DENISON, Circuit Judge.
Infringement suit on patent for electric motor controller regulator, issued January 16, 1906, to Weyand, as No. 810,240, and reissued to Electric Service Supplies Company, January 28, 1908, as reissue No. 12,744, upon reissue application filed March 18, 1907. The first nine claims of the reissue are identical with the nine and only claims of the original; the remaining reissue claims 10-32, are of seemingly broader character. Infringement is alleged of claims 1, 2, 11, 25, 29, and 30. The District Court held claims 1, 2, and 29 to be valid and infringed, and claims 11, 23, and 30 to be void, because broader than the art permitted. Both parties appeal.
The invention pertained, commercially, to that form of controller used by the motormen of electric cars in turning the current on and [458]*458off. All parties accepted, it as essential that the “on” motion should be step by. step, with a distinct stop between each two steps, while the “off” motion should be — or at least the motorman should be left free to make it — a continuous sweep. The purpose of the class of devices to which those of both parties belong is to compel the motorman to use the step by step method for the “on” motion, instead of leaving this result.to his voluntary and skilled manipulation. This result is accomplished by devices in connection with the (customarily) vertical controller rod and the removable operating handle which, when placed in position, suitably engages the rod and so malees it revolve; and more specifically by attaching, rigidly to the top of the controller box and surrounding the rod, a plate called a base-plate, and by attaching to the rod, so as to revolve with it, a similar plate called a cover-plate. Between these two plates were inserted pawl and ratchet devices; the pawl or dog being upon the lower side of the cover-plate, and the ratchet teeth or stops being carried by the upper face of tire base-plate. All these were old; the questions involved pertain to the construction and arrangement of detent and stops, of handle and cover-plate, and of cover-plate and base-plate.
The first question we meet is whether the reissue was invalid, because broadened. It was applied for within 15 months from the original issue, and neither laches nor abandonment can be urged against it from mere lapse of time. The reissue was very frankly for tire sole purpose of broadening the claims. No consequential error in the specification was suggested, and the only “inadvertence, accident, or mistake” alleged was the omission to make the additional claims which were said to have been necessary to make the monopoly or protection of the patent as broad as the invention disclosed by the specification and drawings. The mistake or inadvertence was that of the applicant, induced by a,negligent or unskillful solicitor,
The question is most squarely presented by claim 29, as follows:
“In an electric controller regulator, the' combination with a controller of a base mounted thereon and provided with cams and stops, a rotatable cover on the base having a pawl which co-operates with the cams and stops to intermittently arrest the cover, a jaw on the cover, an operating handle engaging the controller rod and jaw to connect said rod and cover, and a coupling to secure the cover to the base without interfering with its rotation.”
When this is compared with other claims, it is seen that its identifying thought, and so what must be deemed the invention of this claim, is the use of the handle to lock together and cause to turn in unison the cover-plate and controller rod; and this, of necessity, carries the implication that, when the handle is not in position, the rod and cover-plate will- be disconnected. The claims of the original natent did not reach and protect the invention thus described. Two or three of the claims included, as elements, directly or by necessary implication, the handle and the controller rod and the jaws for locking the. handle to the cover; but in such claims these elements were grouped in combination with a large number of other elements not involved in this thought, and it follows that these claims were so narrow as not to protect broadly the idea of making the rod and cover coupling in this way.
[459]*459[ 1 ] Defendant urges that reissue for the purpose of such expansion is invalid — even if there are no “intervening rights” — and presses upon us the Supreme Court and the Seventh Circuit decisions below cited. The question has not been as expressly treated in this circuit as seems desirable, and we proceed to its consideration. If it were an original question, it would, as we see it, present no difficulty. The Patent Act of 1836, by section 13, provided that:
“Whenever any patent * * * shall be Inoperative or invalid by reason of a defective or insufficient description or specification, or by reason of the patentee claiming in his specification as his own invention more than be had or shall have a right to claim as new, if the error has or shall have arisen by inadvertency, accident or mistake, * * * it shall be lawful for the Commissioner * * * to causo a new patent to bo issued to the said inventor, for the same invention * * * in accordance with the patentee’s corrected description and specification.” Act July 4, 1836, c. 357, 5 Stat. 122.
The act of 1870, by section 53, uses practically the same language, save that “specification” is substituted for “description and specification.” The statutory conditions thus become (1) “inoperative or invalid;” (2) “a defective or insufficient specification;” (3) “inadvertence, accident, or mistake;” and (4) “the same invention.”
It is well understood that at the time of the first statute “specification” was an inclusive term, covering what, in later nomenclature, are identified as ‘‘specification” and “claims”; and though in 1870 the distinct identity of claims was recognized, it is not to be supposed that the re-enactment of this part of the act of 1836 contemplated any different definition of the terms used than had been understood and accepted with reference to the same terms in the earlier statute. Indeed, the last edition of Walker on Patents (4th Ed. §§ 111,' 173) says that the claims are part of the specification. See, also, Macomber, pages 711 and 822.2 Thus giving to the statute its necessary interpretation when applied to the present-day subdivision of the statutory word “specification,” we find it distinctly declared that a reissue may be granted when the patent is inoperative by reason of defective or insufficient specification or claims.
[2] The statute does not say “wholly inoperative”; and although a patent may, in the strict sense, be operative if it grants any monopoly of anything, yet if it fails to secure to the inventor the monopoly of his actual invention, it surely does not operate according to the intent of the law. Thomson v. Wooster, 114 U. S. 104, 115, 5 Sup. Ct. 788, 29 L. Ed. 105; Giant Co. v. Nitro Co. (C. C., Sawyer, C. J.) 19 Fed. 509, 510. Further, we find the statute saying, “by reason of a de[460]*460fective or insufficient specification or by reason of the patentee claiming as his own invention or discovery more than he had a right tc* claim as new/’ and this means (interpreting “specification” according to its. true force) inoperative by reason of defective description or insufficient claims or invalid by reason of too broad claims.
[3] “Inadvertence” is a word of broad meaning. The act of the patentee, through his solicitor, in drafting or accepting claims not commensurate with the invention, may be considered a mistake of law or a mistake of fact. In the typical case it is rather of the latter character, and is made because he does hot appreciate the part which certain elements play in the operation and the value of the device described; but whether we call it a mistake of law or fact, it cannot very well escape the appellation “inadvertent,” 3 unless, in truth, it is considered and' deliberate. Leggett v. Avery, 101 U. S. 256, 260, 25 L. Ed. 865; Yale v. Berkshire, 135 U. S. 342, 403, 10 Sup. Ct. 884, 34 L. Ed. 168; Grand Rapids Co. v. Baker (C. C. A. 6) 216 Fed. 341, 351, 132 C. C. A. 485. This latter question — whether the error really was inadvertent or. was deliberate — is primarily committed to the Patent Office for its decision. If the Patent Office, by general rule or by action in a particular case, should hear conflicting evidence upon that question and decide it, the courts would not lightly review that decision. None the less does this result follow when the Patent Office sees fit to accept and act upon an ex parte affidavit. It is only when other parts of tire record (as the former, deliberate abandonment of the broader claim) conclusively show that the error was not inadvertent that there seems to be reason for disregarding the prima, facie case made out by the Patent Office decision.
[4] The further and last statutory condition is that the reissue must be for -“the same invention.” It is true that, for purposes of determining infringement, the identity of the patented invention is fixed by the claims; but to apply .the same test to identity of invention as between original and reissue loses sight of the difference between the real invention and the originally patented invention, and unless there is such a difference, there is no occasion for reissue. To recognize that difference and permit it to be corrected is tire whole purpose of the reissue statute; and so it seems quite destructive of the statute to assume that the identity of the actual invention is permanently declared and fixed 'by the form which the original claims are' inadvertently allowed to take. In the same way as with reference to mistake, the question of identity is submitted to the Patent Office, and for the same reason its conclusion is to be taken as prima facie right. The last sentence of section 53 even permits the Patent Office, in certain cases, to go entirely outside the record to determine what the original invention was. It follows that only when it is clear that the [461]*461reissue is not for the same invention are the Courts justified in reaching that conclusion; and we take this to be the rule of the decisions hereinafter cited.
[5] We thus see that, apparently, the statutory inoperativeness, insufficiency, and inadvertence may exist, even though the only correction to be made is to broaden the claims by omitting unnecessary limitations, and that the particular invention first claimed in the reissue, may be sufficiently disclosed in the original merely by specification and drawing, and without express verbal reference thereto. To the contrary effect, upon one or more of these points, defendant cites the cases in the margin,4 and insists that they establish invalidity in a reissue made for the mere purpose of broadening the claims, if in a case where the first description contains no statement of the broader concept. Some of them do contain language to this seeming effect, yet we think in each the result will be found sufficiently controlled by laches, intervening rights, rewritten and altered description or drawing or other clearly distinguishing fact. Further, it is, in some aspects, a question of fact whether the -reissue is for the same invention, and these decisions are, in some measure, conclusions of fact'applying to the instant case rather than of law making a rule for future cases. However all this may be, if they would otherwise have the effect claimed, they are. inconsistent with Topliff v. Topliff, 145 U. S. 156, 171, 12 Sup. Ct. 825, 36 L. Ed. 658, where it is expressly declared that there may be a valid reissue for the purpose only of broadening claims, and where, if the original description indicated any broader view of the invention than the original claim declared — save by describing the actual structure — that fact was not regarded as important enough to mention in the discussion. We are satisfied to hold that the necessary insufficiency of specification may consist only in that the claims are too narrow; that the necessary inadvertence or mistake may be found solely in the action of the patentee or his solicitor in framing these too- narrow claims; that the original and reissue may be for the same invention although the original specification does not expressly indicate that the invention was regarded more broadly than the claims declared it; that upon all these questions, the decision of the Patent Office makes a prima facie case; and that only in a clear case of (actual or presumed) deliberate acceptance of the narrow claims or actual or presumed intervening rights or a clear case. of departure from the real original invention as it may be found in the description and draw[462]*462ings will the courts say that the reissue is invalid. Seé cases cited in margin.5
If the original specification and drawing disclose a structure with five elements, and it is apparent that three of them make an "Operative combination, the other two being useful but unnecessary additions— though the possible use of three only is not mentioned — and if the claim includes all five, and so is limited to the more complex combination, there is in these facts alone no bar to a reissue securing the simpler combination of three elements of which the patentee was in truth the first inventor. Upon principle, we cannot escape this conclusion, and we find no controlling decision to the contrary. Of course, if the original claims did not touch that part of the structure to which the reissue claim pertains, a somewhat different question would arise. If the opinions cited from the Seventh Circuit (General Co. v. Richmond Co. and McDowell v. Ideal Co.) carry implications against this conclusion, they must be read with the decision of the same court in Moneyweight Co. v. Toledo Co., 187 Fed. 326, 109 C. C. A. 586, in which it is held that the original claims do not fix the identity- of the actuál invention. Sinde the claim constitutes the only statutory medium for declaring the scope of the invention, as distinguished from its .structural embodiment, it seemingly must follow that an omission to specify its true breadth in the place where that is not required — the description — cannot be more fatal than the same omission in the claim, where the formulation is required.
[6] Defendant next urges that there were “intervening rights” which prevented such a reissue, and relies upon Clements v. Odorless Co., supra, and similar cases, as showing that the three patents issued between the date of original issue and the date of application for reissue, sufficiently constitute or evidence those intervening rights which will bar a reissue. The reissue statute fixes no limit of time within which the application must be made, and does not specify that rights, accruing to the public or to individuals, after the first issue and before the application for reissue, shall affect the latter right. Miller v. Brass Co., 104 U. S. at page 350, 26 L. Ed. 783 (see Walker [4th Ed.] § 226), which first declared this additional condition, did not elaborate the foundation reason for so doing, but we think it must be assumed that the principle of estoppel (perhaps in favor of the public generally) is at the bottom of the rule. By the issue of a patent, the inventor dedicates to the public everything which he does not claim [463]*463as his monopoly. Upon this dedication, the public has a right to rely, and if members of the public devote time and money to the manufacture of a device which the inventor has so dedicated, or to the devising, inventing and patenting of structures which embody such a feature, it may be presumed that this is done upon the faith of the dedication; and so the inventor may not be permitted thereafter to enlarge his monopoly to the prejudice of these new rights, — even though, except for them, the reissue would be permissible. The settled doctrine has come to be that from a delay of more than two years, and in the absence of any sufficient contrary evidence, these fatal intervening rights (public or private) will be presumed; in the presence of less delay, they must be proved. But see White v. Dunbar, 119 U. S. 47, 52, 7 Sup. Ct. 72, 30 L. Ed. 303; and Milloy Co. v. Thompson Co. (C. C. A. 6) 148 Fed. 843, 847, 78 C. C. A. 533.
Since a patent grants or creates no right to manufacture the thing patented, but only a right of exclusion (Swindell v. Youngstown Co., 230 Fed. 438, - C. C. A. - (C. C. A. 6—Feb. 8, 1916), it would seem that intervening patents, not followed by manufacture, can furnish the necessary basis for this rule (in a case of less than two years’ delay) only because of the effort and expense involved in making and perfecting the inventions to which they relate; in other words, because the absence of the broad claim in the earlier patent has encouraged or misled the later patentee to make and perfect an invention which he would not have undertaken if the earlier monopoly had been wider.
[7] In this case, each, of these three patents was applied for before the issue of the Weyand original, and so, at the time of such original issue, was a pending application showing an invention then completed and reduced to practice. Such patents, which merely go to issue in the interval, cannot raise the estoppel upon which this doctrine must at last depend — though it may be called “public policy.” There is no evidence that the inventors under any one of them manufactured, and so they do not seem to have done anything in reliance upon the apparent dedication made in the first patent, excepting merely that they paid their final fees. To this action they were measurably committed before the original issue of the patent in suit; and we cannot think it, of itself, enough to bar the reissue. For this conclusion, there is the additional reason that the only one of the broadened claims which we find occasion to consider and enforce has to do with a feature of the device which was not used by any one of these three intervening patents. Their structures do not disclose any infringement of the reissue in this respect, and so it cannot be that they evidence any “intervening right” pertinent to that claim.
We need not determine still another supporting reason urged by the plaintiffs for the same result, viz. that the three intervening patents were either issued or assigned to and are now owned by plaintiffs themselves, and that the general public cannot claim through an intervening inventor who is himself not prejudiced. See Hartshorn v. Barrel Co., 119 U. S. 664, 674, 7 Sup. Ct. 421, 30 L. Ed. 539.
Before coming to the questions of validity and infringement, which we must decide, we note the differences between plaintiffs’ and defend[464]*464ant’s constructions. It does not seem necessary to describe these in detail. Plaintiffs have the main body of the pawl pivoted under the cover, swinging horizontally, engaging stops projecting horizontally-inwardly from the cylindrical wall of the base-plate, and held normally out of engagement by a spring which allows yielding in either direction from the normal. A post rigidly attached to and depending from the pawl, engages a series of cams on the upper surface of the base. In the “on” motion, this part contacts with the outer sides of the cams, and carries the pawl outwardly into engagement with its stops; on the “off” motion, the post passes on the inner side of the'cams, and the entire unitary pawl and post structure yields inwardly and is not stopped. Defendant carries both stops and cams in concentric rows on the upper surface of the base; his main pawl, pivoted between ears depending from the cover, swings vertically, and is held normally up out of engagement with the stops by the weight of its tail, and for the depending cam engaging post, he substitutes a supplementary pawl, horizontally pivoted and depending vertically from the main pawl, and retreating vertically within it on the “off” motion.
[8] Claim 1 of the original and reissue patents (quoted in the margin 6 is a long claim, naming a large number of elements, and, as a matter of first impression, appears somewhat specific; at the same time, if the effect of the reissue were not considered, and if the claim were given a fairly broad range of equivalents, it would be infringed. Due regard for the reissue proceedings and what was accomplished thereby prevents allowing to this claim the liberality of construction that it might otherwise receive. It is, in respect of the features herein involved, the broadest of the original nine claims. The reissue was obtained on the express representation and the sole ground that the claims were inadvertently too narrow, and when this first claim is compared with several of those contained in the reissue, it is too clear for doubt that both "the patentee and the Patent Office regarded the element “pawl,” in the first claim, as entitled only to a narrow range of equivalents, and considered the base and its cylindrical wall as distinct and separate elements. These are the meaning and construction which the patentee deliberately selected for his first claim; he cannot now depart therefrom and say that some of his new claims in the reissue are no broader than his first claim already was; and this, we think, would be the necessary effect of a finding that the first claim is infringed.
The second claim is more specifically confined to the form of pawl and post involved than is the first claim, and it is governed by tire same [465]*465considerations. The decree below must be reversed as to these two claims. Claim 11 is as follows:
“In a controller regulator, the combination with a series of stops, a rotatable member having a pawl movable only in one plane, a series of cams for moving the pawl into engagement with the stops to effect a step by step movement of said member in one direction, and provision for permitting uninterrupted movement of the said member in the opposite direction.”
The. court below held this claim void as anticipated by the patent to Asbury. We agree that it would be void, if it was given a scope sufficient to cover defendant’s form; but we think the claim may well be read so as to have validity over Asbury and still not be infringed. Its force turns on the meaning of the limitation “movable only in one plane.” The claim, containing this limitation, was rejected on reference to Asbury; the applicant then pointed out that in Asbury, while the pawl moved only in what was practically one horizontal plane, yet that its carrying device was swung around the circle, and so, with reference to the part corresponding to the base-plate, it moved in and out in different radial planes. For this reason, the limitation was said to distinguish. This argument did not, in truth present any distinction between Weyand and Asbury. Where two parts are relatively changing, and we speak of the plane of motion of an attachment to one part, it is obvious that though the plane will remain the same as to the attached part, it will constantly change as to the other part. In As-bury, the pawl motion was always in the same plane as to the pawl carrier, but was in successive radial planes as to the stop-carrying part; just so, in Weyand, the pawl motion is always in the same horizontal plane as to the pawl-carrying cover, but is in successive radial and chord planes as to the stop-carrying base-plate; and just so with defendant — the motion of his main pawl is always in the same vertical plane with reference to the carrier, but is in successive vertical chord planes as to the base-plate. The-true distinction between Weyand and Asbury — and this distinction was imported into Weyand’s claim by this limitation in connection with what was already in the claim— was that Weyand provided for his pawl a stop-engaging motion in one plane, and then a further cam-avoiding motion by the same pawl in the same plane continued. We have, then, a case where the applicant was entitled to have, over the reference, the exact claim which was granted, but where he secured the allowance by presenting a view of the meaning of his claim, which, if correct, would have made the allowance improvident. Should wc construe the claim as the applicant did, or should we say that the result obtained by the examiner should be sustained because it was right although he apparently was moved by a wrong reason ?
We conclude that it is unnecessary to pass upon this perhaps novel question, because, if the claim is valid on the suggested theory, defendant does not infringe. It is true that his main pawl moves in a vertical plane, and his supplementary pawl retreats vertically into the main pawl, but neither this supplementary part nor the main pawl has any further motion in the plane of the first motion as extended; the main pawl has no cam-avoiding motion whatever; and it seems inci[466]*466dental rather than characteristic that tire independent motion of the part which corresponds to plaintiffs’ post is in a vertical plane instead of swinging in a right-angled plane. If defendant could avoid infringement merely by changing the pivot angle of this swinging part, there would not' be much left in the claim. The characteristic and meritorious thought formulated by this claim and upon which alone, if at all, its validity must stand, has not been used by defendant.
Claims 25 and 307 clearly must depend for patentability, as compared with confessedly old combinations alone covered by all other portions of the claims, wholly upon the addition to such older combinations of “means for connecting the cover directly to the base to prevent separation thereof without interfering with the rotation of the cover.” The District Court held these claims void because anticipated by Barrett. In fact, the structure of these claims is like Barrett in that it is for the same general purpose and has all the general parts, with the base-plate attached to the top of the controller' box and the cover rotating on the base-plate and a handle attached to the controller rod and actuating the cover-plate. In Barrett, however, the cover-plate was kept from accidental or unintended removal from the-base-plate by the fact that the base-plate, through the' base was fast to die controller rod and the cover-plate was attached to the same rod by a set screw. Weyand discarded the set screw, and made an annular groove in the outside cylindrical wall of the base-plate and attached to the periphery of the cover-plate a depending clip or ear having a lug which traveled in the groove in the base-plate. When he wished to separate the parts, he unscrewed these ears from the cover-plate. If Barrett does not anticipate, it is only because his connection is less direct; but if we assume that there might be invention in the specific means adapted by Weyand for this engagement and 'which might validate some of the claims which refer to such specific means, we can see nothing patentable in the broad ° conception which he attempts to monopolize by claims 25 and 30. , Mechanism for fastening a revolving part to a stationary part, so that the former might revolve upon the latter, but could not get away, was so familiar to every mechanic in every art that we may take judicial notice of it as a common expedient, upon the addition of which to an existing combination of base and cover-plate, invention cannot be predicated. Every wagon wheel and axle responds to this description, and the exhibits in the case show several -familiar methods of getting this result; indeed, one of defendants’ forms uses what is only a cap or retaining flange on an axle.
[9] There remains for consideration claim 29, quoted above. We think this claim valid both as against the defense of reissue and against the defense of noninvention. Every part to which it refers is fully shown in the drawing of the original and sufficiently described in the [467]*467specification. The latter makes clear that the handle is intended to be removable, and although it does not in so many words say that the controller rod is locked to the cover-plate when the handle is in position and that the two are disconnected when the handle is removed, this result is inevitable from the construction shown. The original specification does not affirmatively and expressly indicate that the invention included this feature, nor point out the purposes and commercial advantages of this construction; but as it is clear that the specification is not the appropriate place for declaring scope or equivalency, and that the patentee is entitled to all the benefits inherent in his patented construction, though he did not mention them and even though he did not know of them, it is not apparent how his failure to describe these advantages can be any more effective disclaimer of invention therein than is his failure to claim, or can bar his right, upon a reissue any more than upon an original, to make claim for the actual construction shown. However, if the fact that the original claims do not refer at all to those parts covered by the new reissue claim may tend to show that the latter patent is not for the same invention, that makes no difference here. The original claims, as above mentioned, do — imperfectly — cover this rod-cover-handle construction.
It was not a new thing, broadly speaking, to connect a revolving-part to a central shaft by a bar removably locking to- both so that when it was in position they would revolve together, and when it was removed, they would be disconnected. .This construction was old in turret lathes — not elsewhere, so far as shown — but it had not been used for controller boxes or for any analogous purpose, and we cannot assume that it, like the new subject-matter of claims 25 and 30, was a common expedient. It had special utility for this use, because when the parts have been arranged on the general plan here followed, it is necessary that the cover-plate and controller rod should be fast and revolve together. If they are fastened in some semi-permanent way, as is done in each one of the several earlier controllers shown, it follows that, when the motorman takes out the handle and leaves the controller box, it can still be operated by any one who comes along; the cover-plate can easily be grasped by the hand and turned and the car set in motion; but with the construction which Weyand was the first to use, when the motorman removes the handle, the cover-plate and rod are automatically thereby disconnected, and any trespasser may revolve the cover-plate as much as he pleases without affecting the controller rod. While this is a simple step, we think that it involved invention — for anything shown by this record. The infringement is obvious.
Some complaint is made because, by the final decree, complainant was given only nominal damages; but the final decree recites that this action was taken after “further hearing.” Counsel concede that proofs were taken bearing thereon, and those proofs have not been brought here. We cannot presume error.
The decree below will be modified so as to direct injunction on claim 29 only, so as to find that claims 1, 2 and II are not infringed and that claims 25 and 30 are invalid. The injunction must also be upon [468]*468condition that plaintiff, within 30 days from the mandate, or such further time as may be allowed by the court below, file a certified copy of a disclaimer as to claims 25 and 30. Herman v. Youngstown, 191 Fed. 579, 588, 112 C. C. A. 185. The patent having been invalid in the form in which suit was brought and maintained until final decree, the plaintiff can recover no costs in the District Court (Houser v. Starr, 203 Fed. 264, 275, 121 C. C. A. 462); in this court, since the injunction, although continued, is so far modified, defendant will recover one-half his costs on his appeal.