No. 248.Circuit Court of Appeals, Second Circuit.
March 13, 1933.

Appeal from the District Court of the United States for the Southern District of New York.

Suit by the General Electric Company against the United States Electric Manufacturing Company for the infringement of United States patents No. 1,567,863, to Sargent and De Reamer, and No. 1,567,864, to Sargent. From a decree [56 F.2d 630] dismissing the bill of complaint, plaintiff, the assignee of these patents, appeals.

Decree reversed as to the first patent, and cause remanded, with directions, and decree affirmed as to the second patent.

Charles Neave, of New York City (Harrison F. Lyman, of Boston, Mass., Alfred E. Bobst, of Schenectady, N.Y., and Fish, Richardson Neave, of New York City, of counsel), for complainant-appellant.

Gifford, Scull Burgess, of New York City (George F. Scull and Charles W. Mortimer, both of New York City, of counsel), for defendant-appellee.

Before MANTON, SWAN, and AUGUSTUS N. HAND, Circuit Judges.

AUGUSTUS N. HAND, Circuit Judge.

This suit was brought for infringement of United States patents No. 1,567,863 and No. 1,567,864. Each relates to house wiring, and particularly to the mounting in outlet boxes of electric wiring devices such as switches and plug receptacles. The first patent, No. 1,567,863, has been called the “plaster ear patent,” because the invention consists of an arrangement of plaster-engaging ears on the crossbar of the switch or plug receptacle. The second patent has been called the “weakened ear patent,” because the invention provides weakened lines of joinder between the ears shown in the first patent and the body of the crossbar, in order to render the ears readily detachable therefrom. Infringement of each patent is admitted, so that the only question for determination is that of validity. The District Judge held the “plaster ear patent” void for lack of invention, in view of the British patent to Mitchell, No. 102,167. He also held the “weakened ear patent” void for lack of invention in view of the general prior art.

The specification in the first patent, No. 1,567,863, says: “In wiring houses the outlet boxes are usually fastened to the studding before the building is lathed and plastered, and it often happens that boxes are set crooked so that the plane of the box opening is not parallel to the plane of the plaster surface. As a result, when the switch or other device is mounted on the box it will be crooked with respect to the plaster surface, which means of course that the switch is not straight.”

The patentees, instead of attempting to adjust the crossbar of the switch to any inequalities in level of the wall box, simply extended the ears of the crossbar at each end so that they would rest upon the plaster. The main criticism of the claim of invention is that the step was obvious. But we think its simplicity, in view of the failure of prior inventors to attain the very desirable result achieved by the patentees, is the very basis on which invention may be found.

The difficulty referred to in the specification had long been recognized, and various expedients had been employed in an endeavor to correct it. The traditional way of aligning switches with the surface of the wall was to raise or lower the ends by the use of metal washers. After the lathing and plastering were done, the wiremen proceeded to install the switches and plug receptacles and finishing plates. In setting the outlet boxes, it is impossible to determine just where the line of plaster may come, because with rough lumber the walls are unequal, the laths are a little thicker or thinner in one place than in

Page 765

another, and the boxes are often below the surface of the finished plaster. It has been customary for manufacturers to furnish metal washers with switches so that they could be built up on the edge of the outlet box to the proper height. These washers have been inserted between the ends of the crossbar and the box lugs so as to hold the switch surface in alignment with the wall. The determination of the number of washers necessary to effect alignment and the adjustment of the switch by means of them have been troublesome matters. If in any case the adjustment did not prove satisfactory, the screw would have to be withdrawn, so as to add or take away enough washers to raise or lower the switch to the proper height. At times careless workmen would omit the washers altogether and count on the plate to hold it in position.

Sargent recognized the difficulties inherent in the conventional practice and attempted to better it by the device shown in United States patent No. 1,173,040, granted to him on February 22, 1916. In the specification of that patent, when referring to the practice of using washers, he said: “Such an expedient not only incurs the expense of a number of washers but frequently requires taking the switch out and putting it in several times before a proper adjustment is secured.”

Under the last-mentioned patent No. 1,173,040 Sargent obtained alignment of the switch by means of a “lock nut” provided with a “tang” adapted to engage in a hole in the crossbar. The device was adopted commercially by the General Electric Company and met with temporary success. But it involved furnishing a separate loose part, i.e., the “lock nut” with each switch, which was a somewhat complicated mechanism, and not long after its introduction was displaced by the more simple and convenient devices of the patents in suit.

Not only was Sargent striving to avoid the inconveniences incident to the use of washers and to devise satisfactory ways of aligning the switch, but numerous other inventors were dealing with the problems and seeking to solve them. United States patent No. 754,414 to Bossert shows an extension of the wall box with sides adapted to engage with inwardly extending flanges of the box whereby the extension on which the crossbar rests may be aligned with the wall surface. United States patent No. 835,039 to Slocum provides brackets attached to the box which may be raised or lowered and held in position by screws extending through vertical slots in the ends of the box. United States patent No. 989,854 to Kruse shows an extension which may be raised or lowered telescopically within the box so that its ears on which the switch rests are flush with the wall. United States patent No. 1,038,963 to Roe provides an adjustable extension of the box somewhat resembling the extension in the Bossert patent. United States patent No. 1,069,777 to Fowler shows a box with switch attaching ears which may be raised or lowered by a screw so as to become flush with the wall face. United States patent No. 1,113,762 to Eckman provides racklike members which may be engaged with flanges of the wall box so as to support the crossbar of the switch and align it with the wall surface. In United States patent No. 1,275,725 to Newton horizontal arms are placed flush with the plaster line and screws are tightened up to lock them in place so as to adjust the switch to the height of the wall box at each end.

All of these patents, though attempting to solve the difficulties to which we have adverted, involved the use of additional parts and required considerable additional expense. None of them, so far as the record shows, went into commercial use, and, in spite of their teachings, the trade continued to employ the old method of aligning the switch by the use of washers. All of them provided adjustments of one kind or another between the box and its support on the wall or between the box and the switch.

The great feature of the “plaster ear” patent is that it requires no adjustment whatever. The solution which the patentees obtained consisted merely in extending the crossbar of the switch at both ends far enough and wide enough so that it would rest upon the plaster and automatically support the switch in alignment with the wall. This rendered unnecessary washers, extensions, brackets, and all other adjustment devices. The long struggle to devise something which would conveniently and inexpensively secure alignment of the switch with the wall face shows the need recognized by those familiar with the defects of current methods of installation, and the failure of the mechanisms illustrated in the prior patents to secure commercial acceptance, demonstrates the inadequacy of the earlier methods to satisfy that need. Sargent’s “lock nut” device disappeared in the face of the new invention. The General Electric Company abandoned it, and up to the date of the trial had manufactured 15,500,000 devices under

Page 766

the patent in suit, and had granted licenses to large manufacturers of electrical equipment who had sold 17,800,000 of the plaster ear crossbars. This is certainly an impressive tribute to the novelty and utility of the invention.

The principal attack on the “plaster ear” patent, is based on the prior British patent No. 102,167 to Mitchell, granted in 1916. While the application was pending in the Patent Office, it was at first held that there was “no novelty in the switch support disclosed except in the extension * * * and that to supply such old form of switch support with an extension at each end to rest on the plaster around an outlet box to position the switch would not be invention in view of the British patent.” But the novel advantages of the Sargent and De Reamer invention were finally recognized by the Examiner, and their application for a patent was granted. In view of the great reliance of the defendant on the Mitchell patent to defeat the Sargent and De Reamer claims, it is surprising that the patent to Mitchell was not pleaded as a defense and was only put in evidence at the trial to show the state of the art. Mitchell’s specification and drawings show a square boxlike element for insertion in the wall box, termed, in the patent, a “cradle.” It has four flanges resting quadrilaterally on the surface of the plaster. This cradle is provided at its bottom with a “saddle” or “crossbar” to which the switch may be attached by screws and on which it rests. Thus in Mitchell’s device the switch is supported by an underlying crossbar, while in the “plaster ear” patent it is supported from a “crossbar” above. In the form illustrated in figures I, II, and III of the Mitchell patent, the “cradle” is attached to the wall box by screws at its four corners, and in the alternative form illustrated in figures IV to VII by a central screw which passes through a hole in the “saddle” and engages a central hole in the wall box. Thus the Mitchell device is a two part affair consisting of the “cradle” with the attached “saddle” and the switch. Defendant’s expert, Ray, said that, if the two side members of the cradle with their flanges were cut off, leaving just the end members with their flanges, there would remain the same device as that of the Sargent and De Reamer patent. But that is not so, for, if the switch in the Mitchell device were thus reconstructed, it would still be supported from below and not carried by the crossbar above as in the “plaster ear” patent. Moreover such an adaptation of the Mitchell device to the patent in suit involves a reconstruction of the former, and was not sufficiently obvious or important to appeal to Newton. The latter, in United States patent No. 1,275,725, was dealing with the same old problem, but, like the earlier patentees, made his switch flush with the wall surface by adjusting it to the height of the outlet box. In spite of the teachings of Mitchell, the old method of building up the crossbar with washers, continued for years. That method at least involved the use of only a few small washers, and did not require the cradle and saddle of Mitchell with their cumbersome method of installation and considerable additional expense.

It is true, as was stated in the file wrapper, that “the difficulty the British patentee, Mitchell, seeks to overcome is that with which appellant is dealing,” but the means adopted by Sargent and De Reamer differed from that disclosed in the Mitchell patent. Sargent and De Reamer extended the ears provided for fastening the crossbar to the outlet box so that their ends would rest on the plaster. Thus they were employed not merely for fastening but also for leveling. So far as the extensions served as levelers they were substitutes for the quadrilateral flanges of Mitchell; but they involved a simpler and less expensive device than Mitchell’s, and acted at once both as levelers and as parts of an upper support for the switch. It cannot be said that the Sargent and De Reamer device was an obvious advance over Mitchell. Years passed before its adoption, and in the meantime Newton, in United States patent No. 1,275,725, had continued to grapple with the recognized difficulties by employing a new form of adjustment, and electricians went on using the old washers. The great commercial success of the “plaster ear” device and the tribute which defendant has paid to its worth by infringing the claims of Sargent and De Reamer leave their patent in a far stronger position than when it was before the Patent Office.

It is further argued that United States patent No. 1,161,361 to Tirrill anticipates. The application for the Tirrill patent was filed in 1909, and the patent issued in 1915. If this early patent had really disclosed an invention which has since proved so successful, it is unlikely that it would have lain practically dormant for seven years. Like the patent to Eckman (No. 1,113,762), it shows

Page 767

a crossbar with two screw holes adaptable to different sizes of outlet boxes. This was the ordinary form of crossbar for electric switches. If in rare instances it was long enough to rest on the plaster, it was not properly adapted to any such use. It could only rest on the wall in cases where the electrician had chipped away no more plaster than was necessary just to uncover the screw holes in the flanges of the box. Its use as a leveling member is not disclosed in the Tirrill patent and was undesigned. The instances where such a use is shown to have occurred have been most trifling and were accidental. The ears of the Tirrill patent were in no sense “adapted to engage the surface of the wall,” as required by claim 6 of the patent in suit, nor did they extend beyond the openings that received the fastening screws as required by claim 8. Indeed Tirrill’s crossbar terminates as near the fastening opening as it possibly can. The omission of Tirrill to disclose the invention of Sargent and De Reamer cannot be supplied by oral testimony, especially by testimony about sporadic and accidental modes of installation. Canadian General Electric Co. v. Fada Radio, 1930 A.C. at page 104.

In the “plaster ear” patent we find a practical and commercially successful, solution of a long-felt difficulty which many others had sought to obviate. Such striving and failure followed by ultimate success on the part of the patentee is the strongest proof of invention.

The second or “weakened ear” patent required no inventive faculty. Sargent is supposed to have discovered that in some cases the “plaster ears” were not desired by workmen who preferred to do their leveling in the old ways. He accordingly scored the switch supporting bar so that the ears could be broken off at the place where they were scored. Various patents, among others, United States patent No. 1,225,525, to Sweet, show that this method of scoring metal in order to form points of weakness where an undesired part may be broken off was old. As well might the indentations on billheads or sheets of postage stamps be treated as inventions. This patent is altogether too trifling to endure.

The decree is reversed as to the first patent, and the cause is remanded, with directions to adjudge that patent infringed and to enter the usual decree applicable to such a case. The decree is affirmed as to the second patent. No costs are allowed.