Thomas E. Bryan, D/b/a Bryan Gas Lift Equipment Company, Appellant & Appellee, v. Garrett Oil Tools, Inc., Appellee & Appellant,garrett Oil Tools, Inc., Appellee & Appellant, v. Thomas E. Bryan, D/b/a Bryan Gas Lift Equipment Company, Appellant & Appellee, 245 F.2d 365 (5th Cir. 1957)

These are cross-appeals from a judgment of the district court which had upheld the validity of one patent owned by plaintiff and one patent owned by defendant and had found infringement as to each, while holding that another of plaintiff's patents was invalid and not infringed by defendant; each party appeals from all parts of the judgment adverse to it.

All the patents involved relate to the design and operation of certain valves used in the extraction of oil from wells by the so-called "gas-lift" method. In order to cause oil to flow from a well which by itself does not generate sufficient pressure to push the fluid to the surface it is possible to force gas into the oil under pressure, which results in a mixture of lower density light enough to rise to the well head. Alternately, a slug of gas may be introduced under a mass of oil, which, upon expanding, will push the oil above it to the surface. Customarily the gas is introduced by lowering a string of tubing into the casing to form thus two concentric passages; the outer passage is sealed off at the bottom, permitting oil to rise within the tubing while gas may be pumped into the casing around the tubing, and, through holes in the tubing, into the oil. The problem is to regulate the flow of gas through these holes into the oil in such a manner that the optimum gas-oil ratio is obtained for a given flow — which may be determined either by the capacity of the well or by extraneous limits on the desired rate of extraction. For this purpose a great variety of valves and ancillary control systems had been devised. The valves involved in the present suit are, unlike those of earlier designs that had to be manipulated from the surface, automatic, in that their opening and closing is determined by certain combinations of pressures and velocities of flow in the gas and in the oil, and thus it is possible to control them indirectly by proper adjustment of the gas pumped into the well.

In challenging or supporting the six ultimate determinations of the trial court on the validity and infringement of three patents the parties appear almost to run the gamut of the legal issues that are peculiar to patent litigation. However, the points raised as to each are largely separable and permit an individual examination of each such principal conclusion.

I. Validity of Bryan Patent No. 2,145,918.

Claim No. 9 of the Bryan Patent No. 2,145,918 (hereinafter Bryan '918), the only one of the claims of this patent here at issue, concerns the placing of an automatic, intermittent valve into the tubing in such a manner that it can from time to time be lifted out for inspection, repair, or replacement. It reads as follows:

"9. In combination with a conduit for removing liquid from a well, said conduit having a lateral opening therein in communication with a source of gas under pressure, a valve adapted to be lowered into said conduit for controlling flow through said opening and having a part adapted to removably seat within said conduit and to form a sealed connection with the inner end of said opening when so seated."

In the accompanying disclosures and illustrations this removable valve is always described as being enclosed in a slightly tapered housing which fits into a correspondingly tapered collar fixed in the tubing in such a way that the hole through the tubing and collar corresponds with an appropriate hole and groove in the housing and in the valve through which the gas may pass from outside the tubing into the valve, and, if that is open, into the oil standing in the tubing. At the top of the housing there is an appropriately shaped "spear" that can be engaged by a "fishing tool" from the surface for the purpose of pulling the valve and housing up through the otherwise unobstructed tubing. The valve is so arranged that some oil can flow up through it from the lower part of the well which penetrates to the producing formation.

At the trial much evidence was introduced by both parties bearing on the question whether or not any valves with the tapered housing had ever been successfully used. Plaintiff Bryan testified confidently that a number had been installed, though he was vague as to times and places, and those of his witnesses who definitely testified to the success of the valves were not sure whether they were the ones with the tapered housing or those of a later design. Witnesses for the defendant testified that to their knowledge none of the tapered valves had ever been used, that different type valves had been used on the occasions as to which Bryan had testified that the tapered ones had been used successfully, and that it could be demonstrated from theoretical considerations that the valves were inoperative.

The trial court found that plaintiff Bryan was the inventor of this design and that it had not been anticipated either by prior patents or by prior use or publication. However, he also found as follows:

"9

"One of the essential elements of valve described in claim 9 of Patent No. 2,145,918 was a part adaptable to removable seat and to form a sealed connection when so seated. * * *.

"11

"The valve structure covered by claim 9 of Bryan Patent No. 2,145,918 is inoperative."

and therefore concluded:

"3

"Bryan Patent No. 2,145,918 is invalid as to claim 9 thereof because the valve structure covered by the claim is inoperative."

Garrett insists that the patent is limited by the disclosure and illustrations to a tapered design, and by the terms of the claim itself to a design which simultaneously seats and seals, for it is asserted that the only patentable novelty lay in the latter feature. It is further contended that the tapered design is inoperative because if the taper is designed so as to permit removal of the valve by a pull from the top it will not be able to hold the valve seated against the upward pressure of the oil, only a restricted amount of which can flow freely from beneath — and this inherent defect has allegedly been demonstrated in practice. But even if not restricted to the tapered design, Garrett alleges that no operable alternative involving simultaneous seating and sealing has yet been devised.

On the subsidiary issue of whether any tapered valves had ever been used successfully we must accept the implied finding of the trial court that no such past use has been demonstrated, since this is merely a matter of the credibility of the witnesses and of the weighing of conflicting evidence — and on the record here such a finding by the trial court is not clearly erroneous.

We do not, however, agree with the implied legal conclusion of the district court that claim 9 of the Bryan '918 patent was limited to tapered housings — since the claim itself does not so state and we see no reason for limiting the coverage to the design of the preferred embodiment. Nor do we think that the claim covers only devices in which the seating and sealing functions are performed simultaneously, i. e., by the same mechanical contrivance, as is of course true of the tapered design. A reasonable interpretation of "a part adapted to removable seat * * * and to form a sealed connection" is that the valve housing ("a part") have attachments that will seat or position the valve and also have some that will seal it in the conduit. Thus interpreted Bryan's later valve designs, using latches and sealing cups, as well as Garrett's valves discussed below, show an operable embodiment of the patent which is thus valid.

II. Infringement of Bryan Patent No. 2,145,918.

The valves of the defendant Garrett Oil Tools, Inc., denominated "W-OBP" valves, and illustrated in its trade catalogue, are used to control the entry of gas from the outside to the inside of tubing placed into a well for oil lift purposes, and are designed so that they may be retrieved from the well without removing the entire tubing. To this end their housing is of basically cylindrical design with a snap or spring ring near the lower end which fits into a groove in the collar in the tubing and permits vertical positioning so that the hole through the tubing will line up with an entry port into the valve; several O-rings are placed along the perimeter of the housing to prevent the flow of oil past it when the valve is seated in the socket, i. e., to seal it. The trial court followed Finding of Fact No. 9 supra, with the following determination:

"10

"The `W-OBP' valves of Garrett Oil Tools, Inc., as shown in its 1954 catalog, which are the valves Plaintiff alleges infringe claim 9 of Patent No. 2,145,918, do not have a part to removably seat and form a sealed connection when so seated as required of the valve described in claim 9 of Patent No. 2,145,918. In other words, the `W-OBP' valves do not have a sealing seat."

and concluded:

"4

"Claim 9 of Bryan Patent No. 2,145,918 is not infringed by any of the valves manufactured and sold by the Defendant, Garrett Oil Tools, Inc."

However, as we indicated above in sustaining the validity of the patent, we do not believe that claim 9 is limited to a "sealing seat" as long as some provision is made for both sealing and seating the removable valve within the tubing. In the Garrett valves the sealing is accomplished by several sets of O-rings set in grooves in the valve housing, while the seating is accomplished by means of a snap ring which helps to position the valve and its housing vertically. In another patent case "seat" has been defined as follows:

"The word `seat' when applied to machinery means `the part on which another thing rests, as a valve-seat.' 3 Knight, Mech. Dict. 2084." Safety Oiler Co. v. Scovill Mfg. Co., C.C.S. D.N.Y., 110 F. 203, 204.

which also closely approximates a Webster's Dictionary definition. Though the surface of support may be as narrow as the protruding edge of a snap ring resting on the side of a groove, the valve is seated thereon — in contradistinction to other possible means of support, such as suspension from above.

Garrett, however, contends that unless the claim is severely limited it will be invalid against the prior art, or alternatively, that plaintiff is prevented by file wrapper estoppel from claiming broadly enough to include the accused valves. Though a number of prior patents are relied on, most of these¹  admittedly do not involve automatic valves at all but only such as are manipulated from the surface; Garrett cites particularly Otis Patent No. 1,916,070, which however, provides merely for a retrievable choke which may with the breakage of some minor parts be pulled to the surface for an adjustment of the size of the fluid inlets. Though Garrett claims that the accused valves resemble the Otis design more than they do the Bryan one — which, coming later, can of course not extend into the former's area — in that neither of the former has the "sealing seat" the Bryan patent allegedly calls for, it is clear that claim 9 of the Bryan '918 patent covers precisely the sort of retrievable valve that the Garrett device is. Moreover Bryan is not estopped from claiming broadly enough to include the Garrett valves since in pursuing his application in the Patent Office he did not change the wording of his claim or otherwise limit it, but merely explained to the satisfaction of the examiner wherein the Bryan invention differed from the Otis one — and the distinction is based on the inclusion of the same features that the accused tools also have.²  Cf. Hunt Tool Co. v. Lawrence, 5 Cir., 242 F.2d 347, 353.

We therefore conclude that the Garrett W-OBP valves infringe claim 9 of Bryan Patent No. 2,145,918.

III. Validity of Bryan Patent No. 2,275,345.

The difficulty with the several valve designs discussed above is that when seated in place the valve and its housing seriously constrict the available cross-sectional area of the tubing through which the oil has to flow up. As a result considerable pressure is exerted against the valve and it was necessary to design locking devices that would resist the upward pressure but would still permit the valve to be removed by a pull from above. The design patented in Bryan Patent No. 2,275,345 (Bryan '345) avoids this difficulty by providing in an enlarged section of the tubing several passages paralleling the seated valve, through which there can be a substantially unchecked flow of oil which will then not exert an undue pressure on the valve.

The only issue concerning the validity of this patent is whether the idea of the by-pass passages actually originated with Bryan, who secured the patent in his own name as sole inventor, or with a Mr. R. R. Kyner, a former employee of the Gas Lift Sales Corporation, a licensee of a number of Bryan's patents.

Kyner testified positively that it was he who invented the by-pass design, without any help from Bryan. Garrett also introduced into evidence a contract between Bryan and the Gas Lift Sales Corporation, dated December 28, 1937, in which it was recited that Kyner "claimed" to have discovered the "Kyner By-Pass Drop Valve" and which provided for the assignment to Bryan by the corporation of all patent rights to this development, which had previously been assigned to it by Kyner; to this contract several drawings illustrating the by-pass system were attached, some dated as early as October 30, 1937. It was also shown that three days after the date of the contract Bryan filed a patent application on some improvements of the valve fittings, without, however, claiming the by-pass design, and that in February 1938 Kyner published an article in a trade journal in which, inter alia, he referred to the by-pass design. Soon after receiving the assignment from the Gas Lift Sales Corporation Bryan employed a new patent attorney with no connections with the licensee corporation, through whom he then prosecuted both his own several applications, and later those that had been filed by Kyner; still later, after having made an abortive attempt to have the Kyner claims transferred to one of his own applications, he simply dropped the Kyner applications and filed the pertinent claims as part of the application which became the Bryan '345 patent. Garrett asserts that at no time before Bryan thus incorporated these claims into his own application did he dispute Kyner's status as inventor, and contends that this circumstance is fatal to Bryan's later assumption of that role.

Bryan testified that the by-pass valves had been manufactured in his shop as early as the spring of 1937 and he introduced a witness who remembered working on such a valve as early as April 1937.

From an examination of the file wrappers of the two Kyner applications, Nos. 207,514 and 207,515, and of the Bryan application No. 191,767, which later was granted as the '345 patent, the following somewhat inconclusive chronology appears.

The Bryan application was filed on February 21, 1938. None of the requested claims related to the by-pass passages, though the illustrations showed them and the disclosures apparently mentioned them. As a result the examiner, in his first comment on the application on April 27th, inquired as to the purpose of these passages.

The two Kyner applications were filed on May 12, 1938, and the by-pass system was disclosed and claimed in them, though not with sufficient clarity to be immediately satisfactory to the examiner.

On October 25, 1938, Bryan amended his application to add claims specifically covering the by-pass passages, to which claims the examiner raised only minor objections and indicated that they would be allowed if properly revised.

The assignment to Bryan of the rights to the Kyner application, required by the contract of December 28, 1937, was actually accomplished on January 29, 1939, and the Patent Office was notified that these applications would henceforth be prosecuted by Bryan's attorney. On March 17, 1939, Kyner's February 1938 article was added as a reference of record in both these applications and the examiner thereupon required that Kyner file an oath affirming that his invention preceded the date of the publication.³  This Kyner refused to do, and on July 1, 1940, Bryan's attorney, after having unsuccessfully urged the examiner to waive the oath requirement in view of the inventor's obduracy, requested that the claims of both the Kyner applications relating to the by-pass passages be transferred to still another application that Bryan had filed in his own name on February 9, 1939 (No. 255,395, issued as Patent No. 2,275,346 on March 3, 1942). This the examiner refused to do and the Kyner applications were thereupon abandoned.

On January 10, 1941, Bryan filed five more claims in his application No. 191,767, further revising the coverage of the by-pass invention, stating that these were substitutes for the earlier added claims in the same application as to the amendment of which a deadlock had apparently been reached with the Patent Office. (It is these claims that Garrett alleges were merely lifted from the abandoned Kyner applications, but their wording is certainly not identical.) These five claims became claims Nos. 4-8 of Bryan '345 issued on March 3, 1942, of which claims Nos. 4-7 are here at issue as having allegedly been infringed by Garrett.

The district court found that the devices described in the '345 patent were operative and that the invention had not been anticipated either by prior patents or by prior use or publication and that:

"20

"Although there is some evidence and inferences drawn therefrom that indicate that Plaintiff was not the inventor of the apparatus described in claims 4, 5, 6 and 7 of the Bryan Patent No. 2,275,345, I cannot conclude after a consideration of all the evidence that a preponderance of the evidence shows that Plaintiff was not the inventor of said apparatus. Therefore, I find that the evidence fails to establish that the Plaintiff was not the inventor of the apparatus described in and covered by Patent No. 2,275,345."

It concluded that the claims 4-7 of the Bryan '345 patent are valid.

This finding of fact, depending as it does on an evaluation of testimony of witnesses who gave somewhat contradictory testimony, and on a weighing of the inferences that can be drawn from the various transactions of the parties and from the tortuous paths of the applications through the Patent Office, cannot be disturbed by us unless clearly erroneous. Rule 52(a), Fed.Rules Civ.Proc. 28 U.S.C.A. Since from the above we cannot find it to be that, the conclusion of validity must be affirmed.

IV. Infringement of Bryan Patent No. 2,275,345.

Garrett also insists that the claims of the Bryan '345 patent should be narrowly construed, in view of the existence of prior patents in this field. Reference is made particularly to Boyton Patent No. 1,985,973 (Boyton '973) which provides for by-pass passages around a fixed gas lift valve. The idea of a retrievable valve had also already been disclosed in Langherst Patent No. 1,696,492, Watts Patent No. 1,735,025, and Otis Patent No. 1,916,070.

However, Bryan argues, and the trial court and we agree that the novelty of the Bryan '345 invention lies in the use of by-pass passages to reduce the upward oil pressure on a removably seated valve so that it would not have to be fastened too securely for easy removal by a pull from the surface. But even if we accept a rather narrow construction of the claims there would still be infringement on the basis of the trial court's substantiated findings that:

"22

"The corporate Defendant's `W-OBP' valves * * * fully meet the literal requirements of claims 4, 5, 6 and 7 of the Bryan Patent No. 2,275,345, and said valves function in substantially the same manner and accomplish substantially the same objective as the device and apparatus described in claims 4, 5, 6 and 7 of said Patent No. 2,275,345, and, therefore, said corporate Defendant's `W-OBP' valves infringe claims 4, 5, 6 and 7 of the Bryan Patent No. 2,275,345."

Garrett also asserts that the Bryan '345 claims referred to passages in the plural in order to avoid the single passage provided for by Boyton, and since the Garrett valves have only a single passage they fall in the area of the prior art which cannot be claimed by Bryan. This argument has no merit at all since the mere substitution of a plurality of passages for a single one would clearly not have made his design patentable over the Boyton '973 patent if it did not otherwise contain patentable invention. Cf. Dunbar v. Meyers, 94 U.S. 187, 24 L. Ed. 34; Blackmore v. Ford Motor Co., 6 Cir., 56 F.2d 806; Nestle-Le Mur Co. v. Eugene, 6 Cir., 55 F.2d 854, 858. The Bryan design was patentable because of its novel solution of a particular problem and Garrett's infringement lies in appropriating that answer to solve the very same problem.

V. Validity of King Patent No. 2,339,487.

The King Patent No. 2,339,487 (King '487) is owned by the Garrett Oil Tools, Inc., and its validity and alleged infringement by Bryan were also litigated below as a counterclaim to the original suit. It covers numerous aspects of a system for elevating oil by the gas-lift method operating with a high degree of automaticity. Bryan challenges the validity of each of the eight claims⁴  found valid by the trial court, on the ground that each is substantially covered by the prior art, particularly as disclosed in certain patents preceding the King '487 patent. The trial court, however, found that:

"Prior to the King valve and system, as disclosed and claimed to King Patent No. 2,339,487, the operation of gas lift valves then known to the trade was by manual control from the surface by velocity of gaseous fluids passing through a valve and by velocity flow through a valve loaded by a weighted disk or a spring loaded valve.

"24

"King, as disclosed in Patent No. 2,339,487, devised the first successful gas pressure loaded valve as distinguished from a spring loaded valve. The King valve employed a predetermined charge of gas pressure behind a bellows which held the valve member seated until the charge pressure was exceeded and overcome by well pressure, at which time the well opened and the bellows chamber was sealed and the bellows protected from rupturing by such excessive well pressures. This valve also provided for the first time a construction which would positively operate to open and close at selected predetermined pressures.

"25

"The King system or apparatus, as disclosed and claimed in Patent No. 2,339,487 provided a tubing string with a series of valves mounted thereon, each succeeding valve from top to bottom loaded with a pressure somewhat lower than the valve immediately above, affording an apparatus by which an oil well could be positively unloaded of liquid by sequential operation of said valves with gas supplied from the surface in controlled volumes and at predetermined pressures and the well produced thereafter from any selected valves of the series. The King valve and system enjoyed considerable commercial success."

and that the patent had not been anticipated by prior patents, use, or publication.

An examination of the essential elements of each of the challenged claims as against the coverage of the prior patents cited requires the following comparisons:

Bryan claims that his patent No. 2,008,172 (Bryan '172) exactly anticipates claim 5 (as well as the similar claims 6 and 7) of the King '487 patent. But all that the cited Bryan patent has to duce a predetermined amount of pressay on the subject of "means to introsure fluid into the well as a charge to open such valve or valves as will open by the pressure occurring in the well due to the introduction of such charge" (see claim 5 of King '487) is the disclosure that:

"In case the well goes dead or loses its head of gas, compressed air or gas from an outside source may be admitted by opening the valve 33. By this means a gas pressure may be built up in the casing 10 and the well started to flowing * * *. The valves act as boosters to admit gas to the tubing whenever necessary to lift the load. Their action is entirely automatic and they function when needed to maintain the flow."

This does not set forth by a careful control of the amount of introduced gas it is possible to determine which of a series of spaced pressure sensitive valves should open, and that by appropriate direct or indirect control of the gas pressure it is possible to keep the well operating at a particular valve or at progressively lower levels, in order to achieve the highest efficiency and optimum gas-oil ratios with given speeds of production and extraction.

It is claimed that valves for the intermittent introduction of gas were already disclosed by Bryan Patent No. 1,852,692 (Bryan '692) and by Burke Patent No. 1,803,837 (Burke '837). The Bryan patent, however, merely discloses a mechanical method for having a valve opened and closed at predetermined time intervals, without any relation to a device for achieving a predetermined pressure in a casing to actuate a particular valve therein. The Burke patent uses the time device in order to avoid the introduction of too much pressure into the well which in the there described system would have the effect of driving the oil back into its formation, and in order to permit the most efficient exploitation, in rotation, of a series of wells; the purpose is not to activate a particular unloading valve at a particular level.

One of the particular advantages of the King system is, as the trial court found, that it permitted the positive unloading of a well down to a given level and the continued operation at that level by sequential operation of the valves from the top to the bottom by means of the introduction of gas in controlled volumes and at predetermined pressures. By proper manipulation of the amount of gas admitted, and thus of the pressure in the casing which in turn determined the level at which the well would operate, the most favorable gas-oil ratios could be obtained. The trial court could and did find that the King '487 patent (particularly claims 1, 2, 5, 6, and 7) effectively claimed this invention and that this particular combination and application of separate devices (some new, some old) was a patentable advance over the prior art.

Though the Burke '837 patent describes a bellows type valve with gas at a particular pressure sealed in there is no suggestion of a protective device to shield the bellows from excessive pressure. As the trial court found, this is a novel feature claimed in the King patent, and therefore the validity of claims 3 and 13 must be sustained.⁵  Though in the Bryan '918 patent the bellows chamber is also sealed off under particular conditions of pressure, it is clear from a communication found in the file wrapper addressed to the examiner that the purpose is not to protect the bellows but to prevent their compression by fluid pressure at a time when the operation of the valve requires them to be extended. The same may be said of the claimed priority of Bryan Patent No. 2,179,226 (Bryan '226).

Though both the Bryan '918 and '226 patents show sealed bellows that Bryan asserts must be filled with air at approximately atmospheric pressure, this is not the same as filling a bellows with gas "under" pressure or a series of bellows at "different pressures" as called for by King claims 1, 2, 11, and 13.

The McEvoy Patent No. 906,733 (McEvoy '733) is cited as anticipating the reverse pressure loading recited in one form or another in claims 5, 6, 7, 11, and 13 of the King patent (rather than in claims 1, 2, 7, 11, and 13 as asserted in Bryan's brief). The examiner raised the same point and originally rejected several of the claims as met in or reading in terms of the McEvoy patent. However he, and later the district court were persuaded that there was patentable novelty in the use of the reverse pressure system with gas instead of spring loaded valves, in a deep well gas-lift system rather than in a mere shallow well well-blowing arrangement, and with a method of supplying pressure gas in controlled amounts at predetermined pressures.

Due to the great number of older patents cited and the number of comparisons between their features and those of the several claims of the challenged patent that were advanced or refuted somewhat unsystematically in the briefs, it is of course very difficult to be absolutely certain that each of the eight claims actually presents a patentable advance over the prior art. We note from the file wrapper that the patent examiner was apparently aware of most of these arguments and considered most of the above cited prior patents, and his conclusion that a patent should nevertheless be granted with the above claims is entitled to a strong presumption of validity. As is required we have given great weight to the findings of the district court, here in favor of validity, and we do not find them to be clearly erroneous.

VI. Infringement of King Patent No. 2,339,487.

The trial court made the following findings:

"26

"Plaintiff for some time prior to the institution of this suit was and still is engaged in the manufacture, selling and installing of certain pressure operated valves for use in the production of oil wells denominated as his BPV-1, BPV-2 and BPV-3 valves. Plaintiff has on more than one occasion installed or caused to be installed his said valves in oil wells in a manner which he refers to as `reverse pressure installations' and has issued or caused to be issued certain instructions and recommendations relative to the `reverse pressure installation' of said valves.

"27

"The plaintiff in the manufacture of his BPV-1, BPV-2 and BPV-3 valves employs gas pressure loading and the bellows protection features of the King Patent No. 2,339,487.

"28

"Plaintiff in the manufacture, sale and installation of his BPV-1, BPV-2 and BPV-3 valves and in his `reverse pressure installations' of said valves and the instructions and recommendations he made in connection with `reverse pressure installations' has infringed claims 1, 2, 3, 5, 6, 7, 11 and 13 of King Patent No. 2,339,487."

In effect two questions are presented: (1) is the construction of the accused Bryan valves such as is described in the several claims of the King patent found to have been infringed; (2) has Bryan installed these valves, or recommended the installation, in such a manner as to infringe the gas-lift system described in several of the King claims? Again little aid in systematically matching the coverage of each of the claims with the pertinent features of the accused devices is received from the lengthy briefs, and in our review we will therefore merely follow the patent of the trial court's findings.

As to the design of the valves themselves, the district court found the patentable features to consist of: (a) gas pressure loading, which, according to claims 3, 11, and 13 of the King '487 patent, is accomplished by sealing gas at a predetermined pressure inside a bellows and permitting the pressure fluid to act on its outside; (b) the bellows protection feature which, according to claim 3 consists of a "resilient means" which seals off the bellows as soon as the pressure on it becomes high enough to open the valve.

Apparently the trial court considered it of no import in finding infringement that in the Bryan valves it is the chamber around the bellows that contains the sealed gas under pressure, nor that the pressure fluid in the Bryan valve is not permitted to impinge on the bellows directly but exerts its force on it through mechanical intermediaries. We agree that these simple substitutions are of no import in avoiding infringement. Cf. Hunt Tool Company v. Lawrence, supra, 242 F.2d at page 352. In the Bryan valve as well as in the King patent the pressure fluid first activates the valve and then enters the well tubing through it, the opening or closing of the valve are determined by the balancing of the pressures fluid against the pressure of the gas sealed in the valve, and to protect the delicate bellows no force is exerted against it once the pressure has increased beyond its operative range.

As to the gas-lift system described in the King patent, the court found the salient feature to be the "reverse pressure loading." Of course the coverage of the King patent is restricted somewhat because the prior McEvoy '733 patent prevents us from considering King as a pioneer inventor. However, as pointed out above, the Patent Office and the district court found sufficient patentable novelty in the use of a reverse pressure installation with automatic gas pressure loaded valves and a surface control for the admission of pressure gas to determine which valve is to be operated at a given time, and it appears that the accused Bryan installations and instructions do possess these particular features.

The principal attack made by Bryan on this finding of infringement is that there is no showing that he used any device at the surface to introduce a predetermined volume or amount of gas into the well, since he merely used a pressure controller with an intermittent timing device, both of which were well known prior to the King invention. In particular it is claimed that King is precluded by file wrapper estoppel from claiming any coverage beyond a surface control device that controls the "amount" or "volume" of the entering pressure gas. Without reciting in extenso the considerable correspondence between King's attorney and the Patent Office regarding this feature of the patent, it appears to us from a consideration of this material, as well as from the nature of the patented system, that the critical point is not that the control device measure either the amount or volume of the gas flowing directly, but that particular volumes or amounts of gas be introduced in order to maintain a particular pressure in the casing so that only a particular valve or valves be activated by that pressure. The pressure is obviously a function of the amount of gas in the tubing and of the volume it is permitted to occupy, and any change in pressure must result from the flow of gas through one of the valves into the oil or from the introduction of further gas from the surface. Just as the loss of pressure is the result of gas flowing at a determined rate through a selected valve and is principally a function of the operating pressure (directly by determining the rate of flow and indirectly by selecting the valve) and of time, so the correct amount of gas may be introduced from the surface by permitting flow at a controlled pressure for a given amount of time. That, in substance, is what the Bryan surface control devices accomplish, and they enable the operator to exploit the well at a determined rate at a particular level and at high efficiency by the use of gas loaded valves arranged in a reverse pressure sequence. This is the area protected by the King patent.

Finally Bryan explains that his system and valves may also be used in direct pressure installations which would not infringe the King patent. However, his brief admits that his instructions show both direct and reverse pressure loading, and insofar as they do the latter they infringe. Bryan further complains that the district court's order precludes him from using his valves even in a direct pressure system — but these valves have been found to infringe the King patent per se, particularly claims 1-3, which (contrary to Bryan's contention in his brief), do not specify reverse pressure loading but merely that the "unloader" valves be set to open at a higher pressure than the intermitter valve below them.

The judgment of the district court is affirmed as to its determination of the validity and infringement of the Bryan Patent No. 2,275,345 and of the King Patent No. 2,339,487, and is reversed and remanded for entry of an appropriate order granting relief as to the validity and infringement of the Bryan Patent No. 2,145,918.