after stating the case, delivered the opinion of the court.

In Stover Mfg. Co. v. Mast, Foos & Co., 32 C. C. A. 231, 89 Fed. 333, where, as here, the appeal was from an interlocutory order of injunction granted upon ex parte affidavits, and on the authority of a prior decision in another circuit, we treated the decision of the supreme court in Smith v. Iron Works, 165 U. S. 518, 17 Sup. Ct. 407, as meaning that the review in this court in such cases should go to the merits, and added the suggestion that: “This being the scope of the appeal, the logical inference would seem to be that every application to a circuit court for an injunction or temporary restraining order *682should be considered on its merits, and that a ruling or opinion of another court upon any question involved should be given only its just and reasonable weight according to the circumstances.” We therefore pass to the merits of this case.

It cannot be said that in a disputed case, or on a contested hearing, before the decision below was made, the patent in suit had been adjudged valid. In Blum v. Kerngood, 92 Fed. 992, the patent was denied the merit of a pioneer invention, and, on the narrow margin of novelty conceded to it, the finding and decree were that there had been no infringement. Upon the question of validity it was not necessary to decide definitely, and it is not shown that the decree entered contained anything upon the point. Whatever expression there is in the opinion on that question is only an assumption or concession that the patent was valid to the extent of the very narrow construction put upon it, and at most is entitled to the weight of a dictum. The gist of the opinion as understood below and as contended here is that the word “semicircular,” when used to describe the slot, meant a slot having a straight edge; and it is further contended — as it must be to make the proposition effective — -that the straight edge of the slot must be the edge of the body plate over which the two layers of cloth are to be extended and stitched. Promptly upon the handing down of that opinion, suit was brought against the Eagle Clasp Manufacturing Company in the same court, and that company, according to the affidavit of the appellee, “being satisfied from an examination of the patent and the opinion of Judge Morris in the Kerngood Case that its clasp was an infringement of the patent, and being unable to advance any other defense, etc., was powerless to prevent the issuance of a preliminary injunction against it, and the court * * * entered a decree that the hook manufactured by the Eagle Clasp Company was an infringement of the first claim of the patent.” The affidavit also says “there was no collusion” in the case, but it is evident that there was submission without contest, and that the judge gave the decree without further consideration of the question of the validity and scope of the patent. If the essential feature of the patent was that the edge of the plate next to the slot should be straight, and the slot was described as semicircular for the simple purpose of defining that line, —the form of the slot in other respects being immaterial, — the obvious suggestion is that a remarkably roundabout and questionable way was sought after to express a meaning which might have been easily declared in direct and plain words. It seems clear, however, that the word was used for no such indirect purpose. The file wrapper shows that, as first presented, neither the specification nor claim called for a slot of particular form, but simply for a hook with a slot; but by the same words now found in it the specification required that the center of the hook be cut away to form a slot through which the cloth could be “stitched' substantially along its entire edge,” and it follows that the amendment whereby the slot was described and claimed as semicircular could not have been studied out for the purpose of describing a straight edge for the body plate along which the overlapping cloth could be stitched. The idea of overlapping *683and stitching the cloth, so as to gain all the advantages now deemed possible, was already clearly expressed in the specification; and, a straight edge being the simplest form for that purpose, it is to be presumed, in the absence of evidence on the point, that the original drawings showred a slot with an edge of that kind at its base.

But, while the specification and drawings of the patent indicate clearly enough the intention that the straight side of the semicircular slot should be next to the body of the plate, it is not explicitly so stated, and manifestly that position is not essential to the successful accomplishment of the avowed purpose of the invention. An exact semicircle of course must have a straight side equal to the diameter of the corresponding circle, but, for all that is said in the. specification, that side may be the top as well as the base of the slot, and in either place is equally within the terms of the claim, unless, indeed, by a construction equivalent to a denial of invention, the claim is to be limited to the exact forms of plate and hook shown in the drawing; and, if that be done, the charge of infringement has no founda tion. That the straight edge is not necessary to permit thé stitching of the edges of the cloth inside of the loop, and for the purpose specified, it is easy to see. That could be done where, as in Blum v. Kerngood, the slot is elliptical, even though it extends into the body of the metal as far as into the hook, by stitching not directly across, as in that case it seems to have been thought necessary to do, but along the curved edge of the plate. So done, the stitching, without doubt, would furnish an adequate “bearing to resist the pull.” Especially would this be so if the lower end of the slot be placed, as in one of the illustrative cuts, more nearly in line with the upper edges of the ears on either side of the hook; , and in proportion as the elliptical curve is made, as it might be, to approximate a straight line, the resistance afforded by the stitching will be more nearly the same as if the line were straight. And since the Kerngood device, though manifestly performing, or capable of performing, in an appreciative degree, the function of the patented device, yet did not infringe, because the lower- end of the slot was elliptical and cut into the body plate below the position of the straight line of the patent as much as above that line, is it to be inferred that a change in the position and in the radius of the ellipse, so that the lower part thereof would be more nearly coincident with a straight line, would have been held to constitute an infringement? If so, by what rule is it to be determined when a noninfringing curved line, though performing from the start the requisite function, becomes, by change of position and of degree of curvature, an infringement of a claim for a straight line? The evident impossibility of such a rule demonstrates the fallacy of an attempt to state a substantial distinction between the Kerngood device and that of the patent in respect to the possibility of stitching the two layers of cloth along the lower edges of the two forms of slot in a manner to afford effectual resistance to the pull, and to subserve the other intended functions. The difference, in respect to any of the functions, is only in degree, and does not affect the essential character of the devices. It is evident, too, that fasteners might be constructed in *684various forms so as to have the so-called semicircular slot, but with body plates having other than straight edges within the slots, and yet available for all the supposed functions of that edge. As illustrated in one of the foregoing drawings, ffhe edge might be notched or undulating; and the other forms illustrated afford examples any one of which, in respect to the particular functions in question, is the plain equivalent of the form of the patent, and in all reason should be said to infringe, if the patent had real merit. It is to be •observed, further, that, while the word “semicircular” is one of definite mathematical meaning, it manifestly was not used in this patent in the strict sense. The slot shown in the drawings does not approximate a semicircle, unless reference is made only to the portion of the slot within that part of the loop which is turned back upon the plate; and, if that be the reference, it does not include the straight edge at the base, which alone has been treated as important. The reference, however, it seems clear, is to the entire slot, and that, instead of being semicircular, has a straight depth between parallel lines well-nigh equal to twice the radius of the semicircular top with which it is crowned. The base of the true semicircle, in the drawing of the patent; as well as in the other cuts shown, we have indicated by a dotted line, and there is nothing in the specification and claim which requires that the portion of the slot outside of the semicircular part shall be of any particular shape. As illustrated in the patent, it is a parallelogram, but it may be in any of the forms illustrated, and in many other conceivable shapes, without impairing the efficiency of the fastener in any respect deemed important. By the opinion in Blum v. Kerngood no one of those designs could be deemed an infringement, but, if the patent had genuine merit, it ought to cover them all. Our conclusion is that it contains nothing essentially new. The straight edge is fully anticipated in the Delpy clasp, and if it could, in any stage of the art, have been an inventive act to add to that device projecting ears, the anticipation is found in the Ewig patent of 1887, as well as in other earlier patents. There is certainly no merit in so bending a loop as to leave curved shoulders projecting beyond the body of the plate. It would be difficult to do the bending in a way to leave the necessary space between the plate and the loop, without producing the projecting shoulders. The prior art seems to admit of no theory on which the patent can be deemed valid. The decree below is therefore reversed.

Judge SHOWALTER did not participate in this decision.