At issue in this patent interference case is whether employee Driskell of the Tennessee Valley Authority (TVA) or employee Jones of the Monsanto Company (Monsanto)1 first invented a certain chemical process. The United States District Court for the Northern District of Alabama awarded priority of invention to Monsanto, assignee of the Jones invention. TVA, assignee of the Driskell, invention perfected this appeal. We affirm.

Beginning in 1949, John C. Driskell, who headed a team of TVA research chemists working at TVA's National Fertilizer Development Center at Muscle Shoals, Alabama, conducted experiments on a process for producing nitrogenphosphorus compounds for use as a fertilizer. On October 27, 1954, Otha C. Jones, and employee of Monsanto, filed application No. 465,201 with the Patent Office for a patent on a chemical process similar or identical to the process involved in the Driskell experiments. On January 7, 1955, while the application on the Jones patent was still pending, Driskell filed an application and was awarded patent No. 2,713,536 on the process and the product on July 19, 1955. On April 30, 1956, Jones filed application No. 581,665 for a patent which was in part a continuation of his prior application.2 He also filed a request that an interference with the Driskell patent be established.

Accordingly, the United States Patent Office instituted an interference proceeding under the provisions of 35 U.S.C. 135. Jones was accorded the benefit of his filing date of October 27, 1954, as constructive reduction to practice and consequently he was designated senior party in the interference proceedings. Therefore Driskell, as junior party, had the burden of proving by a preponderance of the evidence an earlier successful reduction to practice.3

The primary examiner found that an interference existed between that portion of the Driskell patent describing the chemical process and the Jones application. The interfering subject matter was set out in four counts. These counts in interference are:

'Count 5. The process which comprises the steps of drying air to a moisture content of less than about 0.0004 pound water per pound dry air; oxidizing elemental phosphorus with the dry air; cooling the products of combustion to a temperature of 450 degrees to 950 degrees F.; reacting the phosphorus pentoxide vapor in the cooled combustion product with anhydrous ammonia; and collecting the solid, finely divided product of said reaction. 'Count 6. The process of claim (Count) 5 in which the proportion of ammonia reacted with phosphorus pentoxide is within the range of 2.1 to 2.7 moles NH(3) per mole of P(2)O(5). 'Count 7. The process of claim (Count) 5 wherein the products of combustion are cooled to a temperature in the range from 550 degrees to 700 degrees F. prior to reaction with anhydrous ammonia. 'Count 8. The process of claim (Count) 5 wherein the products of reaction are cooled to a temperature in the range from 550 degrees to 700 degrees F.; and the phosphorus pentoxide therein is reacted with from 2.1 to 2.7 moles of ammonia per mole of phosphorus pentoxide.'

In proceedings before the Patent Office Board of Patent Interferences, only TVA presented testimony, Monsanto relying on its filing date of October 27, 1954, as constructive reduction to practice. The Board awarded priority of invention to Jones-Monsanto as to the process specifically defined in the aforementioned counts. The Board held that in order for Driskell-TVA, to be awarded priority he 'must not only prove that the process was operated by him, or under his supervision, in conformity with the requirements of the counts but also establish the identity of the reaction product with reasonable certainty as well as the utility thereof.' The Board concluded that 'Driskell has failed to satisfactorily prove the identity of the product and its utility and so cannot prevail here.' That decision was adhered to by the Board on reconsideration.

TVA then brought suit in the district court against Monsanto under 35 U.S.C. 146, 28 U.S.C. 1337, 1338 and 1345, and the court awarded priority to Monsanto. The court found from the testimony that TVA did not establish with the certainty and reasonableness the law requires that all the process steps as specifically defined in the counts of the interference were carried out by Driskell or those working with him. This decision was based on the court's finding that there was no evidence to show that the dryness of the air was measured or that the temperature of the product of combustion of phosphorus and air was determined prior to reaction with anhydrous ammonia. The Patent Office's decision that TVA did not establish the identity of the reaction product was found by the court to be controlling on the ground that the evidence was not of such character and amount as to carry the thorough conviction that such decision was in error, nor was the additional evidence presented to the court competent or sufficient to prove such identity. Consequently, the action was dismissed and TVA's motion for a new trial was overruled.

On appeal TVA contends that the district court erred in requiring proof of the identity or the composition of the reaction product. Also, it submits that the court used an erroneous standard of proof in reviewing the decision of the Board. Finally, TVA contends that the court's finding that there was 'no evidence' that the steps of the process were carried out was clearly erroneous.4

Driskell was issued a patent on a process for the production of nitrogenphosphorus compounds. The steps of this process are described in the patent as follows:

'drying air to a moisture content of less than about 0.00008 pound water per pound of dry air; oxidizing elemental phosphorus with the dried air; cooling the products of combustion to a temperature of 450 degrees to 950 degrees F., preferably 550 degrees to 700 degrees F.; reacting the phosphorus pentoxide vapor in the cooled combustion gases with ammonia in a proportion of from 2.1 to 2.7 moles of NH(3) to 1.0 mole of P(2)O(5); and collecting the solid, finely divided product of said reaction.'

The patent further states that the 'product of this reaction is a new composition of matter consisting of an intimate mixture of ammonium metaphosphate, phosphoronitridic acid, and ammonium phosphoronitridate.' The patent claims that this product, being high in nitrogen and phosphorus, is valuable as a fertilizer material. It is further claimed that the product is useful as an intermediate in the preparation of other nitrogen-phosphorus materials, such as water-softening compounds and fire retardants.

In order for Driskell to prove priority of invention he must prove that he reduced this process to practice before Jones' filing date of October 27, 1954. Even though the counts in interference, quoted earlier, are related primarily to the chemical process involved, not the product resulting from the process, the reaction product is an essential element in proving reduction to practice. Where the invention is a method for making a certain product, it is not reduced to practice until it is established that the steps of the process were actually performed and that such process has utility, namely, that the process produces that certain product and that such product is useful. Brenner v. Manson, 383 U.S. 519, 86 S. Ct. 1033, 16 L. Ed. 2d 69 (1966); Alpert v. Slatin, 305 F.2d 891, 49 C.C.P.A. 1343 (1962); Allen, et al. v. Bertorelli, et al., 100 U.S.App.D.C. 329, 244 F.2d 777 (1957); Reiners v. Mehltretter, 236 F.2d 418, 43 C.C.P.A. 1019 (1956); Birmingham v. Randall, 171 F.2d 957, 36 C.C.P.A. 780 (1948); Schmutzer v. Ayers, 129 F.2d 703, 29 C.C.P.A. 1176 (1942). Consequently, Driskell must prove by a preponderance of the evidence that he carried out each step of the chemical process and that the process actually results in the production of the product described in the patent, nitrogen-phosphorus compounds (identity of the reaction product), and that such product is useful.5

Driskell emphatically contends that it is not necessary to establish the identity or composition of the reaction product. He submits that he is only required to prove that he carried out the process and that the resulting product is useful. He relies heavily on Corona Cord Tire Co. v. Dovan Chemical Corp., 276 U.S. 358, 48 S. Ct. 380, 72 L. Ed. 610 (1928), which involved a dispute over who first reduced to practice a certain process for the vulcanization of rubber. In Corona the court stated that '(a) process is reduced to practice when it is successfully performed. * * *' Such statement seems to indicate that in proving reduction to practice of a process it is not necessary to prove that the claimed product resulted from the process or that such product is useful. However, the opinion in Corona, taken in its entirety, clearly specifies that priority of invention was awarded on the court's finding that the specific process of vulcanization was carried out, that it produced high quality vulcanized rubber and that such product had been successfully tested. See Blicke v. Treves, 241 F.2d 718, 44 C.C.P.A. 753 (1957).

In further defense of his contention that proof of identity is unnecessary, Driskell submits that a determination of whether he has reduced the process to practice should be limited to the counts of interference. Since the counts do not specifically describe the product, Driskell argues that such is beyond the consideration of the court. However, it is quite proper and often necessary in determining whether a reduction to practice has been accomplished, to take into consideration statements beyond the claims appearing in the counts of interference. For example, even though the counts do not specify any use for the claimed product, that fact does not eliminate the necessity of showing the product's utility. See Blicke v. Treves, supra; Birmingham v. Tandall, supra; Muskat, et al. v. Schmelkes, 140 F.2d 984, 31 C.C.P.A. 837 (1944).

But, as a matter of fact, we are not faced with a situation where the counts are entirely silent as to the product involved. The last phrase in Count 5 states: 'and collecting the solid, finely divided product of said reaction.' This product has been specifically defined in Driskell's patent. His patent is a process for producing nitrogen-phosphorus compounds and the patent claims that the process produces a specific product composed of ammonium metaphosphate, NH(4)PO(3), phosphoronitridic acid, (OH)(2)PN, and ammonium phosphoronitridate, NH(4)OOHPN, and that such specific mixture, NH(4)PO(3), (OH)(2)PN and NH(4)OOHPN, is valuable as a fertilizer and as an intermediate product. Hence, establishing the utility of the process in question is, in part, dependent upon showing that the process produces the exact product which Driskell asserts that it does produce.

An examination of one item of evidence presented by TVA to the Board to establish the utility of the reaction product discloses another reason why identifying the reaction product was essential. Such evidence tends to show that treating the soil with certain named chemical compounds will result in an increased yield of millet. Obviously, these chemical compounds are valuable as a fertilizer. However, this does not prove that the reaction product has utility as a fertilizer unless the reaction product is chemically identical to the products used in the experiment. Therefore, the district court was correct in requiring proof of the chemical composition or identity of the reaction product.

The general rule in a patent interference case is that the junior party must prove reduction to practice by a preponderance of the evidence.6 However, where the Patent Office has decided a question of fact from the evidence before it, such decision must be accepted by the court unless the contrary is established by evidence 'which in character and amount carries thorough conviction' that the Patent Office decision was in error. Morgan v. Daniels, 153 U.S. 120, 14 S. Ct. 772, 38 L. Ed. 657 (1894); Aloe Creme Laboratories, Inc. v. Texas Pharmacal Co., 335 F.2d 72 (5 Cir. 1964).7 All parties agree that the above is a correct statement as to the standards of proof which must be applied by the court to the issues involved.

In the proceeding before the Patent Office Board of Interferences, the Board found that Driskell had not established the identity or utility of the reaction product. However, the Board refused to rule on whether Driskell had carried out the steps of the process because, in its opinion, it was unnecessary for decision of the case. Therefore, since the Board did not rule on the evidence before it pertaining to Driskell's performance of the steps of the process, TVA need only prove that it carried out the steps of the process by a preponderance of the evidence. However, since the Board ruled on the question of identity and utility of the reaction product, TVA must comply with the 'thorough conviction' rule announced in Morgan v. Daniels, supra.

TVA contends that the district court applied the same standard of proof, the thorough conviction test, to the issue of proof of identity and to the issue of proof of whether Driskell carried out the steps of the process. If the court did in fact require TVA to prove that it carried out each step of the process by the thorough conviction test we would have to find the court in error. However, we have carefully reviewed the district court opinion and we find it correctly applied the proper standard of proof.

In conclusion of law No. 4 the court states that the Board of Patent Interferences held that the evidence before it did not establish the identity of the reaction product and then refers to the strict standard of proof, the thorough conviction test, as the standard it must use in reviewing the Board's decision. The court then states at the beginning of conclusion of law No. 5 that the evidence presented before the Patent Office was not of such character and amount as to carry the thorough conviction that the Patent Office decision was in error. It is obvious from a reading of the opinion that up to this point the court was dealing solely with the issue of identity as decided by the Patent Office and therefore was applying the correct standard of proof.

Conclusion of law No. 7 also deals with the issue of proof of identity of the reaction product but only with regard to the new and additional evidence on this point which was submitted by TVA to the district court. The court concluded that this additional evidence as to identity was not of such character and amount as to carry thorough conviction that the decision of the Patent Office should be reversed. In conclusion of law No. 8 the district court applied the preponderance test as to new testimony on the issue of identity, a much less stringent test, and concluded that the new testimony was not sufficient to prove identity or to justify the court in overruling the decision of the Patent Office. Accordingly, the decision of the Patent Office that TVA had not reduced the process to practice should not be disturbed.

The remainder of conclusion of law No. 5 and conclusion of Law No. 9 deal with the issue of whether TVA carried out the specific steps of the process. The language employed by the court in deciding this issue makes no mention of the thorough conviction test. Instead the court found that the evidence submitted by TVA to the Board did not establish, with the certainty and reasonableness the law requires, that all the process steps were actually carried out. Also with respect to the additional testimony presented to the court on the question of whether the steps of the process were carried out, the court concluded that this evidence was not competent of sufficient. Obviously, the court was applying the lesser standard of proof.8

Therefore the opinion of the trial court clearly demonstrates that it was aware that different standards should be applied, separated its consideration of evidence ruled on by the Patent Office and that evidence which had not been presented to or ruled on by the Patent Office, and applied the correct standard to each.

Two steps of Driskell's chemical process as described in the counts of interference are as follows:

'drying air to a moisture content of less than about 0.0004 pound water per pound dry air; * * * cooling the products of combustion to a temperature of 450 degrees to 950 degrees F.; reacting the phosphorous pentoxide vapor in the cooled combustion product with anhydrous ammonia;'

The district court concluded that TVA had not proved that it actually carried out each step of the process. In reaching this conclusion the court made two findings of fact which TVA contends are clearly erroneous. Stated briefly they are: (1) there is no evidence that the dryness of the air passing through the drier was ever measured to determine whether the air had less than the maximum water content permitted by the specific terms of the counts of the interference, and (2) there is no evidence that the product of combustion of phosphorus and air in any particular experiment was cooled to a temperature of 450 degrees to 950 degrees F. prior to reaction thereof with anhydrous ammonia as called for by the counts of the interference.

The first experiments carried out by TVA involved the use of glassware equipment and some 33 experiments were made which are referred to as the 'L Series.' Later a larger apparatus was constructed in order to continue the experimentation on a larger scale. Some 46 experiments (designated as the 'LS Series') were carried out in the larger scale equipment. Both parties agree that no specific measurements of the moisture contents of the air in the 'L' and 'LS' runs were made. Since TVA admits it did not measure the dryness of the air, the court's finding tht there is no evidence that the dryness of the air was ever measured is clearly correct. However, we think TVA should not be restricted in its proof of the moisture content of the air to just evidence showing an exact measurement of the dryness of the air. Accordingly, we examine the evidence submitted by TVA on this point. TVA witnesses testified that its equipment was provided with an air drying apparatus containing anhydrous calcium sulfate (sold under the trademark Drierite) or calcium chloride which would ordinarily dry air more than would be necessary to meet the specifications. From this testimony it possibly could be inferred that use of this equipment dried the air to the extent called for in the specifications. However, use of the apparatus does not conclusively show that the air was dried to the required degree. Moreover, Driskell and other TVA employees testified that the drying apparatus would not dry the air properly after it had been in use for a period of time. No evidence was presented that the apparatus was adequately operating at the time of the runs.

Consequently, we read the district court's statement of fact as a finding that TVA has not proved by a preponderance of the evidence that the air was dried to the proper degree, and on the basis of the record before us we cannot say that such finding is clearly erroneous.

Several years later, pilot runs were made by TVA. The district court found that there was not sufficient competent or admissible evidence to establish that the specific process of the counts of the interference were carried out in these pilot runs. Unlike the initial experiments TVA did present testimony tending to prove the fact that two measurements were made on one of the runs. However, one measurement was made by an individual who was never produced as a witness and the other was made shortly before the end of the run and thus did not directly bear on the condition of the air during the entire run. Therefore, we cannot say that the district court's finding that there was not sufficient evidence to establish that the air was properly dried is clearly erroneous.

One of the TVA researchers testified as to temperature ranges during a number of runs in both the 'L' and 'LS' series. Data sheets for several 'LS' runs were also introduced. However, the district court found that the testimony showed that the measurements were not taken immediately prior to the reaction with the anhydrous ammonia as the counts require.

Since TVA did present some evidence regarding whether the product of combustion of phosphorous and air was cooled to a temperature of 450 degrees to 950 degrees F., the district court perhaps used an unfortunate expression in finding that there was no evidence on this point. However, our review of the record convinces us that TVA has not proved by a preponderance of the evidence that the product was cooled to a temperature of 450 degrees to 950 degrees F. prior to reaction thereof with anhydrous ammonia as called for in the counts of the interference. We therefore cannot say that the district court was clearly erroneous.

Having found that the district court committed no error in its requirement of proof of identity, or in its application of the proper standards of proof, or in its findings of fact, we hereby affirm the judgment of the court.

Judgment affirmed.