The appellant is a boiler-maker, who was injured about March 1935, in the course of his employment. This employment was temporary in its nature, and was obtained at irregular intervals and with successive employers between 9th May 1934 and 14th March 1935. The appellant was paid on an hourly basis, and he earned £73 14s. 3d. in fifteen weeks and one day's work during this period. He claimed compensation under the Workers' Compensation Act 1926-1929 of New South Wales, and the question was how his average weekly earnings should be computed under the Act. The Act provides that the average weekly earnings shall be computed in such manner as is best calculated to give the rate per week at which the worker was being remunerated. The appellant contended that his average weekly earnings should be computed at £4 17s. per week, based, as I understand, on the average weekly earnings of a worker in the same grade employed in the same class of work and in the same district (sec. 14 (a), proviso). But the Workers' Compensation Commission was not satisfied that the average weekly earnings of the appellant were best calculated in the manner allowed by the proviso to sec. 14 (a). The appellant was not, apparently, in a comparable position to that of a worker in the same grade &c. No reason has been assigned which leads me to think that the commission was wrong in its conclusion. The commission found that the appellant was a "casual worker," and it applied the provisions of sec. 14 (e) of the Act to his case, and computed his average earnings on the weekly living wage declared in the State of New South Wales, namely, £3 7s. 6d.. The description "casual worker" is not one of precision: it is a colloquial expression, and where, upon all the facts, there is a reasonably debatable question whether the work is casual or regular, the question is one of fact for the commission. The finding of the commission, in the circumstances of the present case, cannot be disturbed. The commission regarded sec. 14 (e) as furnishing the method best calculated to give the rate per week at which the worker was being remunerated. The steps by which it arrived at the living wage, £3 7s. 6d. per week, are not set out, but there is nothing in the case stated to suggest any error on its part.

The result is that the decision of the Supreme Court of New South Wales on the case stated was right, and this appeal should be dismissed.

The average weekly earnings upon which worker's compensation for total or partial incapacity is computed are those of the previous twelve months if the worker has been so long employed by the employer but if not, then for any less period during which he has been in the employment of the same employer (sec. 9 (1) (a) of the Workers' Compensation Act 1926-1929 N.S.W.). The purpose is to obtain the rate of remuneration which at the time of the injury the worker received in that employment. "Average weekly earnings shall be computed in such manner as is best calculated to give the rate per week at which the worker was being remunerated" (sec. 14 (a)). To this end it is necessary to take the period, not exceeding twelve months, during which a nexus continued between the worker and the same employer. Such a nexus is consistent with intervals of idleness through lack of work, or the worker's own absence (Price v. Guest, Keen and Nettlefolds[1]; cp. Scott v. Summerlee Iron Co.[2]). But a change of grade marks a new beginning and perhaps an interruption for some unavoidable cause does so (sec. 14 (c)). For the object is to find the average remuneration obtained by the worker at the time when he suffered the injury and so incurred the incapacity to go on earning it. Accordingly, the weekly average is calculated by dividing the total remuneration received by the worker during the period, not by the number of weeks contained in the period, but by the number of weeks in which the worker was at work. If some of the intervals during which he was not at work and earned nothing are attributable to the ordinary incidents of his employment, the result produced by the calculation will not, it is held, reflect his true average rate of earning. To correct it the product so far obtained must be diminished by an amount bearing the same proportion to the whole as these intervals bear to the total period (Perry v. Wright[3]; Anslow v. Cannock Chase Colliery Co. Ltd.[4]). But it might be found impracticable to compute the remuneration at the date of the injury from the period of the worker's employment by the same employer. This might arise from the shortness of the period, or from the casual nature of the employment, or from the terms of the employment. Thus in the case of workers like stevedore's labourers who are picked up for each ship, it would be impossible to apply the standard. If the impracticability arises from any of the three causes stated, the English Act remits the ascertainment of the average weekly earnings to another test or tests. It provides that regard may be had to the average weekly amount which during the twelve months previous to the injury was being earned by a person in the same grade employed at the same work by the same employer. Again it must be the same employer. But if there is no person so employed, then the average weekly amount earned by a person in the same grade employed in the same class of work must be taken. In New South Wales this provision was at first transcribed without change, but now one of the three grounds has been dropped. The casual nature of the employment is no longer a reason on which the impracticability may be based (sec. 14 (a), proviso). It was omitted as part of what was evidently an attempt to overcome the effect of the decision of the Court of Appeal in Cue v. Port of London Authority[5]. The legislative directions for ascertaining the average weekly earnings include the following provision: - "Where the worker has entered into concurrent contracts of service with two or more employers under which he worked at one time for one such employer, and at another time for another such employer, his average weekly earnings shall be computed as if his earnings under all such contracts were earnings in the employment of the employer for whom he was working at the time of the injury" (sec. 14 (b)).

In Cue's Case[6] the workman was employed on the London Docks for the full period of twelve months before his death. He was a corn porter whose labour was casual. He was sometimes employed by the Port of London Authority and sometimes by a shipowner. In order to aggregate his remuneration from both these sources for the purpose of calculating his average weekly earnings, it was sought, but unsuccessfully, to bring him within the provision relating to concurrent contracts of service. Lord Cozens-Hardy M.R. said: "It is quite clear there were no concurrent contracts; there were successive contracts"[7]. Swinfen Eady L.J. said: "It is not a case of concurrent contracts of service. Perhaps these may be called successive contracts, because technically, even with casual employment, for the time of the actual employment there was a subsisting contract, but it was merely casual employment; that is to say, there were no running contracts under which the employer was entitled to require the labour of the workman or under which the workman was entitled to demand employment from the employer"[8]. The court held that the average weekly earnings must be ascertained on the view that it was impracticable to do so by reference only to the period of the employment by the same employer. The provision to be applied was that remitting the inquiry to a consideration of the earnings of another employee of the same grade employed in the same work.

After this decision, a new provision was introduced into the New South Wales Act. It is par. e of sec. 14. So far as material it provides: - "The average weekly earnings of a casual worker, who has worked under successive contracts of service with two or more employers in the same industry, shall be computed as if his earnings under all such contracts, for a period of twelve months preceding the injury or any less period he may have been engaged in the industry, were earnings in the employment of the employer for whom he was working at the time of the injury. Such average weekly earnings shall be deemed to be not less than the weekly living wages declared by the statutory authority to be payable in the area in which the injury occurs."

It will be seen that it applies to the casual worker, the reference to whom is now dropped from the proviso to par. a. It takes the very words of Lord Cozens-Hardy M.R., and Swinfen Eady L.J., "successive contracts of service"[9]. It repeats in relation to them the language of par. b which had been held inapplicable, viz., "shall be computed as if his earnings under all such contracts ... were earnings in the employment of the employer for whom he was working at the time of the injury."

These two paragraphs demand the making of the same assumption when the conditions which they respectively prescribe are fulfilled. The assumption is that the earnings of the worker proceeded from one and not several employers. When this is assumed, it becomes necessary to apply to the facts, as modified by the hypothesis, the principles prescribed by the other provisions of the section for the ascertainment of the average weekly earnings.

In the case of par. e there is one further qualification. The time is expressly restricted to the period during which the worker has been engaged in the industry, if that be less than twelve months. But, assuming that all his employers were one, it still remains necessary to pursue the method of calculation laid down for such a case. If it appears that intervals occurred in which he earned nothing an inquiry must still be made into the reason. If some of such intervals are ordinary incidents of the employment and some are not, the like calculation must be made as in the case of a continuing relation between a worker and a single employer.

If, because of the shortness of the time during which the supposed employment by the fictional single employer has continued or because of its terms, it is impracticable to compute the rate of remuneration, recourse must be had to the standard of what another worker earns.

In the case of such typical casual work as wharf labouring, all this causes little or no difficulty. But unfortunately what is casual employment is ill defined. Indeed it is scarcely too much to say that it seems open to a tribunal of fact to treat most forms of intermittent or irregular work as casual. Where the employment involves a contract of service lasting some weeks followed by a long interval of idleness and then another such contract of service and so on, more difficulty arises, if the view is taken that the employee is a casual worker. Such a case is before us in the present appeal.

The worker, who is the appellant, is a boiler-maker. After ten months' work out of the State, he returned to New South Wales in May 1934. He succeeded in obtaining a week's work at his trade. This was followed by nearly six weeks' idleness. Then, after a little over two weeks' work for another employer, he remained unemployed for nine weeks. He had two more weeks' work for a third employer and then eight more weeks' idleness. Another fortnight's work brought him to 17th December, when his work ended. But he resumed with the same employer on 17th January and worked two weeks. After an interval of a week's idleness, he obtained employment with the respondent. After nearly six weeks' work, he again became unemployed on 14th March 1935, the date of his injury. Evidence was given of the manner in which boiler-makers are engaged, how they are taken on as occasion arises and put off when the work is finished, and how some are given more permanent employment. The evidence showed the fluctuations in the numbers employed by the respondent during recent years when work was not plentiful. It appears that the appellant was paid the wages of a casual. In these circumstances the Workers' Compensation Commission found that he was a casual worker. I do not know that this finding really operates against him. For even so, upon the facts the ascertainment of his average weekly earnings would depend on the question formulated by Cozens-Hardy M.R. in Anslow v. Cannock Chase Colliery Co. Ltd.[10]. He said: - "In my opinion the true test is this. What were his earnings in a normal week, regard being had to the known and recognized incidents of the employment? If work is discontinuous, that is an element which cannot be overlooked."

That test would be applied if he were outside par. e. There would be this difference, however. A preliminary question would arise under par. a whether the period of employment under the last employer, the period at the end of which the injury occurred, was or was not so short or of such a temporary character that it was impracticable simply by reference to it to compute the average weekly earnings. For without par. e that period alone could be taken for the purposes of the first part of par. a. But, if this question did arise, it would, I think, upon the evidence be necessarily answered that the period was too short. The evidence shows that the worker's employment was always, so to speak, ad hoc and could not be expected to last indefinitely. The computation could not be made by means of any period which was so short that it would not reflect the intermittent character of his employment. Once this was decided, the proviso would apply. But, in any case, I do not think that we can say that the Commission was bound as a matter of law to find that the work was not casual. There is no power to review its decisions except on questions of law.

The operation of par. e upon the case excludes the limitation of the period of employment which the first part of par. a would require and enables the Commission to regard the matter as if from May 1934 the appellant had been accepting such work as was available from the respondent alone and not from several employers.

It still remained for the Commission to consider whether the proviso to par. a should be applied. But the case stated contains a finding that it was not "impracticable" to compute the rate of remuneration. This I understand to mean that the Commission considered that on the evidence the successive periods of employment and unemployment encountered by the appellant provided a reasonably correct measure of the earnings obtainable from his trade. It does not appear precisely what step the Commission next took. Perhaps, as was suggested in the Full Court, the total earnings of the appellant were divided by the number of weeks of the period. If so, I do not think that we are in a position to say that on the evidence such a course was wrong in law. The evidence shows that, for men like the appellant, a very high percentage of unemployment was incident to the trade of a boiler-maker. When the intervals during which a worker is unable to earn remuneration at his trade are all attributable to the nature of the occupation and are ordinary incidents of its pursuit, the total earnings may be divided by the number of weeks in the entire period. For this is the same thing as reducing the average earned during the weeks worked by a proportion bearing the same relation to the whole as all the weeks of idleness do to the weeks contained in the total period. The stated case is, perhaps, open to the objection that at this point it practises too great an economy in setting out precisely how the Commission proceeded with the calculation. But, at the hearing of the appeal, no point was made of this, and in relation to it there is a practical consideration which cannot be disregarded. It arises on the facts disclosed by the evidence read with the conclusions of the Commission. Whatever view we might ourselves take of the question, there can be little doubt that the Commission regarded the pursuit followed by the appellant as especially exposed to periods of unemployment. In such a view of the facts, which is open on the evidence, it does not seem to me to be probable if possible, that a rate of average weekly earnings could be adopted higher than the minimum prescribed by par. e. That minimum is the basis of the commission's order.

This is a case of considerable importance. It involves the proper construction and application of the term "average weekly earnings" in secs. 9 (1) (a) and 14 of the Workers' Compensation Act 1926-1929.

By sec. 9 (1) (a), the compensation payable by the employer to an injured worker is to include a weekly payment not exceeding two-thirds of his "average weekly earnings" for the previous twelve months, if he has been so long employed by the employer; but, if not, then for any less period during which he has been in the employment of the same employer. By sec. 9 (2), the total weekly payment (made up of the weekly payment referred to above and certain payments in respect of dependents) cannot exceed a sum equal to the "average weekly earnings" or £5, whichever is the smaller amount.

Under sec. 14 of the Act, certain rules are laid down in relation to "average weekly earnings." The fundamental principle is contained in sec. 14 (a) - that the "average weekly earnings" shall be "computed" in the manner best calculated to give "the rate per week at which the worker was being remunerated" i.e., was, at the time of his injury. The proviso to sec. 14 (a) permits of reference by the tribunal to the average earnings, first, of certain fellow employees of the injured worker and, second, of others following the same class of employment, though not employed by the same employer. But such a reference cannot lawfully be made unless "it is impracticable at the date of the injury to compute the rate of remuneration." The question whether it is impracticable is a question of fact (Twidale v. London and North Eastern Railway Co.[11]). Here the commission expressly found as follows: -

The commission was of opinion that it should apply the dominant rule prescribed in sec. 14 (a), that "average weekly earnings shall be computed in such manner as is best calculated to give the rate per week at which the worker was being remunerated"; that even if applicant and Kirby be in the "same grade," which the commission doubted (Priestley v. Port of London Authority[12]), the commission did not consider it "impracticable" to compute the applicant's rate of remuneration, and that the method of computation set out in rule e of that section was, on the facts herein, the correct method of computation to apply.

Omitting for a moment the last portion of the above conclusion (i.e., the decision to apply rule e in sec. 14), it is plain that the commission felt itself bound to apply "the dominant rule" in sec. 14 (a); therefore it was not at liberty to look at the average weekly earnings of Kirby, because it could not find that it was "impracticable at the date of the injury" to compute the rate of Doyle's remuneration.

Here let us sum up the position reached by the above analysis. Doyle was injured on March 14th, 1935, while in the employment of the respondent as a boiler-maker. He had been employed there for five weeks and four days only (a period of less than twelve months); so that, under sec. 9 (1) (a) of the Act, the inquiry is: What was Doyle's average weekly earnings at the time of his injury? In each and every week of his five weeks' employment with the respondent, Doyle received £4 17s., no more and no less, working the standard working week at the rate of pay fixed by the Commonwealth award, which has the legal force of a valid Commonwealth statute. Why did not the commission apply the Act according to its clear terms and find that Doyle's average weekly earnings at the time of his injury was £4 17s. per week? That was in fact the rate per week at which he was being remunerated. It must be remembered that what the Workers' Compensation Act is aiming at is a calculation of the average weekly "earnings," which refers to money received for work done. Except where the Act clearly requires otherwise, there seems to be no justification for the introduction into a computation of the average weekly earnings of a worker injured in the service of an employer with whom he has been employed for less than twelve months, such considerations as the worker's earnings in previous employment or his lack of earnings during periods of unemployment.

Then what is the reason for the decision of the commission to apply rule 14 (e)? Their award is as follows: -

Having duly considered the matter submitted, particularly the fact that the applicant's employment with the respondent was always casual; that the industrial award wages paid to the applicant were those fixed for hourly hiring of casual workers in his grade, and that when temporary employment was offering he worked under successive contracts of service with other employers in the same industry for short periods, the commission considered that in computing the applicant's weekly earnings, the manner best calculated to give the rate per week at which he was being remunerated on the fourteenth day of March 1935, is that fixed for casual workers in sec. 14 (e) of the Workers' Compensation Act 1926-1929.

With all respect, none of these findings justify recourse to sec. 14 (e). The appellant's employment with the respondent was only casual in the sense that he was subject to dismissal at an hour's notice. In this respect, his position was like that of tens of thousands of persons engaged in the great engineering trade who are never employed, and never seek employment, outside their chosen vocation.

The appellant's record of employment for some time prior to the accident was as follows: -

Ten months' continuous employment at Port Darwin, working for the Commonwealth Government. Doyle received £6 12s. 6d. for each week of 44 hours;

Garden Island, one week's employment; earning for 44 hours' work £4 17s.

Hodkinson (Alexandria), two weeks two days; average earnings for each week of 44 hours £4 17s.

Mort's Dock, two weeks; average earnings for each week of 44 hours £4 17s.

Hodkinson's (Alexandria), two weeks; average earnings for each week of 44 hours £4 17s.

Hodkinson's (Alexandria), two weeks, average earnings for each week of 44 hours £4 17s.

Sydney Steel Co. (respondents), five weeks and four days; average earnings for each week of 44 hours £4 17s.

During the twelve months prior to the accident on March 14th, 1935, the applicant earned only £73 14s. 3d. as a result of fifteen weeks and one day's work; giving average earnings of precisely £4 17s. per week, the Commonwealth award minimum. The employment at Port Darwin preceded the twelve months, but, during the ten months there, the applicant received about £287. The astounding result of adopting the computation under sec. 14 (e) is this: the commission divides £73 14s. 3d. by fifty-two, and attributes to a highly qualified artisan "average weekly earnings" of less than thirty shillings per week. Then this sum is lifted to the minimum of £3 7s. 6d. per week under the proviso to sec. 14 (e), solely because, at the time of the injury, the New South Wales living wage was £3 7s. 6d.

In my opinion, sec. 14 (e) could not lawfully be applied, having regard to the facts found by the commission and the other undisputed facts of the case. Doyle was not a "casual worker" at all (Cf., per Atkin L.J., Twidale's Case[13]). The only evidence was that he was a trained boilermaker who never engaged himself in any other calling or vocation. Nor did the commission find that he was a "casual worker," merely finding that his employment with the particular employer was "casual" in the sense that he could be dismissed at an hour's notice in pursuance of the court's award. But the object of sec. 14 (e) is to deal with the case of a true "casual worker," who is engaged in different trades, callings or occupations. It provides that, in such a case, the worker shall aggregate his previous earnings in the same trade over a period not exceeding twelve months, during which he has been engaged in such trade. In the result, if a casual worker works in turn (say) as gardener, cleaner and messenger, and is injured while working as a gardener, the computation of his "average weekly earnings" under sec. 14 (e) shall be made by reference to all his engagements as gardener for a period of twelve months preceding the injury, or any less period during which he worked as a gardener. The result is to give a weighted and more accurate average of earnings as gardener, because, being, ex hypothesi, a "casual worker," the worker may be employed only for several hours during each engagement. The general nature of the provision is, for computation purposes only, to associate the "casual worker," who belongs to no one trade or industry, with the trade or industry in which he was engaged when injured. Moreover, sec. 14 (e) does not say that, in making the calculation, the "casual worker" is to be debited with periods of unemployment, or that, where his earnings are to be aggregated over twelve months they are to be divided by fifty-two, so as to get average weekly earnings. But that is what the Workers' Compensation Commission has done in the present instance, and, in my opinion, the course pursued is entirely wrong. As pointed out above, the result of the process is the amazing finding that, at the time of the injury, the "average weekly earnings" of a skilled boiler-maker like the applicant, who was being paid the standard or minimum wage, was less than thirty shillings per week! In my opinion, he was not a "casual worker" at all; but, even if he was, the proper method of calculation under sec. 14 (e) would be to divide the £73 14s. 3d. earned during the twelve months prior to the accident, not by fifty-two, but by 15%p1/5%p, being the number of weeks during which the appellant was making earnings in the same "industry." A man's average earnings per week means the money received on an average during each week worked, and not during a week in which there can be no earning. On the contrary view, a cricketer's average runs per innings would be computed by dividing his total runs, not by the number of innings he completed, but by the innings he might have had if he had played in additional matches.

The conclusion I reach is not inconsistent with any decision of the House of Lords or Court of Appeal.

Although the New South Wales scheme of assessing worker's compensation by reference to "average weekly earnings" follows the general plan of the English Act, there are several important departures from the language of the latter. Accordingly, English cases have to be used with some care. In England, moreover, the special problem created by our Federal and State minimum and living wage system does not seem to have been dealt with. The case of Perry v. Wright[14] is still the leading English authority on average weekly earnings. Fletcher Moulton L.J. was at pains to show that, in certain special circumstances, the calculation must be adjusted because there is a "lack of continuity in employment" which is "inherent in the employment itself" (at p. 461), or arises "from the nature of the employment itself" (at p. 459). The examples Fletcher Moulton L.J. gives include those of (a) a person employed by one employer only, "but in a discontinuous, manner" (at p. 459), e.g. a man employed casually to work a ferry on certain occasions only; (b) a stoppage of work owing to "holidays fixed and recognized by the trade" (at p. 461), e.g. the closing of factories during Lancashire Wakes week. In Bailey's Case[15] (dealt with on the Perry v. Wright appeal[16]) Fletcher Moulton L.J. assumed that

the total of the stoppages from recognized holidays amount to two weeks, and that the remainder of the interruptions from accidents and other causes amount to one week.

In such a case, the sum total of Bailey's year's earnings was £83 2s. 1d. and that sum was earned by forty-nine weeks' work. Dividing £83 2s. 1d. by forty-nine would give a figure of £1 13s. 11d., which, according to Fletcher Moulton L.J., gives the "average wages earned in a week of full work"[17]. He then says that %p2/52%p or %p1/26%p of the figure £1 13s. 11d. should be deducted because "there is incident ... an enforced idleness of two weeks in the year"[18].

Cozens-Hardy M.R. differed from the opinion of Fletcher Moulton L.J., but Farwell L.J. did not choose between the competing views[19]. Cozens-Hardy M.R. held that Bailey's "average weekly earnings" should not be diminished by reason even of the "recognized holidays," i.e., the Wakes Week and Bank holidays (see pp. 447, 454). Cozens-Hardy M.R. added that days in which no work was done and no wages earned were to be disregarded[20]. As an absolute general rule this proposition goes too far (Greenwood v. Joseph Nall & Co. Ltd.[21]), but it is noteworthy that in Bailey's Case[22] the view of Cozens-Hardy M.R. was accepted by the parties.

But in the present case there seems to be no room for the application of the doctrine of Fletcher Moulton L.J. There is no evidence that Doyle's "average weekly earnings" whilst employed by the respondent were, or would be, subject to diminution, or be deemed to be lessened, by reason of any special nature or incident of the employment. Obviously the mere risk of future unemployment is an insufficient basis to found a conclusion that, although a worker is employed and injured, and it is practicable to compute his average weekly earnings while so employed, his average weekly earnings should be deemed reduced. By what measuring rod this risk of future unemployment would be measured I do not know. There was no evidence justifying the Workers' Compensation Commission in adopting rule 14 (e).

The appeal should be allowed, because the evidence demonstrates that, at the time of his injury, the average weekly earnings of the appellant amounted to £4 17s., the minimum wage fixed by the Commonwealth award.

The appellant is by trade a boiler-maker. On 8th February 1935 he was engaged by the respondent company, which carries on the business of steel construction. During the previous twelve months the appellant had worked for a number of employers in the engineering industry with intervals of varying lengths between each employment. On 14th March 1935 the appellant received a personal injury admittedly arising out of and in the course of his employment and resulting in an incapacity for work which lasted until 13th May 1935. The appellant as a married man with three children became entitled to compensation to the extent provided by sec. 9 (1) (a) and (b) and (2) of the Workers' Compensation Act 1926-1929, which involves the ascertainment of his average weekly earnings under the appropriate sub-section of sec. 14 of that Act.

The Workers' Compensation Commission found that the appellant was a casual worker and that sec. 14 (e) of the Workers' Compensation Act should be applied for the purpose of ascertaining his average weekly earnings. The commission fixed his compensation at the rate of £3 7s. 6d. per week. On a case stated the Supreme Court did not disturb this decision.

Now the term "casual worker" is not capable of exact definition. Hamilton L.J. said in Knight v. Bucknill[23]: "I think that casual is here used not as a term of precision, but as a colloquial term." Each case is to be determined on its own facts, consideration being given not only to "the nature of the work but also the way in which the wages are paid, or the amount of the wages, the period of time over which the employment extends, indeed all the facts and circumstances of the case" (Stoker v. Wortham[24], per Swinfen Eady M.R.). The question being one of fact, the commission's finding should not be set aside if there was evidence to support it. In my opinion there was no such evidence.

The appellant was employed as a riveter in the respondent's business. His employment was governed by an award of the Commonwealth Court of Conciliation and Arbitration. Riveting is defined by the award to be within the scope of the work of a tradesman boiler-maker. The award provides for an hourly and a weekly rate. The appellant was engaged at an hourly rate but he was engaged for an indefinite time. In fact, he worked for five weeks and four days. Engagement at an hourly rate is not a criterion of casual employment as distinct from regular employment. The award deals with overtime, Sunday work and holidays and provides for the payment of wages weekly or fortnightly. Many artisans in regular employment are engaged at an hourly rate. The standard working week fixed by the award is forty-four hours and the appellant worked for that period in every week without any interruptions. The respondent regularly employed boiler-makers for the purposes of its trade, and, although the occasion for the appellant's employment was to cope with an increase of work which may or may not have been maintained, the appellant was engaged for an indefinite time on terms applicable to all such artisans in the respondent's employment. Where a skilled tradesman obtains employment at his trade with an employer who regularly carries a staff of artisans belonging to that trade and is employed on terms applicable to all such employees, I find it difficult to say that he is a casual worker, although the occasion of his employment may be the receipt of more orders by the employer. I find nothing in the evidence in the present case which detracts in any way from the prima facie conclusion that the appellant, the nature of whose employment is exactly described by these observations, was not a casual worker. The fact that in the course of the year he had been engaged by other employers in the industry as a boiler-maker and that there were intervals between such periods of employment does not affect this conclusion. There was nothing in the evidence which would, in my opinion, justify the application of the term "casual" to the relations existing between the appellant and the respondent (Cf., per Atkin L.J., Williams v. Haigh[25]).

It follows from this view that sec. 14 (e) was not applicable and that the appellant's average weekly earnings should be computed according to the dominant rule in sec. 14 (a). The wages which the appellant earned was the normal remuneration for a boiler-maker employed during the time that he was working for the respondent.

The aggregate of his wages earned during his fifteen weeks of employment was not swollen by any abnormal payments for overtime or Sunday work or the like. To take the number fifty-two as the divisor for calculating his average weekly earnings is to assume that it was a normal incident of the employment of a boiler-maker to be disengaged for thirty-seven weeks of the year, that figure being arrived at by deducting the total number of weeks during which he was engaged from fifty-two weeks. There is no evidence to justify such an assumption. The evidence does not establish what, if any, of the appellant's idle time was attributable to the normal incidents of his calling, and the respondent has accordingly failed to show that any greater divisor should be adopted than the number of weeks during which the appellant was working for it in order to calculate his average weekly earnings at the time of the accident. There is no period of time calculable by the evidence of which it may be fairly said that the conditions of the industry would compel a boiler-maker to be idle for that period.

Solicitors for the appellant, Rosendahl & Devereux.

Solicitors for the respondent, P. V. McCulloch & Buggy.

[3] (1908) 1 K.B., at pp. 459-462, 465, 466.

[4] (1909) A.C. 435; (1909) 1 K.B. 352.

[9] (1914) 3 K.B., at pp. 897, 900.

[13] (1925) 2 K.B., at pp. 471, 472.

[22] (1908) 1 K.B., at p. 466, footnote 1.

[23] (1913) 6 B.W.C.C., at pp. 164, 165.

[24] (1919) 1 K.B. 499, at pp. 503, 504.

[25] (1925) 18 B.W.C.C. 549, at p. 555.