The specification states that the "invention relates to buttons, more particularly those secured to the fabric or cloth by metallic fastenings, and provided with an open central bore, which adapts them for use, especially with spring.studs, while in the particular ‘button finish,’ so called, combined with the central bore, and in the general arrangement and disposition of the several parts with respect to each other, is embodied the subject of the inven*985tion.” Reference may be bad to the drawing of Mead’s device, infra, which is a reproduction on a larger scale of Fig. 1 of the patent: The other two drawings, Raymond’s and defendants’, will he referred to hereafter.
The specification proceeds:
“The button is composed of a slightly-dished lower disk, B, centrally perforated, * * * and which may be struck from thin metal of any suitable description. G represents a circular disk, also provided with a central bore, with the metal at or near the opening bent or burred, forming a short frustum *986of a cone, and when in place this disk or ‘anvil plate’ resis upon the lower (lisle. B» with the burred portion downward, while the two openings [in B and in Cl form a central bore. F represents an annular ring of filling to protect the upper metal cap, E, by re-enforcement.” “This upper metal cap, E, which is of convex shape, inclosing said disks, B and O, and filling, F, said cap, E, being also imperforate, and provided with a button finish, I consider a very important part of my invention, and, in connection with metal buttons of the class here-inbefore premised, has, X believe, never been used; that is, the metal buttons, employed in combination with spring studs, have generally been flat disks, provided with a central bore, and with the stud, when locked in position, projecting- slightly above or even with the surface of the top -of the button.” “The device is attached to the fabric by a fastening eyelet shown at “d” as an ordinary hollow eyelet, having an internal central bore sufficient to permit the burr or truncated cone formed on the anvil plate, C, to enter therein, and thus the latter spreads and rivets the top of the eyelet down upon the lower disk, B, of the button, and prevents its withdrawal. * * * For this reason [says the patentee] I have termed said disk or plate, C, the ‘anvil plate,’ since it acts like an anvil upon which to rivet or clinch the upper part of the shank of the fastening eyelet. Furthermore, the rough upset end of the eyelet, d, passing between the disk, B, and the plato, C, is completely concealed, and a smooth and nicely-finished surface is given to the interior periphery of the central bore, and no obstruction is ottered to the entrance or withdrawal of the stud.”
A glance at tbe drawing will show that the anvil plate has two functions, — a positive one, in that it acts as an anvil upon which the eyelet is struck up,' and a negative one, in that, having thus contributed to produce the fastening to the fabric, it is so contrived as to give clearance for the inserted stud, which is shown in dotted lines in the drawing. The specification concludes thus;
“When the article is completed, it is provided with a central bore, a,nd ready for immediate application to a glove or other article of wear by simply driving in the fastening eyelet.”
The-claim involved reads as follows:
“(2) A convex imperforate cap, inclosing the interior of a button, in combination with a disk to which the lower edges of said cap are attached, and which has a central opening, a second disk within said cap, and provided with a central tubular lip, which extends downward into the central opening of the first disk, and an eyelet for attaching the latter to the fabric, substantially as set forth.”
In August, 1889, this patent came before Judge Colt, holding circuit court in the district of Massachusetts, upon final hearing on pleadings and proofs, in a suit brought to enjoin a device shown in United States patent to one Raymond, No. 369,882, a drawing of w'hich is given supra. That suit seems to have been vigorously contested. The defendants introduced in evidence the file wrapper, and contents, examined experts, and put in 14 United States and 9 foreign patents. The case is reported as Kent v. Simons (C. C.) 39 Fed. 606, the relevant parts of the opinion reading:
“The invention relates to an improved construction in the button-hole member, and especially in the retention of the button finish in fasteners of this type. * * * The defendants insist that the Mead patent is for an improvement in ordinary metal buttons, and that its use with a spring stud is only one of the forms in which the button may be used. I cannot accept this construction of the patent. * * * The principal object of the Mead invention was the production of an improved button adapted for use with a spring stud. * * * Taking this view of the patent, that the invention of Mead is pri-*987mai'ily an improvement in metallic fasteners to be used with spring studs, I think, the patent is valid. There are many prior patents in this branch of the art, but as to this particular type of button I think Mead made a patentable improvement over prior buttons of this class. All the elements of the Mead fastener, separately considered, may be found in prior patents for ordinary buttons, or for fasteners composed of a metallic button member and metallic button-hole member, but the combination, as arranged by Mead, is new, and produces an improved result; that improvement consisting- largely in the convenient form of the central opening for receiving the spring- stud, while at the same time preserving the "button finish.’ The Mead improvement is manifestly of limited scope, in view of the many prior devices, but I do not think it was anticipated by anything found in those devices, and I believe its production, notwithstanding what preceded it in the art, involved the exercise of the inventive faculty.”
The Raymond device was held to be an infringement, as it quite plainly is. The lower plate, E, is altered in form, but not in function, and the filling, F, is omitted, but all the elements of the Mead second claim (which does not contain the filling) are found co-operating together in the same way to produce the same result. In a subsequent suit before the same judge, upon motion for preliminary injunction, it was held that a defendant who had no eyelet and no anvil plate did not infringe; the court setting forth with more fullness than in its earlier decision the functions of the “anvil plate,” and holding that “an anvil plate to rivet down the eyelet upon the lower disk is the essential feature in the Mead button.” Fastener Co. v. Weisner (C. C.) 90 Fed. 104. There are no other reported adjudications touching this patent, except the one now appealed from. Other suits were brought from time to time against alleged infringers, who made no defense.
In May, 1899, the present suit and one against Bradt and another were begun in the Northern district of New York to restrain manufacture and sale of glove fasteners made under a patent to Eugene Pringle, No. 000,114, March 1, 1898. The question of continuing an injunction pendente lite coming- on before Judge Coxe, he held, upon the record before him, that the validity of the claim as construed by Judge Colt was to be taken as established, and that the only question open for discussion was infringement. That question he decided against the defendants, who have taken this appeal. The various propositions arising upon such appeal may be separately considered.
1. Lt will be remembered that the socket of the Mead fastener was adapted for use with spring- studs. The parts surrounding the central bore are nonresilient. The resiliency which allowed the head of the stud to be compressed when entering- the socket, expanding after entrance, was shown in the stud itself. The second claim does not expressly refer to the spring stud. The same was true of the iiaymond fastening device. The resiliency resided in the stud member. The defendants here, however, have slitted the eyelet so as to render the socket member elastic, while their stud is nonelastic. Of this change Judge Ooxe said:
“This addition may or may not bo an improvement but it is entirely clear that the defendants cannot aijpropi-Iate the Mead invention by adding thereto a new function, which in no way changes the action of the patented, combina*988tion. It will hardly he pretended that after the decision in Kent v. Simons, supra, the defendants therein could have escaped infringement hy malting such a slit in the Kaymond button. Especially is this true in view of the fact that the Pringle patent, under which defendants manufacture, shows a slitted socket with a solid stud, and an unslitted socket with a spring stud, as alternative and equivalent constructions.”
But it would seem to be unnecessary to look for the expression of such an idea in a patent. The transposition of resiliency between stud and socket would surely suggest itself to a mind of ordinary intelligence. We agree with the circuit court in its conclusion. Such a trivial alteration, the fundamental structure of the socket member remaining unchanged, would not avoid infringement.
2. It is next contended that there is no infringement, because defendant’s anvil plate is not perforated nor “provided with a central tubular lip, which extends downwards into the central opening of the first disk.” It will be necessary to examine the defendant’s construction in order to appreciate the force of this contention. It will be seen that there is an outer metal cap, or “button finish,” E, and a bottom plate, B, which, like that in Raymond’s device, is of a different shape from Mead’s, but which performs all its functions. Upper and lower plate interlock, so as to make the complete casing of the button. The top plate is re-enforced by a plate of thin cheap stuff, F, which takes the place of Mead’s filling. The anvil plate, C, is an independent piece, interposed between the upper and .lower plates. Its details are shown in the following Fig. 7 of the Pringle patent:
When the various parts are assembled, the “clench-turning piece,” as it is called in that patent, presents the contour shown in the upper figure, with a depression shaped like a truncated cone extending down towards the opening in B. The hollow fastening eyelet, d, is introduced upwards through this opening. Its internal central bore is of a size sufficient to permit the entrance of the truncated cone on the lower part of the “clench-turning piece.” Pressure being applied, the latter spreads, and rivets the top of the eyelet down upon the lower disk, B, of the button, and prevents its withdrawal. The socket member has thus been assembled and affixed to the fabric, but a part of its central bore is filled by the depending metal of the clench-turning piece or anvil plate, so that the stud piece cannot be inserted to a depth sufficient to insure perfect engagement of the two members of the fastening device. In other words, an “obstruction is offered to the entrance of the stud.” If the metal forming the truncated cone-shaped depression were then cut away on the lines, x, x, we would have a very close copy of the Mead device, in *989which the side walls of the socket or central bore are composed of the anvil piece and the eyelet. The defendants do not cut away this metal, but drive it up into the dome of the button cap, giving to the anvil plate the contour shown in the lower figure, 7, and in the drawing of the completed button, supra, in which the side walls of the central bore are composed wholly of the eyelet, and the raised metal of the anvil plate forms a roof or dome over the inner end of the centra] bore. We are unable to perceive why the opening in the lower plate may not be fairly described as contributing to form a central bore, although that bore, when formed, is lined with the metal of the eyelet. Indeed, in the Mead patent, where it is so described, both the eyelet and the tubular lip of the anvil plate are interposed between the metal of I» and the spring stud. And vve fail to find any sound distinction between an opening which is secured by the entire removal of a portion of the metal and one which is secured by forcing such portion out of the way. When looked at on any plane which the stud passes through to reach the utmost limit of its insertion, the anvil plate, C, presents an opening or central bore. It would seem to make no difference that beyond that limit there is a dome-shaped roof integral with the metal of the anvil plate. The “central tubular lip” of C does not descend quite as far into the opening of the lower plate, B, as it does in Mead, but none the less the concentric central openings which are found within the limits of the planes intersected by the path of the stud together form a central bore; .the main difference between the two devices being that in defendants’ that central bore is wholly lined with the metal of the eyelet. As we have already seen, the clench turning plate of defendants performs the function of upsetting the eyelet in the same manner as does the anvil plate of Mead. The differences between the two are of form, not of function, and, under a construction of claim 2 which would find infringement in Baymond’s device, there is infringement in defendants’.
3. The next proposition advanced is that the state of the art does not warrant the construction given to claim 2 by Judge Colt, that the device of Mead is either anticipated or is devoid of patentable novelty. In appellant’s brief this branch of the case is argued at great length, with a detailed and exhaustive presentation of the state of the art as disclosed in some 30 patents, all of which are analyzed and illustrated by elaborate diagrams. The case has been presented as if this court were sitting in review of Judge Colt’s decision in Kent v. Simons. There seems to have been some misapprehension of the principles regulating' appeals of this character. Shortly af£er its organization this court had occasion to enunciate the proposition that, where there was an appeal from an order granting preliminary injunction, the court below having followed an adjudication at circuit sustaining the patent after contest upon pleadings and proofs, such adjudication is to have the same weight which it should have before the circuit court. “Appeals from orders,” we said, “are not to he confounded with appeals from final decrees, and the rule which we have thus stated will not prevent our review of the adjudication itself, whenever it and the record upon which it was made *990shall be presented upon appeal. The tendency of any different rule would be to produce confusion, and convert the review of the interlocutory order into a review of the final adjudication upon which it was founded.” American Paper Pail & Box Co. v. National Folding-Box & Paper Co., 2 C. C. A. 165, 51 Fed. 229. The propriety of some such rule is apparent. The prior adjudication has been secured by complainant on testimony which has been sifted by cross-examination, and it would seem that he should be entitled to cross-examination of defendants’ experts before being deprived of the advantage of his previous decree. Especially so in view of the ease with which ex parte affidavits of experts exploiting any theory desired may apparently be obtained. This practice has been ever since adhered to by this court. In Curtis v. Wheel Co., 7 C. C. A. 493, 58 Fed. 784, we discussed the merits, and reversed; but in that case “the whole question of patentable invention [was] presented upon the face of the Overman patent.” The patentee’s own description of the state of the art, as set forth in the specification, was accepted, but it was held that the improvement he made in that art did not rise to the dignity of invention. In the case at bar, if we take the art as the patentee states it, we are not constrained to a like conclusion. So, too, in Thomson-Houston Electric Co. v. Hoosick Ry. Co., 27 C. C. A. 420, 82 Fed. 462, we reversed upon the merits, using this language:
“The former adjudication was entitled to great weight upon the application for the preliminary injunction, and justified, although it did not necessarily control, the decision. If it had been founded upon evidence not before the court upon the application 'for the injunction, or not so fully before it, it could not have been intelligently considered by that court, and there would have been no record here upon which it could be re-examined. But the question whether two patents are for the same invention is a question which is to be determined by a comparison of the documents themselves. There may be cases in which it is necessary to resort to extrinsic evidence to ascertain the meaning and the true construction of the documents. The present was not such a case. The patents are unambiguous, and even the file wrappers, which are in the record, are <?f little value as extrinsic evidence.”
In Fastener Co. v. Littauer, 28 C. C. A. 133, 84 Fed. 164, we quoted from tbe Paper-Box Case, supra: “The adjudication upon which the motion for preliminary injunction was based, not being the subject of the appeal, is to have the same weight which .it should have before the circuit court, * * * in the absence of some controlling reason for disregarding it,” — and added:
“No such controlling reason is shown here. No prior patent, or prior use or psior publication, having an important bearing upon the validity or construction of the patent, and which was not before the court in the Columbian Company Case, is now presented; no new authority on patent law is now first cited; there is nothing to show an improvident exercise of legal discretion by the circuit judge; and apparently this is an effort to review the decision in the Columbian Case at final hearing upon a partial presentation.of the evidence then considered, and without, the cross-examination. There is no warrant for such practice, which was expressly condemned in American Paper-Pail & Box Co. v. National Folding-Box & Paper Co., supra.”
It has seemed desirable once more to restate this rale of practice, since the appeal here seems to have been mainly argued on the *991Iheory tliafc the decision of Judge Colt was to be reviewed, as if this were an appeal from that decision upon the record before him.
4. There is some testimony as to the use prior to 3885 of a variety of glove fastener known as the “Schloss Fastener,” and which is claimed to anticipate the Mead device. This is, of course, presented on ex parte affidavits, and it is sufficient to say that we find it unpersuasive. Testimony of prior use, produced after adjudication and unsifted by cross-examination, is always to be accepted with caution. It must be most clear and convincing to warrant what is substantially a reversal of an adjudication at final hearing. The same remarks apply to the attempt made to prove public use of the Mead fastener prior to 3882.
5. There have been several prior patents and drawings introduced in the case at bar which were not in the record before Judge Coit. These include the Schloss English and Hchloss Italian patents, the Hsvppev French patent, and United States patents to Mauvias, Atldns, Mead of 1880, Kraetzer of 1884, Jenk, Newton, and Schneller. The mere introduction of additional patents does not change the situation, unless they-indicate that the prior art knew of devices not indicated by 1he patents in proof before the judge who heard the cause at final hearing. The new patents do not indicate any material difference in the state of the art from that shown by the patents to Oapewell (Fig. 5), .Bates, Burtey, and others, which were before Judge Colt. The patent in suit is undoubtedly one of limited scope, its elements separately considered may be found in prior patents, and whether, in view of the state of the art, it exhibits invention is an arguable question; but the additional patents do not make such a change in the showing as to the state of that art as would warrant the judge who heard the application for preliminary injunction in disregarding the adjudication sustaining the patent at final hearing, and we cannot find that he erred in following that adjudication, and in considering only the question of infringement as fairly before him. The order appealed from is affirmed, with costs.