Laureyssens v. Idea Group, Inc.

OPINION SWEET, District Judge. Plaintiffs Dirk Laureyssens (“Laureys-sens”), I Love Love Company, N.Y. (“ILLC”), Creative City Limited (“CCL) and Extar Corp. (“Extar”) have moved for a preliminary injunction restraining defendant Idea Group, Inc. (“IGI”) from distributing certain toys which the Plaintiffs assert infringe Laureyssens’ copyrights and trade dress. For the following reasons, the motion is granted in part and denied in part. The Parties Laureyssens is a designer of various puzzles and toys including the puzzles which give rise to this dispute (the “Laureyssens Puzzles”). In the United States, the Lau-reyssens Puzzles have been sold under several names, most recently under the name “HAPPY CUBE.” ILLC is a Belgian corporation with its principal place of business in Belgium. ILLC manufactures the Laureyssens Puzzles in Europe and distributes them to certain European countries. ILLC has exported Laureyssens Puzzles for sale in the United States. CCL is Laureyssens’ worldwide representative for licensing the copyrights for the Laureyssens Puzzles. Extar is a California corporation headquartered in Los Angeles. Extar’s Chief Executive Officer is Raphael Berkien (“Berkien”). Extar is the “exclusive United States distributor” of the Laureyssens Puzzles. IGI is a California corporation with principal offices in Palm Desert, California. IGI was formed to develop and market consumer products including the puzzles at issue in this lawsuit (the “SNAFOOZ Puzzles”), and has been investigating the possibility of marketing various other products. Richard A. Frohman (“Frohman”) is the President and Chief Executive Officer of *1039 IGI and Robert D. Stevens (“Stevens”) its Executive Vice President. Proceedings The Plaintiffs’ motion was filed on June 5, 1991. Testimony was heard on July 1 and 2, and the matter was argued and fully submitted on July 10. THE FACTS 1. The Works at Issue At issue in this lawsuit are a series of foam rubber puzzles. The puzzles produced by both parties are described as “flat to cube” puzzles, meaning that their six pieces can be assembled in a flat form in a rectangular frame, and can also be assembled into a three-dimensional hollow cube. The cube is formed by joining the six pieces, the cube faces, with each piece being connected to its four neighbors at right angles. a. The interconnections The pieces are joined together by means of “notches” cut into the edges and “fingers,” the parts of the edge left when the notches are cut out. The fingers fit into the notches and the friction between the two holds the pieces together in the cube form. The notches and fingers are all cut at right angles to the piece edge arid perpendicular to the face of the piece, so that the fingers can interlock with notches in both a planar arrangement, with both pieces lying flat on one surface, and in the perpendicular arrangement necessary to form a cube. 1 For ease of reference, this system of connecting pieces in flat and cube form through the use of right-angled notches and fingers in the piece edges will be referred to as “rectilinear interconnections.” Obviously, in order to connect two pieces in either the flat or cube form, the notches on the edge of the first piece must match up with fingers on the adjoining edge of the second piece. The fundamental challenge of the puzzle is to select the proper pieces to join together and to join them in the correct orientation. While this may appear to be a simple or trivial task, in fact the puzzles, particularly the multi-puzzle combinations described in more detail below, are quite challenging. b. The dimensions The complexity of the puzzles is to a certain extent dictated by the number of notches and fingers on each edge. In order to join two pieces smoothly in the perpendicular alignment, the depth of the notches — and, correspondingly, the length of the fingers — must be equal to the thickness of the material of which the puzzle is made. Assuming that a single notch or finger is square, 2 as is the case in both parties’ puzzles, the length of each edge of the cube can be measured in “notch-widths” or “units,” one unit being equal to the thickness of the puzzle material. The Laureyssens Puzzles are all based on a cube which has five units per edge, while IGI’s puzzles form cubes which are six units in length. 3 2. Laureyssens, Puzzles Laureyssens first began working with cube puzzles in 1985, when he attempted to design a cube which could be assembled from six identically shaped pieces. After some experimentation, he realized that this was not possible, but he did produce a cube formed from five identical pieces and one piece which was only slightly different. Thereafter, Laureyssens began designing cube puzzles in which the pieces could *1040 also be fitted together in flat form in a two piece by three piece matrix. Of course, because of the fingers and notches on the pieces, this flat arrangement did not form a smooth rectangle, but by placing it inside an appropriately cut frame, Laureyssens created a puzzle which could be assembled into both a perfect cube and a flat rectangle. In fact, he designed a series of such puzzles, six of which are the source of this lawsuit (the “Laureyssens Puzzles”). Each was designed by Laureyssens to have a different level of difficulty, and he chose the pieces for the six puzzles so that not only could each individual puzzle be assembled in flat and cube form, but pieces from different puzzles could be assembled to create new cubes and other three-dimensional figures, including a “beam” of two or three cubes joined in a line, a double sized cube, with each side comprised of four pieces, a “cross” of five cubes, and a “star” of six cubes. In selecting the pieces for his puzzles, Laureyssens testified that he was trying not only to satisfy all of the foregoing criteria, but also to select pieces which had “balance,” or were aesthetically pleasing to his senses. In fact, he testified that he rejected a number of pieces because he did not like their appearance. The first commercial puzzles which Lau-reyssens produced consisted of a set of three of what would eventually be his six puzzles. All three were initially manufactured in white foam. Subsequently these puzzles were colored yellow, green, and blue, corresponding to the puzzles presently marketed in those colors. Laureyssens later added orange, red and purple puzzles to his line. Of the thirty-six puzzles pieces which make up the series of six Laureyssens Puzzles, there are thirty-four distinct piece shapes: two shapes are repeated, one in the red and blue puzzles, and one in the orange and purple. The six Laureyssens Puzzles are also identified by the names of various cities: the yellow is called “Tokio”, the orange is “Amsterdam,” green “New York,” blue “Milano,” purple “Brussels” and red “Paris.” Since he first began marketing these puzzles, Laureyssens has sold them under the names I-QUBE, COCOCRASH, CUBE-IT and HAPPY CUBE. 3. Laureyssens’ Attempts to Protect His Puzzles In the first phase of designing his puzzles, Laureyssens created about forty distinct puzzle pieces. He initially sought intellectual property protection for his designs by registering them in 1986 with the World Intellectual Property Organization (“WIPO”), an international organization headquartered in Geneva, Switzerland. Although there were more than one hundred pieces depicted in this registration, there are at least some pieces in the Laureyssens Puzzles which were not registered. 4 a. The 1987 Registration Laureyssens first sought United States copyright protection in 1987, when he personally visited the Copyright Office in Washington and prepared his own certificate of copyright registration, Certificate Number TXU 271 722 (the “722 Certificate”). This certificate contained the following entry: Nature of Authorship Briefly describe nature of the material created by this author in which copyright is claimed. SCULPTURE, TEXT, ARTWORK, PHOTOGRAPHES [sic] AND SAMPLE. The material which accompanied this registration certificate has not been produced by the Plaintiffs, although Laureyssens testified that he submitted as part of the application a white foam rubber version of the puzzle which is now the yellow Laureys-sens Puzzle. b. The 1988 Registrations In June 1988, Laureyssens sought to supplement his earlier registration, filing two new certificates, numbered TX 2 332 524 (the “524 Certificate”) and TX 2 332 525 (the “525 Certificate”). Both certificates *1041 reported the “Nature of Authorship” as “SCULPTURE, TEXT, ARTWORK, PHOTOGRAPHES [sic], & SAMPLES ADDED TO THE WORK.” Laureyssens identified the 722 Certificate as a previous registration for the works covered by both the 524 and 525 Certificates. Under the category of “DERIVATIVE WORK OR COMPILATION,” the 524 Certificate contained the following entry: 6.b. Material Added To This Work. Give a brief general statement of the material that has been added to this work and in which copyright is claimed. NEW TEXTS FOR ARENA; NEW DRAWINGS IQUBE, TATO, AUTOSAFARI; TEERLING AND PHOTOGRAPHES [sic] Attached to the 524 Certificate are twenty-four pages of written material which include diagrams describing how to assemble 120 different cubes as well as twenty-eight more complex puzzles, and several diagrams showing how some of the 120 cube puzzles could also be assembled in flat form. Among this latter category are all six of the Laureyssens Puzzles: the yellow puzzle is identified in the registration certificate as Number 4, the blue as Number 19, the green as Number 22, orange as Number 51, red as Number 55, and purple as Number 58. See 524 Certificate App. at 25. The material attached to the 524 Certificate also includes three photographs of empty frames and assembled and partially assembled cubes. On the 525 Certificate, Laureyssens listed the following as “Material Added To This Work:” TEXT EDUCATIONAL REPORT, NEW MODELS, NEW TEXT, COPY OF INTERNATIONAL MODEL DEPOSIT 1986. There are twenty-one pages of written material attached to the 525 Certificate. Most of the text of this material is in a foreign language (which appears to be Dutch). Included are sketches and drawings of over one hundred puzzle piece shapes, many of which are not pieces in any of the Laureyssens Puzzles. Also, as with the WIPO registration above, there are at least some pieces from the Laureys-sens puzzles which are not depicted in the 525 Certificate, again including the red puzzle piece from the HAPPY CUBE package. On one page there is a diagram indicating that the idea of the puzzles is to convert it from the flat arrangement to a cube. Interestingly, while the pieces displayed are those of the yellow Laureyssens Puzzle, the flat arrangement shown is not the one used for that puzzle. See 525 Certificate App. at 4. A later page includes a diagram of the pieces from the yellow puzzle assembled in yet a third distinct flat arrangement. Id. at 9. None of the other five Laureyssens Puzzles are depicted in the 525 Certificate. c. The 1991 Registrations In March 1991, Laureyssens filed six new certificates of registration, one for each of his puzzles (the “1991 Certificates”). Each certificate indicates that the “Nature of Authorship” is “Shape of pieces (Sculpture).” 5 Although none of the certificates originally referred to any of Laureyssens’ prior registrations, Laureyssens corrected this omission by six amendments filed in June 1991. Laureyssens has always marketed his puzzles with a proper notice of copyright attached. 4- The Plaintiffs’ Marketing Efforts Laureyssens first presented the Laureys-sens Puzzle for sale in the United States at the 1988 International Toy Fair held in New York City. Shortly thereafter, he entered an agreement with Sentinel International Packaging (“Sentinel”) to market the puzzles in the United States. This arrangement collapsed when Sentinel encountered financial difficulties, and Laureyssens decided to market the puzzles himself. a. Laureyssens’ Packaging The Laureyssens Puzzles were originally sold under the CUBE-IT name assembled in *1042 flat form in a clear plastic package with a cardboard insert. The name CUBE-IT was printed on the upper right corner of the insert with each letter a different color (thus incorporating all six puzzle colors) over a black background. The insert identified all six puzzles by color and by city name, although it did not explain that each individual puzzle could be assembled into its own cube, or that each puzzle was designed to have its own level of difficulty. The only explanation of the use of the puzzle was a picture of a person assembling a double size cube — four pieces per side — requiring pieces of all six colors. The reverse side of the insert, which could be read only after the package was opened, explained the various combinations and challenges, from the single color cube through the six-cube “star” configuration, although it did not mention the different levels of difficulty for the six puzzles. When the name was changed to HAPPY CUBE, because of a trademark conflict, the only aspect of the packaging which changed was the name. In the HAPPY CUBE packaging, each letter in the word “HAPPY” is a different color, and the word “CUBE” is entirely in the sixth color. The packages are displayed for marketing in a six-tiered cardboard “step display,” which features the multi-colored HAPPY CUBE logo on a black background. The display also features drawings of several multi-puzzle assemblies and a single picture of a one-color cube. b. Laureyssens’ Marketing Efforts Since the introduction of the Laureys-sens Puzzles into the marketplace, Plaintiffs have sold approximately 103,000 Lau-reyssens Puzzles in the United States and have orders for nearly an additional 250,-000. Approximately 90,000 of these puzzles were marketed in the HAPPY CUBE/ CUBE-IT packaging having rainbow-colored lettering against a black background. In addition to marketing the Laureyssens Puzzles under the HAPPY CUBE line as games of skill, the Plaintiffs have made efforts to sell the puzzles as novelty items, to be imprinted with corporate logos or promotional messages. In these marketing efforts, the Plaintiffs have produced puzzles in colors and designs other than the six colors used in the HAPPY CUBE line. Extar has organized a nationwide sales force which presently consists of forty sales representatives, each of whom is, according to Extar’s CEO Berkien, “committed to selling at least 100,000 Laureyssens puzzles per month,” June 4, 1991 Berkien Declaration at ¶ 5, but the nature of this commitment has not been explained. Ber-kien also stated that since 1988, Extar has sold more than $50,000 worth of Laureys-sens Puzzles in the United States, id., although during his June 11 deposition, taken one week after his declaration was signed, Berkien was unable to confirm this statement or to elaborate on Extar’s sales. Ber-kien Deposition at 65. Plaintiffs have advertised the Laureys-sens Puzzles in trade magazines such as the New York Toy Fair Directory and the National Conventioneer. Since Laureyssens’ original attempt to market his puzzles at the 1988 New York Toy Fair, Plaintiffs have exhibited the Lau-reyssens Puzzles at toy and novelty shows including the Dallas Toy Fair in 1989 and 1990, the New York International Toy Fair in 1990 and 1991, the Chicago Premium Show in 1990, and the New York Premium and Incentive Show in 1989. They intend to exhibit their product at the Dallas Toy Fair, the Chicago Premium Show, the Denver Toy Fair, and the Los Angeles Toy Fair in 1991. During the 1990 Dallas Toy Fair, the Laureyssens Puzzles were filmed and broadcast on a nationwide newscast by NBC Television, although there was no evidence concerning the type or extent of this coverage. The Plaintiffs have distributed their promotional leaflets showing the black and rainbow colored presentation of the HAPPY CUBE packaging at the aforementioned toy and premium shows. They had planned to launch a multimedia advertising campaign beginning on July 4, 1991, which was to include nationwide television advertising. The Laureyssens Puzzles have also been shown on the television show “Family *1043 Feud,” broadcast in early 1991. Plaintiffs have contracted with the National Football League to produce Laureyssens Puzzles bearing the logos of NFL teams on the cubes. Laureyssens testified that ILLC had spent approximately $80,000 in connection with the United States marketing of the puzzles at issue. Plaintiffs altogether claim to have expended approximately $180,000.00 in connection with the advertising and promotion of their Laureyssens Puzzles. In 1990, Extar entered into an agreement with Strottman International, Inc. (“Strott-man”), a supplier of premium and novelty products to packagers such as fast-food and breakfast cereal companies. The agreement granted Strottman exclusive promotional rights to the Laureyssens Puzzle in the premium industry. 5. IGI a. The Proumen/Miner Puzzles In April 1989, Frohman, in his then-capacity as package designer and salesperson for South Coast Packaging Associates, Inc., was approached by Richard Proumen (“Proumen”) and Gary Miner (“Miner”) to quote price estimates for display packaging for a series of flat foam rubber puzzles which they were marketing (the “PM Puzzles”). As it later developed, the PM puzzles were copies of the Laureyssens Puzzles. Frohman provided Proumen and Miner with price quotes for basic packaging of the PM puzzles, but advised them that the product would be more marketable if they used a different presentation with more elaborate packaging and graphics. After considering Frohman’s advice, Proumen and Miner advised him that they could not follow it due to a lack of financing. Froh-man, believing that the PM Puzzles could be a success, expressed an interest in licensing the puzzles from them. In or about August 1989, IGI was incorporated, primarily for the purpose of marketing the PM Puzzles. Since its incorporation, IGI has not marketed any product other than the SNAFOOZ puzzles at issue in this litigation, although it is presently investigating several new products. Negotiations between Proumen, Miner and Frohman resulted in a license agreement dated December 5, 1989 between The Arista Group and IGI, under which IGI was to have the exclusive United States marketing rights to the PM Puzzles in exchange for a royalty on sales and an agreement to sell a certain minimum number of the puzzles per year. IGI planned to market the puzzles in six different colors: red, yellow, blue, orange, green and purple— the same colors used by Laureyssens. According to Frohman, in the fall of 1989 he shopped around to determine whether or not there were any products on the market which would compete with the PM Puzzles. Although he did come across a number of items that assembled into a cube in some fashion, he believed that none of them were competitive with the PM Puzzles. He did not learn of the Laureyssens Puzzles. b. IGI’s Packaging IGI has marketed its SNAFOOZ Puzzles in several different packaging styles. All of its packages feature the SNAFOOZ logo in rainbow-colored lettering on a black background. The logo is displayed prominently across the top of the package in type which is quite distinct from that used by the Plaintiffs for their logo. The simplest packaging style consists of a puzzle assembled in its flat form wrapped in clear plastic “shrinkwrap” with a cardboard wraparound insert. As opposed to the Laureyssens packaging, the SNAFOOZ insert displays a picture of an assembled cube, and the back of the package includes directions and pictures of the different mul-ti-puzzle combinations which can be formed from the SNAFOOZ Puzzle pieces. IGI also markets SNAFOOZ puzzles in blister packs of one, three and six puzzles, with the SNAFOOZ logo printed on the cardboard backing in rainbow colors on a black background. The primary distinguishing feature of the blister packs is that one puzzle is already assembled in cube form in the package, offering a graphic and *1044 readily understandable image of the object of the puzzle. On the back of the cardboard backing are the same directions and pictures of the single and multi-cube assemblies as those on the individual flat package. IGI’s current packaging for its SNA-FOOZ Puzzles is identical to that which it designed for the PM Puzzles. c. The 1990 Toy Fair IGI had two or three dozen prototype PM Puzzles manufactured, as well as prototype packaging, for use at IGI’s display booth at the 1990 New York International Toy Fair. IGI also created and printed a promotional brochure for the fair. During the toy fair, IGI took several hundred dollars worth of orders for the PM Puzzles. During the 1990 Toy Fair, Laureyssens became aware that IGI was manufacturing and marketing puzzles which were similar to his. Upon investigating, he discovered that the puzzles offered by IGI were in fact identical to his own. Laureyssens objected to the IGI personnel at the booth, claiming that the PM Puzzles infringed the copyrights on his puzzles. On February 18, 1990, a cease and desist letter from ILLC was delivered by hand to the IGI booth at the New York Toy Fair. Also, on February 21, the last day of the fair, Plaintiffs forwarded a cease and desist letter directly to IGI, demanding that it immediately discontinue the manufacture and sale of the infringing SNAFOOZ puzzles. On March 19, 1990, IGI admitted to the Plaintiffs’ that “We acknowledge that our puzzles apparently were copied from a sample obtained through your French distributor or licensee and, in its [sic ] present form, cannot be marketed in the United States without your permission.” d. Negotiations Between IGI and Lau-reyssens Upon discovering that the PM Puzzles were, in fact, the Laureyssens Puzzles and that Proumen and Miner were not authorized to distribute and license the puzzles, Frohman attempted to negotiate a license with Laureyssens for United States distribution rights. After negotiations between IGI and Lau-reyssens failed, Frohman assured Laureys-sens and the other Plaintiffs that IGI would cease marketing the infringing puzzles. At some point in March, 1990 IGI also began to consider designing and marketing its own flat-to-cube puzzles, similar but not identical to those of Laureyssens. When it first consulted its attorney about the ramifications of designing puzzles which were based on six units per edge instead of the five units used by Laureyssens, the attorney cautioned IGI that “if five of your six segments match Laureyssens’ five segments it may constitute substantial similarity.” 6 Frohman testified that in early April of 1990 shortly before the negotiations between IGI and Laureyssens broke off, Lau-reyssens’ counsel indicated to IGI that if IGI produced its own flat-to-cube puzzles in which none of the individual pieces were copies of any of Laureyssens' pieces then Laureyssens would have no grounds for a claim of copyright infringement. e.IGI’s Design of the SNAFOOZ Puzzles After the breakdown of negotiations with Laureyssens, IGI proceeded with its plans to design and develop its own flat-to-cube puzzle, using six units per side to ensure that none of its puzzle pieces matched any of those used by Laureyssens. IGI clearly intended to copy the idea underlying the Laureyssens Puzzles, while at the same time making its puzzles different enough to avoid infringing Laureyssens’ copyright. In late April or early May 1990, Frohman approached a computer software programmer about the possibility of creating a series of six-segment-per-side puzzle designs *1045 similar to the Laureyssens Puzzles with multi-puzzle interconnectability and varying degrees of difficulty. The parties entered into an agreement, but the programmer subsequently sought more money and the agreement was cancelled. Thereafter, Stevens, through a friend, contacted Eric Brewer (“Brewer”), a graduate student in computer science at Massachusetts Institute of Technology, about creating a series of puzzles for IGI. On or about July 17, 1990, Stevens called Brewer and described to him the nature of the puzzle that IGI wished to develop, i.e., a flat-to-cube structure with six units per side, rectilinear interconnections and multi-puzzle interconnectability. In response to this description of IGI’s requirement, Brewer, prior to seeing any puzzle samples or prototypes, believed that he knew how to approach the problem, and told Stevens that he could write a program to generate the puzzles without much difficulty. On July 28, 1990, Stevens met with Brewer. He brought to the meeting two of the PM Puzzles (which IGI had admitted were copies of the Laureyssens Puzzles) which had been taken off the market after the New York Toy Fair. Brewer took these puzzles with him when he left the meeting. IGI and Brewer subsequently entered into an agreement effective August 4, 1990, the addendum to which set forth the parameters for the required six single puzzle designs and the additional criteria for more complex multiple puzzle structures. Other than understanding the general concept of the PM Puzzles and measuring the difficulty level of those puzzles by assembling and disassembling them, Brewer asserts that he made no use of the PM Puzzles in creating the SNAFOOZ Puzzles. On August 25, 1990, Brewer sent to IGI a letter describing six six-unit-per-side puzzle designs (the “SNAFOOZ Puzzles”) and including diagrams showing how the pieces could be assembled into cubes and other more complex structures, along with a computer disc containing all of his programming, including the source code. IGI thereafter arranged to have the new puzzles manufactured. From September 1, 1990 to February 1, 1991, IGI ordered and received approximately 300,000 SNAFOOZ Puzzles from its manufacturer in Korea. Associated packaging and display materials were manufactured locally. Although the Plaintiffs have suggested that the foam material of the SNAFOOZ Puzzles may contain toxic substances which would make it unsafe for children, there has been no evidence to support this supposition. Late in 1990; while IGI was in the process of preparing to launch the SNAFOOZ Puzzles at the 1991 New York Toy Fair, IGI’s counsel received a letter from Plaintiffs’ counsel dated November 13, 1990 requesting assurances from IGI that it had no intention of infringing the alleged copyrights in the Laureyssens Puzzles. By letter dated February 7, 1991, IGI’s counsel confirmed that IGI would no longer market the PM Puzzles. Immediately prior to the 1991 New York Toy Fair at which the SNAFOOZ Puzzles were exhibited for the first time, IGI conducted a direct mail campaign and placed advertisements in the 1991 New York Toy Fair Directory and in various trade publications for the SNAFOOZ Puzzles. Apart from a visit by Laureyssens to IGI’s showroom at the 1991 New York Toy Fair, Laureyssens made no effort to stop IGI from selling the SNAFOOZ Puzzles at the fair. At the end of the week of the Toy Fair, Laureyssens requested a meeting with IGI and himself. At the meeting, neither Lau-reyssens nor his counsel asked IGI to stop selling the SNAFOOZ Puzzles nor did they serve IGI with a cease and desist letter. The Issue These facts present the difficulties inherent in the protection of intellectual property: on the one hand, the protection of particular expressions, and on the other, the freedom to develop variations of an idea. As is all too frequently the case, both sides of the controversy are well-advised, even with conflicting advice, and to complicate matters even further, at issue is a puzzle, *1046 intended by its very nature to be difficult to understand and manipulate. Against the accepted background of the authorities with respect to preliminary injunction, unfair competition and copyright infringement, the determinative issue here is that of similarity or the lack thereof. Based on the considerations set forth below, it is concluded that there is some similarity between the trade dress and that the IGI puzzle, designed to differ from the Laureyssens Puzzles, is sufficiently dissimilar to avoid infringing Laureyssens’ copyrights. DISCUSSION 1. The Standard For A Preliminary Injunction The standard for a preliminary injunction in this Circuit is well-known. The plaintiff must demonstrate (a) irreparable harm; and (b) either (1) a likelihood of success on the merits; or (2) (a) sufficiently serious questions going to the merits; and (b) a balance of hardships tipping decidedly in its favor See Wallace International Silversmiths, Inc. v. Godinger Silver Art Co., 916 F.2d 76, 78 (2d Cir.1990) (citing Jackson Dairy, Inc. v. H.P. Hood & Sons, Inc., 596 F.2d 70, 72 (2d Cir.1979)), cert. denied, — U.S. -, 111 S.Ct. 1622, 113 L.Ed.2d 720 (1991). In cases of alleged copyright infringement, irreparable harm is generally presumed. Hasbro Bradley, Inc. v. Sparkle Toys, Inc., 780 F.2d 189, 192 (2d Cir.1985). As for the balance of hardships, IGI has offered virtually no evidence of the harm it would suffer if a preliminary injunction is issued. Based on the facts as found above, including the Plaintiffs’ recent marketing and advertising efforts, it appears that the balance of hardships tips decidedly in the Plaintiffs’ favor. The only remaining question is whether the Plaintiffs have established serious questions going to the merits of their claims. 2. Trade Dress The Plaintiffs claims of infringement of the Laureyssens Puzzle trade dress are based on § 43(a) of the Lanham Act, 15 U.S.C. § 1125(a) (1988), New York common law of unfair competition, and N.Y. General Business Law § 368-d. 7 a. Section 43(a) As the Second Circuit has recently explained: To prevail on a trade dress claim, the plaintiff must demonstrate that the product’s appearance has acquired “secondary meaning” — the consuming public immediately identifies the product with its maker — and that purchasers are likely to confuse the imitating goods with the originals. Even if the plaintiff establishes these elements, the defendant may still avoid liability by demonstrating that the imitated features are “functional”— essential to the basic purposes the article is meant to serve. Coach Leatherware Co. v. AnnTaylor, Inc., 933 F.2d 162, 168 (2d Cir.1991) (citations omitted). Because a showing of functionality by the defendant may obviate the need for a determination regarding secondary meaning, and because IGI here asserts that virtually all of the features of the Plaintiffs’ trade dress are functional, it is appropriate to address this issue first. i. IGI has not established that the Plaintiffs’ packaging form and style are functional. The Plaintiffs assert that their trade dress includes the colors of their puzzles, the packaging form, the packaging style, *1047 and the rainbow lettering on the package. IGI argues that at least the packaging of the puzzles assembled in flat form and the use of simple clear plastic wrapping are functional. The Plaintiffs claim that none of these features is necessary to the product’s use, that none of them affects either the cost or the quality and that “[tjhere are literally thousands of packaging forms and colors” that IGI could have used. As IGI correctly points out, there are in fact three basic ways to package the puzzle: assembled in flat form, assembled in cube form, and unassembled. IGI claims that because the puzzle is manufactured by cutting the pieces from the flat foam rubber which then becomes the frame, it is most economical simply to keep the pieces assembled flat in the frame until it is packaged and sold. However, IGI’s own packaging examples include several in which one puzzle is assembled in cube form. At the very least, this is sufficient to establish a serious question as to whether the assembled flat form ought to be considered functional. It follows that there is at least a serious question as to whether the Plaintiffs’ choice to package their puzzles in clear cellophane with a cardboard insert is functional. Again, IGI’s own samples suggest that other packaging methods are feasible. As for the colors, IGI asserts that it was forced to choose colors similar to the Plaintiffs’ because they were among the few colors which its manufacturer could produce which satisfied the government’s toxicity standards. Moreover, it is quite likely that anyone choosing colors for six items would choose the three primary colors and the three secondary colors, as both parties have done here. Nevertheless, the Plaintiffs are correct in asserting that there are numerous possible colors, and IGI has not carried its burden of showing that the particular shades and hues chosen by the Plaintiffs are necessary or vital to the product’s use or that they affect the cost or quality sufficiently to be considered functional. Finally, as to the rainbow lettering, IGI makes no serious claim that this aspect of the Plaintiffs’ trade dress is in any way functional. ii. Secondary Meaning In Coach Leatherware, supra, the Court of Appeals defined the standard for showing secondary meaning as follows: The trade dress of a product attains secondary meaning when the purchasing public “associates” its design with a single producer rather than simply with the product itself. The plaintiff is not required to establish that all consumers relate the product to its producer; it need only show that a substantial segment of the relevant consumer group makes this connection. Id. (citations omitted; emphasis in original). The Plaintiffs bear the ultimate burden of establishing secondary meaning. Thompson Medical Co. v. Pfizer, Inc., 753 F.2d 208, 217 (2d Cir.1985); Eli Lilly & Co. v. Revlon, Inc., 577 F.Supp. 477, 483 (S.D.N.Y.1983). Factors considered in determining whether a mark has acquired secondary meaning include advertising expenditures, consumer surveys, sales success, unsolicited media coverage, attempts to copy the mark, and length and exclusivity of use. Thompson v. Pfizer, supra, 753 F.2d at 217. On the facts as found above, the Plaintiffs have not met their burden of establishing a serious question as to whether their trade dress has acquired secondary meaning. While the total advertising expenditures have been reported at $180,000, this number was not substantiated by evidence as to how much was spent on each of the various advertising and promotional campaigns. For example, Laureyssens was unable to testify to the amount spent on the promotional television show “Incredible Breakthroughs,” and he indicated that he believed the amount spent on the “Family Feud” promotion was about $5,000. In addition, there was evidence that the Plaintiffs’ figures included salary payments to Laureyssens and his wife. With respect to sales success, while the Plaintiffs asserted that Extar’s sales since 1988 had been in excess of $50,000, neither *1048 Berkien nor Laureyssens could substantiate this claim. The only media coverage of the Plaintiffs’ products, aside from the promotions on “Family Feud” and “Incredible Breakthroughs” — which are properly classified as advertising — is the NBC coverage of the Laureyssens Puzzles at the Dallas Toy Fair in 1991. However, Laureyssens did not describe this coverage either by type of coverage or length in enough detail to permit the conclusion that it was significant. With respect to attempts to copy Plaintiffs’ trade dress, IGI has presented evidence that it created most of its packaging and its trade dress prior to the 1990 New York Toy Fair at which it first learned of Laureyssens and his products. Finally, as to the length and exclusivity of Plaintiffs’ use, they have used their trade dress since early 1990, at a series of toy fairs, and for at least some of that time IGI has been marketing its own products in its allegedly infringing trade dress, so there has not been a lengthy period in which the Plaintiffs have had exclusive use. Therefore, the Plaintiffs have not established a serious question of whether their trade dress has acquired secondary meaning. Hi. Secondary meaning in the making. Alternatively, the Plaintiffs claim that they have shown a serious question as to whether their trade dress possesses secondary meaning “in the making.” The theory underlying this claim is that “where secondary meaning is ‘in the making’ but not yet fully developed, a trademark or trade dress will be protected against intentional, deliberate attempts to capitalize on a distinctive product.” Metro Kane Imports, Ltd. v. Federated Department Stores, Inc., 625 F.Supp. 313, 316 (S.D.N.Y.1985) aff'd mem. 800 F.2d 1128 (2d Cir.1986); see also Jolly Good Industries, Inc. v. Elegra, Inc., 690 F.Supp. 227, 230-31 (S.D.N.Y.1988); but see Cicena, Ltd. v. Columbia Telecommunications Group, 900 F.2d 1546, 1549-50 (Fed.Cir.1990) (concluding that Second Circuit would reject doctrine of secondary meaning in the making). In this regard, while there is evidence that IGI adopted its trade dress before it had even heard of Laureyssens, the evidence does not indicate when IGI developed its flat-form shrink-wrapped package, and therefore the Plaintiffs have barely established a serious question of IGI’s “intentional, deliberate attempts” to copy their trade dress. Therefore, the Plaintiffs have shown evidence which is just sufficient to establish a question as to whether their trade dress should be protected under the doctrine of “secondary meaning in the making,” and have satisfied the first element of their Lanham Act claim. iv. Likelihood of Confusion In deciding the issue of likelihood of confusion in Lanham Act actions, the Second Circuit has long relied on the multi-factor balancing test set forth by the Honorable Henry Friendly in Polaroid Corp. v. Polarad Electronics Corp., 287 F.2d 492, 495 (2d Cir.), cert. denied, 368 U.S. 820, 82 S.Ct. 36, 7 L.Ed.2d 25 (1961). Polaroid’s multi-factor balancing test considers: the strength of [the plaintiff’s trade dress], the degree of similarity between the two products, the likelihood that the prior owner will bridge the gap, actual confusion, and the reciprocal of defendant’s good faith in adopting its own mark, the quality of defendant’s product, and the sophistication of the buyers. Id. at 495. No single factor is determinative; rather, all factors must be considered together according to the degree to which each is implicated. Lambda Electronics Corp. v. Lambda Technology, Inc., 515 F.Supp. 915, 925 (S.D.N.Y.1981). Each factor must be evaluated in the context of how it bears on the ultimate question of the likelihood of confusion as to the source of the product. Lois Sportswear, U.S.A., Inc. v. Levi Strauss & Co., 799 F.2d 867, 872 (2d Cir.1986). (1) Strength of the mark The Second Circuit has said that: *1049 [t]he term “strength” as applied to trademarks refers to the distinctiveness of the mark, or more precisely, its tendency to identify the goods sold under the mark as emanating from a particular, although possibly anonymous source. McGregor-Doniger Inc. v. Drizzle, Inc., 599 F.2d 1126, 1131 (2d Cir.1979). Laureyssens’ trade dress—a flat rectangle of colored foam rubber wrapped in a clear plastic wrapper with a simple cardboard insert—cannot be considered “distinctive,” and the Plaintiffs have presented no evidence that it identifies Laureyssens Puzzles as emanating from a particular source. (2)Similarity of the products and proximity of the markets Clearly, the products sold by both parties are similar; indeed, for trade dress purposes they are virtually identical. Likewise, they are marketed to the same consumers, whether defined as the intermediate retailers or the ultimate end-users. (3)Sophistication of the buyer This factor clearly depends upon who is considered the buyer of the product: the reseller who purchases it from the Plaintiffs or IGI, or the end consumer. As virtually all of the Plaintiffs’ marketing efforts to date have dealt with sales to resellers and promotional marketers, this appears to be the more appropriate market to consider. Given this definition, the buyers of these products, toy industry professionals, must be considered sophisticated. (4)Quality of the products While the Plaintiffs have argued that the SNAFOOZ Puzzles are of inferior quality and that they are made of material which is more toxic than the Laureyssens Puzzles, the evidence establishes that both parties’ products are of comparable quality. (5)Good faith of IGI The parties dispute the good faith of IGI in choosing the colors in which the SNA-FOOZ Puzzles would be manufactured and in adopting a multicolored logo on a black background. While IGI has presented evidence that its color choices were limited by the need to keep the puzzles non-toxic and while there is evidence that IGI adopted its logo before it knew anything about Lau-reyssens, there is at least a factual dispute on these issues sufficient to establish a question as to IGI’s good faith. (6)Actual confusion On the issue of actual confusion, the Plaintiffs have presented little evidence that any of the buyers of these products has been confused. Indeed, the Plaintiffs complain that several potential customers have reported that IGI is selling products similar to the Laureyssens Puzzles, which indicates that the buyers have not been confused as to the source of the SNAFOOZ Puzzles at all. A visual comparison of the Plaintiffs’ products and those of IGI also supports the conclusion that, at least with respect to IGI’s blister packaging there is little likelihood of actual confusion between the two. The Laureyssens Puzzles are marketed in a flat form wrapped in clear plastic with a simple cardboard insert, while the SNAFOOZ Puzzles are marketed in single and multiple puzzle blister packs with one puzzle assembled in cube form. Notwithstanding the virtually identical colors of the puzzles and the similar coloring of the logos, the two products are sufficiently different that there is not a serious question as to actual confusion of the two. However, with respect to IGI’s flat form, clear plastic packaging, visual comparison with the HAPPY CUBE packaging leads to a different conclusion. With respect to this version of the SNAFOOZ trade dress, the similarity in the size and color of the products, the similar coloring of the logos, and the overall simplicity of the clear plastic packaging combine to create a serious question of actual confusion. In summary, consideration of all of the Polaroid factors leads to the conclusion that there is no likelihood of confusion between the HAPPY CUBE trade dress and the SNAFOOZ trade dress with respect to the blister packaged SNAFOOZ Puzzles, but that there is a question of likelihood of confusion with respect to the flat form shrink-wrapped SNAFOOZ Puzzles. Therefore, the Plaintiffs’ request for a pre *1050 liminary injunction under § 43(a) is granted to the extent of restraining IGI from marketing its puzzles in the flat form packaging. b. Unfair Competition With respect to their state law claim for unfair competition, the Plaintiffs argue that they need not establish secondary meaning in order to prevent IGI from using confusingly similar trade dress. In support of this claim, the Plaintiffs cite Perfect Fit Indus., Inc. v. Acme Quilting Co., 618 F.2d 950 (2d Cir.1980). While that case does seem to support the Plaintiffs’ position, more recent authority from the Second Circuit has limited those instances in which secondary meaning need not be shown for a claim of unfair competition. In 815 Tonawanda Street Corp. v. Fay’s Drug Co., 842 F.2d 643 (2d Cir.1988), the court stated Under New York law, secondary meaning is ordinarily a required element of an unfair competition claim; under certain circumstances, however, it need not be proven. For example, “[t]here is no need to show ... secondary meaning, if instances of actual palming off and deception are involved.” Id. at 649 (quoting Polo Fashions, Inc. v. Extra Special Products, Inc., 451 F.Supp. 555, 563 (S.D.N.Y.1978); citations omitted). The court reversed the grant of a permanent injunction on the plaintiff's unfair competition claim, holding that there was no evidence that the defendant’s use of a mark similar to the plaintiff’s “had been intended to trade on goodwill developed by and for the plaintiff,” and that the defendant had not “acted unfairly to usurp the plaintiff’s mark, name and business.” Id. Finally, the court pointed out that even the plaintiff admitted that “this is not a case of defendant trying to borrow plaintiff's established business reputation for its own use.” Id. The Second Circuit has also explained that The essence of an unfair competition claim under New York law is that the defendant has misappropriated the labors and expenditures of another. Central to this notion is some element of bad faith. Saratoga Vichy Spring Co. v. Lehman, 625 F.2d 1037, 1044 (2d Cir.1980) (citations omitted). As discussed above, IGI has introduced evidence that it selected its trade dress before it knew of Laureyssens, which would suggest that the choice was not based on any improper motive. Indeed, several the findings cited above from 815 Tonawanda v. Fay’s Drug seem equally applicable to the trade dress claims at issue here. However, as there is insufficient evidence to establish exactly when IGI adopted its flat form packaging, the Plaintiffs have raised a question as to IGI’s good faith in this respect. In addition, the doctrine of secondary meaning “in the making” seems equally applicable in the unfair competition context, and therefore the Plaintiffs have satisfied the first element of their unfair competition claim. With respect to the likelihood of confusion, the analysis above applies equally well in the unfair competition context, and therefore Plaintiffs’ request for a preliminary injunction on the grounds of common law unfair competition is granted with respect to IGI’s flat form shrink-wrapped package, and is otherwise denied. c. Section 368-d As for the Plaintiffs’ claim for dilution and injury to business reputation under General Business Law § 368-d, the statute itself requires that the plaintiff show “[Ijikelihood of injury to business reputation or of dilution of the distinctive quality” of the plaintiff’s mark in order to obtain relief. As the New York Court of Appeals has commented: The statute prohibits any use of a name or mark likely to dilute the distinctive quality of a name in use. To merit protection, the plaintiff must possess a strong mark one which has a distinctive quality or has acquired a secondary meaning which is capable of dilution. Allied Maintenance Corp. v. Allied Mechanical Trades, Inc., 42 N.Y.2d 538, 546, 399 N.Y.S.2d 628, 369 N.E.2d 1162 (1977); *1051 accord Mead Data Central, Inc. v. Toyota Motor Sales, U.S.A., Inc., 875 F.2d 1026, 1030 (2d Cir.1989). As discussed above, the Plaintiffs have presented no evidence which indicates that their trade dress is distinctive, and have not shown that it possesses secondary meaning. In the dilution context, the plaintiffs are not entitled to rely on the doctrine of secondary meaning in the making, as the essence of a claim for dilution is that the defendant is attempting to “feed[ ] upon the business reputation of an established distinctive trade-mark or name.” Allied Maintenance, supra, 42 N.Y.2d at 545, 399 N.Y.S.2d 628, 369 N.E.2d 1162 (emphasis added). Therefore, the motion for injunctive relief under § 368-d must be denied. 3. Copyright In order to establish a claim for infringement, a plaintiff must show ownership of a valid copyright and unauthorized copying of the copyrighted expression. Folio Impressions, Inc. v. Byer California, 937 F.2d 759, 763 (2d Cir.1991); Wei ssmann v. Freeman, 868 F.2d 1313, 1320 (2d Cir.), cert. denied, — U.S.-, 110 S.Ct. 219, 107 L.Ed.2d 172 (1989); Hasbro Bradley, Inc. v. Sparkle Toys, Inc., 780 F.2d 189, 192 (2d Cir.1985). a. Ownership Although it is not clear exactly what expression Laureyssens intended to protect in his pre-1991 registrations, the 1991 Certificates clearly cover each of the six Laureyssens Puzzles, and he has therefore established a prima fo,cie case of ownership of valid copyrights in each work. However, this does not conclusively establish validity, but merely shifts to IGI the burden of proving that the copyrights are not valid. Folio Impressions v. Byer, supra, at 763; Hasbro Bradley v. Sparkle, supra, 780 F.2d at 192. IGI has set forth three bases upon which to deny copyright protection to the Laureyssens Puzzles: 1)the puzzles are not sufficiently original; 2) the design of the puzzles is not separable from their functional or utilitarian aspects; and 3) the expression of the puzzles is merged into the underlying idea. Each of these grounds will be addressed in turn. i. Originality As the Folio Impressions court explained: “A pervasive requirement of copyright protection is that the work be original, a requirement that follows from the fact that such protection is accorded only to authors.” Folio Impressions, supra, at 763 (citing Feist Publications, Inc. v. Rural Tel. Serv. Co., — U.S.-, 111 S.Ct. 1282, 1287, 113 L.Ed.2d 358 (1991)). In support of its claim of lack of originality, IGI relies on several exhibits of toy cube puzzles and a number of patents relating to cubes and cube puzzles, all of which predate Laureyssens' creation of his puzzles. The exhibits include a puzzle comprised of six virtually identical pieces with rectilinear interlocking projections which can be assembled into a cube; because of the uniformity of the pieces, however, the cube is not a perfect cube, but instead is left with open corners. The patents include Patent Number 3,924,376, issued on December 9, 1975 to Sukeo Tsurumi (the “Tsurumi Patent”), attached as an Appendix to this opinion, which describes a “cu-boidal structure” formed from six pieces with rectilinear interlocking projections— four of one shape, two of another. The pieces are depicted assembled both in cube form and in a flat, three-piece by two-piece arrangement similar to that employed in the Laureyssens Puzzles. The most readily discerned difference between the Tsurumi Patent and Laureyssens’ is that the Tsurumi pieces have more than five “units” per side. 8 Also included among the patents is Patent Number 3,819,188, issued to Gerald Allen Freedman on June 25,1974 (the “Freed *1052 man Patent”), also attached as an Appendix to this opinion. The Freedman Patent describes a system of pieces with rectilinear interlocking projections which can be assembled into a cube, each of the pieces having six segments per side. As with the Tsurumi Patent, the pieces depicted in the Freedman Patent are substantially similar in appearance to the pieces used by Lau-reyssens. However, despite its submission of this evidence, IGI does not seriously contend that Laureyssens did not in fact design his puzzles and select the particular individual puzzle piece shapes himself. As the Supreme Court has recently made clear, “[ojriginal, as the term is used in copyright, means only that the work was independently created by the author (as opposed to copies from another work), and that it possesses at least some minimal degree of creativity.” Feist Publications, supra, 111 S.Ct. at 1287. The Laureyssens Puzzles clearly meet this standard of originality, and are therefore deserving of protection as original works. 9 ii. Utility IGI’s next claim is that the Laureyssens Puzzles are not protectible because they are “useful articles” which are excluded from the scope of copyright by 17 U.S.C. § 101, which provides that: the design of a useful article, as defined in this section, shall be considered a [copyrightable] pictorial, graphic, or sculptural work only if, and only to the extent that, such design incorporates pictorial, graphic, or sculptural features that can be identified separately from, and are capable of existing independently of, the utilitarian aspects of the article. Section 101 defines “useful article” as “an article having intrinsic utilitarian function that is not merely to portray the appearance of the article or to convey information.” Laureyssens responds that a toy can never be considered a useful article, as by its very nature it has no intrinsic utilitarian function, but rather exists for the entertainment of those who play with it. He also contends that, regardless of whether his puzzles are useful articles, the artistic elements of the puzzle — namely, the shape of the pieces which he designed — is conceptually separable from whatever utilitarian function served by the puzzles. In Brandir International, Inc. v. Cascade Pacific Lumber Co., 834 F.2d 1142, 1145 (2d Cir.1987), the Second Circuit explained that where design elements can be identified as reflecting the designer’s artistic judgment, exercised independently of functional considerations, conceptual separability exists. Here the evidence clearly establishes that Laureyssens exercised his creative judgment in selecting only those pieces which he believed had the proper “balance” for his puzzles. The evidence also establishes that there are many alternatives to Laureyssens’ pieces which can be used to form a flat to cube puzzle, therefore Laureyssens’ selection of his pieces was clearly separate from purely functional considerations, and is entitled to protection. Hi. Idea/Expression Merger Finally, IGI argues that the idea of a flat-to-cube puzzle is not separable from Laureyssens’ expressions of that idea, and that therefore under the doctrine of merger Laureyssens may not protect his expressions. The fact that Laureyssens himself has produced six distinct expression of the idea undercuts this argument somewhat, although merger also applies where there are only a limited number of expressions of a given idea and an author attempts to secure a copyright on each of those few expressions. Toro Co. v. R & R Products Co., 787 F.2d 1208, 1212 (8th Cir.1986); Morissey v. Proctor & Gamble Co., 379 F.2d 675, 678-79 (1st Cir.1967); Lotus Development Corp. v. Paperback Software International, 740 F.Supp. 37, 59 (D.Mass.1990). On the record thus far, IGI *1053 has not satisfied its burden of showing that the expression adopted by Laureyssens is inseparable from the idea of a flat-to-cube puzzle. b. Infringement Once the plaintiff has shown ownership of a valid copyright, infringement can be inferred from the defendant’s access to the copyrighted work and substantial similarity between the copyrighted work and the alleged infringement. Walker v. Time Life Films, Inc., 784 F.2d 44, 48 (2d Cir.), cert. denied, 476 U.S. 1159, 106 S.Ct. 2278, 90 L.Ed.2d 721 (1986). i. IGI’s access IGI does not deny that it had access to the Laureyssens Puzzles, nor that it gave two of the PM Puzzles, which it admits were copies of Laureyssens Puzzles, to Brewer for his consideration while he was writing his program to generate the SNAFOOZ puzzle designs. Brewer testified by affidavit that the only use he made of the two PM Puzzles was in estimating the “difficulty levels” for the SNAFOOZ Puzzles. IGI argues that the structure of Brewer’s computer program itself will establish that none of the alleged similarities between the Laureyssens Puzzles and the SNAFOOZ Puzzles are the result of unauthorized copying of Laureyssens shapes, but rather could only arise coincidentally from Brewer’s independent creation of puzzles which shares many of the ideas of the Laureyssens Puzzles. While this presents an intriguing and difficult question relating to the meaning of access, one which might require consideration of the theory which underlies the whole substantial similarity test, it is sufficient for present purposes to conclude that the Plaintiffs have at least shown the existence of a serious question going to the merits on this issue. This is particularly true given IGI concession that after Brewer’s program had generated approximately forty possible puzzle designs which satisfied all of the criteria, Brewer selected, based on his own subjective analysis of what he thought would work best, the six designs which he submitted to IGI for the SNAFOOZ Puzzles. Even assuming that his program was not improperly tuned to copy the protected aspects of Laureyssens’ designs, it is plausible that this final subjective selection was influenced by Brewer’s exposure to the PM Puzzles. ii. Substantial Similarity Therefore, it is necessary to determine whether the SNAFOOZ Puzzles are substantially similar to any of Laureyssens’ copyrighted works. The Plaintiffs, quoting the Honorable Learned Hand, assert that substantial similarity is established when the ordinary observer, unless he sets out to detect the disparities, would be disposed to overlook them, and regard their aesthetic appeal as the same. That is enough; and indeed, it is all that can be said, unless protection against infringement is to be denied because of variants irrelevant to the purpose for which the design is intended. Peter Pan Fabrics, Inc. v. Martin Weiner Corp., 274 F.2d 487, 489 (2d Cir.1960). It cannot be denied that any differences between the individual pieces of the Lau-reyssens Puzzles and those of the SNA-FOOZ Puzzles would likely be overlooked by an “ordinary observer,” and that such a person would tend to “regard their aesthetic appeal as the same.” Thus under the Peter Pan “ordinary observer” test, it would appear that Laureyssens has established substantial similarity necessary to support a finding of infringement. However, as the Folio Impressions court noted, where a copyrighted work incorporates both original protectible material and unoriginal unprotectible material, the plaintiff must “show the copying was ‘illicit’ by demonstrating that the similarities relate to protectible elements.” Folio Impressions, supra, at 765-767 (citing Walker v. Time Life, 784 F.2d at 48 and Amstein v. Porter, 154 F.2d 464, 468 (2d Cir.1946)). The court explained that the requirement that the plaintiff show copying of protectible elements necessitated a refinement of the Peter Pan Fabrics test: *1054 Of course, the ordinary observer would compare the finished product ... and would be inclined to view the entire [work] — consisting of protectible and un-protectible elements — as one whole. Here, since only some of the [work] enjoys copyright protection, the observer’s inspection must be more discerning. At 765. Applying this modified test, the court simply ignored those similarities which reflected those aspects of the copyrighted work which were not original to the plaintiff. Id. at 766. In the present case, IGI has presented evidence of works existing prior to Lau-reyssens’ creation of his puzzles which incorporated a number of features shared by his works. The cube puzzle and the patents discussed above, particularly the Tsu-rumi Patent, are sufficient to establish that Laureyssens was not the first person to conceive and design a hollow cube puzzle formed from six pieces with rectilinear interlocking projections which also can be assembled into a flat three-piece-by-two-piece form, and hence these aspects of his puzzles do not fall within the scope of his copyright. Therefore, to the extent that the SNA-FOOZ Puzzles are also flat-to-cube puzzles with rectilinear interlocking projections, any similarities between them and the Lau-reyssens Puzzles is not the result of “illicit” copying of protectible elements of Lau-reyssens’ works. In other words, the fact that the pieces of both types of puzzles are squares with right-angled fingers and notches cut in the edges may not be taken as evidence of substantial similarity. Rather, the inquiry must focus on whether there is similarity with regard to the original aspect of Laureyssens works, the shapes of the pieces. 10 With respect to the individual piece shapes, Laureyssens has identified eleven of his thirty-four piece shapes which he claims are copied in nineteen pieces of the SNAFOOZ Puzzles. Laureyssens may not prove infringement, however, by amalgamating all of his copyrights and showing similarity to individual pieces of separate copyrighted works; rather, he must prove substantial similarity between any one of his copyrighted expressions and IGI’s alleged infringihg works. Grouping the alleged similarities by individual copyrighted work, there is only one Laureyssens Puzzle, the green one, which has as many as four pieces which are claimed to be similar to those in any of the SNAFOOZ Puzzles, and only this Laureyssens Puzzle has more than two “lookalikes” in any one of IGI’s works, the purple SNAFOOZ Puzzle, with which it allegedly shares three shapes. As this is Laureyssens’ strongest example of similarity, and as it seems somewhat doubtful that similarity in fewer than three out of six pieces could be termed substantial, the infringement claim will effectively be resolved on the basis of whether the purple SNAFOOZ puzzle is substantially similar to the green Laureyssens Puzzle. In determining this question, Judge Hand’s explanation of the theory underlying the Peter Pan Fabrics test is instructive: in his words, if substantial similarity were not judged by the test of the ordinary observer, “protection against infringement [would] be denied because of variants irrelevant to the purpose for which the design is intended.” This indicates that the purpose of the test is prevent infringers from avoiding liability by making nonsubstantial changes to the copyrighted work, but not to overlook variants which are, in fact, relevant to the purpose for which the design is intended. Here Laureyssens’ own testimony establishes that the primary purpose which he followed in designing his puzzles was to create a perfect cube, not merely to create *1055 piece shapes which he found aesthetically appealing. This purpose must be considered when analyzing the differences and similarities between the puzzle shapes used by Laureyssens and those used by IGI. In addition, it must be kept in mind that both parties’ works are meant to function as puzzles. One of the primary characteristics of a puzzle is that all of the pieces are intended to appear similar, at least to the average observer. Otherwise, the puzzle will present little challenge, as it will be too easy to determine which piece fits where. The evidence establishes that, because of the different number of “units” per edge, it is impossible to use both SNAFOOZ Puzzle pieces and Laureyssens Puzzle pieces to assemble anything. While Laureyssens attempts to characterize this as a mere “dimensional difference,” it is not simply a matter of the pieces having incompatible sizes: there is simply no way to form a cube using pieces shaped like those in the SNAFOOZ Puzzles, in any size whatsoever, together with pieces from the Laureyssens Puzzles. In addition, the increase from five to six units per edge redefines the design of the puzzles, meaning that there is no simple or straightforward method by which to modify a Laureyssens Puzzle to create puzzle with six units per edge: such a transformation essentially would involve starting over from scratch. Analysis of Laureyssens’ claimed similarities with respect to his other puzzle pieces likewise leads to the conclusion that there is not substantial similarity between any of the SNAFOOZ Puzzles and any of Laureys-sens Puzzles. Thus it is simply not the case here that the variants between the Laureyssens Puzzles and the SNAFOOZ Puzzles could be described as “irrelevant to the purpose for which the design is intended.” Indeed IGI consciously took the idea that Laureyssens had used and literally redesigned an expression of that idea. Consequently, there is not a serious question of substantial similarity with respect to the elements of the Laureyssens which are protectible. Therefore Laureyssens has not established a serious question that IGI’s SNAFOOZ Puzzles infringe his copyrights in his puzzles. Conclusion For all of the foregoing reasons, the Plaintiffs’ motion for a preliminary injunction under § 43(a) of the Lanham Act and under the New York common law of unfair competition is granted to the extent of restraining IGI from marketing its SNAFOOZ Puzzles in the flat form shrink-wrapped packaging. The claim under § 368-d of N.Y.Gen.Bus.Law is denied. Because Laureyssens has not established a serious question as to whether IGI’s SNAFOOZ Puzzles infringe his copyrights, the motion for a preliminary injunction under the Copyright Act is denied. It is so ordered. *1056 APPENDIX [[Image here]] [[Image here]] *1057 U.S. Patent Dec. 9 1975 3,924,376 [[Image here]] [[Image here]] *1058 3,924,376 CUBOIDAL STRUCTURE CROSS REFERENCE TO RELATED APPLICATIONS This application is a continuation-in-part of my presently pending patent application Ser. No. 200,081 filed Nov. 18, 1971 now U.S. Pat. No. 3,813,841. BRIEF DESCRIPTION OF THE DRAWING FIG. 1 is a perspective view of a perfect cube structure assembled from six face plates, four of one configuration and two of another, according to my invention; FIG. 2 is a plan view of one of four face plates used in the assembly of the cube structure of FIG. 1; FIG. 3 is a plan view of one of two face plates used in the assembly of the cube structure of FIG. 1; FIG. 4 is a perspective view of a cuboidal structure assembled from six identical face plates; FIG. 5 is an edge view of a face plate taken on line 5 — 5 of FIG, 2; FIG. 6 is a plan view of one of six identical face plates used to assemble the cuboidal structure of FIG. 4; FIG. 7 is an edge view of a face plate taken on line 7 — 7 of FIG. 6; FIG. 7A is ji edge view of a modified face plate taken on line * — 7 of FIG. 6; and FIG. 8 is a plan view of the six face plates used to assemble the cube structure of FIG. 1, the six face plates being snown assembled together on a flat surface. DESCRIPTION OF THE PREFERRED EMBODIMENTS Referring to FIGS. 6 and 7, a face plate 10 has a first pair of opposite sides 11 and 12 and a second pair of opposite sides 13 and 14. Sides 11 and 13 each contain a central recess 15 anc 16. Siles 12 and 14 are formed without material at their ends to leave the projections 17 and 18. The projections 17 and 18 and the recesses 15 and 16 are centrally located in relation to the square outer boundaries of the face plates 10 so that there is no top or bottom surface of a face plate 10. A square notch 20 is formed in the comer of each face plate 10 between adjacent sides 11 and 13 containing the recesses 15 and 16. Notch 30 is the same depth as the recesses 15 and 16. The recesses 15 and 16 and the projections 17 and 18 are of substantially the same length. Six plates 10 may be formed into a cuboidal structure to enclose a perfect cubic volume. Since the thickness of the plates 10 is less than the depth of the recesses 15 and 16 or less than the height of the projections 17 and 18, each edge of an assembled cuboidal structure (not shown) will have a projection and edges adjacent to a recess extend beyond the outer faces of the structure so formed. Cuboidal structures formed from plates 10 may be assembled as building blocks with a projection extending beyond a face of one structure extending into a recess of an adjacent structure. As shown in FIGS. 4, 6 and 7A, six identical face plates 10‘ of the same thickness as the depth of the recesses 15 and 16 or the height of the projections 17 and 18 may be assembled into cuboidal structure 22 which has the eight comer notches 23. At each edge of the cuboidal structure 22, a projection 17 or 18 dovetails into a recess 15 or 16. Structure 22 may have face plates 10' formed of concrete slabs to be assembled into cuboidal building blocks 22. If six face plates 10* have calendar months printed on their 12 surfaces 24 and 25, they may be assembled into a calendar cube which can have its face plates reversed after 6 months 5 to display the other 6 months. Structures 22 may be made of wood to be used as shipping containers. These are but a few examples of uses of this invention. Referring now to FIGS. 1,2,3 and 5, a perfect cube structure 30 is assembled from four face plates 10A 10 and two face plates 10B. Face plates !0A have sides containing central recesses 15' and 16’ and have projections 17 and 18'. Face plates 10B have sides containing central recesses 15" and 16" and have projections 17" and 18". The projections and recesses of the 19 face places 10A and I0B are substantially identical to those of face plates 10' as they are the same thickness as the depth of the recesses or the height of the projections. The comers 31' and 31" between adjacent recesses 15' and 16' and 15" and 16" do not contain any 20 notches. The two face plates 10B additionally each have a single square extension 32 and 32" forming a comer of each face plate I0B between a recess and a projection. The square extension 32" has sides equal to the depth of recesses 15" and 16" of the height of projections 17" and 18". 25 The four face plates 10A are assembled as sides of cube structure 30 with two adjacent plates 10A being inverted in relation to the other two. The face plates 10B then form the top and the bottom of the cube structure 30. As shown in FIG. 8, the face plates 10A and 10B can be assembled as shown on a flat surface to have the calendar markings 35, 36, 37 and 38 formed on them or they can have dice markings 39 and 40 formed on them. The cube structure 30 has many uses as a packing container, a building block, a calendar cube, a novelty container for transparent plastic, or a die, for example. 30 33 While this invention has been shown and described in the best forms known, it will nevertheless be understood that this is purely exemplary and that modifications may be made without departing from the spirit of the invention. 40 What is claimed is: 1. A cube structure comprising, in combination, six square face plates each having two pairs of opposite sides, one side of each pair containing a central recess and the other side having a central projection, said projection and said recess being substantially the same ,0 length, and the depth of said recesses being the same as the height of said projections, said face plates being the same thickness as the depth of said recesses and the height of said projections, and two of said square face plates each having a square extension forming a corner between two of said sides of each of said two face gg plates, one of said two sides containing a recess and the other having a projection, said square extensions having sides equal to the depth of said recesses and the height of said projections, said face plates being assembled into a perfect cube structure having edges with each edge having a projection of one face plate fitted into the recess of another face plate. 2. The combination according to claim 1 wherein said face plates have calendar markings formed thereon, said cuboidal structure being a calendar cube assembled to show selectively six of said calendar markings. *1059 United States Patent H9j nw 3,819,188 Freedman June 25, 1974 (54) SECTIONED SHELL PUZZLES [76] Inventor: Gerald Allen Freedman, P.O. Box 503 Federal Station, Worcester, Mass. 01601 [22] Filed: Dec. 29, 1972 [211 Appl. No.. 319,475 [52] U.S. a..273/160,46/24,46/30 [511 Into.A63Í9/12 [581 Field of Search. 273/157 R, 160; 46/24, 46/25, 30 [56] References Cited UNITED STATES PATENTS 2.201.724 5/1970 Gable. 273/160 3,547,444 12/1970 Williams et al. 273/157 R FOREIGN PATENTS OR APPLICATIONS 573,378 11/1945 Great Britain.46/25 1,913,383 9/1970 Gentian;-46/30 Primary Examiner — Anton O. Oechsle [571 ABSTRACT A plurality of pieces constitute the normal edge dissection of a polyhedron shell. Selections of said pieces can be used to reta nstruct the shell perfectly. Male and female keys along the edges of each piece interfit to form the edges and comers of the shell perfectly. Cubic and tetrahral shells are illustrated. Selected pieces can be used to construct variously shaped shells. The pieces may be provided with binding holes through which binding pins may be inserted to hold adjacent pieces together. 9 Claims, 5 Drawing Figures [[Image here]] *1060 PATENTED JUH251974 3.819.188 SHEET 1 Of 2 FIG. I [[Image here]] FIG. 2 [[Image here]] FIG. 3 [[Image here]] *1061 PATENTED JW2 51374 3.819.188 SHEET 2 Of 2 [[Image here]] [[Image here]] [[Image here]] [[Image here]] *1062 3,819,188 SECTIONED SHELL PUZZLES SUMMARY OF THE INVENTION struction block toy where derived puzzle pieces are Geometric shells can be sectioned in a logical way into interlocking puzzle pieces. Many piece selections of puzzle sets are possible and only a few will reform the original shell, in certain ways, perfectly. Some shell puzzles, especially the cube, can also constitute a con-used to build shells of larger size and different shape than that of the original shell. BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a perspective view of the assembled, nor-maJ-edge dissection of a regular cube shell. FIG. 2 shows a perspective view of six interlocking, planar puzzle pieces derived from the cube shell dissection. FIG. 3 shows a binding pin. FIG. 4 is a perspective view of the assembled, normal-edge dissection of a regular tetrahedron shell. FIG. S shows an outline view of four interlocking, planar puzzle pieces derived from the tetrahedron shell dissection. DESCRIPTION OF THE INVENTION - Any geometric solid can be conceived of as a shell, completely enclosing a single or plural hollow interiors. Geometric shells can be sectioned in logical and arbitrary ways into two or more pieces that can be regarded as constituting a puzzle. This invention concerns the regular shells of polyhedron solids, whose interior is a single compartment, whose interior and exterior surfaces are similar in shape, whose thickness is uniform, and whose edges are ail line segments; particular dissections thereof called the normal-edge and the abnormal-edge shell dissection; and a certain, limited, definition of puzzle piece formation. While any regular geometric shell having at least one edge has unique normal-edge dissection, most shells yield largely dull puzzle pieces. Other polyhedron shells, irregular, deformed, and plural compartment, and other dissections as the normal-corner dissection, the face dissection, the unit dissection, and arbitrary dissections have lesser fascination as puzzles or value as construction block toys. The normal-edge and the abnormal-edge dissection of the regular cube and tetrahedron shells are of great interest and are described in detail as examples of this method of shell dissection used to determine this class of sectioned shell puzzles. The normal-edge dissection involves the normal dissection of shells and a further sectioning of shell edge pieces whereby interlocking male and female keys can be provided to derive puzzle pieces. The normal dissection of any regular shell proceeds as follows: Any shell having one or more comers and/or edges has an extension where every shell surface, exterior and interior, is continued beyond comer and edge boundaries to infinity, respective to functions of shape. For every point on the exterior surface of a shell and the exterior surface extension there is a unique tine segment, normal to the shell surfaces, having a length equal to the uniform thickness of the shell, with end points on the exterior surface or surface extension and on the interior surface or extension. Those points of the actual material of the shed ate either (A) common to one and only one normal line segment. (B) common to exactly two distinct normal line segments. (C) common to three or more distinct normal line segments. That material of a shell whose points form a system of adjacency and meet above condition (A) ate called face pieces of the shell; condition (B), ate called edge pieces; condition (C). are called comer pieces. That material of a shell whose points are common to two or more distinct normal line 10 segments is called the shell frame. Usually a regular shell having of its exterior surface X faces, Y edges, and Z comers will, by the normal dissection, determine X face pieces, Y edge pieces, and Z comer pieces. The frame of the shell can also be understood as the union 15 of all edge and comer pieces, and will be disconnected only when the shell embodies a continuous curve sm-face. Further sectioning of the normal dissection of polyhedron shells into the normal-edge dissection proceeds 20 as follows: Every exterior edge of a polyhedron shell, being a line segment, can be sectioned into 3 equal segments, determining 2 section locations on every distinct exterior edge of the shell. Determined section locations wiil be present to the edges of the shell edge 25 pieces (through the normal dissection) only when every exterior shell edge is greater in length than 3 or more times the shell thickness. Final sectioning, after normal dissection, is by way of flat planes, normal to the shell edges at section locations, sectioning edge 30 pieces only into 3 parts. Section locations for the edge piece part dissection can be otherwise determined or arbitrarily positioned, however comprising key symmetry. The assembled normal-edge dissection of a regular 35 cube shell is shown in FIG. 1 and that of a regular tetrahedron shell is shown in FIG. 4. Any piece or connected combination of pieces of the dissection can be regarded as a puzzle piece, connected combinations considered as formed integrally. The Full set of puzzle 40 pieces of the dissection would consist of an infinite quantity of every unique combination possible. More highly defined puzzle piece sets would consist of finite quantities of variously limited combinations. One such outstanding limited set is defined as follows: Those combinations of pieces, face, edge part, and comer that (1) lie in the same plane, (2) include a face piece, (3) have for every distinct edge of the face piece either the central edge piece part adjacent thereto (male key) or the two edge piece parts adjacent to the central edge piece part (female key), (4) intrinsically include a corner piece only at adjacent female keys, (5) extrinsically include a comer piece only at adjacent male and female keys, and (6) include every possible arrangement of comer pieces present and/or absent to every possible piece of above UJ.4.& 5. Comer piece:; can not be present at adjacent male keys, being disconnected to the face piece at these locations. 45 50 55 FIG. 2 shows the pieces of the above definition, parts „ 1.2.3,4, & 3 derived from the cube shell normal-edge j:-J-J -:J-'-6J dissection, FIG. 1. The dotted lines indicate possible locations of comer pieces as of parts 4.5 ,& 6 of the definition. Comer pieces may be present and/or absent to piece A in 6 different ways, to piece B in 10 different ways, to piece C in 6 different ways, to piece D in 7 different ways, to piece E in 3 different ways, and to piece F in 1 way, determining 33 differently shaped pieces in all. Considering construction of a cube shell from these *1063 pieces, there are 48 selections of S<6,6) =462 possible selections (definition parts 1.2J) that provide the necessary 12 male and 12 female keys exactly. Of these 48 selections, 8 provide the proper quantity of 8, intrinsically included, corner pieces (definition parts 1,2,3,4), while the other 40 selections can be adjusted to 8 corners. in many different ways, by adding or subtracting comers (definition parts 5,6) to form the shell perfectly. 3,819,188 The puzzle pieces derived from the above set definition, parts 1,2,3, &4 for the normal-edge dissection of a tetrahedron shell, FIG. 4, are shown in outline, FIG. 5.Extending the definition to parts 5 & 6, there would be 17 pieces in all. From S(4,4) = 35 selections of pieces, there are 5 that provide the necessary 6 male and 6 female keys exactly, and only one that intrinsically includes the proper quantity of 4 comer pieces exactly, that selection being the four outlined pieces K.I-.M, and N of FIG. 5. An important property of a well defined set of sectioned shell puzzles is that the derived pieces be capable of building shells of different size and shape than that of the original dissection. The cube shell pieces are extremely apt in this respect and a large quantity of puzzle pieces can constitute a construction block toy. However many larger built shells could readily collapse. To prevent collapse, the pieces have binding holes, indicated on the pieces of FIG. 2 by the small circles, and can be secured together in the same and at intersecting planes by tapered binding pins, FIG. 3. In addition to thr defined dissection puzzle pieces, a construction block toy can also include a variety of other blocks such as face, edge, and comer pieces of the dissection, as well as posts, columns, partians, doors, stairs, etc.. These additional pieces would also have binding holes and their combination with a large set of puzzle pieces produces a block toy having extensive scope of design. Cube block pieces, buildable in many imaginative ways, can also be combined in many strange and interesting offset patterns. The regular shells of the five regular polyhedron solids all yield fascinating normal-edge dissection puzzles. Of these, the tetrahedron is the simiest puzzle and the cube the most intriguing construction block toy. Every other polyhedron shell has a unique normal-edge dissection and an interlocking planar puzzle piece set, but, for the most part, the derived pieces fit together in few ways and are unable to build differently shaped shells extensively. In the foregoing definition of sectioned shell puzzles only 1 male or 1 female key is provided to the edges of planar puzzle pieces. Shell edge pieces could also be sectioned into any number of parts, in many different ways, to provide any number of keys, in any combination, to the edges of shell face pieces. Special or offset keying can limit the number of ways proper piece selections can be combined perfectly, but, in general, multi-keyed pieces have an unpleasing, unnecessarily complicated appearance. Many various materials can be used to compose the pieces of puzzles and construction block toys. Pieces can be transparent, colored, or opaque. The puzzle container itself can be a piece or pieces of the puzzle. Pieces could also have designs or further sectioned parts, holes, or patterns, not directly related to the nor-3 mal-edge or the abnormal-edge polyhedron shell dissection. Having set forth disclosure of my invention, I claim: 1. A puzzle comprising a finite number of puzzle 10 pieces of a normal edge dissection of a polyhedron shell, wherein selections of said pieces can be fit together to form the original polyhedron shell perfectly, each of said pieces being planar and comprising an entire face of said polyhedron shell, each piece at each 15 edge thereof having either a centrally located male edge key with female edge keys on opposite sides thereof or a centrally located female edge key with male edge keys on opposite sides thereof, and each piece further including at each comer a male or a fe-20 male key, a said selection of said pieces being equal in number to the number of faces of a given polyhedron to form the faces thereof, with the male and female edge keys of said selected pieces together forming the given polyhedron edges perfectly, and with the male 25 and female comer keys of said selected pieces together forming the comets of the given polyhedron perfectly. 2. The puzzle of claim 1 where die section locations of the edge piece part dissection number two and are 20 located at the interior end points of 3 equal edge segments of every exterior shell edge; and where every exterior shell edge is greater in length than 3 time» the uniform shell thickness. 3. The puzzle of claim I where the section locations 35 of the edge piece part dissection number two and are located anywhere on the exterior edges of shell edge pieces. 4. The puzzle of claim I where the section locations of the edge piece part dissection are any number 2 or 40 more and are located anywhere on the exterior edges of shell edge pieces; and where plural edge keys are provided to each edge of each piece in a manner similar to the single-keyed pieces. 5. The puzzle of claim 1 where the exterior edges of 45 said polyhedron shell all have the same length. 6. The puzzle of claim 5 where said polyhedron shell is a cube. 7. The puzzle of claim 5 where said polyhedron shell is a tetrahedron. 50 8. The puzzle of claim 1 where a finite number of said puzzle pieces provides selections of said pieces that are able to build shells of different size and shape than that of the original dissection and constitute a construction block t. 9. The pieces have binding holes sectioned therein and there are binding pins capable of engaging the binding holes to secure pieces together in the same and at intersect- 60 ing planes. Footnotes [1] The Plaintiffs assert that the notches need not be cut at right angles, and have submitted two exhibits, one made with circular notches and one with trapezoidal notches (i.e., dove-tailed). However, examination of these exhibits and consideration of the basic laws of geometry confirms that in order to have the pieces join smoothly in both the planar and perpendicular arrangements, the notches and fingers must be cut on right angles. [2] Of course, if two notches adjoin one another with no intervening finger, the resulting space is double the width of a single notch, and hence is not square. [3] Because IGI uses foam rubber which is roughly five-sixths the thickness of that used by the Plaintiffs, the cubes produced by both parties’ puzzles are roughly the same size. [4] For example, the piece from Laureyssens’ red (Paris) puzzle which is depicted on the HAPPY CUBE package does not appear in the 1986 WIPO registration. [5] To be precise, the first 1991 Certificate, Number TX 3 009 620, which apparently covers the green (New York) puzzle, indicates "Sculpture, shape of pieces” in this field. [6] In February 1991, IGI’s counsel issued a new opinion letter which advised IGI that the SNA-FOOZ Puzzles did not infringe Laureyssens’ copyrights. [7] Injury to business reputation; dilution Likelihood of injury to business reputation or of dilution of the distinctive quality of a mark of trade name shall be a ground for injunctive relief in cases of infringement of a mark registered or not registered or in cases of unfair competition, notwithstanding the absence of competition between the parties or the absence of confusion as to the source of goods or services. [8] However, because the Tsurumi pieces have no more than two fingers or notches on each edge, each one generally spanning several "units,” the overall appearance of the pieces is substantially similar to those of Laureyssens and IGI. [9] However, as discussed in more detail below, Laureyssens is entitled to protect only those aspects of his works which were original to him. [10] This conclusion is consistent with Laureys-sens' statements in the 1991 Certificates that the work for which he sought protection was the "sculpture" represented by the "shape of pieces." In addition, as discussed above, not only are the shapes of Laureyssens’ puzzle pieces the only original aspect of his works, they are also the only aspect of his expression which is separable from the idea underlying his puzzles. Therefore, in order to establish infringement Laureyssens must prove copying of the shapes of the pieces in his puzzles. Durham Indus., Inc. v. Tomy Corp., 630 F.2d 905, 913 (2d Cir.1980) (“the similarity that inevitably stems solely from the commonality of the subject matter is not proof of unlawful copying”).