Full opinion text
WEICK, District Judge. The Consolidated Cases. These four actions were consolidated for trial. They involve the validity, scope and infringement of three patents for improved mills and methods for the rolling of hot and cold metal in strip or sheet form. The patents in suit are: No. 1,744,-016 (’016) issued January 14, 1930 on an application filed June 30, 1923 by Abram P. Steckel, hereinafter referred to as Steckel; No. 1,779,195 (’195) issued October 21, 1930 on an application filed December 9, 1929, as a division of the original Steckel application; No. 2,706-original Steckel application; No. 2,706,-422 (’422) issued April 19,1955 on an application filed May 2, 1947 by William B. Lockwood, hereinafter referred to as Lockwood. Steckel assigned his applications for patent to the Cold Metal Process Company, an Ohio corporation, hereinafter referred to as Cold Metal, and the patents were issued to that company. Thereafter, pending this litigation, the Steckel patents were assigned to the Union National Bank of Youngstown, Trustee of the Leon A. Beeghly Fund. The Lockwood patent was first assigned to the Cold Metal Products Company, an Ohio corporation and thereafter through mesne assignments to the Youngstown Research & Development Company, an Ohio corporation. E. W. Bliss Company, hereinafter referred to as Bliss, is a Delaware corporation having a place of business in Salem, Ohio and is engaged in the manufacture and sale of rolling mills. Greer Steel Company, hereinafter referred to as Greer, is a West Virginia corporation having a place of business at Dover, Ohio and is a user of rolling mills. It purchased several mills from Bliss and others from United Engineering & Foundry Company, hereinafter referred to as United. 1. Equity No. 5402. This action was filed by Bliss against Cold Metal on January 20, 1936 seeking a declaratory judgment with respect to the validity and infringement of Steckel Patent Nos. ’016, ’195 and two other patents which were eliminated from the case at a pre-trial conference. On motion of Cold Metal, the complaint was dismissed on April 12, 1937 for lack of a justiciable controversy. On appeal, the order of dismissal was reversed. 6 Cir., 1939, 102 F.2d 105. Cold Metal filed its answer on May 20, 1939, but asserted no counterclaim at that time. A motion for summary judgment by Bliss as to the validity of Patent No. ’195 was denied. D.C.1942, 47 F.Supp. 897. In 1947, after the patents had expired, and the issues had become moot, Cold Metal and the trustee moved for leave to file a counterclaim charging Bliss with infringement of Patent Nos. '016 and ’195. Bliss then moved to dismiss the complaint. Leave to file the counterclaim was granted and it was filed on May 26, 1950. The motion to dismiss was denied. Bliss thereafter moved to dismiss the counterclaim with respect to all mills sold by it more than six years prior to the date of the filing of said counterclaim. It asserted that such claims were barred by the Statute of Limitations. The motion was denied by the Court for the reason that the filing of the declaratory judgment action had tolled the Statute of Limitations. D.C.1957, 156 F.Supp. 63. This case is the oldest one on the docket of the Court. It was filed at a time when the Court had separate dockets for law and equity cases which long since have been abolished. 2. Civil Action No. 26,542. This action charges Bliss with infringement of Patent Nos. ’016 and ’195. The issues are the same as Equity No. 5402. It is claimed by Cold Metal that this action was a precautionary suit to stop the running of the Statute of Limitations in the event it was held that the Statute was not tolled by the filing of Equity No. 5402. 3. Civil Action No. 24,189. This action charges Greer with infringement of Patent Nos. ’016 and ’195 by reason of its use of mills which it purchased from Bliss and United. 4. Civil Action No. 31,849. This action was instituted by Bliss for a declaratory judgment with respect to the validity and infringement of Lockwood Patent No. ’422. A motion to dismiss for lack of a justiciable controversy was denied by the Court. The trial of the consolidated cases lasted for more than two months. One week was spent by the Court with counsel in viewing rolling mills in six different states. A total of 768 exhibits were offered in evidence, some of which were depositions and testimony of witnesses stipulated from the records of other cases. Mill Types In a 2-high mill there are only two rolls which are at different elevations. A sketch of a model 2-high mill with all its essential parts identified is reproduced below. A similar sketch of a model 4-high mill is also reproduced. It has rolls at four different elevations. The two rolls in the center are called working rolls each of which is supported by a backing roll, A 4-high mill permits the use of smaller working rolls in a wide mill than does a 2-high mill because of the support provided by the backing rolls. If small working rolls were used on a 2-high mill, they would not be able to withstand the heavy pressures required for substantial reductions of metal and the rolls would tend to bow apart. The 4-high mill with large backing rolls provides the necessary support for small working rolls, prevents bowing and permits their use for heavy reductions. The sketch below illustrates the roll arrangement of 2-high, 3-high, 4-high and cluster mills with the strip between the rolls. It will be noted that in the cluster mill the working rolls are nested in the crotch between the backing rolls thus obtaining horizontal as well as vertical support. A mill other than the 2-high mill is usually referred to as a backed up mill, because the work roll or rolls are supported or backed up by backing rolls. Patent No. ’195. In Steckel’s original application for patent filed June 30, 1923 he presented claims for a mill with small working rolls and relatively large backing rolls. The backing rolls were to be mounted on anti-friction bearings (roller bearings). In addition, he presented claims for the application of tension on the strip as a driving force of the mill. In 1929 the Patent Office required division of the claims involving tension as a means for driving the mill and those for the mill stand per se. This resulted in the application, filed on December 29, 1929, upon which Patent No. ’195 issued. Claims 3, 4, 6, 7, 8 and 11 to 17 of that patent are here in suit. It is unnecessary to pass on the remainder. No infringement is charged as to them and, the patent having expired, their validity is moot. The claims in question are all mill claims. Claims 8, 11, 12 and 16 specifically relate to continuous mills, while the remainder describe single mill stands. Basically, they define a 4-high mill with antifriction bearings on the backing rolls. At the time of Steckel, and long prior thereto, antifriction bearings were understood in the art as meaning ball or roller bearings as distinguished from friction or sliding bearings on which one metal surface would slide over another without the interposition of balls or rollers. The claims vary from one another in that some refer to the control of roll neck heating by the application of roller bearings while others refer to the possibility of obtaining greater speeds through the use of the roller bearings. In the claims appears language which, to my point of view, is simply excess verbiage describing a properly designed backing roll or roller bearing. Examples of this are found in Claims 3, 4, 8, 13 and 14 wherein it is stated: * * * the backing rolls having necks of sufficient size to withstand the rolling pressure * * * Any competent mill builder would design his mill so that the roll necks would withstand the rolling pressures. To do otherwise would be sheer folly. By the same token, several of the claims describe the bearings to be used as “of sufficient size to withstand the rolling pressure” (Claims 8, 14 and 17) or “of such character as to withstand speeds which are high relative to those ordinarily employed” (Claims 4, 7, 8 and 13). To say this in a patent relating to a mill intended to be run at a high speed and to make heavy reductions is only to say use an adequate roller bearing. Again, an engineering fundamental. Stripping the claims of the verbiage and looking to their essentials they disclose a 4-high mill with roller bearings on the necks of the backing rolls. The advantages claimed for this mill are expounded in the specifications. “By my invention the limitations Of speed on both hot and cold mills are removed. * * * •» * * * “Where the roller bearings are applied to the backing roll necks they eliminate the heating of such necks, which heat adversely affects the shape of the working roll body. -X * * * * * * “My invention is also highly advantageous in that the amount of power required is markedly reduced * * * •X * •» * * -X * “By the use of my improved mill these high speeds may be readily attained and this with relatively low power requirements.” It stands admitted that the proportions of the 4-high mill stand per se shown in Patent No. ’195 are to be found in the prior art. They are closely akin to those of the Newton Patent (Br. No. 1234/1863). Thus, the question at hand resolves itself to whether invention resides in the application of roller bearings to the necks of the backing rolls of a conventional 4-high mill. Concurrent with this is the issue of whether the advantages claimed in ’195 are a product of a new environment. The use of the roller bearing itself was old in the rolling mill art. In an 1895 article the rolling mills of the Moss-berg Mfg. Co. are described. “These apparatus [rolling mills] are suitable for the rolling of * * even sheet steel; the fact that their trunnions are mounted in roller bearings imparts to them an exceptional ease of operation which manifests itself in a considerable saving of the power consumed as compared with the power usually required. “The substitution of a roller bearing for friction bearings under considerable load is not a recent change; it attracted the attention of research workers for many years. Revue In-dustrielle, No. 44, Nov. 2, 1895.” The Mossberg mills referred to are illustrated and described in that firm’s catalogue of March 1896. Therein it is claimed that: “The Mossberg Roller Bearing is fitted to the journals of the rolls. This latter feature enables us to build a mill which will do its work double as fast as the ordinary mill with less than one-half of the driving power, for with Roller Bearings there is scarcely any friction on the roll journals and consequently no heating of the rolls. ******* “After placing a set of these bearings on any machine we can safely guarantee the total absence of heating from friction and a savings of forty to eighty per cent of the power previously used * * * ” That these claims were not merely sales puffing is reflected in a number of testimonials in the catalogue. Among them is a letter from Browne & Sharp Mfg. Co. That concern stated they had put Mossberg bearings on a mill used for rolling sheet steel, and as a result were able to reduce their mill powering and run the mill ten hours a day with no roll heating. Previously their mills had to lie idle two to three hours a day because of excessive roll heating. This notice to the engineering world of the advantages of roller bearings was repeated in the firm’s 1902 catalogue (p, 19). However, the company recognized that the roller bearings of that day were not a universal panacea for the rolling mill industry’s troubles. The admonition is given: “As yet, Roller Bearings are not universally perfect, nor is their efficiency practically known, except under certain conditions. “A time may come when they can be applied indiscriminately under every condition, but until that time comes, it is necessary to make a special study of each application of Roller Bearings so as to avoid mistakes.” : » Another mill builder, the Standard Machinery Co., also extolled the virtues and advantages of roller bearings, specifically mentioning the increased speeds, decreased power requirements and absence of roll heating. In their 1920 cata-logue several 2-high mills equipped with roller bearings are offered to. the public. The publication “Iron Age” reporting on the use of roller bearings on a rod mill stated: “The introduction of these roller bearings constitutes a distinct and important advance in rolling mill machinery of this type, and it is highly probable that before long it will find application with heavier work. ******* “By the use of roller bearings the product of the mill can be increased in two ways. First, the rolls can be run at a higher rate of speed. Second, much heavier strain can be taken than with plain bearings. “Numerous tests on this class of work, especially cold rolling mills, show a saving in power of from 25 per cent to as high as 75 per cent. Iron Age, April 7, 1898.” Here then in the pre-Steckel publications we see full disclosure of the advantages to be derived from the use of roller bearings on a cold rolling mill. Admittedly, the uses therein described were on 2-high cold mills of a relatively small size which took light reductions. However, each of the advantages claimed to be secured by the Steckel invention are therein described as being a normal attribute of the use of roller bearings on a rolling mill — higher speeds, reduced power requirements and elimination of roll heating. The obvious question then is if this principle of the utility of roller bearings was so well known why wasn’t it applied to a 4-high cold reducing mill capable of taking substantial reductions prior to the 1920’s. Bliss and Greer contend that it was so used. Assuming for the moment that it was not, a partial explanation why not is also to be found in the pre-Steckel publications. As late as 1918 it was written that probably no other feature of steel mill design was more neglected than that of bearings. The Iron Trade Review, Vol. 63 p. 94. Four years earlier, an article appeared in the English publication Engineering entitled “The Design of Rolling-Mills for Cold Metal.” Therein it was observed: “The subject of roll bearings cannot be concluded without a word respecting the use of roller bearings. Bearings of this class are usually considered to be impracticable for mills; nevertheless, the writer is inclined to think that a roller bearing could be designed to take the heaviest loads, and if such is the case, there is no doubt a great economy of power would result from their use. •X- * -X -X- -X- -X- * “The obstacle in the way of experimenting with roller-bearings on a larger scale is that none of the manufacturer’s specializing in roller bearings can, so far as this writer’s experience goes, put forward a bearing of suitable dimensions.” The roadblock in the path of the development of the roller bearings was a mechanical one. “Like most other improvements in mechanical lines, the ball or roller bearing is coming because the time is ripe for it. The principle is a very old one, but years ago, when such bearings were tried the desideratum was accurate and hard balls and bearing surfaces. This desideratum could not be met with the tools and implements then at command. Tin and Terne and the Metal World, Vol. VI, No. 16 (Aug., 1897).” The importance of proper methods and materials in relation to the construction of an adequate bearing is emphasized in a discussion of the Mossberg Bearing. “The efficiency of the bearing depends perhaps as much upon the skill with which it is manufactured as upon the design. The shaft and bearing, as well as the rollers and their bearings in the cage, must be absolutely cylindrical. The slightest variation would impair the bearing and practically eliminate its frictionless quality. The Iron Age, April 7, 1898.” These difficulties in minute accuracy required would necessarily increase as the size of the bearing which one was attempting to build increased, and a large bearing is necessary to withstand the heavier rolling loads. Besides these technological drawbacks a further impediment to the adoption of roller bearings is spoken of — the human factor. “Mechanical men, however, are very skeptical on the life of Roller Bearings as their experience with a great variety of cheap and wrongly applied Bearings has led them to condemn all Roller Bearings. Cata-logue of Mossberg & Granville Mfg. Co., p. 17 (4th Ed., 1902). “Above all, however, the roller bearing has to overcome the strong prejudice which a long, and hitherto uninterrupted, record of failures has naturally raised against it. Old mechanics say:— “ ‘Yes, we tried roller bearings twenty-five years ago. They didn’t work.’ Cassier’s Magazine, Yol. XII, p. 68 (1897)” The most searching statement found relating to the development of the use of roller bearings for industrial machinery is in a 1900 publication. “Nevertheless, the theoretical advantages of the roller bearing have always been too obvious to allow it to be abandoned as impossible, and inventors have constantly been attracted to it. It was not, however, originality that was required to bring it to perfection, but suitable materials and sound mechanical knowledge. “The subject is full of interest for mechanics for it shows that many of the insoluble problems of twenty years ago become possible with modern materials and processes. When shown a device of this kind, it is very natural to say that it has been tried and failed and it is not worth considering. But it has always been tried under more adverse circumstances than the present, and it is possible that the cause of failure lay in some small difficulty which no longer exists. Everyone who has tried, knows that it is the details of progress which give the trouble; the main features are generally fairly easy; it is the little matters which trip us up. This has certainly been the case with the roller bearing; the general design was complete 50 years ago, but the mechanical execution has only recently been perfected. Engineering, Vol. LXIX, p. 133 (1900)” It thus appears that prior to Steckel the principle of the roller bearing was well known, the advantages to be derived from its use widely proclaimed and enjoyed by the users of 2-high mills, but that technological obstacles and human stumbling blocks existed to the development and adoption of a roller bearing for devices which exerted extremely high pressures, such as a large cold reducing mill. With this background of the widespread notice of the utility of roller bearings in mind the prior art patents cam now be considered. Cold Metal concedes, as previously mentioned, that the basic mill structure of ’195 is to be found in the prior art. It is likewise uncontroverted that none of the prior art patents relied upon by Bliss and Greer disclose the identical structure of the ?195 patent — a 4-high mill with roller bearings on the backing rolls. A 4-high cluster mill was the closest approach to it. In Exhibits OX-9, OX-11 and OX-12 nine prior art references are cited as disclosing the principle of the roller bearing as an expedient on a rolling mill. The Dostal patents (Austrian Nos. 77,898 and 81,024 and American No. 1,-453,239) and the Roberts patent No. 1,378,559 relate to an antifriction bearing per se, and are not mill patents. However, Dostal 77,898 is directed to a bearing to be used on the neck of a mill roll, with no qualification as to whether it is to be a working roll or a backing roll. Brightman Patent No. 564,302 is for a wire drawing mill. Although the general term antifriction bearing is used therein, the bearing shown is a ball bearing. The mill is a 2-high type for a special purpose. It cannot be said that this patent discloses the utility of roller bearings as does ’195. Kynoch (Br. No. 12,006/1898) is a 2-high mill for cold rolling metals, particularly brass. The drawings unquestionably show roller bearings on the roll necks and the specifications refer to them as such. The three most important patents cited are Somers (800,950), Kriwan (1,360,-959) and Wilmot (1,071,846). These patents were before the Patent Office and the grounds upon which they were therein distinguished will be considered. ' '* Somers is a patent on a 4-high rolling mill with roller bearings on the work rolls. The Patent Office repeatedly took the position that there was no invention in providing all the rolls (or the backing rolls) of a 4-high mill with roller bearings as shown at 6 of Fig. 2 of Somers. This patent was cited three times in rejection of claims on a 4-high mill with antifriction bearings — once in the prosecution of abandoned Biggert and Johnson Application, Serial No. 103,264 and in the prosecution of Steckel Patent Nos. 1,744,016 and 1,744,018 When Somers was cited as a reference in the prosecution of Steckel ’016 all the claims against which it was asserted were cancelled. In the prosecution of Steckel ’018 the Examiner rejected original Claims 1, 2 and 3 on the ground that there would be no invention in providing the backing rolls with such roller bearings as are provided for the [Somers] working rolls. In answer to this Steckel’s counsel stated: “We recognize the force of the Examiner’s position as to Claims 1, 2 and 3 * * * We take issue with the Examiner’s position that there would be no invention in providing the backing rolls of Somers with anti-friction bearings, but have rewritten the claims so as to recite in terms of structure the driving of the mill by tension on the strip.” In answer to the Examiner’s citation of Somers in Biggert and Johnson, counsel stated that the reference was inapplicable in that the work rolls did not take the rolling pressures in a 4-high mill, and requested reconsideration of the rejection. Before any action was taken on that request the application was abandoned in favor of Biggert and Johnson Application No. 228,176, which was stated to be a continuation in part of Application 103,264. In this latter application the official position of the Examiner on Somers does not appear, as that patent was discussed in an oral interview. The applicant’s contention remained the same, as shown in an amendment of November 22, 1927. Biggert and Johnson Patent No. 1,654,-235 issued on this application, without any prior art references having been cited by the Examiner. It is argued that the ’235 patent was obtained by virtue of the commercial success of the mills installed by United at Rome, New York and Butler, Pennsylvania which embodied the Biggert and Johnson conception. The claim is made that the Patent Office was overwhelmed by the commercial success of these mills and issued the patent notwithstanding the borderline degree of invention shown by the prior art. Consideration of the Biggert and Johnson file wrappers convinces one that this is not altogether without foundation. I feel, therefore, that the granting of ’235 over Somers does not bear the same weight as if the Patent Office had granted it on the basis of a clear cut distinction between the two structures. Thus, the situation with reference to Somers is that of the three times it was cited by the Patent Office as invalidating claims for a 4-high mill with roller bearings on the backing rolls, once the claims Were cancelled, once the claims were amended to include tension drive as a feature of the mill, and the third time the application was abandoned in favor of a continuation application urging commercial success as a basis for patentability. Looking at Somers from the point of •distinction raised by applicants in the Patent Office, their position is only partially sound. It is true, as they contended, that the rolls with roller bearings in Somers did not take the brunt of the rolling pressures and thus had no roll neck heating to overcome. On the other hand, the Somers mill was a “hot roll” mill, and, therefore, roll neck heating was not a detriment. The application of roller bearings to the work rolls would necessarily allow rolling at higher speeds with less power, as the prior art publications teach. Thus, Somers could obtain whatever advantages would accrue to his particular type of mill by the use of roller bearings without putting them on the backing rolls. Kriwan (1,360,959) shows a 4-high mill with roller bearings on the backing rolls. The mill differs from Steekel in that it is a bar, or billet, five-pass mill and the rolls are divided into five sectors, each of which is of a progressively larger diameter. The roller bearings are not on the roll necks (the mill had no roll necks), but are between each backing roll and its arbor. Thus, a series of five backing rings for the five steps of the work rolls results. When Kriwan was cited against claims for a simple 4-high mill with roller bearings in the prosecution of Steekel ’016 the claims were thereupon cancelled. By the same token, in the prosecution of Steekel 1,744,017 the claims describing a 4-high mill with roller bearings on the backing rolls were amended to include the element of tension when Kriwan was cited against them. Kriwan was also relied upon by the Examiner in the Biggert and Johnson prosecution: “Claims 18 and 19 are rejected as not patentable over Somers or Kri-wan. There would be no invention in providing bearing for the backing rolls of Somers of the same type as are provided for the working rolls and in Kriwan each collar b is the equivalent of a backing roller and has anti-friction bearings.” Counsel answered: “The Kriwan patent shows a pair of stepped rolls,'each step being provided with an anti-friction ring. It is not a mill of the type defined in the references and obviously could not be modified to meet applicant’s structure, because it would be impossible to drive the rings shown in Kriwan.” This point of distinction, the inability to drive the backing rings, is valid for Biggert and Johnson, but not Steckel. Biggert and Johnson showed a drive to the lower backing roll, and it is this they claim Kriwan couldn’t do. Steckel had no such backing roll drive. The Patent Office was obviously unimpressed with this argument, for it came back with another rejection, stating: “All that applicants have done is to provide anti-frictional bearings for the brace or backing rolls. Ball and roller bearings are now so common that to supply them in any bearing is not invention as they may be said to be universal to the mechanical arts. The applicants have therefore made no patentable invention.” Faced with this flat rejection applicants changed their tactics. All prior claims were cancelled, new ones filed, and a lengthy argument based on the commercial success of the mill was presented. Therein Kriwan • was distinguished as not having roll necks and consequently having no roll neck heating problem to overcome. It was at this point that the application was abandoned, and as previously stated the application on which the patent finally issued contained no prior art references at all. Thus, the history of Kriwan before the Patent Office shows that when asserted against claims of a 4-high mill with roller bearings on the backing rolls the claims were cancelled in ’016 and amended in '017. As far as Biggert and Johnson is concerned, the distinctions attempted to be made were rejected, although the patent finally issued on the basis of the mill’s commercial success. Further distinctions between Kriwan and Steckel were made during the trial. It was pointed out that the mill could not roll to the gauges and widths of the 4-high sheet mills. The difference in roll construction was pointed to with extreme emphasis. While all the variances would operate to bar Kriwan as an anticipation, they do not have the same effect on the question of invention. Is each person who puts a roller bearing on a different type of rolling mill and utilizes the benefits to be derived therefrom entitled to a patent thereon? Further, viewing only one section of the Kriwan five steps alone, virtually the exact form or replica of the first mill Steckel built, which is described and claimed in ’016, is seen. Steckel’s original mill was of the backing ring variety and had no roll necks. (He did not build a mill with roll necks until 1929). The bearing arrangement on it was identical with that shown for each of the individual sections of Kriwan. Finally, the Wilmot Patent No. 1,071,-846 is relied upon to invalidate ’195. Bliss and Greer contend that Wilmot represents a complete anticipation of Steckel. Cold Metal has taken the position that Wilmot cannot be considered as an anticipation in that it is a cluster mill and, therefore, distinct from the 4-high of the Steckel patent. As before stated, in a cluster mill, each working roll rests in the crotch between two backing rolls thus providing horizontal as well as vertical support for the working rolls. In a 4-high mill only vertical support is provided for the working rolls by a single backing roll. From the testimony given at the trial it appears that the principles of the 4-high and cluster mills are substantially the same. Dispositive of this contention is the fact that the Alcoa Foil Mill (OX-2, pp. 2-3, item 14) and the Stamford Rolling Mills (OX-2, pp. 2-3, item 15), both cluster mills, are charged to infringe ’195. If a cluster mill is capable of infringing ’195 it is also capable of anticipating it. Commercial Mfg. Co. v. Fairbank Canning Co., 1890, 135 U.S. 176, 10 S.Ct. 718, 34 L.Ed. 88. 1 Considerable controversy existed at the trial on the question of whether the Wilmot patent disclosed roller bearings on the backing rolls. No mention of roller bearings, or antifriction bearings, appears in the text of the patent. In the drawings, Fig. 2, a series of small markings appear around the neck of the upper left backing roll. An enlargement of this portion of the drawing was received in evidence and shown to several witnesses. It is around these circles the argument revolves. Most of Cold Metal’s witnesses couldn’t see that they were a representation of roller bearings, while Bliss-Greer’s could. However, Cold Metal’s witness, James T. Reynolds, did interpret the patent drawings as disclosing roller bearings. My conclusion is that the drawing does show a row of circles around the neck of the backing roll from which it can be concluded the backing roll was to be equipped with roller bearings. The mills actually constructed following this patent were in fact equipped with a form of roller bearing on the backing rolls, and the drawings for said mills contain a detailed sketch of a roller bearing. It was not controverted that Cold Metal had represented to the Royalty Adjustment Board in a hearing before said Board that the Wilmot patent did disclose roller bearings on the backing rolls. The Wilmot patent is of record in the file wrapper of ’016. However, Bliss and Greer contend that the Patent Office was not aware of the fact that the patent disclosed roller bearings on the backing rolls. The only time Wilmot is to be found in the ’016 file wrapper is as follows: “Division is required between machine Claims 1 to 13 inclusive and process Claims 14 to 25 inclusive. The process could be carried out by other machinery than that specifieally claimed. The process of Claim 14 could be carried out by apparatus as shown in Wilmot 1,071,846 Sept. 2, 1913 80-38 No action on the merits of claims will be given until the question of division has been disposed of.” (Emphasis added.) Claim 14, to which reference is made, read: “The process of cold rolling strips, comprising pulling a strip past a relatively small reducing roll having anti-friction backing, substantially as described.” It is obvious that Claim 14 does mention antifriction backing. This does not necessarily mean that the Examiner was aware of the fact that Wilmot had roller bearings. His reference to Wilmot was in the category of “other machinery” on which the pulling process could be carried out. It is conceivable that the reference to Wilmot was in relation to the feature of small working rolls, also embraced in Claim 14. The thought that the reference to Wil-mot was meant to cover small work rolls is strengthened by later actions of the Patent Office. When it came to the first consideration of the patent on the merits Claim 14 was rejected on Windlund in view of Mathey or Somers. No mention was made of Wilmot. In the prosecution of Steckel '017 Wilmot ’846 was referred to as disclosing the use and utility of small working rolls. Cold Metal claims that Steckel himself discussed the disclosures of Wilmot ’846 in a personal interview at the Patent Office. Testimony given by him in the Cancellation case was offered to establish that fact. In that case Steckel testified that he recalled that the Wilmot patent was cited against him and the oral interview followed. The file wrappers, as stated, show that it was never raised on the merits. Furthermore, while quite a bit of the patent prosecution was carried out by personal interview, the gist of the discussions and those patents considered were supposedly embodied in amendments following each interview. Nowhere in the file wrappers is there any document showing that Wilmot was ever discussed. In my judgment, Wilmot Patent No. '846 was the most relevant prior art patent. The file wrappers of the Steckel patents do not disclose whether the Patent Office knew that Wilmot had roller bearings on the backing roll necks. If the Patent Office did have such knowledge, it is not understandable why no mention was made of this fact in any of its actions. The mills constructed following the Wilmot patent, referred to as the Bridgeport or Keating cluster mills, are relied upon as a prior public use creating a statutory bar to the validity of Steckel ’195. R.S. § 4886 (29 Stat. 692). These mills were first put into use around 1909-1910. The exact year is unimportant, for it is certain that it was more than two years prior to Steckel’s date of invention. They were installed at the Hancock Avenue Plant of the American Tube & Stamping Co. in Bridgeport, Connecticut. Five stands were put in, four in what is known as a “Belgian Train” and the fifth as a single stand. In 1918 or 1919 three more of these mills were built, so that by 1920 Bridgeport had eight Keat-ing cluster mills in operation. The mill structure itself has been previously described, with each work roll nested between two backing rolls. The mill rolls were 7" wide, and the diameters were 3" for the work rolls and 5" for the backing rolls. The backing rolls were equipped with a form of bearing which has been called “pin”, “needle”, “quill” or “roller” in the course of this litigation. Whatever its label, the physical facts have remained the same. The bearing was composed of a series of rollers 5Ys" long, varying in diameter from to Vie". These rollers ran directly on the neck of the backing roll and were enclosed by a steel shell known as an outer race. Different diameter rollers were not intended to be used at the same time, but rather as a set. As natural wear eroded the surface of the race in which they were enclosed the larger sizes were used. Also included in the construction of the bearing were steel and bronze washers. These were used to take end thrust from the rolls, and are known as thrust washers. More will be said of them later. It is the position of Cold Metal that this bearing was not an antifriction bearing in fact nor in its environment. If it was an antifriction bearing the Keat-ing mills would be entitled to great weight as an anticipation of S.teckel. In its attempt to show that the Bridgeport bearings were not antifriction in fact Cold Metal relies upon the coefficient of friction factor. The coefficient of friction for the Bridgeport bearings was calculated as .0089 by Bliss-Greer’s expert witness and as .0129 by Cold Metal’s expert. Cold Metal contends that neither of these values is sufficiently low to qualify the Bridgeport bearings as antifriction in fact. Reliance is placed upon a definition of antifriction bearing found in a footnote to an article written in 1933. “Anti-friction bearings as applied to roll necks are considered as constituting a class in which the coefficient of friction is less than some relatively low maximum value, perhaps a value of 0.007. Strength of Roll Necks by W. Ti'inks and J. H. Hitchcock” It was also pointed out that the coefficient of a good modern antifriction bearing ranges from .0015 to .0035. I cannot accept these figures as a basis upon which to invalidate the Bridgeport bearing as an antifriction bearing as of the time of its construction. As previously stated, at the time of Steckel’s invention antifriction bearing, meant a roller bearing, and no evidence was offered proving that it meant a roller bearing with any given coefficient of friction. The definition relied upon was given in an article written over 20 years after the Bridgeport bearing was put into use. It. is not a rigid standard — some relatively low maximum value, perhaps a value of 0.007 — and Prof. Trinks admitted that it was intended as a general dividing line and not a precise breakoff. The definition is authoritative for the purpose of the article in which it was used. The important question is whether Bridgeport was antifriction in relation to the sliding bearings of its day, not whether it measures up to today’s standards. As an analogy, let us consider a 1910 automobile, and whether it was a “high speed” vehicle. In relation to, and defined by, the standards of 1933 it would not be. But, in comparison to the other forms of transportation available in 1910 it was. That is the way the Bridgeport bearings must be considered. Cold Metal contends that the coefficient of friction for a 1910 sliding bearing was .01. It is their position that taking .0129 as the correct coefficient of friction for Bridgeport, it was not antifriction in fact. This figure of .01 is derived from an article entitled “Roll Neck Bearings” published by the A.S.M.E. (1935). Examination of the table from which it is taken shows that it is the absolute minimum value given for a sliding bearing. The complete range of values is shown as: In relating the Bridgeport bearings to the sliding bearing it is manifestly unfair to choose either minimum or maximum values for the sliding bearing. We are not concerned with the very best or worst, but rather with that which one would expect to find in everyday operation — the average. The overall average coefficient of friction for sliding bearings based on Cold Metal’s figures is .08. The median of the range with good lubrication is .03. Accepting for the sake of argument Cold Metal’s figure for the coefficient of Bridgeport, it is approximately % of the overall average for sliding bearings, and 2% times smaller than the median value for the best sliding bearings. Of course, if the opinion of Bliss-Greer’s expert is correct, then Bridgeport falls below any value set for the sliding bearing. So far then, it has been developed that the Bridgeport bearing was in physical construction a roller bearing with friction developing qualities 600% less than the average sliding bearing. Here, it should be noted that Cold Metal’s witness Wilson admitted on cross-examination that in principle the Bridgeport bearing was a good bearing. Cf. Atlantic Works v. Brady, 1882, 17 Otto 192, 198, 107 U.S. 192, 27 L.Ed. 438. The last obstacle put up by Cold Metal to this bearing as being antifriction in fact is that it is a cageless bearing. I find no merit in this contention, as the evidence discloses that a well constructed bearing is entitled to be considered as antifriction, be it cage type or cageless. It is my conclusion that the bearings on the Bridgeport mills were in fact anti-friction bearings. If there was a deficiency in the Bridgeport bearings it was one of mechanics, and not of design. The evidence shows that the Bridgeport bearings were a “homemade” job. The rolling elements were cut from lengths of drill rod in Bridgeport’s own plant and the ends were handground to deburr them. This resulted in the production of squared ends on the rollers, whereas the preferable design is with a rounded end. As a result of the rollers being cut as they were, it was sometimes necessary to straighten them out. This was done by pounding them with a lead hammer. It was further stated that the steel from which the rollers were cut was not as hard as that used by commercial bearing manufacturers. Bearing in mind the teaching of the publications of the vital importance of precision in methods and materials in the manufacture of an adequate roller bearing, it is really remarkable that these bearings functioned as well as they did. The claim that these bearings were not antifriction in their environment is based on the actual performance and production of the mills. Testimony was tendered that the mills produced strip that was buckled and riffled in gauges below .020, that strip below that gauge had to be finished on other mills, and that the rolls still heated. Evidence of certain tests made on the mills was shown in support of these claims. The tests shown in this litigation were three in nature. One involved the actual rolling of steel, one is designated the “roll on roll” test, in which the mill was run with no strip in it and the work rolls in contact with each other under heavy compression and the third was for the purpose of measuring heating on the mill. I find that the temperature tests are inconclusive in that they are incomplete. They were intended to compare the relative temperatures of the backing roll necks and bodies, and to show heating in the roll necks. In this regard, continuing readings were made on the roll body of only one of the four backing rolls, only one reading was made on a second roll, and no readings were taken on the other two backing roll bodies at all. Those values which were taken show that except for one highly questionable reading the roll necks were always cooler than the roll bodies. Further, after an initial temperature rise at the beginning of a rolling, the readings on the roll necks were relatively constant, although exhibiting some fluctuation but no steady rise. The compressed roll on roll test, without the strip between the rolls, is pointed to by Cold Metal as proof that the Bridgeport bearings were no good. The reason for this is that the test concluded with the oil on one of the lower rolls catching fire. I find that this test, and its dramatic finale, are unacceptable as proof of the utility, or lack thereof, of the Bridgeport mill as a commercially functioning mill. Cold Metal’s own witnesses admitted that the conditions under which the tests were made were abnormal and did not represent everyday operating procedures. The mill rolls were deprived of their cooling water, the operator was not allowed to lubricate the bearings, and the pressures on the rolls were far in excess of those normally applied. All the test proved was that the mill did not stand up under the procedures agreed upon therefor. These procedures, as has been pointed out, far exceeded the normal operation of the mill. As to the steel rolling tests, the same fact is true in regard to the conditions under which it was made. In considering a test in which reductions were taken from .060 to .020 in four passes on strip 5ys" wide Cold Metal’s witness Reynolds stated: “Well, from experience the bearings would not hold up trying to make these reductions on that mill.” An important sidelight came from the roll on roll test. Frank Pagliaro, the roller on these Bridgeport mills, testified that for years he had been of the opinion that the deficiency in the Bridgeport bearings lay in their inability to absorb end thrust. End thrust is a pressure exerted sidewise by the rolls. It is caused by the rolls being improperly aligned, which in turn may result from the rolls being improperly ground. Excessive end thrust can cause heating at the roll necks if it is inadequately absorbed. The Bridgeport mills, at the time of the roll on roll test, were equipped with bronze and steel washers for the absorption of end thrust. Following the flash fire, Pagliaro observed a condition on these washers which he described as their being virtually welded to the end of the roll. Subsequently, around 1932, several alterations were made in the bearings, but they were principally related to providing thrust rollers instead of washers. Thrust rollers are not shown on the patent drawings in the ’016 application. Steekel, however, did later embody them in his mills. The practice of making their own rollers was discontinued and better quality hardened steel rollers were thereafter purchased from a regular bearing manufacturer. This improved the mills’ ability to take heavier rolling loads. No changes were made in the principle of the bearing so far as its ability to take a rolling load was concerned. It continued as a cageless full complement bearing. In fact, it is still functioning in that same manner today on a full production basis for the Solar Steel Company. After this change was made the mills’ operation improved appreciably. Pagliaro testified that end thrust was a considerable problem in the Bridgeport mills, in that the firm also made their own backing rolls and did a poor job of it. The rolls were often warped and the grinding on them didn’t help matters much. This resulted in poor alignment of the rolls in the mill. The improved ability to absorb the end thrust generated by these rolls would be expected to improve the general performance of the mill, as it did. The Bridgeport mills’ are also attacked on the ground that in order to roll below .020 annealing was necessary, very little finishing work was done on the mill, and strip below .020 was buckled and riffled. The necessity for the anneal at .020 is of no consequence here, for Patent No. ’195 is not directed to the elimination of annealing, and does not mention it at all. The buckling and riffling below .020 and the failure to finish a strip on the cluster mills are inter-related. It is stated that the reason finishing work on strip under .020 was done on other mills was to remove the buckles and riffles produced by the cluster mills. The buckling cannot be attributed to any bearing fault. The testimony established the fact that buckling results from overheating of the center of the work roll from its frictional contact with the strip, and is not the result of roll neck heating. In addition, Pagliaro testified that he purposely “crowned” his rolls to produce a slight buckle. Riffling is a result of roll neck heating. Pagliaro testified that generally only one neck on a given roll would overheat. This would indicate that substantial cause of the roll neck heating was end thrust pressure, as a given roll exerts end thrust on one side of the mill only. The extent to which these strip defects occurred is not agreed upon by the several witnesses. The testimony of Charles Siegel is such that even Cold Metal’s other witnesses disagreed with it and is not of much value. Percy Farwell, for Cold Metal, indicated it was a regular occurrence and that commercial strip off the cluster mills was rare. James T. Reynolds, for Cold Metal, admitted that commercial rolling was done on the cluster mills at .010, but said there were times the strip would show buckling and riffling and it would have to be finished elsewhere. Pagliaro, for Bliss-Greer, said that good commercial strip of .010 could be finished on the cluster mills if the roller had the necessary patience. He, too, admitted to the presence of buckles and riffles. However, he partially attributed the fact that little rolling under .020 was done on the cluster mills to the winding reel in use. The reel had four deep notches cut into it. When the strip got under .020 it was thin enough that the winding pressure might cause it to crimp into these notches, and spoil the entire coil. Hence, such a strip was finished on a mill with a solid reel. This notched reel was abandoned in the early 1930’s in favor of a solid one. In view of the conflicting testimony about the commercial utility of these mills, the cold statistical facts relating thereto are important. They produced several thousand tons of steel a year which was sold to the public. Eight years after they were first put into use the firm built three more of the same design. After the entire Hancock Avenue Plant was shut down in the depression of the early twenties, the cluster mills were the first restored to operation when business picked up. One of the mills is in commercial operation today almost fifty years later, with the same type full complement cageless pin bearing. Even supposing that these mills were as poor as the most critical of the witnesses pictured them to be, where would the designer have learned how to correct them in 1923? Suppose he had been handed Steckel’s original ’016 application. What good would it have done him? It said to use roller bearings on the necks of the backing rolls. That was what he was already doing. Thus, it wouldn’t have helped him at all. On the other hand, suppose he listened to Frank Pagliaro’s ideas about thrust bearings, and read the literature of the art stressing the importance of proper methods and materials. He then, in 1923, could have done what was done in 1932 or thereabouts, and obviated the problems facing him. Steckel’s own view point on the Bridgeport mills is in conformity with the facts previously set forth. In order to view his comments in their proper light a report written by Stephen Badlam of Pittsburgh, in 1927, must be considered. Reporting on The Development of the Backing Roll Mill Badlam wrote: “In 1913 Mr. Frank A. Wilmot secured patents covering certain features of this [Bridgeport cluster] mill * * * Patent No. 1,071,846 dated Sept. 2, 1913. This type of mill proved very successful and a total of nine mills were built between 1909 and 1919. * * * It is certain that the last eight mills were all equipped with roller bearings. * * * * ***** “In 1909 roller bearings were applied to the backing rolls of the cluster type mill as mentioned above, and these were successful; bearings of this same type being in use in these mills today.” Steckel, in writing to Biggert, Presi dent of United, expressed himself thus “As to what was really done at Bridgeport during the past nineteen or twenty years, I found that Bad-lam’s report was substantially true, and if this were all that would bear on the case, we would be in bad. However, I have been trying to account for the fact that neither you nor I ever learned anything, valuable or otherwise, from all this long history at Bridgeport, and the fact that you nor I nor any other engineer with any imagination ever benefited from it can now be explained in a manner that should well influence a Federal Judge to rule in our favor. Blenko strongly urged on me to get the facts as to just how much secrecy there was at Bridgeport up to the time when my own record started at Youngstown nearly six years ago. * * * ****** “Having myself been in the room where all these mills were run, I can satisfy Siebbins that the situation would meet his legal definition of secrecy to the last degree. Anything lacking in the argument of secrecy can well be made up by what the people at Bridgeport will testify as to the limits of their own imagination in improving on the performance of the small mills. *3fr # # “P. S. If the attorneys should find anything lacking in the sole argument of secrecy, they should feel encouraged by the fact that they put your claims .through the Patent Office on the argument of ‘commercial success.’ It can readily be arranged to have Keating testify that he was not on the way toward ‘commercial success’ in the sense that your showing was used at the Patent Office.” In this litigation no claim that the Bridgeport mills were hidden from the public view was pressed with any vigor, and the facts fail to establish any such conclusion. As to the testimony on the Bridgeport mills which Steckel hoped could be arranged, it was not forthcoming as Keat-ing did not testify herein. It is true that the Bridgeport mills were not a “success” in the sense that the United mills at Rome were, but they were a good steady producing mill. With these considerations out of the way it takes one back to Steckel’s original thesis — he is “in bad” on the ’195 patent in light of the Bridgeport mills. It is also asserted by Bliss-Greer that a 3-high mill of the Edgar T. Ward Co. is a complete anticipation of Steckel ’195. Although there was some testimony relating to that mill, it is insufficient to establish anything worthy of serious consideration. The sale of the Rome mill and its use for commercial production and for experimental purposes in December, 1925 is urged as a statutory bar (two years) to the claims asserted in the amendment of May 28, 1928. It is the position of Bliss-Greer that there was no adequate disclosure for these claims in Steckel’s original application. The decision of the Supreme Court in Muncie Gear Works, Inc. v. Outboard Marine & Mfg. Co., 1942, 315 U.S. 759, 768, 62 S.Ct. 865, 86 L.Ed. 1171 is relied upon. The Patent Office ruled that there was adequate disclosure for the claims in the amendment in Steckel’s original application and he was entitled to the filing date of June 30, 1923. This point was also urged and rejected in the Republic case. The defense is without merit. This concludes the survey of the art relevant to Steckel ’195. Its collective import can now be considered. The publications of the prior art widely proclaimed the advantages to be derived from the use of roller bearings in a rolling mill. These are the same advantages enumerated in the ’195 specifications as attributes of the Steckel invention, while in truth they are attributes of the bearing itself. These same publications indicated that it was the inability of the bearing manufacturer to produce an adequate roller bearing that was holding back their adoption in high pressure mills. The prior art patents of Wilmot and Kriwan taught the application of roller bearings, both as to roll necks and as a backing ring as Steckel first utilized them. The Bridgeport-Keating cluster mills used roller bearings on their backing rolls. It is true that these mills had their deficiencies. A major source of the difficulty can be attributed to the inadequate thrust bearings and the homemade rollers and backing rolls with which they were originally equipped. On the overall picture, any shortcomings in the Bridgeport mills in the pre-Steckel era was one of mechanics and not of principle, cf. Atlantic Works v. Brady, supra. I can find no invention in being the first person to get an adequate roller bearing and do what was predicted, advocated, taught, and attempted for many years prior thereto, but which the prior workers had been unable to achieve with a high degree of precision because of mechanical drawbacks. Counsel for Cold Metal has cited the case of Eibel Process Co. v. Minnesota and Ontario Paper Co., 1923, 261 U.S. 45, 43 S.Ct. 322, 67 L.Ed. 523 as authority for the sustaining of Patent No. ’195. In that case the Supreme Court upheld the validity of an improvement patent on a paper making machine. The change in the machine under the patented process was slight, but resulted in a great increase in the operating speed and efficiency of the apparatus. The opinion states: “In administering the patent law, the court first looks into the art, to find what the real merit of the alleged discovery or invention is, and whether it has advanced the art substantially. If it has done so, then the court is liberal in its construction of the patent, to secure to the inventor the reward he deserves. If what he has done works only a slight step forward, and that which he says is a discovery is on the borderline between mere mechanical change and real invention, then his patent, if sustained, will be given a narrow scope, and infringement will be found only in approximate copies of the new device.” 261 U.S. 45, 63, 43 S.Ct. 322, 328. This is a sound principle of law, but the distinction of facts between the Eibel case and that at bar detracts from its value as authority. In Eibel, the improvement was an original conception of the inventor. Here, the improvement was in the utilization of a well-known expedient. In Eibel, the Supreme Court found that the prior art did not suggest the patented improvement. Here, the state of the art taken in conjunction with the prior publications, which the Supreme Court stated should be the first thing considered, shows that the idea of using roller bearings was old, but that the means had theretofore not been available to do so in a precise manner. In Eibel it is said “The defendant’s witnesses without exception refer to that disclosure as something that surprised and startled the paper-making trade.” There was nothing startling about putting roller bearings on the backing roll necks of a 4-high mill. The improvements in both the paper making machine and rolling mill resulted in increased speeds of operation. However, while the increased speed of the paper making machine was totally attributable to the new apparatus, the same cannot be said for the rolling mills. Around the time of Steckel it was necessary to “hand mike” the strip to check on its accuracy to gauge. The roller would put his hand micrometer on the moving strip and move with it as he got his reading. This practice necessarily resulted in a limitation on the speeds at which the mill was run, in that it had to be kept low enough so that “hand miking” was possible. On modern mills a “continuous gauge” is used. This is an automatic, or electronic, gauging device attached to the mill itself. This device checks the accuracy of gauge without any speed limitations. As fast as the mill will run, that is as fast as the “continuous gauge” will operate. L. A. Beeghly, a witness for Cold Metal, admitted that the development of the continuous gauger brought about a substantial increase in the productivity of rolling mills. However, he felt unqualified to comment on a statement made by Steckel himself that the continuous gauger resulted in about a 40% increase in productivity. On the basis of Steckel’s testimony it can only be safely said that the advent of the roller bearing mill itself accounted for about 60% of the modern mill’s increased productivity. Patent No. ’195 has been sustained in prior litigation with great emphasis on the commercial success of the mills based on its design. It is true that when the 4-high roller bearing mill was brought to fruition the old method of pack rolling was thereafter gradually abandoned. This is claimed to establish the commercial success of Steckel. Rather impressive figures of monies collected by Cold Metal were also introduced to establish commercial success. In this regard, it is to be noted that approximately one-third of the total amount collected resulted from licenses and settlements executed on the eve of the expiration of the excess profits tax law. A substantial part of the remainder was collected from other steel companies under threat of suit or in settlement of suits already instituted. This is by no means as impressive as it would be if a willing buyer and seller had entered into the agreement. It is some evidence that the licensees preferred to make their peace rather than fight, and not too much can be judged from it. Ruben Condenser Co. v. Copeland Refrigeration Corp., 2 Cir., 1936, 85 F.2d 537, 540, certiorari denied 1937, 300 U.S. 665, 57 S.Ct. 508, 81 L.Ed. 873. In addition, only 28 of the 52 licenses from which Cold Metal’s wealth was accumulated included ’195 as one of the patents licensed. Of those 2