The Texas Company (sometimes referred to as plaintiff or Texas) brought this action against Globe Oil & Refining Company (sometimes referred to as defendant or Globe), for the infringement of patent No. 1,883,850, issued to O. Behimer October 18, 1932, upon an application originally filed November 21, 1918, subsequently assigned to plaintiff, entitled 'Process for Making Gasoline.' Defendant pleaded numerous defenses, including that of non-invention, invalidity over the prior art, non-infringement, that the equity sought to be asserted was barred by laches and that plaintiff was guilty of unclean hands with respect to the relief sought.

The issues involved were referred to Honorable Joseph F. Elward, a Master-in-Chancery, for a hearing and a report. Such report was in the main approved by Judge William J. Campbell of the District Court, who in connection therewith filed an opinion. Texas Co. v. Globe Oil & Refining Co., D.C., 112 F. Supp. 455. In accordance with the recommendations of the Master, the court decided the issue of infringement adversely to plaintiff and all other issues pleaded as a defense, in its favor. Plaintiff, in No. 10954, appeals from the adverse decision on the issue of infringement. Defendant, in No. 10955, cross-appeals only from that portion of the decree which fails to award reasonable attorney's fees to the defendant as part of its general costs.' The time and labor devoted to this litigation, as is evidenced by the record, is appalling. The Master's report as filed in this court occupies a volume of 436 pages of the printed record, and Judge Campbell's opinion covers some 40 pages of the Federal Supplement. The Master in his report points out that hearings were conducted by him over a period of 71 days, at the conclusion of which printed briefs were submitted, consisting of 1402 pages. And this court has not been neglected, because we have been favored with briefs which cover some 400 printed pages.

In the beginning we think it not inappropriate to recognize the careful and thorough consideration given this case by the District Court and its Master, as well as the manner in which able counsel have presented to this court the contentions of the respective parties.

We shall first consider the issue of infringement. It is pertinent to note that the Tenth Circuit, in Texas Co. v. Anderson-Prichard Refining Corp. (sometimes hereinafter referred to as Anderson), 122 F.2d 829, opinion by Judge Phillips, in a suit by the instant plaintiff involving the same patent affirmed a District Court, 32 F. Supp. 347 which had dismissed the complaint on a holding of non-infringement of a process employed by the defendant in that case of the same type as that employed by the defendant in the instant case. We refer to the drawings of the apparatus shown by the patentee for the utilization of his process, reproduced by the court in Anderson, 122 F.2d at page 832.

'This invention relates to methods of making condensable light oils, such as gasoline. More particularly it relates to certain novel improvements in the cracking of hydrocarbons whereby higher boiling hydrocarbons are decomposed into those of lower boiling point.'

'In prior methods of cracking hydrocarbon oils, considerable difficulty has been experienced on account of carbon formations, which occur on the sides of tubes and stills exposed to the heat required to carry on the cracking operation.'

'It is a broad novel feature of the herein disclosed process that substantially all of the cracking operation occurs in a vessel to which no external heat is applied, except at such times and in such quantities as are necessary to compensate for heat losses, the oil prior to its introduction to such vessel having been subjected to a high degree of cracking heat and the excess heat of the oil itself being used to effect its own decomposition. As a consequence, I avoid substantial carbon formations of the destructive character which have proved such an impediment to prior systems. The removal of the cracking operation from the heating zone also reduces the danger from fires such as frequently occur where cracking is carried on in highly heated stills and coils exposed to direct heat.'

'Thus, although the oil is subjected to cracking heats in the heating zone, this temperature is attained only just previous to the exit of the oil therefrom, and therefore, the oil leaves the heating coil before any substantial decomposition and incident deposition of carbon takes place. Subsequently, the highly heated oil in a state of incipient decomposition, is delivered to the cracking zone, where the desired temperature and pressure conditions are continuously sustained and the cracking of the oil and the incident decomposition of carbon are effected. The bulk of the carbon is removed from the circuit by withdrawing from the cracking zone, portions of the residual oil.'

This process of heating oil in the coils to a high temperature without substantial cracking and delivering the same to a vessel or drum where cracking takes place is known as 'postponed cracking.' The purpose was to obviate the problem of the formation of carbon in the coils where intense heat was applied by transferring the cracking operation to a vessel or drum which was heated only lightly, if at all.

We take it there is no dispute up to this point, but plaintiff contends that the patent discloses not only the process of 'postponed cracking' but another independent process dubbed 'clean circulation.'

"Clean circulation' is associated with 'postponed cracking' in the specific embodiment of the process described in the specification of the original application. The patent in suit, which issued on a division of the original application and hence with essentially the same disclosure, has claims only on 'clean circulation', some limited to the combination with 'postponed cracking' and some without such limitation. It is the latter only which are involved in this suit.'

On the other hand, defendant does not dispute but that the patent discloses 'clean circulation' but contends that it is not a separate and independent process or invention, that it is auxiliary and to be employed only in connection with Behimer's disclosure of 'postponed cracking.'

The Master as to this theory of 'clean circulation' found:

'Plaintiff claims that an essential feature of the patent-in-suit and of the claims-in-suit is 'clean circulation' which plaintiff defines as a continuous coil, and drum pressure cracking operation, whereby cracked residual oils are withdrawn from the drum and discharged from the system; the separated vapors pass from the drum into a dephlegmator, in which vapors heavier than gasoline are condensed as reflux and the reflux from the dephlegmator is directly returned while hot to the heating coil and is forced under mechanical pressure through the coil with the fresh oil supplied thereto.

'Hereinafter, the term 'clean circulation' is to be understood as used in the above sense-- unless, of course, the context should otherwise indicate.'

'Clean circulation' is usable, and of advantage, with the principal cracking in the coil, with the principal cracking in the drum, or with substantial cracking in both the coil and the drum.'

The process employed by defendant charged to infringe is known as the Winkler-Koch coil cracking process for making gasoline. It has been employed by defendant in commercial operation since 1929, at its plant located in Lemont, Illinois. For all relevant purposes, this was the same process employed by the defendant in the Anderson case, supra, which was held not to infringe the instant patent. Defendant's process is described in much detail in the Master's findings, which we think need not be reiterated. For a description of defendant's process see Judge Campbell's opinion, 112 F.Supp. at page 461.

The Master, relative to defendant's process, found:

'The cracking process operated by Defendant in its light oil unit at Lemont is a clean circulation process as hereinbefore defined.'

The Master in detail set forth the elements of each of the five claims in suit and found as to each:

'* * * that there is a substantial, literal correspondence between each of the elements (of all the claims in suit) and Defendant's process.'

'This leaves the question of law-- how far literal correspondence of the accused process with the claims-in-suit is sufficient to establish infringement. I shall deal with this law question later under my 'Conclusions of Law'.'

The Master in his conclusions poses the question of law as follows:

'Does Defendant's operation infringe the claims-in-suit? If (as Plaintiff contends) the claims-in-suit should be construed in accordance with their express language, as including cracking in any zone (either coil or drum, or both), Defendant infringes; if (as Defendant contends) the claims-in-suit should be construed as having an unexpressed but implied limitation to 'cracking in the drum only', Defendant does not infringe. Or stated otherwise, if (as Plaintiff contends) the disclosure of the Behimer specification included several novel features as to each of which Behimer sought, and the Patent Office granted, independent patent protection, and if the claims-in-suit were intended by Behimer and the Patent Office to protect the invention of clean circulation independently of cracking locus, then Defendant infringes; but if (as Defendant contends) the Behimer specification disclosed only one novel feature, 'postponed cracking', and if Behimer and the Patent Office intended by the claims-in-suit to provide patent protection on clean circulation only when used with postponed cracking, then Defendant does not infringe. The answer, therefore, to the question of Defendant's infringement depends upon the construction of Behimer's invention and the claims-in-suit and I discuss that construction hereinbelow.'

Thus, the issue is whether the claims in suit, read in connection with the specifications of the patent, disclose a process of 'clean circulation,' irrespective of the locus of cracking. Plaintiff, of course, recognizes that claims are to be construed in the light of the specifications but argues that the District Court applied this rule of construction in an erroneous manner. Plaintiff in its brief on this phase of the case states the sole contested issue as one of law, as follows:

'Should patent claims be construed as containing an unexpressed limitation to a specific embodiment of the invention shown in the specification when such limitation is not required by the prior art, and was not intended by either the patentee or the Patent Office?'

Both the Master and the District Court considered numerous factors, as this court is urged to do, in construing the claims in suit. These factors, or at least some of them, will be subsequently referred to. The guiding star, however, must be the phraseology employed by the patentee in his claims, read in the light of the specifications. And when so read, could a person skilled in the art comprehend so as to utilize a process which it is now claimed was taught as his invention? It is of no benefit to speculate on what the patentee might have claimed or taught, or what he or his assignee, many years later, particularly in view of subsequent developments, may think could have been stated or claimed.

At one point in our study of this record we were impressed with the soundness of plaintiff's contention that the claims in suit, when read in the light of the specifications, disclosed an invention of 'clean circulation,' that is that 'clean circulation' was taught as separate and independent from the conceded invention of 'postponed cracking' and not merely as auxiliary or incident thereto. However, our doubt on this score has increased with our study of the situation until we have reached the firm conviction that the theory is unsound and must be rejected. This conclusion is reached, irrespective of all other considerations, from an analysis and understanding of the claims, and particularly when considered in the light of the specifications. The most charitable view which can be taken of Behimer's conception of 'clean circulation' is that he taught its use solely in connection with and as a step in furtherance of his process of 'postponed cracking.'

Five claims are in suit: 12, 22, 36, 39 and 40. Judge Phillips in Anderson treated 39 as typical, while Judge Campbell in the instant case treated 12 as such. It is not contended that there is any functional difference among the five claims, which obviates any occasion for separate consideration. It is plaintiff's contention, of course, as noted, that all these claims teach the process of 'clean circulation' as a separate invention. Claim 12, broken into elements, is as follows:

'A process of oil conversion that comprises

(1) passing the oil once through a heating coil where it is raised to a cracking temperature

(2) and into a drum where a substantially constant body of liquid oil is maintained and vapors evolved,

(3) separating the heavier constituents of the evolved vapors

(4) and returning them while still hot and without substantial drop in pressure under mechanical pressure to the heating coil for retreatment

(5) removing residuum from the drum while preventing recirculation of residuum through the coil

(6) maintaining superatmospheric pressure through the coil and drum

(7) and continuously supplying charging stock to the process.'

Thus it will be noted that the claim is entitled 'A process of oil conversion.' The other claims in suit are similarly entitled: claims 22 and 36, 'A process of converting heavy into lighter hydrocarbons'; claims 39 and 40, 'A process of cracking oil.' It is evident, we think, that 'clean circulation' in itself is not 'A process of oil conversion,' 'A process of converting heavy into lighter hydrocarbons,' or 'A process of cracking oil.' Elements 1, 2 and 3 of claim 12, which we have set forth as typical, relate to the cracking process, while elements 4, 5, 6 and 7 relate to 'clean circulation.' It is asserted by defendant, and we think correctly, that 'clean circulation' standing alone performs no useful function. It acquires utility only when used in connection with a cracking process.

We think that plaintiff would not take issue with this reasoning, but admittedly it does not entirely meet its contention that 'clean circulation' is useful, as recognized by the Master, irrespective of the locus of cracking, that is, in the coils, in the drum, or in both. It is contended, and the Master so found, that the claims in suit do not specifically designate the locus of cracking and, as a consequence, the claims read literally on defendant's process where cracking takes place in the coil, and that infringement results. While the claims do not in precise words fix the locus of cracking, they do, in our judgment, fix such locus as certainly and effectively as if express language had been used. As illustrated, claim 12 provides for 'passing the oil once through a heating coil where it is raised to a cracking temperature (2) and into a drum where a substantially constant body of liquid oil is maintained and vapors evolved,' and claim 22, 'passing a stream of oil in a coil through a heating zone where the oil is heated to a cracking temperature (2) the thence to a zone where distillation takes place.' Thus the oil is raised 'to a cracking temperature in the coil or zone' before it passes into a drum where distillation takes place and vapors evolve. In our view, this language cannot reasonably be interpreted as calling for anything other than heating in the tube and cracking in the drum. As Judge Phillips stated in the Anderson case, 122 F.2d at page 841:

'Each of the claims in suit specifically states that it embraces a process of cracking oil. * * * But since the claims cover an oil cracking process, cracking must take place somewhere in the process. In the typical claim, 39, it must be referable, we think, to the language, 'where separation of vapors from residual oil takes place' in the drum.'

We have doubt as to the validity of the conclusion of the Master and the District Court that there is a literal correspondence between defendant's process as employed and the claims in suit. Defendant not only heats the oil to a cracking temperature in the coil but cracking itself there takes place. It is only subsequently that the oil in its cracked state passes into a drum or zone. Behimer by his claims, as already noted, only calls for heating to a cracking temperature in the coil and thence the oil passes to a zone or drum. To state that this does not determine the drum as the locus of cracking is to assert that the claims do not call for cracking as a step in the process. The omission of this essential step would make the claims meaningless and raise a serious doubt as to their validity. More than that, it would indicate that the purpose of the draftsman was to confuse rather than to enlighten those who might be interested in the scope of the invention. Any doubt which we have, however, relative to the conclusion that there is a literal correspondence between defendant's process and that called for by the claims is of little consequence because the District Court, reading the claims in the light of the specifications, construed them as called for 'postponed cracking.'

Any merit in plaintiff's contention relative to its theory as to the proper construction of the claims is dispelled by reading them in the light of the specifications. The isolated portions of the specifications set forth by plaintiff in its brief do nothing more than furnish the basis for the conclusion that Behimer disclosed 'clean circulation.' The specifications at the same time, particularly when considered as a whole, disclose irrefutably that such 'clean circulation' was merely a step to be utilized relative to and in connection with what plaintiff concedes was the 'specific embodiment of the process described,' that is, 'postponed cracking.'

We have heretofore recited portions of the specifications relative to the process of 'postponed cracking.' At the conclusion thereof, the specifications state:

'Although the process, in its broad conception, is complete at this stage of the operation, the light products preferably continue the cycle, leaving the cracking zone in the form of vapor and gas.'

In other words, after the process which he disclosed, that is, 'postponed cracking,' had been completed, Behimer suggested as 'preferable' that certain products continue the cycle, and described what is asserted to be 'clean circulation.' In our view, it cannot be reasonably concluded that his suggestions as to what might be done after completion of the process had reference to any process other than that which he had specified, that is, 'postponed cracking.'

The portions of the specifications upon which plaintiff relies, as set forth in its brief, do nothing more than describe the advantages which flow from 'clean circulation.' They furnish no support for the contention that 'clean circulation' is taught independently of the locus of cracking; in fact, when read in connection with other provisions of the specifications, it becomes clear that 'clean circulation' was disclosed solely in connection with 'postponed cracking.'

We think that this is a fact capable of clear demonstration but which will require rather extensive quotation from the specifications, not only those isolated portions quoted by plaintiff in its brief but those portions preceding and following the quoted language. It is only in this manner that the relied upon language can be placed in its proper perspective and its meaning fairly ascribed. In quoting from the specifications, we place in italics those portions relied upon and quoted by plaintiff in its brief as disclosing 'clean circulation' as a separate and independent process.

As already noted, the patentee, after disclosing his coil heating, drum cracking process, stated that the process 'in its broad conception, is complete at this stage of the operation.' Subsequent portions of the specifications relate to 'clean circulation,' that is, the steps to be taken after the cracking process had taken place which the patentee designated as 'preferable.'

'My process is to be distinguished from those of the prior art, in which a circuit of liquid residuum is kept in circulation. When such oil is kept in circulation, considerable quantities of carbon are formed, owing to the readiness with which this type of oil carbonizes when exposed to high temperatures, the carbon forming in fine particles in the oil, and wherever it comes in contact with heated metal surfaces, it attaches itself thereto and builds up deposits, which often cause the choking of tubes and coils. It also crystalizes the metal surfaces, thereby decreasing the tensile strength of the metal and causing the bulging and breaking of such surfaces. Any residual oil resulting from the cracking operation will necessarily contain a considerable amount of heavy polymerization products and carbon held in suspension in the oil, and to circulate such a residue through heated tubes is to continually pass into or produce in such tubes a large amount of free carbon. In my system, I do not circulate a residuum, but instead keep the residual oil removed from the application of external heat and withdraw the heavier portions thereof, which contain the bulk of the carbon, from the cycle at the cracking zone, and continue the cycle by means of the vapor generated, which contains the condensable product desired and also certain heavier vaporous constituents which are used to complete the cycle. By removing the heavier residuum from the cycle, I keep the circuit free from this source of carbon formation, and since I conduct substantially all of the cracking in a vessel to which either no external heat is applied or only a small quantity of heat is applied to maintain the requisite temperature conditions, I am thus able to effectively remove the carbon formed.'

It will be noted that Behimer commences this paragraph in the singular, 'My process,' which indicates that he is referring to the process previously described, that is, coil heating and drum cracking, rather than to a separate process of 'clean circulation.' Any doubt, however, as to the process about which he speaks is dispelled by the concluding lines, 'since I conduct substantially all of the cracking in a vessel to which either no external heat is applied or only a small quantity of heat is applied to maintain the requisite temperature conditions, I am thus able to effectively remove the carbon formed.' In other words, to effectuate the express purpose of the invention, that is, to avoid the formation of carbon when oil is cracked in the tubes, he stated, relative to his 'clean circulation' suggestion, that the cracking is to take place in a vessel to which no substantial amount of heat is applied.

'The kerosene condensate, preferably at a temperature, not materially below its boiling point, is injected or otherwise suitably forced under mechanical pressure into the heavy oil charge entering and flowing through the heating tubes whereby a foamy agitated mixture is produced by the acceleration of the circulation in the heating tubes resulting from the increased volume of oil therein, thereby preventing the settling and flow of the heavier portions of the oil along the bottom of the highly heated tubes and the consequent local overheating and carbonization of the oil, thus keeping the tubes substantially free from any slight carbon formations. These kerosene constituents are an excellent mechanical conductor of heat, and there is thus produced a heat cycle which materially assists in maintaining the proper temperatures in the heating and cracking zones. By maintaining these bodies in constant circulation, the kerosene constituents, which under existing methods have been found so difficult to decompose into lighter products, are, in my system, ultimately reduced to gasoline.'

Again this language clearly refers to the process which the patentee so frequently described, that is, heating in the coil and cracking in the drum.

The patentee then described in considerable detail the preferred form of apparatus adapted to the carrying out of 'My process.' In the discussion relative thereto he stated:

'The introduction of this hot condensate into the charging line increases the volume of oil in the heating coil and thereby accelerating the flow therein and producing a foamy mixture so that local overheating is prevented and the coil is kept substantially free from any slight carbon formations. This condensate consists in a type of hydrocarbons which when decomposed does not yield any great amount of carbon so that the oil charge is constantly being diluted with oil capable of being decomposed into lighter products without the production of large amounts of carbon. This condensate reaches the jet in a heated condition, preferably at a temperature not materially below its boiling point, so as to heat the charge of oil entering the heating coil. The heat cycle formed by the kerosene constituents contributes to the maintenance of the proper temperatures in the heating coil and the cracking drum. By injecting or otherwise suitably forcing this condensate under applied mechanical pressure into the heating coil I increase the volume of oil flowing therein without increasing the total amount of oil charged into the system at any given time.'

Thus, the forced return of the hot reflux from the drum to the charging line (coil) is specifically stated to be in connection with the process of heating in the coil and cracking in the drum. The statement contained in this last quotation, relied upon by plaintiff (italicized), is an illustration of the fallacy of relying upon a statement removed from its context, because such statement read in connection with that which immediately precedes and follows leaves no room for doubt but that it was made in connection with the process that called for heating in the coil and cracking in the drum. The description of the apparatus recommended for use in carrying out the patentee's process continually refers to the fact that heating is to take place in the coil and cracking in the drum. There is not even a hint that the patentee contemplated that his teachings would be utilized in any other manner. In fact, so much emphasis is placed upon the importance of heating in the coil and cracking in the drum that any different or contrary operation is as conclusively negated as though it had been expressly so stated.

Thus, predicated upon the claims read in the light of the specifications, there is no logical basis for the contention that Behimer disclosed 'clean circulation' as a separate and independent invention. It is shown only as a step auxiliary to and in connection with a coil heating, drum cracking process. We are in agreement with the statement by Judge Phillips in Anderson, 122 F.2d at page 841:

'Behimer solved the carbon problem by inhibiting cracking in the coil, and effecting the cracking in a drum to which substantially no external heat is applied.'

Also we agree with the statement by Judge Campbell in referring to the specifications, 112 F.Supp. at page 467:

'This language seems to indicate that Behimer conceived his invention as prohibiting any substantial cracking in the coil and that coil heating and drum cracking was not only a specific embodiment of the general invention, but was the invention itself.'

As shown, particularly in commercial operation, there is some incipient cracking in the coil but it is not substantial. It undoubtedly was Behimer's objective, however, to avoid as far as possible any coil cracking. Where heating and cracking took place in the same tube or vessel, it was the formation of carbon on the sides thereof which had plagued prior methods. It was to obviate this difficulty that Behimer conceived the process of heating only to a cracking temperature in the tube and passing the oil in its highly heated state into a drum to which little or no heat was applied and in which cracking took place. It was in connection with this process and as auxiliary thereto that the patentee disclosed 'clean circulation.' Plaintiff's theory, if accepted, either that 'clean circulation' was shown as a separate invention or that it was shown in connection with an oil cracking process other than one where the oil was heated in the tube and cracked in the drum, would result in a repudiation of all that Behimer expressly taught and disclosed.

While plaintiff recognizes that patent claims may be read in the light of the specifications, it urges that the District Court applied this rule in an erroneous manner. It is argued, 'The claims in suit read without ambiguity on a 'clean circulation' cracking process regardless of the locus of cracking.' We do not agree with the premise upon which this argument rests. While the claims do not expressly specify the locus of cracking, we think they must reasonably be construed to call for cracking in the drum, and this without reference to the specifications. However, if this be an erroneous view, we have the situation where the claims, designed to fix the metes and bounds of the invention, fail to designate any point as the locus of cracking. More than that, the claims while purporting to describe a cracking process fail to include the essential step of cracking. They only call for heating the oil in a tube to a cracking temperature and then its delivery into a drum.

Thus, if the claims are read literally, as plaintiff would have us do, we find not only a failure to designate the locus of cracking but cracking itself. Such being the situation, we do not think it can be logically asserted that the claims are free from ambiguity. As heretofore noted, the claim elements which relate to 'clean circulation' are meaningless except when employed in connection with the cracking of oil. Assuming, contrary to what we think, that the claims reasonably construed fail to designate a locus of cracking, resort must be had to the specifications for that purpose; otherwise, a serious doubt would arise as to their validity. And when such resort is had, the patentee definitely fixed the locus of cracking in the drum and not elsewhere.

We agree with the statement contained in the Master's report:

'Nowhere, neither in the claims-in-suit nor the specification, was the return of reflux described as a separate and distinct step which might be employed in any oil cracking process, but throughout the specification, the return of reflux is described as an optional adjunct to the process of the specification.

'I am mindful that Behimer, in his specification speaks about continuing the cycle, that this he says is after his process 'in its broad conception' is already complete; his language seems to me to be susceptible of only one interpretation, namely that his invention, as he conceived it, was then complete with the coil heating and drum cracking steps and that the further step of re-cycling of reflux was something that was optional or preferable, but not a necessary part of the invention.'

There is inherent in plaintiff's argument that the court erroneously read the claims in the light of the specifications the premise that 'clean circulation' was disclosed as a separate invention. Obviously, under such a situation it would be immaterial whether cracking took place in the coil, the drum, or both, and there would be no point in resort to the specifications for the purpose of resolving any uncertainty in that respect. Holding as we do, however, that there was no separate invention but that the disclosure relative to 'clean circulation' was only incidental to and in connection with the cracking process, it becomes material to ascertain what the patentee taught relative thereto, and any ambiguity in the claims on this score could be resolved, which was done, by the District Court by reference to the specifications. It follows that the District Court properly construed the claims so that they would read only on a process which called for cracking in the drum. If that resulted in the implied limitation of which plaintiff complains, it was only because such limitation was expressly taught by the patentee. This also disposes of plaintiff's contention that the District Court limited the claims to the primary purpose of the disclosure. The controlling point is that there was no disclosure other than that which plaintiff designates as primary, that is, the coil heating, drum cracking process with 'clean circulation' as an auxiliary step.

In our view, the rule which permits the reading and construction of claims in the light of specifications was not erroneously applied by Judge Campbell. The cases reveal some inconsistencies in the decisions as to the manner and extent of the application of this rule. No good purpose could be served, however, in a discussion of such cases, for two reasons, (1) we are satisfied that the cases adequately support the action of the District Court on this point, and (2) Judge Campbell in his opinion (commencing 112 F.Supp. on page 464) has cited and discussed the cases in considerable detail, and a reiteration of his analysis would unduly prolong this opinion. See also statement by Judge Phillips in the Anderson case, with cases cited, footnote 12, 122 F.2d at page 840.

Moreover, the statute definitely determines the relationship which patent claims bear to the specifications. The claims must point out the subject matter of the invention, the specifications must describe that invention. Title 35 U.S.C.A. § 112 provides:

'The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor of carrying out his invention.

'The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.'

While it is a rule that the scope of the grant is measured by the claim, yet it is true, so we think, that such grant cannot be broader than the invention described in the specifications. And we reiterate that no person skilled in the art, or otherwise, could have obtained the slightest idea from a study of Behimer's specifications that he taught 'clean circulation' as a separate process or that he suggested its use in any manner other than in connection with heating in the tube and cracking in the drum. Therefore, the claims, if their validity is to be sustained, must be confined to the invention described in the specifications. When so construed, they are not infringed by defendant's process.

As previously mentioned, there are numerous factors relied upon by plaintiff in support of its contention that 'clean circulation' constituted a separate invention or, in any event, that it was an invention irrespective of the locus of cracking. All of such factors embody nothing more than rules of construction which in some instances may properly be utilized by a patentee in aid of a favorable construction. None of them, however, will suffice to show an invention which the patentee has not disclosed in his specifications and claims. The most important of these factors so strongly relied upon by plaintiff are what is referred to as the doctrine of claim differentiation and the fact that four separate patents issued upon Behimer's original specifications.

As already shown, Behimer's original application was filed November 21, 1918, which resulted in the issuance of four patents, the original and three divisional, including the patent in suit which issued October 18, 1932, almost fourteen years after the application was filed. All of these patents were issued upon the original specifications, without change here material. The patent in suit contains 44 claims but, as already noted, only claims 12, 22, 36, 39 and 40 are in suit. Some of the claims expressly provide for 'clean circulation' only when used with 'postponed cracking.' Typical is claim 43, which specifically provides that the oil in the coil be raised to a cracking temperature but that it be removed therefrom 'in a state of incipient cracking but before substantial cracking is effected.'

Patent No. 2,027,014 issued January 7, 1936, and its claims call for 'postponed cracking' without the element or step of 'clean circulation.' This patent was one of the four issued upon the original application.

The history of the proceedings in the Patent Office which resulted in the issuance of four patents (including the one in suit) was extremely lengthy and involved. Included in such proceedings was Interference No. 55,610, with the claims of an application by Carbon P. Dubbs, filed subsequent to Behimer's application, which extended over a period of eight years. Both Judge Phillips in the Anderson case (footnote 122 F.2d at page 838) and Judge Campbell in the instant case (commencing 112 F.Supp. at page 467) have set forth much of the Patent Office history, particularly as it relates to the so-called claim differentiation rule and the fact that four patents were issued upon the same specifications. No purpose could be served in an attempt on our part to resolve the many inconsistencies and contradictions disclosed by such history. We think Judge Phillips in Anderson properly summarized the situation, 122 F.2d at page 838:

'The record reflects many instances where, in the patent proceedings, Behimer interpreted the process of the Behimer invention as a coil heating, drum cracking process which solved the carbon problem by preventing cracking in the coil, thereby removing the cause of carbon formation therein; interpreted the Behimer process as permitting no substantial cracking in the coil; analyzed the original Behimer specification and claims and the early prosecution as clearly teaching heating in the coil without substantial cracking; distinguished coil cracking processes such as Dubbs and Trumble from Behimer's invention because they dealt with the carbon problem by limiting the cracking in the processes and by sweeping the carbon from the coil with velocity, whereas Behimer solved the carbon problem by avoiding its cause, namely, cracking in the coil; recognized that the vaporization occurring in a coil cracking operation after pressure release into a drum is not the vaporization occurring in Behimer's drum where a substantial body of oil is maintained and vaporization results from cracking; recognized that the step of returning reflux to the inlet of the coil was a mere auxiliary refinement in the basic coil heating, drum cracking steps; recognized the lack of equivalency between cracking in a coil, and heating in a coil and cracking in a drum as in Behimer, by distinguishing other disclosure and references on that ground; recognized that the Behimer process embodied a flow of oil whereby the cracking temperature is reached only at the time the oil emerges from the coil, as contrasted with the opposite flow in coil cracking processes.'

Judge Campbell, after review of many cases, concluded with reference to the claim differentiation principle, 112 F.Supp. at page 468:

'The application here of this principle seems to me clear; in view of the language in the specification (hereinabove quoted) a broad construction of the claims in suit to include coil cracking-- is not only, not disclosed in the invention of the specification, but is there disclosed as something transcending the invention, indeed opposed to it; therefore, in my opinion the doctrine of 'claim differentiation' does not sustain plaintiff's construction.'

The fact that the claims in suit do not specifically designate the locus of cracking, while there are other claims which do, evidences an intent and purpose both on the part of Behimer and the Patent Office, so it is argued, to award an invention on 'clean circulation,' and this irrespective of the locus of cracking. In support of its argument on this point, great reliance is placed upon Smith v. Snow, 294 U.S. 1, 55 S. Ct. 279, 79 L. Ed. 721. It is true the court in that case applied the doctrine of claim differentiation in aid of the construction sought by the patentee for the claims in suit. The doctrine was not employed, however, to bolster a claim so that it would measure an invention different and contrary to that disclosed in the specifications. The application of the doctrine here, as urged by plaintiff, would result in a construction by which the claims would embody an invention which Behimer not only failed to teach or disclose but one directly contrary thereto. This court, in Burroughs Adding Mach. Co. v. Felt & Tarrant Mfg. Co., 243 F. 861, 870, in considering the rule under discussion stated:

'But this rule of construction of patent claims is entirely subordinate to the fundamental and controlling rule that a patentee's broadest claim can be no broader than his actual invention, no matter how it may be expressed or what other claims his patent may contain.'

We agree with Judge Campbell's conclusion that this claim differentiation is of no aid to the claim construction sought by the plaintiff. In doing so, we think there was incongruity in the action of the Patent Office in allowing one claim which specifically designates the locus of cracking and another which does not, assuming that such was the intent and purpose. It is not reasonable to think, however, that either of the parties intended that the claim should be broader than the invention disclosed. If such a result followed, irrespective of the intention of the parties, a serious question would arise as to the validity of the claims.

As already noted, both the District Court in the instant case and the Court of Appeals in Anderson quoted extensively from the Patent Office proceedings which resulting in the issuance of four patents (including the one in suit) upon the same specifications. The point is made that evidence of the proceedings relative to the patents not in suit was immaterial to the issues in the instant case. Plaintiff indicates that there are no authorities in support of the procedure other than the action of the District Court in the instant case and that of the Court of Appeals in Anderson. We think this is rather respectable authority, particularly in view of the fact that we know of no case which has held to the contrary. As previously shown, the courts in both cases recited at length from the Patent Office proceedings, which demonstrates the numerous times and occasions on which Behimer took a position diametrically opposed to his position in the instant case. While Behimer's statements, attitude and representations made in the course of the proceedings relative to the other patents are deadly to his contention here, we think we need not resolve any doubt which may exist as to the propriety of its use. This is so for the reason that the claims in suit read in the light of the specifications plainly fail to describe the invention now asserted. Thus, our conclusion would be the same, irrespective of whether the testimony relative to the procedure as to the other patents was admitted or excluded. In this connection, plaintiff also complains that while the District Court considered testimony relative to the proceedings in connection with the other patents, it failed to give any consideration to the history of the proceedings in connection with the patent in suit. Plaintiff fails to point out, however, anything in that history which is of any benefit to its cause, and our reading of the history discloses nothing.

The portion of the decree on the issue of infringement, appealed from by plaintiff in No. 10954, is affirmed.

Such being our conclusion, we think there is no occasion to enter a detailed discussion of other defenses relied upon by Globe. Moreover, such defenses are dependent upon factual issues and the findings of the Master as approved by the District Court must be accepted unless clearly erroneous. In addition, the patent in suit has expired. This long and protracted trial was largely a contest between expert witnesses and, notwithstanding Globe's argument to the contrary, we think it is an appropriate situation for the application of Rule 52(a) of the Federal Rules of Civil Procedure, 28 U.S.C.A. The Supreme Court, in Graver Tank & Mfg. Co., Inc., v. Linde Air Products Co., 336 U.S. 271, 274, 69 S. Ct. 535, 537, 93 L. Ed. 672, referring to Rule 52(a) stated:

'To no type of case is this last clause more appropriately applicable than to the one before us, where the evidence is largely the testimony of experts as to which a trial court may be enlightened by scientific demonstrations.'

And as the Supreme Court stated when the same case was before it a second time, 339 U.S. 605, 609, 70 S. Ct. 854, 857, 94 L. Ed. 1097:

'Like any other issue of fact, final determination requires a balancing of credibility, persuasiveness and weight of evidence. It is to be decided by the trial court and that court's decision, under general principles of appellate review, should not be disturbed unless clearly erroneous.'

See also the discussion of this court in Union Carbide & Carbon Corp. v. Graver Tank & Mfg. Co., Inc., 196 F.2d 103, 107, as to the application of the rule which requires a reviewing court to accept the findings of a trial court unless clearly erroneous.

Judge Campbell, commencing 112 F.Supp. on page 463, discusses the issue of laches or equitable estoppel and approves the findings of the Master, as well as his conclusion that this defense was not sustained. Judge Campbell, commencing on page 482, discusses at length the issue of validity and makes an extensive review of the prior art relied upon by the defendant. On this issue he approved the findings and conclusions of the Master. We have studied the situation sufficiently to be convinced that such findings are not clearly erroneous and must be accepted by this court. We make one exception which does not affect the result. Judge Campbell held that the claims were valid irrespective of whether they were construed in accordance with plaintiff's theory or defendant's theory. On this issue he stated, at page 492:

'Summing up on this issue, I agree with the master that whether the claims in suit be construed as limited to heating in the coil and cracking in the drum (as the master held they should be construed and which holding I have adopted) or whether the claims should be construed broadly (as plaintiff contends) to include cracking either in the coil or in any locus-- that, in either case, the claims would not be anticipated by the prior art and would be valid.'

As heretofore stated, we think there is a serious question as to the validity of the claims, if they be construed in accordance with plaintiff's suggestion. Judge Phillips in Anderson, 122 F.2d at page 842, stated:

'While we do not think the claims in suit can properly be construed as embracing clean circulation with a hot oil pump as an auxiliary to a cracking process, were they so construed, they would be void for want of invention.'

We think we need not either agree or disagree with the conclusion of the District Court that the claims are valid, even though they be construed as plaintiff would have us do, because that construction has been rejected. It is sufficient that we agree with the conclusion that they are valid as construed by the District Court and this court.

The cross-appeal in No. 10955 may be disposed of in short order. Globe concedes that its request for an allowance of attorney's fees to be taxed as general costs called for the exercise of judicial discretion by the District Court and expressly refrains from charging an abuse thereof. We discern no reason to disturb the action of the District Court in this respect. Therefore, its refusal to allow attorney's fees to Globe is affirmed.

On Petition for Clarification as to Costs.

The Texas Company petitions for an apportionment of costs on these appeals, particularly as it relates to the printing of the record. As shown in the opinion, the litigation was protracted and the printed record voluminous. The major portion of this record, asserted by Texas to have been unnecessary, was printed at the direction of Globe. Texas suggests that more than half of the pages of the printed record were not referred to by Globe in its briefs filed in this court and particularly complains of the inclusion of thirty-one prior art patents, only six of which were relied upon by Globe. Texas also asserts that a large portion of the record printed at the behest of Globe related to the testimony of witnesses on factual issues and the testimony of experts.

Texas contends that inasmuch as Globe prevailed in appeal No. 10954 and Texas in appeal No. 10955, and as both appeals were heard upon a common record, there should be an apportionment of the printing costs. We think this contention is without merit. As the opinion discloses, Globe in No. 10955 appealed only from that portion of the decree which refused to allow it attorney's fees as part of the general costs. The issue thus raised was of minor importance, in fact it was infinitesimal when considered in connection with the issues relevant to the appeal in No. 10954. This is evidenced by the fact that out of some four hundred pages of printed briefs, one page was devoted by Globe to the issue which it raised in appeal No. 10955, and seven lines by Texas to this issue in its reply brief.

The second contention advanced by Texas is that Globe was responsible for much unnecessary printing and should bear the cost thereof. When the matter is viewed in retrospect, there appears to be merit in this contention. Globe, however, at the time it filed its designation for printing of the record could not anticipate that it would obtain in this court, as it had in the District Court, a favorable decision on the issue of infringement. In this court it had a right to rely upon every defense, which it had pleaded in bar, in support of the judgment of dismissal, even though such defenses other than that of non-infringement had been rejected by the District Court. Such being the situation, we think it was within its right in designating the printing of a record upon which such defenses could be adequately presented. The fact that subsequent developments have shown that such printing could have been avoided is of little if any consequence.

We think that Globe must be regarded as the 'prevailing party,' within the meaning of Rule 54(d) of the Federal Rules of Civil Procedure, as well as Rule 24(a) of this court, and we are not convinced that the general rule as to the taxation of costs in such a situation should be deviated from. The petition of Texas is denied. The costs of these appeals will be taxed to it.