SINKO TOOL & MANUFACTURING CO.
v.
AUTOMATIC DEVICES CORPORATION.

James T. Kline, of Bridgeport, Conn., for appellant.

Bernard A. Schroeder, of Chicago, Ill. (Chritton, Schroeder, Merriam & Hofgren, of Chicago, Ill., and W. Lee Helms, of New York City, of counsel), for appellee.

Before L. HAND, SWAN and FRANK, Circuit Judges.

L. HAND, Circuit Judge.

The defendant appeals from a judgment in an action under § 63 of Title 35, U.S. C.A. awarding to the plaintiff priority in the invention of an improved electric cigar lighter. The action was begun on February 16, 1942, to review two orders of the Board of Appeals of the Patent Office, one entered on January 21, 1942, and the other on September 8, 1941, each affirming an award by the Board of Interference Examiners to the defendant of the invention *975 in suit. The defendant is an assignee of one, Johnson, who filed an application January 17, 1936; the plaintiff is an assignee of one, Sinko, who filed an application on April 25, 1938. At the first trial of the action in 1942 the district judge entered judgment for the plaintiff on the ground that the defendant was estopped by a judgment in an action between the parties in the Seventh Circuit.^[1] Upon appeal we reversed this judgment,^[2] and sent the case back for a trial upon the merits. Upon the new trial nothing was added, and the judge once more entered judgment for the plaintiff, this time upon the evidence which had been before him at the first trial. This consisted of the record in the action of Automatic Devices Corp. v. Sinko Tool & Manufacturing Co., supra^[1]; the record in the Patent Office in the two interference proceedings; and three depositions taken in the action by the plaintiff in Chicago. The judge found that Sinko had "reduced to practice" his invention on October 16, 1933; and he awarded a patent to the plaintiff because Johnson had not carried his invention back of his filing date. No question of diligence arises; the sole issue is of "reduction to practice."

The examiner of interferences, the Board of Interference Examiners and the Board of Appeals all accepted as true the testimony of Sinko and his associates — confirmed as it was by contemporary sketches, dated and signed, and by a sample lighter, made at the same time — Exhibit 5. They appear also to have accepted the plaintiff's evidence that Sinko began commercial manufacture of the lighter early in 1938; and that at least 500,000 had been sold. They refused the plaintiff a patent, however, because they found that Sinko had not tested the lighter, particularly because he had not tested it upon an automobile, for which it was primarily intended; and because without such a test there was no assurance that a short circuit would not be set up which would, or at least might, burn out the "heating element" — a resistance coil. Since this is the critical issue in the case it will be well to add to what we said about it in our first opinion. The lighter consists of two separate members which telescope; one, the "knob," for handling; the other, which we shall call the "heater," containing the resistance coil. These are normally held extended by a spring, which forms part of an electric circuit, which also includes the coil. In the form in which Sinko disclosed the invention in 1933, and in which he had manufactured and sold it, one end of the spring was in electrical connection with a "terminal post," or "abutment," set in the "knob"; and the other end was in electrical connection with two "thermostatic arms" set in the "heater." When the "knob" was pressed down against the force of the spring, and was thus telescoped within the "heater," the resilient "thermostatic arms" spread apart and embraced a bevelled end of the "terminal post" with enough power to prevent its return. This completed a circuit through the resistance coil which included a source of electricity outside the lighter. As the current passed through the coil and heated it, the heat expanded the "thermostatic arms" until they lost their grip upon the bevelled end of the "terminal post"; the spring then snapped the "knob" away from the "heater," and the circuit was broken. The noise of the snap told the user that the coil was hot, he pulled out the lighter from its carrying socket, used it while it remained hot, and slipped it back in place. The defect against which the examiner and the two boards found that Sinko had not sufficiently tested, was that the spring might touch the "thermostatic arms" after they released the bevelled end of the "terminal post." Should this happen, concededly the circuit would not be broken; and, although the "knob" would have snapped away from the "heater," the coil would continue to heat after the lighter had been replaced in the socket.

Sinko made the lighter of 1933 with a clearance of one thirty-second of an inch between the "thermostatic arms" and the spring; and he testified that this was *976 enough to insure against any short circuit. The defendant put in no evidence of any kind; and, so far as we can tell from an examination of Exhibit 5, there appears to have been no chance that the arms could even occasionally touch the spring, to say nothing of keeping steadily in contact with it. Moreover, none of the officials in the Patent Office disclosed any reason for supposing that this could happen. Sinko further testified that, when in 1938 he came to manufacture for the market, he made no changes material to this feature of the lighter; and with two possible exceptions that was undoubtedly true. The first of these exceptions was as follows. In the original lighter the "thermostatic arms" were on the "heater" and the "terminal post" was on the "knob"; in the patent application these were reversed; when Sinko manufactured for the market, they were restored to their original position. Sinko was cross-examined in the interference proceedings about these two changes, and he testified that the change made in the patent application was to cheapen any replacement of the "heater," which was apparently liable to be impaired by use. He swore that he made a few lighters in this form; but that he changed back to the original, when he found that the arms did not release quickly enough unless they were on the "heater." This was a perfectly reasonable explanation which nobody challenged at the hearings, or at the trial. Now, for the first time in its brief upon this appeal, the defendant suggests that Sinko put the arms on the "knob" to avoid any danger of a short circuit because in that position they had the same polarity as the spring. There is not a whisper in the record to support such a suggestion; we must decide the case upon the evidence before us, and we refuse to consider the argument. The second exception was this. In the lighter which Sinko manufactured for sale, he introduced a cup, or sleeve, fastened to the "heater," surrounding the arms, and extending up as far as they. Sinko was not examined about this at all, and naturally did not say why the sleeve was added, of what metal it was made, or whether it would serve as a non-conductor. Again the defendant, for the first time in its brief, ascribes this change as intended to prevent a short circuit, when the arms were put back on the "heater." This is even less excusable than the first argument, for, as we have just said, the sleeve was not touched upon at all. These two arguments being disposed of, there is nothing to throw any doubt upon Sinko's testimony that, so far as concerned any possible short circuit, he made no change in his original lighter — Exhibit 5 — when he began to manufacture and sell. Finally, the plaintiff proved at the trial and the judge found that on December 24, 1933, while Sinko was demonstrating the lighter to one, Ayotte, he pushed in the "knob," that the two then left to examine some sketches, that when they came back, the "knob" had snapped out, and that Sinko took out the lighter and found the coil cold. He then pushed in the "knob" a second time and found the coil hot when the "knob" had snapped out in due course. Thus it appears that at least in the shop, the lighter would not short circuit.

Sinko and four assistants who worked with him in 1933, were all highly expert in these devices — Sinko had been making cigar lighters for over five years, and, although these had not been of the automatic "cut out" variety, he was well acquainted with "cut out" mechanisms, for he had made 57,000 thermostatic switches for light bulb sockets in 1926 and 1927, and 4000 or more switches for flatiron plugs in 1927. All these men — expert in this field, and Sinko apparently as expert as anyone could be — believed, when they had completed Exhibit 5 that it was ready for exploitation without change. It is true that Sinko did not file his application for four and a half years thereafter, but the boards acquitted him of suppressing his invention until he had been "spurred into activity" by Johnson's application; and he had certainly not abandoned it. His explanation for the delay was that the year 1933 was not favorable to new and untried ventures, nor were several of the years that followed. It is indeed true that he might have filed an application, for he was apparently an indefatigable patentee; but we cannot see what inference is to be drawn from his failure to do so, except that he may have *977 thought the invention of less importance than it afterwards proved to be. At any rate delay alone is irrelevant except as basis for the inference that the machine had not been finally perfected; an inference which the evidence dispels. The case really comes down to this; whether, although the invention had been enough tested by persons exceptionally competent in the field, to satisfy them that it was ready for production, and although it in fact went into very extensive production substantially as it was, it was not "reduced to practice" in 1933 because it had not been tested on an automobile.

The law that an inventor's mere conception is not his "prior invention" within the relevant sections of the statute^[3] is judge-made. The courts have written two glosses upon the text: first, that the inventor must disclose his invention, either by description, or physical embodiment; and, second — unless he can show diligence in filing an application after his rival has filed — that the invention has also been tested so far as to make sure that it will operate under service conditions. Why this last condition has been added is not very clear, but it is laid down in too many decisions to be now questioned. So far as we can find, it was first bruited in Gayler v. Wilder, 1850, 10 How. 477, 13 L. Ed. 504. The defendant, Gayler, sought to defeat Fitzgerald's patent for a safe by showing that one, Conner, was a "prior inventor," and at circuit Justice Nelson had told the jury that, if Conner had made his safe and "tested it by experiments," Fitzgerald was not the first inventor. The jury brought in a verdict for Wilder, the plaintiff, and the question was whether the charge had not required too much of Gayler. The court affirmed the charge; but McLean, J., dissented, insisting that the requirement that Conner should have "tested" his safe was an unauthorized innovation. At first blush that certainly suggests that the majority meant that a test was inevitably a condition of "prior invention"; but a reading of the opinion shows that this is not true. On page 498 of 10 How., 13 L. Ed. 504, Taney, C.J., said that the court did not read the instruction to mean "that the omission of Conner to try the value of his safe by proper tests would deprive it of its priority." All that the charge really meant was that the earlier safe must have enriched the art; testing it was merely one factor to be considered with the rest. Although in many later decisions the question has arisen of the need of a test under service conditions, it has never been laid down as an inexorable condition in all circumstances; the courts have again and again upheld inventions, not so tested. Nor has this been confined to chemical cases; it has extended to all sorts of inventions: we annex in the margin a collection of some of the decisions with a description of the article involved.^[4] The doctrine to be drawn from the books, as we read them, is this — and incidentally it is the only doctrine that can find support in reason: a test under service conditions is necessary in those cases, and in those only, in which persons qualified in the art would require such a test before they were willing to manufacture and sell the invention, as it stands. In Sydeman v. Thoma,^[5] the court divided the decisions into three groups, and in that there may be some convenience; but, with deference, we are not sure that it does not tend too much to proliferate a doctrine which had better remain in a single and *978 supple stem. Payne v. Hurley,^[6] appears to us to be a most admirable discussion, especially because the court there pointed out why the laboratory tests had failed to fulfil the conditions which the spark plugs must meet in service, and why those conditions were likely to break them down. Moreover, in that case both sides had put in evidence so that the Board of Appeals had an actual issue of fact to decide, and against its decision any inference necessarily went down which could have been drawn from the inventor's personal satisfaction with his laboratory tests.

In the case at bar the situation was altogether different. As has already appeared there was evidence, and it was the only evidence, that the lighter was ready for manufacture in 1933, and that when it came to be manufactured it was not changed in any way which could have tended to remove the danger of a short circuit. True, it was changed in other ways; but merely to avoid those crudities, always inherent in original models; and such changes do not postpone the date of "reduction to practise."^[7] It is of course true that we are only very rarely justified in reversing a finding of the Patent Office; there has never been any disposition to relax the doctrine of Morgan v. Daniels^[8]; rather the contrary. However, that decision is not in point here; it was like Payne v. Hurley^[9]; the date of the invention had been put in issue; the Patent Office had decided upon an actual conflict in the evidence; and it was that decision which the Supreme Court said must prevail. We do not mean that to prevail a finding of the Patent Office must inevitably be on conflicting evidence. In the case at bar, for instance, if the boards had disbelieved the testimony of Sinko and his assistants, we might be bound to accept the finding; its credibility was for them. But, as we have said, they accepted that testimony as to what Sinko did and when he did it, and as to his belief that the lighter was ready to go into manufacture without further test. What they held was that his belief was wrong; and that was not because they preferred any testimony which the defendant had introduced, but, so far as appears, because they drew upon sources not disclosed; perhaps on their own acquaintance with such devices. Yet it was an issue on which the defendant could have introduced evidence; it could have called persons, expert in the subject, to combat Sinko; that would indeed have created an issue the decision of which would have been substantially invulnerable. Indeed the situation might have been such that, though the testimony stood as it does, it would have been proper to affirm; for an inspection of the lighter might have showed that a service test was required. The difficulty is that it does not; that is, it does not so far as we can form any opinion; for to us it appears that a short circuit was substantially impossible. The arms were firmly fixed upon the "heater," and their movement was tiny; the spring was fast at its top and bottom and was necessarily released as far as it could be, when the "knob" snapped out. What could bring them in contact we cannot see; and the boards have not told us.

It is true that, in spite of all this the law might require us to affirm; and so it would if we are to endue the boards with that specialized acquaintance with the subject matter which enables it to dispense with evidence and draw upon undisclosed sources of information, not available to us and unknown to the parties. Perhaps that view will in the end prevail; but it is hard to see, if it does, what can remain of our jurisdiction; and certainly the doctrine of Morgan v. Daniels, supra,^[10] held nothing of the sort; and the Court of Customs and Patent Appeals has repeatedly reversed the findings of the boards. Our conclusion does not indeed mean that the boards were in fact wrong; perhaps they were right; but it does mean that *979 some basis for their finding must appear in the evidence; it does mean that its support must be ascertainable in such sense that it can become subject to some sort of review.

Judgment affirmed.

NOTES

[1] Automatic Devices Corporation v. Sinko Tool & Mfg. Co., 7 Cir., 112 F.2d 335.

[2] Sinko Tool & Mfg. Co v. Automatic Devices Co., 2 Cir., 136 F.2d 186.

[3] §§ 35 and 52, Title 35, U.S.C.A.

[4] Corona Cord Tire Co. v. Dovan Chemical Corporation, 276 U.S. 358, 48 S. Ct. 380, 72 L. Ed. 610 (process for the vulcanization of rubber); Kyrides v. Bruson, 102 F.2d 416, 26 C.C.P.A. (Patents) 986 (chemical compound used in plastics and resins); Buchanan v. Lademann, 54 F.2d 425, 19 C.C.P.A. (Patents) 836 (metal support); Brydle v. Honigbaum, 54 F.2d 147, 19 C.C.P.A. (Patents) 773 (asbestos shingle); Larson v. Eicher, 49 F.2d 1029, 18 C.C.P.A. (Patents) 1497 (cod-liver oil tablet); Sachs v. Wadsworth, 48 F.2d 928, 18 C. C.P.A. (Patents) 1284 (electric switch box); Gaisman v. Gillette, 36 App.D.C. 440 (safety-razor); Rolfe v. Hoffman, 26 App.D.C. 336 (electric circuit protector); Couch v. Barnett, 23 App.D.C. 446 (horse-collar); Loomis v. Hauser, 19 App.D.C. 401 (ticket holder); Mason v. Hepburn, 13 App.D.C. 86 (clip to close magazine used on fire arms).

[5] 32 App.D.C. 362, 373, 374.

[6] 71 F.2d 208, 21 C.C.P.A. (Patents) 1144.

[7] Coffee v. Guerrant, 3 App.D.C. 497, 499; Burson v. Vogel, 29 App.D.C. 388, 394; Sydemann v. Thoma, 32 App.D.C. 362, 374.

[8] 153 U.S. 120, 14 S. Ct. 772, 38 L. Ed. 657.

[9] 71 F.2d 208, 21 C.C.P.A. (Patents) 1144.

[10] 153 U.S. 120, 14 S. Ct. 772, 38 L. Ed. 657.