delivered the opinion of the Court.

This is a direct appeal under § 2 of the Expediting Act1 from a judgment of the District Court for the Northern District of Illinois,2 dismissing the Government’s action brought in 1949 under § 15 of the Clayton Act.3 The complaint alleged a violation of § 7 of the Act4 resulting from the purchase by E. I. du Pont de Nemours and Company in 1917-1919 of a 23% stock interest in General Motors Corporation. This appeal is from the dismissal of the action as to du Pont, General Motors and the corporate holders of large amounts of du Pont stock, Chris-tiana Securities Corporation and Delaware Realty & Investment Company.5

The primary issue is whether du Pont’s commanding position as General Motors’ supplier of automotive *589finishes and fabrics was achieved on competitive merit alone, or because its acquisition of the General Motors’ stock, and the consequent close intercompany relationship, led to the insulation of most of the General Motors’ market from free competition, with the resultant likelihood, at the time of suit, of the creation of a monopoly of a line of commerce.

The first paragraph of § 7, pertinent here, provides:

“That no corporation engaged in commerce shall acquire, directly or indirectly, the whole or any part of the stock or other share capital of another corporation engaged also in commerce, where the effect of such acquisition may be to substantially lessen competition between the corporation whose stock is so acquired and the corporation making the acquisition, or to restrain such commerce in any section or community, or tend to create a monopoly of any line of commerce.” 6

Section 7 is designed to arrest in its incipiency not only the substantial lessening of competition from the acquisition by one corporation of the whole or any part of the stock of a competing corporation, but also to arrest in their incipiency restraints or monopolies in a relevant market which, as a reasonable probability, appear at the time of suit likely to result from the acquisition by one corporation of all or any part of the stock of any other corporation. The section is violated whether or not actual restraints or monopolies, or the substantial lessening of competition, have occurred or are intended. Acquisitions solely for investment are excepted, but only if, and so long as, the stock is not used by voting or otherwise to bring about, or in attempting to bring about, the substantial lessening of competition.

*590We are met at the threshold with the argument that § 7, before its amendment in 1950, applied only to an acquisition of the stock of a competing corporation, and not to an acquisition by a supplier corporation of the stock of a customer corporation — in other words, that the statute applied only to horizontal and not to vertical acquisitions. This is the first case presenting the question in this Court. International Shoe Co. v. Federal Trade Comm’n, 280 U. S. 291, and Thatcher Mfg. Co. v. Federal Trade Comm’n, 272 U. S. 554, involved corporate acquisitions of stock of competitors.

During the 35 years before this action was brought, the Government did not invoke § 7 against vertical acquisitions. The Federal Trade Commission has said that the section did not apply to vertical acquisitions. See F. T. C., Report on Corporate Mergers and Acquisitions, 168 (1955), H. R. Doc. No. 169, 84th Cong., 1st Sess. Also, the House Committee considering the 1950 revision of § 7 stated that “. . . it has been thought by some that this legislation [the 1914 Act] applies only to the so-called horizontal mergers. . . .” H. R. Rep. No. 1191, 81st Cong., 1st Sess. 11. The House Report adds, however, that the 1950 amendment was purposed “. . . to make it clear that the bill applies to all types of mergers and acquisitions, vertical and conglomerate as well as horizontal . . . (Emphasis added.)

This Court has the duty to reconcile administrative interpretations with the broad antitrust policies laid down by Congress. Cf. Automatic Canteen Co. v. Federal Trade Comm’n, 346 U. S. 61, 74. The failure of the Commission to act is not a binding administrative interpretation that Congress did not intend vertical acquisitions to come within the purview of the Act. Accord, Baltimore & Ohio R. Co. v. Jackson, 353 U. S. 325, 331.

The first paragraph of § 7, written in the disjunctive, plainly is framed to reach not only the corporate acquisi*591tion of stock of a competing corporation, where the effect may be substantially to lessen competition between them, but also the corporate acquisition of stock of any corporation, competitor or not, where the effect may be either (1) to restrain commerce in any section or community, or (2) tend to create a monopoly of any line of commerce. The amended complaint does not allege that the effect of du Pont’s acquisition may be to restrain commerce in any section or community but alleges that the effect was “. . . to tend to create a monopoly in particular lines of commerce . . . .”

Section 7 contains a second paragraph dealing with a holding company’s acquisition of stock in two or more corporations.7 Much of the legislative history of the section deals with the alleged holding company evil.8 This history does not aid in interpretation because our concern here is with the first paragraph of the section. There is, however, pertinent legislative history which does aid and support our construction.

Senator Chilton, one of the Senate managers of the bill, explained that the House conferees insisted that to prohibit just the acquisitions where the effect was “substantially” to lessen competition would not accomplish the designed aim of the statute, because “a corporation might acquire the stock of another corporation, *592and there would be no lessening of competition, but the tendency might be to create monopoly or to restrain trade or commerce.” “Therefore,” said Senator Chil-ton, “there was added . . . the following: ‘Or to restrain such commerce in any section or community or tend to create a monopoly of any line of commerce.' ” 9 This construction of the section, as embracing three separate and distinct effects of a stock acquisition, has also been recognized by a number of federal courts.10

We hold that any acquisition by one corporation of all or any part of the stock of another corporation, competitor or not, is within the reach of the section whenever the reasonable likelihood appears that the acquisition will result in a restraint of commerce or in the creation of a monopoly of any line of commerce. Thus, although du Pont and General Motors are not competitors, a violation of the section has occurred if, as a result of the acquisition, there was at the time of suit a reasonable likelihood of a monopoly of any line of commerce. Judge Maris correctly stated in Transamerica Corp. v. Board of Governors, 206 F. 2d 163, 169:

“A monopoly involves the power to . . . exclude competition when the monopolist desires to do so. Obviously, under Section 7 it was not necessary . . . to find that . . . [the defendant] has actually achieved monopoly power but merely that the stock acquisitions under attack have brought it measurably closer to that end. For it is the purpose of the *593Clayton Act to nip monopoly in the bud. Since by definition monopoly involves the power to eliminate competition a lessening of competition is clearly relevant in the determination of the existence of a tendency to monopolize. Accordingly in order to determine the existence of a tendency to monopoly in . . : any . . . line of business the area or areas of existing effective competition in which monopoly power might be exercised must first be determined. ...”

Appellees argue that there exists no basis for a finding of a probable restraint or monopoly within the meaning of § 7 because the total General Motors market for finishes and fabrics constituted only a negligible percentage of the total market for these materials for all uses, including automotive uses. It is stated in the General Motors brief that in 1947 du Pont's finish sales to General Motors constituted 3.5% of all sales of finishes to industrial users, and that its fabrics sales to General Motors comprised 1.6% of the total market for the type of fabric used by the automobile industry.

Determination of the relevant market is a necessary predicate to a finding of a violation of the Clayton Act because the threatened monopoly must be one which will substantially lessen competition “within the area of effective competition.” 11 Substantiality can be determined only in terms of the market affected. The record shows that automotive finishes and fabrics have sufficient peculiar characteristics and uses to constitute them products sufficiently distinct from all other finishes *594and fabrics 12 to make them a “line of commerce” within the meaning of the Clayton Act. Cf. Van Camp & Sons Co. v. American Can Co., 278 U. S. 245.13 Thus, the *595bounds of the relevant market for the purposes of this case are not coextensive with the total market for finishes and fabrics, but are coextensive with the automobile industry, the relevant market for automotive finishes and fabrics.14

The market affected must be substantial. Standard Fashion Co. v. Magrane-Houston Co., 258 U. S. 346, 357. Moreover, in order to establish a violation of § 7 the Government must prove a likelihood that competition may be “foreclosed in a substantial share of . . . [that market] .”15 Both requirements are satisfied in this case. The substantiality of a relevant market comprising the automobile industry is undisputed. The substantiality of General Motors’ share of that market is fully established in the evidence.

General Motors is the colossus of the giant automobile industry. It accounts annually for upwards of two-fifths of the total sales of automotive vehicles in the Nation.16 *596In 1955 General Motors ranked first in sales and second in assets among all United States industrial corporations17 and became the first corporation to earn over a billion dollars in annual net income.18 In 1947 General Motors’ total purchases of all products from du Pont were $26,628,274, of which $18,938,229 (71%) represented purchases from du Pont’s Finishes Division. Of the latter amount purchases of “Duco” 19 and the thinner used to apply “Duco” totaled $12,224,798 (65%), and “Dulux” 20 purchases totaled $3,179,225. Purchases by General Motors of du Pont fabrics in 1948 amounted to $3,700,000, making it the largest account of du Pont’s Fabrics Division. Expressed in percentages, du Pont supplied 67% of General Motors’ requirements for finishes in 1946 and 68% in 1947.21 In fabrics du Pont supplied 52.3% of requirements in 1946, and 38.5% in 1947.22 Because General Motors accounts for almost one-half of the automobile industry’s annual sales, its requirements for automotive finishes and fabrics must represent approximately one-half of the relevant market for these materials. Because the record clearly shows that quantitatively and percentagewise du Pont supplies the largest part of General Motors’ requirements, we must conclude that du Pont has a substantial share of the relevant market.

The appellees argue that the Government could not maintain this action in 1949 because § 7 is applicable only to the acquisition of stock and not to the holding *597or subsequent use of the stock. This argument misconceives the objective toward which § 7 is directed. The Clayton Act was intended to supplement the Sherman Act.23 Its aim was primarily to arrest apprehended consequences of intercorporate relationships before those relationships could work their evil, which may be at or any time after the acquisition, depending upon the circumstances of the particular case. The Senate declared the objective of the Clayton Act to be as follows:

“. . . Broadly stated, the bill, in its treatment of unlawful restraints and monopolies, seeks to prohibit and make unlawful certain trade practices which, as a rule, singly and in themselves, are not covered by the Act of July 2, 1890 [the Sherman Act], or other existing antitrust acts, and thus, by making these practices illegal, to arrest the creation of trusts, conspiracies, and monopolies in their incipiency and before consummation. . . .” S. Rep. No. 698, 63d Cong., 2d Sess. 1. (Emphasis added.)

“Incipiency” in this context denotes not the time the stock was acquired, but any time when the acquisition threatens to ripen into a prohibited effect. See Transamerica Corp. v. Board of Governors, 206 F. 2d 163, 166. To accomplish the congressional aim, the Government may proceed at any time that an acquisition may be said with reasonable probability to contain a threat that it may lead to a restraint of commerce or tend to create a monopoly of a line of commerce.24 Even when the purchase is solely for investment, the plain language of § 7 contemplates an action at any time the stock is used to *598bring about, or in attempting to bring about, the substantial lessening of competition.25

Prior cases under § 7 were brought at or near the time of acquisition. See, e. g., International Shoe Co. v. Federal Trade Comm’n, 280 U. S. 291; V. Vivaudou, Inc. v. Federal Trade Comm’n, 54 F. 2d 273; Federal Trade Comm’n v. Thatcher Mfg. Co., 5 F. 2d 615, rev’d in part on another ground, 272 U. S. 554; United States v. Republic Steel Corp., 11 F. Supp. 117; In re Vanadium-Alloys Steel Co., 18 F. T. C. 194. None of these cases holds, or even suggests, that the Government is foreclosed from bringing the action at any time when a threat of the prohibited effects is evident.

Related to this argument is the District Court’s conclusion that 30 years of nonrestraint negated “any reasonable probability of such a restraint” at the time of the suit.26 While it is, of course, true that proof of a mere possibility of a prohibited restraint or tendency to monopoly will not establish the statutory requirement that the effect of an acquisition “may be” such restraint or tendency,27 the basic facts found by the District Court demonstrate the error of its conclusion.28

The du Pont Company’s commanding position as a General Motors supplier was not achieved until shortly *599after its purchase of a sizable block of General Motors stock in 1917.29 At that time its production for the automobile industry and its sales to General Motors were relatively insignificant. General Motors then produced only about 11% of the total automobile production and its requirements, while relatively substantial, were far short of the proportions they assumed as it forged ahead to its present place in the industry.

At least 10 years before the stock acquisition, the du Pont Company, for over a century the manufacturer of military and commercial explosives, had decided to expand its business into other fields. It foresaw the loss of its market for explosives after the United States Army and Navy decided in 1908 to construct and operate their own plants. Nitrocellulose, a nitrated cotton, was the principal raw material used in du Pont’s manufacture of smokeless powder. A search for outlets for this raw material uncovered requirements in the manufacture of lacquers, celluloid, artificial leather and artificial silk. The first step taken was the du Pont purchase in 1910 of the Fabrikoid Company, then the largest manufacturer of artificial leather, reconstituted as the du Pont Fabrikoid Company in 1913.

The expansion program was barely started, however, when World War I intervened. The du Pont Company suddenly found itself engulfed with orders for military explosives from foreign nations later to be allies of the United States in the war, and it had to increase its capacity and plant facilities from 700,000 to 37,000,000 pounds per month at a cost exceeding $200,000,000. Profits accumulated and ultimately amounted to $232,-000,000. The need to find postwar uses for its expanded facilities and organization now being greater than ever, *600du Pont continued its expansion program during the war years, setting aside $90,000,000 for the purpose. In September 1915, du Pont bought the Arlington Works, one of the Nation’s two largest celluloid companies. In June 1916, the Fairfield Rubber Company, producers of rubber-coated fabrics for automobile and carriage tops, was taken over by du Pont Fabrikoid. In March 1917, purchase was made of Harrison Brothers and Company, manufacturers of paint, varnish, acids and certain inorganic chemicals used in paint manufacture. Shortly afterwards, Harrison absorbed Beckton Chemical Company, a color manufacturer, and, also in 1917, the Bridgeport Wood Finishing Company, a varnish manufacturer.

Thus, before the first block of General Motors stock was acquired, du Pont was seeking markets not only for its nitrocellulose, but also for the artificial leather, celluloid, rubber-coated goods, and paints and varnishes in demand by automobile companies. In that connection, the trial court expressly found that . .• reports and other documents written at or near the time of the investment show that du Pont’s representatives were well aware that General Motors was a large consumer of products of the kind offered by du Pont,” and that John J. Raskob, du Pont’s treasurer and the principal promoter of the investment, “for one, thought that du Pont would ultimately get all that business . . . .”30

The Company’s interest in buying into General Motors was stimulated by Raskob and Pierre S. du Pont, then du Pont’s president, who acquired personal holdings of General Motors stock in 1914. General Motors was organized six years earlier by William C. Durant to acquire previously independent automobile manufacturing companies — Buick, Cadillac, Oakland and Oldsmobile. ' Durant later brought in Chevrolet, organized by *601him when he was temporarily out of power, during 1910-1915, and a bankers’ group controlled General Motors. In 1915, when Durant and the bankers deadlocked on the choice of a Board of Directors, they resolved the deadlock by an agreement under which Pierre S. du Pont was named Chairman of the General Motors Board, and Pierre S. du Pont, Raskob and two nominees of Mr. du Pont were named neutral directors. By 1916, Durant settled his differences with the bankers and resumed the presidency and his controlling position in General Motors. He prevailed upon Pierre S. du Pont and Raskob to continue their interest in General Motors’ affairs, which both did as members of the Finance Committee, working closely with Durant in matters of finances and operations and plans for future expansion. Durant persistently urged both men and the “Wilmington people, as he called it,” 31 to buy more stock in General Motors.

Finally, Raskob broached to Pierre S. du Pont the proposal that part of the fund earmarked for du Pont expansion be used in the purchase of General Motors stock. At this time about $50,000,000 of the $90,000,000 fund was still in hand. Raskob foresaw the success of the automobile industry and the opportunity for great profit in a substantial purchase of General Motors stock. On December 19, 1917, Raskob submitted a Treasurer’s Report to the du Pont Finance Committee recommending a purchase of General Motors stock in the amount of $25,000,000. That report makes clear that more than just a profitable investment was contemplated. A major consideration was that an expanding General Motors would provide a substantial market needed by the burgeoning du Pont organization. Raskob’s summary of reasons in support of the purchase includes this state*602ment: “Our interest in the General Motors Company will undoubtedly secure for us the entire Fabrikoid, Pyralin [celluloid], paint and varnish business of those companies, which is a substantial factor.” (Emphasis added.)32

This thought, that the purchase would result in du Pont’s obtaining a new and substantial market, was echoed in the Company’s 1917 and 1918 annual reports to stockholders. In the 1917 report appears: “Though this is a new line of activity, it is one of great promise and one that seems to be well suited to the character of our organization. The motor companies are very large consumers of our Fabrikoid and Pyralin as well as paints and varnishes.” (Emphasis added.) The 1918 report says: “The consumption of paints, varnishes and fabrikoid in the manufacture of automobiles gives another common interest.”

This background of the acquisition, particularly the plain implications of the contemporaneous documents, destroys any basis for a conclusion that the purchase was made “solely for investment.” Moreover, immediately after the acquisition, du Pont’s influence growing out of it was brought to bear within General Motors to achieve primacy for du Pont as General Motors’ supplier of automotive fabrics and finishes.

Two years were to pass before du Pont’s total purchases of General Motors stock brought its percentage to 23% of the outstanding stock and its aggregate outlay to $49,000,000. During that period, du Pont and Durant worked under an arrangement giving du Pont primary responsibility for finances and Durant the responsibility for operations. But J. A. Haskell, du Pont’s former sales manager and vice-president, became the General Motors vice-president in charge of the operations committee. The trial judge said that Haskell “. . . was willing to under*603take the responsibility of keeping du Pont informed of General Motors affairs during Durant’s regime . . . 33

Haskell frankly and openly set about gaining the maximum share of the General Motors market for du Pont. In a contemporaneous 1918 document, he reveals his intention to “pave the way for perhaps a more general adoption of our material,” and that he was thinking “how best to get cooperation [from the several General Motors Divisions] whereby makers of such of the low priced cars as it would seem possible and wise to get transferred will be put in the frame of mind necessary for its adoption [du Pont’s artificial leather].”

Haskell set up lines of communication within General Motors to be in a position to know at all times what du Pont products and what products of du Pont competitors were being used. It is not pure imagination to suppose that such surveillance from that source made an impressive impact upon purchasing officials. It would be understandably difficult for them not to interpret it as meaning that a preference was to be given to du Pont products. Haskell also actively pushed the program to substitute Fabrikoid artificial leathers for genuine leather and sponsored use of du Pont’s Pyralin sheeting through a liaison arrangement set up between himself and the du Pont sales organization.

Thus sprung from the barrier, du Pont quickly swept into a commanding lead over its competitors, who were never afterwards in serious contention. Indeed, General Motors’ then principal paint supplier, Flint Varnish and Chemical Works, early in 1918 saw the handwriting on the wall. The Flint president came to Durant asking to be bought out, telling Durant, as the trial judge found, that he “knew du Pont had bought a substantial interest in General Motors and was interested in the paint industry; that . . . [he] felt he would lose a valuable *604customer, General Motors.” 34 The du Pont Company-bought the Flint Works and later dissolved it.

In less than four years, by August 1921, Lammot du Pont, then a du Pont vice-president and later Chairman of the Board of General Motors, in response to a query from Pierre S. du Pont, then Chairman of the Board of both du Pont and General Motors, “whether General Motors was taking its entire requirements of du Pont products from du Pont,” was able to reply that four of General Motors’ eight operating divisions bought from du Pont their entire requirements of paints and varnishes, five their entire requirements of Fabrikoid, four their entire requirements of rubber cloth, and seven their entire requirements of Pyralin and celluloid. Lammot du Pont quoted du Pont’s sales department as feeling that “the condition is improving and that eventually satisfactory conditions will be established in every branch, but they wouldn’t mind seeing things going a little faster.” Pierre S. du Pont responded that “with the change in management at Cadillac, Oakland and Olds [Cadillac was taking very little paints and varnishes, and Oakland but 50%; Olds was taking only part of its requirements for fabrikoid], I believe that you should be able to sell substantially all of the paint, varnish and fabrikoid products needed.” He also suggested that “a drive should be made for the Fisher Body business. Is there any reason why they have not dealt with us?”

Fisher Body was stubbornly resistant to du Pont sales pressure. General Motors, in 1920, during Durant’s time, acquired 60% stock control of Fisher Body Company. However, a voting trust was established giving the Fisher brothers broad powers of management. They insisted on running their own show and for years withstood efforts of high-ranking du Pont and General Motors executives to *605get them to switch to du Pont from their accustomed sources of supply. Even after General Motors obtained 100% stock control in 1926, the Fisher brothers retained sufficient power to hold out. By 1947 and 1948, however, Fisher resistance had collapsed, and the proportions of its requirements supplied by du Pont compared favorably with the purchases by other General Motors Divisions.

In 1926, the du Pont officials felt that too much General Motors business was going to its competitors. When Pierre S. du Pont and Raskob expressed surprise, Lam-mot du Pont gave them a breakdown, by dollar amounts, of the purchases made from du Pont’s competitors. This breakdown showed, however, that only Fisher Body of the General Motors divisions was obtaining any substantial proportion of its requirements from du Pont’s competitors.

Competitors did obtain higher percentages of the General Motors business in later years, although never high enough at any time substantially to affect the dollar amount of du Pont’s sales. Indeed, it appears likely that General Motors probably turned to outside sources of supply at least in part because its requirements outstripped du Pont’s production, when General Motors’ proportion of total automobile sales grew greater and the company took its place as the sales leader of the automobile industry. For example, an undisputed Government exhibit shows that General Motors took 93% of du Pont’s automobile Duco production in 1941 and 83% in 1947.

The fact that sticks out in this voluminous record is that the bulk of du Pont’s production has always supplied the largest part of the requirements of the one customer in the automobile industry connected to du Pont by a stock interest. The inference is overwhelming that du Pont’s commanding position was promoted by its stock interest and was not gained solely on competitive merit.

*606We agree with the trial court that considerations of price, quality and service were not overlooked by either du Pont or General Motors. Pride in its products and its high financial stake in General Motors’ success would naturally lead du Pont to try to supply the best. But the wisdom of this business judgment cannot obscure the fact, plainly revealed by the record, that du Pont purposely employed its stock to pry open the General Motors market to entrench itself as the primary supplier of General Motors’ requirements for automotive finishes and fabrics.35

*607Similarly, the fact that all concerned in high executive posts in both companies acted honorably and fairly, each in the honest conviction that his actions were in the best interests of his own company and without any design to overreach anyone, including du Pont’s competitors, does not defeat the Government’s right to relief. It is not requisite to the proof of a violation of § 7 to show that restraint or monopoly was intended.

The statutory policy of fostering free competition is obviously furthered when no supplier has an advantage over his competitors from an acquisition of his customer’s stock likely to have the effects condemned by the statute. We repeat, that the test of a violation of § 7 is whether, at the time of suit, there is a reasonable probability that the acquisition is likely to result in the condemned restraints. The conclusion upon this record is inescapable that such likelihood was proved as to this acquisition. The fire that was kindled in 1917 continues to smolder. It burned briskly to forge the ties that bind the General Motors market to du Pont, and if it has quieted down, it remains hot, and, from past performance, is likely at any time to blaze and make the fusion complete.36

The judgment must therefore be reversed and the cause remanded to the District Court for a determination, after further hearing, of the equitable relief necessary and appropriate in the public interest to eliminate the effects of the acquisition offensive to the statute. The District Courts, in the framing of equitable decrees, are clothed *608“with large discretion to model their judgments to fit the exigencies of the particular case.” International Salt Co. v. United States, 332 U. S. 392, 400-401.

The motion of the appellees Christiana Securities Company and Delaware Realty and Investment Company for dismissal of the appeal as to them is denied. It seems appropriate that they be retained as parties pending determination by the District Court of the relief to be granted.

It is so ordered.

dissenting.

In June 1949, the United States brought this civil action in the United States District Court for the Northern District of Illinois under § 4 of the Sherman Act and § 15 of the Clayton Act to enjoin alleged violations of § § 1 and 2 of the Sherman Act, and § 7 of the Clayton Act. The amended complaint, insofar as pertinent to the issues here, alleged that du Pont and General Motors have been engaged, since 1915, in a combination and conspiracy to restrain and monopolize interstate trade, and that du Pont’s acquisition of General Motors’ stock had the effect of restraining trade and tending to create a monopoly. In brief it was alleged that, by means of the relationship between du Pont and General Motors, du Pont intended to obtain, and did obtain, an illegal preference over its competitors in the sale to General Motors of its products, and a further illegal preference in the development of chemical discoveries made by General Motors. Appellees denied the charges.

*609The trial of these issues took nearly seven months. The District Court heard 52 witnesses, including most of the principal actors, and received over 2,000 exhibits. The evidence contained in the 8,283-page transcript of record covers in minute and intimate detail the facts bearing on the Government’s charge that du Pont, by coercion, agreement, control or influence, had interfered unlawfully with General Motors’ purchasing and manufacturing policies. On the basis of this evidence, the District Court found that the Government had failed to prove its case and, specifically, that (a) du Pont did not control General Motors, (b) there had been “no limitation or restraint upon General Motors’ freedom to deal freely and fully with competitors of du Pont” or upon its “freedom ... to deal with its chemical discoveries,” and (c) after 30 years in which no such restraint had resulted, there was no “basis for a finding that there is or has been any reasonable probability of such a restraint within the meaning of the Clayton Act.” 126 F. Supp. 235, 335.

The Government’s basic contention in this Court is that du Pont violated §§ 1 and 2 of the Sherman Act in that, by means of its alleged control of General Motors, it obtained an unlawful preference with respect to General Motors’ purchases of materials. In the closing pages of its brief, and for a few minutes in its oral argument, the Government added the assertion that du Pont had violated § 7 of the Clayton Act in that its stock interest in General Motors “has been used to channel General Motors’ purchases to du Pont.”

This Court, ignoring the Sherman Act issues which have been the focal point of eight years of litigation, now holds that du Pont’s acquisition of a 23% stock interest in General Motors during the years 1917-1919 violates § 7 of the Clayton Act because “at the time of suit [in 1949] there [was] a reasonable probability that the acquisition [was] likely to result in the condemned *610restraints.” Ante, p. 607. In reaching this conclusion, the Court holds (1) that § 7 of the Clayton Act applies to vertical as well as horizontal stock acquisitions; (2) that in determining whether the effect of the stock acquisition is such as to constitute a restraint within § 7, the time chosen by the Government in bringing the action is controlling rather than the time of the acquisition itself; and (3) that § 7 is violated when, at the time of suit, there is a reasonable probability that the stock acquisition is likely to result in the foreclosure of competitors of the acquiring corporation from a substantial share of the relevant market.

In applying these principles to this case, the Court purports to accept the carefully documented findings of fact of the District Court. Actually, it overturns numerous well-supported findings of the District Court by now concluding that du Pont did not purchase General Motors’ stock solely for investment; that du Pont’s stock interest resulted in practical or working control of General Motors; that du Pont has used or might use this “control” to secure preferences in supplying General Motors with automobile finishes and fabrics; that the relevant market includes only automobile finishes and fabrics; and that there was, even at the time of suit in 1949, a reasonable probability that du Pont’s competitors might be foreclosed from a substantial share of this relevant market.

The Court’s decision is far reaching. Although § 7 of the Clayton Act was enacted in 1914 — over 40 years ago — this is the first case in which the United States or the Federal Trade Commission has sought to apply it to a vertical integration.1 Likewise, this appears to be the first case in which it ever has been argued that § 7 is applicable to a stock acquisition which took place many *611years before.2 The Court, in accepting both of these contentions, disregards the language and purpose of the statute, 40 years of administrative practice, and all the precedents except one District Court decision. The sweeping character of the Court’s pronouncement is further evident from the fact that to make its case the Court requires no showing of any misuse of a stock interest— either at the time of acquisition or subsequently — to gain preferential treatment from the acquired corporation. All that is required, if this case is to be our guide, is that some court in some future year be persuaded that a “reasonable probability” then exists that an advantage over competitors in a narrowly construed market may be obtained as a result of the stock interest. Thus, over 40 years after the enactment of the Clayton Act, it now becomes apparent for the first time that § 7 has been a sleeping giant all along. Every corporation which has acquired a stock interest in another corporation after the enactment of the Clayton Act in 1914, and which has had business dealings with that corporation is exposed, retroactively, to the bite of the newly discovered teeth of §7.

For the reasons given below, I believe that the Court has erred in (1) applying § 7 to a vertical acquisition; (2) holding that the time chosen by the Government in bringing the action is controlling rather than the time of the stock acquisition itself; and (3) concluding, in disregard of the findings of fact of the trial court, that the facts of this case fall within its theory of illegality.

I.

Section 7 of the Clayton Act, quoted in full in the Appendix, post, pp. 655-656, does not make unlawful all *612intercorporate acquisitions and mergers.3 It does not apply to acquisitions of physical assets.4 It applies only to certain acquisitions of stock, and even then with important exceptions. The first paragraph of § 7, which is the statutory provision primarily involved in this case, provides—

“That no corporation engaged in commerce shall acquire, directly or indirectly, the whole or any part of the stock or other share capital of another corporation engaged also in commerce, where the effect of such acquisition may be to substantially lessen competition between the corporation whose stock is so acquired and the corporation making the acquisition, or to restrain such commerce in any section or community, or tend to create a monopoly of any line of commerce.” 38 Stat. 731-732, 15 U. S. C. (1946 ed.) § 18.

This paragraph makes unlawful only those intercorporate stock acquisitions which may result in any of three effects: (1) substantially lessen competition between the *613acquiring and the acquired corporations; (2) restrain commerce in any section or community; or (3) tend to create a monopoly of any line of commerce. The Government concedes that General Motors and du Pont have never been in competition with each other. Since the substantially lessen competition clause applies only to acquisitions involving competing corporations (generally referred to as horizontal acquisitions), that clause con-cededly is not applicable to this case. The questions before us are whether the other unlawful effects, namely, restraint of commerce in any section or community and tendency to create a monopoly of any line of commerce, are applicable to this case, and, if so, whether the 1917-1919 acquisition of General Motors’ stock by du Pont resulted or may result in either of those unlawful effects.

Section 7 never has been authoritatively interpreted as prohibiting the acquisition of stock in a corporation that is not engaged in the same line of business as the acquiring corporation. Although the language of the Act is ambiguous, the relevant legislative history, administrative practice, and judicial interpretation support the conclusion that § 7 does not apply to vertical acquisitions.

The report of the House Committee on the Judiciary, presented by Representative Clayton, stated emphatically that the provisions relating to stock acquisitions by corporations, which originally appeared as § 8 of the bill, were intended to eliminate the evils of holding companies. H. R. Rep. No. 627, 63d Cong., 2d Sess. 17. Although a “holding company” was defined as “a company that holds the stock of another company or companies,” the one “evil” referred to was that a holding company “is a means of holding under one control the competing companies whose stocks it has thus acquired.” (Emphasis supplied.) Ibid. Two minority statements appended to the House Report evidence a similar understanding that the provisions of the bill were limited to competing cor*614porations. Id., Pt. 2, p. 6; Pt. 3, p. 8. The substance of the House Report was adopted by the Senate Committee on the Judiciary in its report on the bill. S. Rep. No. 698, 63d Cong., 2d Sess. 13, 43, 46.

The extensive debates on the bill in each House of Congress contain many detailed discussions of the provisions relating to intercorporate stock acquisitions. These discussions are devoid of any suggestion that the provisions were to apply to vertical acquisitions.5 On the contrary, these provisions of the bill were repeatedly described as prohibiting the acquisition of stock of competing companies.6 The one specific reference to a vertical acquisition during the entire debate on these provisions ended with a flat statement by Senator Reed to the effect that the bill as then written (containing the tendency toward monopoly clause but not the restraint of commerce clause) would not prevent a steel manufacturing corporation from acquiring stock in an ore producing corporation, a classic type of vertical integration.7 *615A reading of the legislative history of the bill leaves the distinct impression that intercorporate relationships between buyers and sellers which resulted in noncompetitive preferences were intended to be dealt with exclusively by the provision forbidding interlocking directorates (§ 8 of the Clayton Act), if not covered by the specific prohibitions of certain price discriminations (§2), and of certain exclusive selling or leasing contracts (§ 3).8

Forty years of administrative practice provides additional support for this view. Neither the Department of Justice nor the Federal Trade Commission, the two principal enforcing agencies, has brought any action under old § 7 (other than the instant case) that has not involved a stock acquisition in allegedly competing corporations. The Federal Trade Commission repeatedly has declared its understanding that § 7, prior to its amendment in 1950, applied only to competing corporations.9 In a recent report it stated without qualification:

“While the 1914 act applied solely to horizontal mergers, the 1950 act applies not only to horizontal *616acquisitions but to vertical and conglomerate acquisitions which might substantially lessen competition or tend to create a monopoly.” F. T. C., Report on Corporate Mergers and Acquisitions (May 1955), 168, H. R. Doc. No. 169, 84th Cong., 1st Sess.

Beginning in 1927, the Federal Trade Commission included in its annual recommendations to Congress a request that § 7 be amended to remedy its inadequacies. This result was achieved in 1950. 64 Stat. 1125, 15 U. S. C. § 18. As the Court recognizes in its opinion, ante, p. 590, one of the reasons for amending § 7 in 1950 was, in the words of the House Report on the amendments, “to make it clear that the bill applies to all types of mergers and acquisitions, vertical and conglomerate as well as horizontal . . . .” H. R. Rep. No. 1191, 81st Cong., 1st Sess. 11. Forty years of established administrative practice, acquiesced in and recognized by Congress, is persuasive evidence of the proper scope of § 7. Federal Trade Commission v. Bunte Bros., Inc., 312 U. S. 349, 351-352.

The cases cited by the Court, with the one exception of Ronald Fabrics Co. v. Verney Brunswick Mills, Inc., CCH Trade Cases ¶ 57,514 (D. C. S. D. N. Y. 1946),10 *617do not support the Court’s conclusion that § 7 applies to a vertical acquisition. In Aluminum Co. of America v. Federal Trade Commission, 284 F. 401 (C. A. 3d Cir. 1922), the Aluminum Company, which previously had had a monopoly of all sheet aluminum produced in the United States, acquired control through an intermediary corporation of a competing sheet aluminum company established in 1916. A divestiture order of the Federal Trade Commission was upheld, the court holding that the stock acquisition substantially lessened competition and tended to create a monopoly of the sheet aluminum business. In United States v. New England Fish Exchange, 258 F. 732 (D. C. Mass. 1919), two holding companies which had acquired the stock of virtually all the wholesale fish dealers trading on the New England Fish Exchange, which handled about 95% of all the ground fish sold in interstate commerce in the United States, were held to have violated the provisions of § 7. Each of these cases was concerned with the acquisition of directly competing corporations — not vertical acquisitions. Statements in the opinions, not essential to the decisions, merely stand for the proposition that the restraint and monopoly clauses of § 7 are not entirely synonymous with the substantially lessen competition clause.

Assuming that the three unlawful effects mentioned in § 7 are not entirely synonymous with each other,11 such an *618assumption does not require the conclusion that § 7 was intended to apply to vertical acquisitions as well as to horizontal acquisitions. Corporations engaged in the same business activity in different areas do not necessarily “compete” with each other so that their combination would substantially lessen competition between them, even though their combination might result in a restraint of commerce or a tendency toward monopoly violative of § 7. Such a possibility was presented in Transamerica Corp. v. Board of Governors, 206 F. 2d 163 (C. A. 3d Cir. 1953), where a banking corporation through a series of transactions acquired stock in 48 local banking corporations, most of which were located in communities in which no other bank was acquired. A divestiture order of the Board was reversed on the ground that the Board had not proved that the acquisitions of these banks in five western States either substantially lessened competition or tended to create a monopoly.

Finally, this Court has twice construed old § 7 as applying only to stock acquisitions involving competing corporations. In International Shoe Co. v. Federal Trade Commission, 280 U. S. 291 (1930), the Court held that the acquisition of the fifth largest shoe manufacturing company by the largest shoe manufacturer did not violate either the substantially lessen competition clause or the restraint of commerce clause of § 7 because the preexisting competition between the two corporations was insubstantial, and because the acquired corporation was *619in a precarious financial condition. Substantial pre-exist-ing competition was said to be a requisite for violation of either clause of § 7. 280 U. S., at 298, 303. An even more direct holding is found in Thatcher Mfg. Co. v. Federal Trade Commission, 272 U. S. 554 (1926), where this Court affirmed that portion of the lower court’s decree which had allowed Thatcher, a milk bottle manufacturer, to retain the assets of Woodbury, a bottle manufacturer specializing in condiment and whiskey bottles, on the ground that the acquisition did not violate any of the three clauses of § 7 since Thatcher was not in competition with Woodbury. 272 U. S., at 560, affirming in part and reversing in part Federal Trade Commission v. Thatcher Mfg. Co., 5 F. 2d 615 (C. A. 3d Cir. 1925). These holdings apparently will be overruled sub silentio by today’s decision.

The legislative history, administrative practice, and judicial interpretation of § 7 provide the perspective in which the Government’s present assertion that § 7 applies to vertical acquisitions should be viewed. Seen as a whole, they offer convincing evidence that § 7, properly construed, has reference only to horizontal acquisitions. I would so hold. However, even if the opposite view be accepted, the foregoing views of the enforcing agencies and the courts are material to a proper consideration of the other issues which must then be reached.

II.

In this case the Government is challenging, in 1949, a stock acquisition that took place in 1917-1919. The Court, without advancing reasons to support its conclusion, holds that in determining whether the effect of the stock acquisition is such as to violate § 7, the time chosen by the Government in bringing its suit is controlling rather than the time of the acquisition of the stock. This seems to me to ignore the language and structure of § 7, *620the purpose of the Clayton Act, and all existing administrative and judicial precedents.

The first paragraph of § 7 provides that “no corporation . . . shall acquire . . . the stock ... of another corporation . . . where the effect of such acquisition may be . . . .” Yet the Court construes this provision as if it read “no corporation . . . shall acquire or continue to hold . . . the stock ... of another corporation . . . whenever it shall appear that the effect of such acquisition or continued holding may be . . . .” Continued holding, to be sure, is a prerequisite to any action under § 7 because, if the stock is no longer held, the violation has been purged and there is nothing to divest.12 But the fact of continued holding does not allow the Government to dispense with the necessity of proving that the stock was unlawfully acquired. The offense described by § 7 is the acquisition, not the holding or the use, of stock. When the acquisition has been made, the offense, if any, is complete. The statutory language is unequivocal. It makes the test the probable effect of the acquisition at the time of the actual acquisition, and not at some later date to be arbitrarily chosen by the Government in bringing suit.

The distinction carefully made in the several paragraphs of § 7 between an unlawful acquisition and an unlawful use of stock reinforces this conclusion. The first paragraph of § 7, which speaks only in terms of acquisition of stock, is concerned solely with the purchase of stock in “another corporation.” It is the only provision that is applicable in this case. The second paragraph, which expressly prohibits both acquisition and use, is concerned with stock purchases in “two or more corporations.” Concededly, it is not applicable here. When Congress chose to make unlawful the use of stock *621subsequent to its acquisition, it did so in specific terms. The omission of the phrase “or the use of such stock by the voting or granting of proxies or otherwise,” contained in the second paragraph of § 7, from the first paragraph of the section was not inadvertent. The phrase therefore cannot be read into the first paragraph of § 7.13

The Clayton Act was not intended to replace the Sherman Act in remedying actual restraints and monopolies. Its purpose was to supplement the Sherman Act by checking anticompetitive tendencies in their incipiency, before they reached the point at which the Sherman Act comes into play. This purpose was well stated in the Senate Report on the bill:

“Broadly stated, the bill, in its treatment of unlawful restraints and monopolies, seeks to prohibit and make unlawful certain trade practices which, as a rule, singly and in themselves, are not covered by the act of July 2, 1890, or other existing antitrust acts, and thus, by making these practices illegal, to arrest the creation of trusts, conspiracies, and monopolies in their incipiency and before consummation.” S. Rep. No. 698, 63d Cong., 2d Sess. 1.

*622This purpose places emphasis on the probable anticom-petitive effects of transactions or occurrences viewed as of the date of their occurrence. The determination required by the Act is one of predicting the probable outcome of a particular transaction, here an acquisition of stock in another corporation. If, at the time of the stock acquisition, a potential threat to competition is apparent, the acquisition is unlawful under § 7. If, on the other hand, a potential threat to competition is not then apparent, an antitrust violation is not involved unless subsequent use of the stock constitutes a restraint of trade prohibited by the Sherman Act.14

The Court ignores the all-important lawfulness or unlawfulness of the stock acquisition at or about the time it occurred, and limits its attention to the probable anti-competitive effects of the continued holding of the stock at the time of suit, some 30 years later. The result is to subject a good-faith stock acquisition, lawful when made, to the hazard that the continued holding of the stock may make the acquisition illegal through unforeseen developments. Such a view is not supported by the statutory language and violates elementary principles of fairness. Suits brought under the Clayton Act are not subject to any statute of limitations, and it is doubtful whether the doctrine of laches applies as against the Government. The result is that unexpected and unforeseeable developments occurring long after a stock acquisition can be used to challenge the legality of continued holding of the stock. In such an action, the Government need only prove that probable rather than actual anticompetitive *623effects exist as of the time of suit. The Government may thus set aside a transaction which was entirely lawful when made, merely by showing that it would have been unlawful had it occurred at the time of suit, many years later. The growth of the acquired corporation, a fortuitous decline in the number of its competitors, or the achievement of control by an accidental diffusion of other stock may result, under this test, in rendering the originally lawful acquisition unlawful ab initio. Strikingly enough, all of these factors are involved in this case.15

The Court’s holding is unfair to the individuals who entered into transactions on the assumption, justified by the language of § 7, that their actions would be judged by the facts available to them at the time they made their decision.

“The prohibition [of § 7] is addressed to parties who contemplate engaging in merger transactions and is meant, in the first instance, to guide them in deciding upon a course of action. The only standard they are capable of applying is one addressed to the circumstances viewed as of the date of the proposed transaction. Since this is the standard which the parties must apply in deciding whether to undertake a transaction, it seems reasonable to conclude that it is the standard which enforcement agencies should *624apply in deciding whether the transaction violates the statute.” Neal, The Clayton Act and the Trans-america Case, 5 Stan. L. Rev. 179, 220-221.

The Court cites no authority in support of its new interpretation of this 40-year-old statute. On the other hand, examination of the dozen or more cases brought under § 7 reveals that in every case the inquiry heretofore has centered on the probable anticompetitive effects of the stock acquisition at or near the time it was made.16 See, e. g., International Shoe Co. v. Federal Trade Commission, 280 U. S. 291 (1930); Transamerica Corp. v. Board of Governors, 206 F. 2d 163 (C. A. 3d Cir. 1953); V. Vivaudou, Inc. v. Federal Trade Commission, 54 F. 2d 273 (C. A. 2d Cir. 1931); Federal Trade Commission v. Thatcher Mfg. Co., 5 F. 2d 615 (C. A. 3d Cir. 1925), rev’d in part on another ground, 272 U. S. 554; United States v. Republic Steel Corp., 11 F. Supp. 117 (D. C. N. D. Ohio 1935); In re Vanadium-Alloys Steel Co., 18 F. T. C. 194 (1934). The conclusion thus seems inescapable that the unlawfulness of a stock acquisition under the first paragraph of § 7 properly turns on the potential threat to competition created by the acquisition of the stock at the time of its acquisition and not by its subsequent use.

That the time of acquisition is controlling does not mean that the Government is unable to bring an action if it fails to proceed within a few years of the stock acquisition. It means only that if the Government chooses to bring its action many years later, it must prove what § 7 plainly requires — that the acquisition threatened competition when made.

*625Nor does it mean that evidence of subsequent events is necessarily irrelevant. Evidence that anticompetitive effects have occurred since the acquisition, and that these effects are traceable to the original acquisition rather than to other factors, may support an inference that such effects were “reasonably probable” at the time of acquisition. The element of causation is the necessary link with the past. However, if events subsequent to the acquisition indicate that no anticompetitive effects have occurred, that evidence may support an inference that an unlawful potential did not exist at the time of acquisition. Evidence as to what happened after the acquisition is relevant to the extent that it bears on the central question whether, at the time of the acquisition, there was a reasonable probability of a threat to competition.

I agree with the Court that § 7 does not require findings and conclusions of actual anticompetitive effects. Unlike the Sherman Act, § 7 merely requires proof of a reasonable probability of a substantial lessening of competition, restraint of commerce, or tendency toward monopoly. International Shoe Co. v. Federal Trade Commission, 280 U. S. 291; Transamerica Corp. v. Board of Governors, 206 F. 2d 163. When a vertical acquisition is involved, its legality thus turns on whether there is a reasonable probability that it will foreclose competition from a substantial share of the market, either by significantly restricting access to needed supplies or by significantly limiting the market for any product. See Report of the Attorney General’s National Committee to Study the Antitrust Laws (1955) 122-127. The determination of such probable economic consequences requires study of the markets affected, of the companies involved in relation to those markets, and of the probable immediate and future effects on competition. A mere showing that a substantial dollar volume of sales is involved cannot suffice. As the Court says, “The market *626affected must be substantial,” ante, p. 595, and “Substan-tiality can be determined only in terms of the market affected,” ante, p. 593. Section 7 thus requires a case-by-case analysis of the relevant economic factors.

However, when, as here, the Government brings a proceeding nearly 30 years after a stock purchase, it must prove that the acquisition was unlawful when made (i. e., that there was a reasonable probability at that time that du Pont’s competitors would be foreclosed from a substantial share of the relevant market), and also that the effect of the acquisition continued to be harmful to competition at the time suit was brought. Illegality at the time of acquisition is required by the first paragraph of § 7; continuing illegality is a prerequisite for obtaining equitable relief. See United States v. W. T. Grant Co., 345 U. S. 629; United States v. Oregon Medical Society, 343 U. S. 326, 333; United States v. South Buffalo R. Co., 333 U. S. 771, 774. This is particularly true under § 7 since it is a prophylactic measure designed to prevent stock acquisitions which probably will have a deleterious effect on competition. Proof that competition has not in fact been harmed during a long period following a stock acquisition itself indicates that a restraint in the future is unlikely. In such a case, the actual effect of the acquisition largely supplants the conjecture as to its probable effects which otherwise must be relied upon.

In this case, the District Court found that the challenged acquisition, which took place “over thirty years ago,” had not resulted in any restraint of trade “In those many intervening years . . . .” The District Court properly concluded that, when there had been no restraint for 30 years, “there is not . . . any basis for a finding that there is . . . any reasonable probability of such a restraint within the meaning of the Clayton Act.” 126 F. Supp., at 335. If the evidence supports the District Court’s conclusion that there has been no restraint for 30 years, the judgment below must be affirmed.

*627III.

The remaining issues are factual: (1) whether the record establishes the existence of a reasonable probability that du Pont’s competitors will be foreclosed from securing General Motors’ trade, and (2) whether the record establishes that such foreclosure, if probable, involves a substantial share of the relevant market and significantly limits the competitive opportunities of others trading in that market. In discussing these factual issues, I meet the Court on its own ground, that is, I assume that the old § 7 applies to vertical acquisitions, and that the potential threat at the time of suit is controlling. Even on that basis the record does not support the Court’s conclusion that § 7 was violated by this 1917-1919 stock acquisition.

A. FORECLOSURE OF COMPETITORS.

This is not a case where a supplier corporation has merged with its customer corporation with the result that the supplier’s competitors are automatically and completely foreclosed from the customer’s trade.17 In this case, the only connection between du Pont, the supplier, and General Motors, the customer, is du Pont’s 23% stock interest in General Motors. A conclusion that such a stock interest automatically forecloses du Pont’s competitors from selling to General Motors would be without justification. Whether a foreclosure has occurred in the past or is probable in the future is a question of fact turning on the evidence in the record.

The Court, at the outset of its opinion, states that the primary issue is whether du Pont’s position as a substan*628tial supplier to General Motors “was .achieved on competitive merit alone,” or resulted from du Pont’s stock interest in General Motors. Ante, pp. 588-589. In resolving this issue, the Court states that the “basic facts” are not in dispute and hence that it is unnecessary to set aside the findings of fact of the District Court as clearly erroneous. See Fed. Rules Civ. Proc., 52 (a). The basic facts are said to be that du Pont had no standing as a General Motors’ supplier before the stock purchases of 1917-1919, that it gained a “commanding position” after the stock purchases, and that certain items of evidence in this gigantic record tend to indicate that du Pont hoped to get and actually did get a preference in General Motors’ trade. From these alleged facts the Court draws the conclusion that du Pont has misused its 23% stock interest in General Motors “to entrench itself as the primary supplier of General Motors’ requirements for automotive finishes and fabrics.” Ante, p. 606. “The inference is overwhelming,” the Court concludes, “that du Pont’s commanding position was promoted by its stock interest and was not gained solely on competitive merit.” Ante, p. 605. With these words, the Court overturns the District Court’s unequivocal findings to the effect that du Pont was a principal supplier to General Motors prior to the 1917-1919 stock purchases, that du Pont maintained this position in the years following the stock purchases, and that for the entire 30-year period preceding the suit, General Motors’ purchases of du Pont’s products were based solely on the competitive merits of those products. The evidence supporting these findings of the District Court may be summarized as follows:

Du Pont is primarily a manufacturer of chemicals and chemical products. Thousands of its products could be used by General Motors in manufacturing automobiles, appliances and machinery. Despite du Pont’s sales efforts over a period of 40 years, General Motors buys *629many of the commodities produced by du Pont from du Pont’s competitors.18 The Court, ignoring the many products which General Motors declines to buy from du Pont or which it buys only in small quantities, concentrates on the few products which du Pont has sold in large volume to General Motors for many years — paints and fabrics. Before examining the history of those large-volume purchases, it is essential to understand where and by whom purchasing decisions within General Motors have been made.

For many years, General Motors has been organized into some 30 operating divisions, each of which has final authority to make, and does make, its own purchasing decisions. This decentralized management system places full responsibility for purchasing decisions on the officers of the respective divisions. To speak of “selling to General Motors” is, therefore, misleading. A’prospective supplier, instead of selling to General Motors, sells to Chevrolet, or Frigidaire, or Ternstedt, or Delco Light, as divisions. Moreover, when there are several plants within a division, each plant frequently has its own purchasing agent and presents a separate selling job.

*630The record discloses that each division buys independently, that the pattern of buying varies greatly from one division to another, and that within each division purchases from du Pont have fluctuated greatly in response to price, quality, service and other competitive considerations. For example, Oldsmobile is the only division which buys antifreeze from du Pont and one of the two car divisions which does not finish its cars with Duco. Buick alone buys du Pont motor enamel, and Cadillac alone uses du Pont’s copper electroplating exclusively. Thus the alleged nefarious influence arising from du Pont’s stock interest apparently affects the Oldsmobile antifreeze buyer, but not the Oldsmobile paint buyer; the paint buyers at Chevrolet, Buick and Pontiac, but not the antifreeze or electroplating buyers; and the electroplating buyer at Cadillac, but not the Cadillac paint buyer.

1. Paints. — Du Pont, for many years, has had marked success in the manufacture and sale of paints, varnishes, lacquers and related products.19 In 1939, it produced 9.5% of the total dollar value of all finishes produced in *631the United States and, in 1947, 8.1%. In recent years, approximately three-fourths of du Pont’s total sales to General Motors have consisted of industrial finishes.20 Although du Pont has been General Motors’ principal supplier of paint for many years, General Motors continues to buy about 30% of its paint requirements from competitors of du Pont.21 Moreover, the sales of paint from du Pont to General Motors do not bulk large in the respective total sales and purchases of either company. In 1948, du Pont’s finish sales to General Motors were only 3% of its total sales of all products; they were an infinitesimal percentage of General Motors’ total purchases.

Two products account for a high proportion of these finish sales to General Motors: “Duco,” a nitrocellulose lacquer invented and patented by du Pont, and “Dulux,” a synthetic resin enamel developed by du Pont.22 However, Duco and Dulux did not come into commercial use until 1924 and 1931, respectively, and du Pont’s position as a *632principal manufacturer of finishes was attained much earlier.

Du Pont first assumed a leading position in the automotive finish field with its acquisition, in 1918, of a majority of the stock of the Flint Varnish & Color Works at Flint, Michigan. At that time, and for some years before, Flint supplied the finishes used on all General Motors’ cars except Cadillac, and also for many other automobile companies. Du Pont’s acquisition of General Motors’ stock in 1917-1919 did not influence the General Motors’ divisions in purchasing from Flint. In 1921, Flint lost one-half of the Oakland business and, in 1923, a substantial portion of the business at Buick, Oakland and Oldsmobile. 126 F. Supp., at 288.

The invention and development of Duco in the early 1920’s represented a significant technological advance. Automobiles previously had been finished by applying numerous coats of varnish. The finishing process took from 12 to 33 days, and the storage space and working capital tied up in otherwise completed cars were immense. The life expectancy of varnish finishes was less than a year. In December 1921, General Motors created a Paint and Enamel Committee which contacted numerous paint manufacturers in an attempt to find a quicker drying and more durable finish.

Meanwhile, du Pont had been doing pioneering work in nitrocellulose lacquers. In 1920, a du Pont employee invented a quick drying and durable lacquer which contained a large amount of film-forming solids. This patented finish, named Duco, was submitted to the General Motors Paint and Enamel Committee in 1922 to be tested along with finishes of other manufacturers. After two years of testing and improvement, the Paint and Enamel Committee became satisfied that Duco was far superior *633to any other product or any other method of finishing automobiles then available.

The gradual adoption of Duco by some of the General Motors’ car divisions, viewed in conjunction with its proved superiority as an auto finish, illustrates the independent buying of each division and demonstrates that Duco made its way on its own merits. Oakland (now Pontiac) first adopted Duco for use on its open cars in 1924. The new finish was an immense success and was used on all Oakland cars the following year. Buick and Chevrolet adopted Duco in 1925, but Cadillac, which had offered it as an optional finish in 1925, did not abandon varnish for Duco until 1926.23

From the beginning, General Motors continued to look for competitive materials. Letters were sent to other manufacturers urging them to submit samples of their pyroxylin paint for testing. Until 1927, none of the competing lacquers was comparable in quality to Duco. But the strenuous efforts by General Motors to develop competitive sources of lacquer eventually worked a substantial change in the du Pont position. Oldsmobile and Cadillac switched to a competitor, Rinshed-Mason, in 1927, and have continued to buy almost exclusively from that company ever since. Chevrolet, Buick and Pontiac continued to buy Duco, partly because of better service from nearby du Pont plants, and partly because repeated testing failed to disclose any lacquer superior to Duco.

Finally, the success of Duco has never been confined to the General Motors’ car divisions. In 1924 and 1925, nearly all car manufacturers abandoned varnish for Duco. *634By the end of 1925, all cars, except Ford and Cadillac, were using Duco. Nash, Hudson, Studebaker, Packard and Willys have bought, and still buy, Duco in substantial amounts from du Pont. Chrysler bought Duco in large volume until the early 1930’s when, in pursuance of a policy to obtain suppliers to whom it would be the most important customer, it concentrated its purchases on one company, Pittsburgh Plate Glass. Ford has chosen to make a large part of its own requirements. During the 1920’s, when Ford was losing its leadership in the low-priced field to Chevrolet, it continued to finish its cars in Black Japan. Mr. Ford is reported to have said, “Paint them any color, as long as they are black.” Finally, in the 1930’s, Ford was forced to shift to a synthetic enamel finish of its own manufacture. During this transition period, du Pont sold Ford a substantial amount of finishes. In 1935, Ford was making half and buying half from du Pont; by 1937, Ford was making three-fourths and buying one-fourth from du Pont. In 1938, Henry Ford “issued instructions that the Ford Motor Company was not to purchase any more material from the du Pont Company.” From that time until Henry Ford II became active in Ford management, purchases from du Pont practically ceased. Since then, Ford has purchased finishes from du Pont in very substantial amounts.

General Motors has continued to test paints on thousands of cars annually. Du Pont has retained its position as primary lacquer supplier to several General Motors’ divisions because these divisions have felt that Duco best fits their needs. Kettering, who was a leader in General Motors’ research activities and who had been active in the testing and development of pyroxylin lacquers, testified that “one of the reasons” why General *635Motors’ cars had a higher resale value than comparable cars “in a used car lot” “is the paint.”

As the District Court found, “In view of all the evidence of record, the only reasonable conclusion is that du Pont has continued to sell Duco in substantial quantities to General Motors only because General Motors believes such purchases best fit its needs.” (Emphasis supplied.) 126 F. Supp., at 296.

The second largest item which General Motors buys from du Pont is Dulux, a synthetic enamel finish used on refrigerators and other appliances. Prior to the development of Dulux, Duco was widely used as a finish for refrigerators. However, in 1927, Duco began to be replaced by porcelain, particularly at Frigidaire, a General Motors’ appliance division. In 1930 and 1931, in collaboration with General Electric, du Pont developed Dulux, a greatly superior and cheaper product. Since its development, Dulux has been used exclusively by all the major manufacturers of refrigerators and other appliances — General Electric, Westinghouse, Crosley, and many others — except Frigidaire, which continues to finish part of its refrigerators with porcelain. Disinterested witnesses testified as to the superior quality and service which has led them to continue to buy Dulux.24 The District Court did not err in concluding that Dulux—

“is apparently an ideal refrigerator finish and is widely used by a number of major manufacturers *636other than General Motors. Several representatives of competitive refrigerator manufacturers testified that they purchased 100% of their requirements from du Pont. There is no evidence that General Motors purchased from du Pont for any reason other than those that prompted its competitors to buy Dulux from du Pont — excellence of product, fair price and continuing quality of service.” (Emphasis supplied.) 126 F. Supp., at 296.

The Court fails to note that du Pont’s efforts to sell paints other than Duco and Dulux to General Motors have met with considerably less success. Du Pont does sell substantial amounts of automotive undercoats to Chevrolet and Buick but it has failed, despite continued sales efforts, to change the preference of Fisher Body, the largest purchaser of undercoats, for a competitor’s undercoat. The successes and failures of other du Pont finish products at various General Motors’ divisions emphasize the independent buying of each division and negate the notion that influence or coercion is responsible for what purchases do occur. Frigidaire uses large quantities of black finishing and machine varnish, but has not bought these products from du Pont since 1926. At A C Spark Plug Division, located in Flint, Michigan, where du Pont has a finishes plant, du Pont has been consistently successful in selling a substantial volume of the finishes used by that division. Delco-Remy Division, however, *637purchases most of its requirements of insulating varnish from du Pont’s competitors. The Electromotive Division prefers a competitive lacquer for the interior finish of its locomotives, but uses Duco on the exterior because the railroads, most of which use Duco for the exterior of the balance of the train, specify that finish. At Guide Lamp Division, du Pont developed and still supplies a finish for the inside of headlight reflectors, but a competitor developed, and has kept, that division’s substantial primer business. At the Inland Division, which produces steering wheels, du Pont had some of the business. at one time, but has been -completely supplanted by a competitor offering better service.

The du Pont experience at the Packard Electric Division, which uses large quantities of high and low tension cable lacquer, is illustrative. Until 1932, Packard Electric was a separate company wholly unrelated to General Motors, and du Pont was a principal supplier of low tension lacquer and the sole supplier of black high tension lacquer. Now, as a division of General Motors, Packard Electric purchases its entire requirements of high tension lacquer from du Pont competitors, and produces its own low tension lacquer from film scrap bought from du Pont competitors.

The District Court did not err in concluding, on the basis of this evidence, that du Pont’s success in selling General Motors a substantial portion of its paint requirements was due to the superior quality of Duco and Dulux and to du Pont’s continuing research and outstanding service, and that “du Pont’s position was at all times a matter of sales effort and keeping General Motors satisfied. There is no evidence that General Motors or any Division of General Motors was ever prevented hy du Pont from using a finish manufactured by one of du Pont’s competitors; nor is there any evidence that *638General Motors has suffered competitively from its substantial use of Duco.” (Emphasis supplied.) 126 F. Supp., at 296.

2. Fabrics. — The principal fabrics which du Pont has sold to General Motors are imitation leather (du Pont’s “Fabrikoid” and “Fabrilite”) and top material for open cars and convertibles (du Pont’s “Pontop,” “Everbright” and “Teal”).25 Its sales of these materials to General Motors in 1947 totaled $3,369,000, or about 38.5% of General Motors’ total purchases of such materials. In earlier years, before closed cars with all metal tops came to predominate, these materials constituted a larger proportion of the total fabrics used in an automobile than they do today. By 1946 they averaged, apart from the top material for convertibles, only about 1.6 yards, costing about $2.22 per car. They are used principally for seat tops and backs, kick pads, rear shelves, etc. Du Pont does not manufacture the cotton and wool products of which most of the upholstery is composed.

Du Pont entered the manufacture of coated fabrics in 1910, when it purchased the Fabrikoid Company of New-burgh, New York. “Artificial leather,” as it was then *639known, was of poor quality and had very limited areas of acceptance. As du Pont succeeded in improving both its quality and appearance, its use rapidly broadened. By mid-1913, du Pont Fabrikoid, a pyroxylin-coated fabric, had been accepted by the automobile industry for upholstery and interior trim. Three years later, in 1916, almost every automobile company was a purchaser of Fabrikoid, and a contemporary du Pont estimate in that year stated that 60% of all cars produced in the United States would be equipped with Fabrikoid. In that same year, du Pont rounded out its line of fabrics by acquiring the Fairfield Rubber Company, a manufacturer of rubber-coated fabrics. Du Pont thus had achieved, before it purchased its General Motors' stock, a leading position in the automotive fabric field. Before 1917, it was supplying substantially all of the coated fabrics requirements at Chevrolet and Oldsmobile, about half of the requirements at Buick, and about a third of the requirements at Oakland. At the Cadillac division, du Pont supplied all of the coated fabrics for interior trim but none of the top material. 126 F. Supp., at 296-297.

Although there have been variations from year to year and from one car division to another in response to competitive considerations, du Pont generally has maintained its pre-1917 position as the principal supplier of coated and combined fabrics to General Motors. In 1926, General Motors purchased about 55.5% of these fabrics from du Pont, largely because Chevrolet switched entirely to du Pont after an unfortunate experience with competitive products during the preceding year. By 1930, the proportion had declined to about 31.5%, and du Pont was selling more fabrics to Ford than to General Motors. At the time of suit, du Pont’s share had increased to 38.5%, the remainder being supplied by du Pont’s competitors.

*640In addition to the mass of evidence supporting the District Court’s finding that “such purchases of fabrics as the General Motors divisions have made from du Pont from time to time were based upon each division’s exercise of its business judgment and are not the result of du Pont domination” (emphasis supplied), 126 F. Supp., at 301, the record clearly indicates that du Pont’s fabrics can and have made their way in the automotive industry on their merits. Prior to the early 1920’s, du Pont was the principal supplier of coated fabrics to all three of the then major producers — Ford, Willys-Overland and General Motors. After Ford and Willys began to produce their own coated fabrics they still turned to du Pont for much of what they could not produce. Chrysler purchased substantial amounts from du Pont until, in the early 1930’s, it embarked on its policy of one principal supplier for each product and chose Textileather, a du Pont competitor. Du Pont has continued to be Ford’s largest supplier for the material which it does not manufacture for itself. Du Pont likewise has supplied, over the years, a considerable part of the coated and combined fabrics of most of the smaller automobile companies.

The District Court did not err in concluding that “Du Pont, the record shows, has maintained its position as the principal fabric supplier to General Motors through its early leadership in the field and by concentrating upon satisfactorily meeting General Motors’ changing requirements as to quality, service and delivery.” (Emphasis supplied.) 126 F. Supp., at 301.

3. Other Products. — The Court concludes only that du Pont has been given an unlawful preference with respect to paints and fabrics. By limiting the issue to these products, it eliminates from deserved consideration those products which General Motors does not buy in *641large quantities or proportions from du Pont.26 Yet the logic of the Court’s argument — that the stock relationship between du Pont and General Motors inevitably has or will result in a preference for du Pont products— requires consideration of the total commercial relations between the two companies. Du Pont “influence,” if there were any, would be expected to apply to all products which du Pont makes and which General Motors buys.

However, the evidence shows that du Pont has attempted to sell to the various General Motors’ divisions a wide range of products in addition to paint and fabrics, and that it has succeeded in doing so only when these divisions, exercising their own independent business judgment, have decided on the basis of quality, service and price that their economic interests would best be served by purchasing from du Pont. Six such groups of products were considered in detail by the District Court: *642plastics, brake fluid, casehardening materials, electroplating materials, safety glass, and synthetic rubber and rubber chemicals. 126 F. Supp., at 319-324. A few examples drawn from the findings will suffice.

Du Pont’s sales to General Motors of celluloid (du Pont’s “Pyralin”), used as windows in the side curtains of early automobiles, initially declined in 1918 after the stock purchase, and only revived when an improved product was adopted by all the large auto manufacturers. Instead of purchasing brake fluid and safety glass from du Pont, General Motors embarked, during the 1930’s, on its own production of these substantial items. With respect to casehardening materials, General Motors has purchased less than half of its requirements from du Pont, while other auto manufacturers have purchased amounts larger in proportion and quantity. Although du Pont’s new electroplating processes were widely adopted in the automobile and other industries in the 1930’s only Cadillac has used du Pont’s processes exclusively, Oldsmobile and Pontiac have used it occasionally, and Chevrolet and Buick never have used it except for brief periods. Neoprene, a synthetic rubber developed by du Pont, has been used to a much greater extent by Chrysler and Ford than by General Motors. Chrysler also uses, and helped develop, du Pont’s synthetic rubber adhesive for brake linings, but the General Motors’ divisions prefer a more expensive type of synthetic rubber.

The record supports the conclusion of the District Court:

“All of the evidence bearing upon du Pont’s efforts to sell these various miscellaneous products to General Motors supports a finding that the latter bought or refused to buy solely in accordance with the dictates of its own purchasing judgment. There is no evidence that General Motors was constrained to favor, or buy, a product solely because it was offered *643by du Pont. On the other hand, the record discloses numerous instances in which General Motors rejected du Pont’s products in favor of those of one of its competitors. The variety of situations and circumstances in which such rejections occurred satisfies the Court that there was no limitation whatsoever upon General Motors’ freedom to buy or to refuse to buy from du Pont as it pleased.” (Emphasis supplied.) 126 F. Supp., at 324.

Evidence Relied on by the Court. — The Court, disregarding the mass of evidence supporting the District Court’s conclusion that General Motors purchased du Pont paint and fabrics solely because of their competitive merit, relies for its contrary conclusion on passages drawn from several documents written during the years 1918-1926, and on the logical fallacy that because du Pont over a long period supplied a substantial portion of General Motors’ requirements of paint and fabrics, its position must have been obtained by misuse of its stock interest rather than competitive considerations.

The isolated instances of alleged pressure or intent to obtain noncompetitive preferences are four: (1) the Raskob report of December 1917; (2) several letters of J. A. Haskell, written during 1918-1920; (3) certain reports and letters of Pierre and Lammot du Pont during 1921-1924; and (4) a 1926 letter of John L. Pratt. Passages drawn from these 1918-1926 documents do not justify the conclusion reached by the Court. Each of them is a matter of disputed significance which cannot be evaluated without passing on the motivation and intent of the author. Each failed to achieve its specific object. Read in the context of the situations to which they were addressed, each is entirely consistent with the finding of the District Court that, although du Pont was trying to get as much General Motors’ business as it could, there was no restriction on General Motors’ free*644dom to buy as it chose, and that General Motors’ buyers did not regard themselves as in any way limited.27 Moreover, even if isolated paragraphs in these documents, taken from their context, are given some significance, and *645the other evidence relating to the period from 1918 to 1926 is entirely ignored, all of the evidence after 1926 affirmatively establishes without essential contradiction that du Pont did not use its stock interest to receive any preferential treatment from General Motors.

Nor can present illegality be presumed from the bare fact that du Pont has continued to make substantial sales of several products to General Motors.28 In the first place, the record affirmatively shows that the new products which du Pont has sold to General Motors since 1926 have made their way, at General Motors as elsewhere, on their merits. Sales of Duco, Dulux, Fabrilite and Teal are not attributable in any way to dealings in the earlier period. Secondly, the Court’s presumption is based on the fact that du Pont does not sell to all other automobile manufacturers in the same proportion as it does to General Motors. But there is no reason why it should— the Government has not shown that sellers normally sell to all members of an industry in the same proportion. In any event, the record fully explains the disproportion. Since 1930, du Pont’s sales to other members of the industry have proportionately declined, largely because Ford has chosen to make the major share of its requirements of paint and fabrics, and because Chrysler has followed the policy of selecting a single supplier to whom it can be the most important customer. The fact is that du Pont has continued to sell in substantial amounts to the smaller members of the automobile industry. The growth in the *646dominance of General Motors, Ford and Chrysler — companies which together account for more than 85% of automobile production — when combined with the policies adopted by Ford and Chrysler, adequately explains why du Pont sells a larger proportion of paint and fabrics to General Motors than it does to the industry as a whole.

It is true that § 7 of the Clayton Act does not require proof of actual anticompetitive effects or proof of an intent to restrain trade. But these matters become crucial when the Court rests its conclusion that du Pont’s stock interest violates the Act on evidence relating solely to an alleged du Pont intent to obtain a noncompetitive preference from General Motors, and on a finding that such a preference was actually secured through the unlawful use of du Pont’s stock interest. Preference and intent are also relevant because the Government has brought this case 30 years after the event. If no actual restraint has occurred during this long period, the probability of a restraint in the future is indeed slight. Especially is this so when the only change in recent years has been in the direction of diminishing du Pont’s participation in General Motors’ affairs.

Rule 52 (a) Governs This Case. — The foregoing summary of the evidence relating to General Motors’ purchases of paint and fabrics from du Pont, comparatively brief as it is, reveals that a multitude of factual issues underlie this case. The occurrence of events, the reasons why these events took place, and the motives of the men who participated in them are drawn in question. The issue of credibility is of great importance. The District Judge had the opportunity to observe the demeanor of the witnesses and to judge their credibility at first hand. Thus, this case is a proper one for the application of the principle embodied in Rule 52 (a) of the Federal Rules of Civil Procedure, as amended, 329 U. S. 861: “Findings of fact shall not be set aside unless clearly erroneous, and *647due regard shall be given to the opportunity of the trial court to judge of the credibility of the witnesses.” United States v. Oregon Medical Society, 343 U. S. 326, 330-332, 339; United States v. Yellow Cab Co., 338 U. S. 338, 341-342.

This is not a situation in which oral testimony is contradicted by contemporaneous documents. See United States v. United States Gypsum Co., 333 U. S. 364. In this case, the findings of the District Court are supported both by contemporaneous documents and by oral testimony. For example, General Motors’ search for a better automotive finish, the superiority of the product developed by du Pont, and General Motors’ continuous efforts to secure an equally good lacquer from other sources are all proved by letters and reports written in the early 1920’s as well as by the oral testimony of many witnesses. Similarly, contemporaneous exhibits prove that General Motors purchased fabrics from du Pont because of the superiority of du Pont products, and that on other occasions it turned to competing suppliers even though du Pont’s product was just as good. Appellate review of detailed findings based on substantial oral testimony and corroborative documents must be limited to setting aside those that are clearly erroneous. The careful and detailed findings of fact of the District Court in this case cannot be so labeled.29

*648B. Relevant Market.

Finally, even assuming the correctness of the Court’s conclusion that du Pont’s competitors have been or will be foreclosed from General Motors’ paint and fabric trade, it is still necessary to resolve one more issue in favor of the Government in order to reverse the District Court. It is necessary to hold that the Government proved that this foreclosure involves a substantial share of the rele*649vant market and that it significantly limits the competitive opportunities of others trading in that market.30

The relevant market is the “area of effective competition” within which the defendants operate. Standard Oil Co. of California v. United States, 337 U. S. 293, 299-300, n. 5. “[Tlhe problem of defining a market turns on discovering patterns of trade which are followed in practice.” United States v. United Shoe Machinery Corp., 110 F. Supp. 295, 303, aff’d per curiam, 347 U. S. 521. “Determination of the competitive market for commodities depends on how different from one another are the offered commodities in character or use, how far buyers will go to substitute one commodity for another.” United States v. E. I. du Pont de Nemours & Co., 351 U. S. 377, 393. This determination is primarily one of fact.

The Court holds that the relevant market in this case is the automotive market for finishes and fabrics, and not the total industrial market for these products. The Court reaches that conclusion because in its view “automotive finishes and fabrics have sufficient peculiar characteris*650tics and uses to constitute them products sufficiently distinct from all other finishes and fabrics . . . .” Ante, pp. 593-594. We are not told what these “peculiar characteristics” are. Nothing is said about finishes other than that Duco represented an important contribution to the process of manufacturing automobiles. Nothing is said about fabrics other than that sales to the automobile industry are made by means of bids rather than fixed price schedules. Dulux is included in the “automobile” market even though it is used on refrigerators and other appliances, but not on automobiles. So are other finishes and fabrics used on diesel locomotives, engines, parts, appliances and other products which General Motors manufactures. Arbitrary conclusions are not an adequate substitute for analysis of the pertinent facts contained in the record.

The record does not show that the fabrics and finishes used in the manufacture of automobiles have peculiar characteristics differentiating them from the finishes and fabrics used in other industries. What evidence there is in the record affirmatively indicates the contrary. The sales of the four products principally involved in this case — Duco, Dulux, imitation leather, and coated fabrics — support this conclusion.

Duco was first marketed not to General Motors, but to the auto refinishing trade and to manufacturers of furniture, brush handles and pencils. In 1927, 44% of du Pont’s sales of colored Duco, and 51.5% of its total sales, were to purchasers other than auto manufacturers. Although the record does not disclose exact figures for all years, it does show that a substantial portion of du Pont’s sales of Duco have continued to be for nonautomotive uses.31

*651It is also significant that Duco was a patented product. Prior to the expiration of the patent in 1944, only five years before this suit was brought, du Pont issued over 250 licenses — to all that applied — covering its patented process. If Duco is to be treated as a separate market solely because of its initial superiority, du Pont is being penalized rather than rewarded for contributing to technological advance.

Dulux has never been used in the manufacture of automobiles. It replaced Duco and other lacquers as a finish on refrigerators, washers, dryers, and other appliances, and continues to have wide use on metallic objects requiring a durable finish. Yet the Court includes it as a finish having the unspecified but “peculiar characteristics” distinctive of “automotive finishes.” Ante, p. 593.

*652In 1947, when du Pont’s sales of Duco and Dulux to General Motors totaled about $15,400,000, the total national market for paints and finishes was $1,248,000,000, of which about $552,000,000 was for varnishes, lacquers, enamels, japans, thinners and dopes, the kinds of finishes sold primarily to industrial users.32 There is no evidence in this record establishing that these industrial finishes are not competitive with Duco and Dulux. There is considerable evidence that many of them are. It is probable that du Pont’s total sales of finishes to General Motors in 1947 constituted less than 3.5% of all sales of industrial finishes.

The record also shows that the types of fabrics used for automobile trim and convertible tops — imitation leather and coated fabrics — are used in the manufacture of innumerable products, such as luggage, furniture, railroad upholstery, books, brief cases, baby carriages, hassocks, bicycle saddles, sporting goods, footwear, belts and table mats. In 1947, General Motors purchased about $9,454,000 of imitation leather and coated fabrics. Of this amount, $3,639,000 was purchased from du Pont (38.5%) and $5,815,000 from over 50 du Pont competitors. Since du Pont produced about 10% of the national market for these products in 1946, 1947 and 1948, and since only 20% of its sales were to the automobile industry, the du Pont sales to the automobile industry constituted only about 2% of the total market. The Court ignores the record by treating this small fraction of the total market as a market of distinct products.

It will not do merely to stress the large size of these two corporations. The figures as to their total sales— *653$793,000,000 for du Pont and $3,815,000,000 for General Motors in 1947 — do not fairly reflect the volume of commerce involved in this case. The commerce involved here is about $19,000,000 of industrial finishes and about $3,700,000 of certain industrial fabrics — less than 3.5% of the national market for industrial finishes, and only about 1.6% of the national market for these fabrics. The Clayton Act is not violated unless the stock acquisition substantially threatens the competitive opportunities available to others. International Shoe Co. v. Federal Trade Commission, 280 U. S. 291; Transamerica Corp. v. Board of Governors, 206 F. 2d 163; V. Vivaudou, Inc. v. Federal Trade Commission, 54 F. 2d 273. The effect on the market for the product, not that on the transactions of the acquired company, is controlling. Fargo Glass & Paint Co. v. Globe American Corp., 201 F. 2d 534.33

The Court might be justified in holding that products sold to the automotive industry constitute the relevant *654market in the case of products such as carburetors or tires which are sold primarily to automobile manufacturers. But the sale of Duco, Dulux, imitation leather, and coated fabrics is not so limited.

The burden was on the Government to prove that a substantial share of the relevant market would, in all probability, be affected by du Pont’s 23% stock interest in General Motors. The Government proved only that du Pont’s sales of finishes and fabrics to General Motors were large in volume, and that General Motors was the leading manufacturer of automobiles during the later years covered by the record. The Government did not show that the identical products were not used on a large scale for many other purposes in many other industries. Nor did the Government show that the automobile industry in general, or General Motors in particular, comprised a large or substantial share of the total market. What evidence there is in the record affirmatively indicates that the products involved do have wide use in many industries, and that an insubstantial portion of this total market would be affected even if an unlawful preference existed or were probable.

For the reasons stated, I conclude that § 7 of the Clayton Act, prior to its amendment in 1950, did not apply to vertical acquisitions; that the Government failed to prove that there was a reasonable probability at the time of the stock acquisition (1917-1919) of a restraint of commerce or a tendency toward monopoly; and that, in any event, the District Court was not clearly in error in concluding that the Government failed to prove that du Pont’s competitors have been or may be foreclosed from a substantial share of the relevant market. Accordingly, I would affirm the judgment of the District Court.

*655APPENDIX TO MR. JUSTICE BURTON’S DISSENT.

“Sec. 7. That no corporation engaged in commerce shall acquire, directly or indirectly, the whole or any part of the stock or other share capital of another corporation engaged also in commerce, where the effect of such acquisition may be to substantially lessen competition between the corporation whose stock is so acquired and the corporation making the acquisition, or to restrain such commerce in any section or community, or tend to create a monopoly of any line of commerce.

“No corporation shall acquire, directly or indirectly, the whole or any part of the stock or other share capital of two or more corporations engaged in commerce where the effect of such acquisition, or the use of such stock by the voting or granting of proxies or otherwise, may be to substantially lessen competition between such corporations, or any of them, whose stock or other share capital is so acquired, or to restrain such commerce in any section or community, or tend to create a monopoly of any line of commerce.

“This section shall not apply to corporations purchasing such stock solely for investment and not using the same by voting or otherwise to bring about, or in attempting to bring about, the substantial lessening of competition. Nor shall anything contained in this section prevent a corporation engaged in commerce from causing the formation of subsidiary corporations for the actual carrying on of their immediate lawful business, or the natural and legitimate branches or extensions thereof, or from owning and holding all or a part of the stock of such subsidiary corporations, when the effect of such formation is not to substantially lessen competition.

“Nor shall anything herein contained be construed to prohibit any common carrier subject to the laws to regulate commerce from aiding in the construction of branches *656or short lines so located as to become feeders to the main line of the company so aiding in such construction or from acquiring or owning all or any part of the stock of such branch lines, nor to prevent any such common carrier from acquiring and owning all or any part of the stock of a branch or short line constructed by an independent company where there is no substantial competition between the company owning the branch line so constructed and the company owning the main line acquiring the property or an interest therein, nor to prevent such common carrier from extending any of its lines through the medium of the acquisition of stock or otherwise of any other such common carrier where there is no substantial competition between the company extending its lines and the company whose stock, property, or an interest therein is so acquired.

“Nothing contained in this section shall be held to affect or impair any right heretofore legally acquired: Provided, That nothing in this section shall be held or construed to authorize or make lawful anything heretofore prohibited or made illegal by the antitrust laws, nor to exempt any person from the penal provisions thereof or the civil remedies therein provided.” 38 Stat. 731-732, 15 U. S. C. (1946 ed.) § 18.