The plaintiff appeals from a judgment, summarily dismissing its complaint for the infringement of a patent issued to one, Cisin, for an “Amplifying Circuit.” The only question is whether the claims in suit are invalid because they were introduced into the application more than two years after the subject-matter which they cover had been publicly sold. The disclosure is of an electric circuit' which will increase the volume of the electrical pulses produced in a second circuit “in the usual manner by the action of sound waves on the transmitter” (page one, col. two, lines 49, 50). The first circuit contains a “rectifying tube” and an “amplifying tube” in series, set in the usual house circuit, which by hypothesis may carry either direct or alternating current. When a direct current is used, the “rectifying tube” merely allows it to pass through unimpeded. When an alternating current is used the “rectifier” stops, or “strains out” the pulses of current in one direction, letting the current which arrives in the opposite direction pass through to the “amplifier.” Between the “rectifier” and the “amplifier” is a “filter,” designed to “smooth out” the pulses of the “rectified” current. The anode of the “rectifying tube” contains not only the usual anode “plate,” but a “grid” such as is necessary in an “amplifying tube” in order to impress upon the amplifying current the pulses of the current which is to be amplified. In the disclosure this “grid” is connected directly with the input lead of the house circuit to which the “plate” of the anode is connected, so that it is electrically unnecessary. (This connection is apparently what the claims mean when they speak of the “plate” and “grid” as being connected “in multiple.”) The specifications do not disclose why the “rectifying” tube should contain the “grid”; but it was explained at the bar that, as such tubes were on the market for use in “amplifying” tubes, they could be made also available for the “rectifier” by the mere expedient of connecting the “grid,” as we have described. In parallel with the circuit containing the electrodes are two heating “filaments,” one in each tube to heat each cathode, as is well understood; but, since either a direct, or an alternating, current will heat the filaments equally well, there is no need* of a “rectifier” in this circuit. In order “to prevent burning' out of the filaments” in both tubes, resistances in series are added between the filaments and the source of power, and one o$ these “may be a variable resistance, if desired” (page one, col. one, lines 50, 51).
Tlie claims in suit may for convenience be divided into two groups: one, consisting of Nos. 3, 4 and 5; the other, of Nos. 11, 12 and 13. The only difference between claim 3 and claims 4 and 5 is that claim 4 includes the resistance just mentioned and claim 5 specifies it with particularity. There is even less difference between claim 11 and claims 12 and 13: all three contain the same elements, the only distinctions being that the details in the description of each are somewhat varied. The disclosure remained the same —with one altogether trivial exception— throughout the proceedings in the Patent Office, and the defendant relies altogether upon amendments made to the claims. These were, it says, both to expand and to narrow their scope, and both invalidated the claims in suit as they finally emerged. The supposed expansions were two: the first was the change from describing the anode of the “rectifying” tube as consisting of “a grid and a plate connected in multiple,” to describing it merely as an “anode.” The second was that the only claims filed before May, 1936, which mentioned any resistance in the heater circuit — original claims 7, 8 and 9 — -spoke of it as a “variable resistance”; while all the claims in suit speak of it merely as a “resistance.”
So far as concerns the change in the description of the anode of the “rectifying” tube, we are unable to say on this record that it enlarged the scope of the claims at all, if proper allowance *525be made for possible equivalents. As we have said, the “grid” had no function; a “rectifier” which left it out would accomplish the same result in substantially the same way; and every patent is entitled to some range of equivalents. Claude Neon Lights Inc. v. E. Machlett & Son, 2 Cir., 36 F.2d 574. Upon a motion for summary judgment it was improper to deny the plaintiff the right to a trial, for we cannot know to what equivalents Cisin’s original claims would have been entitled; that is a question which can never be determined until the position of the invention in the whole art has been made plain. In this connection it must be remembered that we are concerned with the meaning of the original claims as they were introduced when all of them described the anode of the “rectifier” in the same words. The equivalents, properly allowable for those words at that time, may have been considerably broader than would now be allowed to them as they appear in claims 1 and 2 as issued. That would be true, because after the claims in suit were added, the contrast in locution between them and claims 1 and 2 might compel us to give claims 1 and 2 a more limited scope than they would have had, had all the original claims been allowed as they read. As we are not prepared to say that the original claims would not have covered an anode without a “grid,” we cannot hold that the amendments did more than clarify the meaning of the original claims.
As to the amendment which substituted “resistance” for “variable resistance” in the heater circuit, the disclosure was of a “series of resistances” in that circuit, and the text did not prescribe that any of them should be variable, but only, as we have said, that “one of these resistances may be a variable resistance if desired” (page one, col. one, lines 50, 51). The theory is that since the original claims, which mentioned these resistances uniformly spoke of “variable resistances” they would not have covered fixed “resistances” in the heater circuit, even though these were designed so as to be “sufficient to prevent burning out the filaments” (page one, col. one, lines 48, 49). But once mor'e, we cannot know whether a “resistance,” properly designed for the heater circuit, does not produce substantially the same result by substantially the same means as a “variable resistance” so set that it will prevent the burning out of the filaments. A priori we should suppose that it did.
Moreover, the argument as to both these verbal changes falls to the ground for another reason: i.e., both the supposedly new features appeared in the claims less than two years after the device was sold publicly. On March 23, 1933, in order to institute an interference, Cisin filed two claims — Nos. 10 and 11 — borrowed from the pending application of one, Wuerfel. In these the anode of the “rectifier” was described merely as an “anode,” and in claim 11 the resistance in the filament circuit was described as “a resistance.” The examiner rejected these at first for a purely verbal defect which Cisin corrected on May 3, 1933; the interference was declared on June 22, 1933, and the claims were again amended on July 25. A new application of one, Mavrogenis, was brought into the interference on January 12, 1934, and the invention was eventually awarded to Mavrogenis by the Examiner of Interferences, whom the Board of Appeals affirmed. The record does not contain the date of either ruling; and all we know is that the decision of the Board of Appeals must have been before June 5, 1936. On May 28, 1936, Cisin filed claims which, with unimportant changes, became claims 3, 4 and 5 in suit; and on December 23rd of that year, he filed claims 11, 12 and 13 in suit, and the patent was allowed on January 28th, 1937. Thus it appears that there were pending from March 23, 1933, until an unknown date before June 5, 1936, claims which covered the two putative expansions of the original claims; and that on May 28th new claims were introduced perpetuating these features. We are unable to say upon this record whether there was ever a period from March 23, 1933, forward during which Cisin was not claiming the anode of the “rectifier” merely as an “anode”; and the second resistance in the heater circuit merely as a “resistance.” We cannot understand by what reasoning the defendant argues that because Cisin “can-celled” these claims, they did not stand as part of his invention before he did so. It is true that he must have been satisfied that he could not prevail against Mavrogenis in the courts; and that the claims which he surrendered were for combinations of different elements from those of the claims in suit. But neither of these circumstances was relevant, because, since *526the claims in suit were for combinations of the same elements as the original daims, and differed only in the amended definition of those elements, it was enough if those amended definitions were contained in other claims until the daims in suit were filed. The amended definitions were evidence that Cisin did not understand that the elements so defined were to be confined to his original description of them, and he for that reason was free to incorporate them into his original claims.
We have hitherto assumed that the defendant was right in supposing that an applicant may not enlarge his daim to include any subject-matter whatever not theretofore included, which has been publicly sold for two years before the amendment. That states the doctrine too absolutely.
In Miller v. Brass Co., 104 U.S. 350, 26 L.Ed. 783, the court laid it down that there might be excuses for a delay of over two years in applying for enlarged claims by reissue, in spite of “intervening rights”; i.e., of the fact that the art had meanwhile occupied the field covered by the enlargement. This apparently required more than a tardy awakening to the insufficiency of the original claims, and the doctrine was steadily applied in many later cases; as for example, Mahn v. Harwood, 112 U.S. 354, 5 S.Ct. 174, 6 S.Ct. 451, 28 L.Ed. 665; Wollensak v. Reiher, 115 U.S. 96, 5 S.Ct. 1137, 29 L.Ed. 350; Ives v. Sargent, 119 U.S. 652, 7 S.Ct. 436, 30 L.Ed. 544. Chapman v. Wintroath, 252 U.S. 126, 4 S.Ct. 234, 64 L.Ed. 491, was at one time thought to have put an absolute limitation of two years upon reissues, though that notion was dissipated by Webster Electric Co. v. Splitdorf Electrical Co., 264 U.S. 463, 44 S.Ct. 342, 68 L.Ed. 792. However, that decision likened a divisional application— which is only one form of amendment— to a reissue in this respect, and held that, when broadened claims were filed more than two years after the original application, and when' there had been “intervening rights,” the delay must be explained in the same way as though it were a reissue. Crown Cork & Seal Co. v. Gutmann Co., 304 U.S. 159, 58 S.Ct. 842, 82 L.Ed. 1265, established, as a variant in the case of divisional applications, that no excuse was necessary, if there had been “no intervening rights”; but otherwise left the law as before. We need not .say whether Schriber-Schroth Co. v. Cleveland Trust Co., 305 U.S. 47, 59 S.Ct. 8, 83 L.Ed. 34, and Muncie Gear Works Inc. v. Outboard M. & M. Co., 315 U.S. 759, 62 S.Ct. 865, 86 L.Ed. 1171, have made any change, for in each case there had been “intervening rights.” Moreover . the new claims were for “new inventions,” however that phrase be defined. In both not only had the disclosure itself been amended, but the claims contained new elements — not by way of limitation only — which made the combinations quite different from any originally claimed.
Nor need we say that, though no excuse is given for the delay, the doctrine applies if the change is not to a “new invention” but involves only a minor change necessary to secure complete protection for what the applicant originally intended to reserve to himself. Possibly he may have as much in spite of “intervening rights.” The doctrine is designed to protect the public against abuses, not to deprive inventors of what they plainly never meant to put into the public demesne. It is enough here to say that it certainly does not prevent amendments which go no further than to make express what would have been regarded as an equivalent of the original; or to incorporate into one claim what was to be gathered from the perusal of all, if read together. This is the situation here as it comes to us upon this record.
So much for any increases in the scope of the claims. Little need be added as to the limitations by which their scope was narrowed by the introduction of new features: as, for example, the separate “envelopes” for the rectifying, and amplifying, tubes. The addition of a new element in a combination, while all the rest remain, releases a part of the disclosure to the public demesne, and to that there can obviously be no objection. Applicants ordinarily begin with as broad claims as they can hope to sustain; and they retreat progressively as the examiner forces them by the prior art he turns up in the Office. No other course is really open to them, for, although the invention must in the end be judged in the light of the whole mass of earlier inventions inside and outside of the Office, the inventors themselves can ordinarily know only a very small part of it. If, when they yield any part of what they originally believed to be their due, they substitute a new “invention,” only two courses will be open to them: they must at the out*527set either prophetically divine what the art contains, or they must lay down a barrage of claims, starting with the widest and proceeding by the successive incorporation of more and more detail, until all combinations have been exhausted which can by any possibility succeed. The first is an impossible task; the second is a custom already more honored in the breach than in the observance, and its extension would only increase that surfeit of verbiage which has for long been the curse of patent practice, and has done much to discredit it. It is impossible to imagine any public purpose which it could serve.
Judgment reversed; cause remanded.