This is an appeal by the United States from the judgments in the combined prosecution of four libels (designated in the District Court as 54-C-1754, 54-C-1820, 54-C-1833 and 55-C-70) condemning approximately 10,370 cases of tomato paste as 'adulterated' within the meaning of 21 U.S.C.A. § 334(a). The paste was canned by the Smith Canning Company in Clearfield, Utah, and shipped to Chicago where it was seized by the United States Marshal.

The libels in actions 54-C-1754 and 55-C-70 alleged that the paste seized thereunder was 'adulterated,' in that it had been prepared under 'insanitary conditions' as defined by 21 U.S.C.A. § 342(a)(4). The libels, as amended, filed in cases 54-C-1820 and 54-C-1833 charged that the paste involved in those cases was adulterated because it had been prepared under insanitary conditions whereby it may have become contaminated with filth, Section 342(a)(4); and because it consisted wholly or in part of a 'Filthy, putrid, or decomposed substance' as defined by 21 U.S.C.A. § 342(a)(3): decomposed tomato material in number 54-C-1833, and decomposed tomato material and insect parts in number 54-C-1820.

After an extensive hearing, the record of which comprises some 1700 pages, the trial judge found the issues against the Government with regard to all but a small amount of tomato paste seized in case number 54-C-1820. With the exception of that small amount, the seized tomato paste was ordered released to the claimant owner.

The Federal Food, Drug and Cosmetic Act, as amended, 21 U.S.C.A. § 301 et seq., provides that any article of food in interstate commerce that is 'adulterated or misbranded' may be seized and condemned, 21 U.S.C.A. § 334(a). The Act provides the following definition of 'adulterated':

'A food shall be deemed to be adulterated--

'(a) * * * (3) if it consists in whole or in part of any filthy, putrid, or decomposed substance, or if it is otherwise unfit for food; or (4) if it has been prepared, packed, or held under insanitary conditions whereby it may become contaminated with filth, or whereby it may have been rendered injurious to health * * *.' 21 U.S.C.A. § 342.

Despite the plain language of the section it has been generally held that the two 'if' clauses in subsection (3) above are disjunctive, and that the words 'otherwise unfit for food' do not limit the first part of the subsection which bans food in whole or in part filthy, etc., as adulterated. United States v. 449 Cases, etc., 2 Cir., 212 F.2d 567, 45 A.L.R. 2d 846; 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; United States v. 44 Cases, etc., D.C., 101 F. Supp. 658; United States v. 935 Cases, etc., D.C., 65 F. Supp. 503. However, there have been dissenting voices. See Judge Frank's dissenting opinion in United States v. 449 Cases, etc., 2 Cir., 212 F.2d at page 575.

We find it impossible to agree with the accepted interpretation of Section 342(a)(3), 21 U.S.C.A., without ignoring completely the word 'otherwise' therein. The majority opinion in United States v. 449 Cases, etc., supra (212 F.2d 569) seems to argue that this is just what should be done, because the Meat Inspection Act of 1907 continually uses the phrase 'unfit for human food' sometimes with 'otherwise' and sometimes without it. It has also been suggested that Congress wanted to protect 'the aesthetic tastes and sensibilities of the consuming public,' and therefore intended that food containing 'any filthy, putrid, or decomposed substance' be deemed adulterated whether it was 'unfit for food' or not. United States v. 133 Cases of Tomato Paste, D.C., 22 F. Supp. 515, 516. Congress may also have wanted to set a standard or purity well above what was required for the health of the consuming public, knowing that not every food product can be individually inspected. If the standard is set at the level of what is 'fit for food' or not injurious to health, the occasional substandard item that slips by both industry and Government scrutiny will be hazardous to the health of the consumer. A minimum standard of purity above what is actually the level of danger will, however, allow fewer products to drop below that level. A high standard will also have the same effect by encouraging more careful industry inspection. Therefore, we prefer to follow the general rule in interpreting Section 342(a)(3), although admitting that we are unable to answer Judge Frank as to why Congress put the word 'otherwise' in the section.

The interpretation we have chosen has one serious disadvantage which most courts have recognized. It sets a standard that if strictly enforced, would ban all processed food from interstate commerce. A scientist with a microscope could find filthy, putrid, and decomposed substances in almost any canned food we eat. (The substances which it is claimed render the respondent 'adulterated' were visible only through a microscope.) The conclusion is inescapable that if we are to follow the majority of the decisions which have interpreted 21 U.S.C.A. § 342(a)(3), without imposing some limitation, the Pure Food and Drug Administration would be at liberty to seize this or any other food it chose to seize. And there could be no effective judicial review except perhaps for fraud, collusion, or some such dishonest procedure. Such a position is not indefensible. Congress has obviously found it difficult, if not impossible, to express a definite statutory standard of purity that will receive uniform interpretation. And this court is acutely aware of the fact that it is not the proper body to more narrowly define broad standards in this area so that they can be applied in a particular case. Courts know neither what is necessary for the health of the consuming public nor what can reasonably be expected from the canning industry. Furthermore, this is not a determination that should be made individually for each case on the basis of expert testimony. The Food and Drug Administration should set definite standards in each industry which, if reasonable, and in line with expressed Congressional intent, would have the force of law.

Despite our limitations as a court and the fact that Section 342(a)(3), 21 U.S.C.A., does not give us any power to limit the inescapable force of the words, 'if it consists in whole or in part of any filthy, putrid, or decomposed substance,' we do not think that Congress intended to let the acts of the agency under this subsection go completely without limitation. In Section 346, 21 U.S.C.A., Congress directed that the administrator provide tolerances for amounts of poisonous or deleterious substances that cannot be avoided and are not injurious to health. It would not be reasonable to think that Congress would direct the administrator to set tolerances for the allowance of safe amounts of poisons in food and then declare that the presence of small amounts of filth, etc., which would admittedly have no effect upon health 'adulterates' food and justifies its seizure. We believe that if the fact that almost all food contains some filthy, putrid, and decomposed substances had been called to the attention of Congress, that body would have directed the administrator to provide reasonable and acceptable tolerances for these substances just as it did in the case of poisons.

The spirit of 21 U.S.C.A. §§ 346 and 346a demands that we give effect to what reasonable standards have been set by the Food and Drug Administration in the area involved in this case, and determine them as best we can where they have not yet been established. The decomposed tomato material which the respondent is accused of containing is commonly referred to as rot. A tomato containing rot is simply a tomato parts of which have begun to decompose. This is not at all uncommon and such fruits are perfectly good if all of the decomposed portions can be cut out. Several different things cause tomatoes to decompose but by far the most common cause is mold. Because of this fact there has been developed a technique known as the Howard Mold Count which purports to estimate the amount of decomposition or rotting that has gone on in the tomatoes from which paste or juice has been made by measuring the amount of mold present in the finished product. The presence of mold proves that some rotting has taken place, but the absence of mold does not prove that no rotting has occurred. Add to this the fact that the Howard Mold Count system is susceptible to a fair degree of error in determining the amount of mold in a substance, and we have a rather inadequate method of measuring the amount of decomposition. Nevertheless, it seems to be the only practical method known at the present time, and has received the approval of both the Government and the food industry.1 The Food and Drug Administrator with industry cooperation has arrived at a tolerance for tomato paste which is expressed as 40 per cent under the Howard Mold Count method of measurement. The Administration has announced that it will not seize tomato paste on the basis of mold count alone unless that count is over 40 per cent. We, in our search for standards in this area, accept this administrative tolerance as a proper measure of what approximated amount of decomposition is allowable in tomato paste. A properly obtained mold count of over 40 per cent will, therefore, be considered sufficient grounds for seizing tomato paste if the Food and Drug Administrator chooses to do so.

The record in this case does not disclose any established tolerances for what is termed 'filth' in tomato paste: worm fragments, insects and insect fragments, fly eggs, etc. We can only judge on the basis of the testimony of experts as to what amounts are usual or unavoidable.

Section 342(a)(4), 21 U.S.C.A., has seldom been interpreted by a federal court. We have found only four cases dealing with it directly: Berger v. United States, 8 Cir., 200 F.2d 818; Triangle Candy Co. v. United States, 9 Cir., 144 F.2d 195, 155 A.L.R. 903; United States v. Roma Macaroni Factory, D.C., 75 F. Supp. 663; United States v. Lazere, D.C., 56 F. Supp. 730. Section 342(a)(4) provides that food is adulterated if it is 'packed, or held under insanitary conditions whereby it may have become contaminated with filth.' Whether or not a given factory is insanitary under this subsection is, of course, a question of fact, but the standard is so expressed, perhaps unavoidably, that the decision is likely to be highly subjective. Therefore, when we are dealing, as here, with products that, admittedly, will not affect the public health or sensitivities, we have a natural tendency to equate the standard with the average condition of canneries throughout the country. If the Federal Food and Drug Administration desires to improve that average, it would be more likely to receive the support of the courts if it promulgated regulations which provided detailed standards as to cleaning procedures, screens, hygiene facilities, etc., publishing them to food packers as the requisites for complying with 21 U.S.C.A. § 342(a)(4), and then seizing food packed in plants not meeting the specific standards set. So much for the law which we must apply to the facts of this case. All of the Government's arguments on appeal, with one exception, are that the trial court's findings of fact are clearly erroneous.

The District Court concluded that the Government had not proved by a preponderance of the evidence that the respondent had been prepared under insanitary conditions whereby it may have been contaminated by filth. The Government introduced considerable evidence (both pictures and testimony of Government inspectors) to show that the area surrounding the canning factory, and particularly a migratory labor camp maintained by the company, was in a dirty, 'insanitary' condition. The record contains pictures of piles of trash, pools of water, and dirty, inoperative rest rooms in and around the labor camp. Testimony for the claimant disclosed that wet areas of ground were covered with gravel as soon as possible after the first inspection, rest rooms were repaired as soon as they broke down, garbage in the labor camp was collected twice a week before the first inspection and every day thereafter, and that a local exterminator sprayed the entire area for insects regularly as he himself deemed it necessary. The trial court found that:

'It appeared that the labor camp had been located on the said 13 1/2 acres for the past 14 years, and was for the housing of field workers only, and did not serve as living quarters for any of the employees working in the plant.' Finding of Fact number 17.

In the same Finding the court said, 'the labor camp was too far removed from the cannery to have any possible effect upon the quality or purity of the merchandise being processed therein.' And the Government in its brief contends that 'the sole question on this phase of the case is whether the undeniably filthy conditions may be dismissed as 'inconsequential' because they were 'outside the processing room' or 'remote from the holding vats' * * *.'

The effect to be given the distance of seemingly filthy conditions from the place where the food is 'prepared, packed or held' is itself a question of fact to be determined in the light of the total picture.

Testimony varied slightly on how far distant the labor camp was from the canning plant, but the trial court found it was between 150 and 200 feet at the closest point. The Government's principal argument in its claim that dirty conditions in the labor camp were causing the tomato paste to be contaminated with filth, was that the camp attracted large numbers of flies which carried bacteria, etc., to the factory. Yet Mr. Alvord, Utah State Food Inspector, testified that when he was in the Smith Canning Company plant on September 16, 1954, the fly problem was 'nil.' He said that this cannery was better than most in this regard (pages 1430-1 of the record). The Government had the burden of convincing the trial court that the labor camp was close enough to the canning factory to effect the alleged 'conditions' under which the respondent tomato paste was 'prepared, packed, or held.' The court found that the Government had not sustained that burden, and the evidence is not so one-sided that we can say that the decision was clearly erroneous.

The Government also attempted to prove that insanitary conditions existed inside the plant itself. Evidence on the condition of the plant was based upon inspections made by federal inspectors on September 16 and 17, September 21, and October 15. They testified that during their inspection in September they noticed unscreened openings into the sorting room with many flies present, and dried tomato material on machinery. This testimony was opposed by testimony of State Inspector Alvord that the fly problem at this plant was nil, and testimony by Richard Smith (R. 1777) and Melvin Wood that the machinery was thoroughly cleaned every day with water, detergents, live steam and brushes where necessary (R. 1571-3). We might not have found the facts as the trial judge did, but with testimony to support his finding we cannot say that it is clearly erroneous. The pictures, which the Government claims make it possible for us to decide certain facts as well as the trial judge, do not show whether the deposits on the machinery are due to weeks of accumulation or just a few hours run.

The testimony of the Government inspectors about the activities of the women on the sorting line is best answered by our discussion (infra) of the microscopic analyses that were done on the finished product. Sections 342(a) (3) and (4) are closely related, but evidence as to the sorting and trimming of the tomatoes going into the paste, seems more material to purity of product (subsection (3)) than to cleanliness of the plant (subsection (4)).

When the inspectors visited the plant on October 15, they found animal excreta on the second floor and a bird's nest in the rafters of a corner of the vat room. However, the court found (and it is not challenged) that:'No merchandise involved in this seizure action was packed after October 9. The plant had completed its regular commercial operations on October 12, and was closed down on October 13 and 14. On October 15 it reopened, at the request of some of claimant's growers and local farmers, to run through for them some odd lots of tomatoes, and for local consumption.' Finding of Fact number 12, R. 2103.

It is clear that none of the paste involved in the seizure was packed on October 15, or thereafter. Therefore, there is no evidence that the things seen by the inspectors only on October 15, were in the plant when the respondent paste was 'prepared, packed, or held.'

We cannot say that the trial court was clearly wrong in finding that the Government did not prove by a preponderance of the evidence that the respondent paste was 'adulterated' within the meaning given in 21 U.S.C.A. § 342(a)(4).

In the cases designated in the District Court as 54-C-1833 and 54-C-1820 the Government alleged that the paste seized therein was also adulterated under the definition in 21 U.S.C.A. § 342(a)(3), in that it consisted in part of a 'filthy, putrid, or decomposed substance * * *.' We have already discussed the manner in which we are going to apply this section. The four codes (a code consists of all the cans filled in a certain period of time) which the District Court ordered seized all had average counts over 40 per cent of positive fields, and the judge was therefore correct in finding that they contained 'mold in excess of what is permitted to be shipped in interstate commerce.' Yet, as the Government points out, there are other codes that also have an average mold count above 40 per cent. This court holds that as a matter of law all tomato paste having a mold count (or an average mold count where several valid counts are taken) of over 40 per cent of positive fields found, is adulterated under 21 U.S.C.A. § 342(a)(3). The record shows that all the codes involved in this proceeding which were canned in October 1955, and bear the code letter 'J', have an average mold count above 40 per cent. The Government should be allowed to seize these codes. All of the codes canned in September, bearing the code letter 'I,' have an average mold count of less than 40 per cent, and therefore cannot be seized on that ground.

The claimant complains that the Howard Mold Count system has a large margin of error and that subsequent counts might be well below the tolerance. We acknowledge that this is true, but point out that by the same reasoning they might be above the counts introduced in evidence. Of course, allowances can always be made, but in this case, since there is nothing to indicate that the tests were more likely to be in error in one direction than the other, it seems best to adhere to the accepted standard of 40 per cent. The Government and the canning industry must have taken into consideration the margin of error inherent in the Howard Mold Count system when they set 40 per cent as the tolerance. Any deviation from that figure on our part would be purely arbitrary without any evidence that error was more likely in one direction than the other.

The old maxim that the law cares not for small things which the Government thinks was the principle the trial court used in releasing some of the codes with average mold counts over 40 per cent (Government brief, p. 54) is not here applicable. The tolerance is admittedly a somewhat arbitrary standard, but one that has been agreed upon by all the parties involved. The line must be drawn somewhere, and it has been validly drawn at 40 per cent. Forty-one per cent is not just a slight amount of mold, it is a slight amount over a standard that already has allowed for a large margin of error. A definite line must be drawn, and we will apply the one that has been approved by the industry and the Government.

Various microscopic examinations were also done for insect and worm fragments. Since the 'J' codes will be seized because of their mold counts, we need further consider only the 'I' codes. And of the 'I' codes only those seized in case 54-C-1820, because that is the only case in which the libel alleged that the paste contained insect fragments.2 Although there do not seem to be any acceptable norms for insect and worm fragments, we can get a good idea of how the 'I' codes here compare with other tomato paste from the testimony of Emil Cassidy, a research chemist with the American Can Company. Mr. Cassidy testified that one corn ear worm in a tomato might be responsible for 150-200 worm fragments in the finished paste, and an equal number of seta (the small hairs by which the worms propel themselves). The witness was then shown Government exhibit number 102 which is the report of a microscopic examination of 20 codes all bearing the code letter 'I.' He was asked whether he thought the report on each code in turn indicated that it was a good or a bad pack. The first code had a worm fragment count of 4, and the witness said he considered it a good pack. When asked why, he testified:

'A. Well you have got very little mold count there; your worm fragments are very low. I don't know what the tolerance is, but there isn't any tolerance that I know of.

'This is low compared to a lot of them I have seen.' (Record, p. 1658.)

The highest worm fragment count shown on exhibit 102 is 6. Insect fragments were in the same general range, and the other examinations made do not show worm and insects fragments in 'I' codes appreciably different from those in exhibit 102. The record fully supports the trial court's finding that these counts 'were so low that they are regarded by this Court as insignificant and of no consequence.'

The same is true of other foreign bodies found (fly eggs, etc.). The rodent hairs found in two of the codes in exhibit 102 seemed to be rare, and none of the witnesses knew how to evaluate their presence. But a very small number were found and in only a few of the codes. There is nothing in the record to indicate that the trial court was wrong in finding their presence inconsequential.

For the reasons discussed above, we hold that the codes canned in September and identified by the code letter 'I' are not adulterated, and as to them the judgments below are affirmed. As we said above, we hold that all 'J' codes are adulterated, so the portion of the judgment holding that four 'J' codes are adulterated is affirmed, and those portions of the judgments holding that the rest of the 'J' codes are not adulterated are reversed. The cause is remanded for proceedings in conformity with this opinion.

Affirmed in part, reversed in part and remanded.