Delsanter v. Commissioner

Anthony Delsanter, et al., * Petitioners, v. Commissioner of Internal Revenue, Respondent Delsanter v. Commissioner Docket Nos. 49167, 52463, 52515, 52557 United States Tax Court 28 T.C. 845; 1957 U.S. Tax Ct. LEXIS 134; July 18, 1957, Filed *134 Decisions will be entered under Rule 50. 1. Partnership profits from a gambling casino determined in the light of the applicable burden of proof and evidence in the record. 2. Petitioners are liable for additions to tax for underestimation for 1948 and 1949 pursuant to section 294 (d) (2), I. R. C. 1939, and for failure to file declarations for 1949 pursuant to section 294 (d) (1) (A). However, they are not liable for additions to tax for 1948 under section 294 (d) (1) (A) by reason of having filed "zero" declarations for that year. 3. Depreciation deductions on slot machines denied for failure of proof; deduction for loss resulting from confiscation and destruction of slot machines by Ohio public officials denied. Michael E. Cozza, Esq., and Arlene B. Steuer, Esq., for petitioners in Docket Nos. 49167 and 52515. Protagoras D. Maktos, Esq., and John J. O'Brien, Esq., for petitioners in Docket No. 52557. Russell G. Mock, Esq., for petitioners in Docket No. 52463. James F. Kennedy, Jr., Esq., and R. G. Daha, Esq., for the respondent. Raum, Judge. Murdock, J., dissenting. Pierce, J., agrees with this dissent. Bruce, J., dissenting. RAUM *846 Respondent determined deficiencies in the income *135 tax of petitioners and additions to tax as follows: Additions to tax Year Deficiencies Sec. 294 (d) Sec. 294 (d) (1) (A) (2) Anthony Delsanter [1948] $ 17,360.01 $ 2,020.53 $ 1,212.32 [1949] 8,074.28 991.85 595.11 Estate of Mary Catherine [1948] 47,273.44 5,566.27 3,339.76 Tobin, et al [1949] 16,546.32 2,005.53 1,203.32 John Farah and Shamis Farah [1948] 86,892.10 10,450.50 6,270.32 [1949] 43,132.44 5,223.40 3,134.03 Ralph Coletto and Tessie [1948] 47,321.28 5,605.87 3,363.52 Coletto [1949] 29,780.36 3,669.38 2,201.63 The full amounts of the deficiencies and additions to tax are in dispute. The principal issue involves the determination of the taxable income of a gambling partnership for 1948 and 1949. Also at issue is whether petitioners are liable for additions to tax for failure to file declarations of estimated tax, and for substantial underestimation of estimated tax for 1948 and 1949. An additional issue presented only in Docket No. 52515 is whether petitioners John and Shamis Farah are entitled to deductions for depreciation of slot machines in 1948 and 1949, and for a loss brought about by the confiscation and destruction of those machines by officials of the State of Ohio in 1949. As a result of the deductions thus claimed, these *136 petitioners seek a refund for 1948 and 1949. FINDINGS OF FACT. Some of the facts have been stipulated and are so found. Petitioners John Farah (hereinafter referred to as Farah) and Shamis Farah are husband and wife and residents of Cleveland, Ohio. They filed joint returns for the years 1948 and 1949. Petitioners Ralph Coletto (hereinafter referred to as Coletto) and Tessie Coletto are husband and wife and residents of Cleveland, Ohio. They filed joint returns for the years 1948 and 1949. Edward F. Tobin (hereinafter referred to as Tobin), a resident of Warren, Ohio, is a petitioner for himself in his individual capacity *847 and also as administrator of the estate of his deceased wife, Mary Catherine Tobin. Tobin and his wife filed joint returns for the years 1948 and 1949. Petitioner Anthony Delsanter (hereinafter referred to as Delsanter) is a resident of Warren, Ohio, and filed income tax returns for each of the years 1948 and 1949. All of the above returns were filed with the collector of internal revenue for the eighteenth district of Ohio. In 1947 a partnership known as the Jungle Novelty Company was formed for the purpose of operating a gambling casino called the Jungle Inn (sometimes *137 hereinafter referred to as the Jungle) near Youngstown, Ohio. During 1948 and 1949 the members of the partnership and their respective interests therein were as follows: [1949] (per cent) (per cent) Farah [40] Tobin [20] Coletto [30] Delsanter [10] The Jungle was operated for 45 weeks during 1948, and from January 1, 1949, until August 12, 1949. It was raided by officials of the State of Ohio around 9 p. m. but not later than 9:30 p. m. on Friday, August 12, 1949, and has been closed since that time. Partnership activities conducted on the premises of the Jungle during 1948 and 1949 consisted of the operation of a "horsebook" every afternoon except Sunday, and the operation every evening of dice games, slot machines, poker games, roulette, chuck-a-luck, and bingo games. Poker and dice games were also occasionally conducted in the afternoons. The number of patrons and the amount of gambling at the Jungle increased on a daily basis on Fridays, Saturdays, and Sundays. Patrons came to the Jungle almost exclusively by automobile or taxicab, and there was an increase in patronage upon the cessation of inclement winter weather. The amounts won by the partnership increased when patronage and gambling *138 increased. All of the operations at the Jungle, with the exception of bingo, were carried on by the partnership for and at a profit. The partnership commenced operation with a bankroll of $ 25,000 in cash. This bankroll was under the supervision of Farah. Amounts were taken therefrom and allocated as bankrolls to the separate gambling operations. The employees of the Jungle were paid fixed salaries. *848 Horsebook. Three men to take bets and fill out betting slips, a cashier, a slip sorter, and two supervisors were employed in this operation. A $ 2,000 bankroll was allocated to the horsebook. The horsebook did not accept bets placed by telephone. A patron of the horsebook would place his bet with one of the bet takers. This employee would receive the amount of the bet, make out a slip in duplicate, keep the original, and deliver the duplicate to the person making the bet. Each bet taker used slips of a particular color. In general, a slip contained a serial number, a number identifying the horse involved, and information relative to the amount and type of the bet. No information relating to the person placing the bet appeared on the slip. The Jungle paid track odds up to certain fixed *139 maximums. At the start of the period at issue the maximum odds were 15, 6, and 3 to 1 on win, place, and show respectively. Sometime during the period in controversy the odds were raised to 20, 8, and 4 to 1. Most bets ranged between $ 2 and $ 5. However, there were occasional bets in excess of these amounts. If a bettor won he would take his slip to the cashier for payment. Not all bettors claimed their winnings. Unclaimed winnings were kept separately for 30 days and then added to the thirtieth day's profits. For most of the period at issue the Jungle subscribed to a wire service at a cost of $ 225 per week. The Jungle made no payments for wire service in May, June, and July, 1949. The Jungle did not accept bets on horses running at every track in the United States, but confined its operations to certain major tracks. At the close of a day's operations the total bets taken by each writer, the losses paid out as a result of those bets, and the net winnings, if any, of the whole operation were computed on an adding machine which recorded those figures on a paper tape. Every evening the tape and the day's winnings were turned over to one of the petitioners -- usually Farah. Farah *140 would replenish the horsebook bankroll in the event of a net loss for the day. Betting slips were held 30 days and then destroyed. The average total amount bet at the Jungle horsebook on a weekday during 1948 and 1949 was $ 2,500. The average total amount bet on a Saturday during that period was $ 3,500. The partnership retained as winnings 12 per cent of the total amount bet. Dice. There were three dice tables in the Jungle. Each table was staffed by five employees (four employees and a relief man), and each employee received a salary of $ 20 per night. The Jungle's total weekly salary expense for running the dice tables was $ 2,100. *849 The bankroll for the dice operation was $ 5,000. In the course of an evening tables were opened successively as the number of people playing dice increased. Each table was opened with a bankroll of $ 1,250. It was usual for two tables to be opened each evening. The third table was used infrequently. The odds used were the same as those used in similar operations throughout the country. At the end of nightly operations the money collected from each table was totaled by two or more of the partners. Any amount over $ 1,250 was considered a table's daily *141 win. Amounts missing from that amount were treated as losses. The Jungle's dice operation had average weekly net winnings of $ 5,600. Slot machines. There were 100 slot machines in the Jungle, of which 39 were nickel machines, 24 were dime machines, 31 were quarter machines, 5 were half-dollar machines, and 1 machine was for silver dollars. The machines were the "standard" models turned out by the manufacturers, and were not built to specifications supplied by the Jungle. The bankroll of the slot machine operation was $ 5,000 and was used primarily for making change. Each evening an average of $ 1,080 was given to Jungle employees who made change for patrons. There were four such employees at the beginning of the period at issue, and they were given $ 270 each. Sometime during that period the number of change makers was increased to seven, but the $ 1,080 figure remained constant. At the end of their night's work it was not unusual for change makers to turn back some coins to their supervisor. Each slot machine had four compartments in which coins accumulated -- namely, a tube, two jackpots, and a collection box. Coins would not drop into the collection box until the other compartments *142 were filled. Money in the collection boxes was collected at the end of operations on Thursdays and Sundays. The men who made the collections added the coins and recorded the total on a slip of paper. The coins and the slip of paper were turned over to one of the partners. Coins were not collected from the tubes and the jackpots. These compartments were constructed to hold the following amounts of money: 1 Denomination Double Tube Total jackpot $ 0.05 $ 6 $ 7 $ 13 .10 [50] .25 [90] .50 [145] *850 The dollar machine guaranteed a $ 500 payment to anyone who "hit the jackpot." A collection was made at the close of the night's operations on Thursday, August 11, 1949. No coins were put into any of the machines after 9:30 p. m., Friday, August 12, 1949. At that time the following amounts were contained in the various machines: Nickel Machines. Make Amount Mills $ 20.65 22.20 13.75 15.55 12.80 13.95 15.45 14.55 16.20 23.65 21.00 18.15 13.30 20.60 21.75 16.90 11.35 20.45 14.75 20.70 Mills $ 14.65 12.65 14.10 15.90 25.10 17.40 22.00 23.55 19.95 13.45 22.65 16.50 26.50 Jennings 35.55 8.75 6.15 17.95 26.55 27.55 Dime Machines. Make Amount Mills $ 58.10 32.90 42.30 39.70 32.10 22.20 34.30 35.00 46.60 35.80 22.30 29.80 Mills $ 55.30 43.60 39.30 56.00 45.80 54.70 60.50 Jennings 62.90 43.60 49.90 79.80 46.70 *143 *851 Quarter Machines. Make Amount Mills $ 98.25 31.25 48.00 67.00 56.50 69.50 56.25 65.25 63.25 49.00 68.00 41.00 47.25 61.50 35.50 27.75 27.25 Mills $ 50.75 32.25 22.75 41.75 64.75 56.25 64.00 66.25 31.75 65.50 Jennings 100.00 107.50 108.75 72.00 Half-Dollar Machines. Make Amount Jennings $ 111.00 413.00 453.00 Mills 97.50 72.00 One-Dollar Machine. Make Amount Monte Carlo $ 407 In addition, coins totaling $ 72 were contained in small glass windows set in the front of the machines. A total of $ 926.70 was in the collection boxes at 9:30 p. m., Friday, August 12, 1949. Records prepared by the Jungle's attorney-bookkeeper (hereinafter referred to as the bookkeeper) on the basis of figures supplied him by the partnership show the following results for gambling operations at the Jungle on all of the Wednesdays and Thursdays in February, March, and April 1949: Wednesday Thursday Date Win Lose Date Win Lose Feb. 2 $ 1,482 Feb. 3 $ 4,839 [9] $ 1,413 [10] 4,611 [17] 4,472 [24] 3,634 Mar. 2 1,275 Mar. 3 4,505 [10] 4,472 [16] 1,453 [17] 4,615 [23] 1,746 [24] 4,378 [31] 4,511 Apr. 6 [293] Apr. 7 4,819 [13] 1,519 [14] 4,238 [20] 1,840 [21] 4,110 [28] 4,315 *852 The Jungle averaged $ 1,000 per night as net winnings from its slot machine operations. Poker. There were two poker tables in the Jungle. A poker table *144 was operated by a dealer who was paid $ 100 per week. A dealer did not take part in the games, but dealt the cards to the participants. A dealer was furnished a $ 100 bankroll to be used for making change. One of the main tasks of a dealer was to deduct 5 per cent of each "pot." This amount represented the share of the total amount bet which went to the Jungle. A dealer would deduct the 5 per cent and place it in a locked cash box which was constructed with a small opening for that purpose. At the end of a night's operations a dealer never had the opportunity to count the full amount deducted. When the poker games ended for the night the cash box was turned over to one of the partners. Poker was played on an average of not less than 4 hours per night, and there were occasional games in the daytime. The Jungle's 5 per cent averaged $ 20 per hour. The weekly amount collected from poker games by the Jungle averaged $ 600. Roulette. One roulette wheel was operated in the Jungle by an employee who was paid a salary of $ 140 per week. The odds in favor of the Jungle in this operation were strong. The operator was not furnished a bankroll, but called for money if and when he needed *145 it. An average of 8 people per night gambled at roulette for an average of 3 hours per night. At the end of a night's operations the employee turned all winnings over to one of the partners. The employee kept no record of the amount won by the Jungle. The Jungle had net winnings from roulette of not less than $ 140 per week. Chuck-a-luck. This game was operated by three employees. The employee in charge was paid $ 100 per week, and his assistants were each paid $ 75 per week. A $ 100 bankroll was allocated to this operation. Amounts won by the house were turned over to one of the partners at the end of each night's operations. The three employees did not total or record the amounts won by the partnership. The Jungle had average net winnings of not less than $ 250 per week from this game during the years at issue. Bingo. This game was run by the partnership primarily to attract large numbers of people to the Jungle. On Fridays, Saturdays, and Sundays an average of 600 people per night played bingo. On other nights the average number of people playing bingo was 300. *853 The first game started at approximately 9 p. m. Twenty games were played from that time until 9:30 p. m. There was then *146 a 45-minute intermission which was followed by the final game. The prize for winning the last game was substantially larger than the prize for winning any other single bingo game. A $ 50 bankroll was allocated to this operation. The bankroll was used primarily for making change in the sale of bingo cards. The bingo game was operated by the partnership at a loss which averaged $ 1,000 a week. Gambling activities at the Jungle, excluding the horsebook and occasional afternoon poker and dice games, started each evening between 7:30 and 8 p. m., and ended at approximately 2 a. m. Occasionally activities would not terminate until 4 a. m. At the end of a night's operations two or more of the partners counted the winnings, if any, from each gambling activity. Then, using an adding machine that recorded these figures on a paper tape, they calculated the total winnings, or losses, for the day, which would be added to or subtracted from the bankroll as of the close of the preceding day. On Thursdays and Sundays the amount of the slot machine collections was added to the night's totals. On occasion the poker operation would not terminate until an unusually late hour in which event the night's "take" *147 would be added to the totals of the following night. A copy of the tape showing the net winnings or losses for the day and the new bankroll figure was prepared for each partner, who would then destroy his copy of the tape for the preceding day. In addition one of the partners made out a slip in pencil which was left for the bookkeeper. This slip contained a date and a single figure prefaced by "won" or "lose." This figure was accepted by the bookkeeper as representing the amount of money won or lost by the partnership on the date appearing on the slip. The partnership's bookkeeper visited the Jungle several times a week to pick up the slips left for him. He did not attempt to verify the figure appearing on the slips. Using these slips the bookkeeper prepared the "books," a so-called ledger, which he used in filling out the tax returns of the partnership. The same bookkeeper also filled out the tax returns of the petitioners. In addition the bookkeeper was furnished receipts evidencing other expenses of the partnership. He also kept a detailed record of the payroll. Respondent does not question the correctness of the amounts recorded for these expenses and wages. After making his entries *148 in the ledger, the bookkeeper returned all slips and receipts to the partners. *854 The ledger did not contain any reference to the original cash bankroll of the partnership. Nor did it contain any reference to winnings distributed to the partners or retained by the partnership. The ledger was compiled in the foregoing manner in 1948 and 1949. The ledger for 1948 was, during 1949, kept at the Jungle and was misplaced or destroyed by the officials who conducted the August 12, 1949, raid. The bookkeeper used the single daily figure supplied to him to arrive at two monthly figures. One was the total of all the daily "win" figures, and the other was the total of all the daily "lose" figures. The following figures are the monthly totals of win and loss entries appearing in the ledger for the months of 1949: Win Lose January $ 63,820 $ 9,985 February 52,793 8,158 March 61,566 8,255 April 52,460 8,991 May 57,020 9,092 June 56,530 9,846 July 60,937 10,882 August 19,640 4,299 The ledger of the partnership activities did not correctly set forth the amounts won by the partnership during the years in question. Respondent used the following methods to compute the partnership income from the various gambling operations: Horsebook. *149 From information received from former Jungle employees respondent concluded that total bets placed averaged $ 3,000 per day. On the basis of what he understood to be the experience of other gambling establishments conducting similar operations respondent concluded that 12 per cent of this amount was retained by the partnership as profits. Respondent also understood that the experience of other gambling establishments indicated that wire service expense is usually one-tenth of the profits and he computed the partnership horsebook winnings on this basis. However, he did not use the figure reached by the second method because it exceeded that reached by the first method. Dice. Respondent understood that the experience of other gambling establishments indicated that the net winnings on a dice operation are approximately equal to 10 times the salary paid the employees staffing the operation. Respondent, aware that all the tables did not operate every evening, computed partnership income on the basis of one table and that only staffed by four employees. Respondent used a salary of $ 20 per employee per night in this calculation, and thus arrived at a profit of $ 800 per night, or $ *150 5,600 per week. Slot machines. On the basis of information received from former Jungle employees respondent concluded that $ 1,500 was given to *855 change makers for change purposes each evening. Respondent assumed that all of this amount was played into the slot machines every evening. On the knowledge that another Ohio slot machine operator, using standard machines, retained 79 per cent of the gross play, respondent used that percentage figure in computing partnership income from the slot machine operation. Poker, roulette, and chuck-a-luck. Respondent concluded that these operations would bring the partnership $ 50,000 per year. Respondent prorated this amount to 45 weeks in 1948 and 32 weeks in 1949. Respondent's $ 50,000 figure was not based on any evidence relative to the Jungle's operations, nor was it based on experience in dealing with similar operations. Bingo. Respondent concluded that no profit or loss arose from this operation; accordingly no figure from this operation entered into his determination. Each of the four partners forwarded a declaration of estimated tax to the collector of internal revenue showing an estimated income of zero for the year 1948. These declarations *151 were prepared for the partners by the bookkeeper for the partnership. None of the petitioners filed a declaration of estimated tax for the year 1949. The failure to file declaration of estimated tax for the year 1949 was not due to reasonable cause but to willful neglect. In 1947 Farah acquired 100 slot machines, some new and some slightly used. In 1948 he obtained a few new machines as replacements. The machines belonged to Farah, but he allowed the partnership to use them. The machines were seized by the public officials of the State of Ohio in the raid of August 12, 1949, and were destroyed. The money in the machines was not returned to the partnership and the amount thereof was not included in the $ 424,766 "winnings" shown on its books for 1949. It was used in paying fines that were imposed in connection with the raid. OPINION. 1. Petitioners Farah, Tobin, Coletto, and Delsante were partners in a gambling venture of considerable magnitude. The partnership operated a casino known as the Jungle, near Youngstown, Ohio, which provided its patrons with gambling facilities on an extensive scale. It continued to function during the taxable years until the casino was raided by State *152 authorities in the evening of August 12, 1949. Farah was plainly the dominant figure. The evidence shows that he furnished the physical facilities for the enterprise, that he controlled the bankroll, that he alone knew the combination of the safe in which the funds were kept, and that he had a 40 percent partnership interest. Tobin and Coletto, although playing important and *856 active parts in the day-to-day conduct of the business, appear to have been subordinate to Farah. Initially, their interests were 25 per cent each, but later, when illness caused one of them to curtail his activities for a substantial period, they readjusted their interests to 20 and 30 per cent. The position of Delsanter, on this record, seems to be highly mysterious. The evidence does not show that he contributed any capital to the partnership or what, if any, services he rendered to it. His 10 per cent interest in the venture could give rise to teasing speculation, but since no question is presented with respect to his participation, we pass over the matter as being outside the issues raised in this litigation. Seven separate and distinct gambling operations were conducted by the partnership at the Jungle; *153 Dice games, poker games, roulette, bingo, chuck-a-luck, slot machines, and a horsebook. The Jungle was inaccessible except by automobile or taxi, but it had about 300 patrons a night during the first 4 days of the week, and attendance ranged from 500 to 700 during Fridays, Saturdays, and Sundays. The horsebook was operated 6 days a week, and the evidence shows that there may have been some "action" at the poker and dice tables in the afternoons; all forms of gambling except the horsebook were open to patrons and were actively indulged in 7 nights a week. With the exception of the slot machines, profits from the various operations were determined daily. The net amount of the funds from each operation in excess of the bankroll assigned to such operation was treated as a profit, and any deficiency as a loss. Such profit or loss for each operation was recorded on an adding machine tape. On 2 nights a week the collections from the slot machines were added. These nightly computations took place in Farah's office. At least two partners were always present, in order to insure the integrity of the computations. The results of the day's operations were added to the bankroll for the entire *154 enterprise as of the close of the previous day. A copy of the tape was given to each partner, who thereupon destroyed his copy of the tape for the previous day. There was also recorded by pencil on a slip of paper the date and a single figure preceded by a single word, "win" or "lose," and this slip was picked up by the bookkeeper for the venture, who entered it in a so-called ledger. There was one page in the ledger for each month and the total for each month accurately reflected the information made available to him by the slips. He prepared returns for the partnership and the individual partners, which were faithfully in accord with the ledger. We have no doubt that the tapes prepared in Farah's office each evening for the use of the partners contained reliable data as to the profits of the partnership. And if the penciled slips given to the *857 bookkeeper accurately showed what was recorded on the tapes, there would be no deficiencies here. However, none of the tapes was available to the Government and none was produced in evidence. We have only the word of several of the partners that the bookkeeper was given the correct figure each day. In substance, petitioners have attempted *155 to hide behind an impenetrable wall and they tell us that the Government must accept that daily figure without any opportunity to check its accuracy. Moreover, it was made abundantly clear by the testimony of the partners that they never allowed a partner to count any winnings by himself, in the absence of another partner. It would be asking too much of the respondent to expect him, in turn, to extend more trust to the partners collectively than they extended to each other individually. If we were to believe their story about the accuracy of the now nonexistent slips, that would be an end to this case. But the Judge who presided at the trial and had ample opportunity to observe the witnesses does not believe that they were telling the truth, and we cannot accept their testimony as credible. We do not deem it necessary to set forth the respects in which we find unworthy of belief the assertion that the slips given to the bookkeeper and recorded in the ledger accurately reflected the earnings of the enterprise. Our conclusion is based on the entire record. However, one example will suffice to raise serious doubts as to the reliability of the ledger. The evidence was clear that *156 patronage, and consequently profits, increased as the inclement winter weather ceased. Yet the ledger shows no significant difference in profits as between different seasons of the year. This case is unlike H. T. Rainwater, 23 T. C. 450, where the records produced were shown to have been used by the partners in determining the amounts of their distributive shares as against the dominant partner, a circumstance that provided a reliable check on the accuracy of the figures presented to the Commissioner and to this Court. This brings us then to the question as to what disposition of the controversy should be made here. The answer to that question is to be found in connection with the applicable burden of proof in this litigation. Plainly, the burden of proof was upon the petitioners. And to the extent that we find incredible their testimony that they accurately furnished their bookkeeper with the results of their daily operations, the burden has not been carried. Rather than attempting to show by credible evidence what profits were realized from each of the various gambling operations 2 conducted at the Jungle, petitioners took the offensive and challenged the Commissioner's determination *157 as arbitrary. The trial to a considerable degree involved an attack *858 upon the revenue agent who had made the investigation and an attempt to discredit the formulas that he used in computing profits from the various gambling operations. Thus, they focus upon admissions made by him that he used formulas or factors given to him by his superiors based, as he understood, upon wide experience within the Internal Revenue Service in dealing with such activities. Somehow or other, petitioners seem to think they have discredited the Commissioner's determination by showing that the agent accepted the formulas given to him and that he could not testify of his own knowledge as to their reliability. But the revenue agent is not the Commissioner of Internal Revenue. The Commissioner, of necessity, must rely upon the assistance of many subordinates in his organization, and his determination often represents the product of study and investigation conducted by many persons. It is no answer here to say that the particular agent on the stand was unable to support each formula used, or that he relied upon instructions or information furnished to him by other subordinates of the Commissioner who were *158 not before the Court. Of course, if the burden were on the Government in this case, then plainly enough the agent's testimony here would not be sufficient to carry it. But the burden was not on the Government, and we will not lightly assume that the formulas he used were without rational foundation. The petitioners cannot discharge their burden merely by showing that the agent followed instructions of his superiors in using formulas which they supplied to him and which he understood were worked out on the basis of experience in the field. The burden was not on the Government to produce evidence justifying the Commissioner's determination. The burden was on the petitioner to show that the determination was erroneous. If that burden was difficult to meet by reason of their destruction of their records, it is a situation that they created for themselves. But they cannot shift the burden to the Commissioner and then complain that, since he has not put witnesses on the stand justifying his determination, it is arbitrary and must therefore be disapproved. We cannot say, apart from certain modifications required by the evidence and upon which we will comment below, that the computations *159 made by the agent were unreasonable and that the profits he calculated for each operation did not fairly reflect the gains realized by the partnership. Horsebook. The petitioners have not shown that the agent's method of determining income by taking 12 per cent of the total bets was arbitrary. The agent proceeded upon a reasonable assumption that the 12 per cent formula furnished to him by his superior was based upon a nationwide experience with establishments of this type. Indeed, petitioner Tobin testified to the effect that he thought the partnership retained 10 or 11 per cent. Certainly, the use of the 12 per cent formula has not been shown to be arbitrary. *859 The agent applied this formula to an assumed total of bets averaging $ 3,000 a day. We think, on the evidence, that this figure must be revised. It is our best judgment on the record that the average total amount bet at the Jungle horsebook on weekdays during the period before the Court was $ 2,500, and that the average *160 total amount bet on Saturdays was $ 3,500. We have made a finding to that effect. Dice. Here, too, the respondent used a formula to determine income. It was a formula which the agent understood from his superior to be based upon a nationwide experience with dice operations such as were carried on at the Jungle -- namely, that it was the usual practice in such operations to employ a staff requiring a salary of not more than 10 per cent of the amounts retained by the house as winnings. Petitioners presented no evidence tending to show that the formula used was lacking in validity, or that there were circumstances rendering it inapplicable to their operation. The agent conservatively applied the formula to the staff of only one table, although it was customary to open a second table before the evening was over; and, again with conservatism, he used a staff of four, instead of the full complement of five which included the relief man. Since these employees were paid $ 20 a night, he arrived at the conclusion that the profits amounted to $ 800 a night or $ 5,600 a week. The petitioner Coletto himself indicated that the dice operation brought in $ 5,000 a week. In the circumstances, *161 we do not regard the agent's conclusion in this respect as arbitrary, and we have found as a fact that the Jungle's dice operation had average weekly net winnings of $ 5,600. Slot machines. The revenue agent computed income from the slot machines on the assumption that the amount given to the change makers each night in this operation was $ 1,500, and that the entire $ 1,500 represented gross play; and, on the knowledge that another Ohio slot machine operator using standard machines realized gains equal to 79 per cent of the gross play, he applied that percentage to the $ 1,500 figure to obtain the earnings of the partnership in this department. We think, on the evidence, that the figure reached by the agent must be revised downward. We are satisfied that there is some correlation between the amount of cash given to the change makers and the gross play, but, on the evidence, we have found that the nightly amount given to the change makers was $ 1,080. This does not mean that the 79 per cent formula should be applied to this figure. That formula appears to have been based on a far more limited experience than those employed with respect to other forms of gambling, nor is it clear *162 that the amount given to the change makers should be taken as the equivalent of the gross play. While it appears that they generally did not use all the *860 change each night, and while it is by no means certain that all coins obtained by patrons from the change makers were played, those circumstances could be more than offset by the amount of coins played that did not originate with the change makers and the amount of coins replayed by patrons who had obtained winnings from the machines. Notwithstanding the foregoing difficulties, we are aided in one important respect in arriving at a reasonable estimate of the nightly winnings of the house from slot machines. Our findings show that there was a total of $ 926.70 in the collection boxes of the slot machines at 9:30 p. m., Friday, August 12, 1949, at the time of the raid. The boxes had been emptied, according to the practice of the house, on Thursday, and since winnings were computed nightly at the conclusion of the evening's operations, we must assume that the amounts in the collection boxes represented winnings that were primarily the result of play on Friday. Certainly, petitioners have not shown that they contained any Thursday winnings. *163 At 9:30, the evening was still young, and had the raid not occurred it is plain that the profits would have been considerably higher before the end of the evening. However, taking into account the fact that it was summertime and that the day was a Friday when patronage was higher than average, it is our best judgment that the partnership's earnings from slot machines over the entire period in question averaged $ 1,000 a day and we have so found as a fact. Poker, roulette, and chuck-a-luck. Respondent concluded that aggregate gains from these three operations were realized at the rate of $ 50,000 per 52-week year, and he determined the profits from these games by prorating that amount over the number of weeks in 1948 and 1949 that the partnership was operating the Jungle. The record shows that the agent did not rely upon any specific facts or use any formula in the selection of the $ 50,000 figure. However, the record does contain evidence from which it is possible to make a reasonable estimate as to the profits derived from these games. On the basis of the evidence before us we have found that the Jungle's 5 per cent profit on poker averaged $ 20 per hour. Poker was played not less *164 than 4 hours a night, and there were occasional daytime games. We have found that the partnership's profits from poker averaged $ 600 a week. The evidence as to roulette and chuck-a-luck was meager. Apparently, they were not significant aspects of the over-all operation of the establishment. The odds were in favor of the house, but in view of the limited play, we could not find on the record that the winnings were substantially in excess of the salaries of the employees who were in charge of these games. Accordingly, we have found that the profits from roulette and chuck-a-luck were not less than $ 140 and $ 250 a week, respectively. *861 Bingo. The respondent treated bingo as a "break-even" operation and attributed neither gain nor loss to it. The evidence shows that the bingo games were the lure that attracted the large number of patrons to the casino, that the partnership expected to make its profits in the evenings from other forms of gambling in which patrons would participate once they were present, notably dice and slot machines, and that there was no intention of deriving any profit from the bingo games. Moreover, it is persuasive from the evidence that the bingo games were *165 in fact operated at a substantial loss. Petitioners' counsel, on brief, suggest that the loss might be taken to be $ 1,000 a week. Although the evidence might support a larger figure, we have accepted the suggestion of petitioners' counsel as reasonable, and have found as a fact that the bingo games were operated at a loss of $ 1,000 a week. 2. The next question is whether petitioners are liable for additions to tax under sections 294 (d) (1) (A) and 294 (d) (2), Internal Revenue Code of 1939. 3*166 Contrary to respondent's contention, this issue was raised by all of the petitioners in their respective petitions. Section 294 (d) (1) (A) provides for an addition to tax in the case of a failure to make and file a declaration of estimated tax unless the failure is due to reasonable cause and not to *167 willful neglect. Section 294 (d) (2) provides for an addition to tax for substantial underestimate of estimated tax. This Court has consistently upheld the imposition of additions to tax for failure to file a declaration of estimated tax and concurrently the imposition of the additions to tax for a substantial underestimate of the estimated tax, pursuant to the provisions of sections 294 (d) (1) (A) and 294 (d) (2). See G. E. Fuller, 20 T. C. 308, affirmed 213 F. 2d 102 (C. A. 10); Fred N. Acker, 26 T. C. 107; John R. Rictor, 26 T. C. 913. *862 See also Clayton v. Commissioner, 245 F. 2d 238, (C. A. 6, 1957) affirming T. C. Memo. 1956-21. In the instant case, the parties orally stipulated at the hearing that "Taxpayers John Farah and Anthony Delsanter, Edward Tobin and Ralph Coletto, forwarded declarations of estimated tax for the year 1948 to the collector of internal revenue for estimated income of zero for said period." On opening statement counsel for respondent stated that a search of the records of the director (formerly collector) had failed to disclose any record of the filing of estimated tax by any of the petitioners for 1949. While statements of counsel are not to be taken as *168 evidence, this statement did serve to put petitioners on notice. The burden was on petitioners to prove that declarations of estimated tax were filed for the year 1949. They have not met that burden, and we have found as fact that no declaration of estimated tax was filed by any of the petitioners for the year 1949. The petitioners have not shown that their failure to file declarations of estimated tax for the year 1949 was due to reasonable cause and not to willful neglect. The statute contains no provision excusing an underestimate of estimated tax on a showing of reasonable cause. Considering all the circumstances, we hold that each of the petitioners is liable for the additions to tax imposed by sections 294 (d) (1) (A) and 294 (d) (2) for the year 1949, the amounts thereof to be computed under Rule 50. We likewise hold that each of the petitioners is liable for the additions to tax imposed by section 294 (d) (2) for the year 1948, the amounts thereof to be computed under Rule 50. See G. E. Fuller and Fred N. Acker, both supra. The question remains whether each of the petitioners is liable for addition to tax for failure to file a declaration of estimated tax for the year 1948, *169 pursuant to section 294 (d) (1) (A), where the declaration filed showed an estimated tax of zero. Respondent, in substance, argues that petitioners did not attempt to make a bona fide estimate of their 1948 tax on their declarations filed for that year, and that this amounts to a failure to file a declaration. We do not agree. In section 294 (d) (2)Congress provided for an addition to tax where there is a substantial underestimation of tax, and it is not for respondent to provide for a second addition where the substantial underestimation is deliberate. If the legislature had intended such a result it would have been a simple matter so to provide. Cf. sec. 293, I. R. C. 1939. Additionally, respondent points out that petitioners themselves did not estimate their 1948 taxes or file the required declarations, and that these acts were done by their attorney-bookkeeper. This, respondent maintains, does not satisfy the requirements of section 294 (d) (1)*863 (A). There is little merit to the argument. The declarations were in fact signed by the petitioners, and we would open a Pandora's box if we undertook to go behind them to inquire to what extent each petitioner went through a mental *170 process the results of which were reflected on the declarations. 3. The petitioners John Farah and Shamis Farah allege that either they or the partnership was entitled to deductions for depreciation on 100 slot machines in 1948 and 1949 and for a loss occasioned by the confiscation and destruction of the machines by public officials of the State of Ohio in 1949. In view of our holding that none of the deductions can be allowed on this record, it is unnecessary to decide who is entitled to the deductions. The claimed depreciation deduction must be denied for complete failure of proof. All the record shows is that Farah acquired the slot machines in 1947 and replaced a few of the machines in 1948. The record does not disclose what amount, if any, he paid for the slot machines, their useful life, and the amount, if any, allowed or allowable as depreciation on the machines in 1947. On the record we are unable to determine even the minimum amount which should be allowed as depreciation in 1948 and 1949. Similarly, no loss deduction can be allowed for the confiscation and destruction of the slot machines in 1949, there being no showing that these machines had any basis to any of the petitioners *171 at that time. It thus becomes unnecessary to consider whether the loss must be denied in any event because it was like a fine or forfeiture. Cf. G. E. Fuller, supra. See Boyle, Flagg & Seaman, Inc., 25 T. C. 43, 49-50. Decisions will be entered under Rule 50. MURDOCK; BRUCE Murdock, J., dissenting: The Commissioner determined that the zero declarations were insufficient to avoid the imposition of additions to the tax under section 294 (d) (1) (A). The petitioners had the burden of proving error in that determination. The findings above show nothing on the subject, but the stipulation (adopted as findings) is that each of the four men "forwarded declarations of estimated tax for the year 1948 to the collector of internal revenue in March, 1948 * * * for estimated income of zero for said period." Those documents were not accompanied by the payment of any estimated tax for 1948. Section 58 (a) (2), as applicable to these petitioners for the year 1948, required each to "make a declaration of his estimated tax for the taxable year if -- * * * (2) his gross income from sources other than wages (as defined in section 1621) can reasonably be expected to exceed $ 100 for the taxable year *172 and his gross income to be $ 600 or more." *864 Section 58 (a) also required later filing of the declaration of estimated tax up to January 15, 1949, if the requirements for filing were first met later in the year. No facts are found to show why the petitioners filed the zero estimates "in March, 1948." No declaration is required on an honest estimate of zero income. However, the duty of each did not end there, even if he was acting honestly, because he was required by law to make and file the estimate later if, after March 1, 1948, for the first time it could have been reasonably estimated that his gross income from other than wages would exceed $ 100 and his gross income would be $ 600 or more. Each of these taxpayers had non-wage income for 1948 far in excess of $ 100 and had gross income far in excess of $ 600. There is no proof in this record that any one of them could not have estimated reasonably at some time during the year 1948 or prior to January 15, 1949, that his income would have been substantial. There is no showing that during March, when they filed the so-called declarations, it would not have been reasonable on the part of each one of them to have estimated that he *173 already had or would have a substantial amount of taxable income on which he was required under the law to make an honest declaration of estimated tax and pay a portion thereof at the time of filing the declaration. Obviously it was apparent to each one of them before January 15, 1949, the last date for filing a declaration of estimated tax for 1948, that each had had substantial income and was liable for a substantial amount of tax. None ever filed any further declaration. The evidence indicates to me that these petitioners have flaunted the requirements of sections 58 and 294 (d) (1) (A). The zero declarations were not the required declarations. The majority Opinion in recognizing their zero declarations as avoiding additions to the tax provided by the latter section, has exaggerated a form over the substance required and has allowed these taxpayers to frustrate the intent of Congress in regard to the current payment provisions of the statute and substantially emasculates the provisions of section 294 (d) (1) (A) which provides that: In the case of a failure to make and file a declaration of estimated tax within the time prescribed, unless such failure is shown to the satisfaction *174 of the Commissioner to be due to reasonable cause and not to willful neglect, there shall be added to the tax 5 per centum of each installment due but unpaid, and in addition, with respect to each such installment due but unpaid, 1 per centum of the unpaid amount thereof for each month (except the first) or fraction thereof during which such amount remains unpaid. * * * For the purposes of this subparagraph the amount and due date of each installment shall be the same as if a declaration had been filed within the time prescribed showing an estimated tax equal to the correct tax reduced by the credits under sections 32 and 35. *865 Those provisions are entirely aside and in addition to those provided for substantial underestimate of estimated tax contained in section 294 (d) (2). I think the petitioners in this case have failed to overcome the presumption of correctness attaching to the Commissioner's determination that the zero declarations which they filed in March 1948 were not declarations within the meaning of sections 58 and 294 (d) (1) (A). Cf. Ferrando v. United States, 245 F. 2d 582. I agree with the majority on all other issues. Bruce, J., dissenting: I concur in the Opinion *175 of the Court with respect to all of the issues except the first, which relates to the determination of the taxable income derived by each of the petitioners from the Jungle Novelty Company, a partnership engaged in conducting a gambling casino, during the years 1948 and 1949. With respect to that issue I respectfully dissent. As the trial Judge who presided at the hearing, however, I wish to make it clear that I disagree with the views of the majority principally with respect to the theories by which they have arrived at their conclusions and not as to any evidential matters depending upon the credibility of the witnesses. Specifically I wish to affirm that I do not accept as credible the self-serving and uncorroborated testimony of the petitioners as to the correctness of their books or that the slips, which were made available to the attorney-bookkeeper, and on the basis of which he kept the books of the partnership, accurately set forth the results of their daily computations showing the amounts won or lost by the partnership. With respect to the first and principal issue, it is my view that the majority Opinion does not squarely meet the contentions of the petitioners. To the *176 extent the majority hold that petitioners have not established that the determinations made by respondent are not arbitrary and to be disregarded, I disagree. As I understand them, petitioners' contentions, both at the hearing and on brief, are: (1) That respondent was not justified in disregarding the partnership books and returns and in computing the taxable income by some other method; (2) that the method employed by respondent is unreasonable and the deficiencies determined by him arbitrary, without foundation in fact, and invalid; and (3) that where the evidence establishes that the determination of the respondent is arbitrary and invalid, petitioners are not required to show that they owe no tax or the correct amount. Petitioners cite and rely upon Helvering v. Taylor, 293 U.S. 507; Durkee v. Commissioner, (C. A. 6) *866 162 F. 2d 184; Gaspar v. Commissioner, (C. A. 6) 225 F. 2d 284; and others. Respecting petitioners' first contention, the books and returns clearly do not reflect the income of the partnership and consequently the income of the individual petitioners. Certainly, with respect to the horsebook, the partnership had, in the form of the betting slips, a record of the total *177 amount of bets taken each day, which represented its gross receipts from this source. It also had, in the form of the betting slips on which notations were made of the winning bets and the customers' copies of such slips which were turned in by them upon collecting their winnings, a record of the total amount of "wins" paid out. It retained none of this substantiating data and neither made nor retained any other record of such amounts. While the nature of such games precludes the making of a record of each bet placed, the partnership did have, in the form of the results reported by the operators and recorded on an adding machine tape, a record of the daily amounts won or lost from the operations of the poker tables and the dice, roulette, and chuck-a-luck games, and at least twice a week from the slot machines. It retained none of the adding machine tapes which would have shown its gross receipts or losses from these sources, nor any other record of such amounts. The partnership also knew the number of, and price paid for, bingo cards each day which represented its gross receipts from this source. The difference between the amount received and the amount paid out as prizes was *178 also reported to the partnership each day by the operators in charge of that activity, and the amount of gain or loss recorded on the adding machine tape along with that of the other activities. By recording and reporting only "net winnings" or "net losses" petitioners determined for themselves the amount of their allowable deductions for losses, without any opportunity or possibility of the respondent checking them as the statutes and regulations contemplate he should have. Cf. Fairchild v. United States, 136 F. Supp. 753, reversed on other grounds (C. A. 5) 240 F. 2d 944. "When a deduction is claimed, the government has an undoubted right to demand a full disclosure of the facts on which the claim is based, for otherwise it would be at the mercy of the unscrupulous taxpayer." O'Laughlin v. Helvering, (C. A., D. C.) 81 F. 2d 269. Accordingly respondent was justified in undertaking to compute the income of the partnership and of petitioners by some other method which would clearly reflect the income. Sec. 41, I. R. C. 1939. With respect to petitioners' second contention, in my opinion they have sustained their burden of proof in overcoming the presumptive correctness of respondent's *179 determination. Although this may not have been done on petitioners' case in chief, it is my view that it was *867 certainly done when respondent called the revenue agent as his own witness and opened him up to cross-examination. At this point I think the burden of going forward shifted to respondent to establish by competent evidence, the additional facts upon which he relied in arriving at the deficiencies finally determined by him. This he did not do. Our determination is to be made from the entire record. In the case of the horsebook, respondent's agent first estimated the average amount of bets placed each day on the basis of estimates made by certain employees of the Jungle Inn engaged in that activity. He then estimated that the partnership realized earnings or net winnings equal to 12 per cent of the gross amount bet. By multiplying the average daily take by the known number of days the horsebook operated, he computed what he determined to be the amount realized on the horsebook for each of the years 1948 and 1949. Income from the dice tables was computed on the basis of 10 times the total salaries paid to the employees operating a single dice table. The amount realized from *180 the slot machines was computed by first estimating the average amount placed in the slot machines by players each day. This estimate was based upon estimates given by employees engaged in making change. Respondent's agent then estimated that the machines retained 79 per cent of the gross amount played and that this 79 per cent represented the partnership's net winnings from this activity. Based upon an admittedly arbitrary estimate of the aggregate amount that would be realized over a full year from the operation of the poker tables, roulette, and chuck-a-luck, and the known number of weeks the Jungle Inn was in operation, he computed the amount realized from those three activities during 1948 and 1949. Based upon statements made by certain employees engaged in operating the bingo games, respondent's agent concluded there was no profit or loss by the partnership from this activity. Respondent determined the partnership net income for 1948 and 1949 by adding the estimated profits on the various activities and subtracting the total amount of expenses claimed by the partnership. The formulas, 12 per cent of gross amount bet for computing winnings on the horsebook, 79 per cent of gross *181 amount played for computing winnings from the slot machines, and 10 times salaries paid for computing winnings from the dice games, were furnished respondent's agent by his supervisor, the bases for which were unknown to the agent and are not disclosed by the record herein. Cf. Morris Nemmo, 24 T. C. 583, on appeal (C. A. 6); H. T. Rainwater, 23 T. C. 450. Undoubtedly, because of the lack of adequate records maintained by the partnership or by petitioners individually, respondent was faced with a difficult task of determining with reasonable accuracy the income of the partnership and of petitioners. Lacking information *868 on which he could predicate any other method, such as the net worth and bank deposits methods, respondent resorted to the method described above. While no particular method is prescribed, the method adopted must be one that will clearly reflect the income, Bradstreet Co. of Maine v. Commissioner, (C. A. 1) 65 F. 2d 943, reversing in part 23 B. T. A. 1093, and, as stated in Mount v. Commissioner, (C. A. 2) 48 F. 2d 550, "[there] must be a limit beyond which the presumptive correctness of the Commissioner's determination may not be stretched in order to defeat a taxpayer." Considering *182 all the facts and circumstances presented herein, I would hold that the method adopted by respondent is not reasonable and that the deficiencies determined by him are without foundation in fact, and therefore to be disregarded. Helvering v. Taylor, supra;H. T. Rainwater, supra. See also Thomas v. Commissioner, (C. A. 1) 232 F. 2d 520; Gaspar v. Commissioner, supra;Thomas v. Commissioner, (C. A. 6) 223 F. 2d 83; and Wilson Coal Land Co. v. Commissioner, (C. A. 4) 87 F. 2d 185, in each of which Memorandum Opinions of this Court were reversed and the cases remanded for further proceedings; and Durkee v. Commissioner, supra, remanding 6 T. C. 773. I do not agree with petitioners' third contention, however. Although, in my opinion, petitioners have sustained their burden of overcoming the presumptive correctness of respondent's determination, the deficiencies determined by respondent are merely to be disregarded. The question remains whether, on the basis of the evidence presented, petitioners are liable for any deficiencies in income tax and, if so, how much. The majority Opinion has arrived at certain amounts which they say represent net income received by the partnership from the *183 various gambling activities. I would employ a different approach and arrive at somewhat different amounts which, to me, are more readily supportable. It is well established that where a taxpayer seeks a deduction, the burden is upon him to prove not only that he is entitled to it but also the amount of it. Burnet v. Houston, 283 U.S. 223; Helvering v. Independent Life Ins. Co., 292 U.S. 371; New Colonial Co. v. Helvering, 292 U.S. 435. See also Helvering v. Taylor, supra, wherein, after referring to the burden upon the taxpayer, in a suit to recover taxes paid, to show the amount to which he was entitled, the Supreme Court said: "For like reason the burden is upon the taxpayer to establish the amount of a deduction claimed." Accordingly the burden was upon the petitioners to establish not only the right to deduct the "losses" sustained in the gambling operations of the partnership but the amount thereof. The self-imposed impossibility of proving *869 the material facts upon which the claim rests does not relieve the taxpayer of his burden of proof. Burnet v. Houston, supra. In their individual income tax returns petitioners set forth their distributive shares of the net income shown *184 on the partnership returns. The partnership returns reported as gross receipts at line 1 under the general heading "Gross Income" the amounts of $ 435,242 and $ 355,258, respectively, for the years 1948 and 1949. The same figures were set forth, at line 13 under the same general heading, as its total income. Under the general heading "Deductions," at lines 14 to 24, inclusive, were set forth expenses, such as wages, rent, taxes and licenses, utilities, supplies, race wires, and cab fares, totaling $ 303,032 and $ 259,597, respectively, for the years 1948 and 1949. The net income reported by the partnership for the years 1948 and 1949 was $ 132,210 and $ 95,661, respectively. The amounts returned as gross receipts in fact represented the difference between the total winnings and losses as shown on the books maintained for the partnership by the attorney-bookkeeper. The books for the year 1948, having been destroyed in the raid conducted by the State officials on August 12, 1949, are not available. The books for the year 1949, however, are available and were made a part of the record herein. These books show that the partnership had total winnings during the year 1949, at least*185 in the amount of $ 424,766. The partnership books for the year 1949 also set forth as losses the total amount of $ 69,508. It is the difference between these two figures ($ 424,766 less $ 69,508 equals $ 355,258) which the partnership reported as its gross receipts and gross income for the year 1949. When considered together, it is clear from the partnership books and returns that petitioners are claiming the right to deduct the sum of $ 69,508, as gambling losses, for the year 1949. Section 23 (h) of the Internal Revenue Code of 1939 provides that, in computing net income, losses from wagering transactions, to the extent of the gains from such transactions, shall be allowed as deductions from gross income. Respondent has not questioned the right of petitioners to deduct gambling losses but does contend that petitioners have not sustained their burden of establishing the amount of such losses. In this respect I agree with respondent. The total winnings of the partnership for the year 1949 were at least $ 429,970.70. This is the figure shown on its books (which may be treated as an admission against interest), plus $ 5,204.70 found in the slot machines when seized by the State officials. *186 In finding such amount I am aware of the fact that, due to the practice of deducting losses from winnings, or vice versa, and entering on the books only the resulting difference, the total winnings could well have been, and probably were, larger. *870 The question of the amount of losses sustained by the partnership is essentially one of fact to be determined from the entire record. H. T. Rainwater, supra.I do not doubt that the partnership sustained some losses with respect to the horsebook and perhaps, on some occasions, with respect to bingo, dice, chuck-a-luck, and roulette. Since its take from the poker tables was a percentage of each "pot," it sustained no losses from that activity and it is extremely doubtful that the slot machines ever showed a loss on any day of operation. Implicit in the partnership's method of bookkeeping is the fact that it offset an unknown amount of claimed losses for every day of operation, both on days when its books listed only net winnings and on the days when they listed only net losses. It would serve no useful purpose to repeat the facts here. It is sufficient to say that having carefully considered all the evidence presented, for the reasons *187 heretofore discussed, I find it insufficient to establish, even under the Cohan rule ( Cohan v. Commissioner, (C. A. 2) 39 F. 2d 540), the amount of losses, if any, sustained by the partnership in excess of those thus taken. In so doing I give no weight to the self-serving and uncorroborated testimony of the petitioners as to the correctness of their books. In summary, I would find and hold that, during the year 1949, the partnership, Jungle Novelty Company, received as winnings from gambling operations the amount of at least $ 429,970.70; that the evidence is not sufficient to establish the amount, if any, of its gambling losses in excess of those implicitly taken; that it incurred expenses in the amount of $ 259,597.75; and that its net income distributable to the petitioners in proportion to their respective partnership interest was $ 170,372.95. The situation with respect to the year 1948 is somewhat different from 1949, due to the fact that the partnership books maintained by the attorney-bookkeeper for 1948 were not available. It is possible, however, to make a reasonably accurate estimate with respect to winnings and the amount claimed as losses from the evidence available and *188 by comparison with the operations for 1949. The evidence shows that for 1948, as in the case of 1949, the amount shown on the partnership return as gross receipts represents the amount of its winnings after deduction of its claimed losses. The average weekly net income reported on the partnership returns was $ 2,938 for 1948 ($ 132,210 divided by 45), and $ 2,989 for 1949 ($ 95,661 divided by 32), or substantially the same for each year. The average weekly income reported for 1948 was $ 9,682 ($ 435,242 divided by 45), and for 1949 was $ 11,102 ($ 355,258 divided by 32). This difference could have resulted from the taking of greater losses in 1948 than 1949, especially since it is noted the average weekly expenses reported for 1948 ($ 303,032 *871 divided by 45 equals $ 6,734) was less than those for 1949 ($ 258,597 divided by 32 equals $ 8,081). The average weekly losses claimed for the year 1949 were $ 2,172 ($ 69,508 divided by 32). It is reasonable to assume that the average losses taken for 1948 were at least $ 2,000 per week, and that the total winnings for the 45 weeks the partnership operated in 1948 were at least $ 525,242 ($ 435,242 plus $ 90,000). In summary, I would find *189 and hold that, during the year 1948, the partnership, Jungle Novelty Company, received as winnings from gambling operations the amount of at least $ 525,242; that the evidence is not sufficient to establish the amount, if any, of its gambling losses in excess of those implicitly taken; that it incurred expenses in the amount of $ 303,032; and that its net income distributable to the petitioners in proportion to their respective partnership interests, was $ 222,210. As previously indicated, I concur in the Opinion of the Court with respect to all the issues except the first. For the reasons stated above I respectively dissent with respect to the first issue. Footnotes *. Proceedings of the following petitioners are consolidated herewith: Estate of Mary Catherine Tobin, Deceased, Edward F. Tobin, Administrator, and Edward F. Tobin, Individually, Docket No. 52463; John Farah and Shamis Farah, Docket No. 52515; and Ralph Coletto and Tessie Coletto, Docket No. 52557.↩ 1. No evidence relating to compartments in the dollar machine was presented to this Court.↩ 2. Petitioners did present some evidence as to earnings from some of the games, but it was not of sufficiently comprehensive character to enable us to determine the correct income of the entire enterprise.↩ 3. SEC. 294. ADDITIONS TO THE TAX IN CASE OF NONPAYMENT. (d) Estimated Tax. -- (1) Failure to file declaration or pay installment of estimated tax. -- (A) Failure to File Declaration. -- In the case of a failure to make and file a declaration of estimated tax within the time prescribed, unless such failure is shown to the satisfaction of the Commissioner to be due to reasonable cause and not to willful neglect, there shall be added to the tax 5 per centum of each installment due but unpaid, and in addition, with respect to each such installment due but unpaid, 1 per centum of the unpaid amount thereof for each month (except the first) or fraction thereof during which such amount remains unpaid. In no event shall the aggregate addition to the tax under this subparagraph with respect to any installment due but unpaid, exceed 10 per centum of the unpaid portion of such installment. For the purposes of this subparagraph the amount and due date of each installment shall be the same as if a declaration had been filed within the time prescribed showing an estimated tax equal to the correct tax reduced by the credits under sections 32 and 35. * * * * (2) Substantial underestimate of estimated tax. -- If 80 per centum of the tax (determined without regard to the credits under sections 32 and 35), in the case of individuals * * * exceeds the estimated tax (increased by such credits), there shall be added to the tax an amount equal to such excess, or equal to 6 per centum of the amount by which such tax so determined exceeds the estimated tax so increased, whichever is the lesser. * * *↩