Skelly Oil Co. v. Corporation Commission

GIBSON, J. As a result of rumors and charges that the oil wells of the Skelly Oil Company located on what is known as the Parks College lease in the Fitts pool in Pontotoc county had unusual high potentials, the attorney for the Corporation Commission, after some investigation, instituted a proceeding before • the commission to correct the potentials .and to- adjust the “allowables” allocated to the company for operation of this lease. After a hearing the commission made numerous findings of facts against the company, and entered án order réducing the potentials on 'a'll wells producing from the ' Upper Simpson Series 29.22 per centum, and on all wells producing from the Hunton sand 28.62 per centum, making the order retroactive and requiring the company to run only 50 per centum of its monthly allowable according to the new potentials until the other 50 per centum had made up the “overage.” It was charged that the original high potentials were obtained by the use of a separate hilltop tank, ordinarily of service for water. Into this tank, it was asserted, the company, without the knowledge of the conservation officers taking the potentials, stored oil which, while the potentials were being taken, was allowed by the use of secret valves to run down hill by force of gravity into the lines leading from the well then being . tested and thus to augment the natural flow. The charge amounts to a claim of absolute fraud and deception. According to numerous statements made in the brief of the commission’s attorney, the order is punishment for this fraud and deceit. In fact, the attorney declares that the setup maintained by the company and the attitude of its employees “is absolutely fatal to conservation and proration if it were practiced by the oil industry as a whole. For this condition the Skelly Oil Company is entitled to be punished.” Furthermore, under chapter 131, S. L. 1933 (52 Okla. St. Ann. sec. 115), the employees and company might be guilty of conspiracy. These charges require careful consideration. There is an apparent agreement on the following principles of law involved. In-conducting the hearing and making the order the commission was exercising .judicial functions and could proceed only upon competent, relevant, and material testimony. H. F. Wilcox Oil & Gas Co. v. State, 162 Okla. 89, 19 P.2d 347. Fraud is never presumed; proof thereof must be clear, cogent, convincing, positive, and satisfactory, and must preponderate to the degree of overcoming all opposing evidence and repelling all presumptions of good faith. Brotherhood of Railroad Trainmen v. Brown, 180 Okla. 489, 71 P.2d 742. Where a transaction is fairly susceptible of two constructions, the one which will free it of the imputation of fraud will be adopted-; and where fraud is alleged it must be proved and cannot be inferred from facts consistent with honesty of purpose. Id. See, also, Cromwell v. Ream, 175 Okla. 408, 52 P.2d 752. It is also true that fraud may be established by circumstantial evidence, and that circumstances -altogether inconclusive, if -separately considered, .may by their number and joint operá- *366 tio.n, especially when corroborated by moral coincidences, be sufficient to constitute conclusive proof of fraud. Gordon v. Slate, 169 Okla. 399, 37 P.2d 270. The review here, therefore, resolves itself almost entirely into a discussion of the evidence. The Skelly company had been operating this lease for nearly a year and a half when it decided to discharge one of its roustabouts, Hal D. Turk, because of some violation of its rules. After his dismissal he attempted to be reinstated. Failing in this, he went before an attorney and notary public and made an affidavit which was taken to some of the companies operating offsetting-leases. At the time Turk declared: “They didn’t-treat me fair and I am going to-turn in the unreported oil. I didn’t get a fair deal.” Turk verified the application under consideration here, and was used as a witness at the hearing before the commission. The attorney for the commission here states that the evidence is sufficient if Turk’s testimony is eliminated, but that the commission had the right to believe him. The company declares that the evidence of Turk is not credible enough to establish fraud and that without his evidence there is no semblance of a case. Of course, the mere fact that Turk was a discharged employee seeking revenge does not of itself destroy his testimony, although a proper subject of inquiry as to motive touching his credibility. It frequently happens that information of misdoings of persons, firms, or corporations is not obtainable except through some disgruntled or discharged employee. The inquiry should be broad enough to cover the whole scope of his testimony and to test its reliability as weighed by all corroborating facts and circumstances or lack thereof, .or contradictions, either actual or inherent. We necessarily, therefore,, consider the facts and testimony somewhat in detail. A statement in narrative form of his testimony may serve to present the facts claimed. We state this narrative in the first person, but not as verbatim quotation. I was employed by the Skelly Oil Company in September, 1933, going to the Park College lease November 17, 1935, first as a roustabout, then a switcher. I worked until March, 1937. No. 1 well had been completed when I went there; No. 2 was a gas well. In March, 1937, 18 wells were complete. I remember the taking of potentials on -well No. 3 in January, 1936; if there was a-prior test of No. 3 I know nothing of it. Mr; Hankins, head roustabout, said that the battery tank -was run over in testing for a company test. The second hour of the potential of No. 3 was very high and Hankins said: “Xou know why that happened — that is the reason they get iir a jam, running that tank over while they were testing the well; * * * because they got in a jam on the gate they opened on the hill like they did when they were testing the well.” This was the gate at the oil tank at the water battery. I was there the first two hours on the test of No. 5. The Skelly took a test on No. 5 eight or ten days before the official potential. The well made 160 barrels the first hour and 154 the second. I was at the warehouse cleaning around the tank battery when the potential was taken. Hankins came up, went to the gate at the water tank battery, opened the gate and went down to the tank. I then went to see the gate and saw it was open. Hankins told me a day or so before he opened the gate one-fourth to make 80 barrels going down hill. Later in the day, about 3:00 or 3:30, while I was cleaning out the warehouse, middle ways between the tank battery and separator, I saw Mr. Shrier (another employee) and he appeared to be turning off thq gate. Hankins told me that he had used the gate on wells Nos. 3 and 4. After No. 8 Hankins told me there would be no more run, as Mr. Nunally was coming on. I saw Hankins turn the gate off while they were taking the potential on No. 8; that was the same gate I saw him turn off after the potential was over on No. 5. I was out back of the bunkhouse burning trash when I saw Shrier close the gate. In my affidavit I stated I helped take potential on 5 and 8, that I watched John Shrier close the gate at the tanks when potential was over, and that later on I saw Hankins open and close the secret valve on the hill as potential was started and closed. I believe I’ll change that to what I am swearing to. I am telling the truth now about it. On No. 5 Hankins opened the gate on the hill by the gravity tank and Shrier shut off the tank by closing the block gates. I did not see any valve opened on No. 8; I saw Hankins close the block valve. I was on the dike above the warehouse and saw Hankins stoop over and watched him turn the valve, 200 yards away. The quarter turn on the valve, Hankins said, would make 80 barrels of oil an hour. One time I examined the valve and found it was so turned and was running oil. Hankins said the Tulsa office didn’t know it; that he didn’t want the Tulsa office to know it. I tried to go back to work and *367 get niy overtime and they wouldn’t do it, and I said: “I will just turn them in.” As shaking the credibility of Turk, the following allegations are made, among others: (1) That his affidavit, reaffirmed in his verification of the application here, but made just a few days after his discharge, contains material statements widely at variance with his testimony. In his affidavit he declared that he helped take potentials on wells 5 and 8 and saw John Shrier close the gate at the tanks when the potential was over and saw Hank-ins open and close the secret valve on the hill at the gravity tank. In his testimony he declared more than once that he saw Shrier close the block gate (approximately 190 yards farther down the hill). As to No. 8 he declared he did not see any gate opened, but saw Hankins close the block valve. In his affidavit he asserted that he was watching Shrier’s actions from the warehouse, but several times in his testimony he declared that he was at the trash burner back of the bunkhouse (some distance from the warehouse). In his affidavit he declared that the company had constructed a secret valve gate at the main battery of tanks and two others on the main lead lines from the gravity tanks to the main battery. In his testimony he stated that he did not class the valve at the main battery as a secret gate; that he didn’t know what it was used for until he was fired; that he merely imagined it was on the line. In his affidavit he stated he helped take potentials on Nos. 5 and 8. In his testimony he plainly shows he was not helping take potentials, but was performing other tasks. (2) A civil engineer who took observations proved it would have been impossible for Turk from the'position given to see what Shrier was doing at the time he claims to have seen Shrier manipulate the valve. (3) That his testimony contains contradictions which make it inherently improbable, such as his claim that the private test on No. 3 ran over and that he knew nothing of any private test on No. 3; and placing himself in three different positions widely apart when the No. 5 potential was taken. (4) Other statements are improbable when considered with conceded facts; for example, the' statement that the hill tank gate was opened some 25 or 30 minutes before the potential started. Since the gate at the test tank was opened before the well was turned in — as customary — the oil would have been flowing from the hill tank into the test tank before the well was turned in, a circumstance making detection swift and sure. (5) Other various contradictions and improbabilities are urged, in more or less material statements, such as the statement he was discharged because he knew too much. (6) His motive to get even for being discharged — announced by himself. (7) Impeachment of his reputation for truth and veracity. (8) Contradictions by other witnesses of statements and actions attributed to them. Attorneys for the commission assert that, independently of Turk’s testimony, the other evidence, facts, and circumstances sufficiently prove the fraudulent acts or padding, and that, of course, these other facts corroborate Turk, whose credibility, they urge, was not successfully attacked. First, attention is called to the physical setup which would lend itself readily to padding. The location of the tanks on the hill, from which by force of gravity any oil stored could run into the oil tanks at the battery, if valves were properly opened, and could have been used to supplement the oil when potentials were taken. The company does not deny the use of one of these hill tanks for both water and oil, but asserts that the oil was used • for the purpose of washing in wells. The wells’ logs filed with the commission show numbers of times wells were washed in with oil. “Fraud will not be inferred from proof of the mere opportunity to commit it.” Black v. Post (W. Va.) 67 S. E. 1072. Turk admitted that on several occasions the hill tank had to his knowledge been used for oil for acidizing or washing in. That it was not used for padding potentials is a reasonable conclusion from other facts considered with the testimony of Mr. Thomas, a conservation officer who took some of the potentials. He testified that when potentials were taken on wells Nos. 8, 9, 10, particularly as to 9 and 10, the hill tanks were checked and on one occasion were found to contain only water and on another water in one and a foot of oil in the other. He further testified positively that potentials on Nos. 9 and 10 could not have been padded. In view of the testimony of Mr. Thomas, an employee of the commission and a witness for the attorney, that these wells were' not padded, we cannot accept the contention of the attorney for the commission that Nos. *368 20 and 21 are the only wells known not to be padded.. A consideration of Nos. 20 and 21, in reference to other facts and claims, however, may be' helpful in showing the application of the facts in. relation to- Nos. 9 and 10. Nos. 9 and 10 were in the group of wells belonging to one horizon, and Nos. ‡0' and 21 in another. The statement as to Nos. 20 and 21 is baseij upon the fact that the officers of the commission, wishing to run down rumors, took great precautions to see that these last two wells should not be padded. The potentials on these were respectively as follows for four-hour, periods: No. 20—in barrels per hour—167.96, 127.28, 120.72, 57.67; No. 21—194.28, 158.79, 141.73, 97.11. Based upon these potentials and upon certain potentials in other parts of the field, it is urged that there Should have been in all wells a rapid decline in production hour by hour instead of a somewhat steady flow as appeared in other wells.. And it is urged that these wells are more in line with production from offset wells. ' The commission’s finding No. 13 compares these wells with Nos. 18 and 19, where the potentials were respectively: No. 18 — in barrels per hour — 147, 147, 144, 150; No. 19— in barrels per hour — 220, 213, 191, 177. It will be noticed at once that the first hour’s flow of No. 18 is considerably lower than the 'first hour’s flow of either 20 or 21, while the first hour’s flow of No. 19 is considerably higher than either 20 or 21, while the combined total of the first hour of No. 18 and No. 19 ' is 367 barrels as compared to 362 barrels for Nos. 20 and 21, a difference of only five barrels. It is evident at once that any claim that the first hour of either 18 or 19 had been padded by an additional 80 barrels from, the hill tank is absurd. The’ argument is that the flow in 18 and 19 is too uniform and not at all in keeping with the decline as shown by other wells. Admitting that, we do not find that it supports the testimony of Turk or proves padding. It is evident that if there should have been ■R decline from hour to hour in production, .the addition of 80 barrels per hour or any ■other constant quantity would not have made rthe production per hour equal, but would ’have left the angle of. decline the same. To .have. produced uniformly the Skelly em-;plo.yees would have had to guess pretty accurately, in advance the -number of barrels ■a well on potential would produce the next succeeding hour, and then have had to know pretty accurately how much to turn •the valve to bring that hour’s- production, up to1 a parity, and would have -had. to operate the valve for that purpose every hour during the taking of the potential. We do not think it either feasible that the guess could have been made or possible that the task could have been carried' out over such a period of time without detection. An. illustration here may be apt. We assume that if No. 20 had not been watched, padding would have resulted. If we add 80 barrels to each hour, we have the hourly production as follows: 248, 205, 198, 136. There is still the discrepancy of 112 barrels between the first and last hour — the same angle of decline. If, however, we begin with the first hour as shown and seek to produce uniformity, we find that to the second hour we must add not 80, but 43 barrels, to the third hour 50 barrels, and to the fourth hour 112 barrels. But we find no evidence or hint that the Skelly either had such good guessing employees or opportunity to manipulate the gates hour by hour so as to produce increasing flow. On the contrary, the claim is based on the theory that the valve was open once for all with a substantially even flow hour by hour during the taking of the potential. It is evident, therefore, that the comparative uniformity of production from some of the wells refutes the whole claim that the fraud was perpetrated by padding 80 barrels per hour, or any other constant quantity, if the premise is true that there should have been a rather rapid decline hour by hour. Nos. 20 and 21 were in the Hunton strata. But Nos. 9 and 10 were in the Upper Simpson, the other horizon under consideration here. An application of the 80 barrels per hour theory to them would show the same inferences as heretofore illustrated. Potentials of these for the 6-hour period ran respectively as follows: No. 9 — in barrels per hour — 253.28, 167.92, 180.78, 173.21, 150.96, 174.59 ; No. 10 — in barrels per hour — 217.53, 200.43, 185.07, 165.30, 161.49, 154.48. If, as above, we compare 9 or. 10 with No. 8 or No. 5, particularly referred to by Turk, we find the theory of padding by uniform addition of 80 barrels per hour insupportable. No. 5 is shown of relative uniformity, the second hour showing a slight decrease over the first, the third an increase over the second, the fourth an increase over the third, the fifth an increase over the fourth, the sixth a drop to slightly below the' first. If 80 barrels be subtracted from each hour, the same results would show as to incline or decline.' Here, instead of a decline, we get an incline. The production would show as much uniformity after the. deduction as before. Furthermore, if the decline in these wells *369 should have been from hour to hour, then No. 3 should, have shown a decline of the fourth hour from the third instead of an incline; and a much larger decline of the fifth oyer the third instead of the practically even status actually shown. The first hour of No. 9 shows greater production than of any other well, except No. 17. Here should he mentioned No. 3 in reference to the showing of the first and second hours’ production. The official potential as filed shows 188.9 barrels for the first hour, 314.8 for the second hour, 220.6 for the third, 231.04 for the fourth, 220.4 for the fifth, and 202 for the sixth. Turk’s explanation of this is that Hankins told him the secret valve went “haywire” and ran the production up for the second hour, and attempted by his quotation of' Hankins to draw a parallel with running a tank over on the company’s test. We fail to find any rational explanation of such alleged parallel, and find no intimation of how the valve could have adjusted itself so accurately at the end of the second hour as to give no more trouble. On the other hand, we find that the conservation officer at the time noticed the great discrepancy and inquired about it. The offset witness, the conservation officer, and the Skelly employee reached the conclusion that by error in gauging, one foot of oil was given to the second hour that should have been given to the first. The result was that 63 barrels were taken from the first hour and given to the second. An adjustment would thus give the first hour 252 barrels, the second 252, the third 220, and so on. A comparison of the wells in the Upper Simpson horizon discloses that as to wells Nos. 1, 3, 6, 7, and 10, the decline of the last of the six-hour potential over the first ranges from 48 barrels in No. 1 to 65 barrels in No. 7, the latter being three barrels more than No. 10, and the decline of the last hour on No. 9 over the first hour is 81 barrels, which is the same as in No. 11. It follows that there is nothing to be deduced detrimental to Skelly by comparing these other wells with 9 and 10, whose potentials under the testimony indisputably were not padded. Wells 4, ■ 5, and 8 are practically without decline. In addition to what has heretofore been said as to the inferences to be drawn from these circumstances, we find that in the offsetting well, known as Fleetborn Wooley No. 1, the difference between the last hour and the first is 10.39 barrels, which is comparable. to the 9.3 barrels difference’ between the first and last hours of No. 4. We find also-this Fleetborn well shows a decline of the second hour from the first, an incline of the third over the second, of the fourth over the third, of the fifth over the second and third, and of the sixth over each other hour except the first. This is very like the performance of No. 8 and of No. 5 of the Skelly. Furthermore, the attorney for the commission introduced records of potentials on offsets, which included three Fleetborn wells in the Upper Simpson horizon whose average potential per hour was as follows: First hour, 214 barrels; second hour, 199 barrels; third hour, 194 barrels; fourth hour, 188 barrels; fifth hour, 195 barrels; sixth hour, 174 barrels; total average 24-hour potential, .4,523. Compare this with the average shown by the same evidence for all the Skelly wells, by hours: 216, 208, 194, 191, 184, 177 barrels, with total 24-hour average potential of 4,479 barrels. There is no claim that the Fleetborn wells were padded or fraudulent and no inference of fraud is claimed, either because of their large potentials, their failure to decline rapidly, or their failure to decline hour by hour in all cases, or from the fact that their first hour’s production is large, or that they show less hourly decline and have a less difference between the first and last hours than other wells in the field, more or less offsetting the Skelly lease. If inference of fraud or falsity is not proper in the one case, it is not proper in the other. The average for seven more or less offsetting Magnolia company wells is given as 164, 141, 137, 130, 118, 121 barrels on hour-by-hour measurement, a decline of 43 barrels in the sixth from the first hour, a difference exceeded in amount by Skelly wells numbered 1, 3, 6, 7, 9, 10 and 11. A comparison of three Stahl company wells shows a difference in average of those wells in the decline from the first to sixth hour of 68 barrels. Compare this with Skelly’s wells numbered 10, 7, 9, 11, where the decline of each was respectively. 62, 65, 81, and 81 barrels. From these circumstances it is impossible to conclude that the Skelly wells did not show sufficient decline. It is further evident from these logs that there is no uniformity of declination of production sufficient to give rise to the theory that the Upper Simpson horizon at the time was of such uniformity of geology, pqrosity, and permeability that it should follow as a necessary consequence that any particular well or wells should have the same relative log or declination of production as might be common to any other particular well or group of wells within the same horizon. The differ *370 ence in the production of the sixth hour of the tests from the .first hour in these so-called offsets varied from 19 barrels (Orosbie) to 117 barrels (Amerada). The oral testimony by which it was sought to establish this premise also failed. Mr. Thomas, conservation officer, recalled to testify as to the usual decline hour by hour of the production, declared: “There is no uniform rate for the time * * * There is nothing uniform about it.” Mr. Dave Logan, another conservation officer and expert geologist and oil man, was called to substantiate the assumed premise. The commission had ashed him to examine the fields to determine if there was reason structurally for the 'wells on the Park College lease to be larger than the “average wells on offsets,” and to see if new potentials would throw any light on the question. The taking of new potentials, of course, would throw light only if there was some uniformity of structure. He testified he could see no particular reason why the wells on the Park College should be higher. But he said there were differences that could not be accounted for. “Wells on the same structural strike will vary sometimes 20, 30, or sometimes 40 per cent, where they are completed exactly the same, due to the natural sand body itself,” he said. “You can’t account for it (wide variations in potentials) for any particular reason. Better methods of completion get better potentials. I don’t argue that point.” Previously he had also declared that among factors controlling relative production could be mentioned “the position of the wells on the structure, the time and method of completion of the wells, the sand conditions, how deep drilled into the sand, the size of the drill hole, the use of acid, or washing with oil instead of water ” This court has recognized the fact that no hard and fast rule can be laid down as a premise to determine what production should be. We have said: “By the very nature of oil wells they differ not only in their quantities of production, but in their peculiar qualities. A well of small capacity may be drilled in close proximity to a well of great capacity, depending among other things upon the location of the well on the structure, the porosity of the sand, and many other operative factors.” H. F. Wilcox Oil & Gas Co. v. Bond, 173 Okla. 348, 351, 48 P.2d 820, 823. This general statement of fact is further borne out by conditions in this field. An original potential of Skelly No. 1 was taken December 3, 1935', and was 3,874 barrels; a second potential January 10, 1936, was 4,-355 barrels. Its immediate offset on the south, Fleetbom Wooley No. 1, on potential taken December 8, 1935, showed 4,463 barrels, and on the potential taken January 9, 1936, 3,-789 barrels. On well No. 12 the Skelly potential was 2,988 barrels; its east offset, Magnolia Wooley No. 3, was 3,889, and its north offset, Magnolia Norris No. 4, was 3,633 barrels. In these latter cases the offsetting wells were not only larger than the Skelly well, but in one ease fully 30 per cent, larger. Other comparisons not only with the 'Skelly wells but as to wells on adjoining leases with other wells on other adjacent leases as well as one lease against another were made to the same effect. These disproved any uniformity of production even in the immediate locations. We must conclude, therefore, that we find no basis for an inference of fraud arising from any assumed premise that because of high potentials we must assume padding. As pointed out, potentials varied as to the same well if taken at different times, and no reasonably acceptable ratio of production or decline could be established. In this view we see no basis for presuming guilt from the fact that the Skelly officials did not accede to the request of the commission to retake potentials long after the original potentials were taken, and after Turk had furnished his affidavit. It does not appear that potentials were to be taken as to other wells in the field. Probably other lessees would have declined also. It is asserted that the conservation officers had no knowledge that oil was used in the hilltop tank and that the Skelly employees did not tell them. It is also asserted that persons from surrounding leases did not know of this. Even from Turk’s testimony, however, it appears that the hilltop tank was used frequently for washing in with oil, a process in which he declares he often participated. Three of the seven deputy conservation officers who took potentials testified. Fred Bowles was not asked as to his knowledge of the use of these tanks, but testified that Plankins told him they had a good reason for it; that they loaded the wells with oil by pumping them full of oil before taking the potentials. We do not find any claim made here that such was the method of padding, even were it conceivably possibly by use of a 500-or 600-pound-pressure pump to fill a hole when the pressure at the bottom was probably many times that amount. Perhaps-Hankins should not have given such an equivocal answer, but it may also be true-that he did not consider he had the right to speak, if as claimed the Skelly believed it was using better and more efficient methods-to obtain its production. Hankins testified later that, by loading he was referring to the- *371 pumping of oil into the wells in connection with the acidizing process and that he did not feel at liberty without permission to divulge details. The Skelly, it may be said here, contends that its acidizing processes and methods used were superior to those used by other operators in the field. Much testimony is in the record touching upon this claim. Operators from other .fields testified as to their methods, and on the whole much evidence was offered to dispute the Skelly’s claim. It cannot be doubted, however, that (probably like each other operator for himself) the Skelly believed it had better methods. Some were shown to be different. Hankins, without permission from his superiors, did not need to disclose any legitimate methods they may have been using. Hugh Thomas, another officer, testified that up until taking potentials on Nos. 9 and 10. he did not know that oil was kept in the hill tank. It is inferred from his testimony that Skelly employees took pains to conceal this fact from him. It is urged that when he inquired, while taking potentials, if there were other sources from which oil could come, he was told there were none. A negative answer here did not necessarily indicate bad faith. The question at each time undoubtedly had reference to the immediate potential, and if, as claimed, no valves were open, the answer would naturally be in the negative. An honest man would not necessarily feel called upon to go into detailed explanation of how he might be dishonest if he desired, just because he is asked if at the moment there is any reason why his honesty should not be assumed. According to Turk’s testimony, supported by the records, the washing in with oil was frequent enough to have been discovered by anyone on the premises. Mr. Thomas was on the premises for a while during one of the washings in and observed that there was no tank at the well anfi that the oil must be coming from some other source. Evidently there was no concealment here. It is asserted that there were three secret valves. In Turk’s affidavit he so declared. On the stand, however, he did not claim the one at the tanks as secret. The one on the hilltop seems to have been in a freeze-box. not unlike other valves. This was because the tank sometimes held water as well as oil. One of the gates was on its. side, but still sticking up. Turk did not claim that it had been in this condition all the time, and there was evidence that it had been put into that position by an employee of a pipe laying company to escape possible injury from a tractor. The finding of the commission that Mr. Shrier told Earl Shipman in substance that the purpose of the oil on top of the hill was to aid in getting potentials pg unwarranted. We do not so necessarily interpret his testimony. He says that, after Shrier told him the lease was a two-battery lease, the following occurred: 'Shrier “didn’t say a lot right then. We just walked along. He said, ‘You know we have to take potentials down here, don’t you?’ I said ‘Yes.’ He said, ‘We have got about the best potentials that is in the Eitts area.’ ” In view of the witness’ statement that they talked along after discussing the battery, it is mere speculation to connect the subsequent statements about the potentials with the battery. In inferring fraud the inference must be more direct. Here two somewhat remote bits of conversation. must by some inference be put together to produce an inference from which fraud may be inferred. The rule that fraud may be inferred from circumstances does not mean that mere numbers of circumstances, innocent or otherwise, by the weight of their number establish fraud. No number of innocent circumstances combining can establish fraud. Of course, fraudulent circumstances may combine with innocent, and it is from the entire combination, from the effect of the whole, that the inference must at last be drawn. But there must be sufficient circumstances from which the more reasonable inference is fraudulent rather than honest before a general deduction may be drawn. “ ‘In determining the existence of fraud, any evidence, direct or circumstantial, which is competent by other rules of law, and. has a tendency to pirove or disprove such issue, is admissible.’ * * * “ ‘When fraud is alleged great latitude of proof is allowed, and every fact or circumstance from which a legal inference of fraud may be drawn is admissible.’ ” Berry v. Stevens, 168 Okla. 124, 129, 81 P.2d 950, 955. It will be noticed that the circumstances must have a tendency to prove fraud and must be such from which legal inference of fraud may be drawn. If we apply these principles together with the rules announced at the first of this opinion, we must conclude that ihe circumstances relied on are totally insufficient to establish fraud. Eew, if any, are inconsistent with good faith, and a presumption of honesty must prevail over mere suspicion. Turk, it is true, did testify as to some undisputed facts and admitted -physical conditions. But so did witnesses for the company. Such facts furnish no greater corroboration for him. than for the witnesses for the com *372 pany who disputed his testimony. But without regard to the testimony of these contradictory witnesses, as we have shown, the theory of his testimony cannot be harmonized with the conceded facts as to the potentials taken on the Skelly and surrounding leases. Furthermore, it has highly material contradictions within itself, and is irreconcilable in large part with previous sworn statements made by him, so much so that he, in effect, declared that he had sworn falsely before. It would serve no useful purpose to show further how witnesses for the attorney and others contradicted him. We cannot hold that the evidence measures up to that standard of proof hereinabove referred to upon which such drastic penalty may be inflicted. We have not deemed it necessary to discuss the contention of the Skelly company that its methods of completion were superior to its neighbors. Whatever may be the merits of that contention it is clear from the logs of the wells that Skelly used great care in its drilling, care and operation of the wells obtained by it. This care may have resulted in better production. AVe do not need to decide this question. We have great respect for the findings of the commission and for its efforts to enforce the proration regulations. We cannot accede to the suggestions of the attorney for the commission that an oil company should be punished because the attitude of some of its employees toward the conservation department is not what it should be. AVe merely hold in this case that the evidence is not sufficient to justify the conclusions and order of the commission. The order of the commission is reversed, with directions to vacate said order. O'SBORN, C. J., BAYLESS, V. O. J., and WELCH, I-IURST, and DAVISON, JJ., concur. RILEY, PHELPS, and CORN, JJ., dissent.