The bill of complaint charges infringement of claims 1, 2, 3, 5 and 8 of Letters Patent No. 793,249, granted June 27, 1905, to the appellee as assignee of Libanus M. Todd, for printing apparatus. The District Court found certain of the claims valid and infringed and others not sustained or not infringed. 226 Fed. 791. This is an appeal from that part of the decree adjudging claims 1 and 3 valid and infringed and awarding an injunction and an accounting as to them.
The subject matter of the patent, while described generally as “printing apparatus,” is that class of printing apparatus designed to protect checks and drafts from alteration, by printing upon them, in embossed and serrated letters and numerals, well known words, for example, “NOT OVER FOUR HUNDRED $400$.”
The mechanism of the check protecting machine of the patent is enclosed in a casing, and consists of a metal cylinder, drum or type-wheel, upon whose periphery are rows of projecting type characters extending parallel with its axis; a co-operating platen arranged to contact with the type-wheel; an ink roller serving the double purpose of inking the type characters and holding or centering them in position for co-action with the platen; and appropriate means to bring these parts into action and thereby to produce upon paper, embossed, serrated and inked letters and numerals. Upon the printing or impression surfaces of the type characters is a plurality of parallel ridges and grooves which extend circumferentially of the type-wheel. These have a sharp cutting edge, and form a part of the printing surfaces for-receiving and transferring ink.
The platen is made of hard metal, upon whose impression surface are ridges and grooves corresponding in number, direction and size with those upon the type-wheel. Upon contact, the ridges of the platen enter the grooves upon the printing characters, and the ridges of the printing characters enter the corresponding grooves upon the platen. By this co-action the fiber of the paper is abraded, disrupted and cut on both its upper and under sides. To effect this co-action in the commercial machine manufactured under the patent, the ridges and grooves on the type characters are extended around the type-wheel in a direction parallel with the plane of its rotation and at right angles to its axis, so that when the cylinder comes in contact with die platen at any point in its revolution, the ridges upon the face of the type co-act with *716the complementary grooves upon the platen. The essential principle of this universal co-action is the maintenance of perfect alignment of the complementary ridges and grooves just before and at the instant of contact, and this principle is reduced to practice in the device of the patent by snugly fitting the revoluble type-wheel and the platen against the side walls of th'e casing in a manner to obtain perfect alignment and prevent ¿lateral motion at this critical point. This was thé capital conception of the patentee. There is a clear distinction mechanically between ridges and grooves which encircle a cylinder precisely parallel with the plane of its revolution and ridges and grooves which in any degree deviate from that exact course. In the first, perfect alignment between the ridges and "grooves of the cylinder and those of the platen, when once attained by mechanical means, is never disturbed by the mere revolution of the cylinder; while the revolution of a cylinder, with ridges and grooves deviating ever so slightly from the true circumferential direction, throws its ridges and grooves in and out of alignment with those of the platen, and co-action is uncertain until the movement is arrested by some mechanical centering means and the ridges placed in alignment for contact. Therefore, whether the course of ridges and grooves be truly circumferential or spiral, and without regard to the maintenance or disturbance of alignment during the initial movements of the cylinder or platen, practical and actual co-action and registration in a machine of any type can .be attained only by mechanical means of some sort, and by such mechanical means as will •procure perfect alignment just prior to and at the instant of contact. Hence we are not so much concerned with the theory of complementary co-action of circumferential and spiral serrations, as we are with the means disclosed by the patent in suit and the means employed by the defendant to bring the serrations upon the face of the type, whatever may be their direction, into alignment and contact with their complementary serrations upon the face of the platen. The patent discloses means to procure such alignment and produce such contact. If these means constitute patentable invention and if the defendant employs them, a deviation in the direction of serrations will not save him from the charge of infringement. Let us inquire what is the construction of the defendant’s machine, the result attained by it, and the means employed to produce that result.
The parts of the defendant’s machine which figure in this controversy are the same as those of the patent in suit, namely, a type-wheel or drum carrying upon its periphery rows of projecting type characters, a platen, and an ink roller. In the defendant’s device, the ink roller serves the single purpose of applying ink to tire type characters. The cylinder is put and held in position for printing by other centering means. The printing faces of the type characters are serrated with a plurality of parallel ridges encircling the cylinder. These parallel ridges are not circumferential of the cylinder in the sense of being precisely parallel with its plane of revolution. There is a slight deviation from the true circumferential course, actual, though invisible. The ridges hre like a continuous screw thread, beginning at one end of the cylinder and spirally encircling it until it runs off at the other *717end. The cylinder is not a true cylinder. It is a helix, and its shape is removed from that of a true cylinder in the same minute degree that the direction of the spiral ridges deviates from that of truly circumferential ridges; that is, the pitch of the type* wheel or cylinder conforms to the direction of the ridges. In the defendant’s machine, as in that of the patent, the ridges are .025 of an inch apart. The results desired are two. The first is the printing of words from some desired row of projecting type characters. ■ This is accomplished in the two machines by centering means of different kinds, by which type characters on the cylinder are brought in true printing position with the impression surface of the platen. The second thing desired is a cut or abraded surface of the printed letters. To produce this, the two machines imperatively require precise co-action of the complementary ridges and grooves on the printing surfaces of the letters projecting from the cylinder, and the ridges and grooves upon the platen. This is absolutely necessary not only to avoid collision of the ridges and consequent destruction of their apexes, but also to obtain the registration which produces the desired abrasion and cutting of the paper. To accomplish such exact engagement, it is manifest that at least just before and at the time of impact, the theory of precise alignment must be embodied in mechanism, which must first establish alignment and thereafter prevent its disturbance by any lateral motion.
Both the patentee and defendant employ such mechanism. In the machine of the patent, the alignment of the ridges and grooves is fixedly maintained at all times by the fixed positions of the cylinder and platen in the casing, but in the machine of the defendant, the helical drum is suspended well up in the casing, the ridges of which, when the cylinder is at rest, may be entirely out of alignment with the ridges of the platen. When registration is desired, the platen in the machine of the patent is raised vertically against the clyinder by one straight movement, while in that of the defendant, the operation is reversed, and the cylinder is brought down against the platen by a variety of movements. In the machine of the patent there is no lateral movement of either the drum or platen at any time. The latter is raised in its fixed position to register with the former in its fixed position, and1 so long as their positions remain fixed, the alignment remains perfect and the co-action complete. But a vertical movement of a helical drum containing spiral ridges against a platen containing complementary diagonal grooves, would, without means to procure final alignment, be wholly ineffective to assure registration. To obtain registration by co-action of spiral and diagonal ridges in the defendant’s machine, which are out of alignment when not in action, a cam is employed to revolve and laterally move the spiral dpum on its axis in the upper portion of the chamber of the casing until in its part revolution and succeeding descent its spiral ridges are brought into perfect alignment with the complementary diagonal ridges on the platen. This alignment is procured by bringing one end of the helical drum into contact with a lug on one wall of the casing and by compressing the other end by a spring attached to the other wall of the casing. Thus the defendant’s irregularly shaped drum snugly slips into the very *718casing enclosure which the patent points out as the means for securing perfect alignment just before the instant of contact, from which place the drum, with its spiral ridges held by the wall lug and spring in precise alignment with .their complementary grooves on the platen, descends a small fraction of an inch to contact and registration. After it reaches the arresting points between the walls and starts upon its final motion of descent, no further revolutions of the cylinder occur and no lateral movements of the cylinder are possible. In other words, the. defendant’s machine, unlike that of the patent, does not at all times maintain alignment of its complementary ridges and grooves, but obtains such alignment just before the time of contact, and by the same wall embracing means maintains and continues it until contact is complete. This is admitted by the defendant’s expert, subject to a qualification, which, however, does not disturb the certainty that the defendant’s helical drum is caught in its revolving and lateral movement by the wall lug and spring, and thereafter is directed in a true course to the point of contact, though the distance be minute. It therefore appears that while the initial movements of the moving parts of the two machines are different, both machines disclose the correct theory of registration to be a fixed and true alignment, and both procure such alignment at the only time it is necessary by employing the fixed walls of the casing to prevent lateral motion. Fixed alignment was disclosed by the patentee as a requisite to perfect co-action and accurate registration. As means to attain it, the patentee pointed out the two walls of the casing in which to snugly place the drum and platen. The defendant adopted these two walls as supports for a lug and spring, and used the lug and spring so attached to the walls as his means to procure alignment and prevent lateral movement just before and at* the instant of contact.
Unless the claims of the patent in issue are to be restricted to a device containing circumferential ridges literally parallel with the plane of their rotation, we are of opinion that the means employed in the defendant’s machine to secure registration must be considered full mechanical equivalents of those found in the machine of the patent.
Claim 1 of the patent in suit is as follows:
“In a printing apparatus, the combination with, a revoluble type-wheel provided with printing characters upon its periphery, the impression surfaces of the characters being provided with a plurality of circumferentially arranged, grooves, of a platen mounted to co-operate with the characters of the type-wheel having projections arranged to register with the grooves formed thereon.”
Claim 3 is distinguished from claim- 1, in that it includes a “type support” that is not limited in construction to. “a revoluble type-wheel,” and provides for “a plurality of parallel ridges” without defining their direction. The defendant maintains that the expression “circumferentially arranged grooves” found in the first claim, means grooves arranged circumferentially of the cylinder in position precisely parallel with its plane of rotation, and that the expression “a plurality of parallel ridges” found in the third claim, means the same thing, and therefore as the ridges of the machine of the defendant are spiral, *719neither claim is infringed. This contention is based upon an expression found in the specification, by which the patentee discloses the advantage of a mechanism containing serrations upon a cylinder arranged in planes parallel with the plane of its rotation. This is stated by the patentee as follows:
“Moreover, a material advantage is secured by forming these ridges of the printing characters .so as to exlend circumferentially of the type-wheel in planes parallel to the plwn-e of rotation thereof, for the reason that the printing faces of the type characters will accurately register with the impression surface of the platen at every degree of rotation of the wheel, thereby avoiding the necessity of precisely centering the type characters opposite the platen in order to secure a proper register of the ridges and depressions of the respective parts as would be necessary should these ridges extend spirally or in any direction other than circumferentially of the wheel.”
[1] The defendant seeks to limit claims 1 and 3 by this statement of the patentee in the specification. But we are inclined not to apply this disclosure to claims 1 and 3, but rather to consider it with reference to claim 2, which by identical terms with those found in the specification describes and limits the circumferential arrangement of serrations to a direction "parallel to the plane of rotation.” This unquestionably is the patentee’s preferred arrangement as disclosed by the language of the specification, and while the patent is limited to that arrangement by claim 2, it is not so limited by claims 1 and 3, unless we hold that the three claims mean the same thing. This of course we should avoid, if the language of the claims makes it possible. Zittlosen Mfg. Co. v. Boss, 219 Fed. 887, 135 C. C. A. 551; O’Rourke v. McMullen, 160 Fed. 933, 88 C. C. A. 115; Thomson-Houston v. Nassau (C. C.) 110 Fed. 647; Ramson v. Hillman, 123 Fed. 416, 59 C. C. A. 510; Ryder v. Schlichter, 126 Fed. 487, 491, 61 C. C. A. 469.
Claim 2 contemplates by exact terms a circumferential arrangement of serrations precisely “parallel to the plane of rotation thereof.” Claim 1 is distinguished from claim 2 by providing in general terms for circumferentially arranged serrations without defining or limiting their direction with regard to the plane of rotation. Claim 3 is distinguished from claim 1 in that it provides a “type support” without defining or limiting its shape to that of a revoluble type-wheel and in that it provides a plurality of parallel ridges without requiring them to be either truly or substantially circumferential in their arrangement. Of course we are deciding nothing with respect to claim 2, because it is not in issue, but we are employing that claim, the terms and meaning of which correspond precisely with the defendant’s interpretation of claims 1 and 3, in order to distinguish the claims and discuss the defendant’s contention. Construing claim 2 to mean what its unambiguous language conveys, we feel free to give to claims 1 and 3 an interpretation different from that which must be given to claim 2, and an interpretation, which, in our opinion, is justified by the language of the claims and the disclosures of the specification. We construe claims 1 and 3, with respect to the direction of ridges and grooves, to be broad enough to permit their arrangement in a direction . other than *720■truly circumferential in the sense of being parallel with the plane of their revolution, and in so doing we find that if the claims are not invalid because of anticipation by the prior art,'they are clearly infringed by the full mechanical equivalents employed in the device of the defendant.
[2] The validity of claims 1 and 3 of the patent is vigorously attacked on the ground that they are anticipated by'many patents appearing in the prior art. AVhile we, have carefully considered all of them, we will be able to discuss only a few in this opinion.
The devices of Ocumpaugh patents, No. 734,932 and No. 749,577, are similar. Each shows a type-wheel with rows of projecting type characters mounted upon its periphery, an ink roller, and a platen. The platen is soft and the type characters are smooth.
The Carsley patent, No. 152,329, is for a check stamp designed to emboss ink characters upon a check. The Carsley device is nothing more than a flat die and a counter-die, and the patent discloses nothing new in die and counter-die co-action.
The Pardi patent, No. 675,404, is a seal press with opposing die numbers. These are without serrated printing characters.
The Rogers and Hall patent, No. 592,533, discloses a type disk upon whose surface are mounted numerals. Secured to and revolving with the type disk is a thin spring metal disk having flexible arms corresponding in number and position with each numeral upon the wheel. These arms perform the function of inverted or superimposed platens. The platen arms are perforated with lateral slots. Over the disk is mounted a hammer or thrust, whose impression surface is covered with a type ribbon. In operation, the type disk is revolved so as to bring a desired numeral .under the hammer. The paper is introduced between the numeral and the perforated platen arm. A blow upon tire hammer drives the ink ribbon against the perforated platen, the platen against the paper, and the paper against the ridges of the numeral, with the result that tire paper is embossed in the form of the numeral, abraded on its under side, and colored on its upper side by contact with the ribbon through tire slots of the platen. In this way the paper is embossed, cut and printed. In this device the type characters are not inked nor is there co-action of serrated characters and serrated platens. The result of the action is an embossed check serrated on its under side, and a plurality of parallel lines printed but not cut on its upper side. There is nothing in the device which even'suggests co-action between serrations on type characters and platen, the problem of exact registration or its solution.
Hendrick patent, No. 104,148, discloses a diagonally serrated metal platen designed to cut or abrade the under surface of paper when type heads, arranged in a super-structure, are thrust down upon it. The type heads consist of several disks nested together, upon the edges or peripheries of which are type characters with plain smooth surfaces, across which is suspended a type ribbon. A blow upon the connecting thrust drives the ribbon-covered type characters against the check, inks the check, not the type, and causes a rupture of the paper on its under side by contact with the serrations upon the plate. There is a ser*721rated platen without serrated type, and co-action of ridges and grooves iis not suggested.
Several Beebe patents are confidently relied upon to prove anticipation or prior invention. Beebe patent, No. 576,999, contains a revolu-ble type disk not a cylinder, having mounted upon its face, not upon its periphery, printing characters or numerals, and an ink roller and a platen with depressions intended for co-action with the ridges of the printing characters. The two parts are not formed with a plurality of ridges and grooves, but are formed with one single V-shaped ridge on each, upon the old theory of male and female dies, with the object of producing abrasion of the paper. The. principle suggested by this patent is that of die and counter-die.
Beebe patent, No. 594,319, shows a series of conical printing pins arranged to form the outline of numerals when brought into contact with a platen containing depressions made to correspond with their points. This is far removed from the mechanical problem of the patent in suit. 6
The Beebe patent, No. 783,171, approaches nearest in structure and in theory of operation to those of the patent in suit. It has a cylindrical type-wheel, upon whose periphery are rows of projecting type characters, an ink roller and a platen. The type characters are serrated like those of the patent in suit, running in a direction circumferential of the type wheel. The platen, however, is of hard rubber, and, of course, is not serrated. Therefore, the problem of Todd and the manner in which he solved it, are not suggested by the one similar feature in this Beebe patent.
Todd, the patentee of the patent in suit, holds two patents. The first is No. 784,602, which comprises a cylindrical type-wheel, upon whose periphery are formed projecting rows of printing characters running parallel with , the axis of the cylinder, an inking roller and a platen. The faces of the type characters are plain, having neither ridges nor depressions, and the platen has a surface preferably of soft rubber or other yielding material, and, of course, without ridges and depressions.
Thus the prior art discloses serrated platens with smooth type, serrated type with smooth platens, but does not disclose both type and platen serrated, calling for co-action and precise registration. With these disclosures of the prior art, counsel for the appellant present the question most favorable to their contention, by asking whether there is invention in bringing together in the old machine of the Todd patent just described, Hendrick’s serrated platen and Beebe’s serrated type. In addressing ourselves to this question, it is certain that alone neither Hendrick nor Beebe suggested a device containing serrations upon both type characters and platen, nor together did they reveal the advantages of cutting or abrading both the upper and under surfaces of an embossed letter or numeral, or suggest the means to produce that result. Before Todd, neither the thing which his device accomplishes nor the means for its accomplishment appeared in the art. That the machines of the patent contain merit is evidenced not merely by the vigor of this controversy, but by the fact that in the *722ten years of their manufacture tire public has taken over 200,000 of them at an aggregate cost of about six million dollars. We believe this is due, not to persistence in advertising, but to the novelty and utility of the device, developed and produced by the invention of the patentee. We are of opinion that his invention was not anticipated by the prior art, and therefore hold claims 1 and 3 valid and infringed.
The decree below is affirmed/'