(Delivered by Patanjali Sastri, J.)

This reference arises out of an assessment to Income-Tax and excess-profits duty made on the respondent as the manager of a Hindu undivided family for the year 1940-1941.

The family was carrying on business at Kakatiruppupudur (Ramnad District) in British India and in Ceylon, and its income from both these sources has been taxed on the footing that the family was “resident” and “ordinarily resident” in British India. In each case the income taxed was of the “previous year,” which was the Tamil year Pramathi ending 12th April 1940 for the Indian business and the financial year ending 31st March 1940 for the foreign business, the assessee's accounts in the two places having been made up to those dates respectively. There was no dispute that the family was “resident” in British India within the meaning of S. 4(1)(b) read with S. 4-A of the Indian Income-Tax Act, 1922, as amended by the Income-Tax (Amendment) Act, VII of 1939. The assessee, however, contended that the family was “not ordinarily resident” in British India within the meaning of the second proviso to S. 4(1) read with S 4-B and that accordingly, the income arising to it from the business in Ceylon should not be included in the assessment, there being no suggestion that such business was controlled in India, or that such income was brought into British India.

Now, a Hindu undivided family is deemed to be “ordinarily resident” in British India if its manager is ordinarily resident in British India [4-B(b)], and an individual is “not ordinarily resident” in British India “in any year if he has not been resident in British India in nine out of the ten years preceding that year or if he has not during the seven years, preceding that year been in British India for a period of, or for periods amounting in all to, more than two years” [4-B(a)]. The assessee's attempt before the Income-Tax authorities to prove that his family was not resident in British India in nine out of the ten years prior to the year of account having proved unsuccessful, he shifted his position before the Income-Tax Appellate Tribunal (Madras Bench) and sought to bring the case under the latter part of the Cl. (a), contending that the “seven years” referred to in that clause must be taken to be seven calendar years. The Tribunal ordered a fresh inquiry and it was found (i) that the manager of the family was in British India for 541 (this was subsequently corrected into 606) days on the aggregate during the period of the seven calendar years from 1st January 1932 to 31st December 1938 preceding the year of account, whether of the Ceylon or the Indian business, (ii) that he was in British India for 731 days on the aggregate during the seven Tamil years preceding Pramathi, the year of account of the Indian business; and (iii) that he was in British India for 731 days on the aggregate during the seven financial years preceding the financial year 1939-1940 which was the year of account of the Ceylon business. It will thus be seen that if the “seven years” mentioned in Cl. (a) be taken as seven calendar years, the manager cannot be said to have been in British India during that period for more than two years and the assessee's joint family would be entitled, under the second proviso to S. 4(1) as a person “not ordinarily resident” in British India to claim the exclusion of the foreign income from the assessment. The claim was accepted by the Tribunal and the assessment was ordered accordingly to be reduced. The Commissioner of Income-Tax having challenged the correctness of that view, the Tribunal has referred the following question to this Court for its decision—

“Whether in the circumstances of the case, in computing the period or periods when the manager of the Hindu undivided family had not been in British India, the seven years in S. 4-B of the Income-Tax Act should be taken as seven Calendar years or seven previous years.”

It may be mentioned here that the assessee raised before the Tribunal an alternative contention, viz., that, even if the aggregate period of the manager's stay in British India during the relevant years be taken as 731 days, it cannot be said to be “more than two years” as the seven years, however reckoned, must include at least one leap year. But the Tribunal did not decide the point and it is not before us.

It is necessary, in order to appreciate the contentions of the parties, to describe in brief outline the scheme of taxation laid down in the Act so far as it is material here. The tax is levied for each financial year commencing on the 1st of April at the rate or rates prescribed in the Finance Act in force for the time being and is charged on the “total income of the previous year (S. 3). The “previous year” may be either the financial year next preceding or, at the option of the assessee, any other period of twelve months ending within such year if he has made up his accounts for such period. It was held that an assessee could not have two separate “previous years” for the purpose of Income-Tax [Commissioner of Income-Tax, Bombay v. Abubaker Abdul (1)] but this view was superseded by the Amending Act, and it is now possible for an assessee to have a different “previous year” for each separate source of income. [S. 2(ii)]. “Total income” is the total amount of income, profits and gains referred to in Sub-S. (i) of S. 4 computed in the manner laid down in the Act [S. 2(15)]. S. 4(1) refers to four classes of income and provides for their assessment on a basis which differs according as the assessee is or is not “resident” or “ordinarily resident” in British India during the previous year of which the income falls to be assessed. It is, accordingly, with reference to that year, that the various tests of “residence” or “ordinary residence” as defined in S. 4-A and S. 4-B have to be applied. The question is whether the “seven years preceding that year” mentioned in S. 4-B(a) refers to the seven years ending on that day next preceding the commencement of the previous year or to the period of seven years ending on the 31st December preceding the commencement of such year as suggested for the assessee.

Mr. Srinivasan appearing for the assessee relies on S. 3(59) of the General Clauses Act (X of 1897) which provides that in all Central Acts and Regulations made after the commencement of that Act, unless there is anything repugnant in the subject or context, “year” shall mean “a year reckoned according to the British Calendar.” As the Calendar (new Style) Act, 1750, (24 Geo. 2 C. 23) transferred the beginning of the year in England from the 25th of March to 1st of January in and after 1752, it is urged that, in applying S. 4-B(a) the seven years must be reckoned commencing from the 31st December preceding the year of account and counting backward. We cannot agree. It is obvious that this mode of computation would result in a gap, an interregnum, before the commencement of the “previous year” except, of course, where that year also began from the 1st of January. Such a result is repugnant to the intendment of the provisions of S. 4-B(a) read with S. 4(1) under which the terminus ad quem is the commencement of the “previous year.” We find nothing in S. 50 of the Act to which reference was made in the course of the argument to support the interpretation contended for on behalf of the assessee.

It was said that if the “seven years” mentioned in S. 4-B(a) were taken as referring to the period immediately preceding the year of account, it would give rise to the anomalous position of one and the same person being liable to assessment as “resident” in respect of some of his sources of income and as “not resident” in respect of others according to the “previous year” he has adopted for the respective sources. This argument, appears to have weighed very much with the Tribunal, but we see no force in it. It is true, as has been stated, that under S. 2(ii) as amended it is possible for an assessee to have as many “previous years” as he has separate sources of income. But, as we have pointed out already, under the scheme of taxation laid down in the Act the charge is made on the total income of the “previous year,” and such income has to be computed on the basis of the assessee's residence or non-residence in British India during that year. That is to say, in respect of each separate source of income the year of account is to be ascertained, and if it is found on applying the provisions of Ss. 4-A and 4-B with reference to that year that the assessee was “resident” or “not resident” or “not ordinarily resident” in British India as the case may be, the income of that year is to be computed on the appropriate basis indicated, and the total of the amounts thus computed for all the sources of income, whether British Indian or Foreign, is to be charged to tax. Such procedure may prove somewhat cumbrous but not unworkable in its application to particular cases where the assessee has adopted different years of account for his separate sources of income, and it is possible in such cases that he should be assessed as resident in respect of some of the sources and non-resident in respect of others. But this, in our opinion, is no reason for disregarding a construction to which the language of the provisions in question plainly points. And, after all, the construction contended for on behalf of the assessee, as we understand it, does not help to resolve the supposed anomaly as, even on such construction, an assessee may, in conceivable cases, have to be dealt with as resident and non-resident in respect of his different sources of income.

We are, therefore, of opinion that the expression “seven years” in S. 4-B of the Indian Income-Tax Act should be taken as referring to the period of seven years of twelve Calendar months each immediately preceding the commencement of the relevant previous year, and we answer the reference in that sense. The assessee will pay Rs. 250 as costs of the Commissioner of Income-Tax.

N.R.R