On oral argument, plaintiff’s counsel contended that the alleged confusion resulting from the alleged similarity of the names was injurious to consumers. That contention embodies a frequently encountered misunderstanding of the doctrine of “unfair competition,” a misunderstanding which has led to those instances of undue extension of the doctrine on which plaintiff relies. Much of that misunderstanding seems to stem from the misleading use of the word “competition” in the label “unfair competition.” For, while competition has been cherished in part on the ground that it fosters character traits in competing businessmen deemed socially valuable, its basic virtue is generally regarded as consisting of its benefits to consumers. The magna carta of competition, Adam Smith’s The Wealth of Nations, made it clear that the consumer’s interests were to be the dominant aim of the competitive system: “Consumption,” wrote Adam Smith,1 “is the sole end and purpose of all production; and the interest of the producer ought to be attended to, only so far as it may be necessary for promoting that of the consumer. The maxim is so perfectly self-evident that it would be absurd to attempt to prove it.”2 But the legal protection of trade-names does not engender competition; on the contrary, it creates lawful monopolies, immunities from competition. And the legally forbidden invasions of those monopolies might often benefit consumers.- Thus, if a competitor of the manufacturer of a toothpaste with an established trade-name were to sell that identical toothpaste under that name but at half the usual price, the consuming public would be better off financially ; nevertheless such competition would, of course, be enjoined. In Benton Announcements, Inc. v. Federal Trade Commission, 2 Cir., 130 F.2d 254, we enforced *958an order of the Federal Trade Commission holding that it was “unfair competition” to describe as “engraving” a process the cost of which is cheaper than, and the end-product of which is practically indistinguishable from engraving; very likely the result of that order will be to set up resistance by consumers to the use of a new product less expensive to them.
The protection of such monopolies in names seems, then, to rest on the social interest in protecting primarily, not the consumer, but the businessman who has gained a strategic advantage, through building up of good-will, against unfair practices by competitors who desire to poach on that good-will. The public interest, from that point of view, is primarily in the preservation of honesty and fair dealing in business and in procuring “the security of the fruits of individual enterprise.”3 However, there is also the factor that the possibility of obtaining such monopolies as a reward for their enterprise may have the effect of inducing businessmen to bring out new products which may indirectly benefit the consuming public. A further factor is an assumed social interest in safeguarding consumers from being deceived even if the deception may be to their financial benefit. Federal Trade Commission v. Algoma Lumber Co., 291 U.S. 67, 78, 54 S.Ct. 315, 78 L.Ed. 655. There appears to be a related judicial policy of protecting snobbism; as this court recently put it, “People like to get what they think they are getting, and courts have steadfastly refused in this class of cases to demand justification for their preferences. * * * if the buyers wish to be snobs, the law will protect them in their snobbery.” Benton Announcements, Inc. v. Federal Trade Commission, supra [130 F.2d 255]. In other words, “the public is entitled to get what it chooses, though the choice may be dictated by caprice or by fashion or perhaps by ignorance.” Federal Trade Commission v. Algoma Lumber Co., supra [291 U.S. 67, 54 S.Ct. 320, 78 L.Ed. 655].4
To put the matter differently, there is a basic public policy, deep-rooted in our economy and respected by the courts, resting on the assumption that social welfare is best advanced by free competition, with the consequence that competition affords a privilege to do acts, resulting in financial harm to another person, which would be actionable in the absence of that privileged kind of activity; this privilege, as Holmes said, “rests on the economic postulate that free competition is worth more to society than it costs, and that on this ground the infliction of the damage is privileged.”5 But legal principles do not dwell a la Robinson Crusoe or in an anarchic state of nature in which there is a war of all against all; they must learn to live with one another, sociably, in a sort of democracy of principles in which none is dictator.6 So the privilege afforded by good-faith competition is limited by other conflicting public policy considerations, including that of discouraging certain kinds of business practices regarded as unfair; and among such unfair practices are those which mislead consumers even to their pecuniary benefit.
There are some persons, infected with monopoly-phobia, who shudder in the presence of any monopoly. But the common *959law has never suffered from such a neurosis. There has seldom been a society in which there have not been some monopolies, i.e., special privileges; the legal and medical professions have their respective guild monopolies; the owner of real estate, strategically located, has a monopoly; so has the owner of a valuable mine; and so have electric power companies. No one seriously questions whether there should be' some monopolies; the only question is as to what monopolies there should be, and whether and how much they should be regulated legislatively or curbed judicially.7
The protection of the interest of consumers is an ever-present factor in considering the allowable extent of monopolies in trade-names; this appears in the line of cases holding that a name which is initially entitled to immunity from competition loses that immunity when the name comes to be generally understood as a generic or descriptive designation for the type of goods in connection with which it is used,8 the very effectiveness of the publicity given to a name thus operating to destroy its monopolistic value because the monopoly has become too injurious to consumers.
The failure to keep constantly in mind the divers policy considerations which, in this legal province, come in conflict with one another and the consequent occasional over-emphasis on but one of them' — -the protection of the interest of the businessman who has built a business around a name — has sometimes led to decisions unduly extending the confines of name-monopolies. For a time the courts were remarkably generous in fixing the boundaries of such monopolies. Today the tendency is to be somewhat less generous. Cf. Durable Toy & Novelty Corp. v. J. Chein & Co., 2 Cir., 133 F.2d 853. We approach the case at bar, then, having in mind the basic common law policy of encouraging competition and the fact that the protection of monopolies in names is but a secondary and limiting policy.
Here the plaintiff chose to build its business around the use of a name “Chateau” which was previously in common use in advertising wines. It took the risk that competitors, old and new, would also use that common name. To be sure, plaintiff added the word “Martin,” and, so far as that word contributed a distinctive feature, plaintiff is undoubtedly entitled to a monopoly. But the distinctive monopoly, thus staked off by plaintiff, is not boundless. Already in use before plaintiff commenced business were “Chateau Mouton,” “Chateau Margaux,” “Chateau Mirat.” Plaintiff, therefore, could not and did not pre-empt every use of “Chateau” when coupled with every other name beginning with the thirteenth letter of the alphabet, nor even with one so beginning and consisting of six letters. “Chateau” is no more capable of being monopolized than is “hotel.” If plaintiff had advertised a hotel under the name “Hotel Martin,” no one would suppose Tiat it could enjoin defendant from running a hotel under the name “Hotel Montay.” An eye capable of distinguishing “Martin” and “Mouton” would surely not confuse “Martin” and “Montay”; and an ear which could differentiate “Mouton” and “Martin” — whether given an English or French pronunciation — could not fail to differentiate “Martin” and “Montay.”9
Although the plaintiff made diligent efforts, through an investigator, to find persons who had actually been misled by the alleged confusion of the two names, the evidence on that score which plaintiff obtained was so trifling and unconvincing that the trial judge found that “the evidence as to actual confusion and actual damage is too speculative to support a judgment for accounting.” He did find such “similarity in sound and appearance * * * as to make confusion of the two probable.” Apparently that finding *960was not based on the testimony as to confusion; if we assume that it was, then we must hold that finding “clearly erroneous.” If it was not based on that testimony but solely on the names themselves, we are in as good a position as the trial judge to determine the probability of confusion. For the reasons above noted, we believe there was no such probability. In so concluding, we are indeed confirmed by the inability of plaintiff, despite its diligence, to procure satisfactory evidence on that issue. The issue in such a case as this is “whether an appreciable number of prospective purchasers of the goods * * are likely” to be confused. “That a few particularly undiscerning prospective purchasers might be so misled is not enough.”10 The intent of the defendant in this kind of case is considered pertinent, first because if he .acts with the purpose of deriving a benefit from the plaintiff’s reputation, his purpose is some evidence that there is a confusing similarity,11 and second, perhaps, because the intention itself cuts into the privilege of competition. Here there was no direct evidence of such an intention, and it is significant that, although the plaintiff asserted the existence of a deliberate purpose to poach, the trial judge made no finding on that subject; nor do we see any basis in the evidence for an inference of such a purpose.
Reversed.
SWAN, Circuit Judge, dissents.
FRANK, Circuit Judge.
In its petition for rehearing, plaintiff for the first time urges that, since its trade-name is not registered, we must apply the state rule of decisions under Erie R. Co. v. Tompkins, 304 U.S. 64, 58 S.Ct. 817, 82 L.Ed. 1188, 114 A.L.R. 1487; asserts that the applicable rule here is that of Connecticut ; and now first directs our attention to the Connecticut decisions. Whether, in a suit in equity, the rule of Erie R. Co. v. Tompkins invariably applies has not yet been fully determined by the Supreme Court and is a question not altogether free from doubt, since the doctrine of the Erie case is now perhaps to be taken as founded upon the Judiciary Act of 1789 (28 U.S.C.A. § 725) which, by its literal terms, applies only to “trials at common law.” 1 But, assuming that the Connecticut decisions are controlling, there is no occasion for changing the decision at which we previously arrived. In Middletown Trust Co. v. Middletown National Bank, 110 Conn. 13, 147 A. 22, 25, the court stated: “No inflexible rule can be laid down as to what use of names will constitute unfair competition; this is a guestion of fact. * * * It is not sufficient that some person may possibly2 be misled, but the similarity must be such that any person, with such reasonable care and observation as the public generally are capable of using and may be expected to exercise, would be likely to mistake one for the other.” We find nothing in that or any other Connecticut decision suggesting that our previous conclusion was erroneous.3
We adhere to what we said in our original opinion, namely, that the use of Chateau was like the use of Hotel, and that no one would suppose that, if plaintiff had advertised a hotel under the name “Hotel Martin,” a court, at the suit of plaintiff, would enjoin a defendant from running a hotel *961under the name “Hotel Montay.” That was the fundamental basis of our ruling.
Plaintiff seizes on a subsidiary portion of our opinion in which we referred to the fact that, before plaintiff commenced business, there were already in use in the sale of wines the names Chateau Mouton, Chateau Margaux and Chateau Mirat; plaintiff points out that, as far as the record shows, Chateau Mouton was never sold in the United States or in Connecticut. Perhaps for that reason (although we need not here so decide) so much of our opinion as related to those two particular names should be ignored. The evidence does show, however, that wines were sold in Connecticut under the names Chateau Medallion, Chateau La Mission, Chateau Madrid and Chateau Margaux.4 But even if we were to disregard previous uses of “Chateau” in combination with names beginning with the letter M, there remains the controlling factor, above noted, that the use of Chateau was like the use of Hotel, and that there is not sufficient similarity between Martin and Montay, used in connection with Chateau, to justify an injunction.
The petition for rehearing is denied.