L. HAND, Circuit Judge.
The patent is for a device to bring electric elevators to a stop at the exact level of the floor. It applies primarily to elevators operated on the so-called Leonard system, which all the claims include as an element. This has a cage operated by an electric motor, which coils and uncoils the sustaining cables upon a sheave; the motor being driven by a generator of the usual type. The peculiarity of the Leonard system is that the field winding of the generator is so arranged that the current can be reversed by a switch in the cage, by which the operator, as the ear approaches the floor, can also throw off the current, so that the cage will continue to drift by its momentum, until the floor is reached. If it fails to reach the floor, he must start it again, in the same direction; if it overruns, he must throw the switch in reverse, and bring back the cage. Thus by a series of oscillations, or small advances, he'will eventually bring it to the level, or near enough to allow passengers to come and go.
The invention in suit was to avoid this necessity automatically. To do so, it provided a cam upon the cage which threw a switch in the hatchway, and so reversed' the current. This made the cage return, if it had overrun the floor, until by its movement the cam again threw off the switch. A brake was automatically set and lifted, in alternation with the reverse current, by another aim of the same switch. It braked the ear before the reverse current came into action, was lifted while it did act, and was again set when it was turned off. This reverse current moved the cage at a low speed, so that it was not likely to drift past the floor, when the current was shut off and the brake was on.
In the specifications the only example shown was one in which a separate field winding upon the generator was supplied to carry the reverse current started by the switch. This was in addition to the main winding, and electrically separate. While the current passed through the auxiliary, it did not pass through the main winding, and could not without substantially neutralizing it. The specifications, however, did not declare that an auxiliary winding was essential to the invention. On the contrary, they read as follows: “The reversal of the direction of rotation of the motor, which is preferably driv-
eB by a controlling generator on the Leonard system, is brought about by the field of this generator.being reversed. For example, this generator may be provided with a separate field winding which is independent of the main windings, and is traversed by the current only: .when the cage travels past the desired etad position, and then in an opposite direction' to that in which the current has previously traversed the main winding.” Nevertheless, claim one, and indeed all the .original claims, specified an auxiliary field winding as one element in the combination.
The patentees, were Germans who had filed their application in May, 1913, and' the patent issued on the last day of February, 1916, after the Great War had been in progress for twenty months. On September 29, 1923, more than seven and a half years after the original issue, they filed a claim for reissue,- which was granted in December of that year. The circumstances leading up to this .were as follows. Shortly after the issue of the original patent an officer of the plaintiff learned of its existence, when it was cited against another patent application. The plaintiff’s patent solicitor in April, 1916, recommended its purchase, but at once saw that it required a re-issue to avail of its full scope. .The plaintiff shortly afterwards tried to reach the patentees, but failed, owing to the confused conditions then prevailing in Germany, and the difficulty of communication. We entered the war in April, 1917, and all possibility of negotiation then ended, but in the summer of 1919 the plaintiff’s officers again discussed the purchase between themselves, and early in May, 1920, the plaintiff made an offer to the German owner, an assignee of the patentees. Negotiations continued without result until the parties came to an agreement in November, 1921.
At about this time the plaintiff learned that the Alien Property Custodian had seized the patent on July 28, 1919, and this prevented an assignment until the Custodian was induced to release his claims late in 1922, the plaintiff becoming the owner in January, 1923. Nothing had meanwhile been done towards a re-issue, -although, as we have said, thé plaintiff had at once recognized that necessity in 1916. Soon after January, 1923, it started to move, but some months passed before it could reach one of the patentees, who was in the Occupied Area. Eventually, having prepared the necessary papers, it got them sworn to, and filed, and the Commissioner accepted the excuse for the long delay. Claims seven and eight, in suit, first appearing in the re-issue, omitted the auxiliary field winding as an element.
Meanwhile the art had progressed along the lines of the invention disclosed, but without the use of the auxiliary. The first step was in a building in Los Angeles, where an installation was begun in 1921 under the direction of the Westinghouse Company, the intervener at bar. This was, however, experimental, and the owners rejected it; it was taken out and never went into use. The next effort was in a building in Chicago in 1922, also under the direction of the Westinghouse Company, but removed because experimental and not contracted for. Both these were designed automatically tq reverse the current in the main field winding of the generator by a cam and switch, one in the hatchway and the other upon the cage. A third similar experimental installation in Pittsburgh in 1922 was also abandoned, because the owners thought it unnecessary.
However, in September, 1923, the Westinghouse Company directed the completion of an equipment in the Federal Reserve Bank in Cleveland, which was accepted, and has since then been in continuous operation. This originally had a foot treadle which the operator must press, in order to connect the automatic circuit in which the leveling switch was put, though it was later changed to operate automatically by means of the cage door; it too had no auxiliary field winding. The two infringements in suit are installations of the Atlantic Elevator Company, one in the building of the Campbell Soup Company in Camden, New Jersey; and the other in the Federal Reserve Bank in New York. The second was begun in the summer of 1922, and completed towards the end of 1924. Just how much had been done before October, 1923, does not definitely appear, but some parts at least, though probably not the leveling devices, had gone far enough to be tested by that time. The Camden equipment was not begun until 1924.
Enough has already been said to disclose the outlines of the controversy. The defendants’ position is that they do not infringe claim one, because they do not use an auxiliary field winding. On the contrary the switch on the cage or hatchway is directly connected into the main circuit of the field winding in such wise as to reverse the current there, just as the operator does voluntarily by his switch. As to the re-issued claims, their position is that these were filed too late, and after the art had so progressed as to make an indubitable expansion of the claims invalid,
independently of tho delay. The plaintiff answers that an extra eoil introduced by the defendants into the armature circuit is the equivalent of the Leonard field winding, and makes the main field winding itself auxiliary, or vice versa. Hence the defendants infringe claim one. Furthermore, that the circumstances of the Great War excused the delay; that the doctrine of intervening rights is based upon a personal estoppel 'which the Atlantic Elevator Company cannot assert; and that the intervention of the Westinghouse Company on its own initiative did not change the rights of the plaintiff as to the infringements in suit. The judge found that the two equipments in suit were infringements of claim one, the use of the main winding being'an equivalent of an auxiliary winding; but he thought the delay unexcused, and declared the re-issued claims invalid for that reason. Both parties then appealed.
The infringement of claim one depends on whether the defendants use “an auxiliary field-winding on said generator for reversing the field thereof.” This they certainly do not do, if that phrase means a second field winding, additional to the main field winding of the Leonard system, and capable of driving the motor during the leveling operation. The defendants have no such winding. The coil referred to in the testimony as a “series field winding,” which the plaintiff asserts to' be the auxiliary, is not of the same kind as the plaintiff’s; nor is it used for the same purpose. It is in a closed circuit, which unites the armatures of the generator and the driving motor, and has no connection with any independent source of power. At times it re-inforces the main field winding, at times it counteracts it; but at no time is it used to reverse the current in the field. Its purpose is not that, but to compensate for varying loads, not to actuate the field at all; it always operates at tho same time as the main field winding, and is not, like tho auxiliary of the claims, its alternative in use. By specifying the Leonard system with an auxiliary to reverse the field, the patentees could only have meant two field windings both supplied with power, one of whieh was out of action when the other was in.
No amount of interpretation can disguise this simple fact, so long as we are concerned only with the meaning of the words used. However, the doctrine of equivalents steps in at just this point. In Claude Neon Lights Co. v. Machlett & Son (C. C. A.) 36 F.(2d) 574, we discussed the anomaly SO' introduced into the theory that elaims are the measure of the invention, and unless the doctrine of equivalents is remembered, the cases are in conflict. On the one hand are those whieh hold that the patent is not infringed if the supposed infringer omits any element of a claim in suit. Cimiotti Unhairing Co. v. American, etc., Co., 198 U. S. 399, 410, 25 S. Ct. 697, 49 L. Ed. 1100; Union Water-Meter Co. v. Desper, 101 U. S. 332, 25 L. Ed. 1024; Evans v. Hall Printing Press Co., 223 F. 539 (C. C. A. 2); Hall-Mammoth Incubator Co. v. Teabout, 215 F. 109 (C. C. A. 2); Vibroplex Co. v. Bunnell & Co., 16 F.(2d) 975 (C. C. A. 2); Walkup v. Interborough R. T. Co., 22 F.(2d) 266 (C. C. A. 2); Detroit Showcase Co. v. Kawneer Mfg. Co., 250 F. 234 (C. C. A. 6). On the other hand, when an infringer has combined in one element the functions for whieh the claim prescribes two or more as necessary, he will not escape. Gibbs v. Triumph Trap Co., 26 F.(2d) 312 (C. C. A. 2); Wachs v. Balsam, 38 F.(2d) 50 (C. C. A. 2); Bundy Mfg. Co. v. Detroit Time-Register Co., 94 F. 524, 538-540 (C. C. A. 6) ; Pederson v. Dundon, 220 F. 309, 311 (C. C. A. 9); Proudfit Loose Leaf Co. v. Kalamazoo, etc., Co., 230 F. 120, 138, semble (C. C. A. 6)r The question is the common one, as to how far we may go in disregarding the language of a claim.
The patent was not, certainly in the broader sense, a pioneer. Otis and Smith, 610,197, disclosed an electric elevator which could be operated from the outside, as well as from the cage. When used in the first way, a cam on the cage and a switch in the hatchway were so disposed that when the cage came to a floor the cam threw out the switch and with it the current. If, however, the operator on the floor continued to close the circuit, and if the car passed beyond the floor, the cam threw the switch in the opposite way, and .reversed the current, whieh would bring back the ear towards the floor. All this, it is true, was done from outside the cage and on the floor to which the cage was to be brought; it was not an anticipation of the invention, but it cannot be said that it left a wide scope for patent. Again Ritter, 909,675, though a hydraulic elevator, and therefore outside the zone of the invention, disclosed a mechanism, which automatically would return the cage to the floor level, when the operator, having thrown off the power, pressed a treadle. This was automatic in its action, except for the treadle, which the Cleveland equipment also had when first installed. Thus it appears that there had been earlier contrivances to level the cage when it
passed a floor, and that one of these was by means of a cam and switch, which reversed an electric current, or at least could be made to do so if desired. Moreover, this was accomplished by reversing the current in the main field winding; automatically, in ease the operator continued'to press the button. While, therefore, there was indeed room enough for invention, it seems to us that the art pressed too close to allow the patentees to abkndon the very scheme and pattern which they had selected for their claims. They chose to rest upon the addition of á field winding added to that of the Leonard system; we think it a total desertion of that choice to expand their meaning so as to include sill means by which the current could be reversed in a Leonard field winding. It is strong confirmation of oúr view that the plaintiff’s own solicitor immediately perceived that the patent must be re-issued, if it was to have general application. As it stood, it adyised the art that the patentees in all their claims meant to ■ reserve only the form which they had specifically disclosed.
We • agree that the disclosure was •enough,,to support the re-issued claims, and indeed the German original went so far. However, during the application the patentees had tried to get broader method claims, which were rejected without any effort at substitution. They appear then to have been content to reserve only the auxiliary winding, and their mistake, if 'it properly was one, was discovered not by them, but by the plaintiff. We do not,’.however, depend upon this primarily, because the application was delayed for seven and a half years, and that, standing alone, was much more than enough to defeat it. The usual allowance is only two years, after which the courts look narrowly at such applications, and demand that special circumstances be shown to excuse the delay, when, as here, the claims are broadened. Wollensak v. Reiher, 115 U. S. 96, 101, 5 S. Ct. 1137, 29 L. Ed. 350; Topliff v. Topliff, 145 U. S. 156, 171, 12 S. Ct. 825, 36 L. Ed. 658; Ives v. Sargent, 119 U. S. 652, 662, 7 S. Ct. 436, 30 L. Ed. 544. Moreover, the plaintiff must be content to allow the defendants to tack the successive delays of the owners of the patent during the whole period in question. It is not enough that it was itself as expeditious as reason required, áfter it had secured the patent. Electric, etc., Co. v. Boston Electric Co., 139 U. S. 481, 11 S. Ct. 586, 35 L. Ed. 250; Hartshorn v. Saginaw Barrel Co., 119 U. S. 664, 7 S. Ct. 421, 30 L. Ed. 539.
It would be unfair to charge the patentees with laches from the issue of the patent until the time when it became possible to deal with the Patent Office; that is, from February, 1916, until the spring of 1919, when communication became possible. However, they then showed no disposition to move, nor did their German assignee, a powerful corporation, which presumably had competent legal advice. The Alien Property Custodian seized the patent on July 28, 1919, as we have said, and, held it till about the end of 1922, but we are not inclined to exclude this period either. He was charged by the act with the duties of a common-law trustee, and he had a staff of patent attorneys, employed, in part at any rate, to protect the property seized. Moreover, the seizure was not a confiscation of the property; it was not so proclaimed, and it did not so result. It was by no means impossible for the Custodian to have got the necessary papers from the patentees, just as the plaintiff did later. They themselves, or' their assignees, still had an interest in their invention, both because it had not been confiscated, and because the plaintiff began its negotiations with the assignee early in 1920, nearly three and a half years before the re-issue. There were therefore always persons interested in its preservation, and, the patentees were accessible for at least four years before the application was made. One of them, Osborne, was in the employ of the German assignee, and apparently always subject to its direction. On the whole we are disposed to think that four years of the delay was not excused; that was twice the time ordinarily enough to bar the application. Mahn v. Harwood, 112 U. S. 354, 5 S.
Ct.
174, 6 S. Ct. 451, 28 L. Ed. 665.
Be that as it may, the art meanwhile had not stood still. The installations at Los An-geles, Chicago and Pittsburgh were, it is true, experiments, and did not go into permanent operation, but they were steps forward. They involved considerable expense, whose fruition came later in the building at Cleveland, completed before the application was filed’. We should not ignore them, because they were not used; the analogy of a prior use as anticipation is not apt. The question, on any View, is of the conflicting interests involved ; the patentee’s, to extend his monopoly; the public’s, to be protected in vested rights acquired meanwhile. Earlier experiments, involving large expense, and being the first steps in a continuous development which ends in a completed apparatus, are fairly to be taken as vested interests. Unless we aré
to admit nothing except what the infringer himself may do, the case is one for the doctrine of intervening rights. The art by successive steps had already occupied the area comprehended by the expansion of the re-issued claims.
However, all this had been done by others, whom the plaintiff did not single out for attack, and, if intervening rights are relevant only on the score of a personal estoppel, these earlier efforts must be disregarded so far as concerns the Atlantic Elevator Company, whose two installations alone the plaintiff asks to enjoin. There are some expressions in the books, though no decisions, that it is only the intervening interest of the infringer at bar which counts, and that the general progress of the art is irrelevant [Autopiano Co. v. American Co., 222 F. 276, 282 (C. C. A. 2); Ashland, etc., Co. v. General Refractories Co., 27 F.(2d) 744 (C. C. A. 6)]; although there is a dictum to the contrary in Milloy Electric Co. v. Thompson-Houston Electric Co., 148 F. 843, 847 (C. C. A. 6). Since Keller v. Adams, 264 U. S. 314, 44 S. Ct. 356, 68 L. Ed. 705, the question must at any rate be considered as still open. Nevertheless, it is very difficult in most eases to make out the elements of any real estoppel, on which such a theory must stand, or to accept it without considerable revision of the way in which the matter has been treated hitherto. The argument goes that the patentee, by the abandonment of so much of his invention as he does not claim, may have led others to act upon the faith of his declaration. It follows that, although it is too late for him to retract as to those who have in fact so acted, he may do so as to such as have not. The logic cannot be denied, but its consequences carry one further afield than has hitherto been recognized. In the first place, the critical date should certainly he that of the re-issue itself, not of the application, since the application is not published. Yet the date of re-issue has never been thought material. Again, it ought to he a good reply, even when the infringer has himself occupied the zone of the expansion, to show that he did so in ignorance of the patent, on the faith of whose narrow claims he could not in that event have relied. While at times courts have emphasized the fact that the infringer had in fact so relied (Keller v. Adams-Campbell Co., 287 F. 838 [C. C. A. 9]), in general that circumstance has been ignored. Moreover, estoppel can > certainly have no application to the expansion of claims during the original application, and yet the doctrine of intervening rights is also applied to substantial amendments in that. Webster Electric Co. v. Splitdorf Electric Co., 264 U. S. 463, 44 S. Ct. 342, 68 L. Ed. 792. Finally, it is impossible to reconcile a decision like Hartshorn v. Saginaw Barrel Co., 119 U. S. 664, 7 S. Ct. 421, 30 L. Ed. 539, with any such theory. The first of the two patents there at bar originally had the broader claims; they were transferred to the second by re-issue. The first was apparently valid, and turned out not to be only because it was invented later than the second. The art could not innocently have acquired any intervening rights; an infringer could urge no more than that he supposed he was wronging some one else, not a very good excuse at best. These situations must remain anomalies, if the doctrine is only one of personal estoppel.
That the lapse of time may be fatal to a re-issue appears to have been first held in Miller v. Brass Co., 104 U. S. 350, 26 L. Ed. 783 (Walker, § 276 [6th Ed.]), and.it was by analogy with the statutory period of abandonment. Page 352 of 104 U. S., 26 L. Ed. 783. The re-issue involved broadened claims, as here, and upon the decision has followed the general doctrine that in such a ease a delay of more than two years is presumptively conclusive. But the statute does not require an infringer to show that he has acted upon the faith of the patentee’s delay, when the question is of abandonment, stricti juris. The policy of the law is that the mere lapse of time forfeits the patent. This is not a punitive measure; it is because the public has had this period in which to adopt the invention, and should not he cut off by one who has let the art develop. While the analogy is not complete, because two years is not an absolute bar to the re-issue even of broadened claims, in so far as the doctrine of intervening rights is relevant at all, it appears to us to he rather the positive confirmation of an inference which delay alone might justify; proof is better than presumption, and demonstration than conjecture. The evil anticipated is shown to have actually occurred. So viewed, intervening rights are only another aspeet of the policy which makes laches alone enough, and it is not legitimate to turn it into an independent ground of denial. So it does not seem proper to limit it to those cases in which a personal estoppel can be proved; not only, as wo have said, because this would involve a radical recasting of the law, if it is to be consistent; but because it is not agreeable to its origin. So far as we can find, no
court has yet actually held that a re-issue may be good at large, but bad as to one who has acquired vested rights after the original issue. Such loose expressions as exist in the books, with the exceptions noted, appear to imply that an infringer may vicariously avail himself of any intervening rights (White v. Dunbar, 119 U. S. 47, 52, 7 S. Ct. 72, 30 L. Ed. 303; Dunham v. Dennison Mfg. Co., 154 U. S. 103, 111, 14 S. Ct. 986, 38 L. Ed. 924; Mahn v. Harwood, 112 U. S. 354, 360, 5 S. Ct. 174, 6 S. Ct. 451, 28 L. Ed. 665; Clements v. Odorless Apparatus Co., 109 U. S. 641, 650, 3 S. Ct. 525, 27 L. Ed. 1060), though the language in Topliff v. Topliff, 145 U. S. 156, 169, 12 S. Ct. 825, 36 L. Ed. 658, does perhaps squint in the opposite direction.
Whether the same rule should apply within two years, we need' not say. No doubt consistency would require it, and possibly in any case the effect of intervening rights should be the same. But courts do not usually cut so fine in such matters, and it is probably true that they would weigh more heavily in the ease of an excused period of more than two years, than when the lapse of time is less. At the risk of leaving the matter somewhat equivocal, we need now hold only this, that when the time has been longer, it is not necessary in order to avail of intervening rights that they should be those of the in-fringer. It is enough if the art has progressed, though 'he personally be a late comer.
As to the New York bank, which was well under way before the end of September, .1923, it is not necessary to decide the point. Certainly as to it the Atlantic Elevator Company had acquired vested rights — by its contract and substantial performance. On the .other hand the evidence does not show how far the preliminary work done on the New York bank was a condition upon the Camden installation. If it was, that too is protected, even though the doctrine of personal estop-pel be applied to the fullest. The case in that event is like that of one who has built a machine and proposes continuously to market its output; he is as immune for its product after re-issue as before, Autopiano Co. v. American Co., 222 F. 276, 282 (C. C. A. 2); Ashland, etc., Co. v. General Refractories Co., 27 F.(2d) 744, 746 (C. C. A. 6). A consistent application of the doctrine of estoppel would, indeed, require the Atlantic Elevator Company to supply this defect in proof. On no theory could it be held that an estoppel, affecting one infringement, should give a general license, or protect quite independent infringements, though by the same person. Yet it is doubtful if any court would make the distinction. In any event we decline to require the Atlantic Elevator Company to show that the Camden equipment was made possible only by preliminary work before the date of the re-issue.
If we are right as to the unexcused delay of four years, that alone was enough; if not, that delay, coupled with the intermediate advance of the art, precluded the patentees on September 29, 1923, from broadening their claims as they did. We think that the District Judge was right as to claims seven and eight, though wrong as to claim one.
Finally, the plaintiff argues that under section two of the Nolan Act the time between February 29,1916, and March 3, 1922, should be excluded, in calculating the delay to apply for its re-issue. This is not the purport of the section. All it did was to give persons in the position of the patentees an added year after March 3,1921, within which to take “any action with respect to an application.” Assuming this to apply to the situation at bar, the most that could be said is that, had the time to file a re-issue expired before March 3,1922, the patentees would have had until then to file it. The plaintiff in fact waited for nearly nineteen months after that date. The “equity” of the statute rather makes against its position than for it. Having set a limit to the excuse which the war gave, the section may be better supposed to haVe meant that the war should not be considered as excusing any further delay. We do not find it necessary so to decide; we raise the doubt merely by way of answer to the plaintiff’s argument, drawn from the implied attitude of Congress.
The decree is reversed as to claim one and affirmed as to claims seven and eight; cause remanded with instructions to dismiss the bill.