This is an appeal from a decree which adjudged valid Hosch Design Patent No. 48,248, claims 1 and 4 of Pedersen Patent No. 1,463,589, claim 1 of Pedersen Patent No'. 1,418,162, and claim 2 of Hosch Patent No. 1,434,998, and that further adjudged appellant had infringed as to each upon the rights of appel-lee, the assignee and owner. All of the patents relate to cloth-measuring and pri.ee-computing machines. The trial court adjudged that appellant had .not infringed other claims from which the Measuregraph Company appealed, and that appeal was submitted at a later term. The two appeals, in a broad sense, deal with the same subject matter. Eaeh of the parties manufactures and sells a cloth-measuring and price-computing machine, and it is appellant’s machine, made under later patents issued to A. Vanderveld, which appellee charged infringed upon rights-secured to it by various patents. The ma*498chines are for use in dry goods stores. The mechanism, somewhat complicated, is enclosed in a metallic ease and placed on a counter for use by the clerks. It has a slot or throat in and across one end of the easing into which the clerk inserts the fabric when a measurement is to be made and then pulls the edge of the cloth between two rollers therein, whose circumferential surfaces are in contact, and the pull on the cloth causes them to rotate. By a chain of gearing to one of the rollers, called the measuring roller, its rotation actuates registering devices, bringing into plain view as the cloth is being measured the yardage as it passes between the rollers and its computed cost at the end of the measurement. By simple manipulation of the mechanism the clerk at the end of a measurement sets a brake to hold the registering devices, notches the edge of the cloth at the point it is to be severed from the bolt and separates the rollers. He then withdraws the cloth from the slot and cuts off the yardage purchased as indicated by the notch. Then by another manipulation the indicating or registering devices are-run back to initial or zero position in readiness to begin another measurement. In general appellant’s and appellee’s machines operate in the way that has been described. Neither Van-derveld nor the patentees who preceded him, under whom appellee claims, can be said to be pioneers. They were all preceded by patents, some for cloth-measuring machines and some for cloth-measuring and price-computing machines; but it fairly appears from the proof that the Measuregraph Company was the first to construct a machine that was acceptable to the trade, which was in late 1917 or early 1918. Appellant contends that ap-pellee sold machines to the trade in 1916. Appellee admits it made sales in 1916, but says the number sold was not great, the ma-' chines did not prove satisfactory to purchasers, many were returned and necessary changes had to be made. It spent a great amount of money before it made a machine that stood up to practical use. But much of the prior art, by which its success was achieved, was undoubtedly resorted to by it and its assignors — at least they wei*e charged with knowledge of it. Its efforts were principally directed to bringing the parts of the mechanism .into better co-operation so as to render the machine practically useful to the trade.
Vanderveld’s drawings show the casing which houses appellant’s mechanism, except three levers on its side; but he came after Hoseh No. 48,248) which is a design for housing for measuring and computing machines, and the court held that appellant’s housing infringed his patent. This is the patent drawing of Hoseh’s design:
It will be observed that his design includes not only the casing of the mechanism but also a pedestal on which it is mounted; and the charge of appellee is that appellant’s housing and its supporting pedestal infringed Hoseh, which the trial court sustained. The only ornamental features which we are able to discover are the swelled circular end of the Hoseh housing, the depressed circular inset in the top of the round end of the easing wherein is placed a dial with figures for yardage and moving pointers as the measurement proceeds, and the pleasing blending effect therewith of the pedestal, due to curvation in each throughout. We see nothing else that lends aid to the artistic effect. But appellant’s housing has none of these. Its two ends and four sides are respectively parallel planes. Its pedestal is a straight tubular rod or post. Its easing has nó circular registering device on its top side — that is in the front end of appellant’s machine. Thera are other marked differences, but the ones we have noted seem to wholly eliminate any basis for the contention that appellant is an infringer in appropriating appellee’s ornamental design or its equivalent. Nothing could be plainer or less ornamental as a easing than appellant’s. It has no appeal to the aesthetic emotion, and there can be no infringement when the appearance of the two to the eye of an ordinary observer is not the same and does not arouse the same pleasing sensation. Gorham Mfg. Co. v. White, 14 Wall. 511, 20 *499L. Ed. 731; Kruttschnitt v. Simmons (C. C.) 118 F. 851; R. E. Dietz Co. v. Burr & Starkweather Co. (C. C. A.) 243 F. 592. No ordinary observer who desired to purchase and who had seen appellee's housed machine would be deceived into purchasing appellant’s by any similarity of appearance to him of the housings.
Moreover, on the outside of appellant’s easing are three levers which stand out conspicuously. They are the means by which the clerk operates the machine. Nothing whatever of that kind is shown in the Hosch design drawing. But the machine which ap-pellee makes and sells has three small plates that stand out from the easing and they serve .the same purpose as appellant’s levers. There is a very marked difference in the appearance of the two housed machines because of the wide and striking difference in appearance of the levers and plates; and it would seem to be wholly impossible for anyone, not blind and lost to the sense of touch, to buy appellant’s machine in the belief that it was appellee’s.
Pedersen Patent No. 1,463,589 — Claims • 1 and 4.
Pedersen Patent No. 1,418,162 — Claim 1.
In cloth-measuring and price-computing machines webs on revolving rollers are and long have been used as a means of registering yardage or the computed price of the purchase — more frequently the latter, leaving the yardage to registration by other separate means. Por either purpose the web has figures spaced upon it which come under a window as the web is wound upon a roller by the measuring operation, which actuates the registering devices. This web with the figures on it is called a ehart. It may be a measuring chart to indicate at the window only yardage as the cloth is being measured, and for exactness there may be a fixed pointer under which the figures pass. A ehart may also be used as a means for price computing. In that ease computed figures on the chart on a roller or drum are brought under a window alongside fixed columns of figures indicating prices per yard, so that, for example, if 17 yards which sell at 25 cents per yard has been measured for a customer the computed amount $4.25, will stand by the side of the figures 25 in the price-list column. Pedersen, in Patent No. 1,463,589, proposed a chart that would serve the double purpose, having both yardage and price-computed figures on it; but that is not a point in ■ controversy. Obviously, the diameter of the roller and the ehart upon it as the latter is wound around the former will constantly increase and that necessitates a progressive lengthening of the spaces between the figures on the web; likewise the drum, called the take-up roller, from which the ehart is being unwound and taken, is being progressively decreased in diameter as measured through the drum and chart wound upon it when the measuring operation was begun. The placing of figures on the web having been properly made, it was found at an early date in ehart adjustment, indeed it was apparent to anyone, that the diameter of the drum itself, the thickness of the chart — though as to each the variation might be slight — and the tension on the web would each materially affect correct registration under the window. Exactness was required. A slight difference in drums or thickness in webs cr their tension made a material difference in registration of yardage and computations. As said, this early presented a problem; and Peder-sen dealt with the drum or take-up roller and the chart as it was wound upon it and unwound from it, in both of his patente. His first application, December 9, 1919; was for a process patent, which issued July 31, 1923, as No. 1,463,589, entitled: “Process of Adjusting Measuring Machines,” claims 1 and 4 of which were adjudged valid and infringed. His second application, December 27, 1919, was for improvements in fabric-measuring charts and rollers therefor, which issued May 30, 1922, as No. 1,418,162, entitled: “Pabric Measuring Chart and Roller Therefor,” claim 1 of which was adjudged valid and infringed. The titles do not indicate, except in a most general way, the patented subject matter. The first one might be more appropriately called a process of ascertaining the length of a filler or shim for the drum in a cloth-measuring machine, and the second the filler or shim itself. We think the two patents may be considered together. The claims read thus:
Patent No. 1,463,589.
Claim 1. “In the process of adjusting the size of a roll formed by rolling a measurement ehart past a reference mark and upon a roller, the steps which consist in winding a shim within the roll and shortening the size of the shim until substantially .correct measurement indications are obtained.”
Claim 4. “In a measuring machine having a take-up roller operated by the goods measured and an indicating chart adapted to be wound thereon, the process of adjusting the variable diameter of the take-qp roller *500with the chart wound thereon to cause the calibrations of the chart to indicate exact measurements, which consists in introducing between the chart and the roller a filler continuous in extent and capable of being wound with the chart and of a length such as to cause the progressive diameter of the rolled chart to produce an accurate movement of the chart in exact proportion to the movement of the goods measured and relative to the calibrations of the chart.”
Patent No. 1,418,162.
Claim 1. “In a fabric measuring and cost computing machine, the combination of a casing'having a reading point, a measuring roller, a take-up roller driven by the measuring roller and having a chart with rolls of computed figures attached thereto, and a shim wrapped between the roller and the chart, the shim being of a predetermined length-operating to bring the rolls of computed figures into alignment with the reading point to correspond with the length actually measured by the measuring roller.”
Pedersen’s original application for his process patent 1,463,589 called for six claims. The second claim therein, after being amended in accordance with the suggestions of the examiner, is claim 1 supra. The other five claims were rejected. In making that ruling the examiner said:
“The remaining claims are rejected as being broader than the applicant’s invention. Anyone has a right to put a shim on a measuring roller to increase its effective measuring surface or to pare it down to decrease such surface so as to make it give correct measurements. The claim suggested above seems to cover fully all the patentable novelty in this case.”
Other claims were then presented by Pe-dersen, and after further rulings and arguments the four claims in this patent were finally allowed. Before that point was reached, however, in the Patent Office proceedings the examiner also said:
“The mere winding of a strip of predetermined length on the roller has nothing to do with the process of determining the length of such strip necessary to correct measurement by the roller.”
The rulings of the examiner and his comments at the time, which have been quoted, and in which Pedersen seems to have acquiesced, and the specification and drawings, show that the patent process conceived by Pedersen consists of the steps pointed out by him to be taken in ascertaining the proper length of the shim or filler for the purpose of increasing the diameter of the take-up roller. It seems plain to us, as it did to the examiner, that merely wrapping a piece of paper around the roller, or part-way around it, to increase its circumference, was a simple thing of common knowledge which all had a right to do, and that the filler or shim, so-called, to accomplish that purpose was not a patentable thing. It was obvious to all skilled in that art, and in all others in which drums and rollers are so widely used, and did not constitute invention. Let us see the problem which Pedersen presented for his process patent and the way in which he solved it as shown in his specification. He says:
“In a common type of measuring machines the fabric or cloth to be measured is passed between two rollers including a presser roller which presses the fabric against the face of the other roller which operates as a measuring roller; its rotation is communicated to a take-up roller which draws the chart from a supply roller carrying a tension spring. During the measuring movement the numbers on the chart are presented at a reading point where a fixed pointer may be located. These numbers on the chart may indicate . merely the length of the piece of fabrie measured, or they may be computed numbers and indicate charges to be made for the amount of fabric which the machine has measured. Although the actual diameter of the take-up roller itself may be accurately determined, it is obvious that the effective diameter of the roll upon which the chart is winding, constantly changes as successive layers of the chart are wound upon it. If the chart of such a machine is properly calibrated for a certain kind of paper chart, and for a certain tension of the tension spring of the supply roller, it will be evident that variations in the thickness of the paper stock used for making the charts of the machines, or variations in the tension of the spring, would tend to produce inaccuracies in the measurements indicated by the machines, for example, if a thinner paper were substituted for that which formerly had been used in making these charts, the effective diameter of the take-up roller would be decreased and the machine would have a corresponding error in its indication, and indicate less than the amount actually measured, that is to say, it would have a minus error. The general objeet of my invention is to provide a simple process for overcoming this difficulty in such a way as to enable the effective diameter of the take-up roller to be nicely adjusted to suit special requirements of the paper be*501ing used for the chart or the particular machine being adjusted. The invention consists in those acts to be described hereinafter, all of which contribute to produce a simple and efficient process for adjusting measuring machines. Various factors, principally slight differences in the diameter of the rolls commercially used, differences in the thickness of paper or cloth used for making the charts, affect the movement of the chart as actuated from the measuring roll, so that heretofore it has been found impracticable to make use of the chart directly as a measuring device,, and it has served for purposes of price computing only, and a separate measuring device has been required. These difficulties are obviated by my invention and the chart is made to serve directly as a measuring device. One way of carrying out my invention is by the provision of a shim preferably of paper on the chart roller, which may be made of the proper size to compensate for any inaccuracies in measurement, such as caused by variation in thickness of paper or by variation in the diameter of the roller, etc. The shim may conveniently be made a continuation of the chart itself and rolled up with the chart on the roller. If upon actual test with a measured piece of cloth, the chart is found to overindieate, it is removed from the roll and a section of the paper shim is cut off and the chart retested for measurement, and this is repeated until the indication obtained is substantially accurate. Thus the particular chart is fitted to the particular machine and roll and the computing chart made to serve as a measurement indicator, both simplifying the apparatus and the manner of its use. In practicing my process, the diameter of the take-up roller at the start, that is to say, the actual diameter of the take-up spool or roller should be so small, that when the chart is wrapped upon it in the regular way the machine would have an inherent minus error in its indications, that is to say, the machine would indicate less than the actual length of the piece of fabric measured, but this error should be slight. With a roller of such a diameter, I employ a filler placed between the roller and the chart to give the roller the necessary effective diameter. In doing this I proceed to wrap the chart upon the roller and I simultaneously wrap upon the roller a thin flexible strip which is disposed between the roller and the chart, and may be between the layers of the chart as it winds upon the take-up roller. I continue this winding operation until it is certain that if the flexible strip were cut off, the machine would necessarily have an inherent plus error in its indications. Having cut off the strip at such a point, I observe the character of the indica^ tions and determine approximately the degree of error in the indications. I then partially unwind the chart from the take-up roller until the end of the flexible strip is exposed; and remove a small portion of the end of the strip; I then rewind the chart of the measuring roller and again observe the indications. If the indications are then correct, the machine is passed as perfect in its indications. If, however, the machine still indicates a slight error, I repeat the operas tion just described, or remove another small portion at the end of the flexible strip. In this way, by repeating this process if necessary, a sufficient portion of the strip will eventually be removed to insure accuracy in the indications of the machine. I shall now describe in detail a type of machine referred to, and explain my process specifically. Reference character 10 (in the drawings) designates the chart, and 11, the roll on which the chart is wound and by which the chart is operated for making measurements. Roll 11 is hollow and is provided with slot 12, through which chart 10 passes. Within roll 11 and fixed thereto to turn therewith is the shaft 13. Chart 10 is provided with a looped portion 14, which is slipped over shaft 13 as the chart is inserted in slot 12. This is a simple way to form the shim or strip 15 as a prolongation of chart 10 beyond the looped portion 14, as shown in Rigs. 1 and 2. In carrying out the process, a chart is provided having a shim preferably connected thereto, and the same is put in place on the chart roll with the shim wound on the roll along with the chart as shown in Pigs. 1 and 2, and the other end of the chart is connected to roll 24 (that is, the roll under the window). My invention may be applied to advantage in the practical construction of measuring machines, not only where there is an individual variation in the diameter of the take-up rolls, but it may also be applied to advantage where the diameters of the take-up rolls are uniform, but slightly greater or less than they should be. In case all the diameters are uniform as suggested, it then is merely necessary to employ a flexible strip or shim of a predetermined length, such as will give the take-up roller a virtual or effective diameter which is desired when the flexible strip winds upon the take-up roller. It will be evident that the length of this strip will affect the virtual diameter of the take-up roller. In this way take-up rollers of too small diameter can be utilized without diffi*502culty, and if they are of uniform diameter no tearing off or reducing of the length of the flexible strip is necessary .except to effect the adjustment of one roller; the adjustment of one roller will determine the length that is .to be detached.”
It is observed that his shim or filler is integral with the chart; that at a point where the chart proper and the filler may be said to join, the chart is there looped, that is, the filler is turned back on the chart. The loop is then put through the slot in the take-up roller and slipped over a shaft therein leaving the end of the filler and the chart extending out through the slot. Then by turning the roller the chart and filler are wound together around the roller. If the filler is no longer than one circumference of the roller the chart is wound on the roller over the filler, but if it be longer the excess is wound between layers of the chart around the roller. If the filler proves to be of too great length the end of it is cut off, and this experiment is repeated until the length is found that will give a proper diameter to roller, filler and chart wound upon it. Nothing less than the steps that we have described and to be taken in the manner described would have entitled Pedersen to his process patent. It was this rolling up of the filler on the take-up roller along with the chart, of which the filler was an integral part, and the repetition of that process to ascertain the required length of the filler that constituted the process or art for which the patent was issued.
We come to the undisputed facts, which the court found constituted infringement by appellant and which appellee insists should be sustained. Appellant uses a strip of paper called a shim to increase the diameter of its take-up roller. It is not integral with appellant’s chart. It is not wound on the roller with the chart. If by testing, which is always done before a machine is sent out for use, it be discovered that the chart registers less than the amount actually measured, due to the take-up roller being too small in diameter or from other cause, a strip of paper is pasted on or cemented to the take-up roller. It becomes an integral part of that roller. The employee who makes the test is enabled by experience to estimate the length of the shim needed, though he is not likely to be as accurate as Pedersen. If it then appear that a shim of too great length or thickness has been pasted on to the take-up roller, a part or parts of it are scraped off at the ends or intermediate the ends. This is what the examiner said anyone had a right to do, and is what we believe anyone has a right to do. It was apparent and a matter of common knowledge that the diameter of a roller could be increased in that way, and was not the subject of patent monopoly. We do not hold on this branch of the discussion that the patented process is void when the steps pointed out in the specification are adhered to, but we do hold that the method used by appellant to increase the diameter of the take-up roller by pasting thereto a strip of paper is not Pedersen’s process or its equivalent and does not infringe Pedersen. We further hold that claim 1 of patent No. 1,418,162, for “a shim wrapped between the roller and the chart,” is void for the reasons already indicated. It does not rise to the dignity of invention. It was a mechanical act obvious to all skilled in the art.
Appellant also pleaded public use and sale in this country for more than two years prior to Pedersen’s applications. The trial court did not file an opinion in disposing of the case or assign any reasons, so far as the record discloses, for its rulings. It was content to make bare ultimate findings as the basis of its decree; and we are without the benefit of its views on this issue of fact. We do not know that any were expressed or that the issue was in fact passed on and decided, except as a legal conclusion. And so we are entirely without the help that might have come from that source in estimating the testimony bearing on this issue. The testimony is not lengthy and we have read it attentively several times, and from it we cannot avoid the conclusion that this defense was clearly made out and should have been sustained. That is to say, the proof shows that appellee sold its machines to the trade more than two years before either of Peder-sen’s applications, equipped with shims on its chart rollers, some being integral with its chart and some consisting of a separate piece of paper.
We are not inattentive to the strict rule of proof and the burden placed on one pleading this defense to establish it by the most convincing testimony, and that if there' be a reasonable doubt when all the testimony is considered and weighed the plaintiff should have the benefit of that doubt. We recognize also the settled rule of law that experimental use, that is, such use as is usually made in testing the efficiency of the patented device for the intended purpose, even though carried on in public, is not publie use. Elizabeth v. American Nicholson Pavement Co., 97 U. S, 126, 24 L. Ed. 1000; Smith & Griggs Mfg. Co. v. Sprague, 123 U. S. 249, *5038 S. Ct. 122, 31 L. Ed. 141; Root v. Third Avenue R. Co., 146 U. S. 210, 226, 13 S. Ct. 100, 36 L. Ed. 946. This excerpt from Pitts v. Hall, Fed. Cas. No. 11,192, 2 Blatchf. 229, was quoted with approval in Consolidated Fruit Jar Co. v. Wright, 94 U. S. 92, 24 L. Ed. 68:
“The patentee may forfeit his right to the invention if he constructs it and vends it to others to use, or if he uses it publicly himself in the ordinary way of a public use of a machine at any time prior to two years before he makes his application for a patent. That is, he is not allowed to derive any benefit from the sale or the use of his machine, without forfeiting his right, except within two years prior to the time he makes his application.”
Appellee’s officers testified to these facts: It was not until the latter part of 1917 or the forepart of 1918 that a machine was built by appellee that stood the test; but appellee buüt some machines in 1916 and 1917, though in working them out defects were found and many were taken back, changed and rebuilt. Among the troubles encountered in the machines sold in 1916 and 1917 there was trouble in making the charts come into conjunction with the display opening for the measurement. Appellee made its first sales of its machines, 25 in all, on May 1, 1916, to Nu-gent’s Store in St. Louis, and it made other sales to different parties in that year. Difficulty was early discovered by appellee with respeet to having the figures on the chart always come to the correct position. This occurred almost at the beginning after a few machines were put on the market. There was a shim or paper inserted in the roll to enlarge the body of the chart cylinder, to increase its circumference so that the figures were cheeked in with the sight openings. The addition of the shim simply changed the circumference of the roll. Mr. Kyle was called as a witness for defendant, appellant here. He testified without contradiction that he entered the employ of appellee in St. Louis in May, 1916, that a few machines were out at that time but the company was not selling them to any great extent, that it had a machine fully developed and ready for sale when he entered appellee’s employ. This was Model 115. Numbers were used to designate different models. That the main trouble in installing the chart was to have the proper alignment of figures to correspond with the yardage and the computation on the chart, that Mr. W. E. Hoseh did considerable experimenting and instructed the witness how to take care of this trouble; the thickness of the chart would cause it to vary and in order to get around that “we” used a means of clipping off the end of the chart, or a little flap that lays underneath in order to decrease or throw it in another direction, and if it was necessary to build it up we would either leave it as it was or add thereto, say a little slip of paper, as a rule it was paper something similar to tissue paper, and in this way it would help to make the columns of figures on the chart come up in alignment with the opening and at the same time coincide with the designated yardage on the dial.' He was given those instructions by Mr. Hoseh in the summer of 1916. In the latter part of that year he was sent to Chicago and remained there until the fall of 1917, when he returned to St. Louis and he immediately took up test work on the machines, which included the installation of the charts and the testing of them as to accuracy. He oversaw the work done. Mr. Charles Bacher did the actual chart installation. He would go to Mr. Baeher when an order came in for a machine and have him get it out. He would see Mr. Baeher install the chart, doing it in the same way that he, Kyle, had been in-. structed to do it, “By inserting in the machine and continually running it up to certain places on the dial to see if the chart would have the proper alignment. If it did not he would either cut off to decrease it or build it up by placing a piece of paper in there to make it come to its proper alignment.” He had only been in the employ of appellee a week or two after he first entered in May, 1916, when Mr. Hoseh showed him this method of adjustment. Hoseh was then experimenting with the chart. When he returned from Chicago he tested" the .machines as to accuracy. Mr. Charles Baeher was there at that time. He had entered the employ of the Measuregraph Company and had charge of the chart department when Kyle returned. At the time of the trial Mr. Bacher was still in the employ of appellee and was in St. Louis, where the trial was had. Before the trial came on appellant purchased a Model 115 machine put out by appellee, at Durant, Oklahoma. It was brought into the court room at the trial, taken apart and examined by the witness Kyle. He found on one of the two chart cylinders or rollers in that machine a flap,as he called it, and he also found on the other roller a small strip of paper and said: “This is just exactly the way-they instructed, they cut off her© or added.on here to build *504up. or decrease the rotation of the chart as it turned upon the cylinder.” The flap on one roller had been cut off and on the other “They have placed a slip of paper in there to build it up, as I would call them.” To be sure, there was no testimony as to when the flap in this Durant machine on one roller had been cut off and the strip of paper placed “in there” to build up the other one; but the witness clearly described what had been done in that respect and said that was just what he was instructed to do and what was done in that respect to appellee’s machines that were sent out to purchasers during substantially the whole time he was in its employ. He also testified that Bacher adjusted the charts on the rollers in the same way. This was more than two years prior to Pedersen’s application for either of these patents. Most significant, and of great weight, is the faet that although Mr. Bacher was in St. Louis in the employ of appellee at the time of the trial he was not called as a witness by appellee to refute any of this testimony given by witness Kyle. In no respect was Kyle contradicted. The merchant at Durant, Okla., from whom the machine referred to was purchased or given in exchange for another machine was called as a witness. He identified an invoice for three of appellee’s machines. which he had purchased, dated August 29, 1917. On cross-examination witness Kyle again stated that the strip of paper upon the roller in the Durant machine was placed there in the same way that he was instructed and taught by Hoseh, that he did not know, of course, when the flap on the Durant machine had been cut and the strip of paper placed on the roller, that his duty with appellee on his return frdm Chicago was to test the machines for accuracy, which included the alignment of chart, that he was final inspector, that he had charge of the outgoing machines, that while he was there machines were returned for repairs or adjustment, that sometimes there would he as high as 109 or 150 machines shipped out in a week. He remained with the company until some time in 1921. After he returned from Chicago he did not make the adjustments of charts, that, he assumed, was done by Bacher, he tested them for accuracy.
Witness Marshall was called for appellant. He was with the Measure-graph Company from April, 1919, until the first of July, 1922. When he first went with the company he was given two weeks’ work in the factory and instructed as to what troubles might be expected to develop in the field and how to eliminate the troubles and adjust them. He received his instructions in April, 1919.' He described how a chart would be cut off if too strong and if too weak it would be lifted and some paper pu’ underneath. There are three cylinders in appellees’ machine and sometimes the work of tearing down and building up would involve all three cylinders. The witness had cut off, added on and built up between the three rollers until a perfect alignment of charts was obtained. He knew of no other way by which that could be accomplished. When the witness went to work they'were manufacturing the 116 Model. He thinks it replaced the 115 Model in 1916. The rollers on which the charts are mounted, regardless of any effort to that end, are not exact in size in every machine, there is always a variance. There are two upper rollers and one lower roller. His experience in aligning and testing charts for the Measure-gTaph Company was during the time he was there in 1919 and the three subsequent years, and he had adjusted charts in many machines. He was assigned'to Mr. Bacher, who was then adjusting charts on the fourth floor of the building, and Bacher told him how the adjustments were made. He knew that charts were adjusted at the time machines were sent to customers. He did not know when the adjustments were made upon the Durant machine, Model 115. He knew that two of the machines sold to Durant Mercantile Store were sent back for adjustment, but did not know whether the one in the court room had been sent back. He obtained the machine in the court room from the Durant Company by exchanging one of appellant’s machines for it, and later he obtained the other two. Clearly, this testimony of Kyle and Marshall, uneontradicted, establishes beyond controversy that appellee early in 1916 and in the fall of 1917 was adjusting its charts on its chart take-up rollers by the use of shims or fillers, and that practice was being continued in April, 1919. The two former dates were more than two years prior to either of Pedersen’s applications. In the face of this testimony that Bacher had charge of and was adjusting machines that were sold in' this way, appellee declined to call Bacher, who was in its employ at the time of the trial. No other reasonable conclusion can be drawn from this testimony and the conduct of appellee at the trial than that it was the practice, carried on more than two years prior to Pedersen’s applications, to make adjustments of charts in machines which appellee *505was sending out on sale by the use of shims or fillers, some integral with the chart and some not; and we so find.
Hosch Patent No. 1,434,998.
The trial court held claim 2 of this patent valid and infringed by appellant. It reads this way:
“In a cloth measuring and computing machine, in combination with a frame, a measur--ing roller mounted therein and adapted to be rotated by the goods passed thereover, a circular scale located in one end portion of said frame for indicating a unit of measure and fractions thereof, an indicator co-operating with said scale, said indicator and. scale having a relative rotary movement under actuation by said measuring roller, a web also actuated by said measuring roller and movable transversely of said frame in the opposite end portion thereof, a take-up roller and a drum supporting said web and located above and extending parallel with said measuring roller, the take-up roller being geared to said measuring roller, and a fixed scale bearing numbers indicating different prices of the goods, said web having numbers visible through said frame at a point adjacent to said circular scale for indicating the number of relative rotary movements between said indicator and scale, and having tabulated figures in alignment with the numbers of said fixed scale to indicate the cost of goods for any predetermined number of units or fractions thereof.”
In this patent Hoseh made several material changes of the machine called for in his prior patente. The element in the combination here in controversy found to have been infringed is that part of the claim which deals with the registration of yardage and price computations. In prior patents issued to Hoseh, one as early as October 17, 1916, No. 1,201,735, the yardage, as a measurement was'being made, was. indicated by two revolving hands on a dial in the top of the machine, one pointing to inches or fractions of a yard and the other to whole yards measured up to twelve. This was not new with Hoseh. In McAdams2 measuring and invoicing machine, patented November 26, 1912, moving hands over dials in the end of his machine registered the yards and fractions measured. In Schwartz’ cloth-measuring and cost-computing machine, patented April 7, 1914, No. 1,092,802, his dial or disc in the top revolved and figures on the dial indicated yards and fractions as they passed under a stationary pointer during the measuring operation. Both Schwartz and McAdams had other measuring devices and cost-computing devices on the same faces as their dials and rather near the dials. Appellant has a dial and an are in the front end of its machine on which hands register yards and fractions up to twelve, a hand on the dial registering fractions and a hand on the are registering full yards. There is no controversy about that. From what has been said it would seem there could not be. But Hoseh, in the patent now under consideration, dispensed with the hand on the dial in the top of his machine which indicated full yards, and also with the mechanism which operated that hand. He retained the hand on the dial which indicated inches up to thirty-six, and hence fractions of yards, but there is nothing on the dial which indicates full yards measured. This he accomplished in the patent under consideration by figures on a chart passing under a window adjacent to the dial; and this chart also carries price-computation figures displayed under a window. By this means he achieved an increase in the measuring capacity of his machine, or rather an increase in its registering capacity. His machines before that were limited to measuring and registering twelve yards, and he claimed these changes as improvements. He had eliminated the yardage hand on the dial, the gearing which caused its rotations, transferred the functions of the eliminated hand to his chart and increased the registering capacity. Nothing prevented the rotation without limit of the retained hand around the dial and nothing limited the length of the registering chart except the size of the machine, that is, the capacity of the rollers on which the chart winds. For example, he speaks of the web having a maximum registering limit of 20 yards.
It will be observed from the plan of the top of appellant’s machine hereunder set forth that a small window near the farther or upper end indicates yards, and a window this side of it indicates fractions of yards. The figures 1 in the far window and % in the nearer window are on revolving drums and they register yards and fractions of yards, respectively, as the measurement proceeds up to 12 yards, the capacity of appellant’s machine. There is no controversy about those registering devices. But it will be observed that at the top of the right-hand column of computation figures there are the figures 1%. These figures are on the computation chart and they accord as the measurement proceeds with the figures at the two *506smaller windows on the revolving drums. It is the bringing of these figures to view at the top of the column of computations that is alleged to infringe claim 2.
Appellee’s counsel say the crux of the invention is that as the pointer on Hoseh’s dial, which indicate fractions of yards, makes repeated revolutions over the dial, the numbers of yards on the web appearing through the window will indicate th'e number of times the pointer has revolved around the dial, the capacity of the dial being one yard; that in appellant’s machine this function is sub-served by the numbers at the head of the column of tabulated figures on the chart, which show price computations, and by this means the operator of appellant’s machine is informed of the yardage shown at the two smaller windows. Appellant’s expert put the issue thus: The point of the claim is that the web in the Hoseh patent which shows the yardage has “numbers visible through said frame at a point adjacent to said circular scale for indicating the number of relative rotary movements between said' indicator and scale.”
Counsel for appellant admits that the claim reads on appellant’s machine, but he insists that although the words preferred by a patentee to define his invention may apply literally to another’s device, that does not necessarily prove infringement, that there must be in fact substantial identity between the two to justify a finding of infringement, Edison v. American Mutoscope and Biograph Co. (C. C. A.) 151 F. 767, 773, 774; General Electric Co. v. Allis-Chalmers Co. (C. C. A.) 178 F. 273, 276; Linde Air Products Co. v. Morse Dry Dock & Repair Co. (C. C. A.) 246 F. 834, 838; and he further contends that this element in Hoseh’s combination, yardage registration by figures on the computation chart, disclosed nothing new and was not patentable. Appellant’s expert, of wide experience and high qualification, so testified. In support of these contentions extensive excerpts from the specification, coupled with rulings of the Patent Office on the application, are relied on. It is argued that these show Hosch obtained this patent on simplicity of its construction as compared with his patent No. 1,201,735, issued October 17, 1916, and the increase in its measuring capacity over prior machines; that the Patent Office cited Hosch’s prior patent, which had a chart drum by means of whieh both yardage and price-computation'figures were brought to' view as the measurement was being made. In response, the applicant conceded that his prior patent had a dial in the top of the machine and a pointer there-. on whieh rotated continuously around the dial pointing to figures representing inches up to thirty-six, indicating that one rotation represented one yard measurement and that the drum (whieh is an equivalent of a chart for that purpose) showed in the prior patent the number of yards measured corres*507ponding to the number of rotations of the pointer; but it was argued by the applicant that it was not the function of the drum in his prior patent to record the yardage, that there was another hand on the dial for that purpose in the prior patent and that the reading on the drum was consulted only after the operator had read from the dial by means of the two hands the length of goods measured. It was further argued that the patent applied for represented a very simple and economical machine as compared with the prior patent, that only one hand' was to be employed and all of the mechanism for rotating the other hand was dispensed with; but inasmuch as the retained pointer would not indicate the length of goods measured it was necessary to supplement that pointer with some member such as the web, which contained numbers indicating the number of revolutions of the pointer, and for that purpose it was necessary that the sight opening for the yardage, numbers on the web should be located reasonably close to the dial and in substantially the same line of vision. The specification calls attention to the increase in the machine’s measuring capacity as a new feature. Appellant’s expert referred to this prior Hoseh patent, to Sauermileh No. 662,153, issued November 20, 1900, and to McAdams, No. 1,045,420, issued November 26, 1912, each having a price-computing chart or chart drum which also registered yardage and fractions thereof as the measurement was being made. Sauermileh and McAdams each had two means of registering yardage. And so it is argued there was nothing new in a chart which registered both yardage and price computations, that the only novelty or new features, if any, were the elimination of the yardage pointer on the dial in previous patents, the elimination of the gearing which caused that pointer to rotate, and the increased measuring capacity — none of which have been availed of by appellant in the construction of its machine; and that the measurement figures on its computation chart, which are made to appear at the top of the columns of computations under the windows, is in no sense an equivalent to the new features embodied in claim 2, and is old in the art. We are convinced of the soundness of this contention and believe the court erred in holding that appellant’s machine infringed claim 2.
Appellant’s expert gave it as his opinion on cross-examination that the patents to which he had referred, Hoseh No. 1,201,735 and McAdams and Sauermileh, taJken singly or together, did not completely disclose the structure defined in claim 2 and anticipate it. But that answer, if accepted as a true statement of an ultimate fact, does not settle the question of infringement. A combination may be new and serve a useful purpose and yet a particular mechanical part may be old, which everyone may separately use. Also there may be elements consisting of more than one part, less than all, in a combination which are old as a combination and they may be freely used as a combination, so long as they are not used to make up a new patented combination which contains new and additional parts. Counsel for appellant concedes that Hoseh made a meritorious improvement when he dispensed with one hand on the dial, the train of gearing that actuated it and in connection therewith increased yardage registration, limited only by the capacity of its chart rollers in determining the length of a chart web; and argues that the Patent Office proceedings, which we have summarized and the specification, show this to have been the whole of the subject matter of the patent contract — all else was old. Hoseh in his patent No. 1,201,735 has columns of cost-computation figures on a revolving drum which are brought under a window in the top, and at the head of these columns yardage figures appear as the 'measurement proceeds. As to this drum, he says in the specification, there is “means for operating the same simultaneously and in synchronism with the movement of the hands over the dial,” that is, both hands. A chart drum containing figures that register yards and computations is equivalent to a chart web on rollers with figures that register yards and computations; and certainly thereafter it was not a new discovery to use the chart web in patent No. 1,434,998 to register yards and computations, said web moving in syn-chronism with one of said hands to register its revolutions, that is, show the yards measured, the other hand being removed. When this was called to his attention, that the hand on the dial in No. 1,201,735, retained in No. 1,-434,998, continuously rotated around the dial in both patents, that the chart drum in the 'earlier one and the web in' the later one showed 'the number of yards measured corresponding to the number of rotations of this pointer, his answer was: “This is not the function of the drum, nor is it easily used for that purpose, as the reading on the drum is only consulted after the operator has read from the dial by means of the hands the length of goods measured.” We cannot agree with either of these statements. The yardage figures at the top of the computation *508columns in the first patent serve no other purpose than to register yards measured, which equal the rotations of the hand retained in the second patent, and as to which figures, all in the top of the machine, a clerk might look first on completing a measurement in- the prior patent, would be a matter of choice. Apparently it was intended he should look to both.
Sauermileh has figures on his computation chart, which moves over rollers to register in the top of his machine both yards measured and price computations, as has appellant; and McAdams has the same thing on a chart drum to register yards and price computations in the end of his machine. Near thereto are moving hands on dials that register measurements, the hands and chart drum figures moving in synchronism. In our opinion appellant’s does not infringe.
The decree below, in so far as it adjudged the issues relating to claims of patents herein considered, is reversed with directions to set it aside and enter a decree that Hoseh Design Patent No. 48,248 is valid and not infringed, that claims 1 and 4 of Pedersen Patent No. 1,463,589 are void, that claim 1 of Pedersen Patent No. 1,418,162 is void, and that claim 2 of Hosch Patent No. 1,434,-998 is valid and not infringed, and to allow to defendant below its costs.