LEARNED HAND, District Judge
(after stating the facts as above). The problem before inventors was to prevent the charging of storage batteries by too high a current, resulting in the destruction of those accumulations in the cells upon whose creation the stored energy depends. If the current be too large, which in this instance also means a current of too high voltage, these products, normally aggregated for future use, are turned into gas to the loss and injury of the battery itself. As the battery becomes charged, its voltage rises, and with it the voltage in the charging current, with which varies the amperage.Hence the amperage of that current is always the test. It must for safety be kept within predetermined limits. As the current varies with the energy developed at the armature of the generator, it necessarily depends upon the speed of revolution; i. e., the revolutions per minute (R. P. M.), of the shaft upon which the armature is fixed. Since the speed of the engine cannot for practical reasons be regulated to [586] the needs of the battery, it follows that either there must be a periodic disconnection or slip between the armature and the engine, which will result in the partial disconnection of the system from its prime motor, or there must be some way of damping or disconnecting the current created by the generator from the battery itself. No plan appears in the record for a periodic disconnection of the generator from the battery, and the alternatives are therefore limited to a disconnection or slip between the generator and its prime motor 'and the damping of the electromotive force (F- M. F.) of the generator. One of the points of the case, in our judgment a critical one, is whether these two methods of dealing with the problem are for purposes of invention to be deemed equivalent; but we pass this for the moment, assuming for the statement of the problem that they are to be treated as substantially interchangeable.
No one disputes that it was well understood among electricians that the electromotive force generated by the rotation of an armature within its field depended upon the current in the field and that a common device to reduce that current was by the introduction of a resistance in the field circuit. Such a resistance is shown in a system such as is here in question in the patents to Thomson and Moskowitz, through its effect is certainly gradual, and not inserted and withdrawn all at once, and at opposite phases of the main current. The new thing which Leonard devised,, if anything, did not therefore depend upon the diminution of the current charging the battery, but in the way that that diminution was accomplished. That was as follows: When the cur-, rent in the charging circuit rose to an -amperage which sufficiently energized the magnet, 88, so that it attracted the armature 88', the circuit was broken, 'which energized tire clutch magnet. This mechanically disconnected the clutch, and deprived the armature of all future energy from the engine. Any further rotation of the armature depended altogether upon the mechanical momentum theretofore acquired. This would soon be destroyed by the reluctance within the generator, and with it would disappear the current generated in the charging circuit. When that current had fallen enough to release the armature, 88', it would be snapped back by its spring, and, the clutch coils coming again into circuit, the clutch magnet Would engage that member of the clutch which was always in connection with the engine. The mechanical connection between the engine and the generator would be re-established, and the engine would rotate the armature as before.
Now the magnetization of the magnet, 88, must be stronger to attract the armature than to release it because of the air gap, and the system independently of its inertia, has therefore a maximum current ■of one value at which the clutch is disconnected and a minimum of -another at which it is connected. It is urged that tire full corrective ■effect of the clutch becomes effective at once, either in connection or •disconnection; and this is perfectly true, although it is not true that the effect is immediately translated into terms of current. The effect upon the current, on the contrary, is gradual, resulting in a pulsating ■or vibrating unidirectional flow.
[587] A confusion of issues readily arises over the issue of the vibratory character of the current. From the aspect of the battery a constant current is quite as good as a pulsating; Thomson and M'oskowitz would charge a storage battery quite as well as .Leonard. The supposed advantage of Leonard's device was that his relay operated only at the two critical periods in the current, the maximum and the minimum, and then it operated all at once, and without any possibility of derangement by the jars and shocks constantly happening in a motorcar. The pulsations of the current are therefore taken, perhaps correctly, as a symptom of a system which operates at opposite critical phases in the charging or work current, and that, too, by the entire effect of the corrective. Nevertheless, unless it also appears that such a method of correction necessarily involves a structure which will not be deranged by jars and shocks, it would seem to follow that the advantages of Leonard's disclosure, if any, lie in something else than the mere principle of the sudden interposition of the entire corrective. It may lie in the ‘‘trembler” or “air-gap” relay, which is not of the “plunger” type, and which is sure to operate under all the trying circumstances of motorcar use.
Now it is quite clear that the claims do not cover this kind of relay As we shall later show, the Everett-Bliss regulator, even if it was not as it stood a mechanism which would interject the entire corrective at once, needed only the substitution of metal for carbon points to constitute an anticipation. Yet it had not Leonard’s relay, but was of the plunger type, and was perhaps unfit for motorcar use. We may dismiss, therefore, any consideration of these details of the structure, merely observing that the utility of the patent by no means follows upon that element which eventually became the most prominent in the claims.
A pulsating or vibrating current may be a certain symptom of the sudden interposition of the entire corrective; at least for the purposes of this case we may assume so. The first difficulty which we find in the file wrapper is that no mention occurs of any pulsating current until January 31, 1912. This, it is true, would not be alone enough, if the original disclosure had already contained the suggestion of it! We do not mean to hold that an original disclosure may not be amended to specify features already appearing, even by intimation, though not thoroughly observed at the outset. The difficulty in the case at bar goes deeper, because it appears positively that a constant current, or as nearly constant as possible, was thought to be within the scope of the patent. Thus in Figures 4 and 5 Leonard disclosed mechanism which operated by the gradual interposition of a resistance coil into series with the field circuit. In the specification of Figure S it was expressly stated that the object was to “maintain constant” a current generated by a continuously driven armature. We may therefore be sure that at that time neither the symptom, a pulsating current, nor the cause, the sudden interposition of the entire corrective, were a part of the invention. Nothing of the sort appears anywhere in the file wrapper until January 31, 1912, when Leonard first mentioned the “wide and rapid fluctuations of energy” which accompanied his disclosure of the clutch. On March 27, 1913, he canceled Figures 4 and 5 with their de[588] scriptions. This change of position, while for a time still uncertain in its details, in general Leonard maintained till the patent issued. It was an entire antithesis to the original disclosure and it is doubtful whether it could succeed, if the art had itself remained stationary.
However, the art did not remain stationary, because before January 31, 1912, it had already gone quite as far as the defendant now seeks to go. Whatever the actual date when the Bisfiop-Delano regulator was put into practice, Delano’s application was filed in May, 1910, long before the amendments in the file wrapper, and for that matter before Leonard had changed his own practice from a clutch-to a resistance coil. Furthermore, Turbayne’s regulator was in public use early in January, 1912. Each of tírese regulators was an anticipation of the form which Leonard now seeks to include. The whole corrective was suddenly thrown in and out of series with the field coil by a relay which operated upon the “trembler” principle. We do not understand that the contrary is seriously asserted; but, if so, we are entirely -satisfied that the contention is without basis. The amendments are invalid which sought to anticipate tírese inventions by canceling- out Figures 4 and 5, with their specifications, and by interjecting vague statements about the. fluctuating nature of the current developed by the clutch form. The. general or abstract aspect of the invention upon which the subsequent claims depended had been altogether changed. Perhaps it is better to say that for the first time a more abstract aspect had appeared, but at the expense of much that had earlier been included as a part, thereafter an inconsistent part, of the original conception. This alone in our judgment is enough to limit the patent to the disclosure.
Moreover, we think that Leonard’s own conduct of his application shows that he was aware of the departure which his amendments introduced. He .revised his commercial regulator in 1911, so that it conformed to Delano’s principle, to Turbayne’s, and to the defendant’s. Yet from the beginning of 1912, when he sought to intrude into his application vague expressions of wide fluctuation and the like, he never, avowedly attempted to amend his specifications by any disclosure of the new form which he had actually adopted, and which he then certainly meant to claim. Why did he not do this, if it was not because he recognized that such an avowal would be met at once by a refusal in the Patent Office? Such a refusal must have been based upon the ground that it was a departure from the original disclosure, as it certainly was. Instead of this more candid course, he contented himself with nebulous claims, and one or two rather 'cryptic allusions to possible alternative forms. We cannot,accept the result; the place to disclose the invention, certainly if it has become already known and developed, is the Patent Office itself. After not disclosing it there, we cannot allow it to be introduced here, for our duty is to interpret the instrument, not to rewrite it.
[1]. So far we have considered the amendments introduced into the specifications as though their final form' was sufficient to support tire ‘claims in their wider scope. Assuming, as we do, that taken literally the language is broad enough to cover the defendant’s regulator, the further question nevertheless arises whether the disclosure, as it [589] stands, will serve. An inventor is, of course, not confined to the exact details of his disclosure, els.e his patent would be of small value. The extent to which he may generalize it depends, not only upon the sur-roúnding pressure of the art, but the extent to which the variations which he wishes to cover in his claims, are themselves within the initiative of a journeyman in the art. Tor the inventor’s contribution must be a sufficient guide in itself, and its extent is limited to such substitutes for any disclosed element, as the art needs no help to find. The validity of these claims, when so extended, depends, therefore, upon the readiness with which a journeyman would have substituted a resistance coil for the clutch of the application. Put in another form, we may say that it depends upon whether it would take any invention to change Reonard’s clutch into his second and present commercial form.
We are not ready to agree that to quench the current by means of a rheostat was the equivalent of disconnecting the generator from its prime motor. Bentley’s opinion upon that subject invades the precise issue we must decide. Each method did, indeed, accomplish the same result; but they cannot be said to accomplish that result in the same way, and fail as equivalents under one of the most common tests. However, we should not necessarily treat this as final, doubting, as we do, the nnconditional validity of any absolute tests of the kind. If it was apparent that one might freely and certainly borrow a resistance coil from the art as a substitute, we should not hestitate to call it an equivalent, though it was a new means. Yet we think that there are objective reasons to deny that the supposed substitution was of such a kind. First, we may notice that physically it would not have been enough merely to take out the magnetic clutch and substitute a resistance coil in its place; the coil would not have been in the field circuit, and more readjustment was necessary. Yet the necessary changes may have been immediately obvious, once the idea was suggested, and we shall assume that they were, for it is not respecting them that our doubts arise. In part we rest upon the conduct of Reonard himself, both in his original disclosure and in the development of his machine.
Figures 4 and 5, already mentioned, showed a rheostatic regulator, though it was gradually introduced and removed. But with these actually before him, if Reonard did have the idea of a similar regulator for motorcars, why did he not at least verbally suggest it in his specifications as an alternative to his clutch? We have the answer in his advertisements during the spring of 1910, after he had filed his application, and while he was first exploiting his invention. In these he clearly shows that he thought the rheostat an impracticable regulator for a motorcar. Thus in January he says:
“There is no resistance of any kind inserted in series with the field or armature, so there is no chance of the outfit not working on account of poor contacts, sparking, etc.”
Again in March:
“This [the regulation] should be done without the use of slipping friction clutches, speed gears, iiyball governors, electrically operated rheostats, vibrators, slipping fields, etc., which are inherently troublesome.”
[590] Leonard supposed that he had found something better than a rheostat for use in motorcars. He had thought of the rheostat only to reject it, though he meant to use it for other purposes in a complicated mechanism, which contained both clutch and resistance. He did not think it applicable for a motorcar, and it was only by subsequent experiment that he came' to change his mind. We start, therefore, with the reasonable assurance that Leonard did not find the two means at once substitutes for each other.
[2,3] We shall show later that other inventors attempted to. create a pulsating current by a resistance coil, and supposed they had succeeded. If they failed, which is in doubt, we can safely assume that success was not a certainty. At least, the plaintiff has not shown, and the burden rests upon it to show, that any journeyman of the art could turn from one to the other with any certainty of result. Unless it goes so far, it is not enough merely to say that tire substitute would readily occur to the mind of any one skilled in the art. An inventor must do more than give cues for future experiment. Matheson v. Campbell, 78 Fed. 910, 24 C. C. A. 384. Unless he is dealing with elements whose action and reaction is known and certain, he is bound to disclose how the combination will operate. A patent is the reward of a tested contribution to the art, not'of a pregnant surmise or a promising hypothesis. There is not the least reason to say that, if it had occurred to a journeyman in the art to substitute a resistance coil, he would at once have understood the complete effect of that substitution upon the.system as a whole, whether it would operate as well or worse than the clutch. So far as any evidence is available, it was a matter which required further experiment. We have, it is true, in the final form of the patent certain vague statements that other substitutes for the clutch are open (page 1, lines 46-54; page 2, lines 26-30); but nowhere is the substitute set forth in detail, and we are not satisfied, on the facts in this case, that it could be left so in the air for the speculation of those who should use the patent. Upon this ground, also, we therefore think the patent too limited to cover the defendant’s-regulator.
[4, 5] Finally, the art of acknowledged priority either wholly anticipates the patent, or proves, as we have just said, that the result which Leonard now claims was not a mere corollary of what he disclosed. We refer to the Everett regulator (Bliss type F),- which differentiates from the claims, if at all, only by reason of having carbon, instead of metal, contacts. We do not forget that the relays are of a different type, but for the reasons already given we do not regard that distinction as important. Bentley’s theory of the gradual introduction of the resistance in this regulator depends upon the presence of the carbon contacts. Wray’s college thesis bears him out, though we believe that Wray’was himself probably wrong at that time. Bentley’s theory omits, however, certain facts, which require an explanation. The tension springs which hold up the floating plunger were in practice set at a predetermined amount,, which represented the desirable maximum in the work current, generally 50 amperes. This means that when the coil in the solenoid carried-50 amperes the plunger would be enough attract[591] ed against the spring as only just to remain in contact with its fellow; any further increase in the work circuit effected a separation.
Now the work circuit was always broken until the automatic switch closed, and that could never be in the Everett regulator until the voltage in the generator equaled that in the storage battery. Frequently the current which dosed the switch must immediately put the carbon electrodes just at the point of separation, and in those cases the regulator would at all periods operate by the intrusion of the entire corrective at once. Nevertheless, it is of course possible that the battery might be partially exhausted, in which case there would be theoretically an increase of resistance between the carbon electrodes from the timé the automatic switch closed until the point of separation was reached. But this was a very small matter at most; in Sheldon’s opinion quite too small to effect any regulation whatever. Bentley calculated that at the time of separation the resistance was .8 ohms, and Sheldon only .4 ohms. It makes no great difference which opinion we assume to be correct, though Sheldon’s observations were taken under more nearly working conditions. The difference between the resistance at the time when the automatic switch closed and when the carbons separate is at most all that can be said to be graduated. It depended altogether upon the residual voltage found in the battery at the moment when the engine is put in motion.
At any rate, it follows, after the battery has once reached its proper charge, or, if it is already fully charged, at once after the generator starts, that the whole theory of the plaintiff is inapplicable, and the resistance moves in and out in the same way as though the contacts were of metal. This would in practice be the condition over much the greater part of the time. Even at the beginning the graduated resistance would be less than .5 ohms, while the resistance of the shunted coil was nearly .6 ohms. We regard these considerations as substantially disposing of the theoretical demonstration that the Everett regulator was not an anticipation.
We do not forget the presence of two resistances; but that seems to us of no importance, if there was one resistance continuously thrown in and out over a substantial range of armature speeds. We have, moreover, omitted one element considered by Bentley; i. e., what are the variations of resistance while the carbons are sparking? This is a phenomenon which occurs as well when the contacts are of metal as when of carbon, and there are no data which would allow us to pass upon the differences of the resistance from one spark and another.
There remain only the tests made by both sides. As is not uncommon, the tests of each favored his own side. Sheldon found that the amplitude of the Everett pulsations was nearly as great as, in some cases greater than, those of the plaintiff’s system. On the other hand, the defendant’s oscillograms show a very nearly even current line for the Everett regulator. The suggestion that Bentley’s experiments may have been vitiated by the presence of a gold leaf fuse in the oscillograph ought to be valid for all those taken, including the oscillograms from the defendant’s regulators and the Reonard clutch. It must be conceded that the case in this aspect is too doubtful under [592] the stringent rule of proof required, though we are rather disposed to accept the evidence proving that the pulsations took place than that they,did not.
While, therefore, we are not disposed to hold that Everett mechanism is an anticipation of the pulsating current regulator, we think that it is a final answer to the adequacy of the disclosure actually made. More properly, perhaps, we should have discussed it .with that point. There is no doubt that Everett and Bliss both thought it produced a pulsating current, and that Wray’s suppression of that feature was due to Everett’s injunction upon him. There is also no doubt that so experienced a man as Sheldon believed, and still believes, that it operates the same way. All the inventors meant-to get a pulsating current ; all thought they had; all supposed that the carbon buttons would effect that result. If they do not, the substitution of metal buttons will malee the current pulsate. Now, whatever else all this shows, at least it shows'beyond peradventure that it was not a mei-e journeyman’s job to secure a pulsating current by the interposition of a rheostat. Competent men tried it before Eeonard; if they succeeded, then there is nothing new in his invention, except the clutch and relay ; if they failed, he is hardly in a position to say that his disclosure was enough to advise the art how to secure such a current, except by using his clutch. As we have already said, the precise form of the relays was not a part of the invention. One had no warrant in supposing that the invention lay in them. A “plunger” type is as much within the patent as a “trembler” or “air gap.” In following the patent, therefore, there was no warning that the “air gap” was essential. If one used a “plunger,” one might choose either metal or carbon contacts ; Leonard’s specifications said nothing about it. Even if we suppose that the “air gap” relay was a part of the disclosure certain to be followed, Leonard nowhere states that metal should be used for contact between the keeper £8' and the point 31. If carbon chanced to be used, as might well have been by an imitator, he would have had a graduated increase of resistance up to the point of separation. He would have known as little as Everett, Bliss, and Sheldon that it affected the result for which he was striving. There was not the least indication that one must follow their structure, or that the material made any difference.
We conclude that the patent as now presented to us is a very ingenious and plausible effort to extend a quite limited disclosure into a broad monopoly. It is clear that Leonard had nothing of the sort in mind until by subsequent experiments he had worked out his successful device. Meanwhile the art had already passed beyond him, and his only alternative was to file a new application or proceed with his own. We have, of course, nothing to say in criticism of his actual disclosure, and his claims so far as they are limited by that disclosure. There is not the least reason to question the validity of the invention of a clutch control. Only it is derived from a quite different method, that affecting the transmission of mechanical energy, not the transformation of mechanical energy into electrical.
[593] The decree is reversed, and the bill dismissed for nouinfringement, with costs in both courts.