190 F.2d 543
LANDIS MACH. CO. et al.
v.
PARKER-KALON CORP. et al.
No. 245.
Docket 21952.
United States Court of Appeals Second Circuit.
Argued April 30, 1951.
Decided July 11, 1951.
John Dashiell Myers, Philadelphia, Pa., Kenyon & Kenyon, New York City, George T. Bean, New York City, Chester C. Baxter, Philadelphia, Pa., of counsel, for appellants-defendants.
Drury W. Cooper, and Cooper, Byrne, Dunham, Keith & Dearborn, all of New York City, William A. Strauch, Washington, D. C., of counsel, for appellees-plaintiffs.
Before CHASE, FRANK and L. HAND, Circuit Judges.
L. HAND, Circuit Judge.
The defendants appeal from a judgment holding valid and infringed Claims 1, 9 and 11 of Patent No. 2,010,730, issued to Arthur Harold Lloyd on August 6, 1935, upon an application filed on January 30, 1935. An identical British patent had issued previously. The disclosure is of a machine for the "centerless" grinding of "screwthreads," in which a "work-rest" supports the piece to be ground — the "work-piece" — which is not held at its two ends, as in an ordinary lathe, by a member that advances it in the line of its axis as the grinding proceeds. In the patent in suit the advance may be by means of a corresponding thread on a "backing wheel"; but it may also be because the "work-rest" is set at an angle to the axis of the "backing-wheel" which has a plain surface.1 In that event, owing to the angle just mentioned, there is a sufficient co-efficient of the motion of the "backing-wheel" parallel to the axis of the "work-piece" to move the "work-piece." The consequent rotation, combined with the axial advance of the "work-piece," allows its thread — only the "threaded" pieces are mentioned — to be engaged by the "grinding wheel," whose axis is parallel to that of the "backing wheel," and upon whose circumference are "annular ridges"; which enter the grooves of the "work-piece" and grind them, the "grinding wheel" being rotated faster than the "backing wheel." It is an essential element in the invention that the angle of the "work-rest," and so of the "work-piece," shall be such that the "annular grooves" of the "grinding wheel" "do not foul the thread," (page one, col. one, lines 24, 28); and the optimum angle for that purpose is the "helix angle," a term of art not necessary to describe. However, that angle is only "preferable" (page two, col. one, lines 16-20); it is not necessary to the invention: all that is necessary is that the angle shall be such that the "annular ridges" "mate" with the grooves; and this we read to mean that, although the angle need not be exactly the "helix angle," it must not diverge too far from it. The word "mate," appears only three times in the specifications, (page one, col. one, line 20; page one, col. two, line 20; page two, col. one, line 69), and in each case the meaning is left vague. On the other hand Claims 6 and 7 are for a machine in which "the inclination of the axis of the work-piece to the plane of the axes of the wheels is dependent upon the helix angle of the "screw-thread"; and Claim 3 contains substantially the same language. They are the only claims which prescribe the "helix angle." Claims 1 and 9 in suit use the broader word, "mate," and Claim 11 does not even contain the "mating" feature. Therefore in conformity with the ordinary canon of interpretation of patent claims: i.e., in interpreting a series of claims, a limitation not present in one must not be implied, when the same limitation appears in later claims,"2 we must not imply in any of the claims in suit the limitation that the mating angle need be the "helix angle"; but only that it shall be such as not to "foul the thread". Indeed, the use of the word, "preferably," clearly precludes our identifying that phrase with the "helix angle." It is of course true, since the "helix angle" is one at which the "ridges" of the "grinding wheel" do "mate" with the thread of the "work-piece," that the defendants' machine, in which the "mating" angle is the "helix angle," does infringe Claims 1 and 9 and a fortiori Claim 11. However, in so far as the selection of the "helix angle" can be thought to be the kernel of the invention, none of these claims can rest upon it. Thus we may assume for argument that the selection of the "helix angle" in order to "mate" the grinding "ridges" and "screw-threads" first appeared in Lloyd's disclosure. With this analysis of the meaning of claims in suit we address ourselves to the prior art as a measure of Lloyd's advance, and therefore of the validity of his claims. We shall confine our discussion to two inventions: that disclosed in Hanson's two patents (Nos. 1,640,991 and 1,640,992) which issued on August 30, 1927, and that in Heim's two patents (Nos. 1,579,933 and 1,683,974) which were applied for almost together, but the first of which issued on April 6, 1926. If we assume that both Heim and Hanson had in fact been generally known to the art since their appearance, it had waived for nearly seven years before Lloyd filed his British application on April 12, 1934. In that event, if Lloyd's combination of these disclosures resolved a want which had been felt all along: that is, if it swept the board; there would be reason to conclude that his discovery was beyond the capacities of ordinarily qualified members of the craft.
In order to answer this question we must first take up the two references. Heim's machine was for a "centerless grinder"; it contained a "grinding wheel," a "backing wheel" and a "work-rest," although it did not call all of them by those names. On the other hand, it was concerned only with smoothing the surface of cylindrical "work-pieces"; it neither cut threads, nor ground threads already cut. Taken alone, it was not therefore an anticipation of Lloyd, and we shall further assume for argument that without more it was not even so near that the changes necessary to make it capable of grinding threads, already cut into a smooth cylinder, would not support the claims in suit. Hanson's machine was "for grinding wheels for use in grinding grooved or threaded members such as screws taps or the like, so as to give the grooves or threads accuracy and precision," and it could be used both for forming the threads or "for finishing threads after the same have been roughed out." However, it was a "centered" machine, in which the "carriage, 12, having suitable head and tail stocks 13 and 14," supported the "work-piece" as a lathe supports its work: i. e., the "work-piece" was moved axially and rotated across the surface of the "grinding wheel" by a train of members totally unlike the "backing wheel" of Lloyd and Heim. For threads of small "lead angle" — i. e., pieces in which the threads are close together — in Hanson's machine the axis of the "grinding wheel" and that of the "work-piece" were parallel, apparently because in such cases the angular difference between the sides of the threads and those of the "ridges" on the "grinding wheel" are so small as to be negligible. However, when dealing with a "thread of large lead angle * * * the axis of the grinding wheel is inclined perpendicularly to the lead angle of the thread on the work. In this position the serrations, 55, of the wheel, 56, extend substantially in the direction of the thread angle of the work, 58, instead of across the thread as would be the case if the two axes were placed parallel." (Pat. No. 1,640,991, page 4, lines 125-130; page 5, lines 1-4.) It is true that when the axis of the "grinding wheel" is perpendicular "to the lead angle of the thread on the work," the angle between that axis and the axis of the "work-piece" is not the true "helix angle." However, Hanson's "serrations are so shaped and trued that at the point or line at which they engage the work they will be of such contour and size as to properly and respectively operate on those portions or faces of the thread as has been planned," (Pat. No. 1,640,991, page 5, lines 14-20); and while that does not expressly declare what is the proper angle, it does presuppose an angle which will insure "mating": that is, that the "grinding wheel" shall not "foul the thread," which is just what, and what alone, the claims in suit prescribe.
Thus, if anyone substituted in Heim's machine Hanson's "grinding wheel" with its "annular ridges" in place of Heim's "grinding wheel" with a smooth surface, nothing more would be necessary, save to make the angle such that the "ridges" should not "foul the threads," and that Hanson disclosed. True, the proper angle would have to be fixed experimentally, dependent on the size of the "lead angle," but that too was necessary in Lloyd's machine, because — at the hazard of needless emphasis we repeat it — the "helix angle" in Lloyd was only a "preferable" optimum. It does not necessarily follow that to combine Heim and Hanson was not to create a patentable invention. The length of time between the appearance of the constituent elements of a new combination, each in its own nexus, and their selection as parts of a new combination, is usually regarded as one of the most telling determinants of invention; but "The lapse of time between one invention and another * * * tells very little. * * * until it appears that the art in fact knew the earlier steps, already taken."3 Since a patent is a monopoly not only against those who copy its disclosure but against all the world, it is universally conceded that an inventor's contribution must be measured against the whole prior art, whether he knew of it or not; and it follows that his originality is not to be judged subjectively, but by what attainments were required of a putative inventor who was aware of all that had been done in the past. Not so, however, if we choose to use the period during which the art failed to devise the new combination as an index of its originality. We may not then impute to the art, as we must to the inventor, an acquaintance with all that has gone before; we may not proceed upon the assumption that the invention baffled its ingenuity until the inventor showed it what all along had lain unperceived before its eyes.
Therefore we cannot judge how far Lloyd solved a problem which for over seven years had fettered the progress of the art, until we learn what impress Heim and Hanson had made upon it: i. e., how far these disclosures had become its common property. There are indeed occasions, though they are extremely few, on which we can say with confidence that, even if the craft did not know what preceded, the combination speaks for itself, and only outstanding talent could have produced it. This is not such an occasion. For example, if the record were that very soon after the art had become acquainted with both the Heim and the Hanson machines, Lloyd filed his application, we should not be justified in assuming that the combination demanded high talent. Conversely, if the record was that, although the art had been familiar with these machines for more than seven years, and that no one had thought of combining them; and if in addition the combination met with universal acceptance when made, we should be justified in an opposite conclusion. The issue is always how far beyond commonplace contriving was the foresight necessary to think out the combination. Usually, though not always, it is practically impossible to decide that issue by a mere inspection of the patented disclosure against the background of the prior art. Moreover, it scarcely needs more than the statement of the question to disclose the fatuity of asking judges, undisciplined in the craft and untutored in its inarticulate presuppositions, to say how far the innovation is beyond the powers of merely competent craftsmen. For that reason we have over and over again resorted to the history of what went before, the duration of the period during which the invention was needed but failed to appear and its acceptance when it did. These circumstances have seemed to us, not indeed an absolute determinant of invention, for there is none; but at least the best, and indeed almost always the only, rational approach to a solution.
Accordingly we cannot say in abstractu whether it took, or did not take, more than commonplace abilities to think of substituting in Heim's "centerless" grinder for smooth cylinders the serrated "grinding wheel" of Hanson. Lloyd's patent issued in 1935 and apparently lay barren until 1943. The Landis Machine Corp., its only exploiter, after becoming acquainted with it, did not for four years think it of enough importance to secure any rights under it; it would appear that for more than half its life nobody took any notice of it. (It may of course have been true that during that time the art had not learned of its existence; we are not to assume that all issued patents enter at once into the common knowledge of a craft; that applies as well to Lloyd's patent as to Heim's and Hanson's.) Furthermore, although the Landis Machine Company paid $15,000 for an exclusive license, the record contains nothing to show that, as licensee, it has succeeded in exploiting it, either by selling machines or by granting licenses to others; yet at the time of the trial last October, only two years remained of the seventeen for which it had been granted. It is true that the defendants now do wish to exploit it and have copied it; but the record does not disclose how extensive that exploitation has been, or what has been its success.
All this fails to convince us that the art had been waiting for seven years for someone to divine that a combination of Heim and Hanson into a single machine would be a boon to those who wished to grind "screw threads" in quantity. We do not forget that the two witnesses praised the combination — Harrison and Reimschissel. Harrison was not impressive; all he said was that "it would have been very helpful, had we been smart enough to be able to do the job." Reimschissel said that he thought it impossible to make such a grinder because "we couldn't see any way to control the lead of the thread." That language suggests that he had considered the possibility of constructing a "centerless grinder" for "screw-threads" after the pattern of Lloyd's combination of Heim and Hanson, but had been thrown off because he doubted whether the "ridges" on the grinding wheel could be made properly to "mate" with threads of different "leads." Quite probably he may not have known of Hanson's solution of that difficulty by "tilting" the "work-piece" to accommodate its threads to the "grinding wheel." Be that as it may, at least his testimony shows that such a combination had occurred to others in the art. All that was needed was, either some acquaintance in detail with Hanson, or more persistence in experimenting to overcome what Reimschissel too readily assumed to be an insurmountable obstacle. For these reasons we are not satisfied that the claims in suit are valid; indeed on this record we think them invalid. Yet it is fair to send back the cause for a new trial on the issue of validity where it can be presented in some such way as we have indicated. The plaintiffs may be able to show that Heim and Hanson were well known to the art for a substantial part of the seven years after 1927; they may also be able to show that such a combination would always have been welcomed; and in support of that conclusion to show that the art has adopted it very widely to the exclusion of existing methods. In such a setting our doubts arising from our a priori decision that the combination was not patentable, may disappear. We will not now prejudge the decision of the district court by any present intimations; but we do hold that on the record before us the judgment must be reversed and the cause remanded for a new trial on the issue of validity.
Judgment reversed; cause remanded for a new trial.
Notes:
"In many cases it is immaterial whether the `backing wheel' is provided with a corresponding mating thread, or whether it has a substantially plain, ribbed or other periphery. It is shown in the drawing as having a plain periphery" (page one, col. two, lines 18-22)
Symington Co. v. National Malleable Castings Co., 250 U.S. 383, 39 S. Ct. 542, 63 L. Ed. 1045; Western States Machine Co. v. S. S. Hepworth Co., 2 Cir., 147 F.2d 345, 350; Secretary Pen Co. v. Everlast Pen Corp., 2 Cir., 186 F.2d 575, 577 (C. A. 2)
Condenser Corporation of America v. Micamold Radio Corp., 2 Cir., 145 F.2d 878, 879; Frank B. Killian & Co. v. Allied Latex Corp., 2 Cir., 188 F.2d 940