This appeal concerns a seizure proceeding under the Federal Food, Drug, and Cosmetic Act involving 449 cases of tomato paste allegedly “adulterated” within the meaning of § 402(a)(3) of that act, 21 U.S.C. § 342(a)(3). The government as libelant has sought — so *569far unsuccessfully — condemnation of the food in question upon a showing that it contained tissues rotted, but not necessarily deleterious to health.
The tomato paste, imported from Portugal, was landed in Brooklyn in the Eastern District on April 9, 1951, the entry being in bond. Representatives •of the Federal Security Agency took samples for inspection; and on April 16 claimant, A. Fantis, the importer, received official notice from the Food and Drug Administration that the goods need not be further held. Claimant thereupon paid for the shipment and removed and sold fifty cases from the entire lot. In July, however, a government food inspector, checking the warehouse, noticed that several of the cases had been recoopered; and closer inspection revealed that several cans had been re-soldered. This discovery led to a retesting of the shipment, which disclosed the presence of mold in the tomato paste in ■quantities exceeding administrative tolerances. The instant proceedings ensued.
It is undisputed that mold in tomato products indicates decomposition. It is also undisputed that when, as here, it results from rot in the tomatoes present before processing, it is not visible to the naked eye, but is detectable only by microscopic examination. Libelant at the trial did not offer proof that the paste was deleterious or unfit for food, in any way other than the decomposition, but contended that it was no part •of the government’s case to go beyond the showing made as to decomposition. Thereafter the district judge filed an opinion holding that the government had not sustained its burden of proof and that the shipment should be released to the claimant. D.C.E.D.N.Y., 111 F.Supp. 478. Libelant appeals from the resulting order.
Section 402(a)(3) provides that a food shall be deemed to be “adulterated,” and hence subject to condemnation under § 304(a), 21 U.S.C. § 334 (a), upon shipment in interstate eommerce: “if it consists in whole or in part of any filthy, putrid, or decomposed substance, or if it is otherwise unfit for food.” The district court apparently read “unfit for food” as limiting the entire section by virtue of the word “otherwise,” and as requiring a showing that the product was deleterious. It is this construction which presents the issue on appeal. The point is novel in this circuit, though it has been decided by several other courts which have uniformly held that the government need not prove unfitness for food other than filth or decomposition. Bruce’s Juices, Inc., v. United States, 5 Cir., 194 F.2d 935; Salamonie Packing Co. v. United States, 8 Cir., 165 F.2d 205, certiorari denied 333 U.S. 863, 68 S.Ct. 744, 92 L.Ed. 1142; United States v. 1851 Cartons Labeled in Part H. & G. Famous Booth Sea Foods Whiting Frosted Fish, 10 Cir., 146 F.2d 760; United States v. 935 Cases, More or Less, Each Containing 6 No. 10 Cans of Tomato Puree, D.C.N.D.Ohio, 65 F.Supp. 503; United States v. 44 Cases, etc., Viviano Spaghetti with Cheese, D.C.E.D.Ill., 101 F.Supp. 658, 663; see also United States v. Lazere, D.C.N.D.Iowa, 56 F.Supp. 730; United States v. 184 Barrels Dried Whole Eggs, D.C.E.D.Wis., 53 F.Supp. 652; and the monographic comment, Developments in the Law — The Federal Food, Drug, and Cosmetic Act, 67 Harv.L.Rev. 632, 644. This unanimity of view is itself impressive; moreover, we think the conclusion it represents is required both by the statutory language and by the history and general pattern of the legislation.
The entire subject matter of this subdivision of the statute is covered by two co-ordinate “if” clauses; and the second “if” indicates plainly that the second clause introduced thereby is co-ordinate and independent, rather than a qualification of the antecedent clause. The first clause expressly bans all products composed in whole or in part of any filthy, putrid, or decomposed substance; and the second clause goes on to add to the ban substances which were unfit for food for any other reason.
*570Furthermore, the other subdivisions of § 402(a) make specific' reference to products which are “poisonous,” “deleterious,” “injurious to health,” or “the product of 'a diseased animal.” These provisions cover those cases where danger to health is direct and demonstrable. The specific listed characteristics are clearly essential elements to be proved in actions under those provisions which refer to them'. But in the first clause of § 402(a)(3) the sweeping ban. of products consisting in whole or in part of any decomposed substance without reference to their effect on health is not made to depend on any such additional, props or findings to support the ultimate conclusion requiring the ban. It. may well be that, in the judgment of the legislators^ the presence of any substantial amount of rot in any food product- was a sign of danger sufficiently pointed to justify and require the exclusion of the -product from unrestricted circulation in interstate commerce. Or we may accept an acute suggestion of Judge Maris in United States v. 133 Cases of Tomato Paste, D.C.E.D.Pa., 22 F.Supp. 515, 516, that this section “was designed to protect the aesthetic tastes and sensibilities of, the consuming public,” and that the presence of such material in food, whether “perceptible by the consumer” or not, would offend both. For present decision it matters not which rationale is preferred, since in either event congressional power is clear and is not now challenged.
It should be noted also that the further class of adulterated foods thus added, i. e., “otherwise unfit for food,” is a broad general classification allowing “the widest variety of reasons for condemning a food,” 67 Harv.L.Rev. 632, 645, and not limited to either proof of filth or decomposition or to conditions deleterious to health. See, e. g., United States v. 24 Cases, More or Less, D.C.Me., 87 F, Supp. 826, 827, (canned herring roe of a “ ‘tough, rubbery consistency’”), and cf. United States v. 298 Cases, etc., Ski Slide Brand Asparagus, D.C.Or., 88 F.Supp. 450 (dealing with “stringy asparagus”) ; and Steffy, “Otherwise Unfit for Food” — A New Concept in Food Adulteration, 4 Food Drug Cosmetic L. Q. 552 (1949) (citing a variety of examples). There is therefore no basis for equating unfitness for food with injury to health, and the assumed logical progression from decomposition to unfitness for food to injury to health as showing identic terms thus doubly fails.
The conclusion of these authorities, following the statutory language, that the phrase “unfit for food” is not constrictive, but rather is additional or cumulative, is of controlling importance here. Any attempt to develop a constrictive meaning runs into the difficulty— highlighted by the statutory history developed below — that there is literally no place to which the argument may lead. As appears, complete identification of this phrase with “injurious to health” is universally excluded. But it is manifestly impossible to work out some ter-tium quid, of content sufficient to be grasped and acted upon by government inspectors or courts, of matter which is worse than “filthy, putrid, or decomposed,” but still less than injurious to health. At most, search for such an intermediate ground can only suggest something by way of a greater degree of filthiness or putridity, perhaps along the line of the government tolerance actually allowed, or, if not this, something hopelessly vague arid variable, dependent upon the taste buds or olefactory senses of the inspectors and too shifty a basis to serve as embodiment of congressional intent of drastic prohibition with both civil and criminal sanctions. So it is not. surprising, that no precedent or authority actually supports such a classification. And so the argument for restrictive interpretation of the statute slips insensibly, although perhaps necessarily, into an identification of food unfitness with health injury.' This becomes even more manifest upon an exairiination of the statutory history to which we now turn.
The particular definition of adulterated articles here involved — § 402(a) (3) supra — goes back to the original Federal *571Food and Drugs Act of June 30, 1906, which in § 7, 21 U.S.C. § 8, provided that an article “shall be deemed to be adulterated * * Sixth. If it consists in whole or in part of a filthy, decomposed, or putrid animal or vegetable substance, or any portion of an animal unfit for food, whether manufactured or not, or if it is the product of a diseased animal, or one that has died otherwise than by slaughter.” It will be noted that this limits the phrase “unfit for food” to animal matter and unmistakably separates it from the reference to “filthy, decomposed, or putrid animal or vegetable substance.” And so the uniform construction of the early act afforded “decomposed” an unrestricted meaning, requiring no proof of injury to health, e. g., A. O. Andersen & Co. v. United States, 9 Cir., 284 F. 542; United States v. 133 Cases of Tomato Paste, supra, D.C.E.D. Pa., 22 F.Supp. 515; Knapp v. Callaway, D.C.S.D.N.Y., 52 F.2d 476; United States v. Two Hundred Cases, More or Less, of Canned Salmon, D.C.S.D.Tex., 289 F. 157; United States v. Krumm, D.C.E.D.Pa., 269 F. 848; United States v. Two Hundred Cases of Adulterated Tomato Catsup, D.C.Or., 211 F. 780.1
On the same day that Congress enacted the Food and Drugs Act, 34 Stat. 768, it enacted the parallel and complementary Meat Inspection Act of 1906, 34 Stat. 674 — a temporary measure which was permanently re-enacted in identical language in 1907, 34 Stat. 1260. The provisions for inspection of meat in this companion act continuously use the language “unfit for human food,” sometimes with “otherwise,” 21 U.S.C. §§ 71, 72, 76, 92, and sometimes without, 21 U.S.C. §§ 74, 89. The full phrase, as later used in § 402(a) (3) supra of the present act, therefore goes back to this time; and the several usages in the various sections, taken separately or cumulatively, and in their respective settings, are only rationally to be interpreted as adding an additional — not a delimited— class of banned products. When the phrase appears presently in § 402(a) (3), it clearly should have the same significance.2
*572Thus Congress, when it enacted the present provisions, was aware of the broad construction placed on the earlier wording, as shown by the uniform course of case precedent supra; and the Act of 1938 here follows the earlier act so closely that an intent to restrict the provisions in material respect is not to be inferred. Sen.Rep.No. 361, March 13, 1935, on S. 5, Cal. 375, 74th Cong., 1st Sess., states: “* * * the provisions of Section 301(2), (3) and (5) [later incorporated into 21 U.S.C. § 342(a)] dealing with filthy food and food from diseased animals are essentially the same as those of the present law.”3 Moreover, amplification, rather than restriction, of power, is indicated elsewhere in the 1938 Act, as in the definition of adulterated drugs and devices in 21 U. S.C. § 351(a), which separates even by numbered clause and semicolon those consisting (1) of “any filthy, putrid, or decomposed substance” from those, (2) and (3), which are “injurious to health.”4 So, viewing the key phrase here in the light of history and the overall pattern of the legislative intent, we are clear that it must receive the same construction when applied to tomato paste as when applied to meat or drugs or oleomargarine or butter. It follows that we find correct, and agree with, the-modern current of authority cited at the-beginning of this opinion.
It is of course true, as is often pointed, out, e.g., 67 Harv.L.Rev. 632, at 644, 696, that the power granted is very broad, and “literal application of the statute could lead to unjustified harshness.” But Congress has attempted to meet this difficulty by granting a large measure of discretion: to the administrator, originally the Secretary of Agriculture, later the Federal Security Administrator, and now the Secretary of Health, Education and Welfare. In addition to provisions not here immediately pertinent for regulations making certain exemptions or granting certain: tolerances, 21 U.S.C. §§ 345, 346, there-is a significant provision in the chapter-authorizing penalties, injunctions, and seizures that nothing therein “shall be construed as requiring the Secretary to report for prosecution, or for the institution of libel or injunction proceedings, minor violations of this chapter whenever he believes that the public interest will be adequately served by a suitable written notice or warning.” 21 U.S.C. § 336. Obviously the Congress considered such administrative control a wiser course than the hedging of power by various theoretical restrictions, the negativ*573ing of which might be difficult of proof, in a particular case. And its wisdom is: indicated in this very case, where the product is prepared for and sold in quantity distribution, in cans of “Net Contents 10 Lbs. About” of concentrated paste, thus indicating distribution to restaurants and institutions, where customers and inmates cannot easily, if at all, protest the serving of rotten tomato paste, unlike ordinary retail sales, where housewives do have some possible chance of protecting themselves against unwholesome products by buying first-grade articles at top prices.
Hence the district court was in error in its construction of the governing statute. There was further error in the holding that the government had a “very much heavier” burden of proof here than usual because of its first analysis which resulted in original clearance of the goods. D.C.E.D.N.Y., 111 F.Supp. 478, 480. We have held that in this class of cases the government has no extraordinary burden, but only the usual one of proof by a fair preponderance of the evidence. United States v. 5 Cases, More or Less, Containing “Figlia Mia Brand,” 2 Cir., 179 F.2d 519, 524, certiorari denied Five Cases of Figlia Mia Brand of Oil v. United States, 339 U.S. 963, 70 S.Ct. 997, 94 L.Ed. 1372. Even if there were any basis for applying some rule of estoppel against government agencies in their enforcement of regulatory laws for the general welfare, there is nothing either in the grant of administrative favor or the then less compelling cast of the evidence — discussed below — to change this rule of law.
The order denying condemnation and releasing the goods to the claimant must therefore be reversed. And we think the state of the record is such that we should order judgment for the government, rather than remand for further trial. True, the district judge because of his rulings of law did not go on to make specific findings as to the degree of decomposition of the product. But the evidence was reasonably complete and upon it, taking it in favorable aspects to the claimant and even upon the testimony of his own expert, we think it clear, as a matter of law, that the product was adulterated within the statutory meaning. For a demonstration of this, some further analysis of the record is necessary.
It was agreed by both parties that the test to be employed was the Howard Mold Count as approved by the Association of Official Agricultural Chemists (Book of Methods of the Association of Official Agricultural Chemists, 7th Ed., § 35.64, p. 723). By this test a small amount of the duly mixed tomato paste is put upon a particular slide, the Howard mold count slide, for microscopic examination, the chemist in charge examining 25 fields of view on each slide and 4 slides or 106 fields (a minimum of 50 fields being required under the Howard count). The percentage number of microscopic fields in which mold filaments appear is therefore noted and deductions made as to the existence of mold signifying the use of tomatoes with rotted tissue. The government chemists worked with a tolerance of 40 per cent; that is, as one of the experts succinctly put it, “that the Administrator would not take action on products of this sort where the mould count did not exceed 40 per cent.” 5 It will be noticed that under this test there is no direct gauge or measure of the amount of rot in particular samples, but rather of its prevalence throughout the various fields observed. As the expert testimony showed, under this method variations in result as among the different fields observed would be normal; the experts were looking for the number of fields showing mold filaments according to their prescribed standards, not the amount of mold in each — a point the district judge seems to have overlooked in stressing what he thought to be unusual discrepancy revealed as between the government *574chemists who testified. The tests reported seem to us therefore significant of the presence of rot in the product.
There was a total of 40 mold counts made on 30 cans of tomato paste from the seized shipment. Of those, 24 exceeded the administrative tolerance, as did the average of all mold counts, namely, 43.-6%. The mold counts made by 4 government analysts on 18 different cans varied from a low of 8 to a high of 70, with the last average by Analyst Eisenberg being 56. Claimant’s own expert found an average mold count, as he testified, of 39.-1% ; on a post-seizure sample of 10 cans he found that one-half of his counts exceeded the administrative tolerance. Further breakdown of the figures in the record is set forth in the footnote.6 The claimant’s expert himself answered “Yes” to the question: “On the basis of your counts would you draw the conclusion that there was rotten tissue in this tomato paste?” As indicated in the note, the sampling seems fair and to a considerable extent controlled by the claimant, particularly as to the final tests.
It is therefore clear that there was decomposed matter in all the samples, more substantially evidenced in some than in others, but enough to bring the average well over the administrative tolerance. We think violation of the statute therefore demonstrated, and the goods subject to forfeiture. If we accept the government’s allowance of 40 per cent before prosecution is had, nevertheless the showing- here is adequate to bring a very sub*575stantial portion of the shipment (undifferentiated from the entire mass) above that amount. It is doubtful, however, how far we may accept that tolerance. Here there is no definite provision for administrative setting of a tolerance, as under 21 U.S.C. § 346 for example. Though the Secretary under 21 U.S.C. § 336 would be justified in withholding prosecution, yet, when had, there would seem no authority for us to waive statutory violation perhaps beyond the principle of de minimis, as suggested in 67 Harv.L.Rev. 632, 645 supra. The question here involved is raised in United States v. 935 Cases, More or Less, Each Containing 6 No. 10 Cans of Tomato Puree, supra, D.C.N.D.Ohio, 65 F.Supp. 503, 505, where Judge Jones refers to the discretion given the Administrator not to report or prosecute minor violations and to make regulations of exemption or tolerance under 21 U.S.C. §§ 345, 346, but adds: “No such provision for regulation making exemptions, or for tolerating unavoidable ingredients is provided with respect to Section 342(a) (3).” This is the only case of the group cited above dealing with the present legislation (several of which were tried to the jury) where the question of a tolerance is even mentioned. But it accords with the view taken earlier, as in A. O. Andersen & Co. v. United States, 9 Cir., 284 P. 542, 545, quoted in note 1 supra.
Under either view, therefore, we think the shipment subject to condemnation. The order is accordingly reversed and the action remanded for the entry of a decree of condemnation.
. In examining these cases it is necessary to boar in mind the distinction adverted to in the text between “unfit for food” and “injurious to health” and the inveterate tendency to slip into a complete identification of the two. What these cases are deciding is that the latter is unnecessary for the statutory violation, and occasional references to the former, separated from context, are imprecise and obviously not intended as definition. A good example is A. O. Andersen & Co. v. United States, 9 Cir., 284 F. 542, 544, where, in reaciiing its clear-cut holding, the court did quote from other cases and itself refer to the product affirmatively within the prohibition as that which was “ ‘so far decomposed as to be unfit for food’.” But this is description, not definition or restriction. Thus the court went on to hold that “while a small percentage of adulteration, found only in a small percentage of the product, might not and would not ordinarily satisfy the court or jury that the whole product is adulterated, yet in a case like this, where the jury might properly infer or find that approximately one-fifth of the entire product was unfit for human consumption, and that the adulteration extended to the entire product, no such question can arise.” 284 F. at page 545. Thus as to four-fifths of the product, there was still adulteration, albeit no proof of unfitness for food. So in none of these cases is attempt made to solve the dilemma which would have been presented by the converse ruling as to wliat could be a food unfitness more than putridity, filth, or decomposition and less than injurious to health.
. It is suggested that since, as in 21 U.S. C. § 72, the phrase “or otherwise unfit for human food” follows the words “unsound, unhealthful, unwholesome,” it is restricted in meaning to deleterious to health. But this will not do for several reasons. First, it does violence to the language used. “Unhealthful” may perhaps suggest the urged meaning, but surely neither “unsound” nor “unwholesomo” carries any such connotation. This is made doubly clear by the other usages throughout these sections. These are too numerous to quote in detail, but they support the inclusive meaning ascribed by the precedents. Thus 21 U.S.C. | 72 has several variations, e.g., “sound, healthful, wholesome, and fit for human food”;' “unsound, unhealthful, unwholesome, or otherwise unfit for hu*572man food”; “unsound, unhealthful, unwholesome, or in any way unfit for human food”; and a final duplication of the second phrase quoted; and 21 U.S.C. § 74 ties the phrases to meat containing dyes or preservatives, etc. Second, since meat products were not exempted from the Food and Drugs Act until 1938, 21 U.S.C. § 392, inspection and condemnation under, e.g., § 72 supra must mesh with and parallel the definition of adulteration under the former § 7, 21 U. S.C. § 8, supra. And third, the attempted interpretation runs into the cul-de-sac referred to above of complete identification of unfitness for food with injury to health.
. The report of the House Committee on Interstate and Foreign Commerce on S. 5, H.R. 2139, 75th Cong., 3d Sess., Apr. 14, 1938, declares that the bill “amplifies and strengthens the provisions to safeguard the public health.” The report makes no reference to the change in § 402 (a) (3), 21 U.S.C. § 342(a) (3), under discussion, but, in treating of the section as a whole, discusses several major extensions of the definition of “adulterated” not here relevant. There is no suggestion of any intent to restrict the powers of the administrator.
. In the most recent legislation on the subject, that of 1959 removing previous shackles on the sale of oleomargarine, the same pattern appears. Thus Congress added a new subdivision, (e), to this definition of “adulterated” food, § 402, 21 U.S.C. § 342, viz.: “If it is oleomargarine or margarine or butter and any of the raw material used therein consisted in whole or in part of any filthy, putrid, or decomposed substance, or such oleomargarine or margarine or butter is otherwise unfit for food.” The complete separation of the last clause from the earlier part of the provision is thus clearly manifest.
. He went on to say: “However, a late announcement, in 1951, I think, indicated that other factors would be taken into consideration besides mould count in tbe seizing of goods.”
. The following table shows the results of all the tests from the initial to the final ones:
This table, taken from appellant’s brief and found to be correct on a check of the record and briefs, is slightly more inclusive and accurate than that set forth at 111 F.Supp. 479. It contains additionally the figures used by Analyst Sly, as well as those by claimant headed “Import Sample”; and it contains two variations, a substitution of 27 for 25 in one of the Krinitz counts, and of 42 for 30 in one of the Ohiano counts, changing the averages for those counts from 43% and 38.5% to 43.2% and 39.7% as here correctly set forth.
The claimant actually controlled the sampling for the three final tests, since the government in every case selected a can next that chosen by claimant. The government in its samples ran two tests on each can, with the results indicated.