Full opinion text
FEINBERG, District Judge. This case involves the attempt of a manufacturer of a popular cigarette lighter to keep others from imitating the lighter’s shape and appearance. Plaintiff Zippo Manufacturing Company (“Zippo”), a Pennsylvania corporation, alleges both trademark infringement and unfair competition on the part of defendant Rogers, Inc. 2**(“Rogers”), a New York corporation, by reason of Rogers’ sale of pocket lighters closely resembling Zippo’s. Plaintiff seeks injunctive relief, an accounting, and damages. Rogers counterclaims for a judgment declaring that plaintiff is unfairly competing with it and also seeks injunctive relief. Jurisdiction of the main action is based both on diversity of citizenship and. a claim under the trademark laws joined with a related claim of unfair competition. Defendant also relies for jurisdiction on the Declaratory Judgments Act, 28 U.S.C. §§ 2201-2202 (1952), as amended, 28 U.S.C. § 2201 (Supp. IY, 1957). I Plaintiff Zippo has been primarily engaged in the manufacture of pocket lighters since 1932, and it has grown spectacularly over the years. Its annual national sales of these lighters grew from .about 27,000 units in 1934 to over 3,180,-000 in 1958, the year just prior to suit, and well over 4,000,000 in 1961. Today, Zippo produces more units than any other domestic lighter manufacturer. Its pocket lighters are made in two models, the “standard” and the “slim-lighter.” The latter accounts for slightly less than twenty-five per cent of the number of pocket lighters sold by Zippo. Prior to World War II, Zippo sold only a standard model, the forerunner of its present standard lighter. This pre-war lighter was covered in part by a patent. This model had a two-member outside case with a snugly fitting insert •containing the mechanical features of the lighter. The bottom member of the outside case had a matching flip-top member hinged to it; both were rectangular with square comers and straight lines. The user opened the top to activate the lighter by striking the flint wheel, and closed the top to snuff out the flame produced in the lighter’s “chimney.” To hold the flip-top in a fully opened or •closed position, the lighter incorporated a spring and cam arrangement which enabled the user to operate the top with •one hand. A most important feature of the lighter was an elliptical chimney or windscreen enclosing the wick, with round air holes punched into its sides in horizontal rows in a three-two-three formation. Attached to one side of the chimney were two square “ears” or lugs, which held the cam in position above a vertical coil spring located underneath the cam in a tube inside the lighter case. The flint wheel was pivotally mounted upon another set of lugs projecting from the other side of the chimney. Until 1938, the appearance of the lighter was that shown in the illustrations of the Gimera patent. In that year, Zippo slightly altered the internal mechanical operation of its lighter and, for reasons of economy and efficiency, also changed the shape of its lighter in some respects. No change was made in this standard model until 1942, when Zippo began to devote its entire output of lighters to the production of its “war” model. This lighter, embodying the same mechanical features of the immediately pre-war standard model, was finished in a crackle paint. In June 1946, Zippo again returned to "the production of its peacetime standard model and has continued to manufacture it without major change to the present time. A comparison of the present standard with the early Zippo standard model (1934-1938) indicates the following principal changes: the former model had severely square edges and corners and straight lines while the present lighter features slightly rounded edges and corners, a curvature in the shape of a slight arc in the top of the lighter, and a recess in the bottom of the case; in the early model, the leaves of the hinge attaching the top and bottom members of the lighter case were on the outside of the case, but they are now located inside the case; the ears or lugs holding the cam were previously square in shape, but now they are rounded; and while the cam and spring arrangement embodied a vertical coil spring in the old model, a horizontal flat “leaf” spring is now used. In June 1956, Zippo introduced its slim-lighter, a new model designed to appeal to women. This model included the same mechanical features as the standard, but it was noticeably narrower and, therefore, less square and more rectangular in shape. Its dimensions were approximately 2 Ys" x 1%" x compared to 2%" x V/z" x Yz" for the standard. Over the years, Zippo’s sales techniques have included extensive advertising, unconditional guarantees, and free repair service. Since 1949, the company has spent at least $500,000 each year in advertising its products. In 1958, it spent over $700,000, and in two of the next three years, over $1,000,000 per year. Its advertising has appeared in magazines and periodicals of both national and specialized appeal. Zippo’s advertisements of the standard model have featured, among other things, the word “windproof”; the mark “Slim-lighter” has been emphasized in advertising its slim-lighter Zippo’s unconditional guarantee and free repair service have also been highlighted. Throughout its history, Zippo has consistently promised to repair any Zippo lighter free of charge, regardless of its age or condition, and many purchasers have availed themselves of this service. For example, in 1961, almost 460,000 lighters were sent to the company for free repairs. Zippo markets eighty to ninety per cent of its standard and slim-lighters in individual cardboard boxes. These boxes are distinctively designed, and feature the name “Zippo.” Its lighters are primarily sold to the individual consumer in retail marketing channels, although approximately thirty-eight per cent of Zippo lighters are sold to commercial organizations which use the lighters for advertising or public relations purposes. Zippo also markets a very small percentage of its lighters through the use of display cards, which are changed to some extent each year. Two of plaintiff’s display cards introduced in evidence featured the name “Zippo” imposed in red letters upon a blue background. On one, the word “Windproof” appears in white letters; on the other, the words “Slim Lighter” appear in white lettering. Defendant Rogers or its predecessor has been in business since the first decade of this century. It offers nationally to the public a full line of smoking accessories, including pipes, tobacco pouches, and lighters. Rogers does not manufacture its lighters or pipes, but does produce its own tobacco pouches, which, according to the company’s president, account for seventy per cent of sales of pouches in the United States. The company has grown consistently over the years. In 1932, its gross sales approximated $13,000; in 1958, $1,570,000; and since then, according to its president, gross sales have been in the area of $2,000,000 yearly. Rogers, on a much more limited basis than Zippo, has advertised its products nationally in a variety of publications. From 1932 to 1958, Rogers spent over $1,335,000 on advertising on a gross sales figure of approximately $18,160,000; in 1958, it spent about $72,000 for advertising on sales of approximately $1,570,-OOO. In the fall of 1957, Rogers began to sell imported Japanese lighters with full knowledge that they closely resembled the Zippo lighters. Rogers has marketed these lighters with apparent success. In 1958, it sold between 240,000 and 360,000 lighters of this type; in 1959, approximately 720,000; in 1960, approximately 500,000; in 1961, approximately 380,000; and in 1962, approximately 360,000. Dollar volume from sales of these lighters accounted for about fifteen to eighteen per cent of Rogers’ total gross sales in 1959; in 1962, it accounted for about seven per cent of gross sales. A comparison of these imported “720” lighters with the Zippo models reveals a striking similarity in shape and appearance. Thus, referring to Zippo and Rogers standard models, the dimensions, the configuration, and the finish of the outside case of the lighters are almost the same; the windscreens, even with respect to the three-two-three arrangement of the airholes, are almost identical; the shape of the ears supporting the cam and flint wheel and the shape of the cam itself are the same; and the manner in which the lighters open and close is the same. In short, the Rogers standard is a Chinese copy of the Zippo, except for the inscriptions on the bottom of the lighters’ outside cases. Comparison of the Zippo and Rogers slim-light-' ers compels the same conclusion. Rogers markets these lighters exclusively by the use of display cards which carry the Rogers name. For example, on a 1958 display card, exhibiting the Rogers’ lighter resembling Zippo’s slim-lighter, the name “Rogers” appears in the upper left hand comer in large black logotype on a circular yellow background about 2% inches in diameter. The word “Slim” appears at the top of the card in tall black letters, and underneath, the word “Lighters.” A picture of a “thin” man dressed in evening wear appears on the right hand side of the display card extending from the top to 'the bottom of the card. In the lower right hand corner, the words “Made in Japan” are printed in very small black letters; similarly, in the lower left hand corner, the name “Rogers Imports, Inc.” appears; and printed directly above and across the entire bottom of the card are the words “Windproof — Unconditionally Guaranteed.” The display card is approximately fifteen inches high and ten inches wide. In similar display cards used in 1960, the word “Slim” was eliminated. Defendant has offered to print on its display cards this legend: “This is not a Zippo; this is a Rogers lighter.” The imported lighters in question were marked prominently with the name “Rogers” and the word “windproof” on the bottom of the lighter’s outside case. Commencing in late 1959 or early 1960, the marking on the bottom of the outside case was changed to read: “Rogers, Inc. New York, N. Y. Made in Japan” After Rogers commenced marketing its allegedly offending lighters in 1957, Zip-po began receiving Rogers lighters from consumers who wished to have them repaired through Zippo’s free repair policy. At the time of trial, a total of 191 Rogers lighters had been received by Zippo in this manner. Zippo’s policy was to return the lighter to the person from whom it was received, together with a form letter stating that the lighter was not a Zippo product, and, therefore, the company would not repair it. Zippo then sent questionnaires to a substantial majority of those persons who had sent Rogers lighters to it. Seventy-five of such questionnaires were introduced into evidence at the trial. The questionnaires were designed to find out in each case why the lighter was returned to Zippo even though it clearly bore the Rogers name. Zippo contends that the Rogers “720” lighters are inferior to those it manufactures. To buttress this contention, reports of tests conducted by Zippo on Rogers “720” lighters were introduced at the trial. On all the evidence, it appears that the Zippo lighters are superior to Rogers lighters, both as to material and workmanship. This is reflected in the retail price differential between the lighters. The retail price for the Rogers “720” model is $1.00; the retail price for the standard Zippo in recent years has been $3.50 and up, and for the Zippo slim-lighter $4.75 and up, depending on the model. Plaintiff introduced into evidence lighters manufactured by other companies to show that a lighter with the same windproof characteristics as a Zippo could be made without exactly simulating the Zippo appearance, e. g., the “Wind-lite,” manufactured by Ronson Metal Art Works, Inc. This lighter differs from the Zippo in shape, as well as ornamentation, and has thin vertical slits rather than round holes in its windscreen. Defendant introduced windproof lighters made by other manufacturers simulating the appearance of the Zippo, e. g., the “Storm King,” manufactured by Park-Sherman Co. Although Zippo has not sought injunctive relief against this company, it has taken steps in the past to stop the importation of Japanese lighters imitating the shape and appearance of Zippo models. Thus, Zippo has filed complaints with various agencies in this country and in Japan, and it has sent letters to ten different domestic concerns complaining about the importation problem and asking that it be terminated. Zippo has also brought other suits in federal courts seeking injunctive relief. II Plaintiff’s unfair competition action will be considered first, since, as indicated below, its claim of trademark infringement does not require extended comment. There does not appear to be any conflict between the parties as to the source of the applicable law. Determination of that question, if necessary, would be difficult, since plausible arguments have been advanced to show that state law alone or federal law alone, or some amalgam of both, is the proper source of doctrine. In this case, however, the parties have relied on both New York and federal cases, and do not argue that looking to one source rather than the other would bring about a significant difference in the content of the applicable law. Plaintiff asserts that its case rests, inter alia, upon the establishment of secondary meaning for the shape and appearance of its lighter. While there may be New York cases supporting the view that secondary meaning is not a prerequisite to establishing unfair competition in some situations, on the facts of this case, they would not seem applicable. Therefore, it is not necessary to pursue the matter further. Rather than, in Judge Medina’s words, “unnecessarily rushing in where angels fear to tread” to determine this complex question of the source of the applicable law, it will be assumed that “the indeterminate federal law” applies, with particular reference to expressions of that law in decisions of the Court of Appeals for this Circuit. What, then, is the content of the law? Plaintiff has quoted copiously from the Restatement of Torts, which states: “§ 741. Elements of Unprivileged Imitation. “One who markets goods, the physical appearance of which is a copy or imitation of the physical appearance of the goods of which another is the initial distributor, markets them with an unprivileged imitation,' under the rules stated in § 711 [general rule of liability], if his goods are of the same class as those of the other and are sold in a market in which the other’s interest is protected, and ■*•*•**** “(b) the copied or imitated feature has acquired generally in the market a special significance identifying the other’s goods, and “(i) the copy or imitation is likely to cause prospective purchasers to regard his goods as those of the other, and “(ii) the copied or imitated feature is non-functional, or, if it is functional, he does not take reasonable steps to inform prospective purchasers that the goods which he markets are not those of the other.” This is the view adopted in many cases in this Circuit. A classic statement appears in Crescent Tool Co. v. Kilborn & Bishop Co., 247 F. 299, at p. 300 (2 Cir. 1917), where Judge Learned Hand said: “The cases of so-called ‘nonfunctional’ unfair competition, starting with the ‘coffee mill case,’ Enterprise Mfg. Co. v. Landers, Frary & Clark, 131 Fed. 240, 65 C.C.A. 587, are only instances of the doctrine of ‘secondary’ meaning. All of them presuppose that the appearance of the article, like its descriptive title in true cases of ‘secondary’ meaning, has become associated in the public mind with the first comer as manufacturer or source, and, if a second comer imitates the article exactly, that the public will believe his goods have come from the first, and will buy, in part, at least, because of that deception. Therefore it is apparent that it is an absolute condition to any relief whatever that the plaintiff in such cases show that the appearance of his wares has in fact come to mean that some particular person — the plaintiff may not be individually known — makes them, and that the public cares who does make them, and not merely for their appearance and structure. It will not be enough only to show how pleasing they are, because all the features of beauty or utility which commend them to the public are by hypothesis already in the public domain. The defendant has as much right to copy the ‘nonfunctional’ features of the article as any others, so long as they have not become asso-dated with the plaintiff as manufacturer or source. The critical question of fact at the outset always is whether the public is moved in any degree to buy the article because of its source and what are the features by which it distinguishes that source. Unless the plaintiff can answer this question he can take no step forward; no degree of imitation of details is actionable in its absence.” In that case, the lower court had granted a preliminary injunction enjoining defendant from manufacturing and selling an adjustable wrench similar in shape and appearance to plaintiff’s. The Court of Appeals for this Circuit reversed on the ground that plaintiff had not proved that when defendant began to make its wrenches, “the general appearance of the plaintiff’s wrench had come to indicate to the public any one maker as its source, or that the wrench had been sold in any part because of its source, as distinct from its utility or neat appearance.” More recently, the Court of Appeals has expressed similar views. In American-Marietta Co. v. Krigsman, 275 F.2d 287, at p. 289 (2 Cir. 1960), the Court, again through Judge Hand, said: “The whole basis of the law of ‘unfair competition’ in this aspect is that no one shall sell his goods in such a way as to make it appear that they come from some other source. The simplest form of this is to use the name or trademark of another, but the law goes further than that. If the goods have certain features that have no significance as to the provenance of the goods, yet have come to have a ‘secondary meaning’ which may mislead buyers into supposing that they come from a source on which the buyers rely, those features may not be copied. That is the doctrine of ‘non-functional’ copying, and is limited to those features of the original goods that are not in any way essential to their use. If in addition to being ‘non-functional’ these features have indeed come to signify origin in the minds of buyers, a second comer in the field will be forbidden to incorporate those features into the design of his own wares; the chance that this will deflect buyers who would otherwise buy of the putative other source cancels the second comer’s privilege of reproducing the original, which he has in common with the public at large.” In that case, the Court affirmed the denial of a preliminary injunction against the sale and manufacture of a sponge mop, on the ground that plaintiff had not shown that the copied features of the mop were the reason for alleged confusion of origin among purchasers. In Hygienic Specialties Co. v. H. G. Salzman, Inc., 302 F.2d 614 (2 Cir. 1962), the Court of Appeals again similarly stated the applicable law of unfair competition. In that case, the District Court, 189 F.Supp. 790, had granted an injunction on the theory, inter alia, .that defendants had engaged in unfair competition by selling a soap dish similar in appearance to plaintiff’s soap dish. The Court of Appeals reversed, stating through Judge Kaufman, (302 F.2d at pp. 619-620): “* * * [T]he courts have repeatedly held that the copying of an unpatented design does not in itself constitute unfair competition. Algren Watch Findings Co., Inc. v. Kalinsky, 197 F.2d 69, 70 (2nd Cir. 1952); General Time Instruments Corp. v. U. S. Time Corp., supra [2 Cir.], 165 F.2d [853] at p. 854. “On the other hand, competition must be ‘fair’ both to the consumer, and so far as possible, to the trade. Consumer protection has been limited (in this context) to the prevention of confusion, i. e., the customer should not be misled into purchasing an article from one producer in the belief that it was made by someone else or emanates from some other (but unidentified) source. Norwich Pharmacal Co. v. Sterling Drug, Inc., supra, [2 Cir.] 271 F.2d [569] at p. 571. In order to recover on this theory, the plaintiff must show that a ‘special significance attaches in the public mind to the nonfunctional attributes of its wares’; or, in other words, that they have a ‘secondary meaning.’ Ibid.; American-Marietta Co. v. Krigsman, 275 F.2d 287, 289 (2nd Cir. 1960). To establish such secondary meaning in this case Hygienic must prove that (1) the soap dish design is a mark of distinction identifying its source, and, also that (2) purchasers are moved to buy the soap dish because of its source. Blisscraft of Hollywood v. United Plastics Co., 294 F.2d 697; Lucien Lelong, Inc. v. Lander Co., Inc., 164 F.2d 395, 397 (2nd Cir. 1947). This it has not done. * * * Hygienic has not shown that a single consumer bought one of these soap dishes because of its source. On the contrary, we agree entirely with the trial judge that ‘it is unlikely’ that a buyer would ‘give any thought at all to the identity .of the producer. Manifestly, a purchaser of this simple and inexpensive household item * * * relies not on the reputation of the producer, but upon the apparent quality and eye appeal of the product itself.’ 189 F.Supp. 794. And as Judge Learned Hand stated almost a half a century ago, “it is an absolute condition- to any relief whatever’ on this theory ‘that the public cares who does make them, and not merely for their appearance and structure.’ Crescent Tool Co. v. Kilborn & Bishop Co., 247 F. 299, 300 (C.C.A.2, 1917); Chas. D. Brid-dell, Inc. v. Alglobe Trading Corp., supra, 2 Cir., 194 F.2d 416.” Therefore, plaintiff can obtain relief only if it meets its burden of proving that: (1) Defendant’s lighter copies plaintiff’s lighter; (2) A copied feature has acquired a special significance in the market identifying plaintiff as the source of the lighter, and that purchasers are moved in any degree to buy the lighter because of its source (“secondary meaning”); (3) Such copied feature in defendant’s lighter is likely to cause prospective purchasers to regard the lighter as coming from plaintiff; (4) Such copied feature is nonfunctional. It should be noted, however, that even if the copied feature is functional, plaintiff may still be entitled to relief if defendant has not taken reasonable steps to set its lighter apart from plaintiff’s in the public mind. Kellogg Co. v. National Biscuit Co., 305 U.S. 111, 59 S.Ct. 109, 83 L.Ed. 73 (1938); Vaughan Novelty Mfg. Co. v. G. G. Greene Mfg. Corp., 202 F.2d 172 (3 Cir.), cert. denied, 346 U.S. 820, 74 S.Ct. 34, 98 L.Ed. 346 (1953). In the Kellogg case, after reaching a conclusion that plaintiff could not prevent defendant from making its product (shredded wheat biscuit) in the same pillow shape used by plaintiff, the Supreme Court pointed out (305 U.S. at p. 120, 59 S.Ct. at p. 114) that “[f]airness requires that [defendant’s manufacture of the biscuit] be done in a manner which reasonably distinguishes its product from that of plaintiff.” See Shredded Wheat Co. v. Humphrey Cornell Co., 250 F. 960 (2 Cir. 1918) ; cf. Modern Aids, Inc. v. R. H. Macy & Co., 264 F.2d 93 (2 Cir. 1959). The above analysis of the law of unfair competition is not meant to be exhaustive. Courts have also granted relief in unfair competition cases on the theory that other factors besides secondary meaning are controlling, such as deceptive marketing practices, American Safety Table Co. v. Schreiber, supra note 61; intention to palm off goods as the product of another, Upjohn Co. v. Schwartz, 246 F.2d 254 (2 Cir. 1957); or “violation of plaintiff’s property rights,” International News Service v. Associated Press, 248 U.S. 215, 39 S.Ct. 68, 63 L.Ed. 211 (1918). However, it is not necessary to consider the rationale of these and similar cases because these factors are not involved here. Rogers is an established company attempting to develop its own name and reputation, and is not attempting to trade on the name of its competitor. Plaintiff has argued that Rogers has copied a number of features of plaintiff’s method of doing business. However, the conduct complained of seems unobjectionable (e. g., offering a guarantee), unimportant, or legally defensible. Viewing defendant’s conduct as a whole, I find that it has not been guilty of intentionally deceptive marketing practices or any attempt to palm off its lighter as one manufactured by Zip-po. On the other hand, it has clearly attempted to obtain the commercial benefit of marketing a lighter of a certain shape and appearance popularized by Zippo. Whether Zippo can prevent it from doing so is the question to be decided under the law set forth above. Defendant argues, inter alia, that plaintiff must, in any event, prove actual deception or a clearly calculated plan to deceive in order to obtain relief. This appears to be an inaccurate statement of the law and an inversion of the cases cited above at p. 680. Those cases support the proposition that intention to deceive or actual deception might alone justify relief even if secondary meaning were not present. They represent, if anything, a relaxation of the requirements for a plaintiff in an unfair competition action, and do not purport to require proof of a plan by defendant to deceive in every case. Defendant also claims that plaintiff has not proved secondary meaning or confusion of source and that the imitated features are, in any event, functional. These and other contentions of defendant will be referred to below. Ill Plaintiff has relied heavily on a consumer study to prove the elements of its case. This study was prepared and conducted by the sampling and market research firm of W. R. Simmons & Associates Research, Inc. Mr. Simmons, the head of this firm, and Donald F. Bow-dren, the project supervisor, appeared as witnesses; both are qualified experts in the field of consumer surveys. Mr. Bowdren testified that the purpose of the study was to determine whether the physical attributes of the Zippo standard and slim-lighters serve as indicators of the source of the lighters to potential customers and whether the similar physical attributes of the Rogers lighters cause public confusion. The study or project consisted of three separate surveys. In Survey A, the respondents, or interviewees, were shown a Zippo standard lighter which had all the Zippo identification markings removed and were asked, among other things, what brand of lighter they thought it was and why. In Survey B, the same procedure was followed for the Zippo slim-lighter. In Survey C, respondents were shown a Rogers standard lighter that was being sold at the time of the survey, with all of its identifying markings, and they were asked, among other things, what brand of lighter they thought it was and why. Mr. Simmons’ testimony and the project report made clear the principles and procedures by which the surveys were conceived and conducted. Testimony to this effect is important, because it is well settled that the weight to be given a survey, assuming it is admissible, depends on the procedures by which the survey was created and conducted. Rhodes Pharmacal Co. v. Federal Trade Comm., 208 F.2d 382, 387 (7 Cir. 1953), rev’d in part on other grounds, 348 U.S. 940, 75 S.Ct. 361, 99 L.Ed. 736 (1955); see United States v. E. I. Du Pont De Nemours & Co., 177 F.Supp. 1, 18 (N.D. Ill.1959), vacated in part on other grounds, 366 U.S. 316, 81 S.Ct. 1243, 6 L.Ed.2d 318 (1961). There was no overlapping of respondents in the three surveys so that no one respondent would be influenced in one survey by his answers to another survey. The developmental phase of the project involved preparation of questions that could be handled properly by an interviewer, correctly understood by respondents and easily answered by them. This required several drafts of questionnaires and some pretesting. The “universe” to be studied consisted of all smokers aged eighteen years and older residing in the continental United States, which the research project indicated was approximately 115,000,000 (“the smoking population”). All percentage results in the surveys represent projected percentages of the smoking population. The three separate surveys were conducted across a national probability sample of smokers, with a sample size of approximately 500 for each survey. The samples were chosen on the basis of data obtained from the Bureau of Census by a procedure which started with the selection of fifty-three localities (metropolitan areas and non-metropolitan counties), and proceeded to a selection of 100 clusters within each of these localities— each cluster consisting of about 150-250 dwelling units — and then to approximately 500 respondents within the clusters. The manner of arriving at these clusters and respondents within each cluster was described in detail. The entire procedure was designed to obtain a representative sample of all smoking adults in the country. The procedures used to avoid sampling error and errors arising from other sources, the methods of processing, the instructions for the interviewers, and the approximate tolerance limits for a sample base of 500 were also described. Two of the interviewers testified that they were experienced in interviewing, explained the manner in which the interviews were conducted, and stated that they did not know the purpose of the surveys. All of the original responses to the questions as reported by these interviewers were made available in court. Plaintiff also called Dr. Robert C. Sor-ensen as an expert in the field of survey research. Dr. Sorensen stated that the project was conducted objectively and scientifically. Defendant does not deny this generally, but points to specific procedures and questions as being improper and buttresses its arguments with the testimony of its own expert, Professor Charles Winick. Defendant objects to the admission of the surveys into evidence. It first contends that the surveys are hearsay. The weight of case authority, the consensus of legal writers, and reasoned policy considerations all indicate that the hearsay rule should not bar the admission of properly conducted public surveys. Although courts were at first reluctant to accept survey evidence or to give it weight, the more recent trend is clearly contrary. Surveys are now admitted over the hearsay objection on two technically distinct bases. Some cases hold that surveys are not hearsay at all; other cases hold that surveys are hearsay but are admissible because they are within the recognized exception to the hearsay rule for statements of present state of mind, attitude, or belief. Still other cases admit surveys without stating the ground on which they are admitted. The cases holding that surveys are not hearsay do so on the basis that the surveys are not offered to prove the truth of what respondents said and, therefore, do not fall within the classic definition of hearsay. This approach has been criticized because, it is said, the answers to questions in a survey designed to prove the existence of a specific idea in the public mind are offered to prove the truth of the matter contained in these answers. Under this argument, when a respondent is asked to identify the brand of an unmarked lighter, the answer of each respondent who thinks the lighter is a Zippo is regarded as if he said, “I believe that this unmarked lighter is a Zippo.” Since the matter to be proved in a secondary meaning case is respondent’s belief that the lighter shown him is a Zippo lighter, a respondent’s answer is hearsay in the classic sense. Others have criticized the non-hearsay characterization, regardless of whether surveys are oifered to prove the truth of what respondents said, because the answers in a survey depend for their probative value on the sincerity of respondents. One of the purposes of the hearsay rule is to subject to cross-examination statements which depend on the declarant’s narrative sincerity. See Morgan, Hearsay Dangers and the Application of the Hearsay Concept, 62 Harv.L.Rev. 177 (1948). The answer of a respondent that he thinks an unmarked lighter is a Zippo is relevant to the issue of secondary meaning only if, in fact, the respondent really does believe that the unmarked lighter is a Zippo. Under this view, therefore, answers in a survey should be regarded as hearsay. Regardless of whether the surveys in this case could be admitted under the non-hearsay approach, they are admissible because the answers of respondents are expressions of presently existing state of mind, attitude, or belief. There is a recognized exception to the hearsay rule for such statements, and under it the statements are admissible to prove the truth of the matter contained therein. Even if the surveys did not fit within this exception, well reasoned authority justifies their admission under the following approach: the determination that a statement is hearsay does not end the inquiry into admissibility; there must still be a further examination of the need for the statement at trial and the circumstantial guaranty of trustworthiness surrounding the making of the statement. This approach has been used to justify the admissibility of a survey. Necessity in this context requires a comparison of the probative value of the survey with the evidence, if any, which as a practical matter could be used if the survey were excluded. If the survey is more valuable, then necessity exists for the survey, i. e., it is the inability to get “evidence of the same value” which makes the hearsay statement necessary. When, as here, the state of mind of the smoking population (115,000,000 people) is the issue, a scientifically conducted survey is necessary because the practical alternatives do not produce equally probative evidence. With such a survey, the results are probably approximately the same as would be obtained if each of the 115,000,000 people were interviewed. The alternative of having 115,000,000 people testify in court is obviously impractical. The alternatives of having a much smaller section of the public testify (such as eighty witnesses) or using expert witnesses to testify to the state of the public mind are clearly not as valuable because the inferences which can be drawn from such testimony to the public state of mind are not as strong or .as direct as the justifiable inferences from a scientific survey. The second element involved in this approach is the guaranty of trustworthiness supplied by the circumstances under which the out-of-court statements were made. A logical step in this inquiry is to see which of the hearsay dangers are present. With regard to these surveys : there is no danger of faulty memory; the danger of faulty perception is negligible because respondents need only examine two or three cigarette lighters at most; the danger of faulty narration is equally negligible since the answers called for are simple. The only appreciable danger is that the respondent is insincere. But this danger is minimized by the circumstances of this or any public opinion poll in which scientific sampling is employed, because members of the public who are asked questions about things in which they have no interest have no reason to falsify their feelings. While the sampling procedure substantially guarantees trustworthiness insofar as the respondent’s sincerity is concerned, other survey techniques substantially insure trustworthiness in other respects. If questions are unfairly worded to suggest answers favorable to the party sponsoring the survey, the element of trustworthiness in the poll would be lacking. The same result would follow if the interviewers asked fair questions in a leading manner. Thus, the methodology of the survey bears directly on trustworthiness, as it does on necessity. Since the two elements of necessity and trustworthiness are satisfied, I would admit these surveys under this approach to the hearsay rule, even apart from the state of mind exception. Defendant’s next objection to the surveys is that they should not have been conducted in respondents’ homes but in stores, the actual places of purchase. While it may be that in general the store is the best place to measure the state of mind at the time of purchase, it would be virtually impossible to obtain a representative national sample if stores were used. An interview at a respondent’s home is probative of his state of mind at the time of purchase, although the deviation from the actual purchase situation should be considered in weighing the force of this evidence. Therefore, the surveys are not inadmissible merely because they were conducted in homes. Defendant also objects to the surveys on the ground that they measured only the popularity of Zippo, as compared with the popularity of Rogers, and that this is not relevant to secondary meaning. However, I find that, by and large, the surveys did, as Dr. Sorensen testified, test brand identification and not brand popularity. They are, therefore, not incompetent because of the small elements of popularity which may have crept in, although that possibility should be considered in the weight to be given their results. Defendant’s next objection is to Survey C. In that survey, 34.7 per cent of respondents thought a Rogers lighter was a Zippo lighter, and plaintiff argues that this is probative of likelihood of confusion between the two lighters. However, respondents were not shown a Rogers display card when they made their mistaken identification. Therefore, defendant contends that the percentage of people who would mistakenly identify a Rogers lighter as a Zippo lighter would be smaller if the Rogers lighter were shown to them on a Rogers display card This argument is probably correct. This does not mean, however, that the results of Survey C are inadmissible to prove that the appearance of the Rogers lighter is likely to confuse people into thinking that it is a Zippo lighter. Defendant’s argument is implicitly based on two assumptions: (1) that use of the Rogers display card in the interviews would have caused a great number of the respondents who thought the Rogers lighter was a Zippo lighter to give different answers; and (2) that the number left who, even after a display card was shown to them, would still confuse the Rogers lighter with a Zippo lighter would be statistically insignificant. However, these assumptions at best are too speculative to require exclusion of Survey C. Moreover, they are contradicted by other answers in Survey C, which show that about one-half of respondents who mistakenly thought that the Rogers lighter was a Zippo lighter actually saw something stamped on the bottom of the lighter (where the Rogers name was imprinted), and that over one-third of those who thought the Rogers lighter was a Zippo lighter actually saw the Rogers name stamped on the lighter. These results certainly give rise to the inference that if the Rogers display card had been shown to respondents, a significant number of people would have thought the Rogers lighter was a Zippo lighter anyway: although exposed to the display card, some would not have actually perceived the Rogers name, and others would have seen the Rogers name but would nonetheless think that the lighter was made by Zippo. Therefore, Survey C is not excluded. However, its weight on the issue of likelihood of confusion is less than it would be had a display card been used, and had the same number of respondents nonetheless identified the Rogers lighter as a Zippo lighter. See W. E. Bassett Co. v. H. C. Cook Co., 164 F.Supp. 278, 285 (D.Conn.1958). Defendant has other objections to admissibility of the surveys, e. g., that a survey is not the best way to prove secondary meaning, but none of these merit further discussion. Surveys A, B, and C were scientifically conducted by a competent and professional research firm, substantially in accordance with the recommendations of Recommended Procedures for the Trial of Protracted Cases, 25 F.R.D. 351, 415-430 (1960). I conclude, therefore, that the surveys are admissible to show secondary meaning for the shape and appearance of the Zippo standard and slim-lighters and likelihood of confusion between these lighters and their Rogers counterparts. IV The elements of plaintiff’s case required to be proved will be considered against the background of the applicable law discussed above (Part II). 1. The first element — imitation—is clearly established. While Rogers does not itself copy the Zippo lighters, the lighters it purchases and then markets are, as indicated above, copies of the Zippo lighters. Plaintiff now claims that its lighter has been unfairly copied only as to its external shape and appearance and the windscreen. Plaintiff originally based its claim of unfair competition on alleged copying of other features of its lighter, such as the cover-holding cam, the ears holding the flint wheel, the cam lever spring, and the cutout at the bottom of its internal case. However, it conceded at the close of the trial that these features were functional and, therefore, not protected against copying. Even so limited, I find that the Rogers “720” models, in both their standard and slim-lighter sizes, imitate their Zippo counterparts. 2. The second element of plaintiff’s case is secondary meaning. This term has been called “a triumph of unintended obfuscation” since what it describes is whether a feature of an article, such, as here, its shape or appearance, has acquired a special significance which is anything but secondary in the mind of the purchaser, i. e., identification with source, and impulse to buy because of source. Turning first to the standard model, Survey A is the most important evidence as to the secondary meaning of the shape and appearance of this Zippo lighter. When shown an unmarked Zippo standard, 42.6 per cent of respondents in Survey A were “certain” it was a Zippo lighter. Of the 42.6 per cent who identified the Zippo lighter with its source, a substantial number clearly relied upon its shape, size, or appearance as the reason for identification. A relatively small percentage, 11.3 per cent, identified the unmarked Zippo standard as manufactured by someone else. 42.6 per cent is a significant percentage, indicating that a substantial number of people identify the shape and appearance of the lighter with Zippo. Defendant, as indicated above, minimizes the weight to be given to this survey’s conclusions. Its most persuasive argument in this context is that there is a large element of guesswork indulged in by respondents. A similar point was made by Judge Moore in Norwich Pharmacal Co. v. Sterling Drug, Inc., 271 F.2d 569 (2 Cir. 1959), cert. denied, 362 U.S. 919, 80 S.Ct. 671, 4 L.Ed.2d 739 (1960), where he said (271 F.2d at p. 572) that the results of the public opinion survey in that case “ * * * indicate no more than the popularity of [the product] .”, It is interesting to note that the survey In Norwich was conducted by the same firm involved in this case, and that the phrasing of the key questions in the instant survey was carefully formulated to reduce the importance of the element of popularity found in the Norwich survey. Thus, the first question in Survey A emphasized that respondents should not guess, and asked them to be “certain.” This was characterized by Mr. Simmons and Dr. Sorensen as a “hard” question designed to meet the objection made in the Norwich case. To the second and third questions asked of those who were not “certain” of the brand of the unmarked Zippo lighter, another 16 per cent responded Zippo. However, these questions allowed a greater possibility for guesswork. For this reason, I am not giving much weight to these answers on the issue of whether purchasers identified the Zippo lighter with its source because of its appearance, although the answers do suggest that there was such identification. However, Survey A otherwise supports this aspect of secondary meaning strongly enough, I feel, to alone justify such a finding. Also of some probative value, in my opinion, are Zippo’s vigorous advertising campaign, and the fact that the standard model has been sold in a quite similar form for about thirty years, and in almost identical form since 1946. Accordingly, I find that this element of secondary meaning has been established. The other aspect of secondary meaning required is that the source of the product in some degree motivates the purchase, i. e., the consumer cares about who made the product as distinguished from the pleasing appearance of the product itself. Crescent Tool Co. v. Kilborn & Bishop Co., 247 F. 299 (2 Cir. 1917); Restatement, Torts, Explanatory Notes § 741(b), comment h at 627 (1938). Here again the evidence of Survey A is crucial. With the fourth question of Survey A, a respondent was handed two lighters, one with the Zippo name on the bottom and the other with the Rogers name. He was then accurately told that one lighter was manufactured by Zippo and one by Rogers, and instructed to keep the one he desired. 66.9 per cent of respondents preferred the Zippo, compared to 28.7 per cent who chose the Rogers. A substantial portion of the former group based their choice on the fact that the Zippo company was associated with the lighter they chose. This ratio is clearly probative of the importance of source to respondents. In addition, there is similar probative value in some of the answers to the questionnaires sent by Zippo to those who had sent Rogers lighters to it for repair. Taking into account all the arguments mentioned above affecting the weight of the survey evidence, I find that secondary meaning in both its aspects has been proven as to the standard Zippo lighter as a whole. In this connection, it is important to point out that the copied features of the lighter break down into two categories: those associated with the lighter when closed (its external shape and appearance) and those discernible when the lighter is open (e. g., the windscreen). This division is significant with regard to the issue of functionality, as discussed below. On the record before me, I cannot say that plaintiff has met its burden of proof for secondary meaning specifically with regard to each category, but has only proved secondary meaning for both categories of the lighter’s features together without differentiating between them. This is so because Survey A did not call for answers specifically directed to each category. Interviewers gave a respondent an unmarked Zip-po standard and allowed the respondent to examine it in any way he chose. The interviewers’ notes indicate that a substantial number of respondents actually opened the lighter before stating whether they recognized the brand. Therefore, it is impossible to determine whether most respondents would have identified the lighter as a Zippo if they had been compelled to limit examination solely to the external shape and appearance of the lighter when closed or solely to its shape and appearance when open. Accordingly, I cannot find that secondary meaning has been proved for either category of features alone, although I do find that secondary meaning has been proved for the shape and appearance of the lighter as a whole. With regard to the Zippo slim-lighter, the situation is different. Plaintiff began marketing the slim-lighter in June 1956. By contrast, it began marketing the Zippo standard lighter on a substantial scale in 1934, and, as indicated above, although the appearance of this model was changed somewhat over the years, it has undergone no appreciable change since 1946. Defendant began selling its Rogers slim-lighter in the fall of 1957. Plaintiff has the burden of proving that the appearance of the slim-lighter had achieved secondary meaning within the relatively short period of a little over a year, when defendant began to sell its lighter. See Crescent Tool Co. v. Kilborn & Bishop Co., 247 F. 299, 300 (2 Cir. 1917); Remington Research, Inc. v. Modern Aids, Inc., 170 F.Supp. 7, 11 (S.D.N.Y.1959). Plaintiff appears to dispute this proposition since it now argues that the shape of the slim-lighter is so like the standard that any secondary meaning attaching to the standard over the years prior to 1956 would, of necessity, immediately attach to the slim-lighter as well. However, this point is not well taken. While there is similarity between the Zippo slim-lighter and standard models (e. g., the windscreen), the appearance of the two lighters is, in my opinion, sufficiently different to negate transferral of secondary meaning from the standard to the slim-lighter. The difference in dimensions is striking, and the slim-lighter must stand on its own feet with regard to any secondary meaning. Apart from the alleged carry-over from the standard model, plaintiff relies principally on Survey B to prove secondary meaning for the slim-lighter. The principal questions in this survey, which dealt with the slim-lighter, were identical to those in Survey A. I find that Survey B is important evidence, as was Survey A, but, for various reasons, is far less probative of the existence of secondary meaning. First, with an unmarked Zippo slim-lighter, only 25 per cent of respondents (as against 42.6 per cent for the unmarked Zippo standard) were “certain” that it was a Zippo lighter. It is true, as plaintiff points out, that in a recent decision a court accepted 25 per cent as “credible evidence of confusion on the part of a substantial percentage of the approximately 1,000 persons interviewed * * *.” However, for reasons hereafter set forth, the significance of the 25 per cent in this case is altogether different. Second, 8.9 per cent of the respondents were “certain” that the unmarked Zippo slim-lighter was a Ronson, and another 15.2 per cent attributed the lighter to Ronson although they were not “certain.” It is conceded by both parties that Ronson makes no lighter similar in appearance to either the Zippo or Rogers standard or slim-lighter. This remarkably high percentage (24.1) affirmatively and incorrectly identifying the Zippo slim-lighter with one well-known source was. almost equal to the percentage (25) correctly “certain” that it was a Zippo lighter. This indicates the greater possibility for guesswork with a shape that had not yet achieved a strong identification with source. Third, with the slim-lighter shape, the timing of the survey acquires great significance. The crucial date for Zippo to show secondary meaning is the fall of 1957, when it had been selling the slim-lighter for only a little over a year. The field work in the survey was accomplished in the period from July 30, 1960 to September 19, 1960, almost three years later. It is logical to assume that after those three years, with Zippo popularizing the slim-lighter in the interim, the percentages of respondents correctly identifying the shape would be substantially greater in 1960 than in 1957, when Rogers first began to compete. The 25 per cent who were “certain” that the slim-lighter was a Zippo lighter would undoubtedly have been less in 1957. Finally, it should be noted that on the key question of impulse to buy because of source, as compared to identification of source, Survey B, unlike Survey A, did not include a question which allowed respondents to choose between two lighters, one clearly marked Zippo and • the other clearly marked Rogers. Under all of these circumstances, I believe that Survey B is not sufficiently probative of secondary meaning for Zippo’s slim-lighter for plaintiff to have met its burden. Accordingly, I find that secondary meaning has not been established for the Zippo slim-lighter as of the time Rogers began to sell its own slim-lighter model. 3. The next element of plaintiff’s case is that the Rogers lighter is likely to cause prospective purchasers to regard it as a Zippo lighter. Plaintiff claims that this likelihood of confusion is established for the standard model principally by Survey C and the questionnaires. In Survey C, respondents were handed a standard Rogers lighter with all of its identifying marks, including the name on the bottom, and asked to identify its brand. 34.7 per cent of respondents thought it was a Zippo lighter, and only 14.3 per cent identified it as a Rogers. This would indicate that a substantial number of persons might confuse the Zippo and Rogers standard lighters because of their physical attributes, despite the appearance of the brand name on the lighters. Defendant’s objections to the weight of this evidence must, of course, be seriously regarded, particularly its argument that no such confusion would exist in the store at the time of purchase because Rogers lighters are sold from display cards bearing the name “Rogers.” This contention is buttressed by the fact that almost all of those who did correctly identify the Rogers lighter did so because they saw the brand name on the bottom, and the brand name would be more conspicuous on the display card. However, it is difficult to get away from the fact that the number of confused respondents was almost two and one-half times greater than those who recognized a Rogers as a Rogers, and a good portion of the confusion was due to the shape and style of the Rogers lighter. In addition, the answers to questionnaires discussed above clearly indicate that a substantial number of the consumers who sent Rogers lighters to Zip-po for repairs were confused by the similarity of the two brands, particularly by their shape and appearance. A total of 191 erroneous returns of this sort is a considerable number, even after making allowance for those who did not themselves purchase the lighter (e. g., received it as a gift) or those who knew that they were taking advantage of Zip-po's free repair policy. Evidence of likelihood of confusion accepted as significant in other cases does not appear to be more convincing. See, e. g., Harold F. Ritchie, Inc. v. Chesebrough-Pond’s, Inc., 281 F.2d 755 (2 Cir. 1960) (thirteen letters sent to plaintiff because of defendant’s advertisement). Def end-ant argues that plaintiff's evidence on this point is insufficient, since no person using ordinary care could be confused with the name “Rogers” imprinted on the bottom of a Rogers lighter. Defendant’s assumption that the unwary purchaser is not protected is open to question. But, in any event, I do not think warranted the further necessary assumption that the greater part of the 34.7 per cent who thought the Rogers lighter was a Zippo lighter or the 191 who returned a Rogers to Zippo were negligent. Such wholesale disregard of confusion does not seem justified. On all the evidence then, I find that plaintiff has established the likelihood of confusion — indeed actual confusion— between the Zippo and Rogers standard lighters. 4. The next element plaintiff must prove is that the copied features of its lighter are nonfunctional. But functional in this context can have many meanings. It can mean only those physical features of an article absolutely necessary to its use, e. g., the cutting edge of a razor blade as distinguished from its color, shape, or design. There is support for this strictly utilitarian view. For example, in Haeger Potteries, Inc. v. Gilner Potteries, 123 F.Supp. 261 (S.D. Cal.1954), the Court enjoined the copying of plaintiff’s distinctive art pottery ash tray because, inter alia, the copied features were nonfunctional. It said (at p. 271): “Whether an ash tray is round, oblong, egg-shaped, square, or of free form, makes no difference as to its utility or usefulness in holding ashes. * * * It is the utility to hold ashes, and not a precise shape or form, which is ‘the natural and characteristic action’ of an ash tray.” See McGill Mfg. Co. v. Leviton Mfg. Co., 43 F.2d 607, 608 (E.D.N.Y.1930). A broader definition of functionality is contained in the Restatement of Torts. Section 742 thereof states (at p. 629): “A feature of goods is functional * * * if it affect's their purpose, action or performance, or the facility or economy of processing, handling or using them; it is non-functional if it does not have any of such effects. “Comment: “a. A feature of goods, or of their wrappers or containers, may be functional because it contributes to efficiency or economy in manufacturing them or in handling them through the marketing process. It may be functional, also, because it contributes to their utility, to their durability or to the effectiveness or ease with which they serve their function or are handled by users. * * * “A feature is non-functional if, when omitted, nothing of substantial value in the goods is lost. * * * ” In Pope Automatic Merchandising Co. v. McCrum-Howell Co., 191 F. 979, 40 L.R.A..N.S., 463 (7 Cir. 1911), cert. denied, 223 U.S. 730, 32 S.Ct. 527, 56 L.Ed. 633 (1912), defendant appealed from a preliminary injunction prohibiting it from making and selling a vacuum cleaner identical to plaintiff's in mechanical principles, arrangement of parts, form, and color, but prominently displaying defendant’s name. Plaintiff did not deny defendant’s right to use the exact combination of interior parts if defendant had used a different form and color for the exterior parts. In setting aside the injunction on the ground that the shape of the vacuum cleaner was functional, the Court stated (191 F. at p. 981): “In short, [plaintiff] uses the most efficient and most economically manufactured form into which the mechanical combination can probably be embodied. * * * If [defendant] should be required to give a square or hexagonal or other cylindrical form to the outer surface of the casings, considerations of cost of the superfluous material and labor might prevent them from competing with [plaintiff] in the manufacture and sale of a mechanism that was equally open to both.” The concept of functionality may also be expanded to include anything which contributes to consumer appeal, i. e., a feature which gives the consumer a “substantial reason for purchasing the goods, rather than merely distinguishing them.” In Pagliero v. Wallace China Co., 198 F.2d 339 (9 Cir. 1952), the Court refused to enjoin defendant from copying ornamental designs on hotel china dishes because it classified them as functional. The Court recognized that one of the “essential selling features” of hotel china is its design, and stated (at p. 344): “[F]rom the standpoint of the purchaser china satisfies a demand for the aesthetic as well as for the utilitarian, and the design on china is, at least in part, the response to such demand.” In J. C. Penney Co. v. H. D. Lee Mercantile Co., 120 F.2d 949 (8 Cir. 1941), a similar test of functionality was applied to a “bib-pocket” design on overalls, where the Court stated (at p. 954): “[T]he term ‘functional’ is not to be treated as synonymous with the literal signification of the term ‘utilitarian’. A design, for example, may not be utilitarian in a technical sense, but it may nevertheless be functional in the sense that it will contribute materially to a general sale of the goods.” Cf. West Point Mfg. Co. v. Detroit Stamping Co., 222 F.2d 581 (6 Cir.), cert. denied, 350 U.S. 840, 76 S.Ct. 80, 100 L.Ed. 749 (1955); Vaughan Novelty Mfg. Co. v. G. G. Greene Mfg. Corp., 202 F.2d 172 (3 Cir.), cert. denied, 346 U.S. 820, 74 S.Ct. 34, 98 L.Ed. 346 (1953). The foregoing analysis of different interpretations of fun