No. 294.Circuit Court of Appeals, Second Circuit.
May 21, 1925.

Appeal from the District Court of the United States for the Eastern District of New York.

Suit by the Elyria Iron Steel Company against the Mohegan Tube Company, Inc., and another. Decree for plaintiff, and defendants appeal. Affirmed.

Suit is upon patent 1,388,434 (claims 4, 5, 10, 14, 16, 17, and 19), for a “method and apparatus for butt-welding thin-gage tubing,” and 1,435,306 (claims 3, 5, 6, and 9), for “butt-welded thin-walled tubing.” These two patents grew out of the same application. Under familiar office rulings, division was required, and the earlier grant may be called the method patent and the later the product patent.

The claims in suit and the subject of litigation may be fairly shown by selecting one claim from each patent and comparing them. The nineteenth claim of the method patent is as follows:

“The method of electrically butt-welding thin-walled tubing which consists in applying welding pressure to a transverse, narrow zone of the tube stock intimately to close the seam cleft and applying impulsive welding current transverse to said zone, all in

Page 828

such continuous and rapid progression and with such pressure and current density as to produce a substantially continuous weld exhibiting in the seam a recurrent welding effect synchronizing with the impulses of the welding current.”

And the third claim of the product patent is as follows:

“As a new article of manufacture thin-walled electrically butt-welded steel tubing, wherein the abutting walls are united by a substantially continuous weld presenting recurrent variations in metal texture in closely adjacent narrow zones synchronizing with varying application of welding current.”

When application filed (June 9, 1919), electrical butt-welding was by no means unknown, and especially had a machine capable of performing the feat been invented by one Parpart, to whose assignee a patent issued in 1900 (No. 658,741) and under Parpart’s patent, business, and a good deal of it, was conducted during substantially the whole period of the life of that patent. Consequently the present patentee, Johnston, not only was bound to know what Parpart had disclosed, but he did in fact know about Parpart, his patent, his machines, and their use.

This is not denied, and is inferable from the method in which Johnston discloses his invention. He declares in his specification that his method is “especially applicable to the high-speed production of thin-walled tubing of improved quality”; the welding being done electrically. He asserts that electric welding of tubing as practiced before his application was “relatively slow, difficult, and costly,” notwithstanding a very great “demand for a high-speed economical method of producing * * * thin-walled butt-welded tubing.” Johnston then asserts that he has reached such a “high-speed economical method,” and sets forth, first, why he thinks others have failed; and, second, why he deems himself to have succeeded.

The reasons for the failure of others, and inferentially of Parpart especially, is said to have been a failure “to correlate the factors of stock-feeding rate, current control, and pressure application,” and that what especially militated against earlier inventors’ production of “thin-walled tube stock” was failure “to secure and reliably maintain accurate meeting register of the edges to be welded; lack of knowledge as to how to supply and control proper heating currents; burning away portions of the edges which were to be welded; * * * improper compressing whereby portions of the edges were forced into lapped relation; heating the edges too far back into the body of the blank.”

The inventor then specifies the particular things which he has done to improve the method of electrically welding thin tubular stock:

(1) He organized his apparatus, so as to feed the stock forward at substantially uniform speed.

(2) His electrodes, forming rollers embracing wide segments of tube blank on each side of the seam cleft, are reliably synchronized and maintained in substantially accurate and continuous register with the seam cleft.

(3) His welding throat is formed of members collectively completely inclosing and supporting the blank throughout its circumference, and the size of the throat is sufficiently less than the external size of the tube blank passing through it to insure a pronounced compression, yet only enough compression to produce a small burr.

(4) The electric current supply is made approximately uniform.

(5) The forming rolls transforming the flat steel into the tubular blank are so organized that the edges of the blank are made to register accurately when and as they reach and pass through the welding throat.

(6) The feed rate must be sufficiently fast to remove the welded tubing quickly from the influence of the welding current.

It is not denied that Johnston’s machine or device does all of these six things, nor that his machine, operating in accordance with these six points, has revolutionized the whole business of seamless thin-gage steel tubing, and the product or result of his method Johnston describes thus: “The welded joint made in accordance with my invention is what, for lack of a better term, I call a `recurrent’ weld. The weld, although continuous, is nevertheless nonuniform in the texture of the metal which forms the seam, and usually nonuniform in physical form.”

His seam, it is pointed out, embodies “recurrent variations,” which correspond to the “current alterations and consequently temperature fluctuations along the seam at the progressively differing instants of welding compression.” The burr of the weld frequently exhibits “a very stitched-like pattern,” which is indicative of the fact that the welding conditions are right, for, when other conditions are similar, increasing the current flow and slowing down the rate of travel of the blank “tend to minimize the recurrent effect”; but said “recurrent effect” becomes more pronounced as the speed of

Page 829

the blank is increased, or the current flow lessened until finally the weld is no longer continuous, and is either made only at recurrent points of maximum heat or entirely fails.

Defendant, after the expiration of Parpart’s patent, took a Parpart machine, reorganized it, and proceeded to make upon it tubing which does not look like Parpart’s, does perfectly resemble Johnston’s, and makes the same at a rate never achieved by Parpart. The lower court held that this was an infringement of both the method and the product patents belonging to plaintiff. Defendant appealed.

Livingston Gifford, Oscar W. Jeffery, Harry G. Kimball, and J. Stanley Preston, all of New York City, for appellants.

Drury W. Cooper, of New York City, and George T. May, Jr., of Chicago, Ill., for appellee.

Before HOUGH, MANTON, and HAND, Circuit Judges.

HOUGH, Circuit Judge (after stating the facts as above).

Defenses insisted on in pleading and argument are lack of invention and noninfringement; but both produce a single position for defendants, who insist that Johnston’s method is no more than a function of a well designed and operated Parpart machine, and all they do is to make tube by such a machine. Some study of the history of electric butt-welding of tubular forms should throw light on this contention, which necessarily implies one of two things — either (since Parpart’s and Johnston’s patents are separated by 21 years) there was remarkable stupidity on the part of those who for so long a time made tubes on Parpart machines, or the tales of immediate recognition of and great success by Johnston are false.

Of Johnston’s success there can be no doubt; he designed a machine for making, and showed how to make, thin butt-welded steel tubing, of such excellence, at such speed, and with such economy that it revolutionized a business of no mean importance. Of the history of Parpart’s invention there is also no doubt. It was used during substantially its whole life by one concern, and an unusually large number of the employees of that company have testified in this case. There were efforts to improve the device, and it was somewhat changed in appearance before the patent expired, but the principle and speed of operation remained unchanged to the end; that is, Parpart arranged to have the flat blank formed into tube shape with longitudinal open seam, approach a pair of electrodes “straddling” said seam, receive heat from an alternate current through the electrodes, and then the two sides of the seam were ready to be squeezed together in a compression throat, which welded the butts, extruding a large burr, which marked the line of the once open seam and of the weld.

Why the Parpart machine made tube with so much waste, made it so slowly, and usually left the weakest part of the tube, not in the weld, but alongside of it, is explained by defendant’s expert (Mr. Waterman) in his usual admirable style, and we accept it. Sum of the matter is that the welding heat for Parpart should be uniform, and the upset of metal large, in order to produce a substantially seamless tube, which was what those employing the Parpart machine wanted. The uniformity of heat prevented speed, and the large upset made a large burr; therefore in actual use Parpart’s device during its patent life never made tube commercially at more than about 20 feet per minute, and when the burr was removed the waste of metal was large from that cause alone.

There may be some truth in the assertion that Parpart’s machine was always trying to compete with seamless drawn tube, a very highgrade sort, and that for that reason no effort was directed toward the kind of thin and (it is said) inferior tube commercially produced at present under the Johnston patent. But it remains also true that efforts were made to speed up the Parpart machines, and in a sense that could be done; i.e., the tube could be made to run past the electrodes much faster than 20, or even 30, feet per minute, but the resulting tube was so unreliable, if not worthless, that the machines were locked, so as to prevent the higher speed which was recognized as desirable for economy. We find no evidence that those who handled the Parpart machines for so long failed to attempt improvements, and to get as good results as they could out of them, nor were they (as revealed by their evidence) apparently unintelligent. Our inference from this is that there was room for invention in respect of electric butt-welding of thin tubular stock after Parpart had told the public all he had to tell.

Johnston disclosed in 1919 an evident familiarity with Parpart’s machine, and a theory of action about thin stock butt-welding of which Parpart evidently had no idea, and which the excellent practical men, who for many years had worked with those machines,

Page 830

had not hit upon. The essential and novel thoughts in that theory are to proportion and correlate time or speed of movement, heat both as to kind to be applied and area to be heated, contact of butts as to accuracy and pressure as to amount.

Johnston’s theory of correlation was to move the material in very accurate register at the seam cleft, as fast as the material could be heated to the welding point just at the cleft, in time to be subjected to a pressure that would cause practically no extrusion of metal in burr. He proved that he could move his tubing, through his heat concentrated as nearly as might be at the cleft, at 75 feet and over a minute, and by using alternating current of large amperage and slight voltage upon seam sides in accurate register weld with a minimum of pressure, and therefore with a minimum of wastage in his burr. Result was at least three times as much tubing in the same time, and less than one-third as much wastage, as compared with anything Parpart had ever done, in over 17 years of actual commercial operation.

Since, as has been so often said, patentable invention can never escape from the concept of means by which accomplishment is reached, we assign as the reasons, in our judgment, for Johnston’s success, the ascertainment with considerable accuracy of the kind and quantity of preferred current, the instantaneous heating insisted on, the small amount of upset metal resulting, and the diminution of pressure so rendered possible. The resultant of this correlation of forces or elements was and is the stitch seam; a weld line which looks as if it had been sewed — i.e., the “recurrent” weld of the patent in suit. If any two of the following elements be known, the other can be computed, viz.: Speed per minute of feed, frequency of cycle of alternate current, and number of stitches per foot.

Of course, to validate Johnston, he must disclose a method of operation differing from that disclosed empirically by the continued operation of Parpart’s machine. We think the difference in operation is stated sufficiently by Mr. Waterman (for defendant), whose testimony as to the Parpart device is plain that it intended to and did give the unwelded tube enough time to get heated by accumulation of alternating stabs (so to speak). By such hesitation in passing the tube form by the electrodes, “the more nearly the results of direct current are realized; this being the theoretical and practical desirable effect.” This is what Parpart did; Johnston showed how to do something else.

It is not always easy to approve a method patent, which can, so far as known, be practiced only with one kind of device; the evidence here seems to prove Johnston’s a method patent with singular clarity, because it is admitted all around that Johnston’s method can be practiced with Parpart’s machine, and what plaintiff now calls infringement was first practiced after men skilled in running Parpart machines got hold of a copy of Johnston’s specification, read it, instantly declared they could run a Parpart that way, and did it, after a reorganization not mechanically vital. Defendant’s machines have been built substantially following that reorganized Parpart machine.

Finding, as we have now done, that plaintiff’s method patent is valid, it follows that the object produced by that method is also an invention, because it is a tubing electrically butt-welded and presenting in the weld “recurrent variations in metal texture,” etc. (claim 3), of which the ocular proof is the “stitched” weld seam.

Defendants still insist that, even if the patents are sustainable, they are not infringed. This argument is very difficult to distinguish from the claim of invalidity; for the one easily understood, outstanding, and overwhelmingly proven thing in this case is that both plaintiff and defendants have a commercial product, manufactured on machines whose genesis and construction are known, and both products have stitched seams, and, if of same size and material, ordinary observers can see no difference between them, and neither looks like the usual commercial product of the Parpart machines.

Further, defendants have produced two specimens of bicycle tubing made on Parpart machines before Johnston patent issued, which are asserted to show stitchlike markings on the inner seam burr; the outer one has been smoothed off. Admittedly some markings are there, in at least one specimen. The argument may be thus put: If the Parpart machine ever made any stitchlike seam markings, we are entitled to make them all the time. This is based on the old bicycle tubes. But, continues the argument for non-infringement, we are careful to so place our electrodes that it is impossible for us to obtain the kind of current described in Johnston’s method specification; therefore we cannot produce welding heat by stabs or single impulses of alternating current. We are building up to a welding heat, just as Mr. Waterman explains that Parpart did; therefore we cannot infringe.

But it is admitted, as we have pointed out

Page 831

above, that the Parpart machine can be so reorganized as to utilize Johnston’s method, and the evidence shows that the major part of the way so to reorganize is to give an alternating current of large amperage and speed up the travel of the tube stock; there is no difficulty of accounting for these few markings as the result of what was not the usual course in the factory using Parpart machines. The argument based on defendants’ evidence as to what they do now will not stand criticism by the standards erected by their own expert. They do not do what Parpart did, because they have confessedly speeded up their tubing delivery and made their compression throat looser; they confessedly habitually produce a stitch seamed tube, which Parpart did not. So the position becomes this: We furnish an alternating current; so do both Parpart and Johnston. We feed the tube stock at a rate and through a compression throat similar to Johnston’s, but our electrodes are spaced even farther apart than either Parpart or Johnston; therefore it must follow that, although our product looks like Johnston’s and not Parpart’s, we cannot infringe.

This rather surprising argument is, we think, explained and explained away by plaintiff’s contention that, while separation of electrodes does impair the effectiveness of Johnston’s method, yet the increment in separation is expressible only in thirty-seconds of an inch, and since the tube to be welded is steel, and the electrodes are copper, and therefore of a much smaller resistance, the natural path of current will always be as largely as possible through the copper; so that, though not as well as Johnston’s machine, either Parpart’s or defendants’ will serve as a vehicle for Johnston’s method, and impairing or badly applying a method will not avoid an infringement of a patent therefor.

We consider the patents valid and infringed, and affirm decree, with costs.

HAND, Circuit Judge (dissenting).

From my examination of the electrical and metallurgical evidence, it seemed to me, and indeed still seems, that the plaintiff has at least not proved that the defendant welds by a single condensed shot close to the edges of the seam, but that more probably its method is by a cumulation of heat from several diffused shots; the weakest weld being at the highest temperature. Johnston’s patent is very vague at best; but the notion of a single condensed shot, which alone does the welding, permeates it throughout. If so, I cannot quite see how the defendant infringes, though its method certainly had its origin in some parts of the disclosure.

I have not, however, been able to persuade my brothers to my view, and that inability has, very naturally, I think, added to the doubts which are inevitable under the circumstances. The system which submits such questions to the decision of laymen upon the evidence of partisan experts apparently satisfies the profession. I have said elsewhere what I think of it; it would serve no purpose to repeat. As to this complicated case it seems hardly desirable to spread on the books my reasons in detail; they follow in substance the testimony of Waterman and Campbell. I can only say that the plaintiff has not convinced me, and I suppose that that means that I dissent.