This is an appeal by the United States from an order granting an injunction against some, but not all, of the claims of the appellee used in the interstate sale of her patent medicines, Tri-Wonda No. 1, Tri-Wonda No. 2, and Tri-Wonda No. 3.1
*851These medicines were sold by appellee for use by sufferers of arthritis, rheumatism and bursitis. The Government’s appeal results from the fact that the decree of the trial court enjoined the defendant from distributing the products in interstate commerce when misbranded by representing that they or any similar drug are “a cure or adequate treatment for any form of arthritis or rheumatism” but which decree expressly stated that “she could continue to introduce the drug into interstate commerce provided the labeling thereof was not false and misleading, permitting her to represent that the Tri-Wonda medicines, when taken according to directions, are beneficial in a substantial number of cases in the relief of some symptoms of rheumatoid arthritis, rheumatism and bursitis, such as pain, soreness, the swelling of tissues around the joints, the loss of freedom of motion resulting from pain and soreness accompanying rheumatoid arthritis, the loss of general well being, constipation, and the deficiency of Vitamin B-l associated with arthritis, rheumatism and bursitis * *
The United States complained of the permissive part of the order and the failure of the court to enjoin the representation that “the drugs were beneficial in a substantial number of cases in the relief of some symptoms of rheumatoid arthritis, rheumatism and bursitis such as pain, soreness, the swelling of tissues around the joints, and the loss of freedom of motion resulting from pain and soreness accompanying rheumatoid arthritis” on two grounds: (1) the acts enjoined by the court could not be distinguished from the acts permitted and thus “the decree contains inconsistencies and ambiguities which make it 'unenforceable,” and (2) such representations as are permitted by the Court are wholly unjustified by the evidence of record.
The trial court made explicit findings of fact, which included the following:
“ * * * that the evidence establishes that the Tri-Wonda medicines do not constitute a cure or remedy for any and all forms of rheumatism; and such representation is, therefore, false. * * * ”
Finding No. 29, to which the Government directs its attack, is as follows:
“29. The court finds that the Government has failed to meet the burden of proof and establish that the statements in some of the labeling used and distributed by the defendant, Lela S. Wier, the correspondence and the printed leaflets, are false and misleading, and has failed to prove that the Tri-Wonda medicines, when taken according to directions are not beneficial in a substantial number of cases in the treatment of some of the symptoms of arthritis, rheumatism and bursitis, including: pain, soreness, the swelling of tissues around the joints; the loss of freedom of motion resulting from pain, soreness and the swelling of tissues around the joints; and loss of general well being; constipation, and the deficiency of Vitamin B1 associated with arthritis, rheumatism and bursitis.”
Among the three questions presented by appellee in her brief is the following: “Is Finding of Fact No. 29 in the decision below contrary to the overwhelming evidence, so that it is completely erroneous?” In our view of the case this question must be answered in the affirmative, thus making unnecessary an answer to the first contention of the Government.
In approaching the problem as to the duty and power of the appellate court when called upon to review a finding of fact by the trial court, sitting without a jury, we start with the basic rule:
“Rule 52. FINDINGS BY THE COURT.
“(a) Effect. In all actions tried upon the facts without a jury or with an advisory jury, the court shall find the facts specially and state separately its conclusions of law thereon and direct the entry of the appropriate judgment; and in granting or refusing interlocutory injunctions *852the court shall similarly set forth the findings of fact and conclusions of law which constitute the grounds of its action. Requests for findings are not necessary for purposes of review. Findings of fact shall not be set aside unless clearly erroneous, and due regard shall be given to the opportunity of the trial court to judge of the credibility of the witnesses. * * * ” Fed.Rules Civ.Proc. rule 52, 28 U.S.C.A.
Although the trial court’s finding here was stated in terms of a failure of the Government “to meet the burden of proof and establish that the statements * * are false and misleading,” we consider this simply as a finding against the Government on the evidence.
It will at once appear to anyone dealing with an effort to prove that certain chemicals do or do not have a therapeutic effect on the human body, that the investigator, here the trial court, must do more than simply pass upon the credibility of the witnesses in ascertaining whether there is any evidence substantial enough to support the finding pro or con. For instance, we have held in United States v. Hoxsey Cancer Clinic, 5 Cir., 198 F.2d 273, 280, that the testimony of a layman either that he is suffering from cancer of that he has been cured of cancer, however honestly given and however firmly believed, does not rise to the dignity of substantial evidence. It follows that the same is true as to any disease whose presence or cure can be ascertained only by persons trained in medical science and by the use of scientific aids or surgery.
We think that what has been said as to the diagnosis of disease by a layman, even though he be a sufferer, applies with equal force to an opinion given by a sufferer that his relief from pain or relief from other symptoms of a disease is the result of the taking of specific medicines. Certainly a statement by a patient whose diet is not otherwise controlled or brought into the inquiry, who may be taking other medicines at the same time, and particularly in a disease which has a high rate of remissions, that his pain, swelling or limitation of movement has been helped by Tri-Wonda, cannot amount to substantial evidence, even though it be technically admissible.
Dealing with just such an appeal by the Government from a finding by the trial court that it “had failed to carry the burden of establishing the truth of the allegations of its complaint” this Court, in the Hoxsey case clearly established the law which must guide us here.
There, as here, highly qualified experts in the field of medicine involved (there, cancer, here arthritis) testified uniformly that the disease could not be diagnosed without the use of scientific aids not used by the witnesses for the appellee in either case; that the cause of the disease was unknown and the known treatment of it of very doubtful efficacy at best; that great amounts of research had been carried on to broaden the field of knowledge of the medical profession as to cause and cure. There, as here, there' was testimony on behalf of the Government resulting from clinical studies made by those medical men who had specialized in the field, which tests were made with controls and with care to make most likely the possibility of an objective ascertainment of the truth and there was also testimony for the respondent from others not specialists in the field based on clinical tests made in disregard of the major recognized safeguards to an objective test. There, as here, pharmacologists, those skilled in the knowledge of the effect of chemicals on the human organs and functions, testified to the worthlessness of the drug in question when used as directed for the stated purposes. There, as here, there were doctors and patients who nevertheless testified, obviously without the necessary foundation for basing an opinion, that the medicine did offer relief or cure.
Notwithstanding the testimony which, in another type of case, it might be said would create a conflict, the Court, speaking through Judge Russell, said:
“Thus, even if it be assumed, argu-endo, that there is some measure of *853conflict in the evidence relating to the falsity of the specific representations referred to above, still, it is clear that a finding that such representations are true is not supported by substantial evidence.”
And further:
“We think this so-denominated conflicting evidence is wholly insufficient to cast such doubt upon the testimony adduced in behalf of the Government as to authorize the trial court to find that the Government had failed to carry the burden of establishing the truth of the allegations of its complaint.” 198 F.2d 273, 280, 281.
In this case the evidence on behalf of the United States was impressive. Specialists in the field of arthritic diseases, active in arthritic research and members of the learned societies dealing with this medical speciality, one of whom testified that he had treated approximately 40,000 arthritic patients, one orthopedic surgeon and two research specialists, testified without equivocation that the ingredients of these three patent medicines in the quantities recommended for treatment were without therapeutic value either in the treatment of, or alleviation of the symptoms of arthritis.
In addition to the foregoing evidence three pharmacologists, experts in the field of drugs and their effect on the human system, also testified that the ingredients of Tri-Wonda were not recognized in any of the literature or teachings of the profession for the treatment of, or alleviation of the suffering from arthritic diseases. They further gave their opinion that they had no value in relation to such disease.
This evidence was countered, on behalf of the appellee, by five general practitioners, all of whom professed not to be specialists in the field, and all of whom made disparaging remarks concerning their own qualifications either to diagnose or treat the several types of arthritic diseases. Their testimony was almost without exception based upon tests2 on patients sent to them by the appellee or on patients who had diagnosed their own condition and asked for treatment by obtaining a free course of medicine furnished by appellee.
Appellant’s brief is replete with specific excerpts of testimony relating to cases testified about by these general practitioners and their patients. Strikingly, appellee opens its brief with the following statement:
“Except for conclusions of the pleader as to the effect of evidence adduced, appellee believes the case is fairly stated in a part of the brief for appellant. Only the first page statement is accepted because the remainder of the statement by appellant is argument in advance.”
Yet nowhere in the brief is a single statement of fact contained in the brief of the United States refuted or otherwise attacked by the appellee. We have carefully read the record references to the testimony of the 35 to 40 case histories which are carefully analyzed in the Government’s brief and have found the conclusions drawn by the Government are completely accurate as to the effect of the testimony.
In brief, it must be said that the evidence of no one of the five general practitioners rises to the quality necessary to constitute substantial evidence when considered in the light of the other evidence in the record. This is so because either the medical witnesses thoroughly disqualified themselves as having any skill in either diagnosis or treatment of the arthritic diseases or because their evidence as to the effectiveness of Tri-Wonda in alleviating pain, reducing swelling or improving mobility of joints was either merely a repetition of statements made to them by the patients, or because the *854record clearly discloses that there was no diagnosis of the existence of the disease either before or after the so-called treatment, or for both reasons. For instance, one of the doctors testified: “Everyone of these patients already had their own diagnosis made.” They all completely ignored the important differences in the cause and treatment normally accorded the different types of arthritis. They all testified they knew of no literature in the field that suggested the component parts of these medicines, taken singly or together, as being efficacious in the treatment of the disease.
Even though otherwise not objectionable, the testimony of these witnesses amounted to nothing more than testimony from their individual personal experience. As to this kind of medical testimony Wigmore’s comment is pertinent:
“To allow any physician to testify who claims to know solely by personal experience is to appropriate the witness-stand to impostors. Medical science is a mass of transmitted and collated data from numerous quarters; the generalizations which are the result of one man’s personal observation exclusively are the least acceptable of all. The law must recognize the methods of medical science. It cannot stultify itself by establishing, for judicial inquiries, a rule never considered necessary by the medical profession itself. It is enough for a physician, testifying to a medical fact, that he is by training and occupation a physician; whether his source of information for that particular fact is in part or entirely the hearsay of his fellow-practitioners and investigators, is immaterial.” Wigmore, Evidence, 3d ed., Vol. III, § 687.
In addition to the testimony of the general practitioners above referred to, appellee introduced as a witness a Dr. Mary Gray, also lacking any special qualifications in the field of arthritic diseases, who merely undertook to interview a number of patients to whom she was sent by Mrs. Wier, the appellee. Her “survey” was, of course, based solely on the statements made to her by the customers who had already used the medicine before they were interviewed by her. A number of these persons who were reported by Dr. Gray to have stated that their condition had improved, testified as witnesses at the trial in which they repudiated such testimony.
Finally, a Dr. Nellie Watts testified that she had searched the literature in the arthritic field and, based solely on this search she had made, she concluded that some of the chemicals in these medicines might act as a diuretic and, by lowering the water content of the body generally, reduce swelling of the joints. In view of the fact that two of the writers of the articles on which Dr. Watts principally relied, testified at the trial that the ingredients of these medicines in the quantities prescribed would not have the effect attributed to them, it appears that no reliance can be placed upon testimony based on their theory.
The appellee strongly urges that there is a clear distinction between a contention that a medicine is recommended for the treatment of arthritis and a statement that a medicine will in some cases relieve the pain, swelling and limitation of movement associated with arthritis. We think we do not need to decide whether there is such distinction here as to make permissible the marketing of a product under the second representation which has been found not marketable under the first, because we think it clear beyond any question that the findings of the trial court that the Government had not carried its burden of proving that the Tri-Wonda medicines were not “beneficial in a substantial number of cases in the relief of some symptoms of rheumatoid arthritis, rheumatism and bursitis, such as pain, soreness, the swelling of tissues around the joints, the loss of freedom of motion resulting from pain and soreness accompanying rheumatoid arthritis” was false and misleading is so “against the great preponderance of the credible testimony that it does not reflect or represent *855the truth and right of the case.” Sanders v. Leech, 5 Cir., 158 F.2d 486, 487.
We do not need to question the credibility of any of the witnesses. We assume that the trial court credited each of the appellee’s witnesses with telling the truth. This does not, however, add any weight to testimony which, because of demonstrated lack of opportunity properly to base opinion, relegated such testimony to the class mentioned by this Court in the Hoxsey case as being contrary “to all accepted scientific knowledge” and, therefore, not substantial.
On the entire evidence we are left “with the definite and firm conviction that a mistake has been committed.” United States v. United States Gypsum Co., 333 U.S. 364, 395, 68 S.Ct. 525, 542, 92 L.Ed. 746. The overwhelming weight of the evidence requires a conclusion that the representation that these medicines may relieve the pain or swelling or affect the limitation of movement accompanying rheumatoid arthritis is false and misleading.
The trial court should, upon such evidence, have granted the injunction as prayed for and the court’s failure to do so, as stated by us in the Hoxsey case, “evidences an abuse of discretion.”
Reversed and Remanded for further proceedings not inconsistent with this opinion.
. Tri-Wonda No. 1 is a combination of dilute hydrochloric and dilute and nitric acids with traces of tartaric and acetic acids. Tri-Wonda No. 2, which is a combination of cream of tartar, senna, sulphur, and phenolphthalein, is a mild laxative. Tri-Wonda No. 3 consists of fluid extract of Jamaica dogwood, thiamin hydrochloride (Vitamin Bl) and wild cherry flavoring dissolved in 44% alcohol.
. The circumstances under which these tests were given so far lacked the normal controls recognized even by these witnesses as proper to an objective ascertainment of the worth of a new drug that they cannot really be called “clinical tests.”