The present patent infringement suit, as now narrowed by the appeal, involves solely an issue of infringement. There was no adjudication of validity in the court below and the patent having now expired the public interest does not invite a consideration of that issue.

The patent in suit is Anderson No. 1,952,230 for an automatic profile milling machine, granted on March 27, 1934, upon an application filed in 1927 and renewed in 1931. Of its seventy-four claims but three are now pressed as having been infringed. These claims are 13, 46 and 57. They are printed in the margin.1 The patent is owned by the appellant, a substantial part of whose business is the manufacture and sale of hydraulically controlled milling machines and hydraulic attachments for converting existing milling machine tools to hydraulic operation for duplication of irregular surfaces in metal. The appellee manufactures and sells only the hydraulic attachments.

The Anderson patent lies in the art for mechanically reproducing by means of a tracer device the shape and form of a pattern. The tracer is designed to travel over the pattern and through hydraulic means to control the movement of the cutting tool so that the cutter will respond exactly to the movement of the tracer. Anderson claimed that his object was to provide a milling machine that would automatically cut irregular shapes in metal with only the assistance of a master or template and to provide a simple machine adapted to accurately and rapidly follow the template at any desired rate of speed for the cutting tool without the attention of a skilled operator.

Prior to the Anderson patent, technological advances had been made in response to the need for greater exactitude in machine tools. Early in the century, study was given to the development of hydraulic tracer control duplicating machines to replace the hand-operated metal cutting machines then in use. The latter were impractical because of their slowness, expense of operation and their dependence on skilled artisans. The earlier advances toward automatic duplication are exemplified by the Bontempi Patent No. 709,388 (1902) for reproducing plaster sculptoring, Pierce No. 871,522 (1907) for reproducing fluted marble columns, and Lorant British Patent No. 25,939 (1902) for reproducing wooden shoe lasts. The record fails to show any commercial exploitation of these patents but Bontempi had disclosed the use of two slides moving simultaneously at right angles to each other and operated by hydraulic cylinders; so had Pierce. Lorant had disclosed a tracer designed to open a valve controlling a hydraulic cylinder for operating the slide, and that was also indicated by Bontempi, supra, and in his German Patent No. 126,119. A spring pressed tracer was in Delin No. 560,574 as early as 1896, in Shaw No. 1,506,454 (1924) and Lochman No. 1,774,279 (1930). A hand-operated valve for distributing fluids to two or more cylinders was in Pierce No. 827,993, an oscillating tracer in Delin, supra, and Shaw, supra, and a reproducing machine having three directions of mechanical movement in Bontempi, supra, and Shaw, supra.

While no one of the cited patents discloses the combination of elements in Anderson, the recorded art was available to him and lends support to the court's view, that the claims in suit, with the possible exception of a single element in claim 13, disclosed combinations of old elements and we have held that where the only issue is one of infringement, though all of the elements of the patented combination are not found in a single structure of the prior art, yet in determining the scope of the patent and its place in the art as bearing upon the question of infringement, patents showing separate elements of the combination may properly be considered, Remington Rand, Inc. v. Meilink Steel Safe Company, 6 Cir., 140 F.2d 519, 521.

The embodiment of the invention in the patent discloses a completely automatic hydraulically controlled milling machine capable of reproducing irregularly shaped metal parts throughout 360° without manually operating the changeover valve. The tracer is continuously biased against the surface of the pattern. The pattern is mounted on a conventional work slide so that the pattern and work move as a unit while the cutter and tracer-head likewise move as a unit on a slide which moves at right angles to the other. The cutter and tracer are moved by one hydraulic cylinder and the pattern and work-table by another. These cylinders are so arranged that there is an automatic proportioning of liquid from a constant pressure source simultaneously, so as to control the direction of movement of the respective slides. The hydraulic cylinders are controlled by a series of valve mechanisms which control the pressure conditions. The tracer arm tilts with changes in the surface contour of the pattern. Its deflections actuate a primary valve to change pressure conditions therein and the primary valve, in turn, controls a secondary valve which, in its turn, activates an auxiliary section of a distributor mechanism, causing rotation of the distributor valve by pressure delivered to a pair of hydraulic cylinders. The distributor valve is comprised of an auxiliary valve and a main valve which direct the flow of fluid pressure to the main operating cylinders which move the slides in the required direction so that the forces acting on the slides are automatically varied continuously in exact relation to the direction and amount of tracer deflection as dictated by changes in the surface of the pattern being traced.

The patent summarizes the operation as follows:

"Any displacement of the tracer arm from its neutral position will immediately cause a combined responsive action of the primary and secondary valve in connection with the distributor valve and an immediate change in the direction of the resultant relative motion between the two, such as would limit the displacement of the tracer arm, and that the resultant gradual rotation of the distributor valve stem would eventually give whatever change in relative motion that is required and cause the tracer arm to return to its neutral position."

This sketchy general description of the operation of the machine and its parts is, of course, inadequate but they are more specifically detailed in the opinion of the district judge, Cincinnati Milling Machine Company v. Turchan, D.C., 101 F. Supp. 621, in which will be found the principal patent drawings sufficiently indicating the organization and operation of the Anderson machine, so we need not repeat.

The principal controversy revolves about the "means" clauses in the several claims in suit — and this we have first to resolve. It is conceded that Turchan does not use the primary and secondary valves of the patent nor the several parts of the distributor valve with their associated parts. The appellant insists, however, that the "means" employed by Turchan constitute their mechanical equivalent, so that Turchan infringes. The court found that the "hydraulic means" or "pressure responsive means" claimed by Anderson are, in the light of the specifications and drawings, integral parts of the combination claims in suit without which the structures claimed therein could not perform the functions for which they were designed. The appellant challenges this view, contends that the "means" clauses in the claims are merely background, that the claims are not limited to the means which Anderson disclosed, that they are not part of his inventions because they do not relate to his inventive concept. The question, therefore is, on this phase of the case, whether the hydraulic organization by which the movement of the cutting tool is accurately controlled through tracer deflection communicating through the primary and secondary valves and then through the distributor valve and its associated parts to give that sensitive control of the cutter and that extreme accuracy of the cutting which is within the inventive novelty of the patent.

This leads us to a consideration of the very difficult question, whether the "means" or "mechanism," referred to in the claims, is used, as Judge Denison expressed it in Davis Sewing Machine Company v. New Departure Manufacturing Company, 6 Cir., 217 F. 775, 782, "with reference to the element or sub-combination which is the real point and gist of the invention, or whether it is used with reference to the elements or parts of the combination already wellknown and designed only to co-operate with the new element in order to make a completely operative unit. In other words, where used with reference to the exact point of novelty, `means' or `mechanism' may expose the claim to attack on the ground that it is functional; in that respect, each case will present a problem by itself." We have held that where the novelty that exists in a combination of elements set forth in the claim, resides in the means, the claim must be limited to such means as are disclosed and illustrated in the specification. An inventor cannot by the mere use of the word `means' appropriate any and all kinds of mechanism which may perform the specified function or any other mechanism or device than that which is described in the patent or which is its mechanical equivalent, Ford Motor Company v. Gordon Form Lathe Company, 6 Cir., 87 F.2d 390, 391, Hollingshead v. Bassick Manufacturing Company, 6 Cir., 73 F.2d 543, 548. The word `means' is itself an indefinite statement and as such must be interpreted in the light of the patent drawings and specifications, Remington Rand, Inc. v. Meilink Steel Safe Company, 6 Cir., 140 F.2d 519, 521.

Therefore, to determine whether the claims in suit of Anderson must be limited to the hydraulic means disclosed in the patent or their equivalents, we look to the drawings and specifications and we give weight to the view of those best qualified to interpret them contained in the record. The Anderson specification devotes space and emphasis to the organization of the hydraulic means by which the movement of the cutting tool is controlled, stresses the complete integration of the action of the primary and secondary valve in connection with the distributor valve and the movement of the tracer arm, leading to the conclusion: "This action enables the applicant's milling machine to automatically at a predetermined constant feed follow the master or template and describe any conceivable shape that can be cut by a circular milling cutter." Ray, the leading expert for the appellant, spent much time reviewing the operation of the primary and secondary valves in co-relation with the distributor valve and associated parts. On direct examination, he explained: "You can see that with any kind of valve which is opened or closed and which has to control the flow and consequently the pressure in that cylinder. * * * But to control the movement of this valve stem 24 possibly to 1/10,000th of an inch, that is a surprising thing about these valves, why, this is the only scheme which I know of that could be done practically." And, again, "for carrying on the particular work that has to be done and do it accurately. So that really accomplishes a very unusual result." Once more, he explains: "Of course, this secondary valve also can move to a greater extent than the primary, if necessary, and Anderson speaks of one advantage of this construction, that he will get a magnification of the movement of the primary valve. * * * The whole structure makes an extremely sensitive controlled valve." He calls attention to the language of the specification which reads: "The primary valve and the secondary valve work together as a unit, the displacements of the secondary valve magnify corresponding displacements of the primary valve." And again, referring to the specification, he quotes: "The purpose of the main valve is at all times to so direct the flow of fluid pressure to the main operating cylinders "A" and "B" as to give approximately the required direction to the resultant motion of the tracer and cutter."

Even more significant is Ray's testimony under cross examination:

"Q. Well, let's go a little farther, Mr. Ray, let's tear out the auxiliary valve and throw it away. Now, you haven't anything resembling the Anderson patent, have you? A. That is right.

"Q. Let's go still farther. Let's tear out the main valve of the distributor and throw it away. Then you are away from the Anderson patent entirely, aren't you? A. Certain portions of it.

"Q. Alright. Now then we have thrown out the main valve, the auxiliary valve, cylinder 44, and cylinder 45. Let's go back to Figure 4 and tear out that valve `G' and let's move that piston 23 by some other method. Still farther away from the Anderson invention, aren't you? A. We are farther away from the particular embodiment disclosed by Anderson."

Throughout his testimony, Ray carefully returns to qualify his answers by confining them to Anderson's particular embodiment. He had directed the making of a drawing to show that without the parts that Turchan omits, Anderson can still function well in certain operations with great sensitivity and accuracy. What those operations are he does not disclose. It must be noted also that nothing like the drawing is even remotely suggested in Anderson, and that the drawing itself was made years after the advent of the Turchan machine. Even so, the court doubted that the device so reorganized would be operative, — but upon that we need spend no time.

Ray testified further that one function of the secondary valve was that it increased sensitivity, that the primary valve itself, its design and the hydraulic circuit that it controls is one in itself that gives great sensitivity to the circuit which it controls, — "that is a particularly advantageous circuit, that hydraulic circuit."

"Q. And that particular function is inherent in that structure is it, as Anderson discloses it and operates it? A. Well, if the valve is designed to give those results it is bound to give them, and I suppose the results are inherent in the design."

And more to the same effect.

We must conclude from this cursory review of the expert evidence most relied upon by the appellant that the court was not in error in concluding that Anderson's hydraulic organization of primary and secondary valves co-operating with the distributor valve, under tracer control, was a necessary element of claims 13 and 46 in suit and within the "means" clause thereof, and that claim 57 includes much of that essential organization. Turchan does not use it, or its reasonable equivalent. Compare D'Arcy Spring Co. v. Marshall Ventilated Mattress Co., 6 Cir., 259 F. 236, 240; Hollingshead v. Bassick, supra.

We are supported in this view by the representations made on behalf of Anderson to the Patent Examiner, in distinguishing his invention from prior art. To emphasize the novelty of Anderson's valvular combination over prior art references, it was said: "There is certainly no suggestion anywhere in any of these references of the feature of a tracer controlled mechanism in which there is a primary valve connected with the tracer, a secondary valve which is operated by that primary valve and a distributor valve, such as the Valve Mechanism I of Anderson which is controlled by its secondary valve and which, in turn, determines the relative motion between the carriage and work table. As has been pointed out in the Anderson application, his distributor valve makes possible the proper allocation of the total available hydraulic actuating medium between the control motor mechanism for the movable slides or carriages such that a continuous conforming relative movement of the cutter and work which is in a path — not as within Keller and the references of the prior art, an irregular approximation of — but is actually a reproduction of the path or contour determined by the template or pattern."

It was vigorously contended below that patent office arguments are merely preliminary negotiations and that file wrapper history to ascertain the meaning of claims is looked upon with disfavor. In Wiegand v. W. Bingham Company, 6 Cir., 106 F.2d 546, 548, we fully explained that we have not so tightly closed the door to inquiry upon the precise concept of the inventor, measured by his own representations, with citation of precedents. While conceding that extrinsic aid to construction must be accepted with caution, yet if within the difficult art of claim draftmanship terms are employed in effort to avoid prior art, which are susceptible of construction, there should be no more reluctance to search for precise meaning than in a private contract.

Much importance is attached by the appellant to that element in claim 13 which reads: "And a tracer arm, the axis of which is inclined to the axis of the tracer head when in its neutral position." It is conceded that the electrical patents such as Shaw No. 1,683,581 show a tracer head and a tracer arm capable of lateral deflection but it is urged that these electrical means are devoid of any suggestion of Anderson's hydraulic neutral with attendant advantages. However, it is shown by the evidence that it had been well understood in the art that for accurate duplicating there must be a firm contact between the tracer and the pattern and that this contact must throughout the tracing operation be maintained. It must be presumed, therefore, that given a pivoted tracer arm and the need for continuing firm contact with the pattern, a deflection of the tracer head is inevitable. This is well demonstrated by Ray's response to the question:

"Q. It is, therefore, the thrust delivered to the tracer arm by the pattern that brings about the tilt of the tracer arm, isn't it? A. Yes, I think it could be expressed that way."

But, says the appellant, "These electrical means are devoid of any suggestion of Anderson's hydraulic neutral with attendant advantages." These advantages, however, flow from the integrated action of the inclined tracer with Anderson's novel hydraulic means. Turchan does not employ these means and it is a combination we are considering. Schumacher v. Cornell, 96 U.S. 549, 554, 24 L. Ed. 676; Dillon Pulley Co. v. McEachran, 6 Cir., 69 F.2d 144, 147.

A contention is advanced that Turchan's device employs a hydraulically underlapped valve and it is conceded that if he uses an overlapped valve, claim 46 could not be infringed. Both record and opinion are confusing upon this question and the record is inconclusive that defendant employs an underlapped valve. The court declined to make an express finding that he did so, or to make any finding which would lead to such inference. Referring to one of the exhibits showing the dimensions of an assailed device, Ray testified that it had a physical overlap but, when asked whether the word "physical" had particular significance, he undertook to explain that there might be a physical overlap, but a hydraulic underlap, depending upon whether the shoulders of the valve-spool were tapered or rounded, a distinction which it is somewhat difficult for a technologically unschooled judicial mind to grasp.

To summarize our discussion, a close study of the patent and the record indicates that what Anderson accomplished in achieving his objective, of an automatically controlled machine, was to devise an organization of hydraulic means, novel in the state of the art, and by which he attained great sensitivity and accuracy in the duplicating operation. For this, he was granted a patent, and we may concede that it had great commercial value. The court below found, and we agree, that Turchan, while using much of what was disclosed by prior art, did not use the hydraulic organization which Anderson devised and upon which he was given a monopoly by the claims in suit. Nor did he use reasonable equivalents thereto, in view of the fact that Anderson was not a pioneer and his claims must, therefore, be narrowly construed.

The decree below is affirmed.