Mr. Charles Neave, of New York City, for petitioner. [256 U.S. 668, 669] Mr. Livingston Gifford, of New York City, for respondent.
Mr. Justice CLARKE delivered the opinion of the Court.
This is a suit for infringement of letters patent of the United States, No. 743,206, granted to August Weber, Sr., on November 3, 1903. The District Court held claims 1 and 4 valid and infringed, but the Circuit Court of Appeals, while affirming the validity of the claims, reversed the holding that they were infringed. A supposed conflict of this decision as to infringement with one by the Circuit Court of Appeals of the Second Circuit, with respect to the same patent, serves to bring the case here for review on writ of certiorari.
The patent is a simple and, as we shall see, a narrow one. As described in the specification it relates to new and useful improvements in incandescent electric lamp sockets, the principal object of which is to provide a simple and effective automatic lock or connection between the sleeve and the cap of such a socket.
Nothing new electrically or in the general form of the cap or sleeve is claimed-the invention relates solely to the method of fastening the overlapping cap to the sleeve, or, as it is sometimes called, the shell or casing.
The validity of the claims involved is conceded in argument, and, having regard to the disclaimer with respect to claims 2 and 3, which has been filed by the petitioner since the decision by the Circuit Court of Appeals, we regard the issue as now narrowed to the infringement of the fourth claim.
The familiar incandescent lamp consists of a glass bulb, attached to a screw threaded base, adapted to be screwed into a properly insulated block, through which the necessary electrical connections are made, and which is in closed [256 U.S. 668, 670] by a cylindrical 'sleeve' of sheet metal. In order to compete the lamp this detached sleeve supporting the bulb and inclosing the insulated block must be fitted and securely fastened into the usual cap, which is attached to the wall or fixture, for if the fastening should become loose, the light might fail, or the bulb fall, or a fire hazard be created. To invent a device by which the sleeve and cap could be easily and securely attached to each other and yet be readily detached when desired, was the problem to which the patentee addressed himself, and his solution of it is thus described in the fourth claim of the patent:
To make the device thus described, it was only necessary to cut two transverse slits, each about one-quarter of an inch in length, in opposite sides of the flange of an ordinary cap, and to then, with a punch or die, force outward the upper edges or walls of the slits to the extent desired to create the necessary 'recesses' which must 'terminate' at the lower, sharp edges of the slits. A similar operation on a slotted sleeve would produce a similar result, but when viewed from the outer surface of the sleeve the recesses would become the projections of the quoted fourth claim. When such sleeve is teles copically [256 U.S. 668, 671] applied to-pushed or pressed into-the cap, it is obvious that with the recesses in the cap and the projections in the sleeve properly positioned and of a suitable size, when the projections shall register with the recesses, the resiliency of the metal will cause the two to engage with a snap action and the sharp lower edges of the slots in the cap will then prevent the cap and sleeve from being separated by a longitudinal movement only, until the sleeve shall be manually compressed to release the projections from the recesses.
It is to be noted, for it will be of significance in the interpretation of the claim involved, that it is required that the projections on the sleeve shall be 'beveled or inclined toward said recessed member,' and in the specification it is provided that the projections shall be 'beveled or inclined from the inner end of the sleeve toward the outer end.' With such construction it would seem plain, even without the exhibits, which show it to be true, that, friction aside, projections so beveled or inclined could readily be released by rotative movement from similarly shaped recesses without compression of the sleeve.
It was admitted at the bar that sockets made as specified in the patent in suit were never put upon the market, but the record shows that with a feature added which is covered by another patent owned by petitioner, the socket met with large commercial success. This other patent is No. 916,812 and was granted March 30, 1909, to the patentee of the patent in suit and two others, on an application filed July 18, 1904. Figure 8 from the drawings of this patent, showing the cap and sleeve of the petitioner's socket as manufactured by it, will aid in describing the additional 'improvement' which it made to the device of the patent in suit and will also be of service in comparing this device with what is claimed to be the infringing socket of the respondent. [256 U.S. 668, 672] This patent is also for improvements in incandescent electric light sockets and contains 14 claims, of which only the terms of the first need be noticed. It claims as new and useful 'a pair of tubular, sheet metal members, one adapted to telescopically receive the other, having mutually abutting cut-metal edges on the respective members (the slot 19 in the cap and projection 20 in the sleeve), to prevent relative rotative movement, and automatically interlocking means
for preventing a telescopic movement of separation of one member from the other (the recesses 17 in the cap and the projections 15 in the sleeve ), said means permitting, without manipulation thereof, the telescopic application of the members to each other.' This language, with slight additions and omissions, not significant, is repeated in claims 3, 13, and 14.
This device for preventing 'relative rotative movement' of the cap on the sleeve is the feature which the petitioner added to the socket of the patent in suit before putting it upon the market, and it consists, as the figure supra shows, in the open slot (19) cut to the edge of the [256 U.S. 668, 673] flange of the cap, into which passes the projection on the sleeve (20) when the two are 'telescopically applied' to each other. This projection ( 20) is formed by cutting two longitudinal slits in the sleeve and then pressing outward the narrow strip of metal between them. It is perfectly true, as stated in 4 of the claims and as asserted with much emphasis in the specification, that this added device effectively prevents any relative rotative movement of the cap and sleeve upon each other.
Coming now to the construction of the respondent, which it is claimed infringes the fourth claim of the patent in suit.
This (respondent's) socket has been manufactured since 1909, under patent No. 927,344, and it may most readily be described by reference to the following, Figure 3 of the drawing, which was a part of the specification of the patent.
It will be seen from this drawing that instead of outwardly extending recesses in the flange of the cap, formed [256 U.S. 668, 674] by slits and displacing of the metal, such as are in the petitioner's socket, respondent's cap has four inwardly projecting lugs (19), each of which is a separate piece of metal riveted to the flange. Each of these lugs is circular in outline, with a flat inner end or face and presents abrupt shoulders on opposite sides. In the upper edge of the sleeve are three bayonet slots of the familiar form (14), so positioned that the studs on the cap may enter them when the sleeve is slipped into the cap, but when the sleeve has been telescoped, or pushed, into the cap so that the lugs have been pressed to the end or bottom of the longitudinal part of the bayonet slots (14) rotation of the sleeve is obviously necessary to bring them into the transverse parts of the slots, so that the two members will be locked against longitudinal separation. (No such rotative movement of the two members is mentioned in the patent in suit, it is expressly provided against in petitioner's second patent, and it is not possible in its commercial socket.)
This rotative movement accomplishes a second result.
Reference to the figure, supra, shows a longitudinal slot in the sleeve, which in the key socket affords a passage for the key, but in the keyless socket is narrower, and serves only to render the sleeve compressible. To the left of this slot in the sleeve is cut a round hole ( 15) so positioned that when three of the lugs in the cap are brought by the rotative movement to the ends of the transverse parts of the bayonet slots the fourth end will snap into the hole (15). This hole (15) is placed in such a relative position that when the three lugs enter the longitudinal part of the bayonet slots, the fourth stud rides upon the metal near the edge of the slot in the sleeve, and to facilitate this movement the corner of the slot (16) is bent slightly inward. Thus the adjustment of the studs in the cap is practically a universal adjustment between the cap and shell, so that if any one lug is in [256 U.S. 668, 675] position to enter any bayonet slot two others will be in position to enter the other bayonet slots and the fourth will be in a position to enter the hole (15) when the sleeve and cap have been so pressed together that the ( three) other studs have reached the ends of the longitudinal part of the bayonet slots and the rotation described of the one upon the other has been completed. A slight compression of the sleeve suffices to release the members when desired.
This description shows that the structural features of the two sockets are strikingly different. Instead of slits and outwardly extending recesses in the cap of the one there are inwardly extended studs riveted to the cap of the other. And instead of slits and outward protections of the sleeve to lock by snap action with recesses in the cap against longitudinal movement in the one, there are the bayonet slots in the sleeve to lock with studs in the cap of the other to accomplish the same result without snap action. But these different constructions not only differ radically in structural features but they do not function in the same manner, for locking against longitudinal movement in the one is by snap action upon direct longitudinal thrust of the sleeve into the cap, while the other requires first a longitudinal movement or thrust, and then a rotative movement, without which it is entirely ineffective, whereby it locks against longitudinal movement without snap action.
But it is argued that the infringement lies especially in the locking hole (15) in respondent's socket and the associated lug which operate by snap action and afford a positive lock against rotation of the two members on each other.
The sufficient answer to this is that the patent in suit makes no suggestion of a lock against the rotative movement of the cap and sleeve on each other and contains no disclosure providing for it, but, on the contrary, because [256 U.S. 668, 676] the beveled sides which are required by claim 4 for the projections on the sleeve permitted such movement, a subsequently patented addition was added to prevent it. The petitioner cannot read into the patent in suit the additional slot in the cap and the additional projection on the sleeve of patentee's second patent, and without them there is no lock against rotative movement in the socket of the first patent to be infringed by the stud snapping into the locking hole of the sleeve of respondent's socket.
This is sufficient to dispose of the case, but we also fully agree with the Circuit Court of Appeals that the prior art, the file wrapper and the second patent to the same patentee, No. 916,812, supra, require that the expressions in the fourth claim 'telescopically received' and 'telescopically applied' must be restricted to a direct longitudinal movement or thrust of the sleeve into the cap and that for this reason the construction of respondent requiring a rotary movement to render it effective, does not infringe that of the petitioner.
In the application for the patent in suit this fourth claim, originally the seventh, read as we have quoted it omitting the words in italics. In this form it was promptly rejected by the Patent Office on reference to the Oetting patent, No. 642,825, and to the Kenney patent, No. 712,686, and it is clear that the chief concern of the applicant thereafter was to distinguish his construction from that of Kenney, which is very similar to that of respondent. The Kenney socket has lugs in the cap and bayonet slots in the sleeve, to lock against longitudinal movement, as respondent's socket has, and also slots with open ends and locking holes similarly placed in the sleeve to lock against rotary movement when the lugs in the cap engage with them. The chief difference between the Kenney and respondent's socket is that in Kenney the lug passes down the slot to a position opposite the locking hole and [256 U.S. 668, 677] then with the rotary movement mounts and rides over the narrow strip of metal-'the bridge'-into the hole, while in the respondent's socket the lugs having plain faces ride on the metal near to the edge of the slot during the longitudinal movement and from that position pass into the locking holes with the rotative movement.
When his application was rejected by the Patent Office on reference to the Kenney patent thus described, the applicant, without objection or appeal, amended his claim 7 (numbering it 4) by adding the words in italics, and in his explanatory 'remarks,' when submitting these amendments, it was said:
Thus the patentee, in order to avoid infringing Kenney's construction, voluntarily restricted himself to a 'specific structure' operative when the sleeve was 'simply' inserted in the cap, without suggesting any rotary movement whatever, but, on the contrary, by his reference to Kenney as locking and unlocking by 'simply rotating one member upon the other,' clearly implying that no such rotary movement was necessary in the adjustment of his socket. Having thus narrowed his claim against rotary movement in order to obtain a patent the patentee may not by construction, or by resort to the doctrine of equivalents, give to the claim the larger scope which it might have had without the amendments, which amount to a [256 U.S. 668, 678] disclaimer of rotation as an operative feature of his devise. Shepard v. Carrigan, 116 U.S. 593, 598 , 6 S. Sup. Ct. 493; Hubbell v. United States, 179 U.S. 77, 80 , 21 S. Sup. Ct. 24.
But that no rotary movement was implied in the use of 'telescopically received' and 'telescopically applied' in the patent in suit is further unmistakably shown by the frequent use of the same or equivalent expressions in the specification and claims of patentee's later patent, No. 916,812, supra, in which claims 1, 3, 13, and 14 relate almost wholly to an improvement which renders rotation of the members one upon the other impossible and which petitioner added to the construction of the patent in suit before putting the sockets upon the market. That the inventor did not intend to claim, and that he certainly did not disclose, that any such rotary movement was necessary in his socket as was required to render the socket of the respondent effective, is, we think, for these reasons, so clear that it is not necessary to consider the dictionary definitions of the words used, upon which the Circuit Court of Appeals, with sound reason, relied in reaching this same conclusion.
For the reasons thus elaborated, we conclude that the socket of the respondent does not infringe the fourth claim of the patent in suit because there is no claim made therein of a lock against rotative movement of the cap and sleeve upon each other, and also because the file-wrapper and the subsequent patent show that the disclosure is, and was intended to be, limited to a construction operative by direct longitudinal movement or thrust without such rotative movement.
The contention, pressed strenuously upon our attention, that the added open slot in the cap and projection on the sleeve provided for in the second patent were devised simply to secure the alignment of the projections and recesses of the first patent in making telescopic application of the two members to each other, is not convincing [256 U.S. 668, 679] in the presence of the fact that claims 1, 3, 13, and 14 of the second patent are devoted almost wholly to claiming that this improved construction is intended 'to prevent relative rotative movement' of one member upon the other, while only claims 11 and 12 refer, quite incidentally, to what is now claimed to be the chief function of the added parts, that of a guide to the positioning of the recesses and projections of the first patent with reference to each other. Without a guide, the difficulty should not be very great of aligning projections and recesses distant from each other but a fraction of an inch when the edges of the cap and sleeve are in contact. The constructions held by the Circuit Court of Appeals of the Second Circuit to infringe the patent in suit, were so essentially different from that of the respondent that we regard discussion of them as quite unnecessary.
It results that the decree of the Circuit Court of Appeals as to claims 1 and 4 must be