January 9, 1928.
Appeal from the District Court of the United States for the Western District of New York.
Suit by the Electrical Engineers' Equipment Company against the Champion Switch Company. From an interlocutory decree holding valid and infringed Getts patent, No. 1,212,161, and Jacobs patent, No. 1,515,116, fendant appeals. Modified, and affirmed, as modified.
Both patents are for electrical switches, called "disconnect" switches, and used for carrying large quantities of current. Getts' disclosure presupposed two terminals, or "clips," which the switch proper was to bridge. It had been usual to have each clip in two blades, between which a single blade of the switch proper was wedged, when closed. Difficulties in the contact had been experienced, which it was the object of the invention to avoid. Switches had later been made of two blades, pivoted on the outside of one clip, and falling astride the other upon closure. These were fixed at an inalterable interval from each other by bolts pressing their free ends against a separating block. As the established interval was somewhat less than the width of the clip, the clip pressed the blades apart, the pressure being created by the resiliency of the blades themselves. Difficulties had, however, arisen in this device also because the contacts and pressures were not even over the whole surface of the clip, and the flow of current depends upon the completeness and the pressure of the contact.
Getts' patent was to remedy this defect by making the ends of the blades free to move to and from each other — "free-floating" as it was called. To the outer ends of the blades was fastened a manipulating handle, the nearer end of which constituted the separating block. It was held by bolts loosely fitting in holes in the blades, so as to allow their necessary movement to and fro. In the form here in question two other bolts were inserted through the blades, likewise loosely fitted, one on the far, and one on the near, side of the clip. Pressure upon the clip was exerted by placing spring washers on these bolts at the outside of each blade, so that, when the blades descended astride the clip, it forced them apart against the resistance of the washers. In this way the inner sides of the blades were brought in firm and complete contact with the whole area of the clip.
The Jacobs patent was for a switch like Getts', and for an improvement on it. The features here in suit are two: A latch to hold the switch from slipping off the clip; and means for exerting pressure different from Getts' spring washers. The latch was a complicated mechanism, difficult to describe in words, but generally speaking as follows: The further edge of the clip was notched and the top beveled. Between the outer ends of the blades a latch was placed, moving longitudinally and spring-actuated, so that as the blades descended the bevel pushed out the latch towards the blade ends, until it had descended below the notch, when the spring drove it into place beneath the notch and held the blades firm. To disengage the latch a pivoted member was placed beyond it and between the blade ends, so designed as to pull out the latch beyond the notch, when the pivoted member's outer end was rotated. This member had lugs, which bore beneath ears extending laterally from the outer end of the latch.
The second feature consisted in supplying the place of the spring washers by a saddle on the outside of each blade, straddling the clip. Inside the saddle on each side was a circular concave spring member, held in place by the saddle, and exerting a force at its own edge upon the outside of the blade to press it against the clip. The saddles were held in place by the bolts on either side of the clip, running through each blade and loosely fitted in the holes.
The alleged infringement differed in design from Jacobs' patent only in two respects. The defendant asserted and the plaintiff denied that the blades did not move to and fro as respects each other; both agreed that the spring pressure means, instead of being like the disclosure, was a single piece of metal on either side, anchored by the bolts, and bent convexly to the blade, and depending for its action upon its distortion if the blades were moved apart. This piece was in general triangular in form, and fastened by only one bolt at its near end, as opposed to Jacobs, which had two bolts on either side of the clip. The defense was that the bearing surface was only in a line, or small vertical surface, instead of in a circle, as in Jacobs.
The Jacobs switch was the final step of a series, or, as the plaintiff chose to call it, in an evolution. More than two years before Getts applied for his patent, the plaintiff had made a similar two-bladed switch, called in the record the "Detroit switch." "Free-floating" blades straddled the clip, but they were of unequal length. The end of the shorter blade was just long enough to carry the bolt on the outside of the clip, but the longer blade extended much further, and had on its inner side a block which contained a housed latch and releasing member, not dissimilar to the latch and member of Jacobs. Thereafter Getts' patent appeared, and later, but more than two years before Jacobs' application, another form, Exhibit 2, which had the complete external appearance of the Jacobs switch. The lugs on the latch of this switch were, however, of different shape, and these switches, whenever installed, proved unsatisfactory, and had to be replaced.
Besides the Detroit switch and Exhibit No. 2, the defendant proved the use prior to Getts' application of another two-bladed switch made by one Cartwright. Here the blades when in position were screwed together by a threaded bolt, actuated by a lever and putting pressure upon two springs, held in place by turning back the ends of the blades.
The development of the plaintiff's business was continuous over a period of seven years, culminating in Jacobs' application in 1919. Its first "disconnect" switch had appeared in 1912; it was not free-floating. In 1913 was the first design for a lock, which was that installed in the "Detroit switch." In 1916 Exhibit 2 appeared and was installed, but, as has appeared, was unsatisfactory. It, however, embodied Getts' patent, which the plaintiff had acquired before issue. Jacobs, an expert then in the employ of the General Electric Company, came to the plaintiff in October of 1916, and with his contribution the plaintiff's switch reached its final form. It has had a very considerable success, and has been adopted by the General Electric Company.
The District Judge granted an interlocutory decree on August 5, 1926, and thereafter denied a motion for a rehearing. On February 19, 1927, the defendant filed what it called a petition in the nature of a bill of review, with voluminous affidavits, which the plaintiff answered. The judge consented to entertain this petition, considered the affidavits, and affirmed his previous ruling. The decree then entered, which is that appealed from, provided that upon the pleadings, the original proofs, and "those taken and filed on the reopening of the cause under the bill of review and answer," the original interlocutory decree should be made the interlocutory decree on the bill of review, and that the injunction therein should be continued in full force; further, that the proceedings had before the master should be regarded as had under the decree on the bill of review.
William S. Hodges, of Washington, D.C., Daniel J. Kenefick, of Buffalo, N.Y., and Oscar S. Blinn, of New York City, for appellant.
Charles A. Brown and John A. Dienner, both of Chicago, Ill., for appellee.
Before L. HAND, SWAN, and AUGUSTUS N. HAND, Circuit Judges.
However anomalous may be a bill of review, or a bill in the nature of a bill of review, before final decree, we are content to accept this one as the equivalent of a petition for rehearing upon new evidence. Even so, the proceedings were quite irregular. The motion should have been granted, the first decree vacated, the new proofs taken and a new decree entered, if we are to have jurisdiction. Strictly speaking nothing of the sort was done, but the first decree was made the decree on the bill. However, the affidavits were considered along with the old proof and it seems to us that, however defective the form, in substance the decree appealed from is a new decree, entered upon all the evidence, old and new, and the only interlocutory decree now outstanding in the case, Armat, etc., Co. v. Edison Mfg. Co. (C.C.A. 2), 125 F. 939. In spite of the formal irregularity of what took place, the substance is such as gives us jurisdiction of the controversy.
The only attack upon the validity of Getts' patent worthy the name is from the "Detroit switch" and the "Cartwright switch." The "Detroit switch" did not have two blades, held together by a bolt and held apart by a spacing block between their outer ends. It is true that in operation this switch did clamp the two blades against the clip, just as Getts' switch did; but the two blades could move together only through the two bolts one on either side of the clip. Moreover, as the blades had to be free to approach or recede, they could not be solidly fastened together by the bolts; they must "float." This was an essential feature of the structure, if it was to operate like Getts. It necessarily resulted from this design that, when the switch was pulled down upon the terminal, the longer arm must pull down the shorter only by means of the two bolts that connected them. This tended to twist the two out of alignment, and the plaintiff asserts that it made the switch short-lived and interfered with the contact. To this assertion we need not assent, for it is quite enough that the mode of operation and the structure differed, and that these differences resulted in the entire elimination of the older type, and the substitution of the patented disclosure.
The Cartwright switch is even more foreign to the patent. Here there was nothing at all which automatically pressed the floating blades upon the clip. On the contrary, after the switch had been closed, the blades were in loose contact, and had to be pressed against the clip by a lever operating by a screw thread and two springs. It is apparent that this has nothing whatever to do with Getts' patent. Thus there is no valid anticipation.
The defendant's argument on the issue of infringement is that the blades of its switch do not in fact move to and from the spacing block, but that, on the contrary, they exert pressure upon the terminal only by their own resiliency, as in the early two-bladed switches. It is a complete answer to this that the defendant's own witness, Renke, acknowledged that there was a total movement of about a fifty-third of an inch. The play in the plaintiff's practice is only one thirty-second of an inch, and infringement is not avoided by decreasing the amount, if the operation remains the same. We are not impressed by the attempt to turn this difference in degree into one of kind. If the defendant wishes to lock the ends of the blades to the spacing block so tightly that they will not move at all, it is easy to do so, and it will escape the injunction. While the bolt head is not in close contact with both blades when the switch is raised, that proves nothing. The important action is the movement to and fro of the blades with respect to each other, and with the proof of this was proved the issue of infringement.
Of the two classes of claims contained in the Jacobs patent, the latch and the clamping springs, the validity of the first alone is challenged. The nearest approach by way of anticipation is Defendant's Exhibit 2, which was sold and installed both at St. Louis and at Detroit. If it had worked successfully, perhaps it would have been a good anticipation; at least, we shall assume as much. So far as we can see, the only difference between it and Jacobs' disclosure is that the head of the latch has lugs of slighter strength than the solid T-end disclosed in the patent, and this is not a feature of any of the claims.
However, it is undisputed that all of these switches were unsatisfactory in operation, were taken out by the buyers, and were succeeded by the switch of the patent. Just what the trouble was does not appear with entire certainty, though it is clear from Jacobs' cross-examination that the lugs were not strong enough and broke in service. What were the other difficulties, if any, is left wholly to speculation. The rule is that the defendant has the burden of proving that the alleged prior use was of an operable machine, for neither a prior patent nor a prior use is a good anticipation, if it is inoperable. Keystone Mfg. Co. v. Adams, 151 U.S. 139, 145, 14 S. Ct. 295, 38 L. Ed. 103; Kirch-berger v. Amer., etc., Co. (C.C.A. 2) 128 F. 599, 605. We can see no difference between inoperability due to defective designing, by which the parts will not from the outset co-operate together at all, and that due to weak parts, which will break down in service. Nor do we think that a machine which breaks down under service is within the rule that an anticipation need not work as well as the patent. It will not do the work intended at all, because that is not confined to the initial movements.
However, if the correction is obvious to the ordinary mechanic, the defect cannot be vital. In the case of a prior patent the claims would still be valid, if the correction was at once apparent; a mere mistake in the disclosure is not fatal. Yet, if such a patent would be valid, it can only be because the public gets a good consideration for the monopoly, which means that, despite its defects, it contributes to the art. The same rule must apply to prior uses. Hence we conclude that, since the switches, Exhibit No. 2, were actually sold, not as experiments, but for commercial use, the fact that they were unsuccessful, because of an immediately remediable defect, would be irrelevant; they would still be anticipations.
So the question is whether the defendant has shown that the defects were of a kind to be corrected without invention. It seems to us that the evidence does not justify that conclusion. The plaintiff was well skilled in the art; it had already been working for seven years upon the general subject of "disconnect" switches, and had tried two earlier forms; Exhibit No. 2 was the third. It is apparent that the problems involved were not easy, and that structures thought adequate by competent persons turned bad. The buyers were themselves engaged in electrical development; presumably they had the competence of the ordinary artisan, and, indeed, much more. They did not themselves make the necessary changes, and, so far as appears, they did not suggest them. Jacobs, a new mind in this branch of the art, took the last step, and converted failure into success, and that a substantial success. In the light of the foregoing, we should have to substitute our own uninformed speculation for the history of the art. We are, indeed, well aware of the hazards of such inferences; but, when all is said, in such cases there is nothing more reliable at hand. We can only resort to trite convention, and repeat that, if the change was so apparent, it should have earlier appeared.
The other alleged anticipation is the Detroit switch. The latch was not of the Jacobs form; less like it, indeed, than Exhibit No. 2; and the member which operated it did not extend beyond the ends of the blades, being housed in the block attached to the longer. Verbally, at any rate, it does not read upon claims 1, 2, and 3. The difference is a slight one, as we have said; but the Detroit switch was inherently defective, and never got any currency. The member which operated the latch lay loosely in the block, and was also housed in it. The aperture, being in the body of the housing, admitted of only a small play for the member, and was too small, if the blades were narrow. These were substantial defects, or, if not, substantial differences between the two. Not until after it had passed through the intermediate form of Exhibit No. 2 did invention come to an equilibrium. Again we say that it would be mere assumption to suppose that no invention was necessary to skip that step and to reach the desired form.
The prior patents add nothing closer than what we have discussed, and there remains only the question of infringement. As to the claims for the latch, no point is made; but claims 7, 35, and 38, involving the spring compression feature, are in issue. It is enough for these to say that, though the means is modified, it is in substance the same as is the result. The pressure in the alleged infringement is not confined to a geometrical line perpendicular to the blades. Rather it is an area of pressure, more or less broad on either side of the clip, forcing the blades together through the bearing of the springs upon the both heads. This is an equivalent of the claims whether or not it is an improvement.
There were 9 claims in Getts' patent and 39 in Jacobs', of which all of the first and 27 of the second were here put in suit. Getts' claims are not divided into classes determined by separate features of the disclosure; Jacobs' were, as we have said. It was altogether unnecessary to cumber the case with the mass of verbiage which these claims present in the aggregate. We have long enough protested against the practice; our complaints still remain unattended, and will be so until we take some action actively to discourage it. We mean to do so. Among several claims, the most general, which covers the alleged infringements, alone is necessary to a decision of the controversy; the rest thereafter necessarily present moot issues. If the defendant infringes the more general, a fortiori he infringes the particular. The plaintiff alleges that he has committed a tort in selling the device contrary to his monopoly. We so decide. He asks an injunction; we give it. Whether he has committed other torts is of no consequence. The dialectic which picks its way through the tangled language of 24 abstract statements of the same concept is barren and idle. We decline to engage in it.
We hold claim 1 of the Getts patent, and claims 1 and 7 of the Jacobs patent, valid and infringed, and direct an injunction to go on each. We direct that the decree be modified, by dismissing the bill as to the other claims in suit without prejudice. The appellee will take no costs of the appeal for this reason.
Decree modified as above, and, as modified, affirmed.