LEWIS, Circuit Judge.

Appellant brought this suit charging that appellee was manufacturing and selling a fabric-measuring and price-computing machine, which infringed upon its exclusive rights under certain parents which it owned, to make and vend a machine of the same class. The usual relief was asked. As to some of the claims of some of the patents the plaintiff prevailed, from which defendant appealed; as to other claims the plaintiff failed and it has brought this appeal. The claims here relied on by appellant as infringed are nine in all; claims 1 and 21 in Pat. No. 1,369,663, issued to Hosch February 22, 1921, on his application of March 26, 1915; claims 6 and 7 in Reissue Pat. No. 14,898 to Scott on June 22,1920, his original application being January 29, 1917; claim' 1 in Pat. No. 1,453,300, issued to Wheeler on May 1, 1923, on his application of December 20, 1921; claims 3 and 8 in Pat. No. 1,434,998, issued to Hosch’s admx. on November 7,1922, on his application of November 15,1917; claim 12 in Pat. No. 1,371,028, issued to Hosch’s admx. and Turner on May 8, 1921, on their application of October 12, 1918, and claim 23 in Pat. No. 1,437,207, issued to Stocke on November 28, 1922, on his application of March 12, 1920 — -all said patents being assigned to appellant.

The Wheeler patent is for mountings for measuring machines. Each of the others named is for a measuring and computing machine, and each calls for parts in combination that make up the completed machine, and also teaches the mode of operation. Hosch began his work in 1911 or 1912, and after much experimentation, the appellant put upon the market in 1917 or early in 1918 its commercial machine for which it has since found ready sale. Appellee claims some sales were made in 1916. Others were and had been engaged on the problem and other patents than those named were issued for measuring and price-computing machines — some before those under which appellant claims and some after.

Appellee makes its machine under patents issued to Yanderveld, all subsequent to those assigned to appellant.

In general the two machines are alike in combination and operation. Each is enclosed in a metallic casing which is mounted and placed on a store counter for use. Access to the inclosed machine for use and operation is through a slot or mouth in one end of the casing, into which the edge of the cloth to be measured is inserted; and it is operated by levers or like means connected with the machine through the easing. The edge of the cloth is placed in the slot between two rollers, called the measuring roller and the presser roller, which at the time are apart. In appellant’s machine the measuring roller is under *265 the presser roller, in appellee’s their relation to each other is reversed. The clerk having inserted the edge of the cloth in the position noted, then by manipulation brings the presser roller in contact with the cloth, and its edge is thus held between the two rollers. The clerk then begins the process of measuring hy pulling the cloth between the two rollers from left to right. This causes the rollers to rotate and the rotation of the measuring roller operates through gearing the whole mechanism, causing it to register in plain view by moving pointers on dials the yardage desired and its computed sale price to be displayed under windows in the top of the machine. In appellant’s machine there is a dial on its top and two hands which move with the measurement operation and point to the yardage and the fractions thereof as the cloth is being pulled through. In appellee’s machine a dial and an. arc which serve a like purpose are in the front end of the machine. A moving hand on the dial points to the fractions of a-yard and a moving hand on the arc points to the number of yards as the cloth is pulled through. It also has charts under windows in the top surface which show the yardage. These two exhibits illustrate the arrangement on the top of each machine. Also the front end of appellee’s on a smaller scale.

Appellant’s

Appellee’s,.

The two left-hand columns of figures as shown in each are stationary. They represent prices per yard. The two right-hand columns of figures in each are on webs under windows. These webs are taken up on rollers with the measuring operation and display the price-computations. Appellant’s exhibit, supra, illustrates the results when two yards have been measured; appellee’s when one and a half yards have been measured. It is observed that the computation is brought immediately opposite the yardage prices at which the goods are being sold. When the desired yardage has been measured, the machine is locked and at rest, and must he returned to the initial point before another measurement is begun. By manipulation of means for that purpose the clerk clips or notches the edge of the cloth at the point where it is to be severed from the bolt. The presser roller is then *266 lifted or lowered from its contact with the cloth, the edge of the fabric released, and the measurement and computation are complete.

Having given this general description, we take up in the order named above the claims which appellant charges are infringed by ap-pellee’s machine.

Hosch Patent No. 1,369,663 — Claim 1.

“In a machine of the class described, the combination with a rotary member mounted so as to be actuated by the movement of the goods thereover, of a dial, a pointer moving over the dial, means for effecting a movement of the pointer corresponding to the movement of the rotary member, a pair of stationary price scales, a pair of webs cooperating respectively with said price scales, and bearing rows of figures indicating the cost totals of yards of goods and frgetional parts thereof, a drum on which both of said webs are normally wound, and means for imparting simultaneous and corresponding movement to said webs, which movement corresponds to the movement of the rotary member and the said pointer.”

The infringement charged is in respect of that part of the combination underscored, supra, though the remainder of the claim has a bearing on the inquiry. The pair of webs in appellant’s machine must be wound around the drum, to which one end of each web is attached, before the measurement operation is started, one web throughout and continuously overlying the other on the drum, and each having its other end attached to a roller, one roller being under one of the windows by the side of the stationary price scale and the other roller being under the other window. On one side of each of these webs are the price-computation figures. As the measuring op-, eration proceeds these rollers revolve and simultaneously draw from the drum the two webs, thus bringing the computation figures thereon into proper relation with the price scales. Before another measurement is made, the drum, in the resetting of the machine, takes these webs from the two rollers. They are thus wound back and forth from the drum around the rollers and from the rollers around the drum, being around the drum in greater part when the measuring operation begins and around the rollers in greater part when it stops.

In appellee’s machine the price-indicating mechanism has no drum; it has two rollers, and only one web with computations on each side thereof. One end of the web is attached to the roller under one window and the other end to the roller under the other window. The rollers revolve, when a measurement is being made, and also when the machine is being returned to its initial position, in opposite directions. So, it may be said, the computation figures on the upper side of the web are brought under one of the windows, and those on its under side are brought under the other window.

It is argued on the rule of equivalents that this price-computing element in appellee’s machine infringes claim 1 supra, that appel-lee’s web with computations printed on both sides performs the exact function of appellant’s pair of webs each printed on one side, in that, each causes columns of figures to appear simultaneously in co-operative relation with the two price scales. Appellee admits that appellant’s two-web arrangement was an improvement and patentable, but it claims that the other parts of this element which it uses were old, appellant was not a pioneer, and is not entitled to a broad construction as to equivalents, and that, in fact, appellee’s one-web arrangement is taken precisely from Sauermileh No. 662,153, an expired patent for measuring textiles and registering computed sale prices. It is further contended for appellee that to hold a single web with computations on both sides the equivalent of appellant’s two webs would render the Hosch patent void as to this element in appellant’s combination; for inasmuch as appellee’s web is taken from Sauermileh, Sauermilch’s patent would anticipate Hosch. Hoseh’s purpose in having two webs is plain. He stated it in his patent:

“I make use of two flexible charts (webs) and two price per yard strips in order to materially increase the capacity of the machine without necessitating the provision of an extra long housing, such as would be required to take care of a single long price per yard strip and a corresponding wide single flexible chart.”

He divided a wide web, which required long rollers and which co-operated with a single long list of stationary prices, into two; thus enabling a great reduction in the length of the machine without decreasing its computing capacity. In that it is conceded to be an improvement.

Sauermilch’s patent was issued in November, 1900, and in the present state of the art his machine now would seem crude and cumbrous in size, shape and operation. It was operated by hand in turning a crank attached to a shaft. The goods to be measured was drawn through between two rollers in contact as if fed into the machine and not pulled through, and before the measurement began a *267 stop pin was set for the number of yards and fractions to be measured. Other preliminary manipulations were required. But clearly, he had a list of prices at two windows, through which the yardage measured and the computed amount of the sale price appeared. His price-computing chart had figures on both sides, the figures on one side appearing at one window and the figures on the other side appearing at the other window. To accomplish this he operated it over six rollers. Whatever there was of invention in this has now become public property and may be used by any one. We cannot escape the conclusion that the price-indicating mechanism in appellee’s machine, in the respects now considered, one web with computations on both sides, is that of Sauermileh, and there was-no infringement as to that element of the combination, a pair of webs each with computations on one side. Appellant is entitled to exclude others from the use of its improvement, but it cannot exclude them from using what the art had long taught, a device now dedicated to the public, improvement upon which may be said to have been made by Vanderveld.

Hoseh Patent No. 1,369,663 — Claim 21.

“In a cloth measuring machine, the combination of a rotary member mounted so as to be actuated by the movement of the goods thereover, indicating mechanism including a dial and a pointer movable on said dial, mechanism driven by said rotary member for driving said pointer, and means controlled by said last named mechanism for arresting the rotation of said rotary member when said pointer has made one revolution on said dial.”

The part of the claim underscored is alleged by appellant to be infringed by defendant’s machine. It calls for means which automatically stops appellant’s machine when its measuring capacity has been reached. It will be observed that the “last named mechanism” refers to the mechanism driven by said rotary member which drives said pointer, and the rotation of the rotary member, the measuring roller, is arrested by means controlled by said “last named mechanism,” the mechanism for driving the pointer on the dial. In the chain of that mechanism between the roller and pointer for driving the pointer and as a part of it, there is a control gear wheel, in appellant’s machine. A pin projects upwardly from the edge of this wheel. When it makes one revolution the measuring capacity of the machine, 12 yards, has been reached. This is indicated by one of the moving hands on the dial. As heretofore said, the entire machine in all its parts is operated by the measuring roller as the cloth is drawn over it in the measuring operation. This projecting pin, when the gear wheel has made one revolution, comes in contact with a pivoted arm which causes a pawl attached to the end beyond the pivot to engage with a notched disc by means of which the measuring roller is automatically stopped.

The appellee’s machine is also stopped automatically when its measuring capacity, 12 yards, is reached. But it contends that the means it uses to accomplish that result is different from that of the appellant and operates upon a different principle, that the means it employs is not attached to, operated by or controlled by the “mechanism driven by said rotary member for driving said pointer.” In fact, the mechanism interposed between the measuring roller and the pointer on the dial may be all taken out of appellee’s machine and still the measuring roller could be automatically stopped at a given yardage. The measuring roller and the shaft with which it revolves are rigidly and permanently attached to each other. Appellee’s shaft is threaded. A half-nut is carried along the-threaded shaft while- the measuring operation is being made. When this nut reaches its extreme position a pin earned by it becomes so engaged it stops the roller at the time the pointer indicates 12 yards, its measuring capacity; and in consequence all other parts of the machine, including the gearing which causes the moving hand on the arc of its machine to indicate the yardage, as well as webs on rollers under windows in the top of its machine, automatically stop. The result of the means used by each is the same, but ap-pellee’s is clearly different from that called for in claim 21 and there is no identity in mode of operation between the two. Appellant primarily stops that part of the mechanism which drives the yardage pointer and secondarily the measuring roller, while ap-pellee goes direct to the seat of power which operates the whole machine. Its means for arresting the rotation of the rotary member is controlled by the revolving shaft, integral with the rotary member. We see no similarity in the two, except the result; mechanically and operatively they are essentially different and function differently. Moreover, the trial court found that appellant’s means of stopping the machine was anticipated and appel-lee cites in support of the finding two patents to Schwartz, No. 1,090,796, issued March 17, 1914, and No. 1,092,802, issued April 17, 1914. The first was for a fabric-measuring and cutting machine, and the second for a cloth-measuring and cutting and cost-comput *268 ing machine. In the latter his specification calls for friction rollers to be rotated by frictional contact with the cloth drawn between them in the measuring operation, for automatically locking tbe machine when the length desired has been passed through it. He had a revolving dise and a stationary pointer on the dise which showed the yardage — not a revolving pointer and stationary discs, as in appellant and appellee. He first set the disc so that the pointer indicated the yards and fractions to be measured, and the dise revolved during the measuring operation until it brought zero under the pointer, when the machine is automatically stopped. He has a projection on the revolving disc which comes in contact with a sliding pin controlled by a spring, and these in co-operation automatically lock or stop the roller when the desired yardage is reached. It is difficult to avoid the identity between the two means used and the modes of operation by appellant and by Schwartz. We think the latter anticipated the former. Moreover, in view of Schwartz the claim in the respect now being considered should have a narrow construction, confining appellant to its exact means.

Scott Reissue Patent No. 14,898 — Claims 6 and 7.

Claim 6. “In a cloth measuring machine, the combination with a measuring roller, of a registering device actuated therefrom, a brake for preventing said registering device from returning to initial position, a cutting knife, means for simultaneously operating said cutting knife and setting sadd brake, and manmally operated means for\ releasing sadd brake.”

Claim 7. “In a doth measuring machine, the combination with a measuring roller, of a second roller cooperating therewith, a registering device actuated from said first named roller, a brake for preventing the return of said measuring roller and registering device to initial position, means for simultaneously disconnecting said rollers and setting said brake, and ma/nually operated means for releasing said brake.”

The underscored parts of the claims constitute the elements of the combination alleged to be infringed upon by appellee’s machine. It will be convenient to consider the two claims together.

Scott was not a pioneer in devising combinations in machines of the same class which accomplished all of the results achieved by his combination. Hosch No. 1,369,663 preceded him. Hosch called for a machine in which the cloth was drawn between a presser roller and a measuring roller, which caused the latter to put in action a train of gearing which operated registering devices of the yards that had been drawn through. It has a cutting knife to mark the correct yardage and indicate its point of severance from the bolt. On completion of the measurement manual depression of a pin in a slot in the side of the casing causes through pivotal connections the knife to notch the goods and the rollers to be simultaneously separated; and his gearing was so connected with other means that his registering device did not return to zero when the rollers were separated. The end of a lever extended through the easing and when the operator was ready to return the registering device he pressed down the lever, which caused the gearing to be disengaged with the means that held it against reverse motion. It should be said that in appellant’s machine only the indicating mechanism of yardage and price computation returns to zero, while in Scott’s and appellee’s machines the measuring roller, charts and indicators are all returned to their initial position. Also in Gebhart No. 1,250,843, who preceded Scott, the cloth is drawn between a measuring and presser roller. He has a knife to notch the cloth at the required length. This is done by depressing a handle and the operation simultaneously moves the presser roller away from the measuring roller. His registering device is on a revolving drum under a window and his gearing is so adjusted as to cause a brake to hold the drum with the registered figures from turning back to the initial position. The purpose in all of them in separating the rollers and holding the registering device is to enable the clerk to remove the goods from the machine and note the yardage and computation. It is then returned to zero in readiness to begin another measuring operation. For that purpose, in using the Gebhart machine a stem or rod with one end just through the casing is pushed inwardly by the clerk, which causes disengagement of a clutch and this permits the drum to rotate reversely and return to zero position under the pressure of a spring. Jones No. 443,528 also preceded Scott. His measuring machine has presser and measuring rollers between which the cloth is drawn. The presser is driven by the cloth and it operates a train of gearing which causes two discs to register. It has a knife for notching the goods and holding means which prevents a return of the mechanism to zero until the appropriate time. It has no means for separating the rollers, the spring for returning the mechanism to zero being strong enough for *269 that purpose when the rollers are together. When it is desired to return the registering device, means are provided for that purpose. It seems obvious that a machine of this class would be entirely useless if there were no means for holding the registration when a measurement is completed, and more so if there were no means of -returning it to zero for another measurement. The art made it apparent to Scott at the time of his original application that means had been devised for operating the cutting knife, setting the brakes and disconnecting the rollers at substantially the same time, and manually operated means for releasing the brake had also been provided. It seems clear that all he could obtain was by way of improvement on what preceded him. In Railway Co. v. Sayles, 97 U. S. 554, 556, 557 (24 L. Ed. 1053), it is said:

“If the advance towards the thing desired is gradual, and proceeds step by step, so that no one can claim the complete whole, then each is entitled only to the specific form of device which he produces, and every other inventor is entitled to his own specific form, so long as it differs from those of his competitors, and does not include theirs.”

In Wichita Visible Gasoline Pump Co. v. Clear Vision Pump Co. (C. C. A.) 19 F.(2d) 435, we had occasion to quote this from Walker on Patents:

“Omission of one element or ingredient of a combination covered by any claim of a patent averts any charge of infringement based on that claim, whether or not the omitted ingredient was essential to the combination of the patent, and whether or not it was necessary to the operativeness of the device ; and it makes no difference that another element is made to do the work of itself and of the omitted element. A combination is an entirety. If one of its elements is omitted, the thing claimed disappears. Every part of the combination claimed is conclusively presumed to be material to the combination, and no evidence to the - contrary is admissible in any case of alleged infringement. The patentee makes all the parts of a combination material, when he claims them in combination and not separately.”

And we said, citing eases, that the principle announced was overwhelmingly supported. Among other cases cited was Jewell Filter Co. v. Jackson (C. C. A.) 140 F. 340, wherein we said:

“The absence from a device that is alleged to infringe a patented combination of a single element of that combination is fatal to the claim of infringement.”

Again, in Anakin Lock Works v. Dillon Lock Works (C. C. A.) 292 F. 45, we said:

“The claims of a patent involving mere improvements, in view of the prior art, are tb be narrowly construed and limited to the particular mechanism described, and a device which accomplishes the same result by means of different mechanism is not an ‘infringement.’ ”

With these principles in mind we come now to a comparative consideration of the means used in the two machines, Scott’s and appellee’s, for operating the cutting knife, setting the brake, disconnecting the rollers and releasing the brake. To accomplish the three first-named purposes Scott manipulates an elongated sleeve which extends outwardly through the easing of his machine and has a button or perforated cap on its outward end. When a measuring operation is completed the button is grasped, the sleeve pulled outward, turned upward and the button made to engage with an external catch, causing a spring to press against the measuring roller which holds -it and the registering device in its then position. The mechanism cannot then run back to zero until this spring is released. This manipulation of the sleeve also causes in quick succession, said in the claim to be simultaneous, a separation of the rollers and the knife to press upon and notch the goods as it is drawn out. The yardage and computation of price are noted by the clerk and the goods severed from the bolt at the notch in the edge of the cloth. In order to permit the indicating mechanism to return to zero Scott has a push rod extending entirely through said elongated sleeve and movable lengthwise therein. There is a button on the outer end of this rod. The rod is-pushed inwardly by pressure on the button and this releases the spring which pressed upon and held the measuring roller, thereupon the mechanism all returns to zero. The pressure on the button on the end of the rod does not move or disturb in any way the knife or the presser roller. Scott also says, in effect, in his specification that the brake may be released from the measuring roller without pressing on the button on the end of the sliding rod in the sleeve, that the sleeve may be so manipulated as to release the brake; but even so, it is the inner end of the rod in that manipulation that causes the release.

In appellee’s machine, when a measuring operation has been completed the clerk presses downward with his thumb the three levers shown in the photograph of appellee’s machine, supra. One of them sets a brake which holds the mechanism, one of them notches the cloth, and one of them separates the *270 presser from the measuring roller. These levers can be operated successively, each performs a separate function; but owing to their proximity the thumb can bridge the spaces between them so that they can be pushed downward simultaneously. It appears that they are usually operated in that way, and that appellee in its circulars of instruction teaches that they be operated in that way. This operation of the levers in unison seems to be the basis of appellant’s complaint, that it is mechanically the same or an equivalent to Scott’s sleeve. When the clerk is ready to reset appellee’s machine he raises the reset lever and this causes it to run back to zero. The other levers are all lifted to up position before another measurement is made. Appellee controverts appellant’s contention and it claims that Scott’s “manually operated means for releasing said brake” in his combination is not found in its machine, that one of its levers performs the two functions of setting the brake and releasing it, whereas Scott sets the brake with his sleeve, and adds a separate and independent means, the sliding rod in the sleeve, with which he releases the spring in order to reset the machine. The combinations of Scott and Vanderveld were mere improvements, in view of the prior art, each is to be narrowly construed, each limited to the particular mechanism he describes. While they accomplish the same result their mechanisms are widely different and their modes of operation are not the same: Furthermore, it is our opinion that Scott has an independent and added element in his combination, separate and apart from the means which set the brake, separate the rollers and notch the cloth, not found in appellee’s machine — the push rod which resets Scott’s machine. On the authorities cited we are constrained to hold there is no infringement.

Wheeler Patent No. 1,453,300 — Claim 1.

“In a mounting for a measuring machine to be secured to a counter, and which is operated by pulling the fabric through the machine parallel with.the edge of the counter, the combination of a rail, a shoe mounted to slide along the rail and having fixed bearing cheeks to engage and slide upon the rail on its opposite sides, and a measuring machine supported on the shoe with its center of gravity disposed in a' vertical plane removed from the plane of the rail, whereby the forces developed at the said bearing cheeks of the shoe are reacted to by the said face of the rail, and operate to offer sufficient frictional resistance to prevent the measuring machine from being pulled along the rail by the fabric during the measuring movement.”

The weight of appellant’s machine as it rests on the rail attached to the rear edge of the counter causes it to tend to lean forward over the counter, which, of course, demonstrates that a plane through its center of gravity parallel, with the rail is on the counter side of the rail. It is claimed that the weight of appellee’s machine as it rests on the rail causes it to tend to lean rearward, which likewise demonstrates that its center of gravity is on the other side of the rail. This tilting or leaning tendency causes an impingement of the shoe on the rail; and as the cloth is drawn between the rollers further impingement in the shoe is caused by resistance to a rotary tendency of the machine due to the pull of the cloth. Obviously, these impingements hold the machine from being pulled along the rail during a measuring operation, and the strength of this resistance depends on how far co-action of shoe and rail has permitted it to lean. The measurement completed, it may then be moved along the rail if needed elsewhere. As to whether it will move freely without holding it against its tendency to lean depends on many things —primarily the relative location of center of gravity and rail, and Secondarily the size and shape of the shoe, the closeness of contact between shoe and rail, the nearness of the shoe to the side of the machine-supporting pedestal with which it is integral, and the weight of the machine; and Wheeler gives no rule or formula for their co-ordination in a given ease, or in any case. The distance between a plane through the center of gravity and a parallel plane through the top of the rail on a given level is fixed by the weight of the machine as it rests on the rail. Wheeler does not specify the size of shoe, thickness of rail, nearness of shoe to pedestal, or weight of machine; neither does Pedersen No. 1,463;589, relied on as anticipation. In those respects they are alike. Without stopping to describe the mechanical means of ap-pellee by which it attaches its machine to the counter and along which it is moved laterally for use at different points, it may be conceded for present purposes that appel-lee’s combination is a mechanical equivalent to that of Wheeler. For it seems clear to us, as found by the District Court, that Wheeler was anticipated by Pedersen, whose application was more than two years prior to Wheeler’s, and also that Wheeler was not in fact the first discoverer. Pedersen’s accompanying drawing, to which Ms specification refers *271 and describes, as “thereby providing for the convenient use of the apparatus at any point along the counter,” leaves no discoverable difference between Pedersen and Wheeler. Pedersen shows the rail attached to the rear edge of the counter in the same way as in Wheeler, a shoe shaped exactly like Wheeler’s on the lower end of the supporting pedestal, into whieh shoe the rail is slideably fitted so as to slide the machine along the rail, the shoe having the same fixed bearing cheeks to engage and slide upon the rail on its opposite sides as in Wheeler. It seems clear from Pedersen’s drawing that a plane through the center of gravity would not be coincident with a plane through his rail, but to one side. It would, indeed, be a greater task to balance the machine on the narrow rail with exactness so that it would not overhand rearward or forward. Pedersen says nothing about the center of gravity in a vertical plane removed from the plane of the rail; and so it is argued Pedersen did not teach how his combination should be set up in order that the binding action or impingement in the shoe would hold the machine during the measuring operation and permit its ready movement along the rail at other times. The argument is not convincing. Pedersen taught all that Wheeler taught. If his patent is void on that ground, so is Wheeler’s. Each left to experimentation and practical test a workable adjustment of the combination. Reflecto lyte Co. v. Luminous-Unit Co. (C. C. A.) 20 F.(2d) 607. In substance, all that Wheeler did that Pedersen did not do, was to call attention to a law in physics, that is, in order to cause the impingement the plane through the center of gravity must not pass through the points of impingement — he suggested no method or formula for ascertaining the location of the plane of gravity and ^adjusting the shoe on the rail in relation thereto. And the pull of the cloth, whether the machine was in an upright or leaning position, necessarily would tend to prevent rotation by impingement in the shoe — matters of common knowledge. Granting invention, the combination of Wheeler seems to be identical with that of Pedersen.

The undisputed evidence shows that a combination in every respect like Wheeler’s was manufactured under Pedersen’s supervision, sold and used for the same purpose long prior to Wheeler’s application.

Hosch Patent No. 1,434,998 — Claims 3 and 8.

Claim 3. “In a machine of the class described, the combination of a rotary member mounted so as to be actuated by goods passed through the machine, a fixed pnce scale, a web having tabulated figures alining with the numbers of said price scale, to indicate the cost of different quantities of goods, means for guiding said web past said fixed scale, means for driving said web from said rotary member, means for exerting tension in said web to return the same to its zero position after each measuring operation, means for stopping the return movement at the zero position of the web, cm3, means for regulating the speed, of the return movement to prevent a shock to the web when arrested at its zero position.”

Claim “8. “In a machine of the class described, the combination of a rotary member mounted so as to be actuated by goods passed through the machine, a fixed price scale, a web having tabulated figures alining with the numbers of said price seale, to indicate the cost of different quantities of goods, means for driving said web from said rotary member, means for exerting tension in said web to return the samé to its zero position after each measuring operation, meanss for stopping the return movement at the zero position of the web, and means for regulating the speed of the return movement to prevent a shook to the■ web when arrested at its zero position.”

The underscored parts present the matter in controversy as to these claims. They are alike and may' be considered together. The District Court held there was no infringement and did not pass on the other defenses.

In Hosch’s described machine during a measuring operation the roller whieh takes the web from the chart drum is driven by the measuring roller and a spring in the chart drum from whieh the web is being taken is put under tension, whieh spring being thus put under tension is the power that re-tináis the web and the indicating mechanism to zero position when a measurement has been completed. Gearing connections transmit motion from the measuring roller to the pointer on the dial and the take-up roller. At the end of a measurement' when the clerk is ready to return the pointer and the web to zero position he cuts the line of power from the measuring roller to the chart or web take-up roller and to the pointer by pressing a finger plate whieh disconnects the gearing at a point intermediate the measuring roller and the registering means. The web and pointer run back to zero. The measuring roller and gearing between it and the point where the gearing was disconnected remain stationary. That is the larger part of the. whole mechanism. We have already said *272 that the machine would he useless unless there were means to run it hack to the initial point in order to make another measurement. In being run back under the pressure of a spring, motion is accelerated, and if speed is not cheeked in some way before zero position is reached there will be a shock that might tear the web from the roller from which it is being taken and disrupt other parts. The element of the claims here in controversy deals with means to avoid that hazard.

Appellant checks the return speed in its machine by three brake shoes which come in contact with the inner surface of a drum because of centrifugal force as the speed accelerates, whereby the speed is subdued during the return to initial position. Just as zero is reached on the return motion coaction of a control wheel, a pin, a pivoted lever and a spring causes a pawl to engage the shoulder formed on the control wheel which blocks further reverse motion of the disconnected gearing and stops the registering devices at zero.

In appellee’s machine, when returning the measuring and price-computing mechanism to zero there is no disconnecting of the gear. The lever which sets the brake at the end of a measumig operation is lifted, releasing the brake and the whole machine (except the presser roller), including the- measuring roller, all gearing, chart rollers, chart or web and hands on the dial and arc run back to the initial or starting points where they were when the measuring operation began, moving under the motor power of a large spring on the shaft of the measuring roller. Under the proof appellee has no means of controlling or regulating the return speed during any considerable part of it, as in Hosch. Its machine reaches a very high speed, accelerating until just before it gets back to the starting point, approximately one-quarter yard from the end, as shown by the dial, when, on account of its high speed a friction ring automatically bears against four outwardly-flung fingers pivoted at one end and rotating at high speed and the whole mechanism is safely brought to a quick, safe stop at zero. Only about two- per cent, of the distance to- be covered on the return motion remains when the ring is brought to hear against the fingers. When that occurs the Angers drop quickly and the machine is at rest at the initial point, because of this friction braking action.- Just before the ring-is automatically brought to bear against the out-flung fingers they and the measuring roller fun with unchecked acceleration and attain • a velocity of more than two thousand revolutions per minute. Each machine runs back in about four seconds. It is claimed by appellant that the rapidly revolving fingers, resisted by air pressure, act to govern and control the speed before the ring is in contact with them, hut we think the contention cannot be sustained on the proof. It was testified that tests demonstrated any effect from that source was negligible. It is true that the same purpose in this respect is accomplished by each machine. That is, appellee’s means, as does appellant’s prevents the machine racing to the end under spring pressure and this avoids disruption or breaking of parts. But results are not patentable. It is the devising of means which produce results that entitles one to a grant of the monopoly. National Hollow B. B. Co. v. Interchangeable B. B. Co. (C. C. A.) 106 F. 693, 708. As already seen, Hosch has automatic means for finally stopping the web or chart rollers, and the hands on the dial just as the end is reached, but before that an element in his combination governs and cheeks the speed. This element is not found in appellee’s machine, -and its means for stopping the return motion at zero- is unlike appellant’s applied at that point. Bringing the pawl to hear against the shoulder on the wheel blocks further return motion when zero is reached; while appellee, through' friction-breaking means. eases the speed of its machine to rest at zero. Another obvious difference between the two is that appellant runs back but a small and the lighter part-of his mechanism, whereas the whole of it is put under return motion in appellee’s.

Hosch was not a pioneer, he was not the first to devise means to return mechanism to its initial point under spring pressure- and means to check acceleration. The movement or restraint against movement of part' or all of meehanism by the action of a spring is undoubtedly very old. In Kullberg & Brunzell Pat. No. 573,335, issued December 15, 1896, for a cloth-measuring machine,, means were provided to regulate the reverse or return movement of the gear and parts connected therewith. In not an unrelated art McFarland & Wales obtained patent February 25, 1908, for momentum absorber for' adding machines. Means were provided to-return the actuating shaft to its initial position under tension of a spring, and “means-adapted to prevent a too-sudden return of said devices to- their original position.” The-specification noted the constant energy of the-spring, which furnished the return powex', to increase the return motion; and braking means were devised which wex*e actuated by *273 centrifugal force to control the return speed. There was nothing new in Hosch in providing means to cheek the return speed. Scott preceded him, and he says one of the objects of his invention was “to provide an automatic governor or brake (located at a point where maximum speed is obtained) to prevent excessive speed of the parts in returning to initial position. During the return to zero of all parts roller 10 is revolving at high speed — the governor arms 11 and 12, by centrifugal force, will bear against the bore 16 in casting 14 and thereby act as a brake to prevent excessive speed of all parts in returning to zero.” And so we are led to conclude, on the facts stated, that the principles announced in the cases cited supra, in the discussion of Claims 6 and 7 of Scott Reissue, apply also to these two claims, and that the charge of infringement is not sustained. •

Stoeke Pat. No. 1,437,207 — Claim 23.

Claim 23. “In a machine for measuring textile fabric, the combination of a pair of. rollers, indicating means operated by one of the rollers, means for normally obstructing the roller pass, means for zero-setting the indicating means and a common actuating element for providing for the operation of said zero-setting means and the simultaneous operation of said obstructing means to clear the roller pass.”

The underscored part of the claim is the basis of the charge of infringement. It is that element, or its equivalent, which appellant claims is found in appellee’s machine. As to this element, appellee’s machine is made under the Vanderveld patent 1,443,332. It provides means for obstructing the slot or throat of the machine at the end of a measuring operation, thus preventing the introduction of cloth for another measurement until the indicating or registering devices have been zero-set. A general but sufficient description of the means used in appellee’s machine and the mode of operation is this: On completion of a measurement the reset lever on the side of the easing is pressed down, which applies a brake that holds the registering devices until the clerk is ready to release the brake and let the mechanism run back to its initial zero position. This he does by raising the reset lever to its up position. But when this reset lever is pushed down it performs another function besides setting the brake — it automatically causes a throat lock bar to drop into the slot or throat and completely block entrance to the slot or throat through which goods must pass for engagement between the rollers. Before goods that has just been measured has been taken from the machine it must be notched, which is done by pressing down another lever, and the rollers must be separated, which is done by pressing down another lever. As heretofore said in reference to the claim of another patent, the levers are so arranged that' they may be pushed down together, but whichever way that may be done they each separately perform the separate functions that have been noted. When the brake has been set, the slot obstructed, the cloth notched and the rollers separated, as stated, the cloth is taken out of the machine, but the throat remains obstructed by the throat lock bar until the clerk raises the reset lever, which again performs the double function of automatically raising the throat lock bar out of the throat and releasing the brake, thus permitting the mechanism to run back to initial position.

Prom what has been said it is seen that when a measuring operation is fully finished and the cloth taken out, appellee’s rollers are apart and the slot or throat remains obstructed. Cloth cannot be interposed for another measurement until the mechanism is run back by releasing the brake, which automatically clears the slot or throat into which cloth is inserted for a measuring operation. Although the rollers are apart before the mechanism is returned to zero the lever which separates them and brings them together again has no mechanical connection with the reset lever. It therefore appears that if it were not for the throat lock bar in its obstructive position at the end of a completed measurement, the clerk could insert fabric, operate the lever that separates and brings the rollers together on the cloth between them and proceed with another measurement without returning the registering devices to initial position; hence the necessity of obstructing the throat so that may not be done. The means to that end, which has been described, is charged to infringe claim 23 of Stoeke.

We turn to the specification and drawings of Stoeke to ascertain the thing which constitutes the element of his claim now under consideration, and his mode of operation. He says his invention “includes means for normally obstructing the slots whereby material may not be introduced through said slots and between the rollers until the indicating parts have been zero-set, and whereby on the normal operation of the element which provides for the zero-setting operations, the obstruction will be moved clear of the slots. In the embodiment disclosed the obstruction (figures 2, 3 and 6) consists of a horizontal *274 bar arranged within the carrier (of the presser roller) parallel to one side thereof, said bar having an angular portion which normally projects across the slots in the path of movement of the material and thereby prevents the introduction or movement of the material between the rollers. The operation of the bar is controlled by the element which controls the operations of the zero-setting devices.”

When he separates the rollers, on completing a measurement, his obstructive bar with its up-turned end portion is automatically lifted above and clears the slots, the mechanism runs back to zero, the slots are unobstructed, and material may be introduced between the rollers to begin a new measurement. But when the presser roller is returned to its elevated position to press against the fabric interposed between it and the measuring roller, the arm which carries the bar will drop until the bar rests on top of the fabric and it will lie upon the fabric while it is being drawn between the rollers to effect the measuring operation. “When the measuring operation has been completed and the fabric has been withdrawn from between the rollers the bar will drop (lower) to its normal position wherein its angular portion projects across the slots as above explained.” He also says that he provides means correlated to the bar that prevents spinning of the measuring roller or “further rotation of said roller by momentum, due to the rapidity with which the operator may have pulled the fabric between the rollers.” He states the condition of his machine at the initial point with the rollers separated when the fabric to be measured, or the side edge portion of such fabric, is introduced between them. Thereupon, by manipulation, the presser roller is elevated to its normal position, pressing against the cloth between it and the measuring roller, and by this same manipulation the bar automatically is made to rest upon the fabric. The measuring operation is then initiated by pulling the fabric through the machine. When the required amount has been pulled through as shown by the registering devices a knife is operated to slit the edge of the material and thereby provide an indication 'of the line along which the measured length is to be cut or tom. The operator then takes note of the price computation. Thereupon he operates a lever to separate the rollers and reset the indicating devices to zero. The one operation that separates the rollers resets the mechanism and the machine is then ready for another measuring operation. When measuring a remnant the entire length of the material is drawn through the machine, and having passed from under the bar it drops further down and this lower position of the bar prevents further rotation of the measuring roller that might be due to a sudden jerk upon the material as it passes from between the rollers. When the bar is in the latter position the drawings show the upturned angular end portion of the bar projected upward across the slot or throat. He further says:

“According to the invention, the operations of the zero-setting devices are automatically effected in connection with the spacing or separation of the rollers whereby a single manual operation will serve both for the spacing of said rollers and for the operation of the several zero-setting devices.”

Stoeke has no braking means to hold his mechanism from running back when the rollers are separated. When they are together the contact is sufficient for that purpose. It is insisted that Stoeke was first in the art to conceive the usefulness of a throat lock and that his claim is entitled to the breadth of construction given to a pioneer. A patent is granted for a new and useful realty, a definite thing, and by statute that thing must be described in full, clear, concise and exact terms so as to enable any person skilled, in the art to make and use it. We have just noted Stoeke’s description of his means for obstructing the roller pass and its mode "of operation, found in his specification. In State Bank of Chicago v. Hillman’s, 180 F. 732, these questions were put by the Circuit Court of Appeals for the Seventh Circuit and answered in the negative:

“Can the patentee rightfully include in his claims something that does not emerge from the description? Can a patentee describe something to the world in his letters patent that means just that thing or its equivalents and nothing else,' and, having claimed that, claim in addition something not thus described and not its equivalents?”

The expert witness for appellant and the expert witness for appellee are in direct opposition as to whether appellee’s machine contains or embodies the subject-matter of claim 23 of Stoeke. The latter expressed the opinion that appellee’s machine had no mechanism acting in the same way and serving the same purpose as that of Stoeke, that the two devices were just the opposite of each other, that the presence of cloth in the machine and obstruction of the throat with other means were inconsistent with each other, that cloth between the rollers served that *275 purpose, also that appellant’s obstruction was present when the rollers were together during a measuring operation and until the zero-setting operation, and the contact of the rollers was sufficient to prevent the introduction of cloth for another measurement, and that the beneficial and useful purpose of ap-pellee’s obstruction was at a time when the rollers were separated before zero-setting and cloth might be inserted between them if the throat should then be clear; and he further testified that according to Stoeke’s specification and drawings his claimed obstruction was not an effective obstruction because fabric could be inserted in the slot under the bar with its upturned end. It is conceded Stocke’s bar is useful as an anti-spinning device to prevent the measuring roller from rotating after cloth may have been rapidly pulled through, but that it serves no useful purpose as an obstruction of the throat. Neither drawings nor specification gives information as to the position of the upturned end of the bar relative to the throat while the bar is riding on the cloth before it passes between the rollers. The upturned end during that time must be above the throat. But whether in it or above it during that time we think it apparent that additional cloth, especially if the measurement then being made is of a remnant, could be easily inserted into the slot, and it could also be inserted under the upturned end when the cloth passes from under the bar and it falls to its lower position, if we are to be guided by the appearance of its position in the drawings. Appellee’s expert so testified. The specification gives the impression that it was not Stoeke’s idea that the upturned end alone caused the obstruction, but the whole of the bar ineluding the upturned end; and that evidently would inform the clerk, if he should attempt to insert additional fabric for measurement, that the throat was not clear, even if it were not a complete and unavoidable obstruction, as is appellee’s throat lock bar, which drops into and extends entirely across the throat, and engages in a groove in the bottom of the slot, as soon as the cloth is out of the throat.

Clearly, Stocke had the idea of obstructing the throat to prevent errors, but ideas are not patentable. It is the means, or thing by which they may he accomplished that is within the law. “The real invention is to be found in the specification and drawings, and that the language of the claims is to be construed in the light of what is there shown and described.” Mossberg v. Nutter (C. C. A.) 135 F. 95, 99; American Grapho-phone Co. v. Leeds & Catlin Co. (C. C. A.) 170 F. 327, 331. If the thing so shown and described is not new and useful its discovery is not invention; and if it appear that a machine embodying it would fail to accomplish the intended result there can be no infringement. Ford v. Bancroft (C. C. A.) 98 F. 309. But we pass the application of these principles to the facts here solely because of the presumption which attaches to the issuance of Stoeke’s patent. He has means for obstructing the throat only when the rollers are in contact, of doubtful use at best. Appellee’s means of obstructing the throat performs its function while the rollers are separated and before the mechanism is returned to zero, this is at a time when such means can be used to accomplish its most useful function; and the functions of Stocke and Vanderveld are entirely different. The mechanism of each is different from that of the other and operates on a different principle. They do not accomplish the same result. They are not interchangeable. Stocke obstructs the slot partially when there is but little if any use for the obstruction. Vanderveld, by means operating in a different way, completely blocks the slot at a time when there is the greatest, if not the only, need of obstruction.

Hoseh & Turner Patent No. 1,371,028— Claim 12.

Claim 12. “In a fabric measuring machine, in combination with a measurement indicator, measuring mechanism for operating the same, including movable members for engaging the fabric to be measured, means for returning said indicator to zero position, means for releasing engagement between said members and the fabric at the end of a • measuring operation, and automatic means operating upon the release of said members from engagement with the fabric to provide a warning to the operator to return the indicator to zero position before again using the machine for a measuring operation.”

° The bill also charges infringement of claim 13, which reads this way:

Claim 13. “In a fabric measuring machine, in combination with a measurement indicator, measuring mechanism for operating the same, including movable members for engaging the fabric to be measured, means for returning said indicator to zero position, means for releasing engagement between said members and the fabric at the end of a measuring operation, and automatic means operating upon the release of said members from engagement with the fabric to render said *276 measuring mechanism inoperative until said indicator has been returned to zero position

The underscore of claim 12, supra, is the part of the combination relied on as infringed by appellee. It describes in different language the same thing as the underscore of claim 13, supra. The District Court held that neither was infringed. The assignment of error, however, is directed only to the ruling as to claim 12, and the argument presented here is so confined.

Appellant owns the Hoseh and Turner patent, but the machine it makes and sells does not contain this element. In fact, there is proof tending to show that no machine has been put upon the market embodying the so-called Hoseh and Turner warning deviee. Its mechanism is complicated and it is asserted with apparent confidence that it is wholly impracticable, that in support of this assertion it was shown at the trial appellant had used appellee’s device instead, and was held as an infringer by Judge Hickenlooper in a suit brought by appellee and defended by appellant. These assertions are not questioned by appellant’s counsel, and the effect of them is said to be that Hoseh and Turner conceived a useless and nonpatentable thing; thai there can be no infringement of a machine which fails to accomplish the intended result, Ford v. Bancroft (C. C. A.) 98 F. 309; that the mere existence of an intellectual notion that a certain thing can be done is not enough, that invention does not exist until the inventor’s ideas have been reduced to practical form, or it is clearly obvious they may be. American Graphophone Co. v. Leeds & Catlin Co. (C. C. A.) 170 F. 327. But appellant stands on the presumption that it was new and useful, arising from the issuance of the patent, and insists that the terms of the grant read on appellee’s machine.

The puxpose of Hoseh and Turner was to provide means to prevent the careless beginning of another measurement before returning the indicators to initial zero position, thus preventing inaccuracies in the indicators. They lock the machine by automatic means at the end of a measurement, so it is inoperative until returned to zero. This is done at the end.of a measurement by pressing a button to notch the cloth, which at the same time sets the brake, separates the presser roller from the measuring roller and automatically holds them apart. The presser roller cannot then be brought to bear against the cloth and press it against the measuring roller until the indicators are returned to zero. When .so returned the presser roller is unlocked and the rollérs put in operative relation to each other for the beginning of a new measurement. Hoseh and Turner say in their specification in reference to this element:

“The' safety loek arrangement just described is considered of great importance, as otherwise a careless operator might begin measuring goods for a new customer with the hands in the position they were left from measuring goods for a preceding customer, and thus unintentionally defraud the second customer of a length of goods equivalent to the amount previously measured on the machine. The invention not only provides means for preventing a second use of the machine until the indicating mechanism has been restored to zero position, but, in its broader aspect, it furnishes a warning to the operator to reset the hands before proceeding with a new measuring operation. That is to say, as the first step, after inserting the fabric between the rollers, is to push in the thumbplate, the impossibility of doing this at once apprises the operator that the hands must be reset before proceeding farther. In this regard, therefore, the automatic lock acts in the nature of a signal, or alarm, to warn the operator of the condition of the machine.”

In Hoseh and Turner, although the rollers have been separated and locked at the end of the measurement, the slot or throat of the machine remains open and unobstructed so that cloth may be inserted between the rollers, but being separated and locked they cannot be brought together on the cloth until the indicators are returned to zero, hence no measurement and registration can be made until that is done. The measuring roller, as the cloth is drawn between it and the presser roller, after they are brought together, actuates the registering devices, but while the rollers are separated these devices cannot be made to perform their functions.

Vanderveld’s purpose was the same as that of Hoseh and Turner, to provide means to prevent the eareless beginning of another measurement before returning the indicators to initial zero position, thus guarding against inaccuracies in the indicators. And so appellee makes its machine, according to Vanderveld, with a throat lock which is the alleged infringement. That is, by automatic means it places an obstruction in the slot or throat of the machine, already described, when a measurement has been completed so that cloth cannot be inserted therein between the rollers. That is done at the end of a measurement by pressing' down the lever- *277 which, sets the brake and this causes the throat lock to drop down into and across the throat or entrance slot of the machine, which prevents the insertion of goods therein. The movement of another lever downward notches the goods and still another separates the rollers. Much stress, however, is put by appellant on the fact that the three levers can be and usually are pressed down at the same time, and that appellee so instructed the users of its machine; but even when that is done appellee’s rollers are not locked apart and may be freely brought together. If the throat lock bar were not in the way, precluding the insertion of cloth therein, a measuring operation could be performed; in Hoseh and Turner, the rollers being separated and locked, it could not; although in Hosch and Turner the machine would be freely accessible for placing cloth between the separated rollers. According to Hosch and Turner the machine is inoperative until it is returned to zero, for the reason that the presser roller cannot be released from the lock to press the cloth against the measuring roller. In appellee’s machine, but for the bar across the throat, cloth could be placed between the rollers at the end of a measurement and other measurements made until the reset lever is operated (raised) to release, the brake, which raises the lock bar from the throat. The broad difference between the means of Hoseh and Turner and Vanderveld to accomplish the same result is obvious. Hosch and Turner, at the end of a measurement, separate the rollers and lock them apart, so the machine is inoperative. Appellee’s machine is at no time in-operatively locked, even when all of the levers are pressed down in unison, setting the brake, notching the cloth and separating the rollers; but pressing down the brake-setting lever, whether done separately or with the other two automatically obstructs the throat by dropping the lock bar into it. The problem was approached in a different way by Vanderveld and he solved it on a different mechanical principle from that disclosed by Hoseh and Turner. His mechanism is simple, it embodies a different mode of operation and it is operated in a different way. There is no substantial identity in the means used, but substantial difference between them; and though the purpose to be accomplished and the ultimate results obtained may be the same, this does not constitute infringement. Walker on Patents (4th Ed.) § 340 says:

“The respective results of a machine or manufacture covered by the claim of a patent, and of a machine or manufacture alleged to infringe that claim, do not furnish a criterion by which to decide the question of infringement. Those results may be identical, while the things which produce them are substantially different. Any person may accomplish the result performed by a patented thing without infringing the patent, if he uses means substantially different from those of the patent. To hold the contrary of this rule would be to retard, and not to promote the progress of the useful arts.”

In Anakin Lock Works v. Dillon Lock Works (C. C. A.) 292 F. 45, we quoted with approval this from American Can Co. v. Hickmott, etc., Co. (C. C.) 137 F. 88:

“In a patented combination, a device in one mechanism, to be the equivalent of a device in another, must perform the same function (Rowell v. Lindsay, 113 U. S. 97, 103, 104, 5 S. Ct. 507, 28 L. Ed. 9.06), and perform that function in substantially the same manner, as the thing of which it is alleged to he an equivalent. Walker on Patents, § 354.”

We think it clear appellee’s combination does not include the element relied on in Hosch and Turner’s combination. “The absence from a device that is alleged to infringe a patented combination of a single element of that combination is fatal to the claim of infringement.” National Hollow B. B. Co. v. Interchangeable B. B. Co. (C. C. A.) 106 F. 693, 718.

Prom what has been said it is our opinion that there is no error in the decree below, and it should be affirmed in all respects.

Affirmed.