Appellant Omark Industries (hereinafter “Omark”) appeals from a judgment invalidating its patent for a brush cutting chain invented by appellee Raymond Carlton. The district court found that the brush-cutting chain had been in public use more than one year before May of 1959 when the patent was applied for and held the patent invalid under 35 U.S.C. § 102 . For the reasons stated below, we concur with the findings of the district court, 458 F.Supp. 449 , and affirm its judgment.

FACTS

Omark designs, manufactures, and markets saw chains for use on motorized chain saws. Between 1955 and 1957, appellee Raymond Carlton [“Carlton”], then chief engineer at Omark, designed a saw chain for use in cutting small tree limbs and brush [hereinafter “brush-cutting chain,” “chain”].

In early 1957, Omark’s engineering department undertook substantial field testing of the brush-cutting chain at the site of the North Fork Reservoir on the Clackamas River in Oregon. Members of the engineering department personally placed samples of the brush-cutting chain on the saws of workers cleaning brush on the dam site, and demonstrated to the workers the proper use of the chain. No restrictions were placed on the use of the chain; it was to be used in a normal manner. Omark’s engineering department, through Bill Simpson, kept careful track of the chains which had been given out. The names of the workers were written in a notebook, and as the chains wore out they were taken back to the lab for analysis. Simpson, sometimes accompanied by engineer Russ Carter, made frequent trips to the dam site to monitor the performance of the brush-cutting chain. Simpson made regular reports on the chain’s field performance to Carlton, both through a series of weekly memos, and verbally.

When certain adverse comments were received by Simpson from workers using the chain, the engineering department took steps both to ascertain the source of the problem and to make recommendations for its correction. For example, after receiving comments that the chain “ran hot,” and that the chain would not cut fast in large timber, the engineers determined that the problem was the user’s failure to file the “bumper” or “guard” links in the chain during routine maintenance of the chain. The engineers then instructed the workers concerning proper maintenance of the chain. Similarly, when the workers simply stopped using a brush chain of a certain pitch, the engineering department suggested in its final report that a smaller pitch chain be manufactured.

The overall excellent performance of the product as a brush-cutting chain was documented in Simpson’s memos to Carlton, and was confirmed in Simpson’s deposition testimony.

In January of 1958, Carlton sent a memo to the head of Omark’s sales department, Mr. H. Cutright, recommending that the brush-cutting chain be added to Omark’s product line. In the memo, Carlton stated that while a certain aspect of the brush-cutting chain had not been tested out completely, the engineering department was confident the chain would perform well. Two. weeks later, John Gray, then president of Omark, asked Cutright for an analysis of the “potential market” for that chain. Pursuant to that request, Art Gredler, head of the chain sales division of Omark, made plans to have a limited amount of the chains manufactured in order to obtain customer reaction in a number of states.

In deposition testimony, Gredler stated that “if the engineering department thought they had a marketable product they would give it to [the sales department] and [the sales department] would put it out to test.” After the engineering department had tested out the brush chain in 1957, the sales department gave samples of the chain to Omark distributors for further testing in various types of woods and climates. The sales department conducted a follow-up to see how the user liked the chain, or, in Gredler’s words, to gauge its “general acceptance.” Follow-up was done informally. While the sales department kept records of the reels that were sent out, the reports from the users were generally gathered third hand — the user would give a report to the dealer, who would report to the Omark distributor, who would then relay the information to the sales department. No restrictions were placed on the use of the chain; the only suggestion was that it be tested under the most severe conditions. Customers were asked to let the distributors know how they liked the chain. Samples of the chain were sent to distributors in the Midwest, the Northwest, California, and New England. The chain was distributed to customers without charge, with the exception of perhaps one incidental sale by a dealer in Michigan. Most, if not all, of the sales department’s testing occurred prior to May 1,1958, over one year before Omark applied for a patent on the brush-cutting chain.

Gredler indicated in his deposition that the chain was distributed in order to obtain the users’ reaction to it, “which would mean whether it would be a saleable item.” Customer reaction was not particularly positive, and the sales department was pessimistic about the sales prospects for the chain. Nevertheless, in May of 1959, Omark applied for a patent.

STANDARD OF REVIEW

Appellant contends that there is no presumption in favor of the district court’s findings on the issue of public use, since all of the evidence regarding that issue is in the form of depositions, prepared statements, and documentary evidence. See National Lead Co. v. Western Lead Products Co., 324 F.2d 539, 543 (9th Cir. 1963). Even assuming that on these facts no presumption should be accorded the district court’s findings, we conclude that the court’s finding on the issue of public use was correct, and that the patent is invalid under 35 U.S.C. § 102 .

THE APPLICABLE LAW

35 U.S.C. § 102 provides that if a patented invention was “in public use” more than one year before its patent was applied for, the inventor’s patent is invalid. Incidental use of the invention by members of the public, however, is not considered a “public use” within § 102 if that use was primarily for “experimental” purposes, that is, primarily for the purpose of testing the essential qualities of the invention or for perfecting it. In re Yarn Processing Patent Validity Litigation, 498 F.2d 271, 277-78 (5th Cir. 1974), cert. denied, 419 U.S. 1057 , 95 S.Ct. 640 , 42 L.Ed.2d 654 (1974). See Del Mar Engineering Laboratories v. Physio-Tronics, 642 F.2d 1167, 1169 (9th Cir. 1981).

A defendant who raises the defense of “prior public use” has the initial burden of establishing the facts that would call that defense into play by “substantial evidence” which is “clear and satisfactory”. Tucker Aluminum Products Inc. v. Grossman, 312 F.2d 293, 294 (9th Cir. 1963); see Amerio Contact Plate Freezers, Inc. v. Belt-Ice Corp., 316 F.2d 459 , 462 n.7 (9th Cir.), cert. denied, 375 U.S. 902 , 84 S.Ct. 191 , 11 L.Ed.2d 143 (1963).

Once the defense is adequately raised, however, the burden shifts to the inventor to prove by “full, unequivocal and convincing” evidence that the use was primarily experimental in nature. Smith & Griggs Manufacturing Co. v. Sprague, 123 U.S. 249, 264 , 8 S.Ct. 122, 129 , 31 L.Ed. 141 (1887); Cali v. Eastern Airlines, Inc., 442 F.2d 65, 70 (2d Cir. 1971); see Kalver Corp. v. Xidex Corp., 556 P.2d 966, 968 (9th Cir. 1977); see generally, Robbins Co. v. Lawrence Manufacturing Co., 482 F.2d 426 (9th Cir. 1973).

We find that the appellees have carried their burden of demonstrating that there was use of the invention by the public for over one year before the patent was applied for. In particular, the public’s use of the brush-cutting chain during the sales department’s 1958 testing activity clearly constituted “use by the public” sufficient to shift the burden to Omark to show by full, unequivocal and convincing evidence that the use was primarily experimental.

Moreover, the evidence strongly supports the conclusion that the sales department’s testing program was undertaken primarily to determine the merchantability of the chain, rather than to perfect its essential qualities. See Cataphote Corp. v. DeSoto Chemical Coatings, Inc., 235 F.Supp. 936, 938 (N.D.Cal.1974), aff’d 356 F.2d 24 (9th Cir.), cert. denied, 385 U.S. 832 , 87 S.Ct. 71 , 17 L.Ed.2d 67 (1966). There was apparently little need for further testing of the chain’s essential qualities at the time the Sales Department undertook its market analysis. The main problems associated with the chain — those of slow cutting speed and of proper chain maintenance — had already been identified by the engineering department, which also had made proposals for their correction. Moreover, the engineering department’s testing had already established that, as a brush-cutting chain, the chain performed extremely well.

Secondly, the sales department employed extremely informal procedures for monitoring user reaction during its testing program. The names of the user were not recorded. Customer reaction was generally relayed to the department third hand by Omark’s salesmen and distributors; most reports of user reaction were communicated verbally to the sales department rather than by written report. These procedures suggest that the department’s testing program was designed primarily to determine how well the chain would sell, not to isolate systematically the precise qualities of the chain which remained problematic. The procedures of the sales department contrast markedly with those used during the engineering department’s 1957 testing program, a program whose purpose was clearly to test the essential qualities of the chain as a brush cutter.

Finally, the very involvement of the sales department following the chain’s extensive field testing by the engineering department is in itself some indication that it was the receptivity of the market place that was being probed, not the suitability of the chain for its intended purpose. The deposition testimony clearly indicated that it was the task of the engineering department, not the sales department, to determine the technical causes of and solutions for problems in a product’s performance. Moreover, Gredler indicated that it was customary for the sales department to test the marketability of items once they had been recommended for Omark’s production line by the engineering department. The evidence suggests that this is precisely what occurred here.

Omark supports its argument that the sales department’s activity was experimental testing, designed to test the “qualities” of the brush-cutting chain in two ways. First, Omark argues that the brush-cutting chain had to be tested “in the various woods and climates,” which, it contends, is tantamount to testing the qualities of the invention. Secondly, Omark argues that after 1957, further testing was required to overcome the problems of improper chain maintenance and low cutting speed. It argues that Carlton’s statement in his January 23 memo to the head of the sales department shows that even Carlton conceded that the chain required further testing. Omark’s arguments are not persuasive.

First, Omark’s claim that the chain had to be tested “in various woods and climates” does not establish that this testing was primarily experimental in nature. Indeed, Gredler’s deposition tends to indicate that this testing was necessary to determine the chain’s overall sales potential rather than to probe further its qualities as a brush-cutting implement.

Secondly, the evidence does not support Omark’s suggestion that the 1958 testing was designed primarily to remedy the problems of slow cutting speed and improper maintenance. At most, the evidence indicates that the sales department’s testing merely corroborated the engineering department’s prior finding that these problems existed. The only suggestion made concerning the chain’s operation as a result of the sales department’s activity — that the chain be made available in a smaller pitch— had been made by the engineering department nine months earlier. This evidence falls far short of demonstrating by full, unequivocal and convincing evidence that the sales department’s activity was primarily a program of experimental testing.

Finally, Carlton’s memo of January 23, if anything, suggests that the engineering department, not the sales department, needed to do more testing on the chain. Moreover, Carlton’s memo suggests that the chain already performed well, and that the engineering department simply wanted to do further testing on a possible variation of the chain.

Appellants have thus failed to carry their burden of demonstrating that the public use of the brush-cutting chain during the sales department’s 1958 activity was primarily experimental in nature. For that reason, the judgment of the district court is AFFIRMED.

. U.S. Letters Patent No. 3,180,378. . 35 U.S.C. § 102 provides in pertinent part: A person shall be entitled to a patent unless ... (b) ... the invention was ... in public use or on sale in this country more than one year prior to the sale of the application for patent in the United States. ... The policy underlying this statutory bar “is to prevent an inventor from extending the period of patent monopoly by commercially exploiting the invention before applying for a patent.” Del Mar Eng. Labs v. Physio-Tronics, 642 F.2d 1167, 1169 (9th Cir. 1981); see Pickering v. Holman, 459 F.2d 403, 406 (9th Cir. 1972). . Our disposition of the “public use” issue makes it unnecessary for us to consider the remaining issues raised by appellant. . The “bumper” or “guard” links on the brush-cutting chain were designed to prevent branches or brush from tangling in the saw chain. . The larger *A" pitch chain did not work as well as the Vis" pitch chain on direct drive chain saws, which were becoming increasingly popular at the time the chain was tested. . The memo read: Attached to this memo is a project development booklet which we have introduced in Engineering to be used for the development of projects and new products. This particular booklet covers the design and development for the brush chain, and includes all information and test data relating to this chain. The booklet also points out the tests that have been conducted on this chain and the results of those tests. It also makes recommendations as to how this chain [the brush-cutting chain] is to be used. The advent of the Joe Brady Tree Girdling Chain has made a potential change in this particular brush chain, inasmuch as we are now considering using the bumper that is now being used on the girdling chain, on this brush chain. We have not tested this bumper out completely, but we are confident that it will work just as well as the bumper shown in this booklet. We in Engineering highly recommend that this brush chain be put into our product line and that with a moderate amount of advertising, this chain could add substantially to our overall chain volume. It is interesting to note that if this chain was added to our product line, it would not cost any more to build than our present chain, nor would there be any tooling charges involved. Consequently, the only cost involved would be the inventory that we would maintain on this chain. I am also attaching a 216 form recommending this product be added to our product line. [Emphasis added] . The evidence on this issue consisted primarily of depositions, prepared statements and documentary evidence. Bill Simpson and appellee Raymond Carlton testified briefly at trial about the Engineering Department’s 1957 testing of the brush-cutting chain. It is uncertain whether National Lead is still good law in this circuit. See Del-Mar Engineering Lab. v. Physio-Tron-ics, 642 F.2d 1167, 1169, n.2 (9th Cir. 1981); United States v. Mountain States Const. Co., 588 F.2d 259 , 264 n.5 (9th Cir. 1978). . Omark cites Cali v. Eastern Airlines, Inc., 442 F.2d 65, 70-71 (2d Cir. 1971), for the proposition that a use is also experimental if the inventor’s motive in making the device available is to determine whether his ideas “are worth exploiting.” This dictum in Cali , if read too broadly, would make all test marketing an “experimental use,” since the purpose of test-marketing is to determine whether an idea is “worth exploiting.” We do not read Cali so broadly; such a result would be directly contrary to the purpose of § 102. . The “public use” defense is sufficiently raised by showing, for example, a non-secret, apparently non-experimental use of the invention pri- or to the critical date, In re Yam Processing Patent Validity Litigation, 498 F.2d 271, 277 (5th Cir. 1974), or, as here, by showing a non-secret, unrestricted use by the public of an operative device which substantially embodied the claimed invention. See Kalvar Corp. v. Xidex Corp., 556 F.2d 966, 968 (9th Cir. 1977); Robbins Co. v. Lawrence Mfg. Co., 482 F.2d 426, 432-34 (9th Cir. 1973). . Gredler stated that Omark could not manufacture an item profitably unless they could sell a high volume of the item.