The only issue involved in this workmen’s compensation case is the correct method of computing the average weekly wage of the injured employee under the provisions of T.C.A. § 50-902(c).
The employee worked for the appellant, Oakley Gaw, on a part-time basis. He was regularly employed at another job at which he worked a regular 40-hour week at a rate of $3.00 per hour.
The employer, Mr. Gaw, was engaged in a house building and construction operation and also owned and operated a store, known as Gaw’s Factory Outlet Store, in which lawnmowers were sold and serviced. Mr. Gaw testified that appellee had formerly operated a small shop at home in which he worked on lawnmowers, and from time to time in the past Mr. Gaw would send lawnmowers from his store to the appellee’s home for service or repair. Later, appellee worked at a service station, where he also did part-time work on lawnmowers. Mr. Gaw testified:
“Then in the meantime, he approached me that he would like to come to my place of business. I had a little shop there and serviced my mowers there and we agreed on a certain amount, which was $2.00 an hour, which has been stated and that’s how it came about.”
Appellee started to perform this work for Mr. Gaw in 1971, but “primarily” worked for him in 1972 and 1973. The record also shows that from time to time, during slack seasons, the employee would lay carpet or do other odd jobs for the employer either in the store or at houses being built by Mr. Gaw. There was no itemization of his earnings according to the type of work performed; he was simply paid a flat rate of $2.00 per hour.
The employee worked at his regular job five days each week, Monday through Friday, from 3:30 p. m. until midnight. He did not work there on Saturdays, and his testimony was that he did most of his work for Mr. Gaw on Saturdays, coming to work at 7 or 8 o’clock in the morning and working for nine or ten hours. He also worked during the morning hours of his regular working days, however, depending upon how much work was carried over from the preceding Saturday. He testified that he came by the employer’s store almost every day during the entire year to see whether or not there was any work available for him.
On July 3, 1973, while repairing a lawnmower, the employee sustained an injury to his left eye, resulting in the loss of sight therein. There is no question but that he is entitled to workmen’s compensation benefits for the loss of an eye, together with necessary medical and other incidental expenses, which are not questioned on appeal.
It is undisputed that during the fifty-two weeks prior to the date of his injury, the employee earned a total of $802.50 from this second or part-time job with appellant, representing 401.25 hours at $2.00 per hour. The only hourly records kept, however, were notes made by the employer in a tablet or note pad, in which the number of hours worked by the employee each day were written down. Unfortunately he had lost the notes kept during the year prior to the accident. The record therefore does not show the exact number of days or weeks during which the employee performed work for wages. Only a single page from the earlier tablet or pad was torn out and filed as an exhibit, showing that the employee worked two hours on March 11, 1972, two *578hours March 17,1972, eight hours March 22, 1972 and four hours March 24, 1972. These dates were more than one year prior to the date of injury, however, and the only record of wages paid during the one year in question, July 1972-July 1973, appears on Exhibit 3 in the record, in the form of a wage statement for weeks ending on given dates.
Both the employer and the employee testified that their arrangement was somewhat informal, and that the employer would not pay the employee by the day, but would only pay him when he had accumulated a fairly substantial number of hours, usually at least fifteen hours. A paycheck given on a particular date, therefore, might reflect hours worked during the week preceding the date of the check, or it might represent hours worked over two or three weeks prior to the date of the cheek.1 Mr. Gaw testified very frankly:
“Q And it was a regular, yearly job on the part of Mr. Raymer, wasn’t it? Yearly part-time job?
“A Yes, part-time. It wasn’t a regular job . . . there wasn’t really anything regular about it.”
The employer’s pay record which was filed as an exhibit reflects the following information:
Week Ending Hrs. Worked Rate Total Wage
2.00 72.00 July 22,1972
2.00 72.00 July 29,1972
2.00 66.00 Sept. 9,1972
2.00 72.00 Sept. 23,1972
2.00 66.00 Sept. 80,1972
2.00 60.00 Oet. 7,1972
2.00 65.50 Oet. 13,1972 eS"~
2.00 60.00 Oet. 21,1972
2.00 30.00 Oet. 28,1972
2.00 43.00 Jan. 27,1973 53
2.00 44.00 Mar. 10,1978 S3
2.00 50.00 May 28,1973 S3
2.00 48.00 June 9,1978 c¡¡
2.00 54.002 July 18,1973 S2
Prior to the period reflected above, the employee had last received a check on June 17,1972 for thirty-six hours and total wages of $72.00.
It is apparent from the foregoing that the employee only received payments- in seven of the twelve months involved, but it is not possible to determine from the testimony given in the case or from the records filed whether he worked at all during the five months when he received no payment, or, if so, how many hours. It is fairly clear that during July, September and October 1972 he was working regularly for Mr. Gaw. During these months he was off from his regular job because of strikes or lay-off. During the rest of the year his work for appellant was sporadic. He sometimes did part-time work for other persons than appellant.
The employee testified that on the days when he did work, he would estimate that he averaged about six hours. The employer disputed this, and testified that he would not place the average above four hours. The employee testified that he probably averaged working on two or three days each week.
T.C.A. § 50-902(c) provides several methods, or formulas, for arriving at an average weekly wage of an employee. The usual method is stated in the first part of the section as follows:
“ ‘Average weekly wages’ shall mean the earnings of the injured employee in the employment in which he was working at the time of the injury during the period of fifty-two (52) weeks immediately preceding the date of the injury divided by fifty-two (52) . .
It is the insistence of the employer that this is the method which should have been used by the trial judge in computing the average weekly wage. Using this method, *579the result would be approximately $15.43 per week, entitling the employee only to the minimum benefits provided by the workmen’s compensation statute.
An alternative method is provided in the statute, however, immediately following the language quoted above:
“but if the injured employee lost more than seven (7) days during such period when he did not work, although not in the same week, then the earnings for the remainder of such fifty-two (52) weeks shall be divided by the number of weeks remaining after the time so lost has been deducted.”
The employee insists that this is the method which should be used, and the Chancellor purportedly did so. Taking the total number of hours worked during the year, 401.25, at an average of six hours a day, the Chancellor concluded that the employee worked on 69 days during the year. He computed this to be ten weeks at seven days per week, and therefore arrived at an average weekly wage of $80.25. He reached this unusual result despite the fact that the maximum earned in any one week was $72.00. Further there was no testimony that the employee ever worked as much as seven days in a week. There was no evidence that the employer’s business was open more than six days a week, and, as we have previously pointed out, the employee testified that he worked there on an average of not more than two or three days per week, receiving paychecks after each two or three weeks worked.
We, therefore, simply cannot accept the findings of the Chancellor, and do not find them to be supported by the evidence or to represent a formula deducible from the statutory language above quoted.
The employer relies primarily upon the cases of Carter v. Victor Chemical Works, 171 Tenn. 141, 101 S.W.2d 462 (1937) and New Jellico Coal Co. v. Kenner, 172 Tenn. 185, 110 S.W.2d 476 (1937).
In the Carter case, supra, the employee worked during each of the fifty-two weeks prior to his injury, in some weeks working as much as seven days and in other weeks as little as two days. His weekly earnings varied accordingly. The only explanation for the variation was contained in a stipulation that there was more work for him to do in some weeks than in others. The employment was regular, however, rather than part-time or seasonal, and the Court accordingly concluded that his total wages for the year should be divided by fifty-two.
This holding was followed in the New Jellico Coal Co. case, supra, where the employee also worked regularly in a coal mine. During the calendar year preceding his injury the mine was closed for ten weeks, eight of these being due to strike and the remainder to natural conditions beyond the control of the company. Under these circumstances it was held that the total annual earnings should be divided by fifty-two to arrive at an average weekly wage.
The employee relies upon a different line of cases, in which the alternative method of computation was utilized, deducting the number of weeks “lost” in computing the average weekly wage. The principal case relied upon is Bryant v. McAlister, 202 Tenn. 654, 308 S.W.2d 412 (1957). There, again, the employee was a regular employee at a sawmill, the employer testifying that he was “employed all the time.” Due to shortage of materials, the employee was laid off during part of the fifty-two weeks prior to his injury. The Court held that this period should not be counted in computing the average weekly wage, citing and relying upon Hartley v. Liberty Mutual Insurance Co., 197 Tenn. 504, 276 S.W.2d 1 (1954).
In the latter case, the Court said:
“The average weekly wage of an employee should not, and is not, decreased for reasons over which he has no control, such as closing a plant for repairs (as in the instant case), occasional ‘suspension of operations due to bad weather’, unforeseen shortage of material, ‘lack of orders, lack of cars, slack season’, etc. It is observed that all of the foregoing are occasions and conditions resulting in the cessation of operations by the employer, *580and should not be considered in determining the average weekly wage of an injured employee.” 197 Tenn. at 509, 276 S.W.2d at 3.
It should be observed that all of the foregoing cases involved regular employees, not persons who were working only part time, as was the employee in the present case.
In the case of Toler v. Nashville, C. & St. L. By., 173 Tenn. 378, 117 S.W.2d 751 (1938), however, the employee was employed intermittently or irregularly. He had worked for the employer for more than ten years, but was only assigned work as it became available. During the year prior to his injury, he worked fourteen weeks.
Under those circumstances, the Court held that the employee had not been regularly employed for as much as a year within the meaning of the statute in question, and that his average weekly wages should be computed according to still another method referred to in the statute as follows:
“Where the employment prior to the injury extended over a period less than fifty-two (52) weeks, the method of dividing the earnings during that period by the number of weeks and parts thereof during which the employee earned wages shall be followed; provided, results just and fair to both parties will thereby be obtained.”
In that case it was held that the proper method of computation was to divide the total wages received during the year by the number of weeks during which the employee received wages.
We are of the opinion that this is the method which should have been employed in the present case. Unfortunately it appears that data are not available from which the number of weeks the employee received wages can be computed with precision. The pay records which the employer did introduce show that he received fourteen paychecks during the year, and, according to one of the employee’s own suggested methods of computation, this would represent at least fourteen weeks. Accordingly, the divisor into the total wages should, at the very least, be fourteen rather than ten.
The uncontradicted testimony of both the employer and the employee, however, is that the employee worked in more weeks than he actually received paychecks, and the checks were accumulated for as much as two or three weeks before payment was made, because of the small number of hours worked. It seems clear from an examination of the pay records in evidence, that there were many weeks, and probably several months, in which the employee did not perform any work at all and during which no wages were earned. It would, therefore, in our opinion be unfair to the employee to divide the total wages by fifty-two, since he was conceded to be a part-time and irregular employee. On the other hand it is equally unfair to the employer to divide the total wages by as little as fourteen if the employee did work during more than that number of weeks or parts of weeks.
Accordingly we cannot approve the result which was reached in the trial court, nor are we able to compute with accuracy the average weekly wage from the data now on file. The statute admits of some flexibility, however, and permits the Court to arrive at a “just and fair” result.
The judgment of the Chancery Court is therefore reversed, and the cause is remanded for a new trial with directions that the average weekly wage be computed in accordance with the Toler case. If, however, no more exact evidence is offered on a new trial than presently appears in the record, the average will be computed upon a basis of not less than fourteen weeks. Cost on appeal will be taxed to appellee. All other costs will be fixed by the Chancellor.
COOPER, C. J., and FONES, HENRY and BROCK, JJ., concur.. The employee testified that the employer “never let it go any longer than two weeks.” The employer said it could be carried “for as much as three weeks.”
. Apparently earned prior to the date of the accident, July 3, 1973.