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

On the issue of infringement we think there should be no doubt. A reversal end for end is as readily accomplished by revolving the pad on a diameter through an angle of 180° as by rotating it in its own plane to a similar extent. Owing to the form of the defendant's pad, a lunette or segment, it was necessary to do this to effect the result of the patent; but, even were the specifications more expressly limited than they are, we should regard it as an equivalent of the patent. If the patent is to have any value whatever, it ought not to be possible to evade it so easily.

Nor do we think the supposed prior use of 1913, Exhibit L, is a defense. We of course recognize it as proved, but we think it did not embody the invention. That the pad could be reversed, and the position of the hump changed, is quite true. We agree, as well, that the mere failure of any one to reverse it would not save the patent. In this we cannot agree with the reasoning of the learned trial judge. But the shape of the wedge and pad forbade the idea that the pad should be put in blunt edge first. The insole and pad were sold and used in combination, and the shape of each made it clear that the pad should enter by its sharp end. If any one had reversed it, he would have perverted it from its plainly intended position; the design of the combination would have cried out against it. An earlier device which must be distorted from its obvious design cannot be an anticipation. United Shirt Collar Co. v. Beattie, 149 F. 736, 739, 79 C.C.A. 442 (C.C.A. 2). The pocket was, it is true, made ample enough to allow of lateral movement, but, while that suggested some accommodation, it was not of the necessary kind. For whatever it was worth the invention still remained in nubibus. The fact that the plaintiff during the trial extended his pretensions to cover such a device is not conclusive upon him. Inordinate claims create no estoppels.

The last and most difficult question relates to the date of the defendant's own invention of the infringing soles. The evidence is as follows: Marteau went into the employ of Nathan in November, 1918, and swore that already at that time he had invented the infringing pads, and had disclosed them to Nathan. In this Nathan corroborated him. He produced some drawings, undated, to show what Marteau had made. While some of these may in fact have been intended to show asymmetrical longitudinal pads, they are by no means clear in that respect. This may be due to difficulties of perspective, but the fact remains. They do not confirm his evidence; so far as we can see, they are quite consistent with a later invention of asymmetrical pads.

However, Nathan put in evidence a pad, longitudinally asymmetrical, which he said was the matrix for his molds and had also been made in November. Naturally it bore no date, and the time of its production rests solely on his testimony, in this detail not corroborated by Marteau. This model Nathan swore he sent to two rubber companies, who gave him bids for the manufacture of molds. The bid of the United States Rubber Company was dated December 31, 1918, that of the La Favorite Company, January 2, 1919. The second he accepted on January 9, 1919, a week before Block filed his application. It is clear that no molds were made before that date. These letters, though produced at the trial, were not offered in evidence, and we have no means of knowing how far they corroborate Nathan, or identify Exhibit W as in existence at their date. We can only suppose that they were not offered, because they showed nothing of consequence. Thus the evidence in the end comes only to this: That Nathan and Marteau, each highly interested witnesses, swear that Marteau had conceived of the asymmetrical pads in November, 1918, and had made the matrix from which eventually, apparently in April, they began to manufacture and sell.

When Revised Statutes, § 4886 (Comp. St. § 9430), prescribes that a patentable invention must not be "known or used" by others, it means something quite different from the prior public use by which it may be abandoned. In the case at bar Nathan clearly did not use the pads before January 16, 1919; indeed none were made. Were they "known" in the sense which the statutes uses the term? The word has acquired a somewhat esoteric meaning, imputed by the courts to accomplish the purpose of the statute. Mere acquaintance with the invention, even if disclosed, is not enough; nothing short of "reduction to practice" will do, whatever that may mean. Reed v. Cutter, 1 Story, 590, 599, Fed. Cas. No. 11,645; Coffin v. Ogden, 18 Wall. 120, 21 L. Ed. 821; Stitt v. Eastern R.R. Co. (C.C.) 22 F. 649; Imperial Brass Co. v. Nelson, 203 F. 484, 121 C.C.A. 606 (C.C.A. 7); Buser v. Novelty Co., 151 F. 478, 81 C.C.A. 16 (C.C.A. 6). All this is very old law, though it is by no means always easy to ascertain just when "reduction" has taken place, for the books, perhaps wisely, are reticent of positive definition. We are, however, relieved in the case at bar from deciding whether the earlier invention must always take tangible embodiment and be ready for immediate use; whether no description in words or figures will serve, unless they be in a public document. We think that Exhibit W, although a model made by hand, would have been under any rule adequate "reduction to practice," had the date of its production been certainly fixed. While it is called a "model," it was not truly such, but rather, as we have called it, a matrix from which the working molds were to be made. Taken by itself it was a perfect example of the invention, as complete and perfected as its replicas which were sold.

But, because of the reasons we have already given, it seems to us that the date of its production is not inalterably fixed. While we have no doubt that Marteau and Nathan were at work on their new insoles in November and December, 1918, it is quite impossible, from any documentary evidence, to say when they first made the asymmetrical pads. It is quite likely — indeed, we think it extremely probable — that they are right in saying that Exhibit W was in existence before January 16, 1919; but in such cases probability, even extreme probability, is not enough. The proof must be as absolute as in a criminal conviction; indeed, the rule comes nearly to this, that one must have contemporaneous records, verbal or structural. Furthermore, it is apparent that to introduce such records does not prove the date of their origin, except grossly, as they bear internal evidence of antiquity, or precisely, if they carry their own date upon them. When the issue concerns so short a time as that at bar, the appearance of the exhibit helps not a jot to fix the date of its production, and upon the relevant issue we are as much dependent upon oral testimony as though it had never been introduced at all. Thus we cannot regard Exhibit W as adding any more to the proofs than Exhibit V; it tells us that it was made, but not just when. That is all that is important here.

Decree affirmed.