DEPARTMENT OF HEALTH AND HUMAN SERVICES Food And Drug Administration 21 CFR Parts 103, 129, 165, and 184 [Docket No. 88P-0030| RIN 0910-AA11 BEVERAGES: BOTTLED WATER AGENCY: Food and Drug Administration, HHS. ACTION: Final rule. SUMMARY: The Food and Drug Administration (FDA) is establishing a standard of identity for bottled water. At the same time, the agency is recodifying the standard of quality for bottled water. FDA is revising the definition for bottled water in the quality standard to include mineral water and ingredient uses of this product. In addition, FDA is defining "artesian water," "ground water," "mineral water," "purified water," "sparkling bottled water," "spring water," "sterile water," and "well water." FDA is exempting mineral water from certain physical and chemical allowable levels. FDA is taking these actions, in part, in response to a petition submitted by the International Bottled Water Association (IBWA). FDA finds that the regulations will promote honesty and fair dealing in the interest of consumers as well as the interests of the regulated industry. EFFECTIVE DATE: (Insert date 180 days after date of publication in the FEDERAL REGISTER.) The Director of the Office of the FEDERAL REGISTER approves the incorporations by reference in accordance with 5 U.S.C. 552(a) and 1 CFR part 51 of certain publications at 21 CFR 129.35(a)(3)(ii), 129.80(g), and 184.1563(c), effective 180 days after date of publication in the FEDERAL REGISTER. FOR FURTHER INFORMATION CONTACT: Shellee A. Davis, Center for Food Safety and Applied Nutrition (HFS-306), Food and Drug Administration, 200 C St. SW., Washington, DC 20204, 202-205-4681. SUPPLEMENTARY INFORMATION: I. BACKGROUND In the FEDERAL REGISTER of January 5, 1993 (58 FR 393), FDA published a proposal to establish a standard of identity in  165.110(a) for bottled water (hereinafter referred to as the January 1993 proposal). At the same time, the agency proposed to recodify in  165.110(b), (c), and (d), the standard of quality for bottled water currently found in  103.35. FDA proposed to revise the definition for bottled water in the quality standard to include mineral water and ingredient uses of this product. In addition, FDA proposed to define "artesian water," "distilled water," "mineral water," "purified water," "spring water," and "well water." FDA proposed to exempt mineral water from certain physical and chemical allowable levels if the mineral water contained total dissolved solids (TDS) in excess of 500 parts per million (ppm). Interested persons were given until March 8, 1993, to submit comments. In the FEDERAL REGISTER of March 9, 1993 (58 FR 13041), FDA extended the comment period to April 7, 1993. In addition, the agency reopened the comment period for comments concerning two spring water surveys that FDA received in response to the proposal (58 FR 34010, June 23, 1993). Interested persons were given until July 23, 1993, to submit comments concerning the two spring water surveys. FDA received approximately 430 responses, each of which contained one or more comments, from trade and retail associations, government organizations, manufacturers, consumers, health care professionals, retailers, consumer groups, State groups, private organizations, the U.S. Congress, professional societies, and universities. The comments generally supported the proposal. Several comments addressed issues outside the scope of the proposal (e.g., microbiological quality standards, definitions for multicomponent bottled water beverages) that will not be discussed here. A number of comments suggested modifications and revisions in various provisions of the proposal. A summary of the suggested changes and the agency's responses follow. Elsewhere in this issue of the FEDERAL REGISTER, FDA is proposing to update the methods referenced in  165.110(b)(3) and to exempt mineral water from the allowable level for aluminum in the quality standard. FDA is responding to the comments on the January 1993 proposal that addressed those issues in that proposal. II. THE STANDARD OF IDENTITY A. Coverage The agency proposed in the January 1993 proposal, to move the definition for bottled water from the quality standard to the standard of identity and to revise the definition to include mineral water and ingredient uses of bottled water. Specifically, FDA proposed that bottled water be defined as water that is intended for human consumption and that is sealed in bottles or other containers with no added ingredients, except that it may contain safe and suitable antimicrobial agents. The agency also proposed that bottled water may be used as an ingredient in beverages (e.g., diluted juices, flavored bottled water) but stated that the term did not cover those food ingredients that are declared in ingredient labeling as "water," "carbonated water," "disinfected water," "filtered water," "seltzer water," "soda water," and "tonic water." Finally, FDA proposed that the processing and bottling of bottled water must comply with applicable regulations in part 129 (21 CFR part 129). 1. One comment asked why bottled water is singled out for a source identification requirement (e.g., water from a municipal source), and why soft drinks, beers, reconstituted juices, salad dressings, and other products that contain water as an ingredient are not also subject to this requirement. The agency considered the scope of the bottled water standard, particularly its application to water used as an ingredient in multicomponent foods such as flavored waters and diluted juices, in the proposal to this final rule (58 FR 393 at 395). FDA stated that highlighting the water component of these products is effectively a claim that the water ingredient in the beverage has particular value, and that consumers are likely to purchase these products in large measure because of the claim concerning the water ingredient. For example, in a significant number of situations, the labeling of products stated or implied that the water originated from a source such as a spring or a well. In contrast, in products such as soft drinks or reconstituted juices in which water is simply used as an ingredient, no claim is made about the water. The intent of the proposal was not to require source labeling of all water ingredients from a municipal source, but to require it in the former type of situation, where the finished product is bottled water or the labeling makes an explicit or implied claim concerning the water ingredient. Under section 403(a) of the Federal Food, Drug, and Cosmetic Act (the act)(21 U.S.C. 343(a)), a product is misbranded if its labeling is false or misleading in any particular. To determine whether the absence of information on food labels constitutes misbranding the agency must take into account the extent to which the labeling fails to reveal facts material in light of representations made or suggested with respect to consequences that may result from the use of the article under customary or usual conditions of use (section 201(n) of the act (21 U.S.C. 321(n))). The agency does not deem source labeling of ingredient water from a municipal source as a material fact unless labeling representations are made or suggested that the water possesses particular properties. Thus, the agency finds no reason to include water that is not a highlighted ingredient in the bottled water standards, and the comment has not provided a basis to do so. Therefore,  165.110 applies only to bottled water and ingredient uses of water where the water ingredient is highlighted in the labeling. 2. Two comments expressed concern about sparkling water being regulated under the bottled water standards. They stated that sparkling water has long been understood by consumers and recognized by FDA as a common or usual name for unsweetened and unflavored water containing compressed carbon dioxide. The comments contended that consumers clearly understand "sparkling water" to be in the same category as "carbonated water," "seltzer water," "soda water," and "tonic water," substances that historically have been regulated by FDA as "soft drinks." They stated that as such, "sparkling water" is more appropriately classified as a "soft drink" that is not subject to the proposed bottled water standard. Some comments stated that it was unclear whether "sparkling water" was included under the exemption for "carbonated water" and asked for clarification. One comment stated that if the standard does not encompass the term, FDA should include "sparkling water" in the definition of bottled water. FDA stated in the preamble to the proposed rule that when a beverage is labeled as containing "water," "carbonated water," "disinfected water," "filtered water," "seltzer water," "soda water," and "tonic water," there is no claim that the water ingredient has particular value, and that, thus, these ingredients were not included in the definition of bottled water (58 FR 393 at 395). This proposed exclusion did not extend to the term "sparkling water" or to any other term not specifically excluded by the standard. In the proposal, the agency used the example of the term "sparkling" as indicating that the water ingredient possessed a specific characteristic or had received a specific treatment (id.). FDA tentatively concluded at that time that use of such highlighted terms concerning the water component was effectively a claim that the water ingredient in the beverage had particular value, and that consumers were likely to purchase these beverages in large measure because of the claim. The agency notes, however, that sparkling water was included in the former soda water standard. In the FEDERAL REGISTER of February 16, 1967 (32 FR 2940), the agency amended the soda water standard to add the term "sparkling water" to the standard as an example of a type of soda water generally designated by a particular common name. FDA proposed this change to permit the designation of nonsweetened and nonflavored soda water by names other than those prescribed in the standard (31 FR 11109, August 20, 1966). However, FDA repealed the standard of identity for soda water in the FEDERAL REGISTER of January 6, 1989 (54 FR 398) because some provisions of the standard were being adequately dealt with by other regulations, while other provisions were no longer necessary. Given the traditional use of this term, as evidenced by the repealed standard, the agency agrees that the term "sparkling water" describes a nonsweetened and nonflavored carbonated water, and that it thus, is a term that is synonymous with the term "carbonated water." The fact is that the agency had separate standards at one time for bottled water and for soda water, and that it included sparkling water in the soda water, and not the bottled water standard. Therefore, the comments have persuaded the agency that some types of sparkling water are in the same category as "carbonated water," "seltzer water," "soda water," and "tonic water," and should be regulated as a soft drink instead of as a bottled water. Accordingly, the bottled water standard in  165.110 does not include those food ingredients that are declared in ingredient labeling as "water," "carbonated water," "disinfected water," "filtered water," "seltzer water," "soda water," "sparkling water" (except as explained in this response) and "tonic water." However, the term "sparkling water" may also refer to water that is naturally carbonated (i.e., contains carbon dioxide as it emerges from the source) and that is labeled as "sparkling water." The bottled water standard has traditionally included this type of water because the water has not been carbonated in the same sense that a soda water is carbonated (i.e., with added carbon dioxide). Thus, water that contains carbon dioxide as extracted from the source is not a soda water and must continue to be regulated as a bottled water. Therefore, to differentiate between the two types of sparkling waters, the agency is defining "sparkling bottled water" in  165.110(a)(2)(v) as bottled water that, after treatment and possible replacement of carbon dioxide, contains the same amount of carbon dioxide that it had at emergence from the source. This definition is in accordance with the definition in the European Regional Codex Standard (ERCS) for "naturally carbonated natural mineral water" (Ref. 1). The agency concludes that defining the term "sparkling bottled water" is within the scope of this rulemaking because FDA proposed that sparkling water be included under the bottled water standard. As a consequence of proposing that course of action, FDA has been persuaded that some types of sparkling water should be excluded from the standard and that it should differentiate among the types of sparkling water in its bottled water regulations. 3. Two comments objected to the exclusion of carbonated bottled waters from the bottled water standards. They stated that any product that professes to be, or that has as an important ingredient that is one of the defined bottled water types (e.g., spring water, mineral water), whether noncarbonated or carbonated, should be considered to be bottled water. The comments contended that only those carbonated products with respect to which no reference is made to defined bottled water types should be excluded. The agency agrees with the comment. Products or ingredients described by a term that is defined by the standard of identity (e.g., "spring water") or with a term that makes a claim about the water (e.g., "natural water") are standardized waters and must comply with  165.110 whether carbonation has been added or not. Although terms to describe the water ingredient in a product may sometimes be used in combination with a term that is not included under the standards (e.g., "carbonated spring water" or "filtered natural water"), the product or the water ingredient in the product must comply with the bottled water standards because a claim is being made concerning the value of the water. However, use of only a term specifically excluded from the bottled water standards (e.g., "filtered water" or "carbonated water") means that no claim is being made concerning the value of the water, and, thus, the water is not a standardized food. 4. Several comments stated that it is inconsistent for FDA to exempt carbonated waters from the bottled water standards. They held that carbonated water may be consumed at levels which constitute a major portion of an individual's daily water intake. One comment added that exempting carbonated water from the bottled water category does not provide for consumer safety or confidence. The agency does not agree with the comments, although it acknowledges that carbonated waters may constitute a major portion of some consumers' daily water intake. However, FDA points out that standards of identity and quality are not established because a product is consumed in large volumes or for consumer safety or confidence but are established to promote honesty and fair dealing in the interest of consumers. "Carbonated water" or "soda water" has traditionally not been covered by the provisions of the bottled water quality standard because it has been considered to be a soft drink. Bottled water and soda water, although similar, are two different foods as evidenced by the fact that the agency had quality standards for bottled water at the same time that it had a standard of identity for soda water that included "carbonated water." FDA tentatively concluded in the proposal that it would not include "carbonated water" in the standards for bottled water because it has historically not been considered to be bottled water. In addition, the agency tentatively concluded that the standards for bottled water covered water ingredients that were highlighted because of a claim concerning the water ingredient itself. Labels of foods that claim to contain as an ingredient, or to be, "carbonated water" do not claim or imply any particular properties or characteristics for the water ingredient. Any claims on such foods for the ingredient are simply that carbon dioxide has been added. Thus, "carbonated water" does not fit within the type of food that the bottled water standard is intended to address because no claim is being made about the water itself. The agency finds no reason to include "carbonated water" in the bottled water standard, and the comment has not provided any basis to do so. Therefore, the agency concludes that it has not been inconsistent in the regulation of "carbonated water" and "bottled water." 5. Two comments stated that the terms "carbonated water," "disinfected water," "filtered water," "seltzer water," "soda water," and "tonic water" should be defined to avoid confusion by industry and consumers as to what types of water are covered by the bottled water standards and what types of water are not. One of the comments stated that these terms may have different meanings to the bottled water and beverage industries and consumers. The agency does not agree with the comments. There is general understanding of the meanings of these terms, even though they are not defined in FDA's regulations. "Carbonated water," "disinfected water," "filtered water," "seltzer water," "soda water," and "tonic water" are common or usual names that are in general use by both manufacturers and consumers. In the proposal to this final rule (58 FR 393 at 395), the agency noted that the terms "carbonated water," "seltzer water," "soda water," and "tonic water" have historically been considered to be soft drinks, and that "disinfected water" and "filtered water" described water that has been subjected to a commonly used treatment. The reason for not addressing these terms under the bottled water standard is that they make no claims about the water used in the ingredients that they denominate. The exclusion is not based on the specific source, composition, or processing of these types of waters. The comments did not provide any information to persuade the agency to conclude otherwise. Therefore, FDA is not defining these common terms at this time. However, persons interested in establishing definitions for "carbonated water," "disinfected water," "filtered water," "seltzer water," "soda water," and "tonic water" may petition the agency to do so, providing recommended definitions and justification for the recommendations. 6. Several comments objected to any definition of bottled water that does not allow for the addition of ingredients such as minerals for flavor, flavors that comprise less than 1 percent by weight of the product, and carbon dioxide. They stated that changing the definition of bottled water to exclude established products would: (1) Be costly to the bottled water industry; (2) exempt excluded bottled water products from meeting FDA's proposed health, safety, and quality standards; and (3) confuse rather than unify regulatory authority. One comment declared that it is imperative that the final regulation include all established products of the bottled water industry. One comment stated that some model codes and State regulations provide for the addition of ingredients to bottled water provided that these additives comprise less than 1 percent by weight of the final product. The comment noted that many consumers may supply a major portion of their daily water intake needs with these 1 percent bottled water products. Another comment noted that  129.80(a) states that carbonation, mineral addition, or any other process shall be done in a manner so as to be effective in accomplishing its intended purpose and in accordance with section 409 of the act. It stated that there is no reason to disallow any of these processes or their resultant products as bottled water. FDA does not agree that it is changing the definition of bottled water. The agency acknowledges that some State regulations define bottled waters with added flavors, minerals, and carbon dioxide as described in the comments (Ref. 2). However, the Federal definition of "bottled water" has traditionally been "water that is sealed in bottles or other containers and that is intended for human consumption" ( 103.35(a)(1) and 129.3(b) (1994)). This definition is the basis of the definition that FDA is adopting in this final rule ( 165.110(a)(1)). Although  129.80(a) mentions treatment of product water with carbonation and mineral addition, among other processes, any bottled water product with added ingredients would be just that--bottled water with added ingredients. (However, see the discussion of fluoride in bottled water in comment 8 of this document.) Firms may manufacture nonstandardized bottled water products with ingredients such as minerals for flavor, flavors that comprise less than 1 percent by weight of the product, and carbon dioxide added to bottled water. The common or usual name of the resultant product must reflect these additions. However, only the bottled water ingredient is subject to the standard in  165.110. The other ingredients in the product are subject to regulation under the food additive or other food ingredient provisions of the act. Thus, it is not necessary to include added ingredients, such as minerals for flavor, flavors that comprise less than 1 percent by weight of the product, or carbon dioxide, in the standard for bottled water. Therefore, for the reasons listed above the agency is not persuaded by the comments to include the addition of minerals, flavors, and carbon dioxide in the standard of identity for bottled water in  165.110(a)(1). 7. One comment stated that all bottled waters imported into the United States should meet all of FDA's requirements for bottled waters including mineral water. It added that U.S. standards should not be lessened to meet European standards unless there are compelling public health reasons for doing so. FDA agrees that both foreign and domestic bottled waters sold in interstate commerce in the United States must comply with the act and the regulations issued thereunder, including the specific regulations for bottled water found in part 129 and  165.110. Although the agency attempts, where possible, to harmonize its regulations with the regulations of other countries, there must be appropriate grounds for FDA to amend any of its bottled water regulations, whether or not such action would harmonize international standards. B. Fluoridated Water The agency did not propose a definition for "fluoridated water," although it did request comments on the need to define types of bottled water other than those for which it proposed definitions. Some comments addressed issues on fluoridated water that fall within the scope of this rulemaking. Therefore, the agency is addressing these issues in this final rule. 8. Several comments stated that the addition of fluoride to bottled water should be allowed. One comment noted that many people specifically request and use fluoridated water because of its benefits to dental health. The comment stated that these products have long been established and should not be discontinued. One comment noted an inconsistency between the definition of bottled water and the provisions of proposed  165.110(b)(4)(ii) that discuss the addition of fluoride. FDA agrees that there is an inconsistency between the proposed standard of identity and the standard of quality for bottled water with respect to the addition of fluoride. The agency proposed in  165.110(a)(1) that bottled water contain no added ingredients other than safe and suitable antimicrobial agents. This provision would preclude the addition of fluoride (58 FR 393 at 407). The quality standard, however, provides maximum levels for bottled water to which fluoride is added ( 165.110(b)(4)(ii)). The agency recognizes that water with fluoride added may provide a benefit to consumers. The Surgeon General's Report on Nutrition and Health (the Surgeon General's report) recommends that community water systems contain fluoride at optimal levels for prevention of tooth decay, and that, if such water is not available, other appropriate sources of fluoride should be used (Ref. 3). Bottled water may be used by some consumers as an alternative to community drinking water. Therefore, because of the unique circumstances presented by fluoride, the agency is providing for the optional addition of fluoride to bottled water in  165.110(a)(1) within the limitations established in the quality standard ( 165.110(b)(4)(ii)). Because the agency is providing for the optional addition of fluoride to bottled water, the standardized product may be a multiingredient food, and, as such, its label must bear ingredient labeling. According to  101.4(a)(1) (21 CFR 101.4(a)(1)), all food ingredients are required to be declared on the label. Therefore, bottled water containing added fluoride must list the names of the water ingredient and the fluoride. Therefore, because FDA is providing for the optional addition of fluoride to bottled water, and thus, bottled water may be a multicomponent food, the agency is adding  165.110(a)(4) to require that each of the ingredients used in the food be declared on the label as required by the applicable sections of 21 CFR parts 101 and 130. 9. Several comments requested that FDA define "fluoridated water." Some of these comments recommended that "fluoridated water" be defined as bottled water containing naturally occurring or added fluoride, and that the label specify whether fluoride is naturally occurring or added. One comment stated that the regulation only discusses maximum limits on fluoride addition without setting a minimum, thus opening a loophole that would allow manufacturers to add insignificant amounts of fluoride to their products and call them "fluoridated water." One comment stated that any water that is called "fluoridated water" should contain not less than 0.8 milligram per liter (mg/L) fluoride ion. Another comment stated that any water that is defined as "fluoridated water" should contain not less than 1.0 mg/L fluoride. One comment requested that "fluoridated water" be defined only as water containing added fluoride. FDA has provided for the use of the terms "fluoridated," "fluoride added," and "with added fluoride" on the label or in labeling of bottled water that contains added fluoride in  101.13(q)(8). The agency adopted this regulation in the FEDERAL REGISTER of January 6, 1993 (58 FR 2302 at 2314) and stated that the presence of fluoride in bottled water is of interest to consumers, and its declaration should not be prohibited. However, the agency also stated that it did not wish to encourage unnecessary addition of fluoride to bottled water, and that it was concerned that if it permitted the use of terms like "good source of fluoride" or "high in fluoride," they might encourage such additions. Therefore, the agency has not defined a nutrient content claim for fluoride. Instead, it has provided that a statement indicating the presence of added fluoride could be used, but that the claim cannot include a description of the level of fluoride present. As stated in another final rule in the FEDERAL REGISTER of January 6, 1993 (58 FR 2079 at 2149), the agency has considered the identity statement "fluoridated water" to be misleading if the product is derived from a source naturally containing fluoride. Because the term "fluoridated" represents that fluoride has been added to the water, FDA concluded that the term "fluoridated water" should be used to describe only products to which fluoride has been added in the manufacturing process, and that such products would be required to bear nutrition labeling that complies with the simplified format (id.). FDA also points out that fluoride may not be present in amounts that exceed the limits in  165.110(b)(4)(ii). Although labeling could be used to describe whether fluoride was added or naturally present in bottled water, the term "fluoridated" continues to mean that fluoride has been added. FDA is not establishing a minimum level for fluoride addition because the terms "fluoridated," "fluoride added," or "with added fluoride" have been defined in  101.13(q)(8) and amending that provision falls outside the scope of this final rule. However, if the addition of fluoride to water is so minimal that it would be considered dietarily insignificant, a product that bears a claim about such addition would be misbranded under section 403(a) of the act in that its labeling would be misleading because the use of the term "fluoridated" or any of its synonyms implies that fluoride has been added in a meaningful amount. Thus, FDA concludes that it has not created a loophole that would allow manufacturers to add insignificant amounts of fluoride to their products and call them "fluoridated water." The Surgeon General's Report states that the optimal fluoride concentration of approximately 1 ppm fluoride has been shown to reduce the prevalence of dental caries by more than 50 percent (Ref. 3). In addition, the Surgeon General's Report states that current recommendations for optimum fluoride concentrations vary from 0.7 to 1.2 ppm (id.). Therefore, the comments' suggested values of 0.8 mg/L and 1.0 mg/L fluoride are meaningful amounts of this mineral in bottled water. 10. One comment stated that infant bottled waters may contain fluoride, and that the presence of this mineral may be a problem if parents are not aware that too much fluoride is undesirable, or that an infant should not receive both a physician's prescription of fluoride drops and drink water containing fluoride. For this reason, the comment stated that the label of a bottled water containing 0.3 ppm or more fluoride should include a statement advising parents not to use the product before consulting with their baby's physician if the baby is receiving a fluoride supplement. The comment added that bottled water for general use is also used for infants, so such a label statement should be required on any bottled water containing 0.3 ppm or more fluoride. It stated that this level of fluoride is taken from the current recommendation of pediatricians and pediatric dentists relating to administration of fluoride supplements. It suggested that the statement could read as follows: "Note: If you are giving your baby a fluoride supplement, do not use water with fluoride without consulting your doctor." The agency agrees that an advisory statement such as that suggested by the comment may be appropriate to prevent unwanted aesthetic effects from excessive doses of fluoride, and it encourages manufacturers to provide such information to consumers, especially on products labeled for infant use. However, FDA does not agree that this statement should be mandatory on all bottled waters containing 0.3 ppm or more fluoride. There are allowable levels for fluoride in the quality standard, and bottled water exceeding these levels must be labeled as substandard. The allowable levels are related to secondary levels established by the Environmental Protection Agency (EPA) for public drinking water in 40 CFR parts 141 and 143 and take into consideration excessive infant fluoride intake. In addition, as discussed in the previous comment, the Surgeon General's Report recommends an optimal level of 1.0 ppm fluoride in drinking water. Fluoride supplements are generally prescribed for breast-fed infants because those infants frequently consume little or no water. Human milk contains little fluoride, even in areas with fluoridated water supplies. Physicians may also prescribe fluoride supplements for infants not receiving adequate dietary fluoride. Health care professionals must take into consideration the patient's weight and the exposure to fluoride from dietary and other sources to establish the proper dose (Ref. 4). Therefore, the agency finds no basis to require an advisory statement concerning infant fluoride consumption on bottled waters containing 0.3 ppm or more fluoride. C. Nomenclature FDA proposed that the name of the standardized food meeting the definition of bottled water in  165.110(a)(1) is "bottled water" or one of the following defined terms: "Artesian water," "distilled water," "mineral water," "purified water," "spring water," and "well water." The agency requested comments from interested persons on the definitions for these terms and on other terms that need to be defined. 11. A number of comments requested that FDA define the term "drinking water" because: (1) It is the most commonly used term to describe bottled water and represents 36 to 40 percent of the gallonage of bottled water sold in food stores; (2) the lack of a Federal definition allows States to adopt special, nonuniform definitions for this segment of the bottled water market; and (3) many bottlers would have to revise their drinking water labels to remove this term to come into compliance with the standard, and doing so would impose severe economic hardships to the industry. One comment noted that producers of 5-gallon returnable bottled water products have a very large investment in bottle inventories that are designed to last for a considerable time, and that a high percentage of these packages is permanently labeled as "drinking water." It stated that it would be costly to dispose of these containers, and that the use of stick-on labels would present problems in its manufacturing operation. Some comments recommended modifying paragraph  165.110(a)(2) to define the terms "bottled water" and "drinking water" synonymously. However, other comments stated that "drinking water" is a classification within the bottled water category along with "spring water," "mineral water," and "purified water." These comments suggested the following definition: "Drinking water means bottled water obtained from an approved source that has at a minimum undergone treatment consisting of filtration (activated carbon or particulate) and ozonation or an equivalent disinfection process. Drinking water that has been treated to meet the definitions of distilled or purified water may contain added minerals for taste, provided an ingredient statement 'minerals added for taste' or optionally 'minerals added for flavor' appears on the label." One comment stated that it is important for FDA to define drinking water as only one type of bottled water, and that the terms "drinking water" and "bottled water" not be interchangeable. It stated that "bottled water" includes and describes all types of bottled water products, including bottled "drinking water," but that bottled "drinking water" does not include or describe all types of "bottled water." It stated that it is important that FDA define "drinking water" to prevent the consumer confusion that would result if this product type, already marketed to and accepted by the public, is not recognized by FDA as a specific type of bottled water. It stated that failure to do so could, at worst, mean that products labeled as "drinking water" could no longer be sold in interstate or foreign commerce involving the United States. Conversely, two comments stated that the term "drinking water" should not be permitted on the label because consumers may be misled because they do not understand the meaning of the term. The agency agrees with the comments that stated that it should define the term "drinking water." Consumers are familiar with the term because, as the comments pointed out, products labeled as "drinking water" comprise a significant portion of the bottled water market. In addition, not defining this term would impose an economic hardship on the bottled water industry because products labeled as "drinking water" would have to be relabeled as "bottled water." However, FDA disagrees with the comments that said that "drinking water" should be defined differently than "bottled water." As required by the standard of quality, "bottled water" must meet certain quality requirements, or the water is substandard and must be labeled as such. The definition for "drinking water" suggested by the comments provides an apt description of the method of processing bottled water that is used by many manufacturers. Thus, FDA concludes that a separate definition of "drinking water" is not warranted. In addition, EPA has standards for "drinking water" from public water systems (40 CFR parts 141 and 143) that are nearly identical to FDA's standards for bottled water. FDA is not aware of any reason why its standard for "drinking water" that is sold in a bottle should differ significantly from EPA's standard for "drinking water." Therefore, the agency is including "drinking water" as an alternative name for "bottled water" in  165.110(a)(2). The agency agrees with the comments that pointed out that if minerals are added to bottled water or drinking water, an appropriate statement of identity must appear on the principal display panel of the label of the product to inform consumers of this fact (e.g., "drinking water with minerals added for taste"). An ingredient statement must also appear on the label in accordance with  101.4(a). In addition, if sodium, calcium, or iron are present in the bottled water product in more than an insignificant amount, nutrition labeling is required. 12. One comment suggested that an alternative name for "spring water" or "well water" could be "ground water." The agency agrees that "ground water" is an appropriate name for water from a spring or a well. The term "ground water" encompasses not only "spring water" and "well water" but also "artesian water" and "mineral water" because by definition all of these waters come from an underground source. A geological definition states that "ground water" is water in the saturated zone that is under a pressure equal to or greater than atmospheric pressure (Ref. 5). The saturated zone is the subsurface zone in which all openings are full of water (id.). Because "ground water" is an appropriate alternative term to describe some types of bottled water, and because in the January 1993 proposal, the agency requested comments from interested persons on other terms that need to be defined, the agency concludes that it is within the scope of this rulemaking to define the term "ground water." FDA concludes that the geological definition stated above is appropriate. Therefore, the agency is defining "ground water" in  165.110(a)(2)(ii) to mean water from a subsurface saturated zone that is under a pressure equal to or greater than atmospheric pressure. (Because the agency is establishing an additional definition in  165.110(a)(2), it is recodifying the other terms in  165.110(a)(2) so that they continue to appear in alphabetical order.) The agency is also requiring in  165.110(a)(2)(ii) that "ground water" not be under the direct influence of surface water. EPA defines ground water under the direct influence of surface water as any water beneath the surface of the ground with: (1) Significant occurrence of insects or other macroorganisms, algae, or large-diameter pathogens such as Giardia lamblia; or (2) significant and relatively rapid shifts in water characteristics such as turbidity, temperature, conductivity, or pH which closely correlate to climatological or surface water conditions (40 CFR 141.2). Ground water under the direct influence of surface water is not "ground water" because water that does not meet this definition is mingling with water that otherwise would meet the definition. To clarify that ground water must not be under the direct influence of surface water, FDA concludes that this distinction should be included in the definition of "ground water." 13. One comment stated that FDA should establish a separate definition for "sterilized water." It stated that water for the initial feeding of babies has been called "sterilized water" for decades. The comment held that all water intended for the initial feeding of infants should be commercially sterile, as defined in the low-acid canned food processing regulations (21 CFR part 113). The comment stated that to require a change in the statement of identity from "sterilized water" to "bottled water" or "purified" or "distilled water" would create confusion in hospitals and could result in nonsterilized "bottled water" or "purified" or "distilled water" being fed to newborns. The comment suggested that the following definition be added to  165.110(a)(2)(vi): "The name of the water intended as the initial feeding of infants may be 'sterilized water' provided it meets the definition of commercial sterility contained in 21 CFR 113.3(e)(1)(i)." The agency agrees that the terms "sterile water" and "sterilized water" should be defined as a specific bottled water type. Doing so is the logical outgrowth of FDA's request in the proposal for other terms that need to be defined. Defining these terms will mean that the water must meet a certain minimum standard to be labeled with these terms and will allow firms to prominently label their products in the statements of identity as having been treated to achieve this standard. The definition of commercial sterility in  113.3(e)(1)(i) states that "commercial sterility" of thermally processed food means the condition achieved by the application of heat that renders the food free of microorganisms capable of reproducing in the food under normal nonrefrigerated conditions of storage and distribution and of viable microorganisms (including spores) of public health significance. FDA notes that the United States Pharmacopeia (USP) has official monographs for "sterile water for inhalation," "sterile water for injection," and "sterile water for irrigation." The monographs require that each of these types of water meet sterility requirements set forth by the USP (Ref. 6). These requirements involve microbiological tests to determine the presence of viable microorganisms. If no evidence of microbial growth is observed, the article tested meets the requirements of the test for sterility. The difference between the sterility standard in  113.3(e)(1)(i) and that used by USP is that water that is commercially sterile may contain bacteria, although in an innocuous amount, whereas water that is sterile according to USP standards does not. The agency acknowledges that water for general drinking purposes need not be sterile or even commercially sterile. However, sterile water may be purchased by immunosuppressed individuals, contact lens wearers, infant caretakers, and laboratories with an expectation that the water is free of any bacteria. In addition, FDA finds that it would be confusing to consumers to have two standards for sterility, one for "sterile water" and another for "sterile water, USP." Therefore, the agency concludes that bottled water labeled as "sterile" must meet the USP definition. Thus, FDA is defining the terms "sterile water" and "sterilized water" in  165.110(a)(2)(vii) to mean water that meets the requirements under "Sterility Tests" <71> in the USP, 23d Revision. 14. Several comments requested that the agency clarify whether a bottler may use any name defined in  165.110(a)(2) as long as the bottled water complies with the definition used. One comment asked whether mineral water that contains between 250 and 500 ppm TDS could be labeled as "mineral water," "mineral spring water," or "mineral well water." The agency advises that if more than one term is applicable, bottlers may combine the terms, as appropriate, in naming the food (e.g., "mineral spring water, low mineral content"). Thus, bottlers will be able to label their products in an informative manner because all applicable terms can be presented prominently on the labels in the statements of identity. Because this approach will ensure the informativeness of the statement of identity, FDA finds that it will promote honesty and fair dealing in the interest of consumers. Therefore, FDA is revising  165.110(a)(2) to state that the name of the food is "bottled water," "drinking water," or alternatively one or more of the terms listed in that section as appropriate. 15. Two comments expressed concern that the proposed definitions for "artesian water," "mineral water," "spring water," and "well water" provide an opportunity for unintended mineral content manipulation that could lead to potential consumer deception. To alleviate this problem, the comments requested that FDA revise the definitions for each of these waters to include the following: "Artesian water (or mineral water, spring water, or well water) shall not be altered by addition or deletion of minerals or by blending it with water from a nonartesian water source." FDA disagrees that the suggested revision is warranted or needed. Once a defined bottled water product (e.g., artesian water, mineral water, spring water, or well water) has been blended with water from another source, the product no longer meets the definition of that particular type of bottled water, although it remains bottled water. For example, if artesian water is blended with spring water to reduce the water hardness, the product is "bottled water" or "drinking water," although its labeling may state the percentages of the artesian water and spring water it contains. Mineral water may be labeled as "mineral water" even if it is a blend of one or more waters, as long as each of the component waters complies with the source, composition, and other requirements of  165.110(a)(2)(iii). The agency notes that mineral addition or deletion does not change the source of the water. However, if the water has been altered significantly from the source water, under section 201(n) of the act, that the alteration has been made is a fact material in the light of representations made and must appear on the label of the product. The water is no longer unmodified ground water and differs significantly from the water that was harvested. Therefore, the fact that the water has been altered significantly must be included in the statement of identity so that consumers are aware that the source water has been modified. If minerals have been added, the statement of identity must state that fact. If minerals have been removed from the product, other than those that are removed during normal processing (e.g., filtration to remove precipitates), that fact must be included in the statement of identity of the product (e.g., partially demineralized) ( 165.110(a)(2)(iii)). Therefore, FDA concludes that the requested revision is not necessary. 1. Artesian Water The agency proposed to define "artesian water" as water from a well tapping a confined aquifer in which the water level stands above the natural water table. The agency also proposed to provide for the collection of artesian water with the assistance of an external force to enhance the natural underground pressure so long as such measures do not alter the physical properties, composition, and quality of the water. 16. One comment stated that FDA should not permit the use of the term "artesian" on bottled water labels because it is the most misused term in the bottled water business today. The agency disagrees that it should prohibit the use of the name "artesian." Because FDA is defining this term in the standard of identity for bottled water, manufacturers will have to label their products in accordance with the standard or face regulatory action. FDA expects that misuse of the term will cease as a result. Therefore, FDA concludes that this comment, rather than establishing why FDA should not define "artesian water," only serves to point up why defining this term will promote honesty and fair dealing in the interest of consumers, and, thus, why it is appropriate for FDA to do so. 17. Several comments stated that the original and vernacular meaning of "artesian water" is water that is forced from below the ground to the surface through a well by natural underground pressure and collected at or above the surface. They recommended that this definition be adopted. One comment pointed out that the geologic definition that FDA referenced in the proposal actually states that "the water level in artesian wells stands at some height above the top of the aquifer but not necessarily above the land surface" and does not require that the water stand above the water table. Therefore, the comment added, the water level in an artesian well may be either above or below the water table and still be considered artesian. The comment stated that the distinction in the geologic definition between the water table and the top of the confined aquifer is an important technical one, and that the proposed definition is much more restrictive and not the one that is generally accepted by groundwater scientists. The agency disagrees with the comments that contended that the water in an artesian well must flow to the surface. As mentioned by the latter comment described above, the geologic definition states that "the water in artesian wells stands at some height above the top of the aquifer but not necessarily above the land surface" (Ref. 5). Therefore, the geologic definition does not require that the water flow to the surface, or that, as FDA proposed (58 FR 393 at 398), the water level stand above the natural water table. Because the agency intended that its definition for "artesian water" be the geologic definition, it is revising the definition of artesian water in  165.110(a)(2)(i) to state that bottled water that is drawn from a well tapping a confined aquifer in which the water level stands at some height above the top of the aquifer may be called "artesian water" or alternatively "artesian well water." Concerning artesian water that flows to the surface, FDA notes that a typical geologic definition states that "if the water level in an artesian well stands above the land surface, the well is a flowing artesian well" (Ref. 5). The agency would not object to manufacturers labeling their products accordingly, as long as it is done in a truthful and nonmisleading manner. However, the name of the food remains "artesian water" or "artesian well water." 18. One comment urged that the specific name "artesian well water" be permitted on labels instead of "artesian water" to provide full disclosure to consumers. FDA advises that both "artesian well water" and "artesian water" can be used to identify this product because both terms appropriately describe it, and consumers would recognize either term. "Artesian water" does indeed come from a well and only differs from "well water" in that the water comes from a confined aquifer where the water is under pressure and stands at some height above the top of the aquifer. Therefore, FDA is modifying  165.110(a)(2)(i) to state that the name of water from a well tapping a confined aquifer in which the water level stands at some height above the top of the aquifer may be "artesian water" or "artesian well water." 19. One comment asked how someone who is reviewing the label statement "artesian well water" will be able to verify that the well is actually an artesian well, meeting the definition, after the well has been bored and is in production. The agency agrees that there must be some means of verifying food labeling claims. In specific instances FDA may require that records or other means of verification be provided to FDA regulatory officials, despite the act's lack of express, general statutory records access authority for foods. The Supreme Court has recognized that FDA has authority that "is implicit in the regulatory scheme, not spelled out in haec verba" in the statute. Weinberger v. Bentex Pharmaceuticals, Inc., 412 U.S. 645, 653 (1973). Indeed, "it is a fundamental principle of administrative law that the powers of an administrative agency are not limited to those expressly granted by the statutes, but include, also, all of the powers that may fairly be implied therefrom. * * * In the construction of a grant of powers, it is a general principle of law that where the end is required the appropriate means are given and that every grant of power carries with it the use of necessary and lawful means for its effective execution." (See Morrow v. Clayton, 326 F.2d 35, 44 (10th Cir. 1963)). Under section 701(a) of the act (21 U.S.C. 371(a)), the agency may promulgate regulations for the efficient enforcement of the act. Although it is possible to determine that a source of water is an artesian well after the well is in operation, in some cases it would be onerous for regulatory officials to do so. Therefore, FDA has determined that a verification requirement is necessary for the efficient enforcement of the act. FDA has previously stated that a food manufacturer is responsible for the accuracy of its food labels (58 FR 2079, 2163, and 2165, January 6, 1993). Indeed, placing a claim in food labeling that calls the consumer's attention to a water's source is a representation that the manufacturer has evidence that the product meets the requirements for the claim. See Thompson Medical Co., Inc. v. FTC, 791 F.d. 189, 193 (D.C. Cir. 1986), cert. denied, 479 U.S. 1086 (1987). Making a claim without such a basis would be misleading, in violation of section 403(a) of the act. The agency anticipates that, in some instances, companies will be amenable to demonstrating to FDA the basis for the claim, regardless of the existence of these regulations. The agency considers, however, that, when a product bears a claim based on information available solely to the manufacturer, it is reasonable for the agency to have access to that information. See United States v. An Article of Device, 731 F.d. 1253, 1261-62 (7th Cir. 1984) (upholding regulation requiring makers of prescription devices to be able to prove that their devices work safely for their intended purposes and stating that "[w|here the government's access to the necessary information may be limited * * * it seems not inappropriate to put the burden of persuasion on the party who * * * presumably has better access to the relevant information"); see also Trans-American Van Service, Inc. v. United States, 421 F. Supp. 308, 331 (N.D. Tex. 1976). Therefore, FDA is modifying  165.110(a)(2)(i) to require that plants be able to demonstrate to regulatory officials that the water level of the well stands at some height above the top of the aquifer, and, thus, that the well is an artesian well. Compliance with this provision does not entail the creation of any new information or the compilation of any special records. Rather, the requirement would obligate manufacturers simply to have access to information that they should already possess, or to make a measurement of their well, and to provide FDA with this information. FDA considers this requirement to be the logical outgrowth of its January 1993 proposal. The purpose of this rulemaking is to ensure that terms such as "artesian water" are used in a manner that promotes honesty and fair dealing (see section 401 of the act (21 U.S.C. 341)). Such a result requires not only that these terms be appropriately defined, but that they be used in a manner that accurately describes the product. Thus, how FDA would enforce the definitions was a matter that was within the purview of the proposal, and that the agency would provide for such enforcement in the final rule in this proceeding was reasonably foreshadowed by the proposal. The fact that FDA received comments on how it would ensure that the defined terms are appropriately used evidences that this issue is the logical outgrowth of the proposal. To comply with this requirement, producers may maintain records that demonstrate that the well is indeed an artesian well. The manufacturer may also rely on records from the company that drilled the well. In addition, many States and the United States Geological Survey have record s of some wells and of the geology of the surrounding area. To verify that the water is at some height above the top of the aquifer and is, thus, artesian water, the pump may be shut off, and the height or the pressure of the water in the drilled hole measured. This information can then be used, along with information on the depth of the aquifer, to determine whether the water is artesian water. If the source does not meet the definition of artesian water, the product must not be labeled as artesian water, or it is misbranded under sections 403(a), 403(b), and 403(g) of the act. 20. One comment stated that water chemistry changes as wells are pumped, and that the larger the drawdown, the greater the water chemistry may change. It stated that a mineralogical analysis from a water sample taken at 10 gallons per minute (gpm) may be quite different than one taken at 500 gpm for the same well. The comment added that there would be an ongoing burden on FDA to verify that water produced by bottlers drawing on "artesian" groundwater resources remains constant in water chemistry. FDA agrees with the comment. The use of external force may alter the physical properties, composition, and quality of the water, although usually not significantly, depending on the rate of extraction, because of changes in the pressure of the water as it is extracted. This fact is the basis on which the agency proposed to require that the use of external force not alter these characteristics (58 FR 393 at 398). However, because the rate of extraction from the use of external force could vary from day to day or even hour to hour, the characteristics of the water can be also altered. As discussed in the previous comment, the agency is requiring that the manufacturer demonstrate that the source of the water is indeed an artesian source. However, the agency does not deem it necessary to require that the definition for artesian water extend to the physical properties, composition, and quality of the water. In fact, as long as the source is demonstrated to be an artesian source that meets the definition in  165.110(a)(2)(i), the actual characteristics of the water are a quality control concern for the manufacturer rather than an attribute that defines artesian water. As stated previously, the agency intends that its definition of artesian water be the geological definition. The geological definition does not take into consideration the composition of the water. Therefore, the agency is removing the requirement in  165.110(a)(2)(i) that the use of external force in the extraction of artesian water not alter the physical properties, composition, and quality of the water. 2. Mineral Water The agency proposed that water coming from a source tapped at one or more bore holes or springs, originating from a geologically and physically protected underground water source, may be called "mineral water." FDA further proposed that mineral water be distinguished from other types of water by its constant level of minerals and trace elements at the point it emerges from the source. The agency tentatively concluded in the proposal that it would be contradictory for bottled water that has essentially no minerals and does not perform (e.g., taste) like mineral water to be labeled as mineral water. Consistent with this tentative conclusion, FDA proposed that "mineral water" be defined as water containing not less than 250 ppm TDS. The agency requested comments on the proposed minimum level of 250 ppm TDS in mineral water and stated that if it received substantive data to support another minimum level, it would consider issuing a final rule with a different minimum level. 21. A number of comments objected to FDA establishing a minimum TDS level for mineral water. The comments argued that establishing such a level would be arbitrary and contrary to the most current mineral water definitions, including international definitions which do not include a minimum level, and would prohibit many brands from being labeled as mineral water, thereby adversely affecting consumer sales with no apparent benefit to consumers. Conversely, other comments suggested raising the minimum TDS level to 500 ppm. These comments argued that establishing the minimum TDS at the higher level would: (1) Make it closer to the definition that has been adopted by most States that have bottled water regulations and to the definition that is currently being considered by the Canadian government; (2) provide the basis for identification of the term "mineral water" with the distinctive taste of a higher mineral content; (3) be less confusing to consumers in that it would not allow the same water to be marketed under several names (e.g., "mineral water, low mineral content" or "spring water"); and (4) simplify application of the quality standards and the label identity statement. Several comments expressed the view that mineral water is ground water with at least 50 ppm TDS, while several other comments supported the proposed minimum level of 250 ppm TDS in mineral water. One of these comments stated that waters having a TDS of 250 ppm or more generally have a distinctive flavor. Two comments, however, stated that actual taste thresholds for mineral water are often in the range of 100 ppm TDS. One of these comments stated that consumer identification and differentiation among the flavors of various individual source waters, which result from naturally occurring mineral variations in the water, is clearly possible below the proposed threshold of 250 ppm TDS. The agency acknowledges that many different definitions exist for "mineral water." In the January 1993 proposal, the agency compared several current definitions, including State and European standards, in arriving at its proposed definition (58 FR 393 at 396). International standards vary from requiring at least 500 ppm TDS (Canadian Province of Quebec) to no minimum requirement (ERCS). In a like manner, different States have different requirements. Therefore, no FDA action will harmonize existing State and international regulations. As stated in the proposal (58 FR 393 at 397), the minimum level of 250 ppm TDS for mineral water is based on the apparent consumer expectation that a product identified as "mineral water" will contain at least a minimum level of minerals. The agency tentatively concluded that it would be misleading for bottled water that has essentially no minerals, and that does not perform (e.g., taste) like mineral water, to be labeled as mineral water. The minimum level that FDA proposed, 250 ppm TDS, is in agreement with the Association of Food and Drug Officials (AFDO) definition (Ref. 7) for light mineral water and mineral water. The main characteristic of mineral water is, as its name implies, the presence of a significant quantity of minerals. Other important characteristics (Ref. 8) are that it be from a geologically and physically protected underground water source, and that it contain a constant level of minerals and trace elements at its point of emergence from the source. Mineral water may come from a spring or a well, including an artesian well, but must contain a significant amount of minerals. The agency considers 250 ppm TDS as a significant amount of minerals because at this level, the minerals, depending on the specific mineral content, begin to impart a particular taste to the water. Although minerals may impart some taste below this level, it is not the significant mineral taste that is characteristic of mineral water. FDA recognizes that mineral water from a spring that contains between 250 and 500 ppm TDS may be identified as "spring water," "mineral water, low mineral content," or both. However, FDA disagrees that the availability of these terms will cause consumer confusion because such a product meets the definitions of both "mineral water" and "spring water" in  165.110(a)(2)(iii) and (a)(2)(vi), and both terms appropriately describe the product. As discussed previously (see comment 14 of this document), use of all applicable terms presented prominently on the label in the statement of identity is appropriate because it will ensure the informativeness of the statement of identity. FDA realizes that brands previously sold as "mineral water" that contain less than 250 ppm TDS will not be provided for under  165.110 as "mineral water." However, the brands mentioned in the comments are not being sold in the United States as mineral water but as other types of bottled water (e.g., spring water) because of the many State requirements that mineral water contain greater than 500 ppm TDS. Thus, although some of these brands cannot be labeled as "mineral water," other brands that previously could not be labeled as "mineral water" and sold in some States now meet the definition of "mineral water" in  165.110(a)(2)(iii) and can be labeled and sold as such. Moreover, the agency has not been persuaded that this regulation will adversely influence consumer sales or put some bottled mineral water producers at a disadvantage as compared to others. The comments did not provide any information on such adverse consequences, and the agency is not aware of these adverse effects. Therefore, for the reasons discussed above, FDA concludes that establishing a minimum level of TDS in mineral water is reasonable, and that the proposed level of 250 ppm TDS is the appropriate level. 22. One comment remarked that some bottled waters would not significantly differ from a mineral water slightly above the 250 ppm TDS minimum. Another comment stated that if companies wish to market their products on the basis of the mineral content, they can do so through a TDS disclosure statement on the principal display panel regardless of the amount present. Therefore, there would be no need to establish a minimum TDS level for mineral water. Several comments declared that each mineral water product is unique because of its particular composition of minerals, and that this unique character imparts distinctive flavor. These comments stated that some water products with a constant mineral content of less than 250 ppm TDS might have a distinctive flavor and should be called "mineral water--very low mineral content" or "mineral light." Comments also stated that establishing a level of 250 ppm TDS is contrary to the industry's belief that the overall mineral content is less important than the level of each particular mineral. The agency agrees that some waters that contain slightly less than the 250 ppm TDS minimum would not significantly differ from a mineral water slightly above the minimum. Any minimum level that the agency establishes will preclude some waters from bearing the term "mineral water" even though they do not vary significantly from waters that are above the minimum. FDA also agrees that there is a taste aspect to the presence of minerals, although some minerals may contribute a more distinct flavor to the water than others. For example, in sufficient amounts, sodium chloride gives water a salty taste, and sulfate contributes a bitter taste (Ref. 5). In fact, a common industry practice is to add minerals to some bottled waters for the flavor that they contribute. However, a minimum requirement for TDS is necessary to ensure that when consumers purchase a product labeled as "mineral water," the product contains minerals at a level that justifies calling the product by that name. As explained in response to the previous comment, that level is 250 ppm TDS. The minimum TDS requirement for mineral water will not preclude a product that contains less than 250 ppm TDS from being marketed under another name, such as "ground water," "spring water," or "well water," as applicable, or from being called "bottled water." In addition, manufacturers may include a truthful statement of the TDS level on the label of any bottled water product. Thus, even though FDA has carefully considered these comments, it concludes that it is appropriate to establish a minimum TDS requirement for "mineral water." 23. One comment stated that the definition for "mineral water" should include all water containing over 500 ppm TDS and argued that whether it occurs naturally or is constructed (as are other food products) is irrelevant. The comment added that if mineral solids are added, FDA should require that such additions be noted on the label. Another comment stated that it is essential that the definition be clear that mineral water may not be altered by the addition or deletion of minerals. The comment stated that mineral water should not be water that is derived from a public water supply and to which minerals are then added. FDA disagrees with the comment that advocated that water to which minerals have been added should be eligible to be called "mineral water." The agency has reviewed a number of State and foreign standards, and none define "mineral water" as containing added minerals (Ref. 8). In fact, many of these standards define water with added minerals as a different type of bottled water, distinct from "mineral water." Therefore, the agency concludes that the definition for "mineral water" should not be revised to permit the addition of minerals. The agency agrees with the comment that stated that the addition of minerals should be clearly prohibited in the definition for mineral water. The definition for mineral water has geological as well as compositional factors. The amounts and types of minerals in mineral water is a result of the path that the water has traveled underground. Therefore, to clarify that the minerals present in mineral water must be from the underground source and not added to the water after extraction, FDA is modifying the definition of mineral water in  165.110(a)(2)(iii) to specifically preclude the addition of minerals. However, some mineral waters as extracted from their geological source, may contain high levels of some undesirable minerals (e.g., arsenic, precipitated manganese). In some instances, the water can be treated to selectively remove these undesirable elements. FDA is aware of no reason why it should preclude the removal of these undesirable elements, or why such removal should preclude the manufacturer from labeling the product as "mineral water" as long as all other requirements (e.g., source and composition) of the definition are met. Section 165.110(a)(2)(iii) provides accordingly. 24. Several comments requested that FDA more precisely define "mineral water" in that the agency should require that the level and relative proportions of minerals and trace elements remain constant. Comments stated that such a requirement would harmonize the definition of mineral water with the European Community and Codex concepts. These comments suggested the following definition: "Mineral water shall be distinguished from other types of water by its constant level and relative proportions of minerals and trace elements, at the point of emergence from the source, due account being taken of the cycles of natural fluctuations." One comment added that this wording recognizes that minor natural fluctuations occur with any source water. Some comments requested that the agency clarify that, in the term "constant level of minerals," the "level" is not the characteristic element. They stated that what is fundamental is the "constancy" or "stability" of the mineral composition, which acts more as a fingerprint of the water rather than as a measure of the overall total dissolved solids content. One comment stated that all ground water (well or spring) has a constant level of minerals and trace elements as it emerges from the source. The comment questioned the scientific basis of FDA's approach. The agency agrees that it needs to clarify its definition of mineral water. In the proposal (58 FR 393 at 396), FDA stated that mineral waters may have very different flavors depending on the mineral content and types of minerals and trace elements present in the water. Consumers may purchase a particular mineral water from a particular source because of the flavor contributed by the mineral content. It is important to consumers that the mineral composition of a particular source remain constant. FDA considers that industry and consumers have come to expect that mineral water has a fairly stable mineral composition. Therefore, FDA proposed that mineral water be distinguished from other types of water by the constant level of minerals and trace elements in the water as it emerges from its source. FDA further notes that the ERCS defines "natural mineral water" as being characterized by its content of certain mineral salts and their relative proportions and by the presence of trace elements or other constituents (Ref. 1). The ERCS also states that mineral water is characterized by the constancy of its composition, the stability of its discharge, and its temperature, due account being taken of the cycles of natural fluctuations. As stated previously, the composition and concentration of substances dissolved in ground water depend on the chemical composition of precipitation, on the biologic and chemical reactions occurring on the land surface and in the soil zone, and on the mineral composition of the aquifers and confining beds through which the water moves (Ref. 5). Thus, under constant conditions, the mineral content of ground water will be constant. There are certain natural factors that may affect the constancy of a source such as occurrence of earthquakes and long term climatic changes. These natural factors do not preclude the water from qualifying as mineral water as long as the water continues to meet the compositional requirements in  165.110(a)(2)(iii). Therefore, to clarify the importance of the relative proportion of minerals and trace elements, and to take into account the cycles of natural fluctuations, FDA concludes that modification of the definition of mineral water, along the lines requested by the comments, is appropriate. The modification reflects the fact that there may be some minor variation in mineral water over time, and that absolute amounts of minerals in the water may change slightly. Thus, the agency is modifying  165.110(a)(2)(iii) to state that mineral water shal