This is an appeal by Kenneth L. Berry from the decision of the Board of Patent Interferences awarding priority as to the subject matter in interference No. 93,724 to Watt W. Webb, William A. Wissler and William D. Forgeng (hereinafter Webb).

Berry, the senior party, is involved on application serial No. 696,595, filed November 15, 1957 and assigned to E. I. du Pont de Nemours & Company, and Webb on U. S. Patent No. 3,011,870, granted on an application filed September 5, 1958 and assigned to Union Carbide Corporation.

The invention in issue relates to new types of crystalline alpha-aluminum oxide, namely α-A1₂O₃ (also known as corundum, sapphire and α-alumina), in the form of fibers and platelets. The counts, corresponding to claims 7 and 9 of the junior party Webb's patent, read:

1. Needle-like high strength, virtually perfect single crystals of alpha-aluminum oxide (α-A1₂O₃) having a hexagonal cross-section and a length from about 1 to 30 mm. long and a width from about 3 to 50 microns.

2. Platelet-like high strength, virtually perfect single crystals of alpha-aluminum oxide (α-A1₂O₃) having a length up to about 10 mm. long and a thickness from about 0.5 to 10 microns.

Both parties, in attempting to show priority of invention by earlier reduction to practice, presented considerable evidence in the nature of testimony and physical exhibits briefly summarized below.

Webb, who was in 1957 a research metallurgist at the Union Carbide laboratories in Niagara Falls, testified that in March of 1957 Wissler brought to him certain crystals and told him how they were made since he knew of Webb's interest in these things inasmuch as Webb was engaged in a research program attempting to find high-strength whisker crystals. It does not appear that Webb and Wissler knew at the time what the crystals were; however, Webb consulted with Forgeng, at that time head of the Metallographic Department of the Technology Department, and the two decided to have x-ray diffraction analyses of some of the crystals made. In addition they looked at them with an optical microscope, Webb testifying that the preliminary conclusion reached was that they were alpha; Al₂O₃ commonly called "sapphire."

Webb was assisted in his work on crystalline materials by Durham whose work included the physical testing of the crystals. Durham testified with reference to his own notebook records, Webb exhibit 7, and stated that he received certain crystals from Webb in early March 1957, performed bend tests thereon, and then gave crystals from the group to Faulring to determine the identity of the material. Durham also testified that these crystals had a length in excess of one millimeter and a diameter determined as 9.88 microns.

Faulring testified to running x-ray diffraction patterns before March 11, 1957 on samples of the crystals and finding them to be single crystals of alpha-alumina.

Wissler was asked to produce additional crystals and a second set of crystals was produced in April 1957 in Wissler's furnace. The set of crystals was removed from Wissler's furnace by Webb and Durham and taken to the laboratory while still warm. Durham testified that a specific crystal mentioned in his notebook was selected from this second set and its deformation measured. Its diameter was said to be 3.3 microns with a length in excess of one millimeter as determined from a photograph. Durham testified that he detected no difference in apparent physical properties between the two batches of crystals. Durham further stated that he gave crystals from the second set to Dragsdorf.

Dragsdorf testified that he made x-ray observations on whiskers obtained from the laboratory of Forgeng and found them to be single crystals of alpha-alumina having a single axial screw dislocation. Dragsdorf further testified that the whiskers he tested were 1-4 millimeters long, 5-30 microns in diameter and all had hexagonal cross-sections.

Durham also testified that two types of crystals were present, platelets or blades and straight crystals, and that Webb exhibit 4 is a photograph of a "blade" which was taken by him on April 17, 1957. He recalled that the length of such crystals was in excess of one millimeter and stated that the thickness would be comparable to that of a single crystal, on the order of 3-5 microns.

Saia, an electron microscopist under the supervision of Forgeng, testified that he took diffraction pattern photographs of platelets and that Webb exhibit 4 illustrates the type of platelet upon which he made observations. Forgeng testified that Saia gave him the photographs and from these one could tell whether or not the material was alpha-aluminum oxide.

Berry, an employee in the Chemical Department of the Dupont Company, was involved in early 1955 in a research project dealing with fibrous refractory inorganic compounds including preparation of corundum in a fibrous crystal habit. In September 1955 Berry began a series of experiments designed to obtain fibrous crystals of α-alumina. In the first of these experiments fibers were obtained, personally submitted to the x-ray laboratory for anaylsis and reported to be corundum.

The experiments performed by Berry during a period from September 1955 to January 1957 which followed the initial experiment similarly resulted in the formation of fibers, and in some cases ribbons, found on x-ray diffraction analysis to be corundum. Casey, a laboratory technician, testified to assisting Berry in carrying out the experiments. Casey testified that he had examined some of the fibrous products under microscope and that Berry discussed some of the x-ray reports with him. Simmons, a chemist working in the same laboratory with Berry, testified that he observed the experiments carried out by Berry and witnessed the notebook records of those experiments.

Young, Berry's supervisor, testified that he was acquainted with the work of Berry on fibrous alumina and was personally satisfied that Berry had produced corundum and had approved a report, Berry exhibit 28, covering the period September 30, 1955 to January 31, 1956 and containing information concerning the production of corundum in fibrous form. Balthis, a research chemist, was assigned to a project on fibrous corundum and, after discussion with Berry, performed two experiments in the period November 1956-January 1957, obtaining a product the x-ray analysis of which showed it to be corundum.

A number of x-ray report cards showing the results of analyses run on the products produced by Berry and Balthis have been included as exhibits. Schupp, a research chemist supervising the x-ray laboratory, testified that these were standard x-ray report cards and that they were reviewed by him prior to being returned to the experimenter and that at least in some instances the results were orally reported to the experimenter submitting the sample. Biddle and Babcock, who were technicians doing x-ray diffraction analysis, testified with respect to the interpretation of the results shown on the fiber produced in making the x-ray diffraction analysis. All testified that the x-ray analyses showed that the samples submitted by Berry and Balthis were corundum. Marshall, a technician engaged mostly in photographic and microscopic work, testified that he photographed several of Berry's samples.

Additionally, Berry testified that at the time of the initial preparation of fibers he had in mind that they would be useful for reinforcing plastics, ceramics, and metal, and also as an insulating, ablating and filtering material. The last-mentioned utility was tested on May 3, 1956 and the product found satisfactory.

The board first reviewed a decision of the examiner denying a motion by Webb to dissolve the interference on the ground that Berry had no right to make the counts in that his application lacked support for expressed limitations, particularly the limitations "high strength," "virtually perfect" and "hexagonal cross-section," and further that Berry did not have in his application, within one year of the issuance of the Webb patent, claims corresponding to or substantially the same as the counts of the interference. The board sustained the examiner, noting that the Forshey testimony and affidavits on behalf of Berry, comparing the process disclosed in the patent with that of the application, stated that the products of the two were identical, and concluded that Berry's original claims inherently included the challenged limitations of the counts. Webb does not challenge the board's holding in this regard, nor do we see any error.

The board then considered the evidence of Berry, finding it adequate in all respects except that:

It is noted that count 1 requires that the needle-like crystals of a [sic] alpha-alumina be from 3 to 50 microns wide and count 2 requires that platelet-like crystals have a thickness from about 0.5 to 10 microns. We have carefully considered the Berry record and we find no corroborating evidence that the crystals produced in the experiments in question conform to these limitations.

B. Ex. 24 (Berry's laboratory notebook record of his experiment of 12-23-55) sets out that the fibers produced are up to 3 millimeters long and have a diameter varying from less than one micron to about 10 microns. It is clear that this document cannot be regarded as corroboration evidence that the fiber had the width specified in count 1, being a report or notes written by the inventor, Reed v. Cislak, et al., [175 F.2d 972,] 36 CCPA 1117, 1949 C.D. 400, and Searle v. Glarum et al., [179 F.2d 974,] 37 CCPA 896, 1950 C.D. 206. B. Ex. 17 (Berry's notebook record) states that the fibers produced were "less than 1 micron in maximum cross sectional axis". Count 1 specifies a width of about 3 to 50 microns. B. Ex. 18, the notebook record of another experiment by Berry, refers to ribbon shaped crystals ranging from 5 to 50 microns wide. B. Ex. 25 also records the width of the fibers produced. All of these exhibits, being notes of the inventor himself, do not furnish the required corroboration of this feature. The testimony of Berry that the crystals produced by him in the experiments in question had the dimensions specified in the counts are merely self-serving declarations and does not establish that they had the required thickness or width in the absence of some independent corroboration.

* * * * * *

Berry urges that B. Exs. 8, 9 and 10 can be utilized to measure the dimension of the crystals produced. These are said to be photographs of fibers produced by Berry in September 1955. This Board will not attempt to interpret photographs to determine whether the product shown has the size specified in the counts. Further it is not apparent to us that the photographs are clear enough to enable such measurements to be taken.

Considering Webb's evidence, the board disagreed with Berry that Webb failed to show or appreciate utility of the invention, judicially noting patents of record in one or both of the involved cases disclosing corundum to have utility and finding such utility to have been obvious. The board concluded that:

* * * the record as a whole clearly establishes that Webb et al. reduced the invention of counts 1 and 2 to practice at least as early as June 1957 when Dragsdorf completed his calculations based on x-ray diffraction patterns.

It awarded priority to Webb.

We turn first, for reasons which will become evident hereinafter, to a review of the board's holding that Berry failed to prove reduction to practice of the invention defined by the counts. Although the board agreed that Berry's proofs satisfied the count requirements except as to the limitations of dimensions, the appellee Webb is not so generous, alleging, inter alia, that no witness for Berry can testify that he knew for a fact that any fibrous alumina was produced by or ever in the possession of Berry, that Simmons' over-all testimony should be given little weight in view of mistakes in testimony and that the x-ray exhibits representing the x-ray camera work of operators Sandra Flam and Lynn Forbes should be given no weight due to the failure of the two to testify. Further it is argued, Berry has not submitted any proof specifically directed to count 2, referring to platelet-like crystals having a specified thickness, and hence has not shown reduction to practice. While we agree with the board's holding as to certain of these objections, which were argued before it also, we shall treat them as we proceed.

With the object of making alpha-aluminum oxide fibers, Berry conducted his first experiment on September 28, 1955. The results were recorded in his notebook, in evidence as Berry exhibit 5, which was signed by Berry and by Simmons as witness. Simmons testified that he saw the experiment carried out by Berry, and saw the material produced. Berry coded a sample of the material and took it to be analyzed. The x-ray diffractions on the coded sample were run under the supervision of Biddle and Schupp and identified as alpha-aluminum oxide. The original of the report card and a copy thereof submitted to Berry and attached by him to his notebook are of record as exhibits and were identified by the witness Schupp. Young saw the sample and personally talked with Biddle to satisfy himself that it was fibrous alpha-alumina.

The process steps were repeated on September 30, 1955 and the recorded and witnessed results indicated that the original experiment was reproducible. On December 5, 1955 another attempt to repeat the original experiment, using the same materials and procedure, was made. The witnessed notebook, in evidence as Berry exhibit 17, recites of the crystals produced:

Length to 4 cm. and wide range of type such as long, fine curled fibers less than 1 μ in maxim cross sectional axis to about 4 cm. long. Also long flat & thin, tapered laths ranging in width to about 50 μ.

Two days later, another experiment produced "several dozen sparkling clear crystals rather uniformly about 1.5 cm. long and range from about 5 to 50 μ wide and apparently ribbon shaped."

An experiment run on December 28, 1955 produced a product analyzed by x-ray diffraction to be corundum and described in Berry's witnessed notebook as comprising both ribbons and fine sweeping fibers.

In connection with a December 23, 1955 experiment, Berry noted in his witnessed record: "Fibers are up to 3 mm. long and widely varying in diameter from < 1 &#956; to about 10 &#956;. Wider ones appear to be tapered laths."

Berry testified as to how measurements of the fibers were made, namely by use of a microscope with a calibrated filar eyepiece or by comparing with nylon fibers of a known size.

Simmons testified repeatedly that the notebook entries of Berry were witnessed by him and that he was present in the laboratory when the work was done, although his signatures as witness were usually affixed to the notebook sometime after the experiments were completed. At one point Simmons erroneously stated that an x-ray report dated after the date of his witnessing signature was attached to the notebook at the time he signed; however, the mistake was adequately explained on subsequent redirect examination. We agree with the board that there is no reason to doubt the testimony of Simmons with respect to his observations of the Berry experiments, and that it is consistent with the testimony of Casey, the laboratory technician working in the same laboratory with Berry and Simmons.

With respect to Flam and Forbes, we agree with the board that the record shows that their duties were mechanical in nature and routinely carried out under the supervision of Schupp, Babcock, or Biddle, all of whom testified with respect to the x-ray reports on the products produced by Berry and Balthis. We do not doubt, nor did the board, that the products were alpha-aluminum oxide.

The testimony of Balthis and Young is sufficient, in our opinion, and was so found by the board, to establish that the product produced by Berry met the requirement of being virtually perfect and possessing high strength as well as hexagonal shape.

We are in disagreement with the board, however, as to its conclusion that the extensive experiments carried out by Berry fail to show that the crystals produced conform to the dimensional limitations of count 1. It is not without-significance, we believe, that Berry carried out over a period of time a great number of experiments producing the alpha-aluminum oxide fibers. Many of the runs were repetitious of earlier experiments, thus proving reproducibility. Photographs of crystals obtained were taken under magnification.

In its analysis of the evidence, the board found the only deficiency in Berry's proofs to be an absence of adequate corroboration of his testimony and notebook entries covering the dimensions of the crystals. We do not agree that there is such a deficiency as to the proofs relating to count 1. In our review of the evidence submitted by Berry, we have kept in mind that testimony of an inventor is not by itself effective to prove reduction to practice in the absence of corroboration. However, there is no final single formula that must be followed in proving corroboration. Santelli v. MacMullen, 326 F.2d 1008, 51 CCPA 978. In recognition of the realities of technical operations in modern day laboratories, a rule of reason has been followed. Hurwitz v. Poon, 364 F.2d 878, 53 CCPA 1502. Recently, this court stated, in Beeber v. Krogh, 403 F.2d 743, 56 CCPA ___, that of greater significance are:

* * * certain more recent decisions of this court which have benefited from a larger number of prior decisions to draw from and thus are able to present a more complete and more timely analysis of the law regarding adequacy of corroboration. Those recent decisions direct us to review all of the evidence and consider it as a whole in determining whether the testimony of an inventor is corroborated.

The purpose of the rule requiring corroboration is to prevent fraud. Keeping that purpose in mind, we examine the facts of record in the present appeal. Berry's notebook is a contemporaneous record of his thoughts and actions during the performance of an organized research endeavor. What incentive, one may ask, would there be to falsify the observations regarding the dimensions of the crystals produced by the experiments? Surely, it would not be reasonable to imagine that Berry had in mind, at the time he recorded the experiments' results, this interference with its limited counts. Moreover, Berry's laboratory partner, Simmons, testified to having observed the experimental work being performed, understanding the recorded results, and witnessing the notebook records. This establishes that the notebook was in existence and properly kept at the time alleged. To require more corroboration than that set forth above would not, in our view, serve to dispel the possibility of fraud to any significantly greater extent than has already been done in this instance, and would rather appear only to lead to an unjust result.

Applying the test for adequacy of corroboration quoted from Beeber v. Krogh, supra, and having considered the entire record adduced on behalf of Berry, as well as the parties' arguments relative thereto, we conclude that Berry has proved a reduction to practice of count 1 at least by January 1957.

We are unable, however, to say the same with respect to Berry's proofs as to count 2 defining the platelets. We agree with appellee Webb that it does not appear that Berry has submitted any proof specifically directed to count 2. Berry contends that his exhibit 24, the notebook record of the December 23, 1955 experiment, in reciting "in diameter from < 1 &#956; to about 10 &#956;. Wider ones appear to be tapered laths," establishes the width of count 2. We cannot agree, for count 2 recites not a width but a thickness, measured perpendicular to the platelet surface. Berry's argument that each of his 13 runs produced perhaps hundreds of crystals varying greatly in dimension and that under such circumstances it seems impossible that crystals not having the dimension of the counts would not have been produced is hardly convincing, particularly in the absence of some evidence pertaining to the thickness of platelets.

Turning to the evidence of Webb, we note that Berry's date of reduction to practice of count 1 is earlier than any alleged by Webb and we need consider only Webb's proofs as to count 2.

The most significant evidence with respect to crystals falling within the limitation of count 2 is the testimony of Durham and his identification of certain exhibits. Durham testified that he tested crystals from the second run of April 1957 and recorded the results of the tests in his notebook, Webb exhibit 7. When asked if there were other crystals, in addition to the samples noted in his notebook, selected and tested, Durham replied in the affirmative. Asked if there were different types of crystals, he replied:

Yes, in a run of this nature there were two types of crystals present. There were the straight crystals which we observed here. There were platelets and blades and also some straight crystals which appeared to have hollow centers.

Durham then identified Webb exhibit 4 as a photograph of a "blade" which was taken by him. He further stated that he tested crystals of the type shown in Webb exhibit 4 for deformation "from a scientific curiosity standpoint," however, no official record of this test was made. The following exchange then occurred.

Q104. About how long were the crystals of the type shown in Webb et al Exhibit 4 in length? A. In recollection, I believe they were again in excess of one millimeter long.

Q105. What was the thickness of the crystals shown in Webb et al Exhibit 4 as compared, let's say, to the diameter of the crystals which you tested and the results of which you recorded in your notebook? A. The thickness would be comparable to the thickness of a single crystal.

Q106. About what would this thickness be? A. This would be in the order of three to five microns.

In addition to the above testimony regarding the dimensions of the platelets, there is in evidence as Webb exhibit 2 a research report prepared by Webb and Forgeng, dated May 24, 1957, describing the study of the crystals, showing a platelet and reciting the dimensions of platelets as being from 0.5 to 10 microns thick and up to 10 mm. in length. Hamby, a manager of patent licensing with Union Carbide Corporation, testified that he read and understood the report sometime in June 1957, prior to his writing a letter concerning it in July 1957. This letter is in evidence as Webb exhibit 1. This report is substantially identical to an article by Webb and Forgeng appearing in the December 1957 Journal of Applied Physics. Saia, the electron microscopist, additionally testified with respect to making diffraction pattern photographs and that the platelets were "extremely small and thin and flat." Forgeng testified that from the photographs given him by Saia he concluded that the material represented was a single crystal from which you could tell whether or not the material was alpha-aluminum oxide.

The board found the above evidence sufficient to prove that Webb reduced to practice the invention of count 2 prior to Berry's filing date. Applying the same test of corroboration used above in analyzing the evidence of Berry, we do also. Considered cumulatively, the testimony of Durham, Hamby and Saia adequately corroborates the documentary evidence, Webb exhibit 2. Moreover, while the "Applied Physics" article cannot be regarded as proving the truth of the contents thereof, it shows that Webb and Forgeng made the statements therein. Those statements were made before any priority controversy arose, conform substantially to Webb exhibit 2, and support the credibility of that exhibit. Hasselstrom v. McKusick, 324 F.2d 1013, 51 CCPA 1008.

Berry contends that Webb failed to show any utility for the crystals produced and first suggested that such was obvious in his reply brief before the board. The reply brief, it is argued, is not the proper place for an allegation of obviousness even where there is obviousness in fact. However, the board's response to that argument was:

We are in agreement with Webb et al. that the utility of the sapphire or alpha-alumina crystals produced by Wissler and Cormack in April would be obvious. The patents to Block, No. 2,758,011, issued August 7, 1956 (in the record of both the Webb et al. patent and the Berry application) and to Weiss, No. 2,209,908, issued July 30, 1940, (in the record of the Berry application) disclose that corundum or alpha-alumina was in 1957 known to have utility in refractory and ceramic materials and as a catalyst support. It is our opinion that it would be obvious that the alpha-alumina produced by Webb et al. would be useful for these purposes.

We have considered appellant's arguments but find ourselves in agreement with the board that the utility of these crystals was indeed obvious, if not notorious, at the time they were made by Webb. Moreover, we note that Durham, Webb's assistant, recorded in his notebook, Webb exhibit 7, that "observation and study of these crystals revealed them to possess unusual properties for refractory materials."

We conclude, therefore, for the reasons set forth above, that Berry has established a reduction to practice of count 1 prior to any date alleged by Webb, and that Webb has established a reduction to practice of count 2 prior to the filing date of Berry. Accordingly, the decision of the board is reversed as to count 1 and affirmed as to count 2.

Modified.