ROSE, District Judge.

The plaintiff is the owner of three patents, one to Conradi, No. 1,129,456, February 23, 1915, and two to Kruse, namely, No. 1,081,050, December 9, 1913, and No. 1,244,056, October 23, 1917. It charges infringement of the eleventh claim of Conradi, of the second, third, fourth, and fifth of the earlier Kruse, and the second and fourth of the later. The defendant has purchased and has used the machine alleged to infringe. The case was defended by the manufacturer, the Torres Wold Company of Illinois.

The machine has at one end a hopper, into which a pile of tin blanks, cut to a proper size for making a can body, are put. Out of the other end comes the completed can. In the course of its travels, the blank is subjected to a number of processes. The pending controversy involves two of these.

[1] The eleventh claim of the Conradi patent, and the fourth and fifth of the earlier Kruse, relate to the construction of the hopper from which the machine takes the blanks. The second and third of the first Kruse patent, and the second and fourth of the other to the same inventor, concern themselves with the means for accurately positioning the blanks at the moment at which they are to be notched. The hopper and the positioning devices are found in the same machine, but each successively does its work precisely as it would if the other was altogether absent. They do not form a true combination. Grinnell Washing Machine Co. v. Johnson Co. (Supreme Court, June 10, 1918) 247 U. S. 426, 38 Sup. Ct. 547, 62 L. Ed. -. The claims relating to the hopper will therefore be considered apart from those dealing with the positioning devices.

*905Hopper Claims.

[2] Conradi s Ledges. — Conradi’s patent was issued subsequent to the first of the two to Kruse, but in an interference proceeding he was awarded priority for the claim here in suit. A seli-ieeding machine must take up one blank at a time, out of a bundle or pile in a hopper or other initial receptacle, and only one. The blanks must be so held or supported that they will not fall out of the hopper, aiid the support, must not obstruct the access of the feeding device to the blanks. Obviously both these ends will be attained, if the hopper .be fitted with shelves or plates projecting from its walls, or sides .inwardly towards each other, and towards its center, but which, because they do not meet, will leave, between their innermost edges, space in which the engaging element of the feeding device may operate. Any article, not absolutely rigid, if placed in a position intended to be substantially horizontal, will, although held up at its ends, sag down in its middle, if‘that be unsupported. Means of increasing this central sag or downward bend of the lowermost blank are provided in many machines, including those of the plaintiff and of the defendant. For the purpose of facilitating the engagement of that blank by the feeding device, Conradi thought that, if he could counteract the tendency of the pile as a whole to sag down, there would be much less danger of carrying forward two or more blanks at one time. He believed he could achieve this purpose if he put raised ledges on or near the inner extremity of his shelves or plates. In such a hopper, these portions of the ends of the blanks -which did not rest upon these ledges would not, throughout their whole extent, be in contact with the supporting plates, but would, on each side, bend or sag down from the ledges towards the hopper walls. In this way a cantilever effect would be produced, the result of which would diminish, if not eliminate, the downward sag of the altogether unsupported middle portion of the pile of blanks. It is in the light of this construction that the claim in suit must be read. Its vital element is:

“Means within the hopper for supporting the blanks at points spaced inwardly of the periphery of the blanks.”

These means are the raised ledges which he placed at the ends of his plates or shelves, or some fair equivalent for them. The defendant’s machine has nothing but flat shelves. There is neither a raised ledge nor a substitute for it. The plaintiff, however, contends that the edges of the plates fulfill the same function as Conradi’s ledges, although less perfectly. Conradi did not think so. The claim, if it means what plaintiff now says it does, would be invalid. It did not take the genius of an inventor to discover that a pile of blanks may be supported upon shelves. Conradi’s invention consisted, as he himself plainly declared, in giving to such shelves a peculiar construction, which he thought accomplished a useful purpose. For that he was entitled to a patent, but such construction is not found in defendant’s machine. It follows that the eleventh claim of the Conradi patent is valid, but that it is not infringed.

*906[3] Kruse’s Adjustable Plates. — '-The essential element of the fourth and fifth claims of the first Kruse patent are for:

“Supporting plates in the hopper adjustable on the hopper transversely towards and away from one another, whereby the amount of the sag of tha sheets between their supported ends may be regulated.”

The plates so described differ from those of the prior art, with Conradi's ledges or without them, only in that their inner- edges may be moved towards or away from one another, so as to regulate the extent to which the nonsupported middle portion of the blanks will sag. The year of„our Tord 1913 was rather late in date to discover for the first time that,- the shorter the length of an unsupported center, the less sag there will be. There might have been difficulty in introducing into the self-feeder means for making such adjustments. If so, invention might conceivably have been exercised in overcoming it, although, in view of the innumerable adjusting devices then known to the art, it seems improbable that it would have called for anything more than the aptness of the skilled mechanic. Into that question, however, it is unnecessary to go, for the claims in suit are not limited to any special character of adjusting means, and are for that reason invalid.

Positioning Devices.

[4-7] The machines put each blank through a number of operations. They must have within themselves such co-ordinating devices as will insure that, at the moment they are ready to perform each one of these tasks, the blanks shall be at the right spot. The difficulty of malting certain that it will be increases with the -speed and accuracy required. The call is always for faster machines, and of late years there has been a rapidly growing demand for “sanitary,” as distinguished from the older “hole and cap,” cans. In making the former, each blank must be notched at precisely the same relative position. If the notches are out of place even a few thousandths of an inch, the blank must be thrown aside. In his patents, Kruse provided means by which the blanks, just before they are notched, are moved back against fixed< stops. Absolute accuracy is thus secured, provided that the blanks* have been cut true, as in practice they usually are.

In his earlier patent, Kruse describes a machine in which gripping fingers.with a reciprocating action carry the blank forward, and on their return bring it back against stops. The second and third claims of that patent cover such a device. In his later patent he substituted, for a backward pull of the fingers, a push in the same direction, to be given by a- cam-faced bar or blade, so adjusted and shaped that at the proper moment it rises up and forces the blank back against the stops. The idea common to both is ingenious. To embody it in a working machine called for a number of adjustments, not necessarily obvious. To hit upon it and to reduce it to practice may well have been invention, always provided it was then new in the art. The defendant says that it was not, and that it was embodied in two machines built for the American Can Company, one in 1909 and the other in 1911. The filing date of the earlier Kruse patent was February, 1913. These machines are still in use. If they were originally built as they now are, they *907did in fact anticipate everything in this respect done by Kruse. The difference between the construction and operation of their positioning devices and those shown in the Kruse patents are no greater than between the latter and those found in defendant’s machine, if indeed they a,re as great. The plaintiff has properly stressed the strict measure of proof required of the setter-up of an anticipating device. It still remains true that a patent will be defeated by a prior use, proved beyond a reasonable doubt; that is to say, if the evidence leaves the trier of fact without doubt on the subject. The machines in quest ion were built at the American Can Company’s shops at Maywood, 111." in the years mentioned. So soon as they were, completed, they were, in accordance with that company’s custom, photographed. The negatives and copies of the photographs were preserved in the company’s files, and they have been here produced. The machines themselves were promptly sent to the company’s Richmond factory, and there they have ever since been continuously employed in the daily making of thousands of tobacco cans. I have no doubt that in point of fact the devices used in these machines anticipated the conception described in the Kruse patent. •

The plaintiff argues that, even if so much be found against it, the claims in suit are not thereby invalidated. It says that the American Can Company deliberately concealed or suppressed the knowledge of its invention, and thereby subordinated its claim, in accordance, with the general policy of the law, in the protection of the public interest, to that of another and bona fide inventor, who, during the period of inaction and concealment, gave the benefit of the discovery to the public. It cites the declaration of the Court of Appeals of the District of Columbia, that, viewed in the light of the true policy of the patent laws, the later inventor is in reality the first to invent, and is therefore entitled to the reward. Dieckman v. Brune, 37 App. D. C. 399. There is no doubt that an inventor who makes up his mind not to j latent his invention at all, or not to patent it until he thinks some one else is about to invade his monopoly, forfeits all right to a patent. Macbeth-Evans Glass Co. v. General Electric Co., 246 Fed. 695, 158 C. C. A. 651. To whom does the benefit of the forfeiture inure — the public or a subsequent inventor? The question cannot arise unless the earlier inventor has kept his invention absolutely concealed until after the conception of it by the later; otherwise, the second man would not be the inventor of something not before known or used in this country. We are not in this case concerned with an invention once made, and, even if to á limited extent reduced to practice, after-wards thrown aside, a.nd its úse altogether abandoned. In such cases, the Supreme Court has likened the reinvention to a rediscovery of an old art. Gayler v. Wilder, 10 How. 477, 13 L. Ed. 504. In many cases, in which are to be found expressions relied upon by the plaintiff, there was more or less doubt whether the invention in controversy therein had ever been perfected by him who was said to be the first to make it. The evidence left it doubtful whether it had ever amounted to anything more than an unsatisfactory and abandoned experiment. The fact that it was used but for a short time, if at all, *908and was then relegated to the scrap heap or lumber room, was suggestive, if not persuasive, that- it lacked something found in the second comer’s creation. It may not have been possible from the testimony to discover precisely what was missing, but it was not improbable that it was something which made all the difference between failure and success.

In the case at bar, the anticipating machines have, ever since they were built, been in daily extensive and successful commercial operation. Were they so designedly hidden from the public knowledge that ‘they cannot.be held to have anticipated the subsequent invention of Kruse? If there is such exception as plaintiff claims to the rule that no patent may validly issue for anything known or used in this country before its invention by the patentee, it is one which must be confined within narrow bounds. It will not be usually easy to make sure that the first inventor or user had been successful in keeping that, which he was doing dr using an absolute- secret. In spite of all his precautions, some hint, suggestion, or rumor of it may have come indirectly, if not directly, to the second Richmond. In the instant case, the testimony shows that the very draftsman who had prepared the designs for the anticipating machine subsequently left the can company’s employ and went into that of the plaintiff, and that, too, at a time prior to the filing of the earlier of the Kruse applications. There is no evidence that he or anybody else ever told Kruse anything about the can company’s machine. Nevertheless, if he had, there would have been slight chance of the testimony developing the fact. To find out the real truth in such matters is well-nigh impossible. The second inventor may have himself forgotten, or never clearly appreciated, what originally put the inventive idea into his head, and yet it may well have been something said or suggested by one who, all unknown to him, had some information as to the earlier device. Into such inquiries it will seldom be profitable to go. They will be unnecessary, if we hold fast to the letter of the statute, which forbids the patenting of anything before known or used by others in this country, admitting such exceptions to it only as are clearly recognized and precisely defined by decisions of courts of controlling authority. It is believed that, if any such exception as that contended for by the plaintiff has been so established (and it is not intended to intimate that it has been), it is limited, so far as concerns inventions which are clearly anticipations and have been in continuous use, to cases in which (a) the first inventor had deliberately made up his mind to practice his invention secretly, because he thought he could thereby make out of a monopoly in fact more than he could obtain from one legally given by the patent laws, and (b) in which there was proved to demonstration that his efforts to keep the invention secret had been absolutely successful.

In this case there is no evidence that the American Can Company made any special effort to hide this particular invention. It is abundantly shown that it was its practice to exclude the public from its factory, but none at all that this policy was dictated by a wish to keep the construction and operation of these particular machines hid*909dea. It did not pledge its employes to secrecy. In the course of nine years, many must have left its service, and have gone into that of others, as one of.them, as already stated, went into that of the plaintiff. The American Can Company is not a party to this litigation. No burden of proof is upon it. The issue as to this phase of the case is between the plaintiff and 1he public. If the former would succeed, it must establish every element necessary to bring its case within the limits of the exceptions to the statutory rule, if indeed that rule is subject to exception at all. This it has not done.

The second and third claims of the earlier Kruse patent, and the second and fourth of the later, must accordingly be held invalid, because they are for something before known and used in this country. Six of the seven claims in suit having been held invalid, and the remaining one not infringed, the hill must be dismissed.