After stating the foregoing facts, the opinion of the court was announced by
LURTON, Circuit Judge.If, as contended by the learned counsel for appellant. Plumb’s patent is to be construed as only securing an efficient frictional contact between the driving wheels and the brush shaft, then the field for improvement in frictional devices in sweepers was an exceedingly narrow one. This is abundantly shown by numerous prior patents in evidence. That end was the purpose of so many of these earlier devices, and that result had been secured in so many ways, closely approximating the means adopted by Plumb, that his claim, if sustainable at all, is only to be upheld by limiting him to the specific friction device he has described and claimed. If his patent be thus limited, it is doubtful whether any of the sweepers made by the Goshen Company infringe his second claim. On the other hand, it has been most ably argued that the patent was rightly interpreted by Judge Brown, and that it is entitled to so broad a construction “that any carpet sweeper whose case is carried on its wheels, whether two or four, by springs of any form, or by any equivalent elastic, yielding connection, infringes his second claim.” When we come to interpret Plumb’s patent, and mark out its limits, we are at once confronted with the well-established fact that a carpet sweeper constructed according to his plans, as described in his specifications, possesses, in practical operation, advantages in usefulness which distinguish it in a most decided way from each and all of the alleged anticipatory structures. If we analyze these advantages, we find that in practical use the most obvious improvement in the utility of such machines lies, not in the mere matter of securing a more certain and reliable friction between his driving wheels and the brush roller, but in the adaptability of his mechanism for securing a varying pressure of the brush on the carpet by mere downward *71pressure upon the sweeper case, whereby the brush may be made to •bear heavily or lightly on the surface to be swept, as the conditions of sweeping seem to require*. This adjustability of the brush by downward pressure resembles the action of a broom under hard or light pressure, and has been aptly described by the experts as the “broom-action” advantage of Plumb’s mechanism. This, the chief advantage of his improvement in carpet sweepers, is the very feature which it is said he has not pointed out in his specifications in such definite way as to he regarded within the perception of the inventor, or the limits of his patent. That this advantage does not exist: to so great a degree in a two-wheeled sweeper, such as shown by Plumb’s model sweeper filed with his application, may be conceded. Neither his drawings, specifications, nor claim confine him to a two-wheeled sweeper, though the model filed by him does show” that one end of his sweeper rested upon an unyielding metal shoe, so that no amount of pressure would lower that end of the case, or that end of (he brush roller carried in the case. Sweepers thus constructed were in fact made and sold by Plumb for several years after his patent. But the decided weight of the evidence is that Plumb’s two-wheeled sweepers were exhibited and advertised as sweepers callable of doing hard or light sweeping by downward pressure upon the sweeper case, and that this was called “broom action.” The evidence also clearly show's that this feature of his improved sweeper was then regarded with great favor by both the general public and by manufacturers of sweepers, who soon began efforts to obtain the same function in their several devices. The recognition of the existence and utility of this very feature may in truth be said to have, at an early day, revolutionized (.he carpet-sweeper industry. Though this varying pressure of the brush on the floor was not so beneficial as when both ends of the brush could he lowered, still it is evident, even in thai: form of sweeper, that downward pressure on the sweeper case would lower the wheel end of the case, aud with it the wheel end of the brush roller, so that hard or light sweeping might be done* by a part of the brush. The difference between the advantage of sweeping heavily or lightly with the whole of the brush, and wilh a part only, is, at last, bni: one of degree. It was a new and useful improvement to be able to vary the pressure on but a part of the brush. The interposition of wheels between the brush roller and end of the case necessarily leaves an unswept margin next a baseboard. Any limitation upon the broom-action function of Plumb’s combination in consequence of supporting his ease on wheels at one end only was somewhat offset by the fact that the shoe end of the case was better adapted to sweep close to the baseboard. At the date of Plumb’s invention this was regarded as giving two-wheeled sweepers an advantage over those carried on four wheels, and they were then much the most popular in the market. Plumb himself was the first to extend his principle to four-wheeled sweepers. If the mechanism contrived by Plumb embodied this broom adjon adaptability to a degree constituting it a new and useful improvement upon the old structures, and he has sufficiently de*72scribed or pointed out his improvement to meet the requirement of the patent law, then he is aot to be deprived of the benefit of this feature because, by a duplication of his wheels, this advantage can be obtained to a greater degree. Appellant cannot escape liability as an infringer by a mere duplication of Plumb’s combination. That is a mere carrying forward of his idea, by duplicating the parts, doing the same thing in identically the same way, with better results. Roberts v. Ryer, 91 U. S. 150; Belding Manuf’g Co. v. Challenge Corn-Planter Co., 152 U. S. 107, 14 Sup. Ct. 492.
Passing from a consideration of these practical advantages of Plumb’s sweeper, as observed in its operation, let us contrast its mechanism with that found in the earlier devices, and see how far that indicates a purpose, and what that design was. Here we again find very markéd mechanical differences. This difference lies, not in the case, or brush, or wheels, but in the connection between the case and the wheels. Theretofore the connection between the case and wheels had been rigid. He made that connection elastic. He did this by spring-supporting his case on its wheels, so that both the wheels and the case hove a vertical motion relatively to each other. The combination of old elements, by which he makes the connection between the case and wheels elastic, constitutes the claim here involved. That mechanism provides for the rotation of the brush by frictional contact between the brush roller and the driving wheels, which are held in contact with the brush roller by the action of the springs, I, I. The nécessary mechanical effect of these springs, when arranged as described, is to keep the wheels pressed against the pulley wheel of the brush shaft, and, by the friction thus produced, rotate the brush when motion is given the wheels by driving it over the surface to be swept. Plumb calls attention to the fact that his wheels are allowed “free motion,” attributable in part to. the fastening of these springs to the case in one place only. This reference, looking to the whole language of his specification, and construed as one skilled in the mechanics pertaining to carpet sweepers would construe it, does not refer to the mere rotation of the '.brush roller, but to the “free motion” of the wheels vertically in ¡relation to the case. The only object conceivable in interposing a spring between the case and wheels, each being connected to the spring, is to permit one to move relatively to the other under pressure, and to restore the moved object to its normal position when .the pressure is removed. This vertical motion gives what is called ¡“broom action,” because the brush roller is joursaled in the case, and must move with the case. While the vertical motion of the wheels relatively to the case makes a mechanism adapted to give a varying pressure on the brush, as the necessary law of the structure, the lateral motion of the wheels permits the same spring which supports the case to give frictional contact between the wiieels and the brush roller. That this friction is slightly increased by downward pressure upon the sweeper case is due to the fact that the center óf the brush roller is above the center of the driving wheels, so. that one effect of this “free motion” of the wheels vertically is *73that the descending brush roller crowds the driving wheels further apart, and increases the tension of the springs, I, I. That Plumb makes two independent statements — one in relation to the free motion of his wheels under downward pressure on the sweeper case, and the other concerning an increase of tension in springs, I, I — is made more evident by a comparison of the grammatical construction of his original application with that now forming part of the patent. In his original specification he stated, concerning his spring, that:
“The spring-. I, may be made from a single piece of metal. It is attached to the side of the case, as shown, passes over on a curve, and its lower end supports the journal for the driving wheel. This gives the wheel free play, when pressed both horizontal and perpendicular. The pulley, 10, being above the center of wheels, W, W, the friction may be increased by pressing the sweeper on the carpet. The spring, I, may be varied in form, and may be constructed of any suitable material.”
In the amended specification, now the patent, he stated the second object of his invention to be “to journal the driving wheels to bearings having both vertical and lateral motion.” He then said:
“The spring, I, may be made of one piece of metal, or of more; and the bearing of the wheel may be a flat piece of metal, or may be in the form of a yoke. The spring, I, may also be varied in length. The fastening of the spring to the case of the sweeper in one place only allows free motion of the wheel, and the center of the brush roller being above the centers of the driving wheels enables the operator to Increase the friction between the wheels and the brush roller by merely pressing down on the sweeper. I prefer to attach this spring, I, to the side of the case, and have it pass over on a curve in the form shown in the drawings, but it may be varied in form and construction to suit the operator.”
We quite agree with counsel for appellee in their contention that the quotation above contains two distinct statements, which are substantially the same as the two statements of the original specification above quoted. The first of these statements is substantially that wheels thus journaled in bearings attached to one end of a spring, the other end of which is attached to the side of the sweeper case, would permit the free motion of the wheels when pressure is brought to bear upon the case. The second statement is that, if the center of the brush-roller wheel he above the center of the driving wheels, downward pressure upon the case will crowd the driving wheels further, apart by the descent ,of the brush roller, “and increase the friction between the wheels and the brush roller ” We can but regard the .specification as clearly explaining that the principle of his invention lies in the fact that Ms wheels are so jour-naled as to be capable of both vertical and lateral motion. This principle is an essential to every carpet sweeper having the advantage of “broom action,” for broom action is nothing more nor less than the vertical motion of the wheels relative to a case carrying a brush attached to the case. That this claim does not include, as an element, a brush roller, is immaterial. The claim is not for a complete machine, but is limited to a specific improvement of a part of the mechanism of carpet sweepers, “having the brush roller operated by friction driving wheels.” The specification fully describes such a machine, and points out all the elements necessary to con*74struct it, and describes bow tbe ingredients of bis combination are to be arranged to co-operate with tbe other well-known parts to make a complete working carpet sweeper. A patent is addressed to those familiar with tbe art. A new and useful improvement upon an old machine is just as much tbe subject of patent rights as a new machine. All that the inventor is required to do is to point out distinctly tbe part be claims as new, so as to advise the public as to tbe extent of bis invention, and what is thereby withdrawn from tbe public. Parks v. Booth, 102 U. S. 102; Seymour v. Osborne, 11 Wall. 516-541; Roberts v. Nail Co., 53 Fed. 916; 2 Rob. Pat. §§ 530-534. A distinction exists, though not always observed, between a statement of the mechanical functions of a machine, and the beneficial advantages, uses, or results due to that function, and contributing to the success of the invention. The function of a mechanism is its mode of action or operation proper to the structure, and does not ordinarily refer to the results, effects, or advantages of that appointed power of acting in a specific way. The patent statute recognizes this distinction, for it requires that a patentee shall state the “principle” of his machine, — meaning thereby its mode of operation,- — and nowhere mentions the beneficial effects, results, uses, or advantages. The principal function of Plumb’s spring-supported sweeper case, when arranged to co-operate with the other ingredients of a carpet sweeper, is that, under downward pressure upon the sweeper case, the brush roller is lowered, and thus a varying pressure can be applied to the brush. This capacity of his contrivance he perceived and described when he referred to one of the effects of downward pressure as being an increase of driving fricj tion. The lowered end of the brush is necessarily brought to bear more heavily on the boor. That this would be an advantage in the practical operation of a carpet sweeper would seem obvious. That it is of very great practical utility in the actual operation of his sweeper is established by the evidence, as well as by inspection of the exhibits. That Plumb did not see this advantage resulting from the function of his combination is hardly -to be credited. This capacity of adjusting the brush for heavy or light sweeping is so obvious, so directly the appointed action of his mechanism, that whether he has described it specifically is not material. Neither the function, nor its advantage in the success of his sweeper, was a mere obscure property lurking in some accidental corner of the device. It lay upon the very surface, and was continuously in evidence when the sweeper was in operation. This conibination resulted in an improved sweeper, capable of a new and beneficial use. He was the first to provide an elastic connection between the wheels and case. This new and most beneficial function is the immediate result of the idea of supporting a sweeper case by a spring connection on its wheels. He is entitled to this beneficial function of his invention, whether he then knew all its beneficial uses or not. McCormick Harvesting Mach. Co. v. C. Aultman & Co., 16 C. C. A. 259, 69 Fed. 371; Roberts v. Ryer, 91 U. S. 150; Brown v. District of Columbia, 130 U. S. 87-103, 9 Sup. Ct. 437: Eames v. Andrews, *75122 U. S. 40, 7 Sup. Ct. 1073; Miller v. Manufacturing Co., 151 U. S. 201, 34 Sup. Ct. 310; Stow v. Chicago, 104 U. S. 550; Tucker v. Spalding, 13 Wall. 453; Appleton Manuf'g Co. v. Star Manuf'g Co., 9 C. C. A. 42, 60 Fed. 411; Western Electric Co. v. Sperry Electric Co., 7 C. C. A. 174, 58 Fed. 186; Galt v. Parlin, 9 C. C. A. 49, 60 Fed. 422; Pfeifer v. Dixon-Woods Co., 14 U. S. App. 294, 295, 5 C. C. A. 348, 55 Fed. 390; Thompson Meter Co. v. National Meter Co., 32 C. C. A. 673, 65 Fed. 427. Patents cover the means employed to effect results. Neither an idea nor a function, nor any other abstraction, is patentable in a machine patent. Burr v. Duryee, 1 Wall. 570; Fuller v. Yentzer, 94 U. S. 288. In Eames y. Andrews, cited above, the patent was for an improved method of constructing artesian wills, and was for the process of drawing water from the earth by means of a driven well. Neither the scientific theory nor principle involved, and supposed to constitute the invention, were set forth either in the original or reissued patents. The comment of Judge Blatehford, who had tried the case upon the circuit, in relation to this defect in description, met with the approval of the supreme court. That great patent judge said:
“It. may Iks that the inventor did not know wliat the scientific principio was, or (hat, knowing it, he omitted, from accident or design, to set it forth. That does not vitiate the patent. He sets forth the process or mode of operation which ends in the result, and tlu; means for working out the process or mode of operation. The principle referred to is on>y the why and the wherefore. That is not required to tie set forth. Under section 2(5 of the act of July 8, 1870 (1(5 Ktat. 201), under which this reissue was granted, the specification contains a description of the invention, and of the ‘manner and process of making, constructing, compounding', and using it,’ in such terms as to enable any person skilled in (he art to which it appertains to make, construct, compound, and use it; and, even regarding the case as one of a machine, the specification explains (he principles of the machine, within the meaning of that section, although the scientific or physical principle on which Hie process acts when the pump is used with the air-tight tube is not explained. An inventor may be ignorant of the scientific principle, or he may think he knows it, and yet bo uncertain, or he may be confident as to what it is, and others may think differently. All this is immaterial, if, by the specification, the thing to be done is no set forth that it can be reproduced.”
Roberts v. Ryer, died above, affords an interesting illustration of the scope of a patent which actually embodies a capability of use not perceived or described by the patentee. The patent involved was for an improvement in refrigerators, “whereby live whole of the contained air is kept in rotation, purification, desiccation, and refrigeration, with economy of ice.” The apartments in the device pointed out for accomplishing this end were constructed so as to obtain the full benefit of the descending current of cold air in refrigerating the articles placed therein for that purpose. ” The defense wtts that the invention had been anticipated by a patent to one Lyman. Upon an examination it was found that both inventions adopted substantially the same means for cooling the air, and causing its circulation through the refrigerator. It appeared, however, that Lyman had so arranged his device as to use only the ascending air for the purposes of refrigeration, upon the wrongful supposi-*76lion that the greatest benefit was to be derived from that current. The complainant had seized upon this mistake, and, while adopting Lyman’s method for confining and rotating the air, had arranged his refrigerating chamber so as to get the greatest benefit from the descending current. As to this the court said:
“It being then certain that Lyman contrived a macnine which would produce the desired circulation, and could he used for refrigeration in the ascending current, it remains only to consider whether, if one desired to make use of the descending current for the same purpose, he could claim such a use as a new invention. It is no new invention to use an old machine for a new purpose. The inventor of a machine is entitled to the benefit of all the uses to which it can be put, no matter whether he had conceived the idea of the use or not. * * * With both inventors the circulation, by means of an ascending and descending current, was the principal object to be attained. One considered the greatest benefit for the purpose of refrigeration was to be derived from the use of the descending current, while the other had his attention directed more particularly to the ascending current. They each had both, and could utilize both. It is no invention, therefore, to make use of one rather than the other.”
In McCormick Harvesting Mach. Co. v. C. Aultman & Co., heretofore cited, this precise question as to whether it is always essential that a patentee shall describe the beneficial functions to be accomplished by the parts of his mechanism came before this court. It was a case of great importance, and received most careful attention. It was there said, concerning certain parts of G-orham’s patent, that he had neither perceived nor described the beneficial effect resulting from certain segmental teeth in his improved grain binder, and that his patent should not be construed as including the invention subsequently asserted as within the scope of his patent, and dependent upon the functions of these teeth whose use was unde-scribed. Judge Taft, for the court, said, as to the effect on the patent of this defect in stating the results to be attained by these segmental teeth, that:
“It is not material tbat Gorham did not describe in full all the beneficial functions to be performed by the parts of his machine, if those functions are evident in the practical operation thereof, and are seen to contribute to the success of his device. Eames v. Andrews, 122 U. S. 40, 7 Sup. Ct. 1073. It is difficult to believe that a man of Gorham’s inventive genius did not perceive the useful functions which the parts of his machine áo well performed, even though he did not specifically mention them all.”
In Pfeifer v. Dixon-Woods Co., cited above, the circuit court of appeals had a like question to deal with. Judge Shipman, in delivering the opinion of the court, said:
“The specification closely, and altogether too closely, adheres to mere mechanical features, and creates doubt as to whether Tondeur thoroughly understood his invention. It indicates that the patentee did not understand the philosophical principles which caused bis mechanism to produce an improved annealing. If he had known, they would have been alluded to in Ithe patent; but an examination of the specification and its drawings leaves little doubt that the patentee meant, and that the specification means, to describe bars in such relation to each other that the glass is carried forward «constantly in the same horizontal plane.”
In tbe conclusion be added:
“He did not know, or he did not tell, why the new method would produce better results. He simply told how to construct a machine which carried *77the glass through the leer on a level, and saved much breakage; but he ought not to lose the statutory benefits which would certainly belong to him if he had seen and described the philosophy of his machine accurately.”
That the three sweepers made by appellant infringe the Plumb patent is a matter about which we have no doubt. Mechanically, they do not substantially differ from those held by Judge Brown to infringe in the Wetherbee Case. This was also the conclusion of Judge Severens, who, yielding’to the interpretation placed by Judge Brown on the Plumb patent, found no reason for distinguishing between the devices which were made and sold by appellant, and those enjoined in the Wetherbee Case.
One of the infringing sweepers made by appellant is called the "Perfect.” It is a four-wheeled sweeper. That is, in itself, no defense, if it is but a substantial duplication of Plumb’s invention. At each side a straight spring rod passes through large holes in the ends of the case, this rod serving as the axle for a wheel at each end. Each rod is attached at its center to the top of the case by a bracket. Thus it is but a case spring-supported on its wheels, the spring differing only in form from that preferred by Plumb. There is little difference, mechanically, between appellant’s “Star” and “Banner” sweepers and the “Perfect,” already referred to. Each is spring-supported on its wheels. A straight metal rod, at each side of the case passes through large holes in the ends of the case, and carries a wheel at each end. These rods are controlled and pressed by a spring at each end of the case, so that the weight of the case is carried by these springs. None of these sweepers show any originality. They accomplish the same results attained by Plumb, by almost identically the same means. A mere change in the form of spring, or its place of attachment, by which vertical motion of the wheels relative to the case is obtained, is not invention. Our conclusion is that the decree of the circuit court must be affirmed.