question, concentration of competing

140 DRINKING WATER AND HEALTH metals, and the eRects of other adsorbents such as organic matter and clay minerals. Organic Pollutants These materials encompass a wide range of compounds, including pesticides, polychlorinated biphenyls, aromatic species of various kinds arising from industrial activity, and fluorine compounds in aerosols. Whether or not organic species adsorb or interact with clays depends upon the structure and properties of the compound and the nature of the clay and its exchangeable cations. Several mechanisms of interaction are possible and have been described in a number of recent reviews (Mortland, 1970; Bailey and White, 1970; Theng, 1974; and Rausell- Colom and Serratosa, 1975~. Organic cations adsorb on clays by ordinary ion exchange and are usually preferred over the inorganic ions by the exchange complex because of their large size and high molecular weights. Examples of organic compounds that are cationic and could be considered pollutants if transported outside their areas of application are the herbicides Paraquat and Diquat. These compounds are strong bases and are completely ionized in water. Other organic compounds, while neutral at the ambient pH of the solution phase, may become protonated after adsorption at the clay surface. The surface acidity of clays has been shown to be a considerably stronger proton donor system than pH measurements of the water-clay system would indicate. Thus, organic compounds containing basic nitrogen or carbony} groups may become protonated, and therefore cationic, after adsorption at clay surfaces. Another kind of organic-clay interaction is the coordination or ion- dipole type. Compounds with nitrogen, oxygen, sulfur, or olefinic groups have electron pairs that may be donated to electrophilic exchange cations to form complexes on the clay surface. In natural systems, an important consideration is the competitive eject of water for these adsorption sites. That is, the energy of ligand formation of an organic molecule with an exchange cation must be greater than the salvation energy of the cation in order to displace water molecules and obtain direct organic-cation coordination. In the laboratory these interactions are easily obtained by dehydration; however, in natural systems the competition of water is a major factor in determining whether or not these complexations occur. Nevertheless, it is likely that this kind of interaction does occur with some highly polar, electron-donating organic compounds. Another important factor is the nature of the exchange cation. Thus, for example, transition metal cations on the exchange complex, that have unfilled a'orbitals, will interact strongly with electron-supplying groups of organic molecules.

Solid Particles in Suspension 141 Still another kind of organic-clay interaction is hydrogen bonding. These interactions can be classified into three types: 1. Hydrogen bonding between water molecules directly solvating exchangeable cations and polar functional groups, such as carbonyl, on organic molecules. The water molecules thus act as a "bridge" between the cation and organic species. 2. Hydrogen bonding between functional groups such as alcoholic and amino groups and oxygens of the silicate surfaces. Infrared spectroscopy has indicated that these are relatively weak bonds, being within the lower range of energies where hydrogen bonding is found. 3. Intermolecular hydrogen bonding between two organic species on the clay curface. Other factors involved in clay-organic interactions include physical forces and entropy effects. With this general description of the interactions of organic species with clays, it is now appropriate to mention some special clay-organic properties that have relevance to organic pollutants. Many organic compounds, including aromatics and particularly the halogenated types such as DDT, chlorinated and brominated phenyls and biphenyls, are adsorbed to little if any extent on clay surfaces from aqueous solution. In the natural environment they are more likely to be adsorbed in organic components of soils and sediments. These materials usually have limited solubility in water, since they are hydrophobic. It is thus not surprising that they are not attracted to the hydrophilic surfaces of clays. The above discussion, however, suggests that, in natural systems, clay-organic complexes may act as adsorbing media for some organic pollutants that are not adsorbed at all by pure inorganic clays. Another phenomenon that may take place when organic species are adsorbed at clay surfaces is that of catalytic alteration. This has particular relevance for organic pollutants since there is much interest in their fate in the environment. Much work has been reported on catalytic reactions on clays at high temperatures, but it is only recently that much attention has been paid to catalysis by clays in conditions resembling the natural environment. One mechanism by which clays can act as catalysts is via their Bronsted acidity. Examples of this are the hydrolysis of esters demonstrated by McAuli~e and Coleman (1955), the conversion of atrazine to hydroxyatrazine by Russell et al. (1968), the decomposition of alkylammonium ions by Chaussidon and Calvet (1965), and the hydrolysis of nitrites to amides by Sanchez et al. (1972~. In many decomposition reactions involving Bronsted acidity, carbonium ion formation is undoubtedly involved. On the other hand, Lewis acid sites

142 DRINKING WATER AND HEALTH may exist in clays that also will catalyze many organic reactions. These sites (electron acceptors) may be part of the basic structure of the mineral itself as, for example, ferric iron within the octahedral layer or exposed aluminum on the edges of the minerals. In addition, some cations on exchange sites function in this capacity, particularly those of the transition metal group. Solomon et al. (1968) have demonstrated catalytic properties of Lewis sites located on edges of clay minerals. The activity of some transition metal cations on exchange sites has also been amply demonstrated as, for example, the decomposition of urea to ammonium ion when complexed with Cu2+, Mn2+, or Ni2+ smectite. No such reaction was noted for urea complexed with alkali metal or alkaline earth-saturated clay (Mortland, 1966~. Aromatic molecules such as benzene will complex via pi electrons with clay minerals saturated with Cu2+, under mildly desiccating conditions. Under more vigorous dehydrating conditions, a radical cation of benzene is formed that will react with molecular benzene to give polymers containing phenyl groups as well as fragmented benzene rings (Doner and Mortland, 1969~. Anisole (methoxybenzene) will also form radical cations that react with molecular anisole to give 4,4'-dimethoxybiphenyl (Fenn et al., 1973~. Other cationic species with oxidizing abilities as great as Cu2+, such as Vo2+ and Fe3+, were also found to produce radical cations from some aromatic species with subsequent polymer formation (Pinnavaia et al., 1974~. These reactions suggest the possibility that some pollutant species adsorbed on clay surfaces may undergo similar reactions to form radical cations and subsequently interact with themselves, or other organic compounds with formation of different chemical derivatives. Thus, pollutant degradation or alteration on clays by oxidation-reduction reactions involving ex- changeable transition-metal cations may be a real possibility in nature. In addition to the degradation of atrazine to hydroxyatrazine, mentioned above, a number of other clay-catalyzed pesticide reactions have been reported. For example, Fleck and Halter (1945) report the conversion of DDT to DDE by kaolinite and smectite samples, preheated to 400°K. Also, degradation of heptachlor by palygorskite has been suggested by Malina et al. (1956~. The degradation of the organic phosphate insecticide, ronnel, by clays heated to various temperatures has been reported by Rosenfield and van Valkenburg (1965~. Organic phosphate pesticides have been observed by Mortland and Raman (1967) to be hydrolyzed in the presence of Cu2+-montmorillonite by a coordination mechanism. The much weaker catalytic e~ects of Cu2+- vermiculite, beidellite, and nontronite were attributed to reduced activity of the copper on these minerals, as compared with montmorillonite, due to charge location. While most of the degradation of pesticides in nature

Solid Particles in Suspension 143 has been attributed to biological agencies, the above discussion would suggest that catalysis at mineral surfaces may also play a role. Natural organic material in soils forms complexes with clays that exert important influences on the physical, chemical, and biological properties of the soil (Greenland, 1965, 1971~. Since the exact chemical and physical nature of these organic materials is not known, the kinds of interaction they have with clays are less well known than those of well-defined organic compounds. However, some of the kinds of reactions described above are probably involved. It is obvious that clays eroded from soil surfaces into streams and lakes will probably be, to some degree, complexed with organic matter. Humic acids, a constituent of soil organic matter, may be strongly adsorbed by clays, presumably by interaction with positive sites on the edges of clay particles or with polyvalent cations on the cation exchange complex acting as "bridges." Schnitzer and Kodama (1967) have shown that fulvic acid (another constituent of soil organic matter) adsorption depends on pH, and is greater under acid than alkaline conditions. This is to be expected, since the fulvic acid would be relatively undissociated at low pH but considerably more anionic at alkaline pH. Schnitzer and Kodama (1972) showed that fulvic acid is very strongly bound to Cu2+ on the exchange sites of montmorillonite through a coordination type of reaction. In addition, they have shown such adsorption is typical for any electrophilic cation on the exchange complex, particularly for ions of the transition metal group. Summary Pollutant concentrations are higher in sediments than in the waters with which they are associated. It should be recognized that the consequences of pollutant adsorption by clays may be very important in natural systems and may affect drinking water quality. Clay-pollutant complexes may be mobilized by erosion from the landscape, or form when eroded clay enters a stream containing a polluting species. If the complex survives water treatment and enters the drinking water system, it would then be available for ingestion by humans. In the adsorbed state on the clay surface the pollutant is probably not toxic, but the possibility exists that the pollutant might be released from the clay in the environment of the alimentary tract and thus exert toxic effects. Whether or not such a process might take place would depend on the complex in question, so that no generalities are possible. Information is completely lacking in this area, and thus research should be encouraged and supported.

144 DRINKING WATER AND HEALTH ASBESTOS: NOMENCLATURE, OCCURRENCE AND REDISTRIBUTION IN WATER Structure and Nomenclature Asbestos is the name for a group of naturally occurring hydrated silicate minerals possessing fibrous morphology and commercial utility. This definition generally limits application of the term to the minerals chrysotile, some members of the cummingtonite-grunerite series, crocido- lite, anthophyllite, and some members of the tremolite-actinolite series. Amosite is commonly used to refer to a cummingtonite-grunerite asbestos mineral, but it is a discredited mineral name (Rabbit 1948; Committee on Mineral Names, 1949~. Mode of occurrence and fiber length are important determinants of commercial value. Of the commercially mined and processed asbestos minerals, chrysotile accounts for about 95%, the remainder being amosite and crocidolite (May and Lewis, 1970~. Crocidolite is the fibrous equivalent of riebeckite, and chrysotile belongs to the serpentine group of minerals, which contains other nonfibrous members (Deer et al., 1970~. Noncommercial deposits of asbestos minerals are also relatively com mon. The standard definitions of the Glossary of Geology (American Geological Institute, 1972; second printing, 1973) are given below. ASBESTOS: (a) a commercial term applied to a group of highly fibrous silicate minerals that readily separate into long, thin, strong fibers of sufficient flexibility to be woven, are heat resistant and chemically inert, and possess a high electric insulation, and therefore are suitable for uses (as in yarn, cloth, paper, paint, brake linings, tiles, insulation cement, fillers, and filters), where incombustible, nonconducting, or chemically resistant material is required. (b) a mineral of the asbestos group, principally chrysotile (best adapted for spinning) and certain fibrous varieties of amphibole (ex. tremolite, actinolite, and crocidolite). (c) a term strictly applied to the fibrous variety of actinolite. Syn: asbestos; amianthus; earth flax; mountain leather. ASBESTIFORM: Said of a mineral that is fibrous, i.e. that is like asbestos. ACICULAR (Cryst): Said of a crystal that is needlelike in form. Of: fascicular, sagenitic FIBROUS: Said of the habit of a mineral, and of the mineral itself (e.g. asbestos), that crystallizes in elongated thin, needle-like grains, or fibers. The nomenclature used in this report conforms generally to these definitions, subject only to the further qualifications that the term asbestos will not be used in its most restrictive sense (c, above);

Solid Particles in Suspension 145 asbestiform will not be used; and the terms acicular and fibrous are to be understood as discussed below. The term asbestos has often been used in recent scientific literature to describe individual fibrous or acicular particles of microscopic and submicroscopic size. However, mineralogists and geologists have has- tened to point out that the term should be used only as defined above, in reference to the minerals in bulk. Ampian (1976) considers the terms asbestos and asbestiform to be synonymous and that they then may only be used to apply to the bulb fibrous forms occurring in nature. Asbestiform is often used to define the morphology of a mineral that is similar to asbestos, but does not necessarily occur in nature in a commercial deposit; to avoid ambiguity, the term will not be used here. The terms acicular and fibrous are used here to characterize any mineral particle that has apparent crystal continuity, a length-to-width aspect ratio of 3 or more and widths in the micrometer or submicrometer range. Although the two terms are not strictly synonymous, the use here of either one to describe a mineral particle should be taken to imply the other, unless otherwise qualified. Table IV-2 lists some of the naturally occurring minerals that can have, but do not always have, an acicular morphology. To this list could be added a number of synthetic fibers, although they are not naturally occurring minerals. Many of the minerals in Table IV-2 are common rock-forming minerals. Properties of Asbestos Minerals MINERALOGY The asbestos minerals belong to the serpentine and amphibole groups, and the amphiboles are further divided into those of the orthorhombic crystal system (orthoamphiboles) and amphiboles of the monoclinic crystal system (clinoamphiboles). Table IV-3 summarizes the basic properties of the asbestos minerals. Chrysotile is the asbestos mineral of the serpentine group. Its crystal structure is a double sheet, comprising a layer of silica tetrahedra and a layer of magnesia octahedra, arranged in a manner that is somewhat analogous to the alumina octahedra silica tetrahedra layering of kaolinite. The way in which the sheet structure is modified to develop a fibrous morphology is, in detail, very complex; but in essence the modification can be imagined as a buckling of the double sheet, due to misfits, to form a hollow tube (Deer et al., 1966~. This central tube may or may not be filled with electron-opaque material, and the appearance of its

146 DRINKING WATER AND H"LTH TABLE IV-2 List of Some Minerals That May Occur in Acicular- Fibrous Form. Major Element Stoichiometry Has Been Simplified in Many Cases. Sili~ at~ s Epidote Lawsonite Sillimanite (fibrolite) Pectolite Kyanite Mullite Diopside-hedenbergite Enstatite Wollastonite Jadeite (Nephrite) Anthophyllite Tremolite-actinolite Hornblende Crocidolite (glaucophane-riebeckite) Cu mmingtonite-grunerite Eckermannite-arfvedsonite Halloysite Nontronite Palygorskite-attapulgite-sepiolite Chrysotile (Se~pentine) Talc Prehnite Na~olite Scolecite Thomsonite Stilbite Mordenite Ferrierite Nonsilicutes Cassiterite Goethite-limonite Rutile Nemalite (brucite) Stibnite Bismuthinite Marcasite Jamesonite Gypsum Celeshte Alum Aragonite Lublinite  Siderite Apadte Vivian~te Ca2FeAI2O(Si2O7)0H ;(Si2O4) CaAI2(0H)2Si2O7 H2O Al2SiO,, Ca2NaH (SiO ])3 Al2SiO,, 3AI2O3 2SiO2 Ca(Mg,Fe) (Si2O6) MgSiO 3 CaSiO~ NaAI(Si2O6) (Mg,Fe)7SiHO22 (OH.F)2 Ca2(Mg.Fe2), (S1HO22) (OH,F)2 (Ca,Na.K2)2_3 (Mg,Fe2,Fe3.AI),, [Si6(Si.AI)2] 022 (OH,F)2 Na,,Fe32Fe~ (Si#022) (OH.F)2 (Fe,Mg)7 (S`HO22) (OH)2 Na3 (Mg.Fe)4AI (Si#022) (OH.F)2 Al2Si2O. (OHh (Ca,Na), .:3(Mg, Fe)4(S; `,. 3AI`,.7)#02`,(0H)4 nH 20 (Mg,Al),, Si#022 (OH)6 4H2O Mg3Si2O,(OH)4 Mg6Si#020(0H)4 Ca2AI (Al,Si3O~o) (OH)2 Na2(AI2Si3O,O) 2H2O Ca(AI2Si3O,O) 3H2O NaCa2((AI, Si)50,o)2 6H2O (Ca.Na2.K2) (Al2Si7O,H) 7H2O (Na2,K2,Ca) (Al2Si,0024) 7H2O (Na,K)4Mg2(AI6Si300,2) (OH)2 nH2O SnO2 FeO OH nH2O TiO2 Mg (OH)2 Sb2S3 Bi2S3 FeS2 2PbS Sb2S3 CaSO4 2H2O SrSO4 KAI (SO4)2 1 2H2O CaCO3 CaCO3 FeCO3 Ca5 (PO4,CO3) (OH,F) Fe3 (PO4)2 8H2O

Solid Particles in Suspension 147 image under the electron microscope will be affected accordingly (Yada, 1967~. The chemical composition of chrysotile shows comparatively little of the great variability that is found in the amphiboles. Double chains of silica tetrahedra and metals, octahedrally linked, form the basic structural elements of the amphiboles. The net result is a prominent cleavage parallel to the double chain, and this generally produces an acicular or prismatic habit. Amphiboles show a wide-range of chemical composition, reflecting the temperature, pressure, chemistry, and metamorphic history of formation. This wide variation in composi- tion is due to substitution in the basic silica tetrahedra, and the ability of the crystal structure to accommodate a wide variety of different coordinating cations; the result is that all amphiboles, including the asbestos-forming amphiboles, show overlapping composition. Antho- phyllite and cummingtonite also have overlapping chemical composi- tions, as do crocidolite, tremolite-actinolite, and hornblende. Analytical results in the literature often refer to "amphibole fibers," implying that these are understood to be derived from asbestos. However, many nonasbestos amphibole minerals form fibrous or acicular particles when finely divided. SURFACE PROPERTIES The surface areas of fibrous particles have an important bearing on coagulation and adsorption processes and hence on the ultimate fate of the particulates in the environment. The surface areas of UICC (Unio Internationale Contra Cancrum) reference samples have been measured, using both the nitrogen adsorption and the permeability methods (Rendall, 1970~. The specific surface of chrysotile is about twice that of the amphibole asbestos minerals. This difference may be due to the greater length-to- width ratio of chrysotile and its porous tubular morphology. The isoelectric point, the pH at which the net surface charge of a mineral in an aqueous solution is zero, has been measured for chrysotile and cummingtonite. Chrysotile has an isoelectric point of 11.8, and the isoelectric point of cummingtonite is 5.2-6.0. Anthophyllite may have a value similar to that of chrysotile, and the other amphibole asbestos minerals may have values similar to those of cummingtonite (Parks, 1967~. As the pH of the medium falls below the isoelectric point, the surface charge of suspended particles tends to become more positive. Therefore, in a typical drinking water, chrysotile particles should be more positively charged than those of the amphiboles. However, other suspended or

148 Ed 1 It) ~ N v 7 ~ o ~ ~ ~ ~ E ~ D '8 = $ ~ ~ E Cat ~ E ~ Do ~ C ~ 0. ~ = ~ 9 D § ~ E O E ~ _ ~0 In D o ~0 of o .~ Em - (V N 8 ~o ~ ~ ~ ° e4 o + C) ~ _ ~ V,  o := ~U. o i ~

149 -- II o ~1t 11 ? Ii ~ ~ E E E C) .~ ~o ° ° II o ~1I 11 ? It C D :t E D U. en . ° TV o o 11 co ~1I 11 ? Ii ~ ~ ~ E _ _ ~ ~ I~ ~ARC ·a o o It ~ ~- - ~ o . ~O Ct-D ~ ,, .m ~ ~ ~o ~? It tI it 3 E o ° i- o E ~ ~ O

150 DRINKING WATER AND HEALTH dissolved materials can interact with the asbestos minerals and modify their charge. MORPHOLOGY Fibrils are the individual tubes of single crystals that bundle together to produce a fiber. Chrysotile fibers are usually curved and occur in open bundles, splitting into smaller individual fibrils. Therefore, it is often difficult to carry out repeatable size measurements. The fibrils have various diameters, but the average outer diameter is about 200 A. Typical electron micrographs of chrysotile show cylinders, tube-in-tube, and cone-in-cone forms (Whittaker and Zussman, 1971~. Quite frequently there is a median stripe of greater or less electron density, suggesting that there is a difference in the core compared to the sides of the fiber. A tubular structure, either void or filled with an electron-dense substance, is the simplest explanation of this observation (Yada 1967; Whittaker, 1966; Kehieker et al., 1967; Whittaker and Zussman, 1971; House, 1967~. On the other hand, amphiboles are usually straight and show good cleavage edges parallel to the fiber length and often a second cleavage transverse to the length. The fiber width of UICC samples of anthophyl- lite often exceeds the fiber width of amosite, which, in turn, exceeds the average width of crocidolite. Approximate minimum fiber widths are: anthophyllite 2500 A, amosite 1500 A, and crocidolite 600 A (Timbrell et al., 1970~. Fiber-length distributions have been determined for the UICC reference samples (Rendall, 19701; laboratory experiments have shown that observed distributions depend to some extent on the method used to disperse the fibers for measurement (Timbrell and Rendall, 1972~. Table IV-4 shows the distribution of fiber lengths by number in UICC reference samples, as determined from the analysis of Rendall (1970~. When this distribution is compared to milled dusts, the UICC reference is seen to approach the lower limit of industrial dust in the case of amosite, and to approach the upper limit of industrial dust in the case of chrysotile (Rendall~ 19701. Both in water and in air, the longer fibers (~2 ,um) are much rarer in the general environment than in occupational settings. ~ , , SOLUBILITY Acid dissolution of asbestos minerals has often been studied in order to test resistance to corrosion (Cotterell and Holt, 1972; Spell and Leineweber, 1969; Choi and Smith, 1971~. In general, the resistance to

Solid Particles in Suspension 151 TABLE IV-4 Length Distribution of UICC Reference Samples (after Rendall, 1970) % by Number of Length Greater Than (~m): Mineral Type 0.2 0.5 1.0 2.0 5.0 10 25 50 100 - Amosite 99.9 76.S 45.9 20.3 5.69 1.29 0.21 0.05 0.02 Anthophyllite 99.9 78.2 45.5 22.9 4.79 1.29 0.14 Crocidolite 100 71.6 35.8 13.3 3.01 0.68 0.08 0.01 0.01 Chrysotile A 99.9 79.3 44.4 21.3 6.09 3.26 0.77 0.15 (Rhodesian) Chrysotile B 100 69.4 36.0 16.2 3.00 1.24 0.31 0.07 (Canadian) acid solution is Chrysotile < ~ amosite < actinolite < crocidolite < anthophyllite < tremolite. Occurrence of Asbestos and Fibrous Minerals Fibrous minerals must be considered to be common wherever unconsoli- dated sediment occurs, because many of the common rock-forming minerals can have an acicular morphology (Cralley, et al., 1969) The wide variation in composition, and the crystallographic similari- ties, as well as the ubiquitous occurrence of fibrous or acicular minerals, make analysis of environmental samples most difficult. Most analytical studies consider only minerals of the asbestos group. Asbestos minerals generally occur in metamorphic retrograde deposits. Chrysotile occurs in low-grade metamorphic deposits, and the amphi- boles occur in slightly higher-grade metamorphic deposits. Metasoma- tism may be a key process in the formation of the asbestos minerals. Deposits of commercial grade asbestos are mined in Canada and the United States. Chrysotile accounts for 95% of the world production with amosite and crocidolite accounting for most of the remainder. The largest-known deposits of Chrysotile in the world are found in a belt 125 km by 10 km between Danville and Chaudiere, Quebec. Other Canadian deposits are found in northern Ontario (Matachawan), in northern British Columbia (Cassiar), and Newfoundland (Baie Verte). In the United States, extensive deposits are found in California, Vermont, Arizona, and North Carolina. The North Carolina deposits are amphi- boles, whereas all the other deposits consist of chrysotile. Additional deposits, presently unexploited, are found in other states, notably Montana and Wyoming.

152 DRINKING WATER AND HEALTH Uses and Redistribution There are over 2000 recorded uses of asbestos minerals in the United States (May and Lewis). From this one can conclude that there will be a correlation between population and industrial activity and the concentra- tion of asbestos in the environment. Higher concentrations of asbestos fibers are commonly found in the air and waters around metropolitan areas (Cunningham and Pontefract, 1971; Kay, 1973~. The following tabulation of uses of asbestos fibers, consisting almost entirely of chrysotile, is of the United States in 1968 (May and Lewis, 1970). Construction DEMO Cement products 69~o Floor tile loo Paper products 756 Paint and caulking 2~o Transportation 3% Textiles 2% Plastics loo Other 6% Redistribtion of asbestos minerals to the environment depends on the extent to which the material is conserved or sequestered in the processes of mining, manufacturing and consumption. Emission factors, expressing the fractional loss of asbestos to the environment (on a mass to mass basis) from the various processes and uses, have been estimated in two different studies (Davis, 1970; Environment Canada, 1973), and are shown in Table IV-S. When considering redistribution of asbestos in the environment, it is important to distinguish between the mass of material involved and a more pertinent consideration the number of fibers of specified dimen- sions. Note that the production and use estimates, modified by the emission factors, give estimates of emission on a mass basis that can not be directly related to numbers of fibers of particular size ranges. (Size and number are discussed in the following sections on health effects.) Asbestos fibers occur in bundles, individual fibers, and fibrils and in varying size ranges in the environment. It is not possible, in a general sense, to relate mass to number. The natural background concentrations of asbestos fibers in water and in air are not well known. Some minimum concentrations from remote areas suggest that concentrations of 104 to 106 fibers per liter represent approximate background figures for water, and an ambient air concentra

Solid Particles in Suspension 153 TABLE IV-5 Emission Factors for Asbestos (Chrysotile) for the United States (Davis, 1970) and Canada (Environment Canada, 1973) Emission Factor (g/kg) Canada United States Production Asbestos mining 4.0 4.5 Asbestos milling 6.0 42 Manufacturing Cement products 0.075 0.5 Floor tile 0.075 0~5 Paving 0.075 Coatings/caulking 0.075 0.5 Insulation 0.15 Friction materials 0.45 3.0, Plastics 0 075 Textiles 0.15 1.0 Paper 0 075 0 5 Miscellaneous 0.075 0.5 Consumption Construction industry 0.075 Sprayed insulation 5.0 5.0 Brake linings Installation 5.0 5.0 Wear 20 5.0 NOTE: Emission factor= mass of asbestos (g) emitted to environment per unit mass (kg) of asbestos in use. lion of about 0.01 fiber/cc is an approximate background estimate for air. It may well be that these figures represent a lower limit of detection in a modern analytical laboratory surrounded by materials containing asbestos. In regions that are remote from industrial and populated areas, the -median lengths of fibers found in water are generally less than those found in air. Median fiber lengths of ~ 1 ,um are typical, and fibers greater than 2 ,um are uncommon in such cases (Durham and Pang, 1976; Kay, 1973~. There are reports that asbestos fibers reach the environment by natural weathering (NAS-NRC, 1971), but this hypothesis has not been substantiated for areas where asbestos minerals occur naturally and that have not been disturbed by man. Asbestos cement pipe is used in some localities to transport drinking  water and there has been some concern over the release of fibers from these pipes (Wright, 1974~. Studies have been carried out on the solubility

154 DRINKING WATER AND HEALTH of concrete and cement (Flentje and Schweitzer, 1955; Kristiansen, 1974~. In general, almost all soft waters (alkalinity and hardness about 1O-4 equivalents/liter) should dissolve calcium carbonate in the pipe, whereas hard waters typically would not do so. In addition to solubility studies, fiber analyses are being carried out in a pipe-loop laboratory as well as in asbestos cement pipe in soft water areas (EPA, 19754. Results are tentative at present, but they suggest that some fibers are emitted in corrosive waters. Exposure of asbestos and other fibrous particle-forming minerals to redistribution may result from the presence of the material in a gangue. The best documented case of this kind of emission is that of the release of fibrous cummingtonite from taconite ore tailings into the Western arm of Lake Superior (Cook et al., 1974~. This amphibole occurs together with quartz, tremolite-actinolite, and other minor minerals in both fibrous and nonfibrous forms. A tabulation of the approximate emission rates of fibrous cummingtonite and associated materials gives: Total missions cummingtonite-grunerite concentration Fiber c ant entrati`'n. emissions Annual emissions total cumm~ngtonite-gruner~te no. of fibers weight of fibers 70.000 metric tons/day 40 wt. ho ~ x 109 f bers/g dry wt. 0.006 g of fiber/g dry wt. 10 million metric tons 2.3 x 1023 150.000 metric tons It is interesting to note that the consumption of asbestos in the United States in 1968 was about 840,000 metric tons (May and Lewis, 1970) or about 5.5 times the fiber emissions into Lake Superior. If we assume that the average emission factor for asbestos is between 0.08 and 0.15 g/kg (Table IV-5), the total emissions of asbestos (chrysotile) in the United States can be estimated to be about equal to the fibrous emission into  western Lake Superior cited above, on a mass basis. The ultimate fate of the fibrous cummingtonite-grunerite is not well documented. The concentrations of cummingtonite-grunerite fibers in the western arm of Lake Superior are generally >10 million/liter. The central and eastern portions of Lake Superior have concentrations of 2-3 million fibers/liter, those in nearshore areas being somewhat higher (Kramer, 1976~. In comparison, a number of inflowing streams have

Solid Particles in Suspension 155 concentrations of 1 to 2 million fibers/liter. If the lake were homogeneous- ly mixed, and no fibers were transported out of the lake, one would expect between 3 million and 5 million fibers/liter from measurements of the emission. From these calculations, one may estimate that 50~o or more of the fibers are removed from suspension in the water of Lake Superior, presumably by sedimentation. Topics for Research in Mineralogy One of the more important questions is what differences, if any, exist between fibers derived from asbestos and those that arise from single crystals or cleavages of single crystals. Furthermore, it is important to be able to develop analytical methods to define these differences and to relate these differences to health effects. In a recent summary of the mineralogical aspects of this research, Zoltai (1976) concluded that 1) "the mechanism responsible for the development of asbestiform habit may be related to unique surface structures and properties of individual fibers that make up a bundle, 2) electron microscope studies of 'amosite' asbestos fibers reveal the presence of narrow bands of polysynthetic twinning and of triple chains interlayered with the usual double chain structure of the amphiboles, 3) most of the natural asbestos ('amosite') fibers have apparently orthorhombic optical properties at a scale of several microns, and 4) some natural asbestos contains adsorbed metals and compounds." Reasons for the development of fibrous habit are not known. Several workers have shown details of variations in structure in asbestos minerals, and these variations must be compared with small- scale cleavage fibers from nonasbestos deposits (Chisholm, 1973; Hart- man, 1963; Hutchinson et al., 1975; Ruud et al., 1976~. ASBESTOS FIBER SAMPLING AND ANAYLSIS Introduction Several methods have been used to identify fibrous or acicular particu- lates of asbestos minerals and determine their concentrations in air, water, mineral samples and biological tissue. These include optical and electron microscopy, X-ray diffraction, and differential thermal analysis. Identification of fibers of asbestos origin and their estimation in water and tissue samples is difficult for a variety of reasons:

156 DRINKING WATER AND HEALTH 1. Asbestos mineral fibers in water are generally present in low mass concentrations even though the number density of fibers may be high. 2. Many analytical methods can not distinguish between fibers of asbestos origin and particles of other nonasbestos and nonfibrous minerals. 3. Mineral fibers present in water samples are generally smaller than can be resolved with the optical microscope, hence the electron microscope must be used. Methods of estimating mineral fibers in water, and their limitations, are discussed below. Optical Microscopy Occupational exposure to asbestos fibers in air is monitored, in the United States, by collecting the fibers on contrast optical microscopy. For optical microscopy, asbestos fibers are defined as those particles of length greater than 5 ,um and a length-to-diameter ratio of 3 to 1 or greater. This method may not be specific for asbestos mineral fibers (by fiber definition), nor can it detect fibers less than about 0.5 Am in diameter. Petrographic microscopy may be used to identify asbestos mineral fibers greater than approximately 0.5 ,um in diameter. With the polarizing microscope, various optical crystallographic measurements such as refractive index, extinction angles and sign of elongation may be made and compared with data reported for standard asbestos mineral reference samples. Dispersion staining with polarized light has been reported as a method for identifying asbestos mineral fibers (Julian and McCrone, 1970~. Dispersion staining colors may vary depending on the geographic area in which the asbestos was mined. Fibers less than 0.5 ,um in diameter can not be identified by this method. Asbestos fibers present in water and tissue samples are generally too small in diameter for analysis by any of the above optical microscopic methods. Electron Microscopy Both transmission and scanning electron microscopy have been used for mineral fiber identification and estimation. In addition to morphological observation, selected area electron diffraction and microchemical analy- sis may be used to identify fibers. In addition to their superior resolving power, most modern transmis

Solid Particles il n Suspension 157 sion electron microscopes can be used to observe electron diffraction patterns. Crystalline materials diffract electrons in regular patterns that are indicative of their crystal structures. Visual observation of single fiber (single crystal) electron diffraction patterns may be used to differentiate chrysotile fibers from amphibole fibers (Larger et al., 1974; Timbrell, 1970; Clark and Ruud, 1975~. The hollow "central core" of chrysotile fibers may also aid in fiber identification but beam damage may cause this feature to disappear (Larger et al., 1974~. Other fibrous minerals may also have hollow cores. The amphibole mineral fibers are generally straighter in appearance than those from chrysotile. Selected area diffraction patterns for the amphibole asbestos minerals may be similar in appearance; therefore, casual visual observation of these patterns is sufficient only for clas- sification of the fiber as being "an amphibole asbestos" (Larger et al., 1974; Cook et al., 1974; Clark and Ruud, 1975~. Electron diffraction patterns of asbestos amphibole minerals are sometimes streaked perpen- dicular to the fiber length. It is then necessary to study the intensity distribution of the spots within the layer lines to distinguish them from chrysotile. In addition to visual observation of electron diffraction patterns for fiber identification, photographs can be made of the diffraction patterns and crystal "d" spacings measured from the photographic plates and calculated using the instrument camera constant (Timbrell, 1970~. Both "spot" and polycrystalline patterns may be measured, but the precision of measurement is not nearly that obtainable with X-ray diffraction. It must be born in mind also that intensities are not the same as observed for X- ray powder patterns, and additional reflections may be present. Although some laboratories are confident of their ability to perform positive fiber identification by electron microscopy with electron diffraction, the method is empirical and has not been rigorously tested. The possibility still remains that other nonasbestos mineral particles may sometimes give diffraction patterns characteristic of chrysotile, or of the fibrous amphiboles, in certain orientations in the electron beam. Electron beam microchemical analysis may sometimes be used to distinguish microscopic minerals from other fibrous particles (Rubin and Magiore, 1974; Ferrell et al., 1975; Maggiore and Rubin, 1973; Langer et al., 1975~. The most common system presently in use is the energy dispersive X-ray detector, in combination with a scanning or transmis- sion electron microscope. X-ray wavelength dispersive analyzers and the conventional electron microprobe have been used; however, their routine application is limited due to data acquisition times, small particle size, and to other considerations discussed below.

158 DRINKING WATER AND HEALTH Electron beam X-ray microanalysis coupled only with morphology has often been cited as a method for identifying microfibers (Rubin and Maggiore, 1974; Ferrell et al., 1975~. However, many scientists have criticized this technique as having several major faults (Ruud et al., 1976~. First, the elemental composition of amphibole minerals is variable, and seldom ideally stoichiometric. Furthermore many nonamphibole miner- als are similar in composition to the asbestos amphiboles and sometimes produce microscopic fibers. Second, the chemistry of the environment, e.g., high ion water, can change the actual or apparent composition of the fiber by ion exchange or surface coating. Third, the variables affecting quantitative microanalysis are such that a plus or minus logo elemental variance is not uncommon from a single prepared standard sample (Ferrell et al., 1975~. Electron microscopy with microchemical analysis for fiber identification is reliable only when combined with considerable knowledge of the mineralogy of the fiber source. Such knowledge is generally lacking in the case of water samples. As has already been discussed, possession of proper elemental intensities by energy dispersive X-ray analysis is generally not sufficient for positive identification of fibers. For example, chrysotile, anthophyl- lite, and fibrous talc, which have similar elemental compositions, may be difficult to differentiate (Rubin and Maggiore, 1974; Ruud et al., 1976~. However, these materials may easily be distinguished using selected area electron diffraction (Ruud et al., 1976; Langer et al., 1975~. Transmission electron microscopes equipped with energy dispersive X-ray detectors allow simultaneous observation of morphology, crystal structure, and elemental composition. These microscope systems have been used to study fibers of known asbestos origin as well as environmental and material samples (Cook et al., 1974; Dement et al., 1975~. The probability of positive fiber identification (including amphibole minerals) is greatly enhanced with these techniques. Identification and counting of fibrous or acicular particles in environ- mental and tissue samples has been accomplished by electron microsco PY Langer et al. (1975) have reported a qualitative method for preparing tissue samples for electron microscopy. This method involves first dissolving the tissue in a 40% solution of potassium hydroxide followed by centrifuging. After centrifuging, the residue is dispersed in distilled water and a drop pipetted onto Formvar-coated 200-mesh electron microscope grids. Similar techniques have been used by Pontefract and Cunningham (1973~. Concentrations of fibrous or acicular particulates in environmental samples and biological tissue are usually expressed as fibers per unit

Solid Particles in Suspension 159 volume or weight of sample (fibers/m3,fibers/l, fibers/gm dry lung, etch. These concentrations are determined by counting fibrous or acicular particles within calibrated areas on the electron microscope viewing screen or counting fibers in photographs. Fiber concentrations in water samples determined by laboratories using the same mounting techniques have been reported to vary by a factor of 2 to 3 (Cook et al., 1974~. Much larger variations have been reported between laboratories using different techniques (Brown et al., 1976~. The mass of chrysotile concentrations in environmental samples has also been determined using electron microscopy. This is accomplished by measuring the length and diameter (volume) of each fiber and calculating mass using the appropriate density (Seliko~ et al., 1972~. The accuracy of this method has not been studied in detail. X-Ray Diffraction X-ray powder di~ractometry is one of the standard mineralogical techniques used in the analysis of solid crystalline phases. X-ray diffraction has been widely used for identification and quantitation of fibers in bulk materials such as talc (Stanley and Norward, 1973; Rohl and Langer, 1974) and other industrial materials (Crable and Knott, 1966a, 1966b; Keenan and Lynch, 1970~. X-ray diffraction has also been used to study amphibole asbestos contamination of water samples (Cook et al., 1974~. However application of X-ray diffraction to routine analysis of environmental samples has been limited. Birks et al. (1975) have reported a feasibility study concerning quantitative analysis of airborne asbestos. This technique, however, is still in the research phase and has not been developed to the reliability of a general analytical method. It must be recognized that, with the exception of the technique described above, X-ray diffraction methods are not capable of differentiating between asbestos mineral fibers and their nonfibrous mineral counterparts. This fact, combined with the relatively low sensitivity, suggests that analysis of environmental samples by X-ray diffraction should be confirmed by electron microscopy. Differential Thermal Analysis Differential thermal analysis has been used to determine asbestos mineral fiber levels in talc samples (Schlez,1974~. It has not been used for environmental samples because it has low sensitivity. Moreover, it is not capable of differentiating between asbestos fibers and their nonfibrous mineralogical polymorphs.

160 DRINKING WATER AND HEALTH BIOLOGICAL EFFECTS OF ASBESTOS MINERALS Epidemiological Findings Numerous epidemiological studies have shown that occupational expo- sure to asbestos dust can lead to asbestosis (characterized primarily by pulmonary fibrosis); the formation of pleural plaques; a greatly increased risk of bronchogenic carcinoma; pleural mesothelioma; and peritoneal mesothelioma (Seliko~ et al., 1973; Newhouse, et al., 1972; Elmes and Simpson, 1971; Seliko~ and Churg, 1965; Bogovski et al., 1973; and Lee, 1974~. There is also evidence that the elevated risk of bronchogenic carcinoma as a result of occupational exposure to asbestos dust is largely (though perhaps not entirely) confined to cigarette smokers (Hammond et al., 1975~. Different types of asbestos may vary in their potency in relation to the ejects mentioned above; but this has been difficult to evaluate owing to lack of precise information on degree of exposure, and various other problems. Although there is little evidence concerning the possible ejects of nonoccupational exposure, some cases of mesothelioma have been found among persons working in or living near shipyards and the wives of asbestos workers who handled the dusty clothes of their husbands (Newhouse and Thompson, 1965; Wagner et al., 1960; Anderson et al., 1976). There has been much discussion of the possibility that the ejects of inhaling asbestos mineral dust valor greatly with the length of the particles. However, no epidemiological evidence is available on this matter. The dust to which industrial workers are exposed typically consists of fibers varying in length from very long to extremely short. Fibers sufficiently large to be seen under a light microscope are invariably accompanied by large numbers of far smaller fibers. It should be noted that all of the epidemiological evidence noted above has come from studies of people who have been exposed to dust from types of asbestos minerals that have been used commercially. There is no direct epidemiological evidence on the ejects of various other types of fibrous minerals, some of which may perhaps find their way into drinking water. The fact that exposure to air heavily polluted with asbestos mineral fibers often leads to the diseases mentioned above does not necessarily indicate that drinking water contaminated with an equally large number of such fibers may lead to the same diseases or perhaps some other diseases. (Site of cancer can vary depending upon the way in which individuals are subjected to a carcinogenic agent, for example: skin

Solid Particles in Suspension 161 TABLE IV-6 Observed vs. Expected Number of Deaths from Cancer of Several Sites among Two Groups of Workers Occupationally Exposed to Asbestos Dust. Data Provided by Selikoff, Hammond, Seidman, and Churg. Group #1. Asbestos Insulation Workers (United States and Canada) Group #2. "Amosite" Asbestos Factory Workers (New Jersey) Obs. Exp. Obs. Exp. Site of Cancer Deaths Deaths Ratio Deaths Deaths Ratio Buccal and pharnyx 15 7.87 1.91 6 2.04 2.94 Esophagus 14 5.38 2.60 0 1.36 TOTAL 29 13.25 2.19 6 3.40 1.76 Stomach 18 11.23 1.60 10 4.89 2.04 Colon-Rectum 47 28.64 1.64 16 7.65 2.09 TOTAL 94 53.12 1.77 32 15.94 2.01 exposure, inhalation, or ingestion.) However, the hypothesis is tenable to the degree that it cannot be ruled out of consideration without evidence to the contrary. It is important to note that workers exposed to air containing large numbers of asbestos minerals fibers inevitably ingest such fibers. Many of the fibers that are inhaled are later propelled upward from the lungs and trachea, enter the mouth, and are then swallowed. Thus, fibers are brought into direct surface contact with the epithelial lining of the buccal cavity, esophagus, stomach, and intestines. For this reason, it is pertinent to inquire whether death rates from cancer arising in these tissues are higher among asbestos workers than in the general population, age, sex, and calendar-years of exposure being taken into consideration. This has been investigated by Seliko~, Hammond, Seidman, Churg, and their associates. The data shown in Table IV-6 were provided by these investigators. It shows, for each of two groups of workers, the observed and expected number of deaths from cancer of sites directly exposed to asbestos fibers by way of ingestion: buccal cavity and pharynx, esophagus, stomach, colon, and rectum. The two groups are described below. Group # 1. The entire membership of the insulation workers union in  the United States and Canada was registered as of January 1, 1967 (17,800 men). Many characteristics of the men were recorded, including date of

162 DRINKING WATER AND HEALTH birth and onset of work with asbestos. These subjects have been traced through December 31, 1974. Group #2. This group consisted of the entire workforce (1941~5) of a factory in an eastern U.S. city that manufactured "amosite" asbestos insulation materials (including Insulating block and pipe covering and asbestos mattresses), for use by insulation workers in the construction industry, especially in ship construction and repair. The plant opened its doors in June 1941 and remained in business until November 1954. Very few of these production workers studied had had prior occupational exposure to asbestos. Between 1941 and the end of 1945, a total of 933 men were hired; of these, some worked for as little as one day, while others worked until 1954, when the plant closed. Of the 933 men, 881 (94.4~0) were traced through December 31, 1973, and the remaining 52 were traced for a part of the period but later lost to follow-up. For both groups, "observed," as shown in Table IV-6, means the number of men who were found to have died of the indicated cause during the period of observation. "Expected" numbers of deaths are based upon age-specific death rates for white males in the United States during each year of observation. Group # 1 was by far the larger of the two groups, and in consequence the figures are more stable statistically. In this group, the observed number of deaths was substantially higher than the expected number for cancer of each of the four sites. The findings were similar for Group #2, except in respect to cancer of the esophagus (zero observed versus 1.36 expected deaths). In each of the two groups, the excess of observed over expected number of deaths (for all four sites of cancer combined) was highly significant statistically. It may be argued whether U.S. white males are an appropriate control group in relation to Group #2 (New Jersey "amosite" workers). Therefore, the expected numbers for Group #2 were recomputed based upon New Jersey mortality data. These expected numbers were a little higher than those shown in Table IV-6. Similar findings in respect to gastrointestinal cancer have been reported by Mancuso (1965) and by Elmes and Simpson (1971~. This, taken together with the data presented in Table IV-6, strongly suggests that ingestion of asbestos mineral fibers can result in an increased risk of cancer of several sites. If so, the degree of risk is probably highly dependent upon the number of fibers ingested and the duration of exposure. It may also depend upon the type and size of the fibers, as well as upon other agents to which the individuals have been exposed. For example, both smoking and alcohol increase the risk of cancer of the buccal cavity; and it is possible that exposure of the buccal mucosa to

Solid Particles in Suspension 163 asbestos mineral fibers simply increases the risk of cancer occurring among persons also exposed to one or both of these other two factors. As described elsewhere in this report, mineral fibers-varying in concentration from trace amounts to hundreds of thousands per liter, or more are present in drinking water in various locations in the United States. The situation in Duluth, Minn. (as described elsewhere), is of special interest from an epidemiological point of view. Studies of time trends in cancer death rates and incidence rates in Duluth (as compared with other cities) do not show any increases that could reasonably be attributed to the mineral fiber pollution of the drinking water of that city (Mason et al., 1974; Levy, et al., 1976~. Among persons occupationally exposed to asbestos dust, the increases in death rates from cancer of the stomach, colon, and rectum do not reach measurable proportions until many years after onset of exposure. The lapse of time since Duluth water was first heavily polluted with mineral fibers is not yet sufficient to have produced a significant increase in death rates from these cancers, even if it be assumed that the risk for residents of Duluth is as great as the risk for workers occupationally exposed to asbestos dust. If the risk is not so great but still of important proportions it will probably become apparent within another 5, 10, or 15 yr from now. An even lower risk might not become apparent for a much longer period of time and then it would be difficult if not impossible to pinpoint the cause. Experimental Studies As discussed above, epidemiological studies of the population exposed have failed to detect any increase in gastrointestinal cancer that might be ascribed to the asbestos mineral fibers contained in some drinking water. But, since these results do not indicate conclusively that ingestion of water containing such fibers is without risk, and because these substances are widely distributed, it is important also to consider experimental evidence. An Advisory Committee on Asbestos Cancers to the Director of the International Agency for Research on Cancer, meeting in 1972, consid- ered whether there was evidence of an increased risk of cancer resulting from asbestos minerals present in water, beverages, food, or in the fluids used for the administration of drugs. They concluded that "such evidence as there is does not indicate any risk" (IARC, 1973~. The group recommended that the effect of long-term ingestion of fibers of various sizes, shapes, and chemical compositions should be studied. The advisory committee's conclusion can hardly be taken as a categorical denial of

164 DRINKING WATER AND HEALTH risk, but rather reflected the inconclusive character of the evidence then available. Published results attest to the difficulties experienced by many investigators in studying, by means of experiments with animals, the biological effects and behavior of inorganic fibrous materials in general, and asbestos mineral fibers in particular. Several different approaches have been taken to the design of such experiments, including administra- tion of different types and preparations of asbestos minerals and glass fibers to several species of experimental animals, by ingestion (feeding), inhalation, surgical implantation and injection. Experiments have also been conducted to test the response to fibers of cell cultures in vitro. The results of some representative experiments of these kinds are reviewed below. Westlake et al. (1965, 1974) fed chrysotile to rats and noted the presence of fibers of the material in many sites in the colonic epithelium and lamina propria; and Webster (1974), after feeding crocidolite to baboons, showed that small numbers of fibers appeared in the ashed tissue of the gut wall. Gross et al. (1974) reported no such penetration. Bolton and Davis (1976), reporting on the short-term effects of chronic ingestion of asbestos minerals by rats, found no sign of cell penetration by fibers or damage to the gut mucosa. They concluded that penetration of the gastrointestinal wall, if it occurred, must have been on a very small scale. Pontefract and Cunningham (1973) reported that chrysotile, adminis- tered to rats intragastrically by direct injection, was later found to be distributed to other organs, and suggested that the material penetrated the digestive tract. Gross (1974) questioned this interpretation of the study on the grounds that the method of administration was likely to have produced contamination of other organs by leakage. Cunningham and Pontefract (1974) reported the appearance of asbestos mineral fibers in fetal organs after injection of asbestos into the femoral veins of pregnant Wistar rats. While none of these papers demonstrated transport of fibers across tissue boundaries, some included the suggestion that the reported deposition of the material in other organs implied that transport had occurred. Reports such as these have, without resolving the matter, lent interest to the question whether or not human ingestion of asbestos may be accompanied by similar effects. Experimental investigations with animals of the toxic effects of asbestos mineral fibers have not yet led to the development of an experimental model system that reproduces the putative effects of ingestion of such fibers by man. Nevertheless, various animal studies on

Solid Particles in Suspension 165 the toxicity of asbestos have been reported, and some large-scale feeding experiments are now in progress. In the experimental studies by Wagner et al. (1974), using the UICC reference samples, no association was found between exposure to asbestos by inhalation and excess gastrointestinal cancers. Thus, this animal study did not duplicate the positive epidemiological findings previously discussed. The possible ejects of ingested asbestos also were studied experimentally. Several animal feeding studies with asbestos are reported to have shown no increase in cancer with this route of exposure (Bonser and CIayson, 1967; Smith, 1973; and Gross et al., 1974~. The asbestos minerals tested included crocidolite, ch~ysotile, and "amosite." However, none of these studies would be considered adequate under modern constructs of chronic toxicity testing (National Academy of Sciences, 1975; Sontag et al., 1976), and so the possibility of false negatives remains. More recently, Gibel et al. (1976) fed filter material containing asbestos (20 mg/day) to 50 Wistar rats. This material was composed of asbestos (chrysotile, 53~O), sulfated cellulose and a condensation resin. The authors reported 12 malignant tumors in the group that was fed the filter material, 2 malignant tumors in a control group, and 3 in a similar group of rats that were fed the same amount of a standard grade of talc. The 12 malignant tumors appearing in the group receiving the filter material comprised 1 lung carcinoma, 3 reticular cell sarcomas, 4 kidney carcinomas, and 4 liver carcinomas. In the control animals and the talc- fed animals the malignant tumors were all liver carcinomas. The average survival times were 441 days in the filter-fed group, 702 days in the controls and 649 days in the talc-fed group. The presence of other substances besides asbestos in the filter material and the manner of reporting the malignant tumor counts suggests that further studies are needed, with asbestos alone, before firm conclusions can be drawn from the results of this study. A fundamental difficulty associated with these studies is that they seek to duplicate an eject in man by means of animal models that have neither been validated for the route of administration nor for the materials of interest. Negative results may be interpreted as indicating either the absence of the eject under investigation or, simply, as further proof that the model was not valid to start with. At present, the existing animal data base leaves the matter unresolved. Results obtained from some experiments with animals by direct application of fibrous materials, and with in vitro systems, have been adduced to support the view that ingestion of drinking water containing mineral fibers should not be expected to lead to any observed increase in

166 DRINKING WATER AND HEALTH gastrointestinal cancer in the general population. These experiments are interpreted to show the dependence of biological erect on the size of fibers. In this approach, large quantities of mineral fibers, of various size distributions, are injected into the pleural or peritoneal cavity. Stanton et al. (1969) wrapped a plaque of fibrous material around the lung in their animal model studies. This discussion will be limited to the results reported with respect to a carcinogenic response. The results obtained from these experiments, in which the material under investigation is deposited or applied directly, are limited by the statistics of their individual designs, methods of fiber preparation, and determination of fiber size distribution. They may be summarized as follows: 1. Tumor induction is related to fiber size, shape and durability, and not to the source of the fiber or its chemical composition (Stanton and Wrench, 1972; Stanton, 1973; Stanton et al., 1977; Pott and Fredricks, 1972; Pott et al., 1976). 2. In cell cultures, particles less than 20 Am in length do not induce growth of fibroblasts (Maroudas et al., 1973~. 3. Chrysotile containing particles less than 5 Em in length did not elicit a carcinogenic response (Smith, 1974; Smith et al., 1972~. 4. Defining a "significant" asbestos fiber as one less than 0.5 Am in diameter and greater than 10 Am in length, there was good agreement between the number of these "significant" fibers in an asbestos sample and its degree of carcinogenicity (Wagner et al., 1973~. 5. Man-made mineral fibers less than 1.5 Am in diameter and greater than 8 Am in length have the highest probability of inducing a biological response. The response appears to increase with increasing fiber length at fiber diameters less than 1.5 Am (Stanton et al., 1977~. In contrast to the results summarized above, two other investigations appear to show a different dependence of biological effect on fiber length. In one of these, Pott et al. (1972) injected 100 mg of two different preparations of chrysotile intraperitoneally, one sample contained 95% of particles less than 5 Am in length and the other (a milled sample) contained 99% less than 3 ,um in length. Both groups gave approximately the same tumor incidence, although the latency period was greater in the latter group. The data subsequently were reported in greater detail by Pott et al. (1974~. In this, for the unmilled ch~ysotile, 78.7% of the fibers were less than 2 ,um in length and 93.9% were less than 5,um in length; while for the milled sample,*97.4% were less than 2 ,um in length and

Solid Particles in Suspension 167 99.8% were less than 5 ,um in length. After administration of the unmilled chrysotile sample, the elapsed time to appearance of first tumor was 270 days with a tumor incidence of 37.5%, while for the milled chrysotile sample the time to appearance of first tumor was 400 days, with a tumor incidence of 30%. This study would indicate that both short and long fibers were carcinogenic, though the milled sample seems to have been somewhat less active than the unmilled one. But in the absence of information about the distribution of fiber diameters and the number per unit weight it is impossible to assign activity to a particular size class with certainty. Another possible exception is contained in a paper by Wagner et al. (1976) describing experiments with glass fibers. In these experiments two different preparations of glass fibers were injected intrapleurally into experimental animals. The characteristics of the fiber preparations and the results are summarized below. Characteristics of Glass Fibers Finer Coarser Diameter <O.5,um 99~O Diameter < I,um 17% Median diameter (,um) 0.12 1.8 Length >20,um Do Length >50,um logo Median length (,um) 1.7 22 Results No. with mesothelioma 4/32 0/32 Absence of hyperplasia in mesothelial cells 1/32 12/32 The same weight (20 ma) of each material was used per animal, and there were about 1,000 times more fibers in the finer glass than in the coarser glass (Wagner et al., 1976~. The results indicate, therefore, that the finer sample contained more "active" fibers than did the coarser. On the basis of the animal experiments noted above, it is possible to infer that, in these experimental systems, short fibers may be biologically unreactive, or less reactive, than longer ones. Since the majority of the fibers that are found in water are short (<S ~m), these may be similarly unreactive, and this may account for failure to detect any increased incidence of various gastrointestinal cancers related to ingestion of mineral fibers in water.

168 DRINKING WATER AND HEALTH Another hypothesis can be offered for the inability to detect, in the general population, any increase in gastrointestinal cancers attributable to ingestion of asbestos mineral fibers in water. Assuming that the heavily exposed asbestos workers (who showed an excess gastrointestinal cancer rate) swallowed a large fraction of the fibers, of all sizes, that they inhaled, the dose so acquired would have been several orders of magnitude greater than that ingested in drinking water by members of the general population. It would therefore be expected that any excess gastrointesti- nal cancer due to fibers in water would be much smaller in the general population than that found in asbestos workers, and if sufficiently low might be very difficult to detect against the usual background incidence of cancers of the colon, sigmoid, and rectum, and less so for cancers of the esophagus and stomach. The results of experiments on the chronic toxicity of asbestos mineral fibers fall very far short of rigorously clarifying the risk to health that may be associated with potable water either qualitatively or quantitatively. Although no harmful ejects of fibrous contamination of drinking water on the present state of public health have been demonstrated, it is possible, as noted in the discussion of the epidemiology, that they may be observed in the future. The development of gastrointestinal cancers among the heavily exposed asbestos workers has been slow (20 to 30 yr or more), and the possibility of long-delayed ejects of mineral fiber ingestion through water cannot be ignored. Research on the unresolved problems of the chronic ejects of ingesting asbestos mineral fibers therefore merits strong support. ORGANIC PARTICULATES IN WATER The organic particulate matter found in natural water systems may be classified as follows: 1. Organic matter associated with soil particles. 2. Organic particles from effluent of sewage treatment facilities and industrial waste disposal facilities. 3. Plant and animal debris. 4. Microorganisms. 5. Organic colloids. Microorganisms are discussed in Chapter III; the other categories are discussed below.

Solid Particles in Suspension 169 Organic Matter Associated with Soil Particles The major part of the suspended material in most natural bodies of water is made up of soil particles derived from the land surface by erosion. The coarser sand and silt fractions consist mainly of inorganic rock and mineral fragments, many of which are, at least partially, coated with organic material. The clay-size fraction of most soils consists of clay minerals, metal oxides, hydrous metal oxides, and soil organic material (humus) in intimate contact with one another. As Gorbunov, Yerokhina, and Schchurina (1971) and Greenland (1971) have shown, the clay minerals in the A horizons of most soils are coated with organic matter, forming a so-called "clay-organic complex." Greenland also found that hydrous oxides provide even more sites for adsorption of organic matter in soils than do micaceous clay minerals. Hydrous oxides also coat surfaces of clay mineral grains. It is therefore to a certain extent artificial to attempt to separate the ejects of the organic and inorganic components of the soil particulate matter suspended in water. However, it is likely that in many instances the eject of the organic components is predominant. In those instances in which the soil organic material (humus) forms coatings on the clay and oxide surfaces of soil particles, reactions between the coated particles and the environment external to the particles will to a large extent be governed by the physical and chemical properties of the humus. For example, the ion-exchange capacity and the adsorption properties of the particles will be strongly influenced by the ion-exchange capacity and the adsorption capacity of the humus coatings. Both soluble and insoluble organic compounds are sorbed by humus. Ladd and Butler (1971) have shown that the humic and fulvic acid components of soil humus inhibit proteolytic enzymes by binding them. In addition to enzymes, humus strongly binds some pesticides and other organic compounds that are found in soils and natural waters. This sorption reduces the phytotoxicity of herbicides and often inhibits the degradation of many pesticides (Wershaw and Goldberg, 1972~. In order to understand the interactions of soil particles in natural water systems, it is necessary to consider briefly what is known about the chemical structure and physical-chemical properties of soil humus. Humus has been traditionally fractionated into three compounds: humic acid, fulvic acid, and humin. Humic acid is the component that is soluble in strong base but insoluble in strong acid, fulvic acid is soluble in both acid and base; and humin is insoluble in both acid and base. Humic acids contain phenolic, quinoid, polycarboxylic aromatic

170 DRINKING WATER AND HEALTH acidic, and aliphatic groups, the aromatic groups predominating (Kono- nova, 1966; Felbeck, 1965, 1971; Schemer and Ulrich, 1960; Steelink, 1963; Flaig, 1970~. In addition to benzene carboxylic acid and phenolic groups, some authors have included fatty acid, protein, and polysaccha- ride groups. It has been assumed that these groups are joined together covalently by linkages such as ether, ester and carbon-carbon to form three-dimensional polymers of high molecular weight (Flaig, 1970~. Several recent studies indicate that this structural interpretation is not correct. Wershaw and Pinckney (1971, 1973a, 1973b) have shown that the most abundant humic acid fractions form aggregates in solution that are made up of relatively small molecules (molecular weights of several thousand or less). The degree of association of these molecules is a function of pH. Monahan, DeLuca, and Wershaw (1972) have found that the degree of aggregation of some of the fractions is also a function of concentration and that at low concentrations the molecules are dissociat- ed. The presence of these aggregates will have a marked eject on the way humic acids interact with pollutants in natural water systems, and this will be discussed later. Fulvic acids are generally believed to be similar in structure to humic acids but of lower molecular weight. It is apparent, however, that other groups of compounds, such as polysaccharides, proteins, and amino- sugars are also included in many fulvic acid preparations (J. A. Leenheer, oral communications, 1974~. Alternating treatment of the humin fraction with strong mineral acids and strong bases generally renders most of the humin soluble in basic solutions (Kononova, 1966~. From this Kononova has concluded that humins are humic acids that are bound by the mineral constituents of soils. Humic substances have a large capacity for base exchange. Kononova (1966) has reviewed the exchange capacity of several different soils and has reported values from about 200 to over 500 milk-equivalents per 100 g for humic acids. The ion exchange capacities of humic acids depend on pH and are smaller at low pH values than at high pH values. There has been a great deal of study lately on trace metals in natural water systems because of the eject that variations in the concentration of those metals can have on plant and animal health. Harding and Brown (1975) have studied the distribution of some selected trace metals in sediments of the Pamlico River estuary in North Carolina. They found the highest concentrations of trace elements in the mixed clay organic matter sediments in the center of the estuary. These high trace metals concentrations are apparently due to human activity. Concentrations 4 to 1,750 times normal background were detected. Their evidence suggests

Solid Particles in Suspension 171 that the trace metals are scavenged from water entering the estuary by the clay and organic matter in the sediments. The suspended clay and organic matter in the water would also be expected to have high trace metal concentrations. Shimp et al. (1971) have found that the concentra- tions of bromine, chromium, copper, lead, and zinc increase with increasing organic carbon concentrations in recent sediments from southern Lake Michigan. They pointed out that: "High concentrations of these elements correlate more closely with the amounts of organic carbon present than with clay-size material, water depth, iron oxide or manga- nese oxide." A significant portion of the cation exchange capacity of most soils and marine sediments is due to organic polyelectrolytes; in soils and sediment of high organic content, most of the exchange capacity can be attributed to organic polyelectrolytes (Schnitzer, 1965; Rashid, 1969~. Schnitzer and Skinner (1965) found that both carboxyl and phenolic hydroxyl ligands react with' polyvalent cations. Infrared studies by Rashid (1971) on marine humic acids confirmed the participation of carboxyl groups in exchange reactions with cations, but no evidence for the participation of phenolic hydroxyl groups was found. Koshy and Ganguly (1969), Malcolm (1969), and Ong et al. (1970) have studied the formation of soluble humic acid-metal complexes. They found that the amount of metal ions that is complexed by humic acid increases with increasing pH. Although these authors did not over an explanation for this behavior, it is reasonable to suppose that, with increasing pH, more disaggregation of the humic acid takes place, and more ligand sites are exposed for complexing with the metal ions. The metal ions may be bound both by ligands on a single molecule or ligands on more than one molecule. The metal ions may therefore act as bridges between molecules, causing aggregation (Sipos et al., 1972~. In addition to interactions with metals, humic acids and other natural organic polyelectrolytes can interact strongly with other organic compo- nents in natural water systems. These interactions can be of two types: incorporation of organic molecules into humic acids during their formation, and reaction of organic compounds with humic acids after their formation. Humic acids are generally considered to arise from the microbial degradation of organic debris. Martin and his co-workers (1972) have shown that some soil fungi can convert a wide variety of different organic substrates into phenolic and quinoid compounds and aromatic acids, which are then converted into humic acids. Some studies have also revealed that some humic acid fractions have hiah free radical content. Steelink and Tollin (1967) and others have

172 DRINKING WATER AND HEALTH found free radicals both in soils and in humic acids isolated from soils. The evidence suggests that the radicals detected are quinoid radicals. Martin has not postulated a mechanism for the conversion of the phenols, quinones, and aromatic acids into humic acids, but it is likely that free radical polymerization reactions play a major role in this conversion. The free radicals that polymerize into humic acid molecules probably arise either from enzymatic reactions or photolysis. Compounds that are either oxidized or reduced to free radicals may be derived either from plant and animal remains, or from sewage, pesticides, or other pollutants. The conditions for these oxidation-reduction reactions are undoubtedly present in many soil-water systems. Su and Zabik (1972) have studied the photolysis of the miticide m-(n,N-dimethylformamidine) phenyl N-me- thylcarbamate in water. They have shown that decomposition proceeds through the formation of free radicals. Similarly, Mazzocchi and Rao (1972) have found that radical intermediates are formed in the photolysis of the herbicides monuron and fenuron. Bordeleau et al. (1972) have studied the degradation of phenylamide herbicides in soils. They found that azobenzenes and polyaromatic compounds are formed by a combination of enzymatic and chemical reactions, from the phenylamide herbicides. Chloroaniline groups are formed by acyl-amidase degradation of phenylamide herbicides. These chloroaniline groups are then transformed to stable chloroazobenzene residues by peroxidases. During the enzymatic degradation by peroxi- dase, free radical intermediates were detected by electron paramagnetic resonance. The free radicals produced in the above reactions could be incorporated into humic acids. The work of Khan and Schnitzer (1972) strongly suggests that different organic molecules are indeed incorporat- ed in humic acid. They have succeeded in releasing a variety of hydrophobic organic compounds from a humic acid after exhaustive methylation with diazomethane. They were able to identify alkanes, fatty acids, dialkyl phthalates, and butyl adipate in the mixture of released compounds. These compounds comprised about 1% of the total humic acid weight. Previous extraction with organic solvents had released only a very small amount of the hydrophobic compounds. Khan and Schnitzer have proposed that humic acid has a molecular sieve type of structure that traps insoluble organic compounds. This mechanism implies that humic acid molecules form a three-dimensional network with suitable voids in which the hydrophobic molecules are adsorbed. Khan and Schnitzer apparently assume that methylation disrupts this structure and releases the entrapped compounds. The humic acids that were used by Khan and Schnitzer were extracted from soil with sodium hydroxide solution. During this type of extraction, disaggregation of the humic acid

Solid Particles in Suspension 173 molecules will take place and most of the molecules trapped in the voids should be released. The compounds that they identified were apparently not released and therefore another mechanism must have been responsi- ble for binding these molecules. The fact that methylation was required for release of the molecules strongly suggests that they were chemically bonded to functional groups in the humic acid polymers and could well have been incorporated into the humic acid polymer when it was formed. The humic acid polymer, after formation, may participate in reactions which may be conveniently divided into the categories tabulated below. When considering these types of interactions it must be kept in mind that humic acids form molecular aggregates; changes in the degree of aggregation will probably modify these interactions. Mechanisms of Interaction Physical Interactions Adsorption a. Hydrogen bonding b. Van der Waals bonding Chemical Interactions Chemisorption Coordination reactions Radical and other chemical reactions Most organic pesticides are adsorbed by soil organic matter (Wershaw and Goldberg, 1972~. A few studies of the mechanism of adsorption of particular pesticides by soil polyelectrolytes have been conducted. Sullivan and Felbeck (1968) and Li and Felbeck (1972) have proposed that the sorption of triazine herbicides is due either to hydrogen bonding or to ionic bonding or to both of these taking place simultaneously. Hsu and Bartha (1976) found that up to 90Yo of the 3,4-dichloraniline (DCA) derived from the biodegradation of phenylamide herbicides is adsorbed so strongly by soil organic matter that it is not extractable by solvents. The bound DCA is susceptible to acid and alkaline hydrolysis. The authors have concluded that this suggests that the nitrogen atom of the DCA is covalently bound to the carbon of a carbonyl group or to a quinoid ring of a humic acid molecule. Much more work is necessary, however, to elucidate the general principles of adsorption of pesticides by soil organic polyelectrolytes. The evidence cited above for the strong adsorption of many pesticides by soil organics suggests that one would expect these pesticides to be

174 DRINKING WATER AND HEALTH associated with the sediments in rivers and lakes, and indeed this has been found to be the case. Manigold and Loral (personal communication, 1976) have found in a study of the distribution of pesticides in rivers that the chlorinated hydrocarbon insecticides such as DDT are concentrated in the suspended sediments and that very little of these materials is found in solution in the water. A similar situation would be expected with the polychlorinated biphenyls and other hydrophobic pollutants in natural waters. An example of chemisorption by humic acids appears to be found in the work of Perry and Adam (1971~. They studied the incorporation of peptides into humic acid by allowing glycylglycine to react with humic acid at pH 8.5. They found that part of the amino nitrogen introduced into humic acid by this reaction could not be removed by hydrolysis with 6N HCL at 110°C for 24 hr. At least part of this unreleased amino nitrogen is probably bound to the humic acid by N-phenyl linkages. Although the mechanism of formation of these bonds has not been elucidated, this example illustrates the reactivity of humic acids and the strengths of the bonds formed. It also suggests that chemisorption accounts for much of the binding of amino acids, peptides, and proteins that has been observed in soils. Anderson (1958) has shown that the hydrolysis products of deoxyribonucleic acid (DNA) are released from humic acid by perchloric acid hydrolysis; he concluded from this that DNA is present in humic acid. The binding of enzymes to humic acids (Ladd and Butler, 1971; Mato et al., 1971) is probably also due in part to the formation of bonds between amino nitrogen groups and reactive sites on the humic acid molecules. Undoubtedly, other types of reactions can also take place, some perhaps involving shared metal cations. A careful study of the interactions of humic acids with enzymes should give a new insight into the biochemical reactions that take place in soil-water systems. Municipal and Industrial Wastes Municipal and industrial waste disposal is a major source of organic particulate matter in natural waters that serve as sources for drinking water. The amount of particulate matter added by any particular waste disposal site is highly dependent on the amount of treatment that is given to the wastes (Litchfield, 1975; Soderquist, 1975; Gove and Gillman, 1975; Jewell et al., 1975; Talbot, 1975; Pico,1975; Macauley, 1975~. Large quantities of organic and inorganic particulate wastes are also added to natural water from urban runoff and from combined sewer overflows (Field and Knowles, 1975~. These sources supply large quantities of

Solid Particles in Suspension 175 pollutants to natural waters during periods of heavy rainfall, so that the amount of pollution is highly variable. This makes it particularly difficult to predict the quality of the feed water into a drinking water plant that is being supplied with water from a source that is subject to intermittent pollution caused by storm runoff. In combined sanitary storm sewage systems untreated sanitary sewage may be discharged into a receiving stream during periods of high storm sewage flow. Organic matter supplied to natural waters by sanitary sewage-treat- ment plants and treatment facilities for food-processing plants undergoes degradation relatively rapidly. However, some industrial products are much more persistent. Giger et al. (1974), in their study of the sediments in Lake Zug, Switzerland, found high concentrations of petroleum- derived hydrocarbons in the sediments near densely populated areas. They concluded that biodegradation of these hydrocarbons is retarded in the sediments. Shelton and Hunter (1974) have studied the occurrence of oil pollutants in sediments. They found that the heavier petroleum fractions are concentrated in sediments. In the event of an oil spill, petroleum is adsorbed by the sediment in the area of the spill and slowly released. As time passes, the petroleum components remaining in the sediment will become heavier and heavier. The fine-grained particulate matter found in rivers and lakes that drain heavily populated areas are generally high in organic carbon and in trace elements. Much of this material is derived from domestic and industrial waste disposal. For example, high concentrations of metals are often found in sewage sludge and in the organic particulates released by sewage treatment plants (Bruland et al., 1974; Leland, Copenhaver, and Wilkes, 19751. An example of high trace metal concentrations in suspended sediments of a heavily industralized drainage basin is found in southern Lake Michigan. Leland et al. (1973) have studied the distribution of trace metals in the bottom and suspended sediments in southern Lake Michigan. Their data show markedly higher concentrations of arsenic, bromine, chromium, copper, mercury, lead and zinc" . . . near the sediment water interface of fine-grained sediments than in underlying sediments of Southern Lake Michigan." They found that with the exception of bromine the trace element concentrations in the suspended sediments were equal to or higher than the concentrations in fine-grained surficial sediments. They concluded from this evidence that the high trace metal concentrations in the sediments are due to high concentrations in the suspended sediments that settle out in the lake. The highest concentrations of trace elements are in the central basin of the lake,

176 DRINKING WATER AND HEALTH where the finest sediments are found. Thus the smallest suspended particles, which do not settle out until they have reached the center of the lake, have the highest concentrations of trace metals. These sediments also have the highest organo-chlorine insecticide concentrations. Leland, Shukla, and Shimp (1973) found a positive correlation between the organic carbon concentration, iron oxide concentration, and trace metal concentrations in the sediments. Since many of the communities around Lake Michigan take their drinking water from the lake, the high trace metal and pesticide concentrations in the suspended sediments could pose a problem if the coagulation and filtration processes used in water treatment are not adequate to remove them from the water. When organic particulate matter containing high trace metal concen- tration is introduced into a natural body of water a new set of equilibrium equations will in general be required to represent the system. If the chemical composition of the water is different from that of the solutions in which the particulate matter acquired the trace metals, redistribution of metal ions between the water and the sediment phases will take place. This sediment will serve as a source of trace metals in water even after the source of pollution is removed. Release of metals fixed to the suspended and bottom sediments may also take place by decomposition of the organic matter binding the metals in the sediment. DeGroot, DeGoeij, and Zegers (1971) have reported this phenomenon in the Rhine and Ems rivers. They found that most of the mercury and other heavy metals transported to the sea by the rivers are fixed to suspended particles of less than 16 ,um in diameter. The metal ions remain fixed to the suspended particles in both of the rivers until the chemical composition of the river water is changed by mixing with seawater. Downsteam from the freshwater tidal zone of each of the rivers mercury, zinc, lead, chromium, arsenic, cobalt, and iron are lost from the sediments. The authors have attributed this to decomposition of the organic matter of the suspended sediment. Experiments performed in their laboratories suggest that the metals are released as organometallic complexes. Muller and Forstner (1974) have questioned the conclusions of DeGroot, DeGoeij, and Zegers and claim that most of the reduction in heavy metal concentration in the sediments of the Rhine estuary is due to dilution of the sediments derived from the Rhine with cleaner North Sea sediments. However they do not totally discount the solubilization mechanism. At this time, however, it is not at all clear which conclusion is correct.

Solid Particles in Suspension 177 Organic Debris Very little work has been done on debris from living organisms in streams. Lammers (1967) has used ultracentrifugation to isolate the suspended organic components of streams. In his later work (rammers, 1975) he isolated the organic debris from the gut of living filter-feeding organisms. Lammers chose several different bivalves for this work. The use of the bivalves allows one to obtain integrated samples of the debris in the stream. The material isolated from the bivalves was then fractionated by ultracentrifugation. In both of the above papers Lammers deals mainly with the viruses, algae, and bacteria that he isolated; however, he does indicate that other organic particles, such as ribosomes, were also isolated. A number of studies have shown that organisms in natural bodies of water concentrate hydrophobic pollutants, such as chlorinated hydrocar- bon insecticides, in their tissue (Sodergran et al., 1972~. It would therefore be expected that some of the organic debris in natural waters would contain elevated levels of hydrophobic pollutants; however, there do not appear to be any studies in the literature that deal with this matter. Organic Colloids Organic polyelectrolytes that are very similar to the soil humic and fulvic acids have been isolated from surface and groundwaters. As Black and Christman (1963) point out, these materials will impart a brownish color to water if present in high enough concentration, and indeed many potable water supplies are colored. Day and Felbeck (1974) have shown that the fungus Aureobasidium pullulans, which is common in sewage, soils, and surfaces, exudes a substance similar to fulvic acid. Martin and his co-workers in a series of papers (see Haider et al., 1972, and Martin et al., 1972, for reviews of this work) have shown that soil microorganisms can form humic substances in two ways: by extracellular transformation of plant and animal constituents into humic compounds and by synthesis of humic precursors from aliphatic compounds. The most common precursors that Martin and his co-workers have detected are phenols that are polymerized by autoxidation or enzymatic reactions, or more likely by both pathways, into humic materials. Similar transformations in sewage and in surface waters would be expected. Thus we may expect that the degradation of many organic wastes will result in the formation of humic materials. During the polymerization process, it is possible that resistant pollutant molecules that have not been degraded by the

178 DRINKING WATER AND HEALTH microorganisms will be incorporated into the humic and fulvic acid polymers. The first part of this discussion has been principally concerned with the relatively insoluble humic acid associated with soils. However, the humic acid salts of the more common monovalent cations are soluble in water. Wershaw and Pinckney (1973b) have demonstrated that these humic acid molecules form molecular aggregates in solution, the degree of aggrega- tion depending on both pH and concentration. This aggregation phenomenon was detected both in humic acid fractions and in unfrac- tionated salts. Fractionation apparently increased the chemical homoge- neity of the fractions by comparison with the unfractionated salts. The aggregates that have been detected in the fractions therefore probably comprise molecules that are chemically more or less similar. However, in an unfractionated sample, the aggregates will contain a diversity of different types of molecules. Even molecules that are quite different from humic acids can be incorporated into the aggregates; Wershaw et al. (1969) have shown that the sodium salt of humic acid can solubilize DDT by incorporating DDT molecules into the sodium humate aggregates (micelles). Undoubtedly, this mechanism is responsible for the transport of other relatively insoluble pollutants, besides DDT, in natural water systems. Ballard (1971) reported an apparent example of this mechanism at work in a natural system. He found that the urea salt or the ammonium salt of humic acid solubilized DDT and carried it down through a soil profile. In addition to solubilization reactions, soluble humic acid salts and fulvic acids interact with metals to form colloidal organometallic complexes. Much of the early work on colloidal humic material-metal complexes was on iron complexes. Several workers have found that the oxidation of ferrous iron is greatly retarded in many natural waters that contain humic substances (Theis and Singer, 1974~. Whatever the reaction mechanism, it has been well established that humic material stabilizes both ferrous and ferric iron in natural water systems. This has important implications for water treatment, since the standard method of removing excess iron from water is to oxidize it with oxygen and allow it to precipitate as Fe(OH)3. The presence of humic material inhibits these reactions. A number of more general studies (Rashid and Leonard, 1973; Orlov and Yeroshicheva, 1967; Ong et al., 1970) have shown that humic materials form complexes with copper, cobalt, manganese, nickel, zinc, iron, and aluminum, and solubilize these metals in natural water systems. Cheam has reviewed the literature on the complexation of copper (II) by fulvic acid. The available evidence indicates that copper is chelated at

Solid Particles in Suspension 179 a bidentate site consisting of a phenolic hydroxl group and an ionized carboxyl group. Potable water supplies containing humic materials are often decolo- nized by chlorination. Rook (1974) has shown that chlorination of humic acid solutions in water results in the formation of haloforms, some of which are known or suspected to be carcinogenic (Nicholson, 1977~. A survey for haloforms in the water of 80 cities, conducted by the EPA, showed that occurrence of trihalomethanes is widespread in finished drinking waters and is a direct result of chlorination (Stevens et al., 1975; Symons, 1976~. The risks that ingestion of these compounds may pose to human health are discussed in Chapter VI. Summary Natural organic polyelectrolytes Pubic materials) play a major role in the binding and transport of pollutants in natural water systems. These reactions include: (1) adsorption of pesticides, enzymes, and other organic compounds; (2) free radical reactions including charge-transfer reactions between free radicals; (3) ion-exchange and complexation reactions between metal ions and ligands or trace organic poly- electrolytes; (4) solubilization reactions in which relatively insoluble compounds are rendered more soluble. These reactions will lead to the binding of toxic metals and organic compounds to suspended colloidal particles in raw water supplies. MICROORGANISMS AND SUSPENDED PARTICLES IN WATER The tendency of microorganisms to form aggregates and to become concentrated at the surfaces of solid particles, rather than to be uniformly and individually dispersed, may have important consequences for their survival and for their reactions to the various processes of water treatment. It is doubtful that many of these microbial agglomerates will pass through an efficiently operating water-treatment process, but a large segment of the population ingests surface water that has had only partial treatment (i.e., disinfection). Under such circumstances these microbial agglomerates constitute potential health hazards. Unfortunately, there is a lack of scientific information on either the survival of the bacterial component of these microbial-agglomerates after disinfection of surface waters or on their ability to disaggregate and produce infection in the human host after ingestion. Viral aggregates, on the other hand, have a higher resistance to disinfection than free virions

180 DRINKING WATER AND HEALTH and produce infection in susceptible hosts. (See Chapter III, "Microbiol- ogy," section on viruses.) Although there is a large body of scientific information from laboratory and field studies on raw water sources with regard to bacterial and viral agglomerates, the same level of scientific attention has not been given to studies of the viability of these agglomerates in finished drinking water or, more importantly, on infection of the human host after ingestion of these agglomerates. Particulate-Bacterial Interactions A knowledge of aquatic and terrestrial habitats of microorganisms is essential to the understanding and resolution of problems that arise when man must use the same environment. Potable water, which may be regarded as an environment, has been developed by man out of natural aquatic microbial habitats that are continuously fed, in a microbial sense, from the terrestrial habitat. Pollution of these habitats by higher organisms adds to the natural microbial population already present. This pollution component of the microbial population supplies the more hazardous aspects to the potable water environment because these are the microbial members that have passed through man and other organisms and have, in some instances, been responsible for debilitation or death. Studies of microbial aggregates in terrestrial environments that ultimately seed the raw water sources have been carried out by Brock (1966, 1974), Cameron (1965-1969), Rice et al. (1975), Henrici (1948), Schmidt (1968), Mallette (1963), and Kononova (1966~. These studies demonstrated that the most extensive microbial growth takes place in nature on the surfaces of particles and inside loose floes of solid particles. This occurs because the nutrients required for microbial growth are also adsorbed at the surfaces of these particles. Only a few microorganisms are found free in the soil solution or in raw water because of the lack of dissolved nutrients. Thus, chemical and microbiological analyses of water per se will not reveal the higher concentrations of organic nutrients and microorganisms present on the surfaces and inside sand, silt, or clay particles that have settled to lower water strata or to the bottom of the raw water sources. Some of the aggregates will be sequestered by sedimentation, but microbial particulates in lake and reservoir sediments remain viable and grow rapidly at the time of the spring inversion, when mixing with upper layers will cause some redistribution (Henrici, 1948; Boylen and Brock, 1973; Hendricks, 1973~. Mutual microbial interactions also occur, not only in association with inorganic and organic particulates, but also in their absence to form

Solid Particles in Suspension 181 mixed microbial aggregates. Such mutualistic relationships may benefit both organisms in such ecological associations or show a graded series of symbioses of variable stability and specificity. Studies of these particulate interrelationships have been reviewed by Lederberg 0952) and investi- gated by Sharp and Church (1963) and Rice et al. (1975~. These mutualistic relationships allow microbial growth to take place in environments unsuitable for rapid growth of either member alone; this concept has implications for the survival and rapid growth of micro- orgnisms in nutrient-starved aquatic systems, e.g., potable water supplies. Investigations of the physical-chemical attachment of microorganisms to sand, silt, and clays, the difficulties encountered in disaggregating these particulate complexes, and the conditions promoting migration of the aggregates into potable water supplies (deep wells in particular) have been reported by Stotzky (1965, 1966), Lammers (1967), Boyd et al. (1969), and Gray and WiLkinson (1965~. Of the physical-chemical characteristics studied, Stotzky, in a series of publications, showed that cation exchange capacity (CEC) had an important effect on the binding of organisms to particles. Lammers isolated and fractionated various organic and inorganic particle species occurring in natural waters. Studies of this kind could help to elucidate the composition and concentration of microbial particulates in water. Particulate-Viral Interactions in Water River silt adsorbs viruses with moderate efficiency and does not relinquish them very easily. Berg (1973) showed experimentally that silt adsorbed up to 94% of poliovirus 1 from distilled water, but only 0.3 to 0.6% of the adsorbed virus could be eluted from the silt. He also showed that viruses were recovered more frequently from the silt filtered from river water than from the river water itself. Since viruses were not recovered so efficiently from silt as from water, there would appear to be much more virus adsorbed to the silt than suspended in the water. These studies on viral adsorption to sand, silt, clays, and organics (feces) to form particulates are consistent with what is known for bacterial aggregates. To infect by ingestion of water, viruses must pass from their source (human and animal excrete) through two formidable barriers: sewage treatment and water treatment. In both processes, formation of large particulate-viral floes are required for settling and filtration. Sewage or water treatment may inactivate the virus, but viruses may also survive in the settled sludge or on filters.

182 DRINKING WATER AND HEALTH Aggregation and Survival Several studies have demonstrated that the presence of particulate matter in water interferes with disinfection. Neefe et al. (1967), Walton (1961), Hudson (1962), Sanderson and Kelly (1962), Tracy et al. (1966), and Symons and Hop (1975) have all shown this eject. The association between microorganism and particle appears to produce a resistant complex that is not easily dissociated. Some aspects of the resistance to disinfection of viruses that have become associated with organic particulate material are illustrated by experiments reported by Symons and How (1975) on the inactivation of aqueous suspensions of poliovirus 1. Purified suspensions of the free virus, washed suspensions of virus on the debris of cells in which it had been grown, and mixtures of suspensions of virus and uninfected cells were treated with HOC1 at 3 mg/liter. The results indicated that preparations of free virus, either alone or mixed immediately with cell suspensions, were inactivated rapidly. By contrast, virus associated with cell debris, and virus mixed with a cell suspension and held overnight, were inactivated much more slowly. The authors note that the nature of the association between virus particles and cell debris, and the extent to which these preparations represent the conditions in which viruses occur in nature and in treated water, are unknown. Melnick (1975) discusses the viral inactivation problems associated with the use of heat, formalin, chlorine, and other agents, in cases where tissue-culture techniques demonstrate viral inactivation (no PFU). If however, the virus is inoculated into an animal host, it may produce infection although no PFU's were demonstrated in tissue culture. (See also Chapter III, "Microbiology," section on viruses.) Viral particulates and aggregates -form protective barriers to some of the internal virions of the aggregate. Thus the nucleic acid component of the virus is doubly protected against disinfecting (inactivating) agents by the viral protein coat and by the outer mass of virus particles in the aggregate. Petrilli et al. (1974) used an experimental treatment plant to show that enteroviruses survived even though coliforms, E. colt, and fecal strepto- cocci were eliminated after treatment of raw water by coagulation with ferric chloride, sedimentation, filtration through activated carbon or sand, and terminal chlorination. He also demonstrated survival of enterovirus in spring water that had been chlorinated sufficiently to destroy the coliforms and fecal streptococci that were also initially present. A review of viral problems facing water treatment plant operators is presented by Taylor (1974~.

Solid Particles in Suspension 183 Conclusions Investigations are required of the physical-chemical attachment of microorganisms to sand, silt, clays, and organic particles, and disaggrega- tion of these particulate complexes. Viral aggregates are more resistant to disinfection than are separate viral particles. Fundamental information is needed on the interactions between the viable and nonviable components of particulates in drinking water and particularly on their resistance to disinfecion and to other water-treatment processes. PARTICULATE REMOVAL AND TURBIDITY Definitions and Occurrence Some terms commonly used in water treatment practice are described here. Turbidity in water is caused by the presence of suspended matter such as clay, silt, nonliving organic particulates, plankton, and other microscopic organisms. Operationally, turbidity measurements are ex- pressions of certain light scattering and absorbing properties of a water sample. Color in water is due primarily to the presence of natural organic matter; it may also be caused by certain industrial wastes and by some metallic complexes. Color is measured by determining light adsorption. Hence, colloidal particles can produce some color as it is operationally determined. Filtration has been defined as the passage of a fluid through a porous medium to remove matter held in suspension. Particles ranging in size from a small fraction of a micrometer to several hundred micrometers can be removed by conventional packed-bed filters used in water treatment if appropriate chemical pretreatment is provided. Coagulation is a process in which colloidal particles are destabilized by the addition of suitable chemicals and then assembled into larger aggregates primarily by gentle fluid motion. Sedimentation is a process by which solids are separated from water by gravity. In conventional treatment for the removal of particulates, the solids in water are first coagulated into compact, fast-settling flees. Most of these are then removed by sedimen- tation, after which the water is filtered to remove additional particulates. Softening involves the removal of calcium and magnesium ions from water. Usually these are removed as solids, viz., calcium carbonate and magnesium hydroxide. Hence, water-softening plants have sedimentation and filtration facilities for the removal of these precipitates. Other solids in the raw water (e.g., clay) and other precipitates that may be formed in

184 DRINKING WATER AND HEALTH the softening process can then be removed by these solid-liquid separation processes. A survey by the USPHS of municipal water facilities in communities with populations of 25,000 or more was summarized by Jenkins (1963~. The population covered by this survey totaled 100,940,020 people. A minimum of 62% of this population received filtered water and hence was delivered water treated for the removal of Articulates. This information has been revised and expanded to include more communities in a facilities survey by the EPA in the early 1970's. The results are presently being analyzed and summarized by EPA. Removal of Particulates MINERAL FIBERS Pilot-scale filtration experiments for the removal of amphibole and chrysotile fibers from the Duluth water supply have been conducted (Black and Veatch, Consulting Engineers, 1975; Logsdon and Symons, 1975~. Amphibole fibers were easily removed by coagulation and filtration. Considerable difficulties were encountered in removing the chrysotile fibers. These were small and may have been positively charged. Since conventional filters are able to remove submicroscopic particles (Yao et al., 1971), and since conventional pretreatment chemistry (coagulation) is designed to destabilize negative particles, it is plausible that the surface characteristics of the chrysotile particles may hinder their removal by conventional filtration practice. Finally, it is important to note that the design, operation, and evaluation of water-treatment facilities for the removal of fibrous particles is seriously impaired by the time, effort, and expense required to detect these particles in water supplies. CLAY Direct evidence of the removal of clays from water supplies by coagulation and filtration is lacking, since clay particles are not detected directly in routine water analysis. It can be implied that clays are electively removed since they comprise a significant portion of natural "turbidity," and conventional treatment plants are very elective in turbidity removal. This conclusion is substantiated by laboratory studies of the coagulation and filtration of clay suspensions. Studies of this type are numerous. For example, Packham (1965) and Black and Hannah

Solid Particles in Suspension 185 (1961) have demonstrated elective coagulation and sedimentation of kaolinite, montmorillonite, and other clay suspensions using aluminum sulfate for destabilization under conditions similar to those encountered in potable water treatment. Among many others, Ling (1955) and Adin and Rebhun (1974) demonstrated elective removal of clays from suspension by filtration after appropriate chemical pretreatment. ORGANICS Conventional processes for the removal of particulates are elective in removing color and many other organic macromolecules and particulates from water. Here again, laboratory studies provide the bulk of the available evidence. Hall and Packham (1965) have demonstrated that humic and fuivic acids can be coagulated by iron(III) and aluminum(III) salts. Humic acids were readily removed, but a significant fraction of the fulvic acid component was not removed by either coagulant. Similar results have been reported by others, including Black and Willems (1961) and Rook (1975~. MICROBIOLOGICAL PARTICULATES Large microorganisms including algae and amoebic cysts are readily removed by filtration from properly pretreated water. Bacterial removals in excess of 99% are also achievable (ASCE et al., 1969~. More than 98% of poliovirus type 1 was removed by conventional coagulation and filtration (Robeck et al., 1962~. Complete (i.e., loopy) removal of microorganisms is not feasible, so that filtration is followed by disin- fection with chlorine in conventional water treatment practice. Possible Concerns HALOMETHANES AND OTHER CHLORINATED ORGANICS The natural humic substances are known to be precursors for the formation of chloroform and other halomethanes in water treatment. These may be colloidal, or adsorbed on other colloids such as clays. A recent study (Stevens et al., 1975) has indicated that the rate and extent of chloroform production is reduced when chlorine is added to water containing humic substances after filtration, rather than before coagula- tion. Presumably, some organic materials (particulates, or solutes adsorbed on particulates) are removed by coagulation and filtration prior

186 DRINKING WATER AND HEALTH to chlorination. Hence, water treatment for particulate removal can lead indirectly to reduced chloroform production. ENCASING OF MICROBIAL PARTICULATES It has been proposed that pathogenic bacteria and viruses can be encased in gelatinous metal hydroxides during conventional coagulation process- es. If these particles then pass through subsequent settling and filtration facilities, it has been further proposed that the hydroxide gel coatings can protect the pathogens from being inactivated by chlorination. It has also been proposed that natural organic substances can coat and protect pathogens from disinfection. These problems are addressed by Symons and How (1975~. Their preliminary evidence suggests that adsorption of virus particles on clays or coagulation of viruses by aluminum dulfate do not impair disinfection of poliovirus type 1 by chlorine. The presence of cell debris, however, significantly slowed the rate of disinfection. This problem requires additional research. COAGULANT PRECIPITATES The amorphous hydroxide floes (e.g., aluminum hydroxide, ferric hydroxide) formed in many coagulation processes are reactive solids. They can, for example, adsorb trace metals and organic compounds. Hence, if they pass through filters and into the treated water, they may carry other substances with them. This is plausible, but conjectural. POWDERED CARBON Powdered activated carbon is frequently used for controlling taste and odor. If some of these carbon particles pass through the filters, they exert a chlorine demand and may also carry adsorbed substances with them. Some of these adsorbed substances might have been on the carbon before it was added to the water. EPA does not at this time have standards for carbon used during water treatment. The problem is at present hypotheti- cal, i.e., no data are available. ORGANIC COAGULANTS Many synthetic organic materials have been approved for use in water treatment for removal of particulates. These have been examined and

Solid Particles in Suspension 187 certified as nontoxic. The problem of the safety of polyelectrolytes is currently under review by EPA and FDA (Symons, 1976~. Turbidity This consideration of turbidity as an indicator of potable water quality begins with an evaluation of the test itself. Two recent reviews are excellent (McCluney, 1975; Pickering, 1976~. Turbidity determinations involve measurements of the light scattering and-absorbing properties of a suspension. A variety of definitions, methods of measurement, instruments, standards, and units of measure have been used. Typically, the light scattered at an angle of 90° to an incandescent source beam is determined. The amount of light scattered depends on the number, size, shape, and refractive index of the particles, the wavelength spectrum of the incident light, and the geometry and detection characteristics of the turbidimeter. A variety of standards (American Public Health Association, 1976) and instruments are in current use. However, each type of instrument will respond differently to the diversity of sizes, shapes, and types of particles found suspended in both raw and treated waters. Furthermore, depend- ing on the method of standardization used, the same instrument can indicate different turbidities on a single water sample. Hence, it is important that the turbidity test be standardized to provide a sound base for the possible use of turbidity as an indicator of water quality and for other uses. Since many pollutants in water are either particles themselves or are adsorbed on particles, and since particles absorb and scatter light, it is advantageous to consider turbidity as an indicator of water quality. In addition, there are indications that normally innocuous particles.can, at least under some circumstances, protect pathogens from a disinfecting agent (Symons and How, 1975~. Finally, it is important to note that turbidity determinations are rapid, relatively inexpensive, and can be performed continuously by detectors in situ. Measurements of turbidity do not give complete information about the size, number, mass, or type of the particles that scatter or absorb light. Small particles (those less than about 0.1 ,um in maximum dimension, single viruses and many asbestos mineral particles) do not scatter much visible light and so are not detected by conventional measurements. Furthermore, larger particles can scatter light electively but may be harmless; the presence of ice crystals in Arctic waters illustrates this point.

188 DRINKING WATER AND HEALTH Conclusions Many harmful substances in raw water supplies and in treated potable waters are either particulates themselves or adsorbed on solid particles. These particles can be removed efficiently by conventional water treatment using coagulation, settling, and filtration processes. Particles that pass through conventional coagulation and filtration plants in significant amounts do so because of inadequate design, operation, or control of these facilities. Some of these particles may be harmful in themselves. Most are innocuous, but have the potential of containing harmful substances. The absence of a detectable turbidity does not guarantee that a water is free from harmful particulates. A determination that a water has a high turbidity does indicate the presence of particulate materials and is a cause for concern. Turbidity is a collective parameter. It cannot replace specific tests for individual pollutants (e.g., Pb, asbestos); it can provide an indication that such tests should be performed. The nature of the particulates in a raw water supply should be determined, and removal accomplished when necessary, by treatment of the supply and appropriate source control. The rapidity, low cost, and "in situ" features of turbidity measurement make it valuable for monitoring and operation of water-treatment plants. When high turbidities are observed in potable water supplies, particular attention should be given to the chlorine demand of the water and to the disinfection process. Finally, the test itself must be standardized. This will involve the selection of a single light-scattering standard material or a single optical unit, and a standardized light source, instrument geometry, and detection system. SUMMARY-SOLID PARTICLES IN SUSPENSION Materials suspended in drinking water include inorganic and organic solids as finely divided particles of sizes ranging from colloidal dimen- sions to over 100 ,um. Such particles may also have other substances and microorganisms attached to them. Small particles of some materials, such as the asbestos minerals, may have the potential to affect human health directly when they are ingested, and there is widespread concern over the biological effects of such substances. Many kinds of particles, though apparently harmless in themselves,

Solid Particles in Suspension 189 may indirectly affect the quality of water by acting as vehicles for concentration, transport, and release of other pollutants. Water treatment can often be elective in removing most of the particles suspended in it, but conventional methods of detecting the presence of particulate material by measurement of turbidity have serious deficiencies. Direct Effects on Health Particles of asbestos and other fibrous minerals occur in raw water; usually in a range of sizes from fractions of a micrometer to a few micrometers. Generally, there are fewer than 10 million fibers per liter, but waters are found with from less than 10,000 to more than 100 million fibers per liter. Some of the highest counts have been found near some cities. Fibers in drinking water are typically less than 1 ,um long and fibers longer than 2 ,um are uncommon. Identification and counting of fibers is difficult and time consuming, usually requiring the transmission electron microscope. The reported counts are highly variable, often differing from one count to the next by a factor of 10 or more. Epidemiological studies of workers exposed to asbestos by inhalation have shown an increase in death rates from gastrointestinal cancer. With respiratory exposure it is likely that more fibers are swallowed than remain in the lungs. The workers studied were exposed to asbestos with a large range of fiber lengths. It is not clear whether fiber length is pertinent to the development of cancer in the digestive tract in humans. Epidemiological studies of cancer death rates in Duluth, Minn., where the water supply has been contaminated with mineral fiber, have so far not revealed any increase with time, in comparison with death rates in other areas. Contamination of the Duluth drinking water began less than 20 yr ago, however, and since many cancers have long latency periods, these negative epidemiological findings do not exclude the possibility that an increase might appear within the next 5 to 15 yr. Animal deposition model studies have shown that fiber length and diameter affect the carcinogenic response seen, the long thin fibers appearing to be the active ones. However, the relevance of these animal deposition models to the human experience is not clear. While some animal studies have shown penetration of the gastrointestinal epithelium, others have not. It is not known whether other inorganic particulates that occur in water produce any direct ejects on human health.

190 DRINKING WATER AND HEALTH Indirect Effects on Health The concentration of inorganic, organic and biological pollutants is usually much higher in the suspended solids and sediments of streams and lakes than in water. Clay, organic, and biological particulates, alone or in combination, are the materials chiefly responsible for such concentrations. Clay and organic particulates have large surface areas and strongly adsorb ions, polar and nonpolar molecules, and biological agents. Occurrence of these materials in water is a consequence of natural events, as well as human activity, and is common in many waters that people drink. Although many of the clay or natural organic particulates, in themselves, may not have deleterious ejects when ingested by humans, they may exert important health ejects through adsorption, transport, and release of inorganic and organic toxicants, bacteria, and viruses. The clay or organic complex with a pollutant may be mobilized by erosion from the land, or complexes may form when eroded particulate matter enters a stream containing pollutants. The atmosphere is also an important pathway. In the adsorbed state, organic and inorganic toxicants may be less active; however, the possibility exists that the toxicants may be released from the particulate matter in the alimentary tract and then exert toxic ejects. It is not clear how complexes of particulate matter with viruses and bacteria behave in the gut. It is known, however, that some enzymes retain their activity when adsorbed on clay, and that viral-clay particulates are infectious in tissue culture and in animal hosts. Turbidity as an Indicator A high turbidity measurement is an indication that a water may produce an adverse health eject; however, a low turbidity measurement does not guarantee that a water is potable. Turbidity measurements do not indicate the type, number, or mass of particles in a water supply. Where particulates in water are suspected of being harmful, the particulate content should be identified and counted by more specific techniques. Such techniques may include biological, organic, inorganic, and fibrous particulate surveys. Turbidity measurements are valuable for process control in water- treatment plants. However, the results obtained with present instruments, procedures, and units of measurement are not well correlated with particle concentrations and size distribution. The test itself must be standardized and refined to facilitate its use for this and other purposes.

Solid Particles in Suspension 191 Conclusions and Recommendations Certain mineral fibers found in water are suspected of being harmful upon ingestion. The available data with respect to asbestos orally ingested through drinking water do not suggest an immediate hazard to public health. They do suggest that additional research, both experimen- tal (using animals) and epidemiological, is required to determine the degree of hazard. Until new results become available, contamination of drinking water by mineral fibers should be kept to a minimum through the use of effective coagulation and filtration processes and other appropriate measures. Because particulates are vehicles for concentration, transport, and release of pollutants, they may have indirect ejects on health. Coagula- tion and filtration are elective methods of reducing particulate concen- trations. Measurement of particulate content by turbidimetry is imprecise and cannot be relied upon as a sole indicator of the safety of an uncharacterized drinking water source. Recommendations for Future Research 1. A survey of suspended particulate matter in raw and treated drinking water supplies in several "typical" communities is urgently needed as background information. This must be coupled with analysis of accompanying organic and inorganic material and microorganisms, as well as characterization of the particulates with respect to size, shape, composition, and adsorbed constituents. 2. Ingestion studies should be carried out with fibers of various types and size distributions in validated animal model systems. 3. Epidemiological studies of time trends in death rates should be conducted in areas that have high concentrations of mineral fibers in drinking water. 4. Electron microscopy procedures for detecting and counting asbes- tos fibers should be scrutinized with respect to their specificity, precision, and accuracy. 5. Information is required on the ejects of inorganic, organic, and biological toxicants adsorbed on clay and organic particulates. 6. Development of improved and standardized methods for determin- ing particle concentrations and size distributions by optical techniques, such as light scattering and absorption, should be supported. 5. Information

192 DRINKING WATER AND HEALTH REFERENCES Clay Particles and their Interactions Bailey, G.W., A.P. Barnett, W.R. Payne, Jr., and C.N. Smith. 1974. Herbicide runof~from four coastal plain soil types. EPA 660/2-74017. Bailey, G.W., and J.L. White. 1970. Factors influencing the adsorption, Resorption and movement of pesticides in soil. Residue Rev. 32:29-92. Blatter, C.L. 1973. Hg-complex intergrades in smectite. Clays Clay Min. 21:261-263. Brown, G. 1961. The X-ray Identification and Crystal Structures of Clay Minerals. Mineralogical Society, London. Chaussidon, J., and R. Calvet. 1965. Evolution of amine cations adsorbed on mont- morillonite with dehydration of the mineral. J. Phys. Chem. 69:2265-2268. Doner, H. E. , and M . M . Mortland. 1969. Benzene complexes with Cu(lI) montmorillonite. Science 166: 1406-1407. Farmer, V.C. 1975. The Infrared Spectra of Minerals. Mineralogical Society, London. Farmer, V.C., and J.D. Russell 1967. Infrared adsorption spectrometry in clay studies. Clays Clay Min. 15:121-142. Fenn, D.B., M.M. Mortland, and T.J. Pinnavaia. 1973. The chemisorption of anisole on Cu- montmorillonite. Clays Clay Min. 21:315-322. Fleck, E.E., and H.L. Haller. 1945. Compatibility of DDT with insecticides, fungicides and fertilizers. Ind. Eng. Chem. 37:403-405. Frink, C.R. 1969. Chemical and mineralogical characteristics of eutrophic lake sediments. Soil Sci. Soc. Am. Proc. 33:369-372. Gard, J.A. 1971. The Electron-Optical Investigation of Clays. Mineralogical Society, London. Greenland, D.J. 1965. Interactions between clays and organic compounds in soils II. Soils Fertil. 28:521-532. Greenland, D.J. 1971. Interactions between humic and fulvic acids and clays. Soil Sci. 111:34. Grim, R.E. 1969. Clay Mineralogy. McGraw Hill Book Company, New York. Gupta, G.C., and W.V. Malik. 1969. Transformation of montmorillonite to nickel chlorite. Clays Clay Min. 17:233-240. Hodgson, J.D. 1963. Chemistry of the micronutrient in soils. Adv. Agron. 15:119-159. Holdridge, D.A. 1969. The sorption of heavy-metal cations by ball clay. Proc. Int. Clay Conf. Israel 1:341-349. Jenne, E.A. 1968. Controls on Mn, Fe, Co, Ni, Cu. and Zn concentrations in soils and water: the significant role of hydrous Mn and Fe oxides. Adv. Chem. 73:337-387. Jenne, E.A., and J.S. Wahlberg. 1968. Role of certain stream-sediment components in radioion sorption. Geol. Prof. Pap. 433-F. Lietzke, D.A., and M.M. Mortland. 1973. The dynamic character of a chloritized vermiculitic soil clay. Soil Sci. Soc. Am. Proc. 37:651-656. Mackenzie, R.C. 1957. The Differential Thermal Investigation of Clays. Mineralogical Society, London. Malina, M.A., A. Goldman, L. Trademan, and P.B. Polen. 1956. Deactivation of mineral  carriers for stable heptachlor-dust formulations. J. Agr. Food Chem. 4:1038-1042. McAuliffe, C., and N.T. Coleman. 1955. H-ion catalysis by acid clays and exchange resins. Soil Sci. Soc. Am. Proc. 19:156-160.

Solid Particles in Suspension 193 Mortland, M.M. 1966. Urea complexes with montmorillonite: An infrared absorption study. Clay Min. 6:143-156. Mortland, M.M., and K.V. Raman. 1967. Catalytic hydrolysis of some organic phosphate pesticides by Copper (II). J. Agr. Food Chem. 15:163-167. Mortland, M.M. 1970. Clay organic complexes and interactions. Adv. Agron. 22:75-117. Nicholson, H. P. 1969. Occurrence and significance of pesticide residues in water. J. Wash. Acad. Sci. 59:45. Nicholson, H.P., and D.W. Hill. 1970. Pesticide contaminants in water and mud and their environmental impact. In Relationship of Agriculture to Soil and Water Pollution. Cornell Univ. Conf. Agr. Waste Mangement. pp. 171-179. Pinnavaia, T.J., P. Hall, S. Cady, and M.M. Mortland. 1974. Aromatic radical cation formation on the intracrystal surfaces of transition metal layer lattice silicates. J. Phys. Chem. 78:994999. Rausell-Colom, J.A., and J.M. Serratosa. 1975. Reaction of clays with organic substances. In Chemistry of Clays and Clay Minerals. Mineralogical Society, London. (in press) Rosenfield, C., and W. Van Valkenburg. 1965. Decomposition of O,mdimethyl ~2, 4,5- trichlorophenyl phosphorothioate (ronnel) adsorbed on bentonite and other clays. J. Agr. Food Chem. 13:68-72. Russell, J.D., G.W. Bailey, W.R. Payne, J.D. Pope, and J.I. Teasley. 1968. Mode of chemical degradation of s-tri~nes by montmorillonite. Science. 160:1340-1342. Sanchez, A., A. Hidalgo, and J.M. Serratosa. 1972. Adsorption des nitriles dans la montmorillonite. Proc. Int. Clay Conf. Madrid 617:626. Schnitzer, M., and H. Kodama. 1972. Reactions between fulvic acid and Cu2+-montmoril- lonite. Clays Clay Min. 20:359-367. Schnitzer, M., and H. Kodama. 1967. Reaction between a podzol fulvic acid and Na- montmorillonite. Soil Sci. Soc. Am. Proc. 31:632-636. Solomon, D.H., and B.C. Loft. 1968. Reactions catalyzed by clay minerals. III. The mechanisms of spontaneous interlamellae polymerizations in aluminosilicates. J. Appl. Polymer.Sci. 12:1253-1262. Tamura, T. 1962. Strontium reactions with minerals. Ground Disposal of Radioactive Wastes Conf. 2d, Chalk River Canada, Sept. 26-29, 1961. U.S. Atomic Energy Comm. TID-7628, Book 1, 187-197. Theng, B.K.G. 1974. The Chemistry of Clay-Organic Reactions. Halstead Press, New York. Weaver, C.E., and L.D. Pollard. 1973. The Chemistry of Clay Minerals. Elsevier Scientific Publishing Co., Amsterdam. Whittig, L.D. 1965. Methods of Soil Analvsis. vol. 1. oo. 671-696. Am. Soc. APron. Inc., Madison, Wis. ---~- ~rr ~ Asbestos: Nomenclature, Occurrence, and Redistribution in Water Ampian, S.G. 1976. Asbestos minerals and their nonasbestos analogs. In Review of Mineral-  Fibers Session, Electron Microscopy of Microfibers. Pennsylvania State University, University Park, Pa. Chisholm, J.E. 1973. Planar defects in fibrous amphiboles. J. Material Sci. 8:475-483. Choi, I., and R.W. Smith. 1971. Kinetic study of dissolution of asbestos fibers in water. J. Colloid Interface. Sci. V. 40: 253-262. Cook, P.M., G.E. Glass, and J.H. Tucker. 1974. Asbestiform amphibole minerals: detection and measurement of high concentrations in municipal water supplies. Science 185:853- 855.

194 DRINKING WATER AND HEALTH Committee on Mineral Names. 1949. Am. Min. 34:339. Cotterell, K., and P.F. Halt. 1972. An examination of crocidolites from North West Cape and Transvaal Mines. Inst. Min. Metall. Trans. Sect. B. 81:169-171. Cralley, LJ., R.G. Keenan, J.R. Lynch, and W.S. Lainhart. 1968. Source and identification of respirable fibers. J. Am. Ind. Hyg. Assoc. 29:129-135. Cunningham, H.M. and R Pontefract. 1971. Asbestos fibers in beverages and drinking water. Nature 232:332-333. Davis, W.E. 1970. National inventory of sources and emissions of asbestos. NTIS PB 192252. Deer, W.A., R.A. Howie, and J. Zussman. 1966. An Introduction to the Rock-forming Minerals. Longmans, London. Durham, R.W., and T. Pang. 1976. Asbestiform fiber levels in Lakes Superior and Huron. Scientific Series no. 67. Inland Waters Directorate, Canada Centre for Inland Waters, Burlington, Ontario. Environment Canada. 1973. National inventory of sources and emissions of asbestos. Report APCD 73-4, Air Pollution Control Directorate, Environment Canada, Ottawa. EPA. 1975. Preliminary assessment of suspected carcinogens in drinking water. Interim Report to Congress. Environmental Protection Agency. Flentje, M.E., and RJ. Schweitzer. 1955. Further study of solution effects on concrete and cement in pipe. J. Am. Water Works Assoc. 49:1441. Hartman, P. 1963. Structure, growth and morphology of crystals. Z. Kristallog. 119:65-78. House, R.F. 1967. Dispersion of asbestos. U.S. Pat. Office no. 3,586, 639. Hutchinson, J.L., M.C. Irusteta, and E.J.W. Whittaker. 1975. High-resolution electron diffraction studies of fibrous amphiboles. Acta Crystallog. 31:794-801. Kay, G. 1973. Ontario intensifies search for asbestos in dunking water. J. Water Pollut. Control Fed:33-35. Kehieker, D.M. et al. 1967. Determination of elementary fiber size of chrysotile asbestos. Sov. Phys. Crystallog. 12:430-435. Kramer, J.R. 1976. Fibrous c~mmmgtonite in Lake Superior. Can. Min. 14:91-98. Kristiansen, H. 1974. The extraction of calcium by soft water from prestressed concrete pipes. Vatten, 1:70. May, T.C., and R.W. Lewis. 1970. Asbestos. In Mineral Facts and Problems. U.S. Bureau Mines Bulletin 650:851-865. NAS-NRC. 1971. Airborne asbestos. Committee on Biologic Effects of Atmospheric Pollutants. National Research Council. (See also Environment Canada, 1973.) Parks, G.A. 1967. Aqueous surface chemistry of oxides and complex oxide minerals. In W. Stuum, ed. Equilibrium Concepts in Natural Water Systems, Adv. Chem. no. 67, Am. Chem. Soc. Rabbit, J.C. 1948. Am. Min. 33:263-323. Rendall, R.E.G. 1970. The data sheets on the chemical and physical properties of the UICC standard reference samples. In H.A. Shapiro, ed. Pneumoconiosis. Oxford University Press. Ruud, C.O., C.S. Barrett, P.A. Russell and R.L. Clark. 1976. Selected area electron diffraction and energy dispersive X-ray analysis for the identification of asbestos fibers, a comparison. Micron 7: 1 15-132. Timbrell, V. 1970. Characteristics of the International Union Against Cancer Standard  Reference Samples of Asbestos. In H.A. Shapiro, ed. Pneumoconiosis. Oxford University Press. Timbrell, V., F. Pooley, and J.C. Wagner. 1970. Characteristics of respirable asbestos fibers. In H.A. Shapiro, ed. Pneumoconiosis. Oxford University Press.

Solid Particles in Suspension 195 Timbrell, V., and R.E.G. Rendall. 1972. Preparation of the UICC reference samples of asbestos. Powder Tec. 5:279-287. Whittaker, E.J.W. 1966. Diffraction contrast in electron microscopy of chrysotile. Acta. Crystallog. 21:461-466. Whittaker, E.J.W. and J. Zussman. 1971. The Serpentine Minerals. In J.A. Gard, ed. Electron Optical Investigations of Clays. The Mineralogical Society, London. Yada, K. 1964. Study of chrysotile asbestos by a high resolution electron microscope. Acta Crystallog. 23:704707. Zoltai, Tibor, and J.H. Stout. 1976. Comments on asbestiform and fibrous mineral fragments, relative to Reserve Mining Company taconite deposits. Report to Minn. Pollution Control Agency, Minneapolis. Wright, G.W. 1974. Does the use of asbestos-cement pipe for potable water systems cause a health hazard? J. Am. Water Works Assoc. 66:4-22. Asbestos Fiber Sampling and Analysis Beaman, D.R., and D.M. File. 1975. The quantitative determination of asbestos fiber concentrations. The Dow Chemical Company, unpublished report. Berkley, D., J. Churg, I.J. Selikoff, and W.E. Smith. 1965. The detection and localization of asbestos fibers in tissue. Ann. N.Y. Acad. Sci. 132:48-63. Berkley, C., A.M. Langer, and V. Baden. 1968. Instrumental analysis of inspired fibrous pulmonary particles. Trans. N.Y. Acad. Sci. 30:331-350. Birks, L.S., M. Fatemi, J.V. Gilfrich, and E.T. Johnson. 1975. Quantitative analysis of airborne asbestos by x-ray diffraction. Feasibility Study AD-A007530, Naval Res. Lab., Washington, D.C. Brown, A.L., Jr., W.F. Taylor, and R.E. Carter. 1976. The reliability of measures of amphibole fiber concentration in water. Environ. Res. 12:150-160. Clark, R.L., and C.O. Ruud. 1975. Transmission electron microscopy standards for asbestos. Micron 5:270. Cook, P.M., J.B. Rubin, C.J. Maggiore, and W.J. Nicholson. 1974. X-ray diffraction and electron beam analysis of asbestiform minerals in Lake Superior waters. In Trans. Inst. Electrical Electronic Eng. (in press). Crable, J.V., and M.J. Knott. 1966a. Application of x-ray diffraction to the determination of chrysotile in bulk and settled dust samples. Am. Ind. Hyg. J. 27:383-387. Crable, J.V., and M.J. Knott. 1966b. Quantitative x-ray diffraction analysis of crocidolite and amosite in bulk or settled dust samples. Am. Ind. Hyg. J. 27:449-453. Dement, J.M., R.D. Zumwalde, and K.M. Wallingford. 1975. Asbestos fiber exposures in a hard rock gold mine. In Proc. N.Y. Acad. Sci. Conf. Occup. Carcinogenesis. Ann. N.Y. Acad. Sci. 271:345-352 (1976). Ferrell, R.E., G.G. Paulson, and C.W. WaLlcer. 1975. Evaluation of an SEMEDS method for identification of chrysotile. Scanning Electron Microscopy:537-546. Julian, Y., and W.C. McCrone. 1970. Identification of asbestos fibers by microscopical dispersion staining. Microscope 18:1-10. Keenan, R.G., and J.R. Lynch. 1970. Techniques for the detection, identification and analysis of fibers. Am. Ind. Hyg. J. 31:587-597.  Langer, A.M., A.D. Mackler, and F.D. Pooley. 1974. Electron microscopical investigation of asbestos fibers. Environ. Health Perspect. 9:63-80. Langer, A.M., I. Rubin, and I.J. Selikoff. 1975. Electron microprobe analysis of asbestos bodies. Histochem. Cytochem. J. 20:735-740.

196 DRINKING WATER AND HEALTH McCrone, W.C., and I.M. Stewart. 1974. Asbestos. American Laboratory, April. McMillan, L.M., R.G. Stout, and B.F. Willey. 1977. Asbestos in raw and treated water: an electron microscopy study. Environ. Sci. Tech. 11:390-394. Nicholson, W.J. 1974. Analysis of amphibole asbestiform fibers in municipal water supplies. Environ. Health Perspect. 9:165-172. Pooley, F.D. 1972. Electron microsope characteristics of inhaled chrysotile asbestos fibre. Br. J. Ind. Med. 29: 146-153. Otiz, L.W., and B.L. Isom. 1974. Transfer technique for electron microscopy of membrane filter samples. Am. Ind. Hyg. Assoc. J.35:423-425. OSHA. 1975. Occupational Safety and Health Standards. U.S. Department of Labor, Occupational Safety and Health Admmistraion. Fed. Reg. 29 CFR 1910.1001. Rohl, A.N., and A.M. ~ anger. 1974. Identification and quantitation of asbestos in talc. Environ. Health Perspect. 9:95-109. Rubin, I.B., and C.J. Maggiore. 1974. Elemental analysis of asbestos fibers by means of electron probe techniques. Environ. Health Perspect. 9:81-94. Ruud, C.O., C.S. Barrett, P.A. Russell, and R.L. Clark. 1976. Selected area electron diffraction and energy dispersed x-ray analysis for the identification of asbestos fibers, a comparison. Micron 7: 115-132. Schlez, J.P. 1974. The detection of chrysotile asbestos at low levels in talc by differential thermal analysis. Thermochemica Acta 8: 197-203. Selikoff, I.J., W.J. Nicholson and A.M. Langer. 1972. Asbestos air pollution. Arch. Environ. Health 25: 1-13. Stanley, H.D., and R.E. Norward. 1973. The detection and identif~cation of asbestos and asbestiform minerals in talc. In Proceedings Symp. on Talc, Washington, D.C., May 8, 1973. Bureau of Mines Information Circular 8639. Timbrell, V. 1970. Characteristics of the UICC standard reference samples of asbestos. In H. Shapiro, ed. Pneumoconiosis. Oxford Univ. Press, London. Biological Effects of Asbestos Minerals, Epidemiological Findings Anderson, H.A., R. Lilis, S.M. Daum, A.S. Fischbein, and I.J. Selikoff. 1976. Household- contact asbestos neoplastic risk. Ann. N.Y. Acad. Sci. 271:311-323. Bogovski, P., V. Timbrell, J.C. Gilson and J.C. Wagner, eds. 1973. Biological Effects of Asbestos. IARC Scientific Publication no. 8. Elmes, P.C., and M. J. C. Simpson. 1971. Insulation workers in Belfast. 3. Mortality 1940 1966. Br. J. Ind. Med. 28:226-236. Hammond, E.C., I.J. Selikoff, H. Seidman. 1975. Multiple interaction effects of cigarette smoking. Extrapulmonary cancer. In Proc. XI Int. Cancer Congress, Florence 1974, vol. 3, Cancer Epidemiology. Environ. Factors: 147-150. Excerpta Medica, Amsterdam. Lee, D.H.K., ed. 1974. Proceedings of the Joint NIEHS-EPA Conference on "Biological Effects of Ingested Asbestos," Durham, North Carolina, November 1973. Environ. Health Perspect. 9:113-338. Levy, B.S., E. Sigurdson, J. Mandel, E. Laudon, and J. Pearson. 1976. Investigating possible effects of asbestos in city water: Surveillance of gastrointestinal cancer incidence in Duluth, Minnesota. Am. J. Epidemiol. 103:362-368.  Mancuso, T.F. 1965. Asbestos and neoplasia: Epidemiology: Discussion. Ann. N.Y. Acad. Sci. 132:589-594. Masson, T.J., F.W. McKay, R.W. Miller. 1974. Asbestos-lilce fibers in Duluth water supply: Relation to cancer mortality. J. Am. Med. Assoc. 228:1019-1020.

Solid Particles in Suspension 197 Newhouse, M.L., G. Berry, J.C. Wagner, and M. E. Turok. 1972. A study of the mortality of female asbestos workers. Br. J. Ind. Med. 29:134141. Newhouse, M.L., and H. Thompson. 1965. Mesothelioma of the pleura and peritoneum following exposure to asbestos in the London area. Br. J. Ind. Med. 22:261-269. Selikoff, I.J., and J. Churg, Co-chairmen. 1965. Biological effects of asbestos. Ann. N.Y. Acad. Sci. 132:1-766. Selikoff, I.J., E.C. Hammond, and H. Seidman. 1973. Cancer risk of insulation workers in the United States. In Biological Effects of Asbestos. IARC Scientific Publication no. 8:209- 216. Wagner, J.C., C.A. Sleggs, and P. Marchand. 1960. Diffuse pleural mesothelioma and asbestos exposure in the North Western Cape Province. Br. J. Ind. Med. 17:260-271. Biological Effects of Asbestos Minerals, Experimental Studies Ampian, S.G. 1976. Asbestos minerals and their nonasbestos analogs. In Review of Mineral- Fibers Session, Electron Microscopy of Microfibers. Pennsylvania State University, University Park, Pa. Bolton, R.E., and J.M.G. Davis. 1976. The short-term effects of chronic asbestos ingestion in rats. Ann. Occup. Hyg. 19:121-128. Bouser, G.M., and D.B. Clayson. 1967. Feeding of blue asbestos to rats. 45th Annual Report, British Empire Cancer Campaign:242. Cook, P.M., G.E. Glass, and J.H. Tucker. 1974. Asbestiform amphibole minerals: detection and measurement of high concentrations in municipal water supplies. Science 185:853- 855. Cunningham, H.M., and R. Pontefract. July 30, 1971. Asbestos fibers in beverages and drinking water. Nature 232:332-333. Cunningham, H.M., and R.D. Pontefract. 1974. Placental transfer of asbestos. Nature 249: 177-178. Cunningham, H.M., R.D. Pontefract, and R.C. O'Brien. 1976. Quantitative relationship of fecal asbestos to asbestos exposure. J. Toxicol. Environ. Health 1:377-379. Fears, T.R. 1976. Cancer mortality and asbestos deposits. Am. J. Epidem~ol. 104:523-526. Gibel, W., Kh. Lohs, K.H. Horn, G.P. Wildner, and F. Hoffmann. 1976. Tier experimentelle Untersuchungen uber eine kanzerogene Wirkung von Asbestfiltermaterial nach oraler Aufaahme. Arch. Gesch~vulstforsch. 46:437442. Gross, P. 1974. Letter to ea.: Asbestos fibers in drinking water. J. Am. Med. Assoc. 229:767. Gross, P., R.A. Harley, L.M. Swinburne, J.M.G. Davis, and W. B. Greene. 1974. Ingested mineral fibers. Do they penetrate tissue or cause cancer? Arch. Environ. Health. 29:341- 347. Hammond, E.C., and I.J. Selikoff. 1973. Relation of cigarette smoking to risk of death of asbestos associated disease among insulation workers in the United States. In Biological Effects of Asbestos. IARC Scientific Publication no. 8:312-377. Harwood, C.F., and G. Yamate. 1975. The detection and quantification of asbestos present in the environment. Presented at Third International Conference on the Physics and Chemistry of Asbestos Minerals, Quebec.  IARC 1973. Report of the Advisory Committee on Asbestos Cancers to the l~irector of the International Agency for Research on Cancer. Br. J. Ind. Med. 30:180-186. Kay, G. September, 1973. Ontario intensifies search for asbestos in d~nking water. Water Pollut. Control 72:33-35.

198 DRINKING WATER AND HEALTH Levy, B.S., E. Sigurdson, J. Mandel, E. Laudon and J. Pearson. 1976. Investigating possible effects of asbestos in city water: Surveillance of gastrointestinal cancer incidence in Duluth, Minnesota. Am. J. Epidemiol. 103:362-368. Maroudas, N.G., C.H. O'Neill, and M.F. Stanton. April 14, 1973. Fibroblast anchorage in carcinogensis by fibres. Lancet I :807-809. Masson, T.J., F.W. McKay, and R.W. Miller. 1974. Asbestos-like fibers in Duluth water supply. Relation to cancer mortality. J. Am. Med. Assoc. 228:1019-1020. NAS 1975. Priniciples for Evaluating Chemicals in the Environment. National Academy of Sciences, Washington, D.C. Nicholson, W.J., and F.L. Pundsack. 1973. Asbestos in the environment. In Biological Effects of Asbestos. IARC Scientific Publication no. 8: 126-130. Pontefract, R.D., and H.M. Cunningham. 1973. Penetration of asbestos through the digestive tract of rats. Nature 243:352-353. Pott, F., and K.H. Friedrichs. 1972. Tumors in rats by intraperitoneal injection of fibrous dusts. Naturwissenschaften 59:318. Pott, F., F. Huth, and K.H. Friedrichs. 1972. Tumors of rats after L.P. injection of powdered chrysotile and benzo[a]pyrene. Zentrabl. Bakteriol. Parasitenkd. Infektionskr. Hyg., Abt. 1: Orig. Reibe B 155 :463469. Pott, F., F. Huth, and K.H. Friedrichs. 1974. Tumorigenic effect of fibrous dusts in experimental animals. Environ. Health Perspect. 9:313-315. Pott, F., F. Huth, and K.H. Friedrichs. April, 1976. Result of animal carcinogenesis studies after application of fibrous glass and their implications regarding human exposure. In Occupational Exposure to Fibrous Glass. A Symposium, Univ. of Maryland, College Park, Maryland, June 26-27, 1974:183-191. U.S. Department of Health, Education, and Welfare. NIOSH 76-151. Shabad, L.M., L.N. Pyleo, L.V. Krioosheeva, T.F. Kulag~na, and B.A. Neminko. 1974. Experimental studies on asbestos carcinogenicity. J. Cancer Inst. 52:1175-1187. Smith, W.E. 1973. Asbestos, talc and nitrites in relation to gastric cancer. Am. Ind. Hyg. Assoc. J. 34:227-228. Smith, W.E. 1974. Experimental studies on biological effects of tremolite talc on hamsters. In Proc. Symp. on Talc, Washington, D.C., May 8, 1973. Bureau of Mines Information Circular 8639:4348. Smith, W.E., D.D. Hubert, and M.S. Badollet. 1972. Biologic differences in response to long and short asbestos fibers. Am. Ind. Hyg. Assoc. J. 33:A162. Sontag, J.M., N.P. Page, and U. Saffiotti. 1976. Guidelines for Carcinogen Bioassay in Small Rodents. NCI Carcinogenesis Technical Report Series No. 1. Department of Health, Education, and Welfare Publication No. (NIH) 76-801. Stanton, M.F. 1973. Some etiological considerations of fibre carcinogenesis. In Biological Effects of Asbestos. IARC Scientific Publication no. 8:289-294. Stanton, M.F. 1974. Fiber carcinogenesis: Is asbestos the only hazard? J. Nat. Cancer Inst. 52:633-634. Stanton, M.F., R. Blackwell, and E. Miller. 1969. Experimental pulmonary carcinogenesis with asbestos. Am. Ind. Hyg. Assoc. J. 30:236-244. Stanton, M.F., M. Layard, A. Tegeris, E. Miller, M. May, and E. Kent. 1977. Carcinogenici- ty of fibrous glass: Pleural response in the rat in relation to fiber dimension. J. Nat.  Cancer Inst. 58:587-603. Stanton, M.F., and C. Wrench. 1972. Mechanisms of mesothelioma induction with asbestos and fibrous glass. J. Nat. Cancer. Inst. 48:797-821. Wagner, J.C., G. Berry, J.W. Skidmore, and V. Timbrell. 1974. The effects of the inhalation of asbestos in rats. Br. J. Cancer 29:252-269.

Solid Particles in Suspension 199 Wagner, J.C., G. Berry, and J.W. Skidmore. April 1976. Studies of the carcinogenic effects of fiber glass of different diameters following intrapleural innoculation in experimental animals. In Occupational Exposure to Fibrous Glass. A Symposium, Univ. of Maryland, College Park, Maryland, June 26-27, 1974:193-197, U.S. Department of Health, Education, and Welfare. N1OSH 76-151. Wagner, J.C., G. Berry, and V. Timbrell. 1973. Mesotheliomata in rats after inoculation with asbestos and other materials. Br. J. Cancer 28:173-185. Webster, I. 1974. The ingestion of asbestos fibers. Environ. Health Perspect. 9:199-202. Westlake, G.E., H.J. Spjut, and M.N. Smith. 1965. Penetration of colonic mucosa by asbestos particles. An electron microscopic study in rats fed asbestos diets. Lab. Invest. 9:2029-2033. Organic Particulates in Water Anderson, G. 1958. Identification of derivatives of deoxyribonucleic acid in humic acid. Soil Sci. 86:196-174. Ballard, T.M. 1971. Role of humic carrier substances in DIET movement through forest soil. Soil Sci. Soc. Am. Proc. 35:146-147. Black, A.P., and R.F. Christman. 1963. Characteristics of colored surface waters. J. Am. Water Works Assoc. 55:753-770. Bordeleau, L. M ., J. D. Rosen, and R. Bartha. 1972. Herbicide-derived chloroazobenzene residues: Pathway of formation. J. Agr. Food Chem. 20:573-578. Bruland, K.W., K. Bertine, M. Koide, and E.D. Goldberg. 1974. History of metal pollution in Southern California coastal zone. Environ. Sci. Tech. 8:425-432. Cheam, V. 1973. Chelation study of copper (II): Fulvic acid system. Can. J. Soil Sci. 53:377- 382. Day, H., and G.T. Felback, Jr. 1974. Production and analysis of a humicacid-like exudate from the aquatic fungus Aureobasidium pullulans. J. Am. Water Works Assoc. 66:484 488. De Groot, A.J., J.J.M. De Goeij, and C. Zegers. 1971. Contents and behavior of mercury as compared with other heavy metals in sediments from the River Rhine and Ems. Geologie en Mijobouw. 50:393-398. Felbeck, G.T., Jr. 1975. Structural chemistry of soil humic substances. Adv. Agron. 17:327- 368. Felbeck, G.T., Jr. 1971. Structural hypotheses of soil humic acids. Soil Sci. 111:43~8. F7ield, R., and D. Knowled. 1975. Urban runoff and combined sewer overflow, J. Water Pollut. Control Fed. 47:1352-1369. Flaig, W. 1970. Contribution a la connaissance de la constitution et de la synthese des acides humiques. Sci. Sol 1:39-72. Giger, W., M. Reinhard, C. Schaffner, and W. Stumm. 1974. Petroleum-derived and idigenous hydrocarbons in recent sediments of Lake Zug, Switzerland. Environ. Sci. Tech. 8:454-455. Gove, G.W. and I. Gellman. 1975. Paper and allied products. J. Water Pollut. Control Fed. 47:1402-1446.  Gorbunov, N.I., G.L. Yerokhina, and G.N. Shchurina. 1971. Relationship between soil minerals and humic substances, Pochvovedeniye 7:117-128. Greenland, D.J. 1971. Interactions between humic and fulvic acids and clays. Soil Sci. 111:34- 31.

200 DRINKING WATER AND H"LTH Haider, K., J.P. Martin, Z. Filip, and E. Fustec-Mathon. 1972. Contribution of soil microbes to the formation of humic compounds. Proc. Int. Meet. Humic Substances:71-85. Harding, C., and H.S. Brown. 197S. Distribution of selected trace elements in sediments of Pamlico River estuary, North Carolina. Environ. Geol.:181-191. Hsu, T-S., and R. Bartha. 1976. Hydrolyzable and nonhydrolyzable 3,4-dichloroaniline- humus complexes and their respective rates of biodegradation. J. Agric. Food Chem. 24:118-122. Jewell, W.J., J.B. Petersen, E.G. Srinath, W.T. Tseng, E.J. Kroeker, and E.C. McGriff, Jr. 1975. Agricultural wastes. Water Pollut. Control Fed. 47:1446-1465. Khan, S.U., and M. Schnitzer. 1972. The retention of hydrophobic organic compounds by humic acid. Geochem. Cosmochim. Acta 36:745-754. Kononova, M.M. 1966. Soil Organic Matter. Pergamon Press, New York. Koshy, E., and A.K. Ganguly. 1969. Organic materials in the marine environments and their interactions with some metal ions. Bhabha Atomic Research Centre, Bombay, India. Ladd, J.N., and J.H.A. Butler. 1971. Inhibition by soil humic acids of naive and acetylated proteolytic enzymes. Soil Biol. Biochem. 3:157-160. Lammers, W.T. 1967. Biophysical limnology, separation of suspended and colloidal particles from natural water. Environ. Sci. Tech. 1:52-57. Lammers, W.T. 1975. An investigation of colloidal organic particles isolated from different populations of bivalves. Verh. Int. Verein. Limnol. 19:154~1545. Leland, H.V., S. Shukla, and N.F. Shimp. 1973. Factors affecting distribution of lead and other trace elements in sediments of southern Lake Michigan. In C. Singer, ed. Trace Metals and Metal-Organic Interactions in Natural Waters, pp. 89-129. Leland, H.V., E.D. Copenhaver, and D.J. Wilkes. 1975. Tidal pollution. Heavy metals and other trace elements. J. Water Pollut. Control Fed. 47:1635-1656. Li, Gwo-Chen, and G.T. Felbeck, Jr. 1972. A study of the mechanism of atrazine adsorption by humic acid from muck soil. Soil Sci. 113:140 148. Litchf~eld, J.H. 1975. Meat, fish, and poultry processing wastes. J. Water Pollut. Control Fed. 47:1381-1389. Macauley, D.C. 1975. Chemicals and allied products. J. Water Pollut. Control Fed. 47:1515- 1520. Malcolm, R.L. 1969. A comparison of conditional stability constants of North Carolina humic and fulvic acids with Co(II) and Fe(lII): Southeast. Geol. Soc. Am. Meet. Proc., April, Columbia, S.C. Martin, J.P., K. Haider, and E. Bondietti. 1972. Properties of model humic acids synthesized by phenolaxidase and autoxidation of phenols and other compounds formed by soil fungi. Proc. Int. Meet. Humic Substances:171-186. Mato, M.C., R. Fabregas, and J. Mendez. 1971. Inhibitory effect of soil humic acids on indoleacetic acidoxidase. Soil Biol. Biochem. 34:285-288. Mazzacchi, P.H., and M.P. Rao. 1972. Photolysis of 3-~chlorophenyl)-1,1-dimethylurea (Monuron) and 3-phenyl-1,1-dimethylurea (Fenuron). J. Agr. Food Chem. 20:957-959. Monahan, A.R., A.F. DeLuca, and R.L. Wershaw. 1972. Spectroscopic characterization of humic acid fractions in aqueous media. Am. Chem. Soc. Meet., Aug. 27-Sept. 1, New York. Muller, G., and U. Forstner. 1974. Heavy metals in sediments of the Rhine and Elbe  estuaries: Mobilization or mixing effect? Environ. Geol. 1:33-39. Nicholson, A.A., O. Meresz, and B. Lemyk. 1977. Determination of free and total potential haloforms in drinking water. Anal. Chem. 49:814819. Ong, H.L., V.E. Swanson, and R.E. Bisque. 1970. Natural organic acids as agents of chemical weathering. U.S. Geol. Sur. Prof. Pap. 70~C:C13~C137.

Solid Particles in Suspension 201 Orlov, D.S., and N.L. Yeroshicheva. 1967. Interaction of humic acids with the cations of some metals. Soviet Soil Sci. S(570) Sa34E: 1799-1806. Perry, D.R., and W.A. Adams. 1971. The incorporation of glycylglycine into humic acid. Biochem. J. 125:29-30. Pico, R.F. 1975. Dairy wastes. J. Water Pollut. Control Fed. 47: 1513-1516. Rashid, M.A. 1969. Contribution of humic substances to the cation exchange capacity of different marine sediments. Maritime Sediments 5:4450. Rashid, M.A. 1971. Role of humic acids of marine origin and their different molecular weight fractions in complexing di- and trivalent metals. Soil Sci. 111 :298-306. Rashid, M.A. and J.D. Leonard. 1973. Modifications in the solubility and precipitation behavior of various metals as a result of their interaction with sedimentary humic acid. Chem. Geol. 11 :89-97. Rook, J.J. 1974. Formation of haloforms during chlorination of natural waters. Water Treatment and Examination 23:234243. Schemer, F., and B. Ulrich. 1960. Humus und humusdungung. In Lehrbuch der Agrikultur- chemie und Bodenkunde. F. Enke, Stuttgart. Schnitzer, M. 1965. Contribution of organic matter to the cation exchange capacity of soils. Nature 207:665-668. Schnitzer, M., and S.I.M. Skinner. 1965. Organo-metallic interactions in soils: 4. Carboxyl and hydroxyl groups in organic matter and metal retention. Soil Sci. 99:278-284. She1ton, T.B., and J.V. Hunter. 1974. Acrobic decomposition of oil pollutants in sediments. J. Water Pollut. Control Fed. 46:2172-2182. Shimp, N.F., J.A. Schleicher, R.R. Ruch, D.B. Heck, and H.V. Leland. 1971. Trace element and organic carbon accumulation in the most recent sediments of southern Lake Michigan. Environmental Geology Notes no. 41. Sipos, S., I. Dekany, and F. Szanto. 1972. Investigation of humic acids and metal humates with analytical ultracentrifuge. Acta Phys. Fasc.34:253-257. Sodergren, A., Bj. Svensson, and S. Ulfstrand. 1972. DDT and PCB in south Swedish streams. Environ. Pollut.3:25-36. Soderquist. 1975. Fruit, vegetable, and grain processing wastes. J. Water Pollut. Control Fed. 47: 1389-1398. Steelink, C. 1963. What is humic acid? J. Chem. Ed. 40:379. Steelink, C., and G. Tollin. 1967. Free radicals in soil. Soil Biochem. I: 147-169. Stevens, A.A., C.J. Slocum, D.R. Seeger, and G.G. Robeck. 1975. Chlorination of organics in drinking water. In Proceedings of Conference on Environmental Impact of Water Chlorination. Oak Ridge National Laboraory, Oak Ridge, Tenn. Su, G.C.C., and M.J. Zabik. 1972. Photochemistry of bioactive compounds. Photolysis of (N,N-dimethylformamidine) phenyl N-methylcarbamate hydrochloride in water. J. Agr. Food Chem. 20:642-644. Sullivan, J.D., Jr., and G.T. Felbeck, Jr. 1968. A study of the interaction of S-triazine herbicides with humic acids from three different soils. Soil Sci. 106:42-52. Symons, J.M. 1976. Interim Treatment Guide for the Control of Chloroform and other Trihalomethanes. U.S. Environmental Protection Agency. Talbot, R. 1965. Textile wastes. J. Water Pollut. Control Fed. 47: 1465-1473. Theis, T.L., and P.C. Singer. 1974. Complexation of iron(II) by organic matter and its effect  on iron(II) oxygenation. Environ. Sci. Tech. 8:569-573. Wershaw, R.L., P.J. Burcar, and M.C. Goldberg. 1969. Interaction of pesticides with natural organic material. Environ. Sci. Tech.3:271-273. Wershaw, R.L., and D.J. Pinckney. 1971. Association and dissociation of a humic acid fraction as a function of pH. U.S. Geol. Sur. Prof. Pap. 75~D: D216-D218.

202 DRINKING WATER AND H"LTH Wershaw, R.L., and M.C. 1972. Interaction of organic pesticides with natural organic polyelectrolytes. Advances in Chemistry Series no. 111, Fate of Orgnic Pesticides in the Aquatic Environment, pp. 149-158. American Chemical Society. Wershaw, R.L., and D.J. Pinckney. 1973a. The fractionation of humic acids from natural water systems. J. Res. U.S. Geol. Sur. Wershaw, R.L. and D.J. Pinckney. 1973b. Determination of the association and dissociation of humic acid fractions by small angle X-ray scattering. J. Res. U.S. Geol. Sur. Microorganisms and Suspended Particles in Water Berg, G. 1973. Removal of viruses from sewage, effluents, and water. Bull. WHO 49:451~60. Boyd, J.W., T. Yoshid, L.E. Vereen, R.L. Cada and S.M. Morrison. 1969. Bacterial response to the soil environment. Sanitary Engineering Papers, Colorado State Univ., Ft. Collins, No. 5:1-22. Boylen, C.W., and T. D. Brock. 1973. Bacterial decomposition processes in Lake Wingra sediments during winter. Limnol. Oceanogr. 18:628. Brock, T.D. 1966. Principles of Mcrobial Ecology. Prentice-Hall, N.J. Brock, T. D. 1974. Biology of Microorganisms, 2d ed. Prentice-Hall, N. .J. Cameron, R.E. 1965-1969. Soil Studies-Micobial Habitats. NASA Publ. Space Programs Summaries, vol. IV. Jet Propulsion Laboratory, Pasadena, California. Gray, G.W., and S.G. Wilkinson. 1965. The action of EDTA on Pseu~lomonas aeruginosa. J. Appl. Bacteriol. 28:153-164. Hendricks, C.W. 1973. Measurement of baseline levels of enteric bacterial activity in nver water. Louisiana State Univ. Rep. LSU-SG-73-01:245. Hennci, A.T., and E.J. Ordal. 1948. The Biology of Bacteria. D.C. Heath, New York. Hudson, H.E. 1962. High quality water production and viral disease. J. Am. Water Works Assoc. 54:1265-1272. Kononova, M.M. 1966. Soil Organic Matter, 2d Engl. ea., pp. 51-52. Pergamon Press, New York. Lammers, W.T. 1967. Separation of suspended and colloidal particles from natural water. Environ. Sci. Tech. 1:52-57. Lederberg, J. 1952. Cell genetics and hereditary symbiosis. Physiol. Rev. 32:403-430. Mallette, M.F. 1963. Validity of the concept of energy of maintenance. N.Y. Acad. Sci. 102:521-535. Meloick, J.L. 1975. Proceedings 13th Water Quality Conference, Virus and Water Quality: Occurrence and Control. Univ. of Illinois and Illinois EPA. Neefe, J.R., J.B. Baty, J.G. Reinhold, and J. Stokes. 1947. Inactivation of the virus of infectious hepatitis in drinking water. Am. J. Public Health 37:365-372. Petrilli, F.L., P. Crovari, S. DeFlora, and A. Vannucci. 1974. The virological monitoring of water. I. Drinking water. Boll. 1st. Sieroter, Milan. 53:43~442. Rice, C.W., I.L. Uydess, W.P. Hempfling, and W.V. Vishniac. 1975. Isolation of microorgan- isms from soil of the Antarctic "Dry Valleys". Abstr. Annual Meet. Am. Soc. Microbiol., New York City. Sanderson, W.W., and S. Kelly. 1962. Discussion of Human Enteric Viruses in Water:  Source, Survival and Removability, by N.A. Clarke, G. Berg, P.K. Kabler, and S.L. Chang, Int. Conf. Water Pollut. Res. London 1962. Pergamon Press, New York, 1964. Schmidt, E.L. R.O. Bankole, and B.B. Bohlool. 1968. Fluorescent antibody approach to study of rhizobia in soil. J. Bacteriol. 95:1987-1992.

Solid Particles in Suspension 203 Sharp, J.J., and B.D. Church. 1963. Molecular mutualism among the marine protista. Bact. Proc. 996:47. Stotzky, G. 1966a. Influence of clay minerals on microorganisms. II. Effect of various clay species, homoionic clays, and other particles on bacteria. Can. J. Microbial. 12:831-848. Stotzky, G. 1966b. Influence of clay minerals on microorganisms. III. Effect of particle size, cation exchange capacity, and surface area on bacteria. Can. J. Microbiol. 12:1235-1246. Symons, J.M., and J.C. Hoff. 1975. Rationale for turbidity maximum contaminant level. 3d Water Quality Tech. Conf., Am. Water Works Assoc., Atlanta, Gal, December 8-10. Taylor, F.B. 1974. Viruses: What is their significance in water supplies? J. Am. Water Works Assoc. 66(5):306-311. Tracy, H.W., V.M. Camarena, and F. Wing. 1966. Coliform persistence in highly chlorinated waters. J. Am. Water Works Assoc. 58: 1 151. Walton, G. 1961. Effectiveness of water treatment processes as measured by coliform reduction. U.S. Department of Health, Education, and Welfare, PHS Publ. no. 898. Particulate Removal and Turbidity Adin, A., and M. Rebhun. 1974. High-rate contact flocculation-filtration with cationic polyelectrolytes. J. Am. Water Works Assoc. 66:109-117. American Public Health Association. 1976. Standard Methods for the Examination of Water and Wastewater, 14th ea., pp. 131-139. American Society of Civil Engineers, American Water Works Association, Conference of State Sanitary Engineers, 1969, Water Treatment Plant Design, p. 122. American Water Works Association, Inc., New York. Black and Veatch, Consulting Eng~neers. 1975. Direct Filtration of Lake Superior Water For Asbestiform Fiber Removal. Report No. EPA-670/2-75-0500, EPA, National Environmental Research Center, Cincinnati, Ohio. Black, A.P., and S.A. Hannah. 1961. Electrophoretic studies of turbidity removal by coagulation with aluminum sulfate. J. Am. Water Works Assoc. 53:438-452. Black, A.P., and D.G. Willems. 1961. Electrophoretic studies of coagulation for removal of organic color. J. Am. Water Works Assoc. 53:589-604. Hall, E.S., and R.F. Packham. 1965. Coagulation of organic color with hydrolyzing coagulants. J. Am. Water Works Assoc. 57:1149-1166. Jenkins, K.H. 1963. 1962 USPHS summary of municipal water facilities in communities of 25,000 or more. J. Am. Water Works Assoc. 55:1485-1492. Ling, J.T. 1955. A study of filtration through uniform sand filters. Proc. Am. Soc. Civit Eng., Sanitary Engineering Division, 81 : Paper no. 751. Logsdon, G.S., and J.M. Symons. 1975. Removal of asbestiform fibers by water filtration. Water Supply Research Laboratory, EPA, Cincinnati, Ohio. McCluney, W.R. 1975. Radiometry of water turbidity measurements. J. Water Pollut. Control Fed. 47:252-266. Packham, R.F. 1965. Some studies of the coagulation of dispersed clays with hydrolyz~ng salts. J. Colloid Sci. 20:81-92.  Picker~ng, R.J. 1976. Measurements of 'turbidity' and related characteristics of natural waters. U.S. Geol. Sur. Open-File Rep. 76-153. Robeck, G.G., N.A. Clarke, and K.A. Dostal. 1962. Effectiveness of water treatment processes in virus removal. J. Am. Water Works Assoc. 54:1275-1292.

204 DRINKING WATER AND H"LTH Rook, J.J. 1975. Formation of and occurrence of haloforms in dunking water. Presented at 95th Annual Conference of the American Water Works Asociation, June 8-13, Minneapolis, Minn. Stevens, A.A., C.J. Slocum, D.R. Seeger, and G.G. Robeck. 1975. Chlorination of organics in drinking water. Presented at Conference on the Environmental Impact of Water Chlorination, Oak Ridge Naional Laboratory, Oak Ridge, Tenn. Symons, J.M. 1975. Personal communication. Symons, J.M., and J.C. Hoff. 1975. Rationale for Turbidity Maximum Contaminant Level. Water Supply Research Division, U.S. Environmental Protection Agency, Cincinnati, Ohio. Yao, K.M., M.T. Habibian, and C.R. O'Melia. 1971. Water and wastewater filtration: Concepts and applications. Environ. Sci. Tech. 5: 1 105-1 1 12.