This is an appeal by the plaintiff from a final decree in favor of the defendant in a suit brought for the infringement of two patents, viz., reissue No. 18576 to Sedgley for method of, and apparatus for, making coils dated August 23,1932, and No. 1886671 to Gottsehalk and Grater for metallic- bunch and method of making it dated November 8, 1932. The District Judge held that some of-the claims in suit were not infringed and the rest were void, and dismissed the bill.
The art involved is that of making “metal sponges” so called. Those in question are composed of thin flat wire, about 1/2000’s inch thick, which by the action of the Sedgley machine has been converted into long curls. The patent refers to the wire as “coiled,” but “curled” is a better term. The diameter of the curl is about an eighth of an inch; it comes from the machine in a continuous length and is loosely reeled or coiled into a size convenient for use; in this form it is packaged and sold. This product, when made of bronze wire, was found to be very useful in scouring and cleaning, serving many of the purposes for which steel wool is used, but in a safer and more efficient manner. It was admittedly first introduced by the Gottsehalk Company about 1926, and in the eight years since more than twenty-five million of these metal sponges have been sold. The product itself was not patented, as apparently it might have been. The Sedgley patent is for the method of making such curled wire and the machine for carrying out the method. The Gottsehalk and Grater patent is for the form in which the sponge is marketed.
At the time when the Gottsehalk sponge was first manufactured, there was upon the market a variety of cleaning devices made of material having a core of cotton or jute around which a thin flat wire was wrapped. These were objectionable because the core absorbed grease and was not easily cleaned. The Gottsehalk sponge, the wire curls having no core, is free from this objection; and there is evidence that it is more efficient and retains its efficiency better than those previously made.
The principle involved in making the Gottsehalk curled wire is that certain thin flexible materials, when drawn under tension over an edge like a dull knife blade, acquire internal stresses which, when the tension is released, cause them to take the form of a curl. The principle can be illustrated by drawing a narrow strip of paper across a dull knife or even across the end of a finger nail. When the strain on the paper is released, it takes the shape of a coil. This fact was well known in a general way before Sedgley entered the field. It was also known that any wire, flat or round, with or without a central fibrous core, would be waved or curled when drawn over a curling edge. But nobody before Sedgley, unless Gottsehalk, a point which will be discussed later, ever seems to have appreciated that a thin flat wire could be made into a continuous curl in this way. The defendant has *585introduced many prior patents; but none of them show anything like the Gottsehaik material, or any machine on which it could be made unless the machine was alterecj. The idea of this curled wire yam — so to term it — originated as has been said with Gottsehaik; and, of course, there was no prior art of making it. The learned District Judge regarded this material as a mere modification of the previously known wire-wound yarn. We are unable to agree. It seems to us that it was radically different from anything that had preceded it.
The defendant contends that the continuous metallic curl was Gottschalk’s invention, not Sedgley’s, and that Sedgley’s machine was also Gottschalk’s invention. The facts on this point appear to be as follows: About 1922, Gottsehaik was manufacturing covered wire, tinsel, and other similar products. He was familiar with cleaning cloths and sponges made of wire-wound yarn, sometimes called “gimp,” and he knew the objection to such cleaners, lie conceived the idea that it might bo possible to make a flexible scouring yarn of curled wire without any core. There is no doubt that this conception originated with Gottsehaik. He then began to experiment in efforts to produce such a coreless wire yarn. These experiments are described by himself and by Davidson, a disinterested witness called by the defendant, who worked with him. They set up a machine which had delivery rolls and a curling edge, but they failed to appreciate the necessity of proper tension of the wire as it was drawn over the curling edge. The product was not satisfactory. Sometimes, probably when the tension happened to be right, it was marketable; at other times it was not. Gottsehaik testified that 2,000 to 3,000 pounds of it was scrapped. Some of it seems to have been sold.
Sedgley had a machine shop and made machinery. He was in and out of the Gottsehaik plant; and Gottsehaik told him what he was trying to do and had done and asked his assistance. This led Sedgley to attack the problem. Working in consultation with Gottsehaik, Sedgley eventually succeeded in devising the machine of his patent. He discovered the part which tension played in producing a satisfactory result, and arranged! his machine accordingly. When it was put to Sedgley by the defendant on cross-examination whether he claimed to be the inventor of the material of the Gottsehaik sponge, he answered that ho did not, but that he did claim to be the inventor of the machine for making the material and of the method for making it. It should be noticed that in the testimony the word “curl” is used in different meanings; and this must be borne in mind in considering the evidence. Sedgley himself speaks of “an inherent tendency to curl,” as shown in the product made under the Akin patent. An examination of that patent shows that it had no eurling edge and that the Akin product was a waved flat wire, having no curl at all as that word is used in the present proceeding. In Sedgley’s cross-examination which is strongly relied on by the defendant, he said, as quoted in the defendant’s brief, that “he knew it was very old to curl wire under tension over an edge.” But he immediately added, “The material that had been curled in that manner before was an entirely different material to what we were using or what we experimented on. The coiling of springs was very old and other material likewise, but I claim that the art of eurling this particular material which is so fine, being less than 1/1000’s in diameter is new.’’ These statements are, we think, in accord with the facts.
The Sedgley machine appears to have solved Gottschalk’s problem. It is still used by his company. In it Sedgley combined the eurling process with a preparatory process of flattening round wire by running it through rolls. In so doing, he found a good deal of difficulty in adjusting the pull of the delivery rolls, by which the tension was maintained, to the speed with which the wire came through the flattening rolls. Possibly other and more subtle difficulties were involved, because, as has been said, it appears to be necessary for the wire to be under a certain tension when passed over the eurling edge in order to curl properly. At any rate, much experimenting was done on the delivery rolls of the Sedgley machine before a satisfactory arrangement was obtained.
It is said by the defendant that Sedgley’s machine is but an obvious application of well-known principles and did not involve invention. Gottschalk’s testimony is strongly against this view. He made determined efforts to get the machine, but without commercial success, — which was the reason why Sedgley was called in. As has been said, Gottsehaik used draft rolls and a curling edge, but he failed to appreciate the necessity of proper tension. Davidson says in his testimony, “The wire was not drawn over the blade under tension according to my recollection”; and he showed no tension device in his sketch of the machine. Sedgley *586says that the essence of his invention lay in the idea of running this thin flat wire over a curling edge under a tension maintained by rolls on the delivery side of it; that he discovered that such an arrangement would produce the curled metallic yarn which Gottschalk desired. This statement appears to us to be correct. It is true that the machine is extremely simple, and at first sight seems obvious in view of the prior art on gimp machines. But inventions are often not quite so obvious before they are made as they afterwards appear. No previous machine would do the work of Sedgley’s machine, without rearrangement (see Topliff v. Topliff, 145 U. S. 156, 12 S. Ct. 825, 36 L. Ed. 658); and the alleged prior uses in the Montgomery and the Harris machines are not in our opinion an anticipation of Sedgley’s machine. The Sedgley machine must be considered not as an aggregation of parts, but with reference to the problem which it solved. We think it involved invention.
The defendant’s' next contention is that even if the original Sedgley patent was valid, the reissue is not, at least as to the last four claims, on the grounds that they were not Sedgley’s invention and were inserted to broaden the patent in order to cover the present defendant’s machine. The District Judge so held.
The claims in suit of the Sedgley reissued patent read as follows:
“19. The method of making curled wire which comprises rolling wire to flatten it, maintaining thq flattened wire while moving under tension, and causing the flattened wire while moving and under tension to pass over a curling edge which imparts to the wire an inherent tendency to curl. * * *
“28. The method of making a metal strand in the form of a helical coil, which consists in feeding a metal strand under tension, continuously working the metal at one side of the strand while it is under tension to stretch the opposite side of the strand, thereby causing said strand when released from tension to assume the form of a helical coil.
“29. The method of making a metal strand in the form of a helical coil, which consists in feeding a metal strand under tension, and, while it is under tension, passing it over a curling-edge to impart to it an inherent tendency to form a helical coil when released from tension, and releasing the strand thus treated from tension to permit it to assume the form of a helical coil.
“30. A machine for making a metal strand in the form of a helical coil, comprising means for feeding and tensioning a metal strand, and means to stretch the metal at one side of the strand while it is under tension to cause the metal strand when released from tension to assume the form of a helical coil,
“31. A machine for making a metal strand in the form of a helical coil, comprising means for feeding and tensioning a metal strand, and a curling-edge over whieh said stand is fed while under tension to cause the metal strand when released from tension to assume the form of a helical coil.”
The District Judge held that claim 19, which is the same in the original patent and in the reissue, was not infringed, and that the other claims in suit were invalid.
The defendant makes the Gottschalk material and uses the Sedgley machine except that, in an effort to avoid claim 19, its machine is divided, the wire being flattened in one machine and curled in another. The question is whether this avoids infringement. The answer depends in the first place on the interpretation of the claim. The defendant contends that the expression in the claim “maintaining the flattened wire while moving under tension,” refers to wire moving between flattening rolls and delivery rolls; and that this element does not appear in the defendant’s machine. The plaintiff interprets the claim as meaning, in effect, maintaining a flattened wire under tension and causing it, while so moving, to pass over a curling edge, etc. If this view.be correct, the defendant’s curling machine infringes because it moves the wire under tension over a curling edge.
We are dealing with a patent on the first machine devised to make a new product. Under well-established law it is entitled to a liberal interpretation in favor of the patentee. Flattening the wire is a mere mechanical process and was no part of Sedgley’s invention, as was said by the Examiner when the patent was in the Patent Office. His inventive idea is well expressed in claim 19, viz., that the thin flat wire must be maintained under tension while being drawn by rolls over a curling edge. If this claim be not limited by the language of the specifications, we think it should be given a construction broad enough to include the defendant’s machine.
The defendant insists that the language of the specifications precludes such a construction. In the opening paragraph of the specifications it is said, “The object of *587this invention is to deviso a novel method of and apparatus for making coils (i. e. curls) wherein a continuous travel is imparted to a wire and during such continuous travel the wire is subjected first to a forming operation and thereafter to a coiling (curling) operation.” But the original patent also states, “It will of course bo apparent that it is within the scope of this invention to form coils from lengths of material without subjecting the material to a flattening operation.” (Page 4, lino 55). There wás no statement by the applicant with reference to the meaning of claim 19 while the original patent was in the Patent Office. We do not think that the general statement relied on by the defendant, as cutting down this elaim, is sufficient to do so in view of the explicit statement to the contrary above quoted. We conclude therefore that elaim in the original patent was valid and infringed.
The final question upon it is whether it was lost to the patentee by the reissue proceedings. Only minor changes were made in the specifications in connection with the reissue. The defendant contends that Sedgley’s statements in those proceedings create an estoppel against a construction of this elaim broad enough to include the defendant’s machine. The statement particularly relied on occurs in Sedgley’s oath to the application to reissue. He there said:
“[The] claims are unduly limited by the inclusion of the step of passing the wire between forming rolls, or, in other w'ords, the flattening of the wire or changing its contour in cross section. This is an unnecessary limitation because if a formed wire is passed through the macfiine the essential features of applicant’s invention could be used without coming within the scope of the claims.
“The essential feature in applicant’s invention is imparting to a strand of wire, while being fed under tension, an inherent tendency to curl, so that as soon as the wire is released from tension the curling action will be effected. It will thus be clear that the claims are unnecessarily restricted with immaterial details of the construction and the steps of the method and do not give applicant the protection to which he is entitled and which it is necessary for him to have to properly protect him in the advance in this art which, he has, in fact, made.”
It is contended by the defendant that these statements amount to an interpretation of claim 19 as applying only to a continuous flattening and curling machine, and that they estop Sedgley and those claiming under him from maintaining that the claim applies to a divided machine. Statements relating to a matter of law like the construction of a written instrument do not ordinarily create an estoppel unless made under circumstances which render it inequitable to permit the party to repudiate them. A patentee will not be permitted to repudiate a construction which he has put upon a elaim in a communication to the Patent Office in order to obtain the allowance of it. But we do not think this principle applies when the patent has issued and the applicant for reissue is contending that the claims are not as broad as required to protect the invention disclosed. See Spalding & Bros. v. John Wanamaker (C. C. A.), 256 F. 530, 534. No decision going so far has come to our attention. A limitation on claims imposed by the Patent Office and accepted by the applicant stands, of course, on a very different footing. We conclude that claim 19 was not lost by the reissue proceedings.
The District Judge was of opinion that Sedgley was limited as to the claims in suit by the cancellation of the original claim 1 in the application. That elaim, however, made no mention of tension or of a “eurling edge,” and was much broader than claim 19. Sedgley’s acquiescence in the rejection of it does not therefore estop him or his assignees from pressing this claim which we accordingly hold to be valid and infringed.
With reference to the other claims in suit, namely, 28, 29, 30, and 31, we may say briefly that 28 and 30, which do not include a curling edge, go beyond Sedgley’s invention. They are too broad and are as the District Judge rightly held invalid. The twenty-ninth and thirty-first claims are in substance the same as claim 3 9 as we construe it. They are valid and infringed.
The other patent in suit, that to Gottsehalk and Grater No. 1,886,671, is upon a form of metal sponge. The single claim in suit reads as follows:
“3. A flexible and resilient metallic bunch, comprising a eoreless metallic strand having an inherent tendency to form a helical coil, said strand being wound upon itself with juxtaposed strands interlocking with each other throughout its mass to form a flattened, flexible and resilient metallic bunch, with the outer strand portions circumferentially disposed around the inner strand portions.”
*588The District Judge was of opinion that this claim described nothing but a reeled coil of the Gottsehalk metallic yam. He accordingly held the claim invalid for lack of invention.
In making the sponge covered by this patent, the Gottsehalk yarn is gathered on a reel or into a coil, which is cut off when a proper amount has been obtained. The cutoff portion is then taken up on the fingers of the operator, and the coil is turned inside out. It is said by the plaintiffs that this action brings the shorter loops from the inside to the outside of the buneh and puts them under a tension which holds the sponge together. It was an obvious thing to manipulate the eoreless wire yarn so as to hold the sponge together. We do not think that this particular way of« doing so amounted to patentable invention.
The decree of the District Court is vacated and the ease is remanded to that court for further proceedings not inconsistent with this opinion; the appellants recover costs of appeal.