The bill filed in the District Court by appellants here charged an infringement of patent granted to them on March 11, 1919, as assignees of Nesbitt et al., No. 1,297,226, for a separable core used in building automobile tires. The District Court gave a narrow construction to the claims, whereby it resulted that they were not infringed. Claims 1, 2, 5, and 9 were in issue.

The device consists of two primary parts — the separable core and the locking ring. ■ The core is of solid metal, annular in form. Its *693main part is cylindrical in cross section. It carries an inwardly projecting continuous radial flange, with a flat surface on one side, and having on this side a complete annular groove sunk into the flange surface to a substantial depth. This core and flange are divided into at least two parts by cuts, leaving between them a separable key section substantially less than half, and the cuts are at such an angle that the section is removable inwardly. The locking ring is of a size and form to fit the groove in the flange. It follows that, when the two sections of the core are placed together and the ring dropped into the groove, the two sections will be held together, and when the ring is , removed they will come apart. Thus the tire can be built and vulcanized upon the core, and when it is finished and the locking ring is withdrawn, the separable section can be moved inward out, of the tire, and the tire can then be taken off from the remainder of the core.

[1] It is obviously necessary that the ring should be held firmly against displacement while th.e tire is being formed. This had been done by a series of nuts and bolts, registering holes being provided through the ring: and the bottom of the groove to the other side of the flange. In these holes threaded bolts were inserted and the nuts placed thereon were turned down to make the device rigid, and when it was desired to have the core collapse, the nuts and bolts were taken out and the ring removed. It occurred to Nesbitt (we refer to both the joint inventors by the name of one) that the locking ring would be held with sufficient permanence in its effective position, if it were so constructed.as to make with the groove a driving fit, so that as it was forced into the groove it would bring the core sections tighter together endwise, and would stay in position until it was driven out again by a suitable reverse driving or other force. He experimented with this until he built it in an effective and operative form and found it successful. The'specific form in which he made his first working device, and which is shown in his patent drawing, provides complementary tapers upon the inner side of the groove and the contacting inner side of the ring, whereby in driving in the ring there is a wedge action between the ring and the flange, operating to force the coré sections together; but it is the patentees’ theory, essential to a finding of infringement here, that this tapered form of groove and ring was the specfic invention covered by claims not in suit, while the more generic invention consisted in the use of the ring with a driving fit, with no aid from other parts, and no matter whether the contacting edges of the ring and groove were tapered or were at right angles to the plane of the core — a self-retaining ring, effective by itself alone, as a core section retainer lock.

We have concluded that this broader idea involved invention, and was patentable over the old form, which used the bolts, and over other forms shown in the published art. The reasons which lead to that conclusion rather than the other are never demonstrative, and it is not profitable to attempt to state them at length. The change was undoubtedly of substantial utility; the older and less rapid and more complicated forms had been long in use; several different and apparently commercially unsuccessful attempts had been made to get something better; and when this improvement was made it received very *694substantial, although not universal, recognition and acceptance. We think it should be deemed, under all the environment shown by the record, invention rather than mere mechanical skill.

We are satisfied, also, that claims 1 and 2 should be interpreted as not limited to the tapered form, but as satisfied by the straight-sided form. This conclusion results from a study of the language of the respective claims and of the proceedings in the patent office. For this purpose claim 1 may be taken as typical of the broader class. It is:

“A core of the character described comprising a plurality of separable sections having inwardly extending portions provided with grooves, which, in the assembled condition of the sections, provide an annular channel, and an annular locking ring adapted to be forced into and out of the said channel by substantially direct axial movement and to he retained in locking position by frictional engagement.”

Claim 8, on the other hand, expressly calls for “an endless taper ring,” and claims 3 and 4, among others, specifically call for “an annular wedge ring.” Unless there is something in the specification or patent office proceedings which forbids, it is a reasonably clear inference that the “annular locking ring” of claim 1 should not be affected with the limitation carried by the reference to the “wedge” and the “taper” form in the other claims cited.

The Patent Office proceedings rather confirm than dispute this natural inference. As the application was filed, all the claims were plainly confined to and recited the wedge elements. The application went into interference with three others, and after some months of delay the decision was in Nesbitt’s favor. Following the usual practice, and taking the benefit of the education gained by the interference, Nesbitt then added five more claims (4 — 8). One of these called for a taper, but the other four did not, and the references to the ring were substantially in the same form as in issued claim 1. Thereafter, and when the Nesbitt application had been pending for about two years, Yemiker filed an application, the only claim of which expressly called for a ring and groove with a complementary taper in each in the precise form which Nesbitt had stated to be his preference. An interference was declared between Nesbitt and Yemiker.- At the suggestion of the Patent Office, Yemiker had copied claims 1 and 8 of the .Nesbitt application, and these became counts 1 and 2 of the interference. Count 1 was the same as claim 3 of the issued Nesbitt patent (not now in suit). It expressly called for the wedge ring. A similar call and limitation in count 2 of the interference, being claim 8 of the Nesbitt application at that time, and claim 9 of the Nesbitt patent as issued, were not express, but it was eventually held by the Patent Office that they were implied. The interference between Nesbitt and Yemiker thus progressed, with issues limited to the specific form of taper ring, while all the time Nesbitt had in his application, and substantially allowed, the broader claims to the locking ring regardless of taper. That these underlying claims were possibly forgotten in some of the interference decisions cannot change the fact or its effect.

At this point we may well observe the distinction there is between the two forms of rings and their functions in this connection. When *695the separable core sections are placed' together endwise, as closely as they naturally would be, and as they must be in order to form the continuous annular groove, very little further motion is possible; but whatever tends to pull the sections inward radially will tend to force them further together endwise, and to bring them exactly into the predetermined perfect circle. Hence it is evident that the tightening and holding effect of driving in the ring will come from the resistance of the sections against moving inwardly and from the friction thus resulting between the inner edges of the groove and the close-fitting ring. Since the total of permitted motion is almost negligible, the effect would be practically the same whether these edges are tapered or straight. In either case there is a driving fit and nothing more. The difference is that, with the wear which comes from use, the straight-edged form will have'a tendency to cease to be sufficiently tight-fitting, while with the tapered form this wear is compensated by driving in a little further, and the taper permits more inaccuracy in assembling the parts before starting to drive the ring. We think it is clear enough that there would be no invention in merely substituting a taper edge for a straight one in this combination; nor do we understand that the Patent Office held to the contrary. Nesbitt’s claim to the specific form was upon the basis of his other broader claims, and the differentiation justified the specific claims; while from the standpoint of the Yemiker application he was the first to form either the straight or the tapered edge, and therefore had a right to make the claim in interference as well as the broader one, unless the straight-edged form was disclosed by his patent of August, 1914, in which case he had abandoned any right to the broad claim.

In the Yemiker-Nesbitt interference, it appeared that they were workmen in the same factory; that Nesbitt finished the first full-sized working core embodying his friction ring about January 20, 1913; and that he had been engaged for a considerable time before that in making the necessary plans, drawings, molds, and castings. He dates conception back to the early part of December, 1912. He has some corroboration. Yemiker claims conception and disclosure back in June and September, 1912; but his corroboration was held by the Patent Office tribunals to be insufficient so far as it pertained to a wedge ring. We agree with this conclusion, though for additional reasons. The case turned in the Patent Office tribunals and in the District Court upon the effect of an admission by Nesbitt in his testimony that Yemiker had made a certain disclosure to him about January 14, 1913. The Patent Office rejected this disclosure, because it did not reach the wedge form, and therefrom the District Court drew the inference that, not only the Nesbitt claims, which were in interference, though not now in suit, are limited to the wedge form, but that the broader Nesbitt claims, which were not in the interference, but which are now in suit, are also so limited.

If this disclosure of January 14 were out of the way, it would still not be easy to see why the testimony of Nesbitt and his witnesses should not be accepted to the effect that he already had his complete invention before this date; but we proceed to a further study of this *696disclosure, assuming that, if it had covered the subject-matter of the claims now in suit, the Nesbitt patent must fail.

Yemiker had two earlier patents, one filed in April, 1910, and one in May, 1912, covering methods of reaching the desired result by a contractile ring fitting in a groove practically like that of the Nesbitt patent and the old art. In August, 1912, he filed an application on a device which had no groove or ring, but in which each two segmental core sections were held together by a pair of wedges inclined away from each other, connected by a bridge at their heads, and driven into corresponding inclined slots in the respective sections. Thus, as the wedges were driven in, the sections were pulled endwise together arid were also held against relative radial motion. At some later time the thought occurred to him to unite these wedges with the groove and ring which were old in the art. This was obviously by way of improvement upon his form applied for in August, and a court is not to be easily convinced that he had this improvement in June and July, and made no reference to it in his August application. However, at some time he did get this idea, and this was what he explained to Nesbitt early in January. It will be remembered that the old art did not disclose any ring, which, because of its taper or any kind of driving fit, served effectively as a self-retaining lock, and there is no reason to think that the device or idea of Yemiker explained to Nesbitt contemplated any such ring. It was rather the precise thing shown in Yemiker’s August, 1913, application, viz. a ring with four of these peculiar compound tapered wedges projecting from its bottom, and the flange had not only the groove, but had corresponding slots extending through its bottom to receive these tapered wedges.

It is quite apparent that, unléss the parts were made with watch-like precision, the ring could not have any binding or locking effect; but the full functions of the wedge drawing would have been accomplished by the wedges in the earlier part óf the operation, and the ring would not be in contact at all upon its inner edge with the groove. The wedges were so made that their only direct effect was to pull the sections together circumferentially to the limit of the power of the wedges and hold them there, with any radial motion forbidden. The locking ring could function only by its tendency to cause radial motion, and this was impossible becarise there was no room for further motion, and if there was it would break off the wedges, if the ring and wedges were rigidly attached together. If they were separate (as Yemiker says they may be), and the wedges were driven in first, then any possible further effect by driving in the ring would loosen the wedges. The model which is produced as representing Yemiker’s disclosure and his August, 1913, application shows precisely this result, and when the wedges are tightened down the ring does not bind upon the inner side of the groove. This result might be charged eithér to inadvertence or intention in the proportioning of the parts of the model; but we are satisfied that it is inherent in the form of the structure, unless there is an accuracy in construction and proportion between the various parts and the circumferential and radial tendencies which would be impracticable in structures of this class.

*697Nesbitt’s testimony is confirmatory. He says that what Yemiker showed him had the wedge-carrying ring, but the ring did not fit tight in the groove. Yemiker does -not dispute this in his testimony. He told Porter the purpose of the ring was to give the wedges strength and position them. In-his contemporary letter to his patent attorney he says that the ring acts as a protector to prevent the wedges from falling out, and that the core with the ring alone, and no wedges, would not be very durable. True, this same letter contains a sketch indicating a taper ring, and he says the taper ring will draw the sections inwardly. But this letter was not written until after he had seen the finished Nesbitt ring. He does not claim there was any taper on the sketch he made for Nesbitt, and this sketch sent to the attorney and the other ones that are said to be earlier all indicate at most that Yemiker had only an idea — and that an erroneous one — as to what would be the joint'action of his wedges and a taper ring.1

We are thus led by what is nearly, if not quite, a demonstration to the conclusion that the straight ring which Yemiker explained to Nesbitt did not contain the essence of the patented invention, but was only the loose ring of the old art, held in position by wedges projecting from its bottom through the bottom of the ring, instead of by bolts and nuts. This conclusion is destructive of the theory upon which the defense has been made, and upon which the decree below rests.

In the District Court it was argued that the Yemiker application of August, 1913, merely embodied the disclosure of January, 1913, and described and claimed the straight-sided, locking ring which would thus anticipate any claim thereto by Nesbitt. We do not so understand this application, which went to issue in 1914. It does not picture or describe any taper ring, thereby inferring a self-retaining function, nor does it indicate any self-retaining or locking capacity, save by the casual statement that the ring is “forced into” the groove— an ambiguous statement, but in its pertinent aspect a substantial impossibility, as above pointed out. It makes no claim to any locking ring, as such, but only to the wedges and the ring carrying them. Every claim is to the combination.

There is little reason to think that Yemiker even believed himself to be the inventor of the self-retaining locking ring, efficient without wedges, bolts, or contractile means. Though now reaching back to June, 1912, he filed an application in August, 1912, which did not disclose it. In October, 1912, he had a mold built, which did not disclose it. Learning in January, 1913, that Nesbitt was claiming it, he complained to his attorney that Nesbitt had taken his invention, but leaving off the wedges, and thought it would not be successful. The manufacturers of his old device, on a royalty, discontinued and *698began to make Nesbitt’s. Yemiker made no complaint, except informally to a shop foreman. Then, in April, 1913, he applied for patent on another form quite distinct. In August, 1913, he filed his application, above referred to, and still made no claim to the ring alone, though it was being sold for Nesbitt in.large numbers. Never until January, 1915, did he make any such claim (by the application which went into interference). This claim was too belated to be accepted, unless upon convincing proof, and the proof is not of that character.

Claim 5, sued upon, contains the express limitation to a wedge ring, and is not infringed. In the bill of complaint, claim 6, which does not seem to carry the limitation, was also sued upon, and there may have been confusion between 5 and 6. However, appellants’ brief does not now ask anything under claim 6. Claim 9, also sued upon, we should be inclined to construe as of the broader class; but it formed count 2 of the interference, it was held by the Court of Appeals of the District of Columbia to be limited to the wedge ring, the interference as to this count was decided on that theory, and it 'is apparently quite immaterial to plaintiff in this case whether claim 9 is or is not infringed. Neither injunction nor accounting can be thereby affected, since, if it is narrow, the claims upon which injunction and accounting will rest are broader.

As to claims 5 and 9, the decree is affirmed; as to claims 1 and 2, it is reversed, with directions to enter the usual decree for an injunction and accounting.

On Petition for Rehearing.

[2] Upon an application for rehearing it is urged that one who is beaten in an interference upon a specific issue, may not afterwards, by amendment or reissue, obtain broader claims which dominate the interference issue and thus in a measure reverse the result (In re Curtiss, 46 App. D. C. 183); and it is assumed that this is what Nesbitt did and that in the opinion we have approved this practice. Not so. Nesbitt was the successful party in the interference, and obviously the rule does not apply to that party. The concession and assumption in the interference that Yemiker had disclosed a straight-sided ring was not the thing adjudicated; it was a reason given for a distinction that was specifically important; and we have been convinced, in so far as the reasoning assumed that Yemiker was entitled to a patent on the straight ring, it was both immaterial and erroneous. He was not misled, to his prejudice. This reasoning first became definite with the Examiner’s decision in February, 1917. Yemiker could not then apply for this broader claim through a reissue of the patent of his August, 1913, application, issued in August, 1914, more than two years before; and he could not apply for it as for something not disclosed in the August, 1913, application, because it had been in public use since early in 1913.' No doubt there are cases where the conduct of a party and his positions taken in an interference litigation may estop him, beyond the operation of any technical rules and merely upon equitable principles; but this is not a case for permitting *699in estoppel to defeat a patent granted to the true inventor. We have concluded, as the opinion shows, that the substantial advance here made was the conception of an independent locking ring efficient by Itself and without bolts or wedges, that this involved inventive merit, that Nesbitt’s specifications and drawings permitted claims of that breadth, and that claims 1 and 2 should be so construed. Yemiker had full opportunity to make such a claim in the two applications which he filed, including the one in interference, and he did not.- The record is wholly unconvincing that he ever had that conception until long after Nesbitt’s application.

The application is denied.