COXE, Circuit Judge.
This is an equity action for the infringement •of letters patent, No. 42,920, granted May 24, 1864, to James Knibbs, assignor to himself and Marcus P. Norton, for “Improvements in Pumps for Steam Fire and other Engine Pumps.”
The inventor conceived the invention April 21, 1860, and applied the device to a steam fire engine April 29, 1860. The application was filed May 13, 1864.
The Patent.
As stated in the description the nature of the improvement consists in the employment of a pipe by means of which the force or discharge part of said pump is connected to and-with the suction or supply part, so that one or more discharge pipes or hose may throw streams of water at the same time and stroke of the piston. This is accomplished [846]*846without waste of water, hy the opening of a valve or discharge pipe,, to enable the pump to work successfully and without injury in throwing streams of water.
The description states that prior to the invention great difficulty had' been experienced when an engine was required to throw a less number of streams than its total capacity permitted. In such circumstances, the suction being greater than the discharge, relief could only be obtained by opening a waste valve and discharging the water upon the ground. If this were not done the pump would become strained and flooded and the engine would cease to work. This primitive method of relieving the strain was injurious to the engine, wasteful, inconvenient and resulted in flooding the street, thus making it difficult for the firemen to operate, particularly in winter when ice was liable to-form. All these difficulties and disadvantages were obviated by the invention. By connecting the force part with the suction part of the pump the extra quantity of water thrown into the force part and not discharged through the hose is conducted back into the supply or suction pipe, thus relieving the force part from any excessive quantity of water..
The connection between the two sections of the pump, as shown in the drawings, is made by means of a pipe placed on the outside of the-pump, the excessive quantity of water to be returned through this pipe to the suction part, being regulated by a valve which remains closed if all the hose pipes are discharging. If less than the whole number of pipes are discharging at the same time or stroke of the piston this valve must be opened sufficiently to allow the return of the water which cannot be discharged by reason of the closing of some of the hose pipes.
The claim involved is the second, which is as follows:
“The connecting of the discharge or force part or section of a steam fire- or other engine pump to and with the suction or supply section thereof by means of the tube GG and the regulating-valve H, or an equivalent therefor, substantially as and for the purposes herein described and set forth.”
The invention, though simple and easily comprehended, was one of unusual merit and contributed largely to the success of the steam fire-engine. The testimony shows that prior to the application of the Knibbs device to the engine it was not an effective and reliable machine; afterwards, there was little difficulty in operating with a limited number of hose pipes without waste of water or damage to the engine.
The Defenses.
The usual defenses of anticipation, lack of invention and noninfringement were interposed and are reasserted by the assignments of error. We do not deem it necessary to discuss any of these defenses further than to say that were they the only obstacles in the complainant’s path it would not be difficult to sustain the decree. Infringement, though not admitted, is not seriously disputed.
The principal controversy, during the quarter of a century that this-litigation has occupied'the attention of the courts, arises over the defense of public use for more than two years prior to the filing of the application. The application was filed May 13, 1864, so that a well-established case of public use, prior to May 13, 1862,'will invalidate the patent. The defendant insists that the proofs show that before May [847]*84713, 1862, the invention was publicly on the fire engines “Arba Reade” and “Jason C. Osgood,” in the city of Troy, N. Y., on the “Phoenix,” in the city of Hartford, Conn., on the “Atlantic,” in the city of Lawrence, Mass., and on the “Gov. Hill,” in the city of Concord, N. H. That the Knibbs device was placed on each of these engines and publicly used, prior to May 13, 1862, is established by convincing proof. The facts, however, differ in each instance, and, as it will subserve no useful purpose to state the evidence in its entirety, we have selected the prior uses in connection with the Osgood and the Phoenix — and incidentally the Arba Reade — as presenting all the facts necessary to a complete determination of the controversy.
The Arba Reade.
Knibbs, the patentee, was the engineer in charge of the stearn fire engine Arba Reade, manufactured by the Amoskeag Manufacturing Company, of Manchester, N. H., and delivered to the Troy fire department March 28, 1860. On April 21, 1860, while the Reade was operating at a fire in Troy, it became necessary, in order to relieve the pressure on the pump, to open one of the gates and discharge the water onto the street, thereby causing some damage. The idea immediately occurred to Knihbs of relieving the pressure by the means described in the patent. Before the 3Qth of the same month he had connected the discharge section with the supply section of the pump by means of a crooked pipe with a globe regulating valve seated therein, the mechanism being the exact mechanical counterpart for that described in the patent four years afterwards. On April 30,1860, the Reade was taken out and publicly tested, the official entry being “All worked well during trial at Fulton Street.” The device originally placed on the Reade remained unaltered for nearly three years, during which time the engine was in constant use by the fire department of Troy and was several times the subject of public exhibition. On one occasion, about the middle of July, 1860, when the agent of the Amoskeag Company was at Troy, for the purpose of receiving pay for the Reade, there was a successful public exhibition at Green Island during which the relief valve and water passage was the subject of examination, the agent remarking “that it seemed to operate very nicely.” Knibbs replied that they could hardly have gotten on without it and that as soon as he had it patented he would like to open a correspondence with the Amoskeag Company.
As early, then, as July, 1860, the necessity for a patent had occurred to Knibbs, not for any new or improved device, but for the mechanism then on the Reade, which, after a trial of nearly three months, had proved to be a practically successful operative device. If the apparatus were inchoate it is evident that Knibbs did-not so consider it, for in his judgment it -was ready for patenting in 1860. From the time it was placed upon the engine, in April, 1860, until February 12, 1863, when a larger pipe and an automatic valve were substituted, the engine was in constant use in circumstances which required the frequent operation of the relieving mechanism and seems to have given entire satisfaction. This use [848]*848was frequent, satisfactory, successful, public, open and notorious. The device was explained by the inventor to many interested persons and among others to the employés of the Amoskeag Company, including its mechanical engineer. This company, it will be remembered, built the Reade and was engaged in the business of manufacturing steam fire engines. That the change of February 12, 1863, — when a two-inch pipe one foot in length was substituted for a one-inch pipe two feet in length and an automatic valve was substituted for a hand valve — did not go to the essence of the invention or change its character is demonstrated by an examination of the patent itself, which makes no reference to the diameter or length of the tube or the character of the valve. The valve shown in the drawing is a hand valve. Any valve-regulated connection between the two sections of the pump, w'hether the tube be large or small, inside or outside, is within the claim. The change was an improvement, but nothing more.
The Osgood.
The chief of the fire department of Troy, who, until March 23, 1861, had been captain of the Arba Reade Company, wrote, under date of March 26, 1861, to the general agent of the Amoskeag Company, asking how soon and at what price he could furnish the city another engine, “same as the Arba Reade.” The negotiations thus commenced culminated in an order for the Jason C. Osgood. The language of the order, so far as it is necessary to refer thereto, is as follows:
“City of Troy, N. Y„ Aug. 22nd, 1861, by J. C. Osgood.
“One steam fire engine like the ‘Arba Reade’ except the following, viz. steam cylinder little higher, water connection inside,” etc.
There is a dispute as to who suggested the inside water connection. Knibbs says that he proposed the change, but there is evidence to the effect that the suggestion came from the mechanical engineer of the Amoskeag Company. In strong corroboration of the defendants’ contention in this regard is the letter, written August 12, 1861, by Starbuck — the Troy Chief — to the company, in which he says, speaking of the pump connection, “Mr. Bean suggested having that an inside fix.” Bean was the mechanical engineer of the Amoskeag Company.
We do not deem it important to determine who made the original suggestion, as in our view it is immaterial. The inside connection was placed on the Osgood with the knowledge and concurrence of Knibbs and all interested parties. So much is certain. It probably was an improvement, as all of the apparatus, except the valve wheel, was safely out of the way and the entire structure was more symmetrical and compact. But, whether outside or inside, the connection was a perfect embodiment of the invention.
The Osgood was completed and sent to Troy in January, 1862, with the Knibbs invention thereon in its latest and most improved form. The pump sections were connected by a two-inch inside passage or opening. The improvements made a year later on the Reade only increased the diameter of the connection passage to con[849]*849form to that of the Osgood. If, then, the improved engine operated, if she did her work satisfactorily, it would seem that the proposition that her use thereafter by the fire department for the four months preceding May 13, 1862, was a prior public use cannot be controverted.
On January 17, 1862, the Osgood was tested and the Knibbs device worked successfully. Knibbs was present and ran the engine. The entry in the official record book is as follows:
“Friday, Jany. 17.
“Quiet. Steamer Osgood tried by crew of Read (fuel fumisb. by Read) from new hydrant corner of Third & Division. Work well, made 25 lbs. steam 5 minutes; run around pump worked satisfactory: steam valve too much lap.”
On the 27th of January, 1862, the Osgood was put in commission, with her own engineer, and remained in the service of the city for years, performing her work efficiently and well. The Knibbs device was on the engine at least as late as December, 1878, when the testimony of her engineer was taken, and had worked successfully without change during the entire period. The official record shows that she was actually engaged in extinguishing fires on February 22, April 2 and May 10, 1862, the latter being the most extensive fire in the history of the city, destroying a vast amount of property.
The Osgood was ordered, in the due course of business, by the city of Troy for use in the city’s fire department. She was built, delivered and used by the city as the other engines were built, delivered and used. That she should have the pump connection, iron wheels “and other obvious improvements” was part of the specifications for her construction.
Knibbs knew the details of the Osgood’s inside pump connection, of her arrival at Troy, her satisfactory test in January, her subsequent public employment by the city and her complete and unquestioned success.
It is not easy to see how attaching the Knibbs device to the Osgood in such circumstances can be regarded as an experiment; but assuming it to be such, the moment its success was demonstrated the experiment became merged in the invention. Nothing more was needed, all doubt was then removed, the use ceased to be experimental, the invention became an accomplished fact. No one ever suggested any change in or addition to.the Osgood’s pump after the trial test and no change or addition was ever made. Knibbs testified that from the time she was accepted and put into commission as one of the city engines he never ran her or made inquiries regarding her.
The Phoenix.
It is admitted by the complainant that the steam fire engine Phoenix was built by the Amoskeag Company for the city of Hartford, Conn., delivered to the city on November 9, 1861, and used as part of the Hartford fire department for at least 25 years thereafter. The defendants admit that the prior use of the Phoenix was without the knowledge or consent of Knibbs. Unquestionably the Phoenix had upon it the patented device for may years subsequent to June, 1865, but the complainant contends that it was then placed upon the engine, when [850]*850she was sent back to the Amoskeag Company for repairs. The defendants insist that the relief mechanism was a part of her original construction and was never removed or changed. The only question of fact, therefore, relates to this controversy.
The Phcenix was ordered in July, 1861, the order appearing on the Amoskeag’s books August 21, 1861, the day before the date of the Osgood order. The construction of the two engines being substantially contemporaneous it might be inferred that so obvious an improvement as the relief mechanism would not be placed upon one and omitted from the other, but the matter is not left to inference. The engineer of the Phcenix, who represented the city, and the engineer of the Amoskeag Company both concur in stating that the engine when delivered was provided with the relief mechanism. The Phoenix was originally constructed to be drawn by hand and was delivered with the hand rigging upon her. This remained until the department was reorganized and transformed from a volunteer to a paid department in December, 1864. Two photographs of the Phcenix are in evidence which must have been taken prior to December, 1864, for they show the hand rigging which was not removed until that date. They also show the relieving mechanism, or such portion thereof as was visible, viz.: 'the valve stem and wheel. The demonstration is complete that the Knibbs device was on the engine prior to June, 1865, when, the repairs were made, and as it is not pretended that it was added intermediate the date of delivery and the date of the general repairs, the presumption is conclusive that the device was on the Phoenix when she first arrived in Hartford.
The repairs authorized in June, 1865, were new boiler tubes and new tires, but not a new pump or changes in the old pump. In some way the relief device got on the Phoenix, for it was there in 1888 when the testimony was taken. How did it get there if not when the engine was built in the autumn of 1861 ? In the absence of any reliable testimony to the contrary the court has no alternative but to find that relief mechanism was part of the original construction. Indeed, the court understands that complainant does not seriously contend to the contrary but does maintain that the evidence shows that the device was placed there secretly and without authority.
The judge of the Circuit Court reviews with care all the proof relating to the Phoenix and reaches the conclusion, in which we fully agree, that the Phcenix as originally constructed had upon it the Knibbs device. Campbell v. Mayor (C. C.) 47 Fed. 515, 517.
We have, then, leaving out of view the other engines built by the Amoskeag Company, three plain and unmistakable instances of prior use — the Reade, the Osgood and the Phcenix. The first was in public use four years and thirteen days, the second two years, three months and twenty-six days and the third two years, six months and three days before the application for the patent. The use of the Reade is complicated by a number of considerations which do not apply to the others and hereafter we shall confine the discussion to the Osgood and Phcenix. The principal difference between these two is that the use of the Osgood was with the knowledge and consent of Knibbs and the use of the Phcenix was not.
[851]*851The Law of Prior Public Use.
Before proceeding further it is well to ascertain what is the law as applicable to this situation. The patent was applied for and granted in May, 1864. At that time the act of 1839 — explaining, supplementing and superseding the act of 1836 — was in force. Act March 3, 1839, c. 88, 5 Stat. 354. The seventh section of that act is as follows:
“Sec. 7. * * * That every person or corporation who has, or shall have, purchased or constructed any newly invented machine, manufacture, or composition of matter, prior to the application by the inventor or discoverer for a patent, shall be held to possess the right to use, and vend to others to be used, the specific machine, manufacture, or composition of matter so made or purchased, without liability therefor to the inventor, or any other person interested in such invention; and no patent shall be held to ,be invalid, by reason of such purchase, sale, or use prior to the application for a patent as aforesaid, except on proof of abandonment of such invention to the public; or that such purchase, sale, or prior use has been for more than two years prior to such application for a patent.”
The last clause of this section, which is particularly applicable to this controversy, has been condensed and tersely stated by Mr. Walker in his work on Patents (Fourth Ed. § 93) as follows:
“A patent is void if the invention covered thereby was in public use or on sale earlier than two years before the application for the patent.”
The entire subject has been thoroughly examined and the conclusion of the Supreme Court stated in Andrews v. Hovey, 123 U. S. 267, 8 Sup. Ct. 101, 31 L. Ed. 160, and reaffirmed, after a second exhaustive examination, in the same case, 124 U. S. 694, 8 Sup. Ct. 676, 31 L. Ed. 557. The question now at issue was directly involved in that case. The court says, speaking of the numerous defenses:
“It is necessary to consider only one of them, which, in our view, is fatal to the validity of the patent, and that is, that the invention was used in public at Cortland, in the state of New York, by others than Green (the inventor) more than two years before the application for the patent.”
The court says, in construing the section quoted supra:
“The evident intention of Congress was to take away the right (which existed under the act of 1836) to obtain a patent after an invention had for a long period of time been in public use without the consent or allowance of the inventor; it limited that period to two years, whether the inventor had or had not consented to or allowed the public use.”
After an examination of every reported case upon the rehearing the court was confirmed in its opinion that its former decision was correct and closes the elaborate discussion in the following language:
“The second clause of the seventh section seems to us to clearly intend, that, where the purchase, sale, or prior use referred to in it has been for more than two years prior to the application, the patent shall be held to be invalid, without regard to the consent or allowance of the inventor. Otherwise the statute cannot be given its full effect and meaning.”
Among the authorities thus examined and overruled was Campbell v. The Mayor, 20 Blatchf. 67, 9 Fed. 500, which was the original decision holding the Knibbs patent valid.
[852]*852Surreptitious Use.
As before stated the Phoenix was sold and delivered to the city of Hartford and paid for November 9, 1861. She was used by the city thereafter for many years publicly, as part of the Hartford fire department. It cannot be pretended that the Phoenix did not contain a perfect and complete development of the invention. The only contention as to the Phoenix is that this use of the relief mechanism was unknown to and unauthorized by the inventor and that the sale to and use by the city, though open and notorious, was “fraudulent, surreptitious and piratical.”
In distinguishing the Driven Well Case from Kendall v. Winsor, 21 How. 322, 16 L. Ed. 165, the court, in Andrews v. Hovey, 124 U. S. 708, 8 Sup. Ct. 680, 31 L. Ed. 557, makes use of this language:
“It may be that a fraudulent, surreptitious, and piratical purchase or construction or use of an invention prior to the application for the patent would not affect the rights of the patentee under either clause of the seventh section.”
The complainant seizes upon this language as creating a broad exception to the general rule and insists that the facts regarding the Phoenix establish such a use. The doctrine thus enunciated, though not advanced until the argument at the rehearing, may be a fair statement of the law where the facts justify a finding of fraud and piracy. The court, immediately after the statement quoted, says:
“But the present is not such a case as that which existed in Kendall v. Winsor.”
The driven well was invented by Green in 1861. The application for the patent was filed in March, 1866. A public exhibition was made in 1861. More than two years prior to the application Suggett and Mudge put down nine wells having obtained their knowledge from Green and having acted without his knowledge, allowance and consent.
“There is nothing,” says the court, “that indicates in regard to these wells fraud or piracy or surreptitiousness, in the sense of the decision in Kendall v. Winsor.”
In other words, a person may obtain his knowledge of an invention direct from the inventor and may practice.it publicly without his knowledge or consent and such use will invalidate a patent unless the application is filed within the statutory period thereafter. In order to relieve the inventor from the consequences of such use, assuming that relief is possible, it must appear that the knowledge was obtained by deception and that the use was fraudulent or piratical. The meaning of “fraudulent” is too well known to require definition, but it may be wise to recur to the meaning of the other adjectives used by the Supreme Court. “Surreptitious” means, “Fraudulently obtained. Falsely crept in. Obtained by falsehood, fraud or stealth, by suppression or concealment of facts.” “Piratical” means, “Acquired by piracy or robbery.” It is entirely clear that when the courts have used these words in patent causes they intended them to apply to acts done mala fide, clandestinely, [853]*853treacherously and by means of falsehood, fraud or breach of trust. It is equally clear that they have never been used to characterize knowledge obtained openly in the due course of business or applied to an act which is neither malum prohibitum nor malum in se.
Kendall v. Winsor was an action at law. The jury found for the plaintiff. The defendant sued out a writ of error. One of the instructions to the jury sustained by the Supreme Court was:
“That if Aldridge, under a pledge of secrecy, obtained knowledge of the plaintiff’s machine — and he had not abandoned it to the public — and thereupon, at the instigation of the defendants, and with the knowledge, on their part, of the surreptitiousness of his acts, constructed machines for the defendants, they would not have the right to continue to use the same after the date of the plaintiff’s letters patent.”
There was testimony to show that the inventor did not work the patented machines in public and that he pledged his employés to secrecy. No one who did not take such a pledge was allowed to view the machines. One of these workmen was Aldridge who obtained his knowledge by promising not to divulge the inventor’s secret and, thereafter, deserted his employer and entered into an arrangement with the defendants to copy the patented machine for them, and did so. In plain language Aldridge stole the invention and sold it to the defendants with their knowledge and procurement. At least the testimony warranted such a conclusion and the jury so found.
The controversy arose wholly under the first clause of the seventh section of the act of 1839, and it will be more and more manifest as we proceed that neither on the facts nor the law has the decision any but a remote bearing on the questions now in issue.
Pennock v. Dialogue, 2 Pet. 1, 7 L. Ed. 327, was decided before the act of 1836 and has no bearing upon the case at bar apart from the fact that Mr. Justice Story-.asserts, tentatively, what may for the purposes of this case be conceded, that “in respect to a use by piracy, it is not clear, that any such fraudulent use is within the intent of the statute; and upon general principles, it might well be held excluded.”
Shaw v. Cooper, 7 Pet. 292, 8 L. Ed. 689, arose under the act of April 17, 1800, c. 25, 2 Stat. 37, and simply reiterates the opinion of Mr. Justice Story, just quoted. It throws little light upon the present controversy although it may not be inappropriate to note that the judgment for the defendant was affirmed although the plaintiff gave evidence that for the purpose of making experiments he imparted the secret of the invention to his brother who subsequently divulged it for a reward. The court observed:
“Whatever may be the intention of the inventor, if he suffers his invention to go into public use through any means whatsoever, without an immediate assertion of his right he is not entitled to a patent; nor will a patent obtained under such circumstances protect his right.”
In Smith & Griggs Mfg. Co. v. Sprague, 123 U. S. 249, the court, at page 256, 8 Sup. Ct. 122, at page 126 (31 L. Ed. 141), says:
“The thing implied as excepted out of the prohibition of the statute is a use which may be properly characterized as substantially for purposes of experiment. Where the substantial use is not for that purpose, but is otherwise public, and for more than two years prior to the application, it comes within the prohibition.”
[854]*854A not unreasonable construction of this language is that an experimental use is the only one excepted, the fair implication being that a surreptitious use is not excepted.
Pierson v. Eagle Screw Co., 3 Story, 402, Fed. Cas. No. 11,156, was an action at law, tried in 1844 before Mr. Justice Story and a jury. There was no dispute as to the facts; but what the facts were relating to the surreptitious use of the invention by the defendant’s vendor, does not appear. It may be safe, however, to assume that he was guilty of grossly unfair conduct, otherwise the court would hardly have been justified, when explaining the law to the jury, in using such expressions as the following:
“A m'ere wrongdoer, who by fraud or artifice, or gross misconduct, had gotten knowledge of the patentee’s invention.”
“A person who pirates the invention of any party.”
Again, in speaking of the first clause of section 7 of the act of 1839, the judge charged the jury that:
“It could never have been the intention of this clause to confer on a fraudulent purchaser, or a purchaser with full notice a right to use an invention pirated from the original inventor by wrong.”
The question, what constitutes a fraudulent, piratical or wrongful appropriation, sale or use of the invention, is left very much as the earlier authorities leave it, in obscurity. We have been referred to no case since-the clear exposition of the law in Andrews v. Plovey where a plain case of public use earlier than two years before the application for the patent has been held to.be ineffectual as a defense because the use was surreptitious. When the question is fairly presented it may be that the courts will hesitate to introduce exceptions to the rule as broadly stated by the Supreme Court, with the confusion and uncer tainty incident thereto.
If it be once understood that the object, of the act “was to require the inventor to see to it that he filed his application within two years from the completion of his invention, so as to cut off all question of the defeat of his patent by a use or sale of it by others more than two years prior to his application,” the courts will no longer be vexed by the perplexing questions which must frequently arise when the intent of the user and the bona fides of the use are questions to be determined on oral testimony. Isolated instances of injustice to inventors may result, but the remedy is certain and sure. The inventor is master of the situation and has it in his power by prompt action to make the defense of prior public use impossible. Surely two years after he has completed his invention is ample time in which to file his application. If he fails to take so simple and reasonable a precaution why should it not be said that the risk is his and that he cannot complain of the consequences of his own supineness ?
Phcenix Use not Fraudulent.
We do not deem it necessary to decide the question just stated for the reason that we are unable to find that the sale and use of the Phcenix was either fraudulent, surreptitious or piratical. The purchase and use by the city of Hartford was in. the usual course of business and was [855]*855in perfect good faith so far as the city officials were concerned. No unfair conduct can be, and, indeed, none is imputed to them.
Was the Amoskeag Company guilty of fraud or surreption? The judge of the Circuit Court, in the opinion dismissing the bill, 35 Fed. 504, after the decision in Andrews v. Hovey, finds expressly that there was nothing fraudulent, surreptitious or piratical in the conduct of the officers of the Amoskeag Company and he sees no way to distinguish their acts from those of Suggett and Mudge in putting down the driven wells without the knowledge of the inventor. Subsequently, after additional proofs were taken, the judge still refused to find that there was anything fraudulent in the use of the Phoenix and other engines, but he says that it was “a breach of trust and confidence not wicked and probably not thought to be wrong.” He also finds that the Amoskeag people appear to have thought that Knibbs had no right “which they were invading.”
It may be doubted whether this finding is sufficient to warrant the conclusion that the use was surreptitious, fraudulent and piratical, but it is unnecessary to decide the point for the reason that we are unable to find that the officers of the Amoskeag Company were guilty of a breach of trust and confidence.
Where a breach of contract is alleged the first step is to prove the contract. The same is true of a breach of trust. Where there is no relation of trust there can be no breach of trust. Knibbs was engineer of the Arba Reade in the employ of the city of Troy. The city needed another engine and contracted with the Amoskeag Company to build it. Knibbs was never employed by the company or asked to furnish information or advice in the construction of the new engine — the Osgood. He had no relations of any kind with the company. At the request of the chief engineer and with the consent of the city’s fire commissioners the invention was placed upon the Osgood. Knibbs did not employ the Amoskeag Company “to embody the -invention to enable him to continue his experimental use.” He explained his invention in the summer and autumn of 1860 to the officers of the company, after the invention had been on public exhibition, and he testifies that-he then referred to the subject of a patent. He also testifies that he told Bean when the Osgood was under discussion that he wanted his device placed on the engine for experimental purposes only. The natural interpretation of this remark is that he desired to know whether “an inside fix” were feasible. Bean could not have understood him to mean that he wished a device, the utility of which had not been proved or tested, placed upon the engine in order that he might make it successful by subsequent experiment. Neither Bean nor the fire commissioners would have consented to any such use of the city’s property. The city bought the engine, not as an experiment but as a completed and efficacious machine, ready for immediate use. At no time were any disclosures made by Knibbs under the seal of secrecy; at no time was any information imparted in confidence. Indeed, when the Osgood was ordered the invention had been in open and notorious use upon the Reade for more than a year and the invention was as well understood by Bean as by Knibbs. [856]*856There could, therefore, be no secret information imparted in the summer of 1861; there was none to impart. At no time did Knibbs request the officers of the Amoskeag Company to regard their knowledge of the invention as secret or confidential.
We are unable to distinguish the case from Andrews v. Hovey. In both cases there was a public exhibition of the invention; in both the user derived his knowledge from the inventor; in both the use was without the consent of the inventor. It is impossible on any rational theory to hold the one honest and innocent and the other fraudulent and piratical.
The fact that the officers of the Amoskeag Company knew that Knibbs thought of having the invention patented and that he wished the device applied to the Osgood for the purposes of experimentation by him does not alter the legal aspect of the case. If a patent were obtained they took the risk of being treated as infringers in case the application was made prior to the use, but until then they were as free as any other person to use information which was open to any one who had curiosity enough to examine the Arba Reade.
If the mere announcement of an intention to patent has the effect contended for by the complainant it will only be necessary to make the announcement general to extend indefinitely the monopoly of the patent. The patent law recognizes the rights of the public as well as of the inventor and at all stages of the proceedings requires that he act with reasonable promptness. The contention that the Amoskeag Company was debarred from doing what any other manufacturing company could have done with propriety cannot be maintained. It is not doubted that another manufacturer to whom the device had been explained by Knibbs could have placed it on a new engine without being subjected to a charge of fraud. It is argued that “a surreptitious use is a secret invasion of another’s rights, and it is in evidence that Mr. Knibbs knew nothing of this misuse of his invention.” But the Supreme Court has said not only that knowledge is unnecessary, but that proof of the absence of knowledge of the inventor does not render the use surreptitious.
The contention that the invention was not completed till February 13, 1863, and that a use prior to that date was during the experimental period, can be treated more appropriately when we come to consider the use of the Osgood.
Osgood Use Not Experimental.
There is no pretense that there was anything surreptitious, fraudulent or piratical in the use of the Osgood. If we accept the testimony of Knibbs, and we do so for present purposes, the mechanism was placed upon her at his request for the purposes of experiment. From her arrival at Troy in January he was familiar with all the details of her construction and made the first test on January 17, 1862, which demonstrated the success of her relief connection.
That the Osgood contained the invention in its completed condition is admitted. Knibbs testified:
[857]*857“Question. Did you regard this valve of 1860 at the time, or subsequently down to 1863, as a perfected or practical contrivance? Answer. It was not perfected. Question. Do I understand you that it never was. perfected or practical down to 1863? Answer. Not in the Arba lieade steamer. Question. Was it ever on any other steamer? Answer. It was on the J. C. Osgood No. 3, of Troy, put on that engine at my request.”
Again, as showing that in the opinion of the inventor it was a complete invention and ready for patenting in January, 1862, Knibbs testifies:
“As soon as I saw the successful working of it on the J. 0. Osgood I applied to Mr. Norton (the patent solicitor) as to what course to pursue.”
The only question, therefore, regarding the use on the Osgood which is even debatable, is whether it was an experimental use or occurring within the experimental period.
Law of Experimental Use.
The law upon the question of experimental use is well settled and counsel do not disagree regarding it. It is the duty of the inventor to file his application within two years from the completion of his invention. He is permitted to take the time necessary to complete the invention and to make experiments for that purpose, but the moment the invention is completed the two-year period begins to run. The leading case is Elizabeth v. Pavement Company, 97 U. S. 126, 24 L. Ed. 1000. An experimental public use of six years was there held not to be unreasonable. The invention was for a pavement and durability was the principal object sought to be attained. An invention must not only be new, it must also be useful. A pavement that would not last six years, was useless, no one would want it, no one would incur the expense of laying it down. There is but one way to ascertain whether a newly invented pavement is durable and that is to test it by public use. A trial of two years will demonstrate nothing. It may stand the test for that period and then suddenly disintegrate. These-were the considerations which induced the court to uphold the Nicholson patent. As we shall presently see there were no such problems to be solved in the case in hand.
Smith & Griggs Mfg. Co. v. Sprague, supra, is authority for the following propositions: .
First. Where it is clearly shown that there was a public use of an-invention by the inventor for more than two years prior to the application the burden rests on him to establish by convincing proof that the use was for the purpose of perfecting an incomplete invention by tests and experiments.
Second. Where the invention is one of many embodied in a single machine or where the device contains features not included in-the invention or covered by the claims, experiments intended to-produce more perfect working of these extrinsic features are not such as will prevent the running of the statutory limitation. In other words, the experiments must be made for the purpose of developing the invention as described and claimed and nothing else. When the invention is completed the time begins to run and it is-[858]*858of no moment that something else, not a part of the invention, is incomplete and requires tests and experiments to perfect it.
In Egbert v. Lippmann, 104 U. S. 333, 26 L. Ed. 755, the Supreme Court held that the use of one of the patented articles in public was sufficient to constitute public use. The court further observes that:
“Whether the use of an invention is public or private does not necessarily depend upon the number of persons to whom its use is known. If an inventor, having made his device, gives or sells it to another, to be used by the donee or vendee, without limitation or restriction, or injunction of secrecy, and it is so used, such use is public, even though the use and knowledge of the use may be confined to one person.”
When Egbert v. Lippmann was in the Circuit Court, Judge Blatchford said:
“The policy introduced by the act of 1839, and thus continued, is, that the inventor must apply for his patent within two years after his invention is in such a condition that he can apply for a patent for it.” 15 Blatchf. 295. 297, Fed. Cas. No. 4,306.
In Root v. Railroad, 146 U. S. 210, 13 Sup. Ct. 100, 36 E. Ed. 946, the court, while quoting at length from the Pavement Case and fully recognizing its doctrine, observes:
“It cannot be fairly said from the proofs that the plaintiff was engaged in good faith, from the time the road was put into operation, in testing the working of the structure he afterwards patented. He made no experiments with a view to alterations, and we are of opinion, on the evidence, that sufficient time elapsed to test the durability of the structure, and still permit him to apply for his patent within the two years. He did nothing and said nothing which indicated that he was keeping the Invention under his own control.”
In Worley v. Tobacco Co., 104 U. S. 340, 26 L. Ed. 821, the court says:
“The invention was made by Worley alone. He at once began using his invention in McCabe’s factory. He testifies that it was complete, and he became satisfied with the results, in 1871. It is true that after that date he made experiments to decide upon the best mode of constructing his finishers so as to secure the requisite strength; but the finisher constituted no part of his patented invention. In 1871 his invention was complete, and in his opinion successful, and was adhered to from that date, without change.”
In Perkins v. Paper Co. (C. C.) 2 Fed. 451, the court says:
“An improvement has now been made, but it is not described in the specification or shown in the model. At all events, a machine which, whether entirely satisfactory or not, has been run in the ordinary course of business for 20 or for 30 years, and which is patented precisely as it was used, cannot be properly called an experimental machine.”
See, also, as bearing on the questions involved: Hall v. Macneale, 107 U. S. 90, 2 Sup. Ct. 73, 27 L. Ed. 367; Harmon v. Struthers (C. C.) 57 Fed. 637; Swain v. Holyoke Co., 109 Fed. 154, 48 C. C. A. 265; Lettelier v. Mann (C. C.) 91 Fed. 917.
From these authorities we deduce the following propositions, as applicable to the present controversy:
First. An inventor has a reasonable time in which to experiment for the purpose of perfecting the invention' and demonstrating its utility.
Second. The time thus spent, if in good faith, is no part of the two-year statute of limitations.
[859]*859Third. The experiments must be made in perfecting the invention as described and shown.
Fourth. Experiments made in testing parts of the machine not covered by the invention will not have the effect of extending the two-year period.
Fifth. As soon as the invention is completed, viz.: “in such a condition that the inventor can apply for a patent for it,” the two-year period begins to run and the application must be made within this period.
Sixth. The fact that the invention has been improved since its original embodiment does not demonstrate that it was then embryonic or incomplete.
Seventh. When a clear case of prior public use is established the burden is on the inventor to prove by convincing proof that the use was experimental.
Long Period of Experimentation Unnecessary.
The Knibbs invention, though one of unusual merit, was an exceedingly simple one. All agree upon this proposition. One of complainant’s counsel says of it:
“The device is undoubtedly a very simple one, but does not, on that account, any less exhibit invention of a high order.”
Another counsel says:
“Mr. Knibbs’ perfected invention resulted simply in a hole in the partition between the supply and discharge chambers of a pump. Nothing can be more easily understood that this.”
And yet it is contended that it required over two years and nine months of experimentation to demonstrate its utility. Invention, in so far as it relates to machines, may, for the purposes of illustration, be divided into two general classes: First, where the invention is the result of a happy creative thought; and, second, where it is developed by arduous and patient toil. To the first class belong the safety lamp and the driven well; to the second the “Linotype” and the typewriter. The moment the idea of surrounding the lamp with wire gauze occurred to Davy and its practicability was demonstrated by a simple test, the invention stood revealed. Mergenthaler, on the other hand, was not rewarded so much for the inspiration and originality of his idea as for the ingenuity and skill with which he built up his intricate and marvelous machine. The invention of Knibbs belongs to the first class. Undoubtedly his conception showed the genius of the inventor as distinguished from that of the mechanic. It did not, however, require a long series of laborious experiments to prove its efficacy. As soon as the two sections of the pump were connected by a hole on the inside or a tube on the outside and satisfactory tests demonstrated that the old difficulties were thus obviated, the invention was in a condition for patenting. That the application for the patent could have been made before the delivery of the Osgood, in January, 1862, is not disputed. Indeed, the patent as granted describes and shows the apparatus as originally placed on the Reade without, so [860]*860far as we are able to see, the slightest change either in the details of the mechanism or the principles of its operation.
Ample Time for Experiments.
But we now assume, for the purposes of argument, that the entire time from April, 1860, to January, 1862, belonged legitimately to the period of experimentation. Knibbs says that the inside connection was placed on the Osgood at his suggestion and a few days after her arrival she was tested by him and worked satisfactorily. After this test what more was needed? It convinced Knibbs that the invention was then complete, for he says as soon as he saw its successful operation on the Osgood he applied to his patent solicitor. It convinced the solicitor for he advised Knibbs to make the apparatus work on the Reade as successfully as it already worked on the Osgood.
It is admitted by the complainant after the enlarged pipe was placed on the Reade, February 12, 1863, a single test was sufficient to demonstrate its success and complete the invention. The record of this test was entered in the following words;
“Friday February 13 3863; trial, engine taken at 3 P. M. to well on Third Street, South Troy, for the purpose of testing arrangement put on yesterday; relieve pressure at sixty pounds with gates shut, valve open, work freely, keeping up speed all satisfactory, worked about half an hour, made 3,378 revolutions.”
With the exception of an automatic valve which had been added to the Reade, and which we will refer to later on, the relief mechanism was identical in principle on the two engines. In such circumstances it is difficult to understand upon what theory the test of February 13, 1863, should be accepted as final and conclusive, and that of January 17, 1862, should be rejected as insufficient and indeterminate. The mechanism was the same, the test was the same and the result was the same in each instance. If one was sufficient to end the experimental period so was the other.
Suppose that Knibbs had conceived his invention January 7, 1862, had applied it on the 16th to the Osgood, in the form of an inside connection, and on the 17th had tested it and found it worked satisfactorily in every particular, can there be a doubt that the invention would be complete on that day and ready for patenting? If so, how is the case altered because the machinists of the Amoskeag Company did the work at his request rather than the machinists of Troy? The alterations made on the Reade did not advance the invention a hair’s breadth any more than if Knibbs had taken the Osgood’s pump and applied it bodily to the Reade.
But it is said that it was impossible to tell whether the device possessed the requisite utility until it had been tested by actual use at fires. This contention seems inconsistent with the complainant’s position regarding the Reade where, as we have just seen, the two-inch pipe connection was attached one day and the next day after an experimental test, but no test at a fire, the invention was pronounced complete and ready for patenting. But even if this were not so the position seems to us untenable. When the ques[861]*861tion is whether a device on the inside of an engine’s pump works successfully what possible difference can it make whether a fire is or is not burning in the immediate vicinity? Will not the result be precisely the same whether the water from the hose is discharged on a hot surface or a cold surface, a bank of fire or a bank of snow ? If any difference at all can be suggested it would seem to be in favor of the experimental use when the conditions are entirely in the control of the engineer and when tests can be applied which the exigencies of a fire might make impossible.
But if tests at actual fires were necessary such tests were had. The device had been in use on the Reade since the spring of 1860 and its general utility had been proved at a large number of fires. It was thought, however, that the connection between the two sections was too small and the question whether or not a two-inch aperture would furnish the necessary relief was the only question of importance to be determined when the Osgood arrived at Troy. Prior to May 13, 1862, there were three occasions when the Osgood was actually tested at fires. Counsel are not in accord as to the date of one of these fires, but they all agree that the Osgood participated in three fires during the period in question, one of them being the great fire of May 10, 1862, which destroyed a third of the city. And yet Knibbs waited for more than two years after this supreme test before filing his application.
It thus appears that when the Osgood was delivered at Troy she was a complete and successful machine, embodying the invention in its most approved form, a form which has never been changed, and that every test which the ingenuity of counsel can suggest was applied without revealing a single imperfection. Two years passed after the last test was made and still no application was filed.
Automatic Valve no Part of Invention.
But it is said that Knibbs was experimenting upon an automatic valve to take the place of the hand valve and that this valve was not completed until placed by him upon the Reade in February, 1863. Several answers are suggested but a sufficient one is found in the fact that an automatic valve is no part of the invention. Upon this question the patent is the best evidence and no hint of an automatic valve is found either in the drawings, description or claim. On the contrary “the regulating valve H” is plainly shown as an ordinary hand valve. The new valve was probably an improvement although Knibbs testifies:
“Question. Did you regard the automatic form of the valve, which you made in February, 1863, as an improved and better form than the simple hand relief? Answer. Well, I am not prepared to say as to that; it was merely an experiment, that spring was; it was not essential to the successful working of that valve.”
Concede .it to be an improvement, the situation is not altered. Knibbs was not justified in delaying his application until the invention had reached a stage where improvements were no longer possible. Few inventions spring Minerva-like into instantaneous and absolute perfection. The inventive spirit engendered by the rest[862]*862less activity of competition in all the mechanical arts is ever watchful to suggest changes which will enable existing machines to do-their work to better advantage. It is safe to say that in all the great mechanical inventions of the present day the primitive apparatus of the patent can hardly be recognized in the perfected machine which a few years of actual use has developed. To declare that the inventor may withhold his application until the last improvement has been made will extend his monopoly indefinitely and nullify the plain provisions of the statute. The period of experimentation must end at some time and that time arrives when the invention, as described and claimed in the patent, is complete.
Solicitor’s Mistake.
Finally, it is suggested that Knibbs acted throughout under the direction of his solicitor and is not responsible for his solicitor’s mistakes. That the advice thus received was unwise cannot be doubted. Having a complete embodiment of the invention on the Osgood there was no necessity for duplicating it on the Reade. It must be remembered, however, that this advice was given in the early part of 1862. Had it been followed promptly and had the application been filed within a reasonable time thereafter no question of prior public use could have arisen. But leaving out of view every other consideration it will hardly be contended that the mistaken advice of a patent solicitor can override a statute of the United'States. We cannot resist the belief that to the inventor’s procrastination and delay can be attributed the disaster which has overtaken his patent.
Conclusion.
We entertain no doubt as to the correctness of our conclusion that in at least two instances — the uses on the Phcenix and the Osgood — the invention was in public use for more than two years prior to the application and that the patent is therefore invalid.
The decree of the Circuit Court is reversed with costs and the cause is remanded to the Circuit Court with instructions to dismiss the bill with costs.