HAMILTON, Circuit Judge.

This is an appeal by appellant, the W-R Company, from a decree of the District Court restraining it from the alleged infringement of appellee's mechanical patent number 1,571,813, for which he applied April 30, 1925, and which was issued February 2, 1926. The patent in suit relates to a fixture which appellee devised and attached to a piston ring grinding machine. His appliance removed the burr or wire edge from the periphery of the ring while being ground and with the same machine.

It is claimed by appellant that appellee had dedicated his appliance to public use more than two years before he filed his application. R.S. § 4920, 35 U.S.C.A. § 69. If this be true a description of his patent and its alleged infringement become immaterial.

Appellee, Alfred L. Sova, was employed by the Buick Motor Company in 1921 as a utility foreman and shortly thereafter became foreman of its piston ring department and had supervision of the grinding of all the piston rings used by the Company. When he started working for this Company it used an automatically operated ring grinder, which was old to the art and which left a burr on the ring.

In June, 1922, the appellee made a rough sketch of his appliance and constructed it of scrap metal, and placed it on one of the Buick grinding machines. It worked, but imperfectly, and he attempted to perfect it by making another which was completed before April, 1923, and placed on the machines and worked satisfactorily. Blue prints were made of it by the Buick Company before April 3, 1923, and orders issued by it for the construction of the devices in sufficient quantities to service five grinding machines which the Buick Company then had in operation. To this point the facts are not in dispute.

Appellee testified that the second device magnetized because made of cast iron and in contact with a magnetic chuck in the old machine; that it loaded up with metal shavings causing it to clog and break or scar the rings. He further testified that while blue prints were made of this device and additional ones manufactured conformable to them and of a non-magnetic metal they were not placed in operation until after April 30, 1923, and that their uses until that time were experimental. Appellee had no written record to substantiate his testimony as to the date of the completion and testing of the device. He undertakes to fix the date by the birth of one of his children, which occurred on April 6, and its baptism May 6th. His own birthday fell on May 5th, and he and his family attended services in the cemetery at Saginaw, Michigan, Decoration Day in May, 1923. Appellee says he was in charge of all the machines on which his device was used and he knows they were not thus equipped until after all of the above events had taken place. He also undertook to refresh his recollection from a written report of the efficiency engineer of the Buick Motor Company, dated July 3, 1923, and which he says was made with reference to this appliance after his machines had been in use for thirty days.

The factory manager of the Buick plant testified that the first appliance devised by appellee did not work satisfactorily, but that the second one did and that it was placed on one of the machines during the shut down inventory period between December 23, 1922, and January 2, 1923, and was continued in use until the middle of March, 1923, during which period approximately 150,000 rings were finished on it and passed into production along with the others. He testified the drawings of this device were completed about the first week in April, and additional devices manufactured from the blue print and placed on the other machines and rings ground on them before April 29, 1923.

The Master Mechanic for Buick testified he was in charge of all tool designing, ordering of tools, jigs and fixtures. He testified that he saw the second fixture made by appellee in successful operation before April 3, 1923, and that he saw the original tracings and blue print of this fixture before that time, and he saw the additional fixtures in operation before April 23, 1923, and that there was no delay in making the blue prints and manufacturing the device in his department. The witness identified written orders dated April 4, 1923, issued by him for making the castings and identified receipts dated April 16th and 17th executed in his Department for them. This witness also testified that the factory manager was pressing him to get the additional fixtures built quickly and he put the second fixture constructed by appellee on one of the machines before its drawing was completed. A tool designer for Buick testified he made tracings for the blue prints of the second appliance before April 3, 1923, and produced and identified these tracings with his initials marked thereon, and dated April 3, 1923.

The Manager of the machine and tool repair room of Buick testified he helped appellee build his first device in the Fall of 1922. He observed it in operation and instructed one of the men working under him to build the second fixture, which was completed early in January, and that thereafter it was continuously used for six weeks in the manufacture of rings, when it was removed and drawings made from it and then replaced on the machine for continuous use. He testified that this helper worked overtime to build four additional fixtures, which were placed on the machines, and some of them were in use before April 28, 1923.

A tool maker testified that he was instructed by the manager of the tool room to assist appellee in building his first device, and he saw it attached to the grinder and in use, and that he assisted appellee in building the second device, and it was completed and attached to the machine before the company's 1922 inventory was completed. He also testified that he took this device off the machine to have drawings made and afterwards made additional fixtures from them, and that he installed the third or fourth one on April 28, 1923; that all of the others were manufactured and placed on the machines by working overtime. He says he is able to positively remember the date because he took time off of his overtime for the purpose of purchasing and receiving a new automobile, which was on the installment plan, and the mortgage was executed April 28, 1923. The witness produced and filed the mortgage.

A machine operator testified that appellee's first fixture was tried out on his machine in the fall of 1922; that it worked imperfectly and that it was taken off and another made and placed on the machine late in 1922 or early in 1923; that it worked perfectly but would magnetize and load up with grindings because of its contact with the magnetic chuck. He said it was taken off and returned to the machine after drawings were made, and three or four additional fixtures were manufactured and put on the machines and used in production in the early part of 1923.

An employee of Buick, whose duty it was to receive and deliver machine parts in the Buick plant, testified he received for delivery inside the plant on April 4, 1923, four bronze castings for Sova burring fixtures, and another employee received for delivery inside the plant April 17, 1923, four aluminum castings of Sova burring fixtures. The witness identified receipts for these parts made at the time delivered. None of the devices covered by appellee's patent and used by the Buick Company were produced at the trial. They were discontinued in March, 1925, and a dissimilar device substituted and the old ones scrapped.

The presumptive validity of a patent is so weighty that the defense of invention by another must be established by the clearest proof, perhaps beyond a reasonable doubt and a fortiori the same rule applies to the defense of prior public use or sale by the inventor. It is also true that human recollection is not always dependable. It is weakened by the frailty of memory and tainted by personal interest. Consequently, when dealing with dates or details long since past of the development of a mechanical device, and oral testimony of when the device emerged from the experimental to the finished stage, the Court cautiously weighs such testimony. If, however, all the facts and circumstances convince a reasoning mind that such testimony is unimpaired by faulty memory or unimpeached by personal interest, it is as effective to establish prior invention or public use as written documents. American Roll-Paper Company v. Weston, 6 Cir., 59 F. 147.

In the case before us, appellee admits that he used his first device by attaching it to the grinding machine before December, 1922, and he discovered that it would not work satisfactorily and took it off. He says: "We done but very few rings because with that positive stop the tool would come down and grab the ring and draw it down over that chuck, which would make a mar or scar * * * so we took it off and put it back on again in the last part of January or the first part of February."

Appellee says he was thinking up some way to improve it, and then he states that he completed it shortly after February, 1923, and "the tool did not work perfectly due to the fact that this was all iron or steel. It was subject to be magnetized from the magnetic chuck. That was detrimental because the magnetism would load this up and draw the metals into the point." He then states that this tool was delivered to the draftsman from which drawings were made, and additional appliances constructed made of non-magnetic metal. It follows that the evidence shows, beyond a shadow of a doubt that before March, 1923, the appellee had fully and intelligently expounded the theory of his invention and described the constituent parts and functions of the mechanism by which it was to be reduced to practice, and he had constructed an apparatus capable of practical operation and use. It had ceased to be an experiment or a trial of an incomplete mechanical structure. He had ascertained what changes or additions were necessary to make it accomplish perfectly his designs. He knew that the metal out of which it was constructed when used on the old machine with a magnetic chuck would magnetize and that this was a defect. He knew that this could be obviated by constructing it of bronze or aluminum. By his experiments he had proved the capacity of his appliance to effect what he proposed. When he had done this the patent law assigned to him at that time the merit of having produced a complete invention.

Where the idea of an appliance has been conceived and the conception carried into effect by its construction, which is used or capable of being used for the purpose for which designed, it is no longer an experiment but an invention. Buser v. Novelty Tufting Machine Company, 6 Cir., 151 F. 478; Smith Griggs Manufacturing Co. v. Sprague, 123 U.S. 249, 267, 8 S.Ct. 122, 31 L.Ed. 141; Thomson-Houston Electric Co. v. Lorain Steel Co., 2 Cir., 117 F. 249.

More than two years before the filing of appellee's application, he turned his appliance into the factory branch of the Buick Motor Company, where he was foreman and took no precautions, nor expressed any desire to keep his invention secret. There was little or no difficulty for anyone who chose to see its operation and the employees associated with him were under no obligation of secrecy and the plant in which the appliance was used was large and accessible to many people. Electric Storage Battery Co. v. Shimadzu et al., 307 U.S. 5, 22, 59 S.Ct. 675, 83 L.Ed. ___. His appliance was removed from the machine under his supervision and blue prints made of it in another department of the Buick Company, all without any complaint or protest from him.

At the time the appliance was removed and taken to the drafting department of the company, no further experimentation was necessary to perfect it. Nothing had to be done in a substantial way to change its character. It was a workable, commercial, valuable appliance. Jenner v. Bowen, 6 Cir., 139 F. 556.

More than two years before the filing of appellee's application, the Buick Company retained and used his device with his knowledge and consent. It also made and used blue prints which were available to all who manufactured piston rings for the Buick Company. Early in 1924 it exhibited appellee's device in operation to the officers of appellant and delivered to them drawings of it from which they constructed the devices which are the subject of infringement in this action. Appellee testified that he knew more than two years before the filing of his application that his second device was delivered to the drafting department of the Buick Motor Company for the purpose of making drawings of it. From this, the conclusion is inescapable that appellee knew that the full details of how to make his device were being disclosed to many people. It follows that the claims of the appellee are invalid because of prior public use.

Appellee cites numerous cases discussing the legal concept of experimental use and insists they are controlling. We will not discuss all of them, but in applying them we make the general observation that every opinion must be read as applicable to the particular facts proved or assumed to be proved, since the generality of expressions which may be found in opinions are not intended to be the exposition of the whole law, but are governed and qualified by the particular facts of the case in which such expressions are found, and that a case is only an authority for what it actually decides.

The case of Austin Machine Company v. Buckeye Traction Ditcher Company, 6 Cir., 13 F.2d 697, so strongly relied upon by appellee as controlling, involved an improvement in trenching machines which consisted of a spring cleaning device by which the buckets on a continuous chain would be automatically cleaned by a resilient scraper. It was necessary to test this improvement under varied conditions that would arise in the field, with different soils and unknown difficulties.

The legal principle stated in the case that the period of experimentation ends when it is no longer necessary to determine by use of the device whether the invention is complete or requires modification or change before final adoption is correct. All of the cases cited by appellee are clearly distinguishable on the facts. In the Austin case, as in the others, the experiment was necessary to determine whether the invention was complete. In the case at bar no further experiment was necessary after appellee had completed and tested his second device, which was done more than two years before he applied for his patent.

The decree of the District Court is reversed with directions to dismiss appellee's petition.