Envirotech Corporation (Envirotech) appeals from a judgment, after a jury trial, of the United States District Court for the Western District of Louisiana holding that claims 1, 14 and 15 of its U.S. Patent 4,110,210 ('210) ("Dispersed Gas Flotation Process") and claims 1, 10, 11 and 12 of its U.S. Patent 4,226,706 ('706) ("Dispersed Air Flotation Machine") are invalid and not infringed by appellees Al George, Incorporated and Monosep, Inc. (Monosep). We affirm in part, and vacate and remand in part.

* Background

The Envirotech patents involve separation of materials having different densities by ejecting minute gas bubbles into a tank containing a two-component fluid or solid/fluid "slurry". The bubbles attach to the particles to be separated and provide sufficient buoyancy so that both particle and bubble float to the tank surface to form a froth that is skimmed from the surface--thus, the term "flotation separation". Typical flotation separation devices use a nozzle extending beneath the tank's surface in order to introduce bubbles into the tank by ejecting a mixture of air and water as a two-phase (gas and liquid) effluent.

In order to have successful flotation separation, two principal conditions must exist. First, the surface of the liquid in the tank must be relatively smooth because turbulence will dislodge the particles from their bubbles causing them to sink back into the fluid. Second, the bubbles must be dispersed throughout the tank in order to come into contact with all of the particles to be separated. In the past, dispersion was achieved by using various types of impellers placed at the bottom of the tank, using multiple nozzles in each tank, or using baffles.

The '210 patent discloses a flotation separation method in which a two-phase effluent is ejected from a single nozzle into a tank in such a fashion that the desired conditions of good bubble distribution and a smooth non-turbulent surface may be obtained throughout a range of tank sizes without the use of baffles, impellers or multiple nozzles. The invention recognized that one could obtain these optimal conditions with a certain effluent density (gas-liquid ratio) and a certain energy rate per tank volume (velocity of the effluent). Figure 2 of the '210 patent graphically illustrates this energy/density relationship for a wide range of tank sizes:

NOTE: OPINION CONTAINS TABLE OR OTHER DATA THAT IS NOT VIEWABLE

The curves set forth in Fig. 2, supra, divide the graph into Regions I, II and III. If a nozzle and tank combination are designed to operate within Region I then the two conditions for optimal separation can be obtained. Operation within Region III lacks these two conditions, and Region II is a gray area. Claim 1 of the '210 patent is representative:1. A dispersed gas flotation process wherein hydraulic effects are used to disperse gas bubbles throughout a contained liquid body with a free surface, said process comprising pumping a two-phase fluid into the liquid body through an ejection device with the density and the kinetic energy rate of the ejected fluid per unit volume of the contained body at the point of ejection being defined by a point on the graph of FIG. 2 within the area encompassed by Region 1.

According to the teachings of the patent, to calculate effluent density the designer need know only the flow rate of gas through the nozzle (QG ) and the flow rate of the liquid through the nozzle (QL ). To calculate the effluent energy rate per tank volume, the designer must know the tank volume (V), the flow rate of the gas through the nozzle (QG ), the flow rate of the liquid through the nozzle (QL ), and the area of the effluent being ejected into the tank at the "point of ejection" (Ae ).

The '706 patent discloses a flotation separator having a series of tanks arranged in line so that the contaminated fluid can be treated in stages from one tank to another. As shown in Figure 2 of the '706 patent, reproduced below, each flotation tank 13 has one nozzle 20 (extending below the surface of the liquid) which ejects the two-phase effluent. The froth, which is subsequently formed by the bubble-carrying particles, is then skimmed off by paddle wheels 14, 14a.

NOTE: OPINION CONTAINS TABLE OR OTHER DATA THAT IS NOT VIEWABLE

The claims of the '706 patent require that each nozzle have a "hollow tubular expansion chamber member" through which the two-phase effluent is discharged. The claimed pipe-within-a-pipe configuration comprises a smaller inner pipe which carries the liquid into a larger outer pipe creating a vacuum which sucks air down the outer pipe into the expansion chamber where it is mixed with the liquid from the inner pipe. The '706 patent expressly incorporates in the specification the types of "converging-diverging" nozzles included in the '210 patent. (Column 4, lines 31-33.)

D. The Defendants' Alleged Infringing Device

The defendants' accused device, marketed under the name "Multisep", has a series of tanks, each having a single off-the-shelf nozzle (also called an "eductor") where liquid is pumped and air is drawn in. The resulting two-phase effluent (air and liquid) is then ejected into the tank filling it with small bubbles. Defendants' eductor nozzle, illustrated below, normally used to pump two liquids, is of the converging-diverging type with a cylindrical throat portion and a flared end portion.

NOTE: OPINION CONTAINS TABLE OR OTHER DATA THAT IS NOT VIEWABLE

The defendants stipulated that if their device operated within Region I of the '210 patent graph, they would be infringing that patent. As already mentioned, to ascertain the effluent energy rate, the area (Ae ) at the "point of ejection" from the nozzle must be known. However, the parties disagree about the measurement of the exit diameter of the effluent stream from the defendants' nozzle--a critical difference which affects infringement. Because the defendants' nozzle has a flared end, the diameter in the throat is smaller than the diameter at the end of the nozzle. Defendants say that, if the effluent fills up the flared end due to back pressure, then the larger end diameter should be used, and thus their process would fall completely off the '210 patent graph. The flared end is said to be considered "functional" if the effluent fills up the flare, meeting the tank fluid at the end of the nozzle. Under this construction, the "point of ejection" is at the nozzle's end. On the other hand, if the flared end is not functional, as plaintiff says, the throat diameter is to be used because the effluent exits in a column which does not contact the walls of the flared end, and, thus, defendants' machine falls within Region I of the '210 patent graph. Under this latter construction, plaintiff contends that the flared end is merely non-functional and cosmetic, and the "point of ejection" is at the throat.

Defendants also stipulated that their machines included many of the elements of the '706 patent. They contend, however, that neither plaintiff's claimed "hollow tubular expansion chamber member" nor the incorporated '210 nozzles read on their converging-diverging off-the-shelf eductor nozzle.

Plaintiff appellant Envirotech sued defendant appellees Monosep for infringement of claims 1, 14 and 15 of the '210 patent and claims 1, 10, 11 and 12 of the '706 patent. Defendants Monosep counterclaimed for a declaratory judgment of invalidity and non-infringement. The action was tried before a judge and a jury from March 7, 1983 through March 11, 1983. The judge gave very general instructions on the law of patents and evidence, and then gave very general special interrogatories, simply asking whether each invention was or was not "novel", "new",1 "useful", "obvious", whether each patent "distinctly claims the inventions", whether each invention "has been adequately described," and also whether defendants infringed the patents. In answer to these summary interrogatories the jury found the claims at issue in the '210 patent "novel", "new", "not useful",2 "obvious", "fails" to "particularly point out and distinctly claim the subject matter",³ "fails" to adequately describe the subject matter,⁴ and not infringed. Further, the jury found the claims at issue in the '706 patent "novel", "obvious", and not infringed. Without further comment, the judge adopted the jury's findings and adjudged the claims at issue in the '210 patent invalid under 35 U.S.C. Secs. 101, 103, 112, and not infringed, and the claims at issue in the '706 patent invalid under 35 U.S.C. Sec. 103 and not infringed. The proposed findings of fact and conclusions of law of both parties were rejected.

The plaintiff subsequently made a motion for judgment notwithstanding the verdict and, in the alternative, for new trial based on the lack of substantial evidence to support the jury's verdict. By order, but without any memorandum, the district court denied the motion.

On this appeal, both the issues of validity and infringement are controverted. We consider them in the reverse order.

In general, a finding of infringement depends on whether the accused device falls within the scope of the asserted claims as properly interpreted. Kalman v. Kimberly-Clark Corp., 713 F.2d 760, 770, 218 U.S.P.Q. (BNA) 781, 788 (Fed.Cir.1983). The patented invention as indicated by the language of the claims must first be defined (a question of law), and then the trier must judge whether the claims cover the accused device (a question of fact). See Fromson v. Advance Offset Plate, Inc., 720 F.2d 1565, 1569, 219 U.S.P.Q. (BNA) 1137, 1140 (Fed.Cir.1983); SSIH Equipment S.A. v. USITC, 718 F.2d 365, 375, 218 U.S.P.Q. (BNA) 678, 688 (Fed.Cir.1983). The patent owner must show by a preponderance of the evidence that the accused has infringed his patent. Hughes Aircraft v. United States, 717 F.2d 1351, 1361, 219 U.S.P.Q. (BNA) 473, 480 (Fed.Cir.1983); SSIH, supra; Chisum, Patents, 18.06 (1983).

This was a jury case, and the judgment below must be scrutinized by the rules applicable to such jury cases. As the appellate court, we review the jury's findings of fact, (i.e., here that plaintiff failed to meet this burden) in light of the district court judge's denial of plaintiff's JNOV motion. A court cannot merely substitute its view for that of the jury's when reviewing questions of fact. Instead, the guidelines for considering motions for judgment notwithstanding the verdict are: (1) all of the evidence must be considered; (2) in a light most favorable to the non-moving party; (3) drawing all reasonable inferences favorable to that party; (4) without making determinations of credibility of the witnesses. Connell v. Sears, Roebuck & Co., 722 F.2d 1542, 1546, 220 U.S.P.Q. (BNA) 193, 197 (Fed.Cir.1983); Railroad Dynamics, Inc. v. A. Stucki Company, 727 F.2d 1506, at 1512 (Fed.Cir.1984). The judge on a JNOV motion (and this court on review) must ascertain whether there was "substantial evidence of such quality and weight that reasonable and fair-minded men in the exercise of impartial judgment could reasonably return a verdict for the non-moving party". Wyatt v. Interstate & Ocean Transport Co., 623 F.2d 888, 891 (4th Cir.1980). As this court pointed out in Connell and Railroad Dynamics, supra, these guidelines are applicable to patent infringement suits. Thus, the precise issue before us on infringement is whether there was substantial evidence to support the jury's factual finding that Envirotech failed to prove infringement by a preponderance of the evidence.The '210 Method Patent

It is elementary that resort must be had in the first instance to the words of the claim which define the metes and bounds of the invention. If the accused matter falls clearly within the terms of the claim, infringement is normally made out. Graver Tank & Mfg. Co. v. Linde Air Products Co., 339 U.S. 605, 607, 70 S. Ct. 854, 855-56, 94 L. Ed. 1097 (1950); Smith International, Inc. v. Hughes Tool Company, 718 F.2d 1573, 1579, fn. 2, 219 U.S.P.Q. (BNA) 686, 691 fn. 2 (Fed.Cir.1983). Said another way, what is patented must first be defined. Words in a claim "will be given their ordinary and accustomed meaning, unless it appears that the inventor used them differently". Universal Oil Products Co. v. Globe Oil & Refining Co., 137 F.2d 3, 6, 54 U.S.P.Q. (BNA) 504 (7th Cir.1943), aff'd, 322 U.S. 471, 64 S. Ct. 1110, 88 L. Ed. 1399 (1944).

Representative claim 1 of the '210 patent describes, inter alia, "[a] dispersed gas flotation process wherein ... the density and the kinetic energy rate ... per unit volume ... at the point of ejection...." (Emphasis added.) The area ("per unit volume") of the effluent discharged from the nozzle is measured at the "point of ejection". The diameter of defendant's nozzle used to measure that area depends upon the question of where the "point of ejection" exists in the diverging section of the nozzle; thus, infringement here depends on this "point of ejection".

Plaintiff appellant argues that the "point of ejection" describes the point where the effluent meets the tank fluid. Defendant appellees argue that ejection occurs at the end of the nozzle. In effect both are saying the same thing. The claim language may be understood, as appellees contend, to define the "point of ejection" as the point where the effluent effectively leaves the piece of hardware. However, this does not rule out an interpretation that the "point of ejection" may be at the throat area as Envirotech contends. The flared end of defendant's nozzle may not be functional, and thus the effluent effectively leaves the nozzle at the throat area. We therefore construe the claim to describe the "point of ejection" as the point where the effluent effectively meets the tank liquid.

It was up to the jury to determine whether the defendant's flared end is in fact functional or whether the effluent effectively meets the tank fluid in the throat area of the nozzle. As applied to the evidence, the significance of the latter determination if made, as we have pointed out, would be that defendants' process would fall within Region I of the '210 graph.

Envirotech asserts that the flared end is functional only when there is enough back pressure in the flare to fill up the flared end. To support its contention, Envirotech called as its expert witness, Dr. Bourgoyne, professor and chairman of the petroleum engineering department at Louisiana State University. During his testimony, Dr. Bourgoyne offered a demonstration of a plastic replica of defendants' device in an attempt to show that the effluent ejected in a column never touched the sides of the flared end.5 Bourgoyne stated that the plastic replica would behave the same as defendants' cast-iron eductor. The entire jury stepped up to observe both the eductor and the fluid flow seen through the plastic.

Defendant called Dr. Muster, professor of mechanical engineering at the University of Houston. He asserted that one cannot tell precisely where the "point of ejection" would be because of changing conditions in a turbulent zone. If anything, he argued, the most sensible "point of ejection" would be where the effluent exited at the end of the nozzle. In addition, defendants read into the record the deposition of Dr. Colbert, one of the patentees, which appears to say that, with a nozzle of defendants' type, the "point of ejection" is at the end of the nozzle.

The jury, faced with a clear conflict in evidence, returned a finding that defendants had not infringed the claims of the '210 patent. This is plainly a factual issue. Based on the testimony of the two experts and the in-court demonstration we have to agree that there was substantial evidence on which the jury could make a finding of non-infringement. Accordingly, we hold that plaintiff failed to prove by a preponderance of the evidence infringement of claims 1, 14 and 15 of the '210 patent.

Once again we must start with the language of the claims. The element in the claims which is dispositive requires, inter alia, a:

"fluid ejection device including: (i) a hollow tubular expansion chamber member which has an open end through which the mixed fluid is ejected into a liquid..." (emphasis added)

Thus, the precise issue for literal infringement of the '706 patent is whether defendants' converging-diverging eductor comprises a "hollow tubular expansion chamber member" within the meaning of the claims. The terms of claims are best construed in light of the specification and the circumstances which surround the patent at its inception. Fromson v. Advance Offset Plate, Inc., 720 F.2d 1565, 1569, 219 U.S.P.Q. (BNA) 1137, 1140 (Fed.Cir.1983); Autogiro Co. v. United States, 384 F.2d 391, 397, 155 U.S.P.Q. (BNA) 697, 702 (Ct.Cl.1967). Adverting to the specification, appellant argues that the fluid ejection devices (nozzles) claimed in the '210 patent are incorporated by reference into the '706 patent specification, and accordingly the claims read on defendants' devices. Claim 12 of the '210 patent recites a converging-diverging type of nozzle to be used with the process of claim 1 in the '210 patent. Claim 14 of the '210 patent recites an expansion chamber type of device also to be used with the process of claim 1. Those recitations of the '210 patent are part of the '706 patent specification. See General Electric Co. v. United States, 572 F.2d 745, 758, 198 U.S.P.Q. (BNA) 65, 76 (Ct.Cl.1978); In re Schaumann, 572 F.2d 312, 317 n. 11, 197 U.S.P.Q. (BNA) 5, 9 fn. 11 (Cust. & Pat.App.1978); Velo-Bind, Inc. v. Minn. Mining & Mfg. Co., 647 F.2d 965, 968, 211 U.S.P.Q. (BNA) 926, 929 (9th Cir.1981). However, in connection with infringement of the '706 patent, two essential points must be remembered: first, it is still the claims of the '706 patent which must be found infringed; and second, incorporation by reference can only aid in the construction of the '706 claims.

On the basis of the incorporation by reference, appellant argues that its pipe-within-a-pipe configuration which requires a "hollow tubular expansion chamber member" includes the converging-diverging type of nozzle.6 But defendants' expert, Dr. Muster, testified that the Monosep eductor does not have a hollow straight tubular member. Even the throat portion of the eductor, which is straight over a small length relative to the entire nozzle, would not qualify. Not only was the jury aware of Dr. Muster's testimony but there was the in-court demonstration of a purported replica. Although the purpose of the demonstration was to show where the "point of ejection" was within the meaning of the '210 patent, the jury could use the demonstration to ascertain whether there was an expansion chamber within the meaning of the '706 patent. Thus, there was substantial evidence of such quality and weight to support a non-infringement verdict in favor of Monosep. Plaintiff did not meet its burden to prove literal infringement of the '706 patent.

Alternatively, plaintiff alleges infringement under the doctrine of equivalents. This doctrine is usually asserted when actual literal infringement is not present. See Hughes Aircraft Company v. United States, 717 F.2d 1351, 1361, 219 U.S.P.Q. (BNA) 473, 480 (Fed.Cir.1983). An accused device may infringe "if it performs substantially the same function in substantially the same way to obtain the same result" as the claimed apparatus. Graver Tank & Mfg. Co. v. Linde Air Products Co., 339 U.S. 605, 608, 70 S. Ct. 854, 859, 94 L. Ed. 1097 (1950). A finding of equivalence is a determination of fact which may be based on proof by testimony of experts, and should only be disturbed as allowed by the general principles governing appellate review. Id. at 609-610, 70 S.Ct. at 856-57.

Even if we assume that the '210 claimed nozzles were incorporated by reference into the '706 patent, plaintiff has still failed in its burden to prove infringement by equivalence. The jury had before it both patents and was aware of the incorporation by reference. In addition, the judge gave the correct definition of the doctrine of equivalents to the jury. Testimony by Dr. Muster showed, contrary to appellant's assertions, that it was disputable whether a pipe-within-a-pipe nozzle and converging-diverging nozzle perform the same function in the same way to obtain the same result. Accordingly, we cannot say that there was not substantial evidence on which the jury could base its verdict of non-infringement. The result is that Envirotech failed to prove by a preponderance of the evidence Monosep's infringement of claims 1, 10, 11 and 12 of the '706 patent.

Our affirmance of the judgment of non-infringement may render academic (for this case) the validity of the '210 and '706 patents. But defendant-appellees have sought a declaratory judgment of invalidity, and may wish to pursue it on the remand which we consider necessary on the issue of validity.7

We consider that serious errors vitiated the district court's determination of invalidity. The judge gave erroneous jury instructions which marred the jury's consideration of the factual and legal issues bearing on validity.

"as to information or documents (prior art) which you find from the file wrapper to not have been considered by the examiner and which you find to teach an invention closer to or more like that claimed ... no such presumption (of validity) exists..."

This instruction is not proper. The "presumption does not change upon introduction of that art, or at any other time". Connell v. Sears, Roebuck & Co., 722 F.2d 1542, 1549, 220 U.S.P.Q. (BNA) 193, 200 (Fed.Cir.1983). See also, American Hoist & Derrick Co. v. Sowa & Sons, Inc., 725 F.2d 1350 at 1358-1361 (Fed.Cir.1984). It is upon the introduction of more relevant art that the challenger's burden of persuasion may be more easily carried.8

The district court also erred in its handling of the utility defense. Defendants argued that the alleged critical limitations represented by Region I of the '210 patent graph do not truly delineate the desired operating conditions described in the patent. In addition, they alleged a lack of exact correlation between experimental graphs and the actual patent graph. As we understand them, these defenses are based on the specification's alleged failure to disclose adequately to one ordinarily skilled in the art "how to use" the invention without undue experimentation--usually considered the "how-to-use" defense under 35 U.S.C. Sec. 112. See, Chisum, Patents, 7.03 (1983). However, such assertions have been applied to an argument for lack of utility under 35 U.S.C. Sec. 101 when there is a complete absence of data supporting the statements which set forth the desired results of the claimed invention. See, e.g., In re Ruskin, 354 F.2d 395, 148 U.S.P.Q. (BNA) 221 (Cust. & Pat.App.1966) (specification devoid of any quantitative data). The trial judge did nothing to dispel any confusion between Sec. 101 and Sec. 112. He even added to the problem by first allowing defendants' last amendment of the pleadings to include a lack of utility defense,9 and then merely quoting from Sec. 101 in his charge. Even if defendants' argument of inadequate experimentation or inexact correlation were to apply, the fact that an invention has only limited utility and is only operable in certain applications is not grounds for finding lack of utility. Raytheon Co. v. Roper Corp., 724 F.2d 951, 958-59, 220 U.S.P.Q. (BNA) 592, 598 (Fed.Cir.1983); Carpet Seaming Tape Licensing Corp. v. Best Seam, Inc., 694 F.2d 570, 578, 216 U.S.P.Q. (BNA) 873, 880 (9th Cir.1982). Some degree of utility is sufficient for patentability. E.I. du Pont de Nemours & Co. v. Berkley and Co., 620 F.2d 1247, 1260, fn. 17, 205 U.S.P.Q. (BNA) 1, 10 (8th Cir.1980). Further, the defense of non-utility cannot be sustained without proof of total incapacity. Id.

Perhaps the most glaring defect arises from the defective charge on the issue of obviousness. In his instructions to the jury the judge did not set forth the trilogy of factual inquiries outlined in Graham v. John Deere Co., 383 U.S. 1, 86 S. Ct. 684, 15 L. Ed. 2d 545, 148 U.S.P.Q. (BNA) 459 (1966): (1) the scope and content of the prior art; (2) the differences between the prior art and the claims at issue; and (3) the level of ordinary skill in the art. Nor was the jury called upon to answer special interrogatories on these inquiries, or to give special verdicts thereupon.

Submission of the obviousness question to the jury should preferably be accompanied by specified interrogatories designed to elicit responses to all the factual inquiries enumerated in Graham. Proper instructions based on those inquiries should always be given.10 See Connell, supra, 722 F.2d at 1547; American Hoist & Derrick Co. v. Sowa & Sons, Inc., supra, at p. 1361; Railroad Dynamics, Inc. v. A. Stucki Company, supra, slip op. at 17 ff. In the absence of such special findings and of a proper charge on which a conclusion of obviousness can be reviewed, and because the parties here are not yet in substantial agreement as to facts bearing on the obviousness issue, it is appropriate to remand the present case.

Accordingly, in light of the above errors, we cannot affirm the lower court's denial of Envirotech's JNOV motion regarding invalidity. "(This) court must be satisfied ... that the party challenging validity has [properly] carried its burden of overcoming the presumption (of validity)". Medtronic, Inc. v. Cardiac Pacemakers, Inc., 721 F.2d 1563, 1567, 220 U.S.P.Q. (BNA) 97, 100 (Fed.Cir.1983). We are not satisfied. As a court never "declares" a patent valid, we merely vacate and remand on the issue of validity. We leave initially to the district court the extent to which a new trial is required if the issue of patent invalidity is pursued.11 However, if the parties can stipulate all the relevant facts, the judge can decide for himself the legal issues relating to validity. See White v. Jeffrey Mining Machinery Co., 723 F.2d 1553 (Fed.Cir.1983).

We affirm the judgment that the inventions set forth in claims 1, 14 and 15 of the '210 patent, and in claims 1, 10, 11 and 12 of the '706 patent are not infringed; we vacate and remand on the issue of the validity of these claims. Each side will bear its own costs in this court.

AFFIRMED IN PART, VACATED AND REMANDED IN PART.

BALDWIN, Circuit Judge, specially concurring.

I only wish to clarify that in this case the claims at issue were submitted to the jury with proper instructions. The jury in its deliberations construed the claims in accordance with those instructions and then determined there was no infringement.

Further, regarding the '210 patent, we have not construed claim 1 de novo in this appeal. We are presuming this particular construction as it is consistent with the jury verdict and it is supported by substantial evidence.