Plaintiffs-Appellants Patricia Hennon and New Bern Molding Company ("NBM") appeal the district court's decision to deny Appellants' motion for a preliminary injunction to enjoin Defendants-Appellees World Bazaars, Inc., and Kirkland's, Inc., from infringing Appellants' copyrights. We affirm.

Patricia Hennon and her husband, James Hennon, own and operate NBM in New Bern, Virginia. Through their company, the Hennons produce hand-made Dickensian Christmas caroler figurines, which they primarily sell at wholesale to dealers at craft fairs and by mail. They also sell their carolers at retail stores in Wytheville and Roanoke, Virginia, and at a temporary Christmas store in Charlottesville, Virginia.

Early in 1994, the Hennons learned that World Bazaars was merchandising inexpensive imitations of the Hennons' carolers at trade shows and at stores operated by Kirkland's. World Bazaars, a major importer and national retailer of merchandise, purchased all of the designs at issue as open stock items from an overseas supplier. The company then sold the carolers to Kirkland's and to other retailers. Both World Bazaars and Kirkland's have retail stores across the country. Neither World Bazaars nor Kirkland's specified the design details for the accused figurines at issue in this case.

After the Hennons compared the two lines of carolers, Patricia Hennon and NBM filed this copyright action against World Bazaars and Kirkland's on October 11, 1994, in the United States District Court for the Western District of Virginia. Appellants claimed that Appellees' unauthorized copying, display, and sale of their figurines violated Appellants' exclusive rights under 17 U.S.C. Sec. 106. Appellants alleged that their business has decreased by as much as thirty percent because they cannot compete with the low price of Appellees' Chinese-made figurines. Although the Hennons' carolers retail from $30.00 to $50.00 each, Appellees' carolers retail at $10.00.

On October 11, 1994, Appellants also moved for a temporary restraining order ("TRO") against Appellees; and the district court entered a TRO against Appellees on October 24, 1994, which would have become effective upon Appellants' posting of a bond in the amount of $20,000.1 The matter of Appellants' motion for a preliminary injunction was set for a hearing on November 18, 1994. Following a hearing on that date, the district court denied Appellants' motion for a preliminary injunction in an order dated November 28, 1994.

Appellants argue that this Court should reverse the district court's decision because the court abused its discretion in denying Appellants' motion for a preliminary injunction. See Hughes Network Sys., Inc. v. Interdigital Communications Corp., 17 F.3d 691, 693 (4th Cir.1994) (holding that preliminary injunctive relief is a remedy that lies within the sound discretion of the district court). This Court established the standard governing the issuance of a preliminary injunction in Blackwelder Furniture Co. v. Seilig Manufacturing Co., 550 F.2d 189 (4th Cir.1977). Under Blackwelder, the issuing court considers: (1) the likelihood of irreparable harm to the plaintiff if an injunction is not granted; (2) the likelihood of harm to the defendant if an injunction is granted; (3) the likelihood of success on the merits; and (4) the public interest. Id. at 195-96. In this case, the district court denied the preliminary injunction because it concluded that Appellants had not satisfied their burden of showing that they would likely succeed on the merits in their copyright infringement action.

In order for Appellants to meet this burden, they must make a prima facie showing of copyright infringement. Such a showing establishes a presumption that Appellants have suffered irreparable harm and will likely succeed on the merits. Service & Training, Inc. v. Data General Corp., 963 F.2d 680, 690 (4th Cir.1992). To establish a prima facie case of copyright infringement, the plaintiff must prove that the plaintiff owns a valid copyright and that the defendant copied the original elements of the copyrighted work. Keeler Brass Co. v. Continental Brass Co., 862 F.2d 1063, 1065 (4th Cir.1988). Because direct evidence of copying is seldom available, courts presume copying if the plaintiff shows that the alleged copier had access to the work and that the original work and the alleged copy are "substantially similar." Id. The parties in this case do not dispute that Appellees had access to Appellants' figurines; rather, their debate centers on whether the two sets of figurines are substantially similar.

A prima facie showing of substantial similarity requires that the plaintiff establish the substantial similarity of both the ideas of the two works and the expression of those ideas. Dawson v. Hinshaw Music Inc., 905 F.2d 731, 732 (4th Cir.), cert. denied, 498 U.S. 981 (1990). Courts refer to this first prong as an "extrinsic" or "objective" inquiry while they characterize the second prong as an "intrinsic" or "subjective" inquiry, which analyzes the "total concept and feel" of the works. Id. (citing, inter alia, Litchfield v. Spielberg, 736 F.2d 1352, 1356 (9th Cir.1984), cert. denied, 470 U.S. 1052 (1985)). In addressing this second prong, courts separate the protectible expression unique to the allegedly infringed work from the unprotectible expression that is dictated by the idea upon which the work is based. Aliotti v. R. Dakin & Co., 831 F.2d 898, 901 (9th Cir.1987). Courts then look to the entire work, including the unprotectible expression, to determine if the two works are substantially similar in expression. Courts, however, will not find substantial similarity "where analytical dissection demonstrates that all similarities in expression arise from the use of common ideas" because protecting the expression when the idea and expression are inseparable " 'would confer a monopoly of the idea upon the copyright owner.' " Id. (quoting Herbert Rosenthal Jewelry Corp. v. Kalpakian, 446 F.2d 738, 742 (9th Cir.1971)). From these principles, courts have rejected copyright infringement claims when the similarities in expression necessarily followed from the underlying ideas of the works. See, e.g., id. at 901-02 (holding that intrinsic test was not satisfied when similarities in stuffed dinosaur toys were due to either the physiognomy of dinosaurs or the nature of stuffed animals designed for children).

Applying these principles to the figurines at issue in this case, the district court concluded that the similarities between the Appellants' carolers and the Appellees' carolers were due to the "inevitable congruences" of expression that are dictated by the types of impoverished Dickensian carolers the figurines portray and by the common conventions of clay molding used in creating the figurines. Beyond these inevitable similarities, the court found that the details of the two collections of carolers differ in many respects. Specifically, the district court found several differences in the color, texture, and size of the carolers.2 The district court also found specific differences in the characters the figurines portray. Although both sets depict four types of individuals--a woman caroler, a singing "uncle," a drummer boy, and a singing boy--the court detailed certain distinctions in the clothing, songbooks, and facial features and expressions of the characters.

To support their contention that the district court abused its discretion in denying the preliminary injunction, Appellants rely on the testimony of their expert witness, who testified at the hearing about the similarity of the figurines. The expert, a professional sculptor, testified that Appellees' collection "unquestionably" had been copied from Appellants'. He reasoned that a caroler by Patricia Hennon has "a variety of textures and markings on it which are very specific to [Hennon's] vision and talent" and that the similarities between the collections could not be a coincidence because "there are too many indications visually, formally, proportionally, and in terms of caricature, that are more than similar. To me, they are really very exact." Joint Appendix 55. On cross examination, the expert conceded that some differences in coloration existed between the two collections, but he maintained that the "caricatures" and "portraitures" (the form of the face and body) of the two sets are very similar.

Appellants contend that, because Appellees did not offer any testimonial evidence to rebut the expert testimony, the district court abused its discretion in not concluding that the expert testimony established substantial similarity. Appellants' argument is flawed, however, because expert testimony cannot conclusively satisfy both prongs of the substantial similarity inquiry needed to find copyright infringement in this case. While expert testimony is useful to establish the first prong regarding the similarity of ideas, techniques, and conventions in two works, the second prong ultimately hinges on the perception of the "intended audience." Dawson, 905 F.2d at 733, 736; see also Hartman v. Hallmark Cards, Inc., 833 F.2d 117, 120 (8th Cir.1987). In this case, because the intended audience for the carolers at issue is the lay public who purchase the figurines, the second prong depends on the perception of the "ordinary lay observer." See Dawson, 905 F.2d at 733, 736. Appellants therefore did not need to offer testimonial evidence to rebut Appellants' expert. Appellees introduced the figurines into evidence and the district court properly conducted a "lay" examination of the figurines in order to determine whether the expressions of the two collections are substantially similar.3 The district court's determination of similarity of expression depended on its factual analysis of the figurines, and we review for clear error the district court's factual findings for substantial similarity. Wildlife Express Corp. v. Carol Wright Sales, Inc., 18 F.3d 502, 506-07 (7th Cir.1994). Applying this standard, we conclude that the court did not commit clear error in finding that an ordinary lay observer would not find the two collections substantially similar. The district court conducted a detailed examination of the figurines at issue. The court found several differences in the color, texture, and size of the carolers, as well as specific differences in the characters the figurines portray. These differences between the two collections certainly could be such that a lay person, unskilled in identifying the similarities alleged by the expert, would not find the sets of carolers substantially similar. We therefore hold that the district court did not abuse its discretion in concluding that Appellants failed to satisfy their burden of demonstrating a likelihood of success on the merits.

For the foregoing reasons, we affirm the decision of the district court denying Appellants' motion for a preliminary injunction.4

AFFIRMED.