delivered the opinion of the Court:
What is meant by the term “on sale” as used in the section referred to [U. S. Eev. Stat. § 4886] has been a subject of judicial consideration, and it will be found that the special facts of each case have been controlling in the conclusion arrived at in the special ease under advisement.
We may take as established certain general rules, and the main difficulty that arises in this, or any other, case lies in the application of the rules to the facts disclosed. Soane of these propositions have been laid down in cases where the question to be determined relates to sale, but the majority of them have related to 'the question of use. The general principles, however, are in the main applicable, whether the case relates to *383“public use” or to “on sale,” and in most cases of sale it will be found that the sale had been followed by more than two years’ public use. Confining our consideration as closely as possible to the question here directly involved, we consider these two propositions deducible from the enunciations of the courts.
First. A single unrestricted sale by the inventor of his invention is a public sale, or puts it “on sale,” within the meaning and intent of section 4886 of the Revised Statutes.
Second. A single sale of the invention by the inventor for experimental purposes, where he is unable otherwise to make proper tests, does not put the invention “on sale” within the meaning of the section.
Third. Where a clear case of “on sale” is established, the burden is on the inventor to prove that the sale was for the purpose of having proper tests made, and that the sale was, at least to that extent, a restricted sale.
The right of an inventor to have a reasonable time for making experiments for the purpose of perfecting his invention or demonstrating its utility has been recognized by all the Federal courts from 1889 down to the present time, and Judge Coxe, delivering the opinion of the circuit court of appeals for the second circuit, in Eastman v. New York, 134 Fed. 844, has stated this as an established proposition.
In Winans v. New York & E. R. Co. 4 Fish. Pat. Cas. 1, Fed. Cas. No. 17,864, Mr. Justice Nelson, in charging the jury said: “If the use be experimental, to ascertain the value, or the utility, or the success of the thing invented by putting it into practice by trial, such use will not deprive the patentee of his right to the product of his genius.”
In Birdsall v. McDonald, 1 Bann. & Ard. 165, Fed. Cas. No. 1,434, it is said: “Public use in good faith for experimental purposes and for a reasonable period, even before the beginning of the two years of limitation, cannot affect the rights of the inventor. The objection rests upon the principle of forfeiture, and is not to be favorably regarded.”
The Supreme Court in Elizabeth v. American Nicholson Pav. Co. 97 U. S. 126, 24 L. ed. 1000, recognized this right, and did *384not deny it, even in cases like Egbert v. Lippmann, 104 U. S. 333, 26 L. ed. 755, where it applied the rule of public use in the most drastic manner. In passing, it may be said that in the cases where the courts have found the statutory bar of public use and on sale to exist such use or sale will, almost without exception, be found to have extended over a long period (ten years in Egbert v. Lippmann, 104 U. S. 333, 26 L. ed. 755), and with many sales as in Consolidated Fruit Jar Co. v. Wright, 94 U. S. 92, 24 L. ed. 69; and where the invention had meanwhile been generally adopted and the rights of the public had intervened and been appropriated.
In Swain v. Holyoke Mach. Co. 48 C. C. A. 265, 109 Fed. 154, where many cases on the subject of public use and sale are collated, Judge Colt, speaking for the circuit court of appeals said: “We should hesitate to lay down the broad proposition that a single sale of a patented device for experimental purposes works a forfeiture of the patent under the statute. We do not understand that it has ever been so expressly decided by the Supreme Court.”
It will be remembered that in the case at bar but one sale had been made, and that claimed to be a restricted sale and for the purpose of experiment, and that there is no claim of bar of public use, for the application for the patent was made within about a year and a half after the engine was set up and tested. It is also recognized by the authorities that the experimental use of the invention need not necessarily be by the inventor himself, nor at his shop. Recognizing, as he must, that a machine can best be tried under the conditions of ordinary use for which it is designed, it would be absurd to say that to test an engine, or other machine, intended for use in making pulp or paper, an inventor is not at liberty to have the actual trial made in some one else’s paper-mill, but he must buy or erect his own mill. It would also be unfair to hold that an employee who had not the means to build a machine, except by permitting his employer to build one containing his invention, and to sell it conditionally and for the purpose of experiment, should be deprived of his invention where he filed his application within *385two years from the time of the experimental trial of such machine. Elizabeth v. American Nicholson Pav. Co. 97 U. S. 135, 24 L. ed. 1005.
Another proposition seems to have been lost sight of in the consideration of this case by the Commissioner of Patents. He considers the fact that the engine was, upon use by the Nashua ■company, found to be satisfactory so far as the invention is concerned, as tending to negative the claim that the sale was for ■experimental tests. That no changes in the parts constituting appellant’s invention were found to be necessary is wholly immaterial. As the Supreme Court said, in Elizabeth v. American Nicholson Pav. Co. 97 U. S. 135, 24 L. ed. 1005, in speaking of experimental use, the inventor “may see cause to alter it and improve it, or not. * * * And though * * * he may not find that any changes are necessary, yet he may be justly said to be using his machine only by way of experiment.”
Before passing to the application of the law to the facts in the case, three or four minor questions, specially applicable to the correct decision of the question here in issue, may be set ■out as sustained by authority and based on reason. In Harmon v. Struthers, 57 Fed. 637, it was held that the statutory bar did not exist, where the patentee derived no profit from the sale, and the device was a mere appendage to a large machine sold by his employer. In Graham v. McCormick, 11 Fed. 859, the court held that the statutory bar did not exist because the machine was sold for the purpose of trial and was warranted. In Innis v. Oil City Boiler Works, 22 Fed. 780, it was held that the statutory bar could not be successfully invoked where the machine was sold at an under price, and without profit to the inventor, and for the purpose of securing a fair test of the invention. In Jones v. Sewall, 3 Cliff. 563, 6 Fish. Pat. Cas. 343, Fed. Cas. No. 7,495, it was held that a use Avas experimental when it was had to ascertain .the value, utility, or success of the invention by putting it into practice; that it was not ■such use or sale as would deprive an inventor of his title.
The burden is on Mills to prove that the sale of the machine to the Nashua company was of such a nature as to take it out of *386the statute and bring it within the exceptions recognized and stated by the courts in construing section 4886 of the Revised Statutes, U. S. Comp. Stat. 1901, p. 3382. In our opinion he has succeeded, and we believe that no statutory bar exists to the. granting of a patent to him through the sale of the engine in question. He is entitled to be heard • as a witness on his own behalf in public-use proceedings, and. the only question is as to the weight to be given to his evidence. He is corroborated on substantially all material points, and the facts proved have already been set forth. Those facts clearly bring him within the exceptions to the general rule. He was an employee of a company making, but not using, machinery used to make paper-pulp. His employer could not put the device to a test in actual use. It was but natural that the employer should be unwilling to make a machine embodying the invention, without first obtaining an order which should provide that if the machine proved satisfactory it would be paid for. Natural, too, that when it had to warrant the machine and replace it, if found unsatisfactory, with one of its standard machines and at a less price, that it should provide that no royalty should he paid Mills for the machine, or for others if sold to the purchaser of' the first machine. After that machine was sold nothing more was done, save to build and ship it, and then wait for the results of actual, every-day use. Very unlike other cases where the statutory bar has been successfully invoked. Here was no unlimited number of sales, no indefinite time of use. Had he waited more than two years after the test of September, 1899, the statutory bar of public use might have been successfully invoked, for the machine, so far as the invention goes, was found to be a success. The sale was for experimental purposes, and we think may fairly be found to be a conditional one. Had the machine not proven to be a success, we think it could have been returned, and the extra price paid for the incorporation of the invention into the machine been recovered. To find the statutory har of “on sale” to be proven in this, case within the meaning of the statute would compel us to go beyond the adjudicated cases, and this we are not prepared to do. We are not *387prepared to say that an inventor who is compelled to sell a machine, or allow it to be sold, and without any profit to him, and with a warranty given by the manufacturer, and an obligation to substitute another machine for it, should it prove inoperative, all in order that it may be tested under the conditions of every-day use, has not sold it for experimental use, and that the sale is not a conditional one. In arriving at our decision we have taken it as a fact that the sale was made more than two years before the filing of the application. It might be argued with considerable force that the sale was not even a conditional, completed sale until the receipt of the machine by the Nashua, company, or until after its acceptance after actual test. In the former case it may be said that the sale would not have been proved beyond a reasonable doubt, and in the latter case there would have been no sale, conditional or otherwise, more than two-years prior to the filing of the application. In the absence o£ proof we have considered a conditional sale for experimental purposes to have been made as of the date of the shipment of the engine.
Forfeiture of the invention cannot, and in fairness should not be, predicated upon the sales relied upon. There was- manifest error in the decision of the Commissioner of Patents, and we must reverse his conclusion that appellant is not entitled to a patent for the reason assigned.
It follows from what we have said that the decision of the Commissioner must be reversed, and it is so ordered.
The clerk of the court will certify this opinion and the proceedings of this court in the premises to the Commissioner of Patents, according to law. Reversed.