Corrugated Paper Patents Co. v. Paper Working Mach. Co.

237 F. 380 | S.D.N.Y. | 1916

LEARNED HAND, District Judge.

Four points are raised in this case; (1) Jurisdiction; (2) validity; (3) infringement; (4) plaintiffs inequitable conduct.

[1] I do not believe that Judge Lacombe, in Streat v. American Rubber Co. (C. C.) 115 Fed. 634, meant to hold that, although the point of jurisdiction is taken after general appearance, nevertheless it is good. Judge Beatty, in U. S., etc., Co. v. Phœnix, etc., Co. (C. C.) 124 Fed. 234, suggests with much justice that in that case the only suggested infringement anyway was in the Southern district of New York, and that therefore, when Judge Lacombe found that the defendant was not responsible for that supposed infringement, he could do nothing but dismiss the bill. I attach no significance to the circumstance that it is said to have been dismissed for lack of jurisdiction. If there had been an acknowledged sale, outside of the Southern District of New York, and if, after general appearance, Judge Lacombe had held that that sale was insufficient jurisdictionally, that would have raised the point. While I can find no exact authority under the section in question except that of Judge Beatty, yet it seems to me that the question must be controlled by the general principle with which we are all so familiar when the jurisdiction depends upon diverse citizenship. There it is so well established as to need no citation that, when the controversy is one over which the Constitution gives us jurisdiction as to subject-matter, the provisions regulating the district in which the suit may be brought are for the protection only of the defendant, and that he must assert his right to dismiss the bill for lack of jurisdiction by motion to quash, if it appears on the bill, or by special plea, if it does not. Nobody can think it reasonable that the defendant should let the cause proceed to trial, reserving his right to assert that he ought never to have been haled into this particular court, because it imposed upon him the hardship of fighting his case away from home. Of course, if the court has no power to hear the case at all, no consent will serve; but concededly this court has that power.

[2] The only real question as to the validity of this patent arises under Smith, 457,676, granted in 1891. Ferres, 746,807, granted 1903, is not suggested as an anticipation of more than the tension devices, which I think it is. As to Smith,' I think: First, that even if only a new use, the patent in suit is a good invention for the reasons I shall show; but that, second, it is not a new use, because the disclosure was not adapted to the work of making corrugated paper without changes involving a very radical difference in purpose, if not in structure. As to the first point, we are all familiar, I believe, at least since Potts v. Creager, 155 U. S. 597, 15 Sup. Ct. 194, 39 L. Ed. 275, with the idea that the rule regarding new use depends upon what kind of new use may be in question. This, indeed, the Supreme Court has recognized even when holding that a patent was without invention. Mast, Foos & Co. v. Stover Mfg. Co., 177 U. S. 485, 20 Sup. Ct. 708, 44 L. Ed. 856. While there must, of course, be physically a new combination of elements to support a patent, as in all combinations, each element is simply put to a new use, and that may require the highest originality. Now, the facts here are very strong. This art was at least eight years old when Langston filed his application, because already Ferres in 1899 had devised a highly complicated machine to do just this work. I think it within the reasonable range of inference that such a completely differentiated machine does not appear till an art is somewhat mature. Therefore for those eight years and longer the need had existed, and Ferres represented its supply. Then came Langston, and Langston has had a very substantial success, whether or not his results are better than Ferres’. Is there no reason to suppose that, if it were ail obvious thing to use Smith for the purpose, some one during those eight years would have done so ? Especially, if we are permitted to assume the art was older, the argument is stronger. It is true that we must by a fiction suppose that Langston had Smith before him, though not suggested for the purpose, and only in the Sense that the whole of kindred arts were before him. Whether the process is one of ironing or not, certainly the purpose is different from ironing, and, as I believe, there is no easy road of suggestion between the two, for Langston did not have to smooth out a fabric; he had only to keep it under yielding pressure while the paste set. That one should have looked to the ironing art to solve that problem does not seem obvious; and, if it did, I should hesitate so to hold in the face of proof that it did not seem obvious to any one else for eight years.

But I believe that, even if it be thought an obvious step to go to such a machine, much readjustment was necessary, and that, too, with an underlying purpose which Smith did nqt supply. It will be noticed in that patent that the pressing and drying are alternate, not simultaneous. While the web is under the rollers, FF', etc., it is not exposed to heat, except in so far, of course, as the plate may be heated by conduction. The steam pipes are withdrawn from that part of the plate. It is only when over the parts, a°, between the rollers, that the web is dried. Moreover, there is no reason at all to suppose that the apron has any, pressure of itself, for it is kept taut by the feed-roller, I, and, if taut, its weight does not freely fall upon the plate over the spaces,.a°, a°, a°.' In this machine, therefore, we have only an alternate presser or roller and a drier, the web being kept moving and held from flying upwards by the taut apron, but not by the weight of the apron while the drier is at work. I can see in this no suggestion of Langston. Furthermore, I think it quite clear that to suppose the rollers are in the least analogous to the rollers added commercially to Langston is to misconceive their function, even if they have a vertical play, which is nowhere indicated. No one would think of this machine as capable of doing the work, who had not already conceived of a depending belt whose mere weight would hold the web in yielding pressure. Then he would only think of it to see that it might be reformed under the guidance of the constructive insight of an inventor into what the inventor had already discovered.

As to infringement, I shall certainly not decide now whether the Raffel patent infringes. The plaintiff had the opportunity of allowing the proof in at the trial of the defendant’s other structures, and refused. Having limited the case strictly, it is certainly bound to stand on its proof. On the other hand, I shall not hear the defendant suggest that the Baltimore machine may originally have been constructed under the Raffel patent and then changed. It could have told us all about the machine when sold, and it refused; therefore, I shall take the plaintiff’s proof as true. I attach no importance to the word “resilient,” which, though perhaps not the best possible word, is perfectly clear. Nor can I see that the defendant’s belt is not in the same sense resilient. Certainly it does not hold the web at a rigid and inflexible distance from the plate, like Ferres’ plates. It may be that the accommodation to variation in the height of the corrugations is less close then the plaintiff’s—I should think so—but certainly even as between two adjacent plates some change in angle is permissible and so some accommodation to variations in the corrugations. When one compares the defendant’s machine with Ferres’, the prior art, one at once sees that it has borrowed frankly from the plaintiff.

I shall not allow infringement of claim 6, however, as I think the last words refer to the lower belt. As to claim 4, I have considerable doubt whether the proof shows that the lower web is under tension in the defendant’s machine. Peterson does not indicate on his drawing, and apparently paid little attention to it; but he does say it came from a drum whose inertia at least and journal friction created some tension. Here, too, moreover, the defendant, having a complete disclosure within its own power, cannot too nicely scrutinize the plaintiff’s proofs. Claims 1, 2, 3, 4, and 5 are infringed and valid.

[3] The last question is of the collateral agreement of Langston not to engage in the business of making or selling corrugated paper machines or their product, while the patent lasts. I shall not decide whether such a covenant accompanying the sale of a patent is or is not too broad to be valid. Gamewell Fire Alarm Tel. Co. v. Crane, 160 Mass. 50, 35 N. E. 98, 22 L. R. A. 673, 39 Am. St. Rep. 458. Rather 1 shall assume that these covenants are illegal, in that they go farther than is reasonably necessary to protect the grant—a matter I have much doubt about. Even so, it must be conceded that the plaintiff could sue at law for infringement. Such an action does not involve the illegal agreement at all. The present theory is that any one so tainted as to take such an agreement from the patentee should be refused all equitable relief. Now, I do not understand that the so-called doctrine of unclean hands means that you may not have recourse to equitable remedies because in the past you have even committed a crime, dr shown yourself otherwise immoral. A court of equity will refuse to enforce any part of an illegal contract, but it will -not refuse to give the usual remedies to property which has been in the past acquired by-means which involved the violation of law. Sharp v. Taylor, 2 Phil. Ch. 801; Harvey v. Varney, 98 Mass. 118; Heath v. Van Colt, 9 Wis. 516; Trice v. Comstock, 121 Fed. 620, 57 C. C. A. 646, 61 L. R. A. 176; Ownes v. Ownes, 23 N. J. Eq. 60; American Association, Ltd., v. Innis, 109 Ky. 595, 60 S. W. 388; Pitzele v. Cohn, 217 Ill. 30, 75 N. E. 392. The nearest cases to the case at bar are those in which it was alleged that the plaintiff had formed a combination of patents which was a monopoly, but in most of these the usual equitable relief was given. Strait v. National Harrow Co. (C. C.) 51 Fed. 819; American Soda-Fountain Co. v. Green (C. C.) 69 Fed. 333; Columbia Wire Co. v. Freeman Wire Co. (C. C.) 71 Fed. 302; General Electric Co. v. Wise (C. C.) 119 Fed. 922. Judge Baker appears to have thought otherwise in National Harrow Co. v. Quick, (C. C.) 67 Fed. 130, but that is the only case I have found. However that may be, the Circuit Court of Appeals authoritatively settled for this circuit, in Edison Electric Light Co. v. Sawyer-Man Electric Co., 53 Fed. 592, 598, 3 C. C. A. 605, that, in a suit for infringement the plaintiff combination would not be investigated to ascertain whether or not it violated the law in this respect.

Moreover, it should be remembered that the supposed illegality of this contract is much more remote than if the plaintiff were engaged in a trust or combination. In those cases it was always urged that the suit was in aid of the very trust itself; that an injunction secured the very restraint of trade which the whole combination was designed to procure. There was force in that contention. Here, however, the injunction has not the remotest relation to the plaintiff’s contract with Langston. If that contract is void, part of the plaintiff’s consideration for its payment of $40,000 has failed and it has only got its patent. I cannot see upon what theory so much of the consideration as remains should be denied protection, except it be that any one who has once made such a contract has become such an outlaw as not to be entitled to come into any court of equity for any relief of any kind. If such a rule is to obtain, then the assignee of a mortgage could not foreclose if he took an illegal collateral covenant, nor the grantee of an interest in land have partition, nor could the grantee of a reversion stop waste.

Let the usual decree pass; no costs.

June 5, 1914.

[4] The only question I shall consider is whether the Lacaux patent, assuming that it is an anticipation, antedates-the date of actual invention of Langston’s patent; but I do not mean that I have decided the Lacaux patent a good anticipation. That question I pass over. The effective date of the Lacaux patent is July 10, 1906, nearly a year before the date of Langston’s application, June 27, 1907. No one disputes that if Langston, before July 10, 1906, disclosed his invention to any other person, Lacaux is not an anticipation, provided that the disclosure was made with enough particularity to enable the person in question to understand it and to describe it. Moreover, it is settled that this fact must be established by a high degree of proof (Clark Thread Co. v. Willimantic Lumber Co., 140 U. S. 481, 11 Sup. Ct. 846, 35 L. Ed. 521), “with equal certainty” to the degree of conviction necessary for the proof of prior use. Brooks v. Sacks, 81 Fed. 403, 26 C. C. A. 456, says that the rule requires proof “beyond a reasonable doubt,” and I shall accept that canon.

In most of those cases in which the patentees have failed it has been because of the absence of some contemporaneous document inevitably fixing the datq of the invention; courts have been very.unwilling to trust in such matters merely to oral testimony, especially to that of the inventor. I do not regard the case of Westinghouse Electric & Mfg. Co. v. Catskill Ill. & P. Co., 121 Fed. 831, 58 C. C. A. 167, as an authority against Langston, because the court especially emphasized the absence of the inventor Tesla’s testimony, and of any document by which to fix Page’s dates. Subsequently, upon fuller proofs, the Circuit Court of Appeals for the First Circuit, in Westinghouse El. & Mfg. Co. v. Stanley Instrument Co., 133 Fed. 167, 68 C. C. A. 523, found the prior publication was successfully antedated, as had District Judges in Westinghouse Electric & Mfg. Co. v. Mutual Life Ins. Co. (C. C.) 129 Fed. 213, Same v. Roberts (C. C.) 125 Fed. 6, and Same v. Jefferson Electric Light, Heat & Power Co. (C. C.) 128 Fed. 751.

In the case at bar we start with two dates of unquestionable authenticity, the contract with Howell on June 19, 1906, and the offer of a similar contract made to MacMullen on May 7, 1906. Nobody can reasonably question that Langston at this time had already completely conceived and elaborated what he supposed to be a practical automatic machine for double-facing single-faced corrugated paper, because he could not otherwise have undertaken its manufacture or guaranteed its performance. Nor can any one question that on May 10, 1906, he had asked quotations of J. H. Jolley & Co., upon six hard-rolled yellow metal plates which were of the same size as the heating plates eventually put into the two machines and which were unquestionably ordered for them. We know that the stationary heater afterwards furnished was therefore definitely decided on and its exact size already known. Mac-Mullen did not close the offer of May 7th till the 25th day of July, 15 days after the critical date here in question; but when closed it followed the offer made on May 7th. Two days after MacMullen closed, Langston ordered the yellow metal plates, about which he had inquired on May 10th. As the dimensions are the same, we know that there was no change in plan as to the stationary heater meanwhile. Of these plates, four were billed to him on October 6th, and two later, while the machiñes were being made. The belting he ordered also on July 27th, and we have the bills. This I shall consider later. Furthermore, the correspondence during the autumn of 1906 shows that the machines were actually being built at the'time, and that the buyers were constantly pushing Langston for delivery, which was made in one case at the end of the year and in the other during the early part of 1907.

There remains, therefore, only this question: To what extent did Langston at the time he made the offer to MacMullen, or the contract with Howell, disclose the machines which he delivered some eight months after, and were they the same as the patent in suit ? I quite agree that the undisclosed conception of the invention alone would not be enough. The case must depend upon some talk, or diagram, or sketch, or the like. That Langston had a talk with MacMullen on May 7, 1906, appears beyond contradiction from the letter written by him that day when he got back to his office. That he also had a -talk with Howell is also proved by the contract of June 19th; but I disregard the latter, since he does not rely on it. If in the talk with MacMullen he in fact described those elements of the machine which are the essentials of the patent in suit, he made an adequate disclosure.

As to the details of what Langston described, they are in agreement. Langston’s version was that he drew a sketch at the time showing that they would use a flat heating surface with belts to carry the web over it and a carrier at the end of it to pull it through. He did not go into the cutting mechanism, as that had already been built for a single-faced paper machine, but he described definitely how his machine worked, and what he described to him was identical with the machines which he is building now; that he had never departed from that first conception, except in the matter of changing the brass plates to iron steam chests. As to the talk with Howell, he was dead, and his clerk, who was present when the contract was signed, Langston had not been able to find. He went over the matter with Howell, but his recollection was not as clear about what he told him, as it was about what he told MacMullen, who was a trained mechanic, or the extent to which .he described the machine.

MacMullen testified that he remembered the conversation of May 7th in which Langston quoted the price on the machine. He could not place it independently of the letter, but the letter confirmed his impression that it was in the early spring of that year. He'was under the impression that Langston drew some sketches of the machine in the course of the discussion, which sketches, however, he did not keep. Langston told him that he had a machine'which was different from any one’s else, in that it ironed the paper; explained the use of the belts, and why they were put in front of what wás then intended to be a brass bed-plate; showed him how the belts operated, by means of which the paper was carried over the brass plates at a certain speed and then cut off in various lengths; and said that he would use the old gluing apparatus on the hand machine. When the machine was delivered, it corresponded with what Langston had told him on that day and with Fig. 1 of the Langston patent, except as to the heating arrangements. On cross-examination he described what the sketch showed, as follows: A top belt traveling endlessly over two pulleys,' the heater directly underneath, and directly in front of the heater two pulleys with an endless belt running around them.

I count for little the discrepancy between these witnesses as to where this talk took place. It was certainly about the time when MacMullen moved his shop, but they might well differ as to just when. It is inherently most improbable, assuming that any such talk ever took place, that it should be at any other time than when Langston made his offer, which we know to be May 7th. Then was the time to explain what he proposed, to deliver. .This strongly corroborates the evidence of the two men, though it does not, of course, pnake it like a proposition in Euclid, an unreasonable degree of proof to ask.

We know from photographs that Howell’s machine was in fact precisely like the patent in suit, except for the matter of the steam chests, and that it was well under way in the autumn. Rowand corroborates Langston in saying that there was no change in design after they began.

What, then, have we by unquestioned proof, and how much does it corroborate the oral story? We know that as early as January Mac-Mullen was dissatisfied and wanted a different machine from the hand gluer. We know that by May 7th Langston had in a talk described to MacMullen an automatic machine, and had worked it out enough to make a guaranteed offer. We know that he had fixed the details of the stationary heater by May 10th, as they eventually were made, and that he made a second guaranteed contract on June 19th. We know that, two days after MacMullen closed with him, he' ordered the heater plates and the belting, and that from then on he was at work building the machine.

While we have it only by word of mouth, we may, I think, from antecedent probability, accept without reservation Langston’s statement that he had before him Ferres, which he meant to avoid in whatever he did, and of course we know by the result what differentiation he in fact made from Ferres. With this assumption we may infer, I think, what elements of his machine were nec'essarily in his mind before he offered to make a contract. What were the possible variations from Ferres at once consistent with the eventual result which we know, and also with patentable differentiation? That he retained Ferres’ joint belts in the front of the machine is'too obvious for coinment. That he kept Ferres’ lower stationary" heater and omitted his upper heater appears from his inquiry for quotations on May 10th and from the necessity .of that omission to escape infringement. What was there left for patentability but the very element upon which I have found patentability to rest, the holding of the web upon the lower stationary heater by making the upper belt run back to the end of the machine? If it be suggested that he might simply have taken off Ferres’ upper stationary heater and let the two belts in front pull the web across the lower heater, the answer is that to let the web run over the heater without any pressure whatever was so obviously 'a futile method that we may assume Langston did not entertain it. Moreover, we have an interesting documentary confirmation of this at least as early as July 27th, in the belting which he ordered. That consisted of 160 feet of 7-inch belting and 160 feet of 16-inch, 80 feet for each machine. The upper belts were three in number, two strands of 16-inch and one of 7-inch. The plates were 15 feet long, and the lower belts, one-third the length, “approximately 6 feet,’’-making 20 feet or 21 feet in all. It therefore took for the upper belts of one machine 80 feet or 84 feet of 16-inch belting and 40 or 42 feet of 7-inch, or, for two machines, 160 feet or 168 feet, and 80 feet or 84 feet respectively. If we assume for the lower belts, of which we do not know the structure, three strands of 7-inch belting,¹ we have 30 feet or 36 feet for each machine, or 60 feet or 72 feet for both. Thus we find that the proper belting was 140 feet or 156 feet of 7-inch belting and 160 feet or 168 feet of 16-inch, approximately the amount ordered of each. If, on the other hand, we assume that the upper belts originally did not extend back over the plates, the amount of belting was more than four times that needed. Indeed, it is incredible that Langston should have ordered anything like what he did, if at that time he had not thoroughly understood that the upper belt did cover the stationary heater. Of course, theoretically it remains possible that he should have thought out this structure between May 7th and July 27th, but I think I have shown that it is very hard, knowing what we do of the way Langston was working, to suppose that, when once he had in mind any completed machine which was on the path from Ferres towards the completed machine he made, it should have been any other than the final product.

The chief attack of the defendant is against the disclosure on the theory that a brass-plate heater is not a stationary heater. Surely this is a most meticulous objection at this stage of the case. I am now in no position to say that it was impossible theoretically for brass plates with steam ■ pipes under them to' work as a heater; not. the slightest suggestion of the sort was made at the hearing and Langston stated that they worked perfectly well, except that the paper adhered to them. To allow such an issue now to be raised would be extremely unjust and purely fanciful. Obviously such plates would constitute some kind of stationary heater, and if the defendant meant to insist that it was unpractical or ineffective some indication of that position must appear in the evidence. There is not the slightest such; all the proof on the subject is quite to the contrary. Both Langston and Rowand say that the MacMullen machine ran successfully fot a time before Howell’s machine was delivered.

Finally, the defendant attacks Langston’s honesty. I am sorry for it, because my own impression was so strongly the opposite that I cannot help thinking the attack to be unwarranted. He in no sense attempted to suppress the change from brass plates to a steam chest, which nobody at the trial thought of any consequence. It was perfectly apparent all along, and he himself brought it out at once, that the original structure was of brass plates, and, as for the subsequent steam chest, it was there in the patent to speak for itself. As ten his seeing the brass plates at Vincennes in 1912, he may be right and Rowand wrong (I should incline to accept his statement); but, right or wrong, there is not the slightest ground to attribute the difference to dishonesty, because, as I have said, nobody at the trial suggested that the change was significant or that the original machines were inoperative. His difference with MacMullen as to the place where the talk took place operated to his own disadvantage. On the other hand, so far as one is justified dn judging of a witness’ honesty from his bearing, Langston appeared to me an exceptionally straight and honorable man. I agree that he had an interest, which ought to be kept in mind—though a somewhat remote one—but his answers were fair, he made no effort to state a clear recollection when he did not have it, and he did not evade cross-examination. My personal judgment was strong in favor of his honesty; so far as I could see, there was every reason not to impugn it.

I therefore conclude that Langston did on May 7th disclose to MacMullen all the essentials of the machine afterwards patented, and that he at once proceeded, after July 25th, to make both machines with as much diligence as could reasonably be expected of him. There is, of course, always room for spme doubt in these cases. Absolute certainty is out of the question, but I do not think it extravagant to ¿ay that many men have gone to prison, and indeed have been executed, upon no stronger proof than has been made in this case. I will therefore not change the ruling I made before this patent was introduced, and the same decree may be entered as w^.s entered upon the original hearing.

December 29, 1914.

This cause comes up upon notice of motion to compel the defendant to make and file with the master a statement showing the number of so-called double-facing machines, including and substantially like the machine sold by the Baltimore Paper Box Company, made and sold’ by it, and also the number of such machines substantially like the machine shown in the Raffel patent and in a certain cut or advertisement on pages 4 and 5 of the American Box Maker of August, 1914.

An interlocutory decree had been entered herein adjudicating the patent valid and infringed by a machine known as the Baltimore machine and directing the usual reference for an accounting. The master, on October 14, 1914; passed an order directing the defendant to file a statement in writing to show the extent of its infringements and what profits it had made, and, in particular, to set forth the number of so-called “cellular board double-facing or double-backing machines made dr sold by the defendant” and the profits made by them, and also to attend with its books and papers and by its president to be examined. The defendant, on the advice of its counsel, filed a paper on the return day, the 16th day of October, 1914, setting forth: First, that it had never infringed the patent; second, that it had never made any “so-called cellular board double-facing or double-backing machines” and had never made any profits thereon; third, that it had no books or papers bearing upon such a sale, or any evidence that it ever had so infringed; fourth, that it produced a letter from its attorney stating his advice not to answer otherwise, and a copy of the details of the-Baltimore Paper Box Company’s machine in question. The plaintiff did not ask the master to certify defendant for contempt, but at once applied by this motion.

[5] Having proved one infringement, the plaintiff is entitled to a sworn statement from the defendant as to what other infringements he has committed and the profits on each. The defendant files a statement under rule 63 that he has committed no other infringements, particularly that he has-made no “so-called cellular board machines.” It is not credible that the defendant should not have known what the phrase really meant, and I cannot accept his explanation as ingenuous ; but I do not think the question is here material. I think it will serve no-useful purpose at this stage to require any further statement from the defendant until the facts be ascertained regarding the structure of the machines which it makes and .their number. Obviously better progress can be made by an examination viva voce than by further statements. Rule 63 provides for a statement, and the defendant has made a statement denying all infringement. The rule next provides that any party, if dissatisfied, may proceed to examine the other viva voce. It is this examination which must proceed, and it is the scope of this which is the real controversy. Is the plaintiff to be confined to the single Baltimore machine, or may the examination include all machines made by the defendant according to the Baltimore machine, the Raffel patent, or the cut in the American Box Maker?

Both parties insist that the defendant makes and has made but one kind of machine. The defendant says that the Baltimore machine was misdescribed in the suit, and, conceding that the decree is an estoppel as to that single machine, that the plaintiff cannot prove an infringement as to any other. The case is therefore not one where the defendant has changed his infringement after bill filed; it is not even one where there were two forms at the outset. Yet in this circuit it is well settled that the reference is the place to determine the extent of the infringement even in such- cases. Welling v. La Bau (C. C.) 32 Fed. 293; Westinghouse Air Brake Co. v. Christensen Engineering Co. (C. C.) 126 Fed. 764; Brown Bag-Filling Mach. Co. v. Drohen (C. C.) 171 Fed. 438. A fortiori it is the place to determine the extent of the infringement where there is but one type. It was certainly not the duty of the plaintiff to present on the hearing all the machines as infringements, when the defendant tendered that issue; one infringement was enough for a decree and all that the plaintiff may have perhaps been prepared to prove.

Since, however, there is a genuine dispute here about the extent of the infringement, I will have the master report first upon the number of machines which the defendant has manufactured and which are infringements. If there be found such, I will then direct an account for profits to be filed, and that account to be stated by the master. The proceeding before the master must now continue to accomplish this preliminary purpose, its scope will include all machines made by the defendant like the Baltimore machine, the Raffel patent, or the cut in the American Box Maker. The defendant, if called as a witness, must answer all questions relating to such machines, and the plaintiff may put in such proof as it wishes. The master will then report.

As Judge Geiger, in Beckwith v. Malleable Iron Range Co. (D. C.) 207 Fed. 848, 854, suggests that the plaintiff in such a case should indicate his dissatisfaction under rule 63 by an exception, it may be as well for the plaintiff to except to the statement filed, for it is obvious that this case is to be fought with the extreme of technicality.

If it is thought necessary, an order may be entered, which ip view of the general prayer for relief is not limited to the relief specifically asked. Settle on notice.

September 16, 1916.

[6] The questions presented upon the exceptions to the master’s report are two: First, whether the defendant’s machines, as now finally found to be originally constructed, infringe the patent; and, second, whether the Lacaux machine, assuming its délivré date is what counts, and not its publié date, is a valid anticipation. A preliminary question also arises which should be first disposed of; that is, whether there should be a further rehearing to allow in evidence other patents not discovered before. This question I can dispose of in the plaintiff’s favor, without the least compunction. The case has, for one reason or other, already dragged out over much too long time. After the proofs were once all closed, I allowed the cause to be reopened to put in the Lacaux patent, but only upon the ground -that, owing to the indexing of foreign patents in the Patent Office, it was not fair to hold any one responsible for upturning even with reasonable efforts all relevant foreign references. Now, after the cause has proceeded for another year, .the defendant proposes to introduce a number of added United States patents and one added British patent, concededly an equivalent of one of the United States patents. The defendant offers no ground for this unusual relief under such circumstances, except that the relevancy of these patents must some day be decided. I should disregard every rule applicable to rehearings, were I to allow the application. The references are not foreign (omitting the Lake British patent), and there is no suggestion of a reason why they should not have been originally discovered. The plaintiff refuses to allow the cause to go on without the right to offer rebutting evidence at least by the explanation of experts, and it is within its rights in so objecting. If trials are to conclude anything, the counsel must be held responsible for their preparation at the outset. I shall therefore decline to consider the new references.

I therefore proceed to the question of infringement. The defendant’s machine is so made that at least at the outset the belt, which holds the paper by a resilient pressure upon the heater, moves forward at a slower rate than the upper and lower rollers advance it through the ma;chine. This is accomplished by a difference of gearing which insures .a slight backward drag upon the top surface of the paper against the forward pull of the two wheels. The belt does not, therefore, advance, but, on the contrary, retards the paper in its passage, because the differential between the speed of the belt and the pull of the rollers necessarily results in a backward wipe of the belt along the upper face of the paper, a function which is claimed to have advantages in smoothing out air bubbles in the cement. The patentee no doubt supposed he effected this same result by the “ironing effect,” which he speaks of on page 1, lines 92-94. He assumed, and it necessarily follows from his disclosure, that the paper and the belt, 12, travel at the same rate, but any results which he secured by the “ironing effect” on the lower surface the defendant has retained. Even if the backward wipe improves this result, it is only by an added function; the whole effect of the original disclosure remains. i

The defendant’s position is that the claims in suit—the first six—do not admit of a construction which will include such a machine. Claim 1 has the phrase, “means spaced therefrom and movable in relation thereto for advancing the paper and holding it in resilient engagement with said heater.” The question is: First, whether it must be a single “means” which at once advances the paper and holds it in engagement; and, second, whether, if the means may be double, the rollers of the defendant are “spaced” from the heater and “movable” in relation to it. I do not think that the “means” need be single. If the belt,, 12, were cut in half and the left-hand half ran on idlers, while the right-hand half advanced the paper, I should not think infringement would be avoided. Second, are the defendant’s means for advancing the paper and for holding it in engagement spaced from the heater and movable in relation thereto? The defendant’s belt which holds the paper in engagement fulfills both these conditions in exactly the same way'as Langstons’. What of the advancing means ? The upper belt of Langston does not alone advance the paper; it is the grip between the right-hand part of that belt and the lower belt, 13. Only a part, therefore, of the advancing means is vertically spaced from the heater, and these two belts are in precisely an analogous position to the heater as the two rollers of the defendant. Moreover, two such rollers are exactly the equivalent of two such belts; they operate in the same way, and they do the same thing. If the defendant had divided the upper belt as suggested above, I cannot suppose that any one would have thought the claim avoided. The further change of the upper and lower gripping belts into rollers would certainly have been án equivalent. I therefore find that the advancing and holding means are spaced from and movable in respect of the heater in exactly the same sense as Langston’s, and that claim 1 is infringed.

I also find claim 3 infringed. The paper is advanced in Langston, as I have said, by the grip between the right-hand part of the upper belt and the lower belt. The paper is advanced in Raffel by that part of the rubber rollers slightly flattened by pressure which is parallel'with the paper. As the rollers rotate, this part, even though it were a mere tangential point, necessarily moves parallel with the surface of the heater. As the rollers are a precise equivalent, as I have said, for the belts, the language of the claim serves to cover the defendant’s device.

However, claim 2 seems to me not to be infringed because the “pressure-applying mechanism” does not advance the paper. It is an undue extension of the words when the several claims are read together, to hold that the rollers of the defendant are part of the “pressure-applying mechanism,” as they must be to infringe.

Verbally taken, claim 4 is clearly infringed, except for a doubt arising from the phrase “traveling wdth the corrugated strip.” The defendant’s belt does not travel at exactly the same speed as the paper, but nothing in the patent or in the art requires such a limited interpretation. .If, as above suggested the defendant’s belt traveled upon idlers, obviously the claim would be literally included. Instead, being positively actuated, it travels at a slightly lower speed. Yet the words had best be held to exclude only a fixed folding means, below which the paper is drawn frictionally. Taken literally, the more natural meaning of the language would, it is quite true, cover only a belt traveling at exactly the same speed; but all the results of the plaintiff’s mechanism are accomplished by the defendant’s and in the same way, because, as I have said, although the belt has a differential of speed as against the rollers, the “ironing effect” of wiping the paper under pressure forward across the face of the-heater still remains, even if it is improved upon. It may.be said that the effect would be as good if the belt were stationary, but the means would be different, and no one can know whether the result would be the same or not. The point here is that the defendant’s upper belt retains the forward wipe relative to the face of the heater, and does so by traveling in the same direction, and with substantially equal speed, as the paper itself.

Claim 5 I am disposed to hold not infringed because the belt ,does not advance the paper. If that had been the only claim, I should not have hesitated to think the doctrine of equivalents applicable, because the rollers are an equivalent for the right-hand part of the belt; but where there are, as in this patent, multiplied claims, some differences must be attributed to each, and such a term as “endless belt” must be more literally applied than if it stood alone.

At the hearing I held claim 6 not infringed, but I must say that without that opinion before me I cannot now recall why I should have done so. The fact of claim 6 is not of any practical moment, and I shall not disturb the finding. Costs must be divided in any case owing to my decision upon claims 2 and 5, though my present disposition would be to hold claim 6 infringed.

I therefore hold infringed claims 1, 3, and 4, and not infringed claims 2, 5, and 6. I have throughout disregarded the question whether in actual operation the differential of speed between Raffel’s belt and his rollers disappears. In the view I take of the claims, this is a matter of no consequence. It is at best a very vexed matter depending upon how far the differential in speed in fact disappears and how far I must impute to.the defendant, as a contributory infringer, a knowledgé of this inevitable practice which users will make of the machine. A much firmer basis for decision appears to me to exist in the character of the.machine as it is put out. In no event can I see that the result would be changed as regards claims 2 and 5.

[7] I next take up what amounts to a reargument of my decision regarding Lacaux’s French patent, upon the theory that it is the délivré date and not the publié date which counts. I have, it is true, not héard the plaintiff upon this question; but the defendant has stated its position fully, and >as I deem the patent irrelevant, even if the délivré date does determine the question, there is no occasion for waiting to hear the other side. Claims 1, 3, and 4 call for a stationary heater, and claim 4 that the facing strip shall pass in engagement with it. In Lacaux there is no stationary heater and if the felted table, /, were made into such, the facing strip would not come into engagement with it and there would be no “ironing effect” as referred to on page 1, 91/., above mentioned. . The defendant, of course, recognizes this and only asserts that the changes necessary to conform Lacaux into Langston are all suggested in Ferres, and that the necessary changes in Ferres are suggested in Lacaux. At thb outset it must be noted that some of the defendant’s assertions about Lacaux are inaccurate. Thus, it is far from true that:

“If ttie lower endless belt of tbe French, patent were stationary, we would have the precise and exact structure of the Langston patent.”

The belt (conducteur, i) is the means of conveying heat from the revolving heáter (cylindre, h), and, if this belt were stationary, there would be no drying means at all, except so far as the mere tangential point of the cylinder at the very entrance of the paper might most ineffectively dry the paste. The whole theory of Lacaux was of a belt which should be heated, and give up its heat during its passage synchronously with the paper, to be reheated upon the revolving, steam-filled cylinder after ifihad left the paper. Nor is it true.to say that:

“The French patent shows every single element of every one of the Langston’s claims except the stationariness of the lower heater.”

The means for advancing the paper comprise the lower flexible strip (conducteur, i), which is in no sense spaced from the heater, if the felted table were turned into a heater. Moreover, as already said, the facing strip would not come into engagement with the heater in any possible readjustment of the machine.

[8, 9] It is quite true that one could take Lacaux’s belt and, by combining it with Ferres, make a machine which would anticipate the Langston, if Raffel infringes. The floating upper heaters of Ferres (for the defendant is quite right in saying that these heaters float upon the paper) would be taken off, and there would be substituted a heavy belt such as Lacaux shows. The question of patentability, therefore, may fairly be said to rest upon whether such a combination of several elements requires invention, and the patent has no presumption in its favor, since Lacaux was not discovered by the examiner. Whether a combination of several elements requires invention is the commonest issue in patent causes, and depends upon the degree of originality which the court chooses to set up to protect the public against the monopolization of what ordinary artisans with ordinary incentives would have accomplished anyway. We must not suppose that only he is an inventor who has the insight of genius. A person who by a persistent series of experiments eliminates one after another of all possible combinations may be ,an inventor, though each combination is obvious enough as a possible permutation. Indeed, it is such patient work by trial and error that the inducement of a patent would be most likely to call out, for genius often works without any incentive but its own expression. Therefore we are so familiar with the dictum that a good invention may seem obvious once it has proved successful.

If we are to suppose a person with Ferres and Eacaux before him, setting out to experiment with the possible combinations of their elements, it would take no talent out of the common to select as one combination Lacaux’s belt with Ferres’ heater; it would be one of a number of possible machines. But the final selection of the successful combination would involve just the experimentation which may constitute invention, and, when the experimenter has by trial finally ascertained which combination is in fact fruitful, he has made an advance, though he is no son of Archimedes. -It is only when the result of the combination is familiar in analogous situations, so that its appropriateness at once jumps to the eyes of an ordinarily competent artisan, that we should-deny to an inventor the fruits of his work.

The suggested combination of Ferres and Eacaux, might, or might not, have proved such a fruitful experiment. It is impossible to speculate upon it, now that it has shown its value. Even if it took an original insight merely as a possible hypothesis, its trial and demonstration among all the alternatives are enough to constitute it an invention, so long.as past experience did not at once show that it must inevitably work successfully. There was no such experience, and no ordinary artisan would have known that it would necessarily result in a successful answer to what the art required. I therefore find that the addition of Eacaux to the art does no,t invalidate the patent.

A decree will be entered adjudging the 42 defendant’s machines manufactured to be infringements of claims 1, .3,' and 4, and directing the master to take an account of the profits and damages upon the same, which will in turn come on for confirmation.

An order may also go denying the defendant’s application for a second rehearing to introduce the United States patents mentioned in the brief and the Lake British patent.

1 It is reasonable to suppose that all the 16-inch belting was used above where weight was of the essence; while three strands of 7-inch would have been enough below to help pass along the finished web.