Rajagopalan, J.:— These two references, one under section 66(1) and the other under section 66(2) of the Income-tax Act, arose out of proceedings of the assessment years 1947-48 and 1948-49 to assess the income of the assessee company from its life insurance business which it carried on in the corresponding years of account 1946 and 1947.

2. As directed by section 10(7) of the Act, the profits of the assessee company had to be assessed each year either under rule 2(a) or under rule 2(b) of the Schedule to the Act, whichever was more favourable to revenue.

3. The assessee company commenced its business in 1935. The first actuarial report was for the period May 24, 1935, to March 14, 1941, and it showed a surplus of Rs. 5,131-5-1. The assessments for the years of account between 1935 and 1940 had necessarily to be only on the application of rule 2(a). There was and there could have been no actuarial report for any antecedent period, without which there could be no recourse to rule 2(b). The next inter-valuation period for which an actuarial report was obtained was from March 14, 1941, to December 31, 1945. That showed a deficiency which, when adjusted under rule 2(b) to furnish the basis for assessment to income-tax, was ascertained as Rs. 86,708. The average annual deficiency for that period 1941 to 1945 worked out to Rs. 18,096. The second actuarial valuation report could furnish a basis for assessing the income of the assessee company under rule 2(b) only for 1945 and not any of the antecedent years. Subsequent to 1945 there were annual actuarial valuations for 1946 and 1947. The assessments for the years of account 1941 to 1944 were based on the first valuation report for the period that ended on March 14, 1941. That showed a surplus of Rs. 5,131. The computation of the surplus under rule 2(b) worked out to Rs. 7,280, and the annual average surplus was Rs. 1,238. Under Tule 2(b) therefore Rs. 1,238 was the computed profit for each of the years 1941 to 1944, and it was on that basis the assessments were completed for the relevant assessment years 1942-43 to 1945-46. That the subsequent actuarial valuation for the same period included in the inter-valuation period 1941 to 1945 revealed a deficiency did not affect the validity or the finality of the assessments already completed for the assessment years 1942-43 to 1945-46. In other words, though there was a deficiency or loss in the years of account 1941 to 1944, which the actuarial report disclosed in 1946 the application of rule 2(b) resulted in the company being taxed on an annual profit of Rs. 1,238 computed under that rule.

4. The assessment of the income for 1945 was on the basis of the second actuarial valuation report, and under rule 2(b) the average of Rs. 18,096 was computed as the loss for 1945. That was more favourable to revenue than the computation of the loss for that year under rule 2(a).

5. When the income for the year 1946 had to be assessed in the assessment year 1947-48, the valuation report for 1946, the third valuation report, was available. The actuarial deficiency disclosed by the second report for the period that ended on December 31, 1945, was Rs. 90,851, and the deficiency at the end of 1946 as disclosed by the third report was Rs. 77,165. Therefore there was an actuarial surplus of Rs. 13,686 for 1946. When the assessment was first completed in 1949 the Income-tax Officer committed a mistake. Instead of deducting Rs. 77,165 from Rs. 90,851, which would have resulted in an actuarial surplus of Rs. 13,686 for 1946 he added both the sums and arrived at a deficiency of Rs. 1,68,016. The computation of the loss in that manner was more favourable to revenue than a computation under rule 2(a).

6. A similar mistake was committed when the income for 1947 was first assessed in 1949 for the assessment year 1948-49. The actuarial deficiency at the end of 1947, for which there was a separate actuarial valuation, was Rs. 32,124. Instead of deducting this from Rs. 77,165 which would have shown an actuarial surplus of Rs. 44,951 for 1947, the Income-tax Officer added both the figures and computed an actuarial deficiency of Rs. 1,09,379. That again was more than the loss computed under rule 2(a) in 1947.

7. The Income-tax Officer subsequently discovered the mistakes in the computation, and on November 17, 1951, he issued notice to the liquidator, who represented the assessee company, which had meanwhile been ordered to be wound up, to show cause why the errors should not be rectified and the assessment revised under section 35 of the Act. The liquidator objected. The Income-tax Officer dropped the idea of applying section 35 and he issued notices under section 34 of the Act on December 29, 1951.

8. The assessee company contended that the initiation of the proceedings under section 34 was not valid. The further contention of the assessee was that, even if the finality of the assessments could be validly reopened under section 34, the assessee was entitled to set off under section 24 the loss of Rs. 86,708, which represented the total deficiency at the end of 1945 computed under rule 2(b) on the basis of the second actuarial report for the inter-valuation period which ended on December 31, 1945.

9. Both the contentions were negatived by the Income-tax Officer. The appeals the assessee successively filed to the Appellate Assistant Commissioner and the Tribunal failed. The Tribunal upheld the validity of the reassessment under section 34. The Tribunal also upheld the contention of the Department, that all the loss the assessee was entitled to carry forward at the end of 1945 was Rs. 18,096, against Rs. 16,308, the profits for 1946 computed under rule 2(b), leaving an unabsorbed loss of Rs. 1,788. Rs. 1,788 was deducted from Rs. 47,779, the profits for 1947 computed under rule 2(b), and the assessee was taxed on the balance.

10. The questions that were referred to this court were:

“1. Whether the reopening of assessments of 1947-48 and 1948-49 under section 34 is legal?

2. Whether the refusal to carry forward the unabsorbed loss of Rs. 86,708 and to set off the same against the profits in the year of assessments 1947-48 and 1948-49 is valid?”

11. The comparatively wide scope of section 34, with the use of the formula “if for any reason”, was restricted when it was amended in 1939, requiring a correlation of the discovery of escape from assessment to definite information which had come into the possession of the Income-tax Officer. What section 34(1)(b) as amended in 1948 required was that the Income-tax Officer should, in consequence of information in his possession, have reason to believe that the income had escaped assessment.

12. Section 35 permits rectification where the mistake is apparent on the face of the record of assessment. The learned counsel for the Department could not challenge the correctness of the plea of the assessee, that the original assessments completed in 1949 for both the assessment years under consideration now could have been rectified under section 35. The mistakes were apparent on the face of the assessment orders themselves. What should have been subtracted was added to ascertain the deficiency disclosed by the actuarial reports. It should be remembered that the Income-tax Officer first contemplated recourse to section 35 but dropped it as the assessee did not agree.

13. The further contention of the learned counsel for the assessee was that where rectification was permissible under section 35, recourse to section 34 was barred: they were mutually exclusive.

14. In Commissioner of Income-tax v. Khemchand Ramdas [1938] 6 I.T.R 414, 426. their Lordships of the Privy Council observed:

“In their Lordships' opinion the provisions of the two sections (sections 34 and 35) are exhaustive, and prescribe the only circumstances in which and the only time in which such fresh assessments can be made and fresh notices of demand can be issued. In the present case it is a debatable question whether the circumstances were such as to bring it within the provisions of section 34. It is not necessary to determine that question inasmuch as, in their Lordships' opinion, the case clearly would have fallen within the provisions of section 35 had the Income-tax Officer exercised his powers under the section within one year from the date on which the earlier demand was served upon the respondents… The Income-tax Officer took no further step, however, until May, 1929, and by then he was hopelessly out of time whichever of the two sections was applicable.”

15. Their Lordships of the Privy Council did not decide the question whether in a given case section 34 and section 35 could be viewed as alternative remedies open to the Department, the choice being left to the Department, or whether they were mutually exclusive. In fact their Lordships had no occasion even to consider whether the two sections were mutually exclusive.

16. This passage in Commissioner of Income-tax v. Khemchand Ramdas was referred to by the Supreme Court in Maharana Mills (Private) Ltd. v. Income-tax Officer, Porbandar [1959] 36 I.T.R 350, 358.. But there there was no occasion then either for their Lordships of the Supreme Court to decide or even discuss whether section 34 and section 35 were mutually exclusive in their operation.

17. In Commissioner of Income-tax v. D.R Naik [1939] 7 I.T.R 362, 367. Beaumont, C.J, observed:

“It is also suggested that this is really a mistake which ought to have been remedied under section 35; but even if that be so, the fact, that a mistake might be remedied under section 35, is no reason why the assessment should not be altered under section 34, if the case falls within that section. I see no reason for supposing that sections 34 and 35 are mutually exclusive.”

18. Section 34 that was considered in that case was as it stood before its amendment in 1939 which permitted the reopening of assessments under the wider formula “if for any reason”. That decision is not direct authority on the scope of sections 34 and 35 as they now stand.

19. The real contention of the learned counsel for the assessee that we have to consider is that neither the test of “definite information that had come into the possession of the Income-tax Officer” nor the requirement of “information in his possession” that is prescribed by the 1948 amendment of section 34, can be satisfied, unless the information is extraneous to the record of the original assessment which the Income-tax Officer seeks to reopen under section 34. Another form in which the same contention was put forward was that a mistake apparent on the face of the order of assessment itself, as in this case, could not constitute information within the meaning of section 34, and that what is already on record, especially if it is a mistake readily discoverable by a mere scrutiny, cannot be information which has come into the possession of the Income-tax Officer, or information in his possession within the meaning of section 34.

20. The learned counsel for the assessee sought support for this contention in the observations of their Lordships of the Supreme Court in Maharaj Kumar Kamal Singh v. Commissioner of Income-tax [1959] 35 I.T.R 1, 14;. Referring to Benoy Kumar Sahas Roy…Applicant; v. Commr. Of Income-Tax, West Bengal…. [1953] 24 I.T.R 70. their Lordships pointed out that in that case it was held that information as to the true state or meaning of the law derived freshly from an external source of authoritative character was definite information within the meaning of section 34. Their Lordships proceeded:

“It appears that in construing the scope and effect of the provisions of section 34, the High Courts have had occasion to decide whether it would be open to the Income-tax Officer to take action under section 34 on the ground that he thinks that his original decision in making the order of assessment was wrong without any fresh information from an external source or whether the successor of the Income-tax Officer can act under section 34 on the ground that the order of assessment passed by his predecessor was erroneous, and divergent views have been expressed on this point. Mr. Rajagopala Sastri, for the respondent, suggested that under the provisions of section 34 as amended in 1948, it would be open to the Income-tax Officer to act under the said section even if he merely changed his mind without any information from an external source and came to the conclusion that, in a particular case, he had erroneously allowed an assessee's income to escape assessment. We do not propose to express any opinion on this point in the present appeal.”

21. We should like to emphasise even at the outset that we are not dealing with a case of a change of opinion on the part of the assessing authority, but with an error in computation obvious on the face of the order of assessment itself. That the real deficiency or surplus in each of the relevant years was to be ascertained by subtraction and not addition of the deficiency for each of the two years could never be challenged. Whether the discovery of such a mistake, which in its turn led to the further discovery of escape from assessment or under-assessment, constitutes information within the meaning of section 34(1) is what we have to decide in this case. We do not propose to cover any wider ground.

22. We are unable to accept the extreme proposition, that nothing that can be found in the record of the assessment, which itself would show escape of assessment or under-assessment, can be viewed as information which led to the belief that there has been escape from assessment or under-assessment. Suppose a mistake in the original order of assessment is not discovered by the Income-tax Officer himself on further scrutiny but it is brought to his notice by another assessee or even by a subordinate or a superior officer, that would appear to be information disclosed to the Income-tax Officer. If the mistake itself is not extraneous to the record and the informant gathered the information from the record, the immediate source of information to the Income-tax Officer in such circumstances is in one sense extraneous to the record. It is difficult to accept the position that while what is seen by another in the record is “information” what is seen by the Income-tax Officer himself is not information to him. In the latter case he just informs himself. It will be information in his possession within the meaning of section 34. In such cases of obvious mistakes apparent on the face of the record of assessment, that record itself can be a source of information, if that information leads to a discovery or belief that there has been an escape of assessment or under-assessment.

23. The real question is not whether section 34 and section 35 are mutually exclusive in their operation, but whether in a given case, the statutory requirements are satisfied. If in a given case the requirements of both section 34 and section 35 are satisfied, the Income-tax Officer can have recourse to either. That in such a case there is overlapping will not bar recourse to either section at the choice of the assessing authority.

24. So the real question is whether the requirements of section 34 and in particular the requirements of information were satisfied in this case. We see no justification to accept the contention of the learned counsel for the assessee that to constitute information within the meaning of section 34, it must be wholly extraneous to the record of the original assessment. We hold that the mistake apparent on the face of the order of assessment itself constitutes information: whether someone else gave that information to the Income-tax Officer or whether he informed himself is immaterial. We are further of opinion that, in the circumstances of this case, the availability of the powers vested in the Income-tax Officer by section 35 did not bar recourse to the jurisdiction vested in him by section 34. The initiation of proceedings under section 34 was, in our opinion, valid.

25. The second question is, what is the quantum of the loss that the assessee company could carry forward when its income for 1946 and 1947 had to be reassessed. Section 24 is not one of the sections excepted by section 10(7) of the Act, and the statutory right to carry forward the loss could not be and was never denied. It was the extent of the relief permissible under section 34 that was in controversy.

26. The profits or losses of an assessee have to be ascertained with reference to each year of his account. Before the profits of any year are ascertained the losses of the previous years have to be set off or deducted within the limits prescribed by section 24. Did the assessee company sustain loss in the years that preceded 1946 and what was the extent of that loss have to be determined first.

27. In the case of the assessee the profits and losses in each of the years that preceded 1946 were computed and ascertained in the course of the assessment proceedings of the relevant assessment years. That computation was, and had to be on the application of rule 2(a) or rule 2(b), at the discretion of the Department, the Department being given the right to choose the method of computation more favourable to revenue. We have already pointed out that the assessment for each of the years of account 1941 to 1944 was based on the first actuarial report for the inter-valuation period that ended on March 14, 1941. The factual position was as follows. In 1940 there was an assessed loss of Rs. 2,870, assessed under rule 2(a), which the assessee was allowed to carry forward. The profits of 1941 were assessed again under rule 2(a) (Rs. 1,726) which was more favourable than the assessment under rule 2(b). That left an unabsorbed loss of Rs. 844, which was set off towards the assessed profits of 1942. The profits of each of the years 1942, 1943 and 1944 were assessed under rule 2(b), and it had necessarily to be the same, Rs. 1,238, in each of those years, because that was the annual average of the actuarial surplus of the preceding inter-valuation period that had ended on March 14, 1941, adjusted under rule 2(b). There was no assessed loss to be carried forward in 1943, 1944 or 1945. When the income of 1945 had to be assessed, the second actuarial report furnished the basis of assessment and, as we have pointed out already, Rs. 18,096 was assessed as the loss of that year under rule 2(b). That that was the correct computation of the loss for 1945 could not admit of any doubt. The assessee was not entitled to treat the entire deficiency disclosed by the second actuarial report for the period that ended on December 31, 1945, as loss sustained in the year of account 1945. The deficiency adjusted under rule 2(b) was Rs. 86,708, but only the annual average of Rs. 18,096 could be taken as loss for one year, 1945, even as Rs. 1,238, the annual average of the surplus of the preceding period, was taken as the annual profits of each of the three years that preceded 1945.

28. What has to be borne in mind is that what is computed as profits and losses of a life insurance business for purposes of assessment of income-tax may have no relation to the actual trading profits or losses as disclosed in the balance-sheet. The assessment had to be under rule 2(a) or 2(b). In either case it is a notional profit or notional loss, computed only for purposes of assessment to tax. But that is the only basis for assessment. The assessment of profits and losses for each of the years 1941 to 1944 was, and necessarily had to be, with reference to the actuarial report of the period that preceded 1941, whatever were the actual trading profits as disclosed in the balance-sheets of those years. What happened in this case was that, while the actuarial report for the period 1941 to 1945 disclosed a deficiency, that could not be, and was not, the basis of assessment for any of the years 1941 to 1944. It was on a profit computed under rule 2(b) that the assessee was liable to be assessed and was in fact assessed up to the end of 1944. That notional profit could not be converted into a loss, again a notional loss, on the basis of the actuarial report, which could not be the basis of the assessment for any of the accounting years to the end of 1944.

29. There was no loss that could be assessed as a loss under the special rules of computation applicable to the life insurance business in 1943 or 1944. There was nothing to be carried forward as a loss when computing the assessable income of 1945. The loss for 1945 was correctly computed under rule 2(b), and that loss of Rs. 18,960, alone was available to be carried forward in assessing the income of 1946. No doubt section 24 allowed the loss of more than one year to be carried forward; but in this case “factually” there were assessable profits in 1943 and 1944, which meant there was no loss to be carried forward to 1945, to be set off against the assessable income of that year.

30. The learned counsel for the Department pointed out that the assessment of profits and losses for each of the years that preceded 1946 had become final, and he relied on section 24(3) to bar the acceptance of the contention of the learned counsel for the assessee that the entire loss during the period that preceded 1946 should be allowed to be carried forward. It is really unnecessary to pronounce any concluded opinion of ours on the scope of the finality of the assessment of a loss communicated to an assessee under section 24(3). We have said that though there was a computed deficiency of Rs. 86,708 at the end of the period that ended on December 31, 1945, the loss that the assessee could carry forward was only the computed loss of 1945, viz., Rs. 18,096.

31. We answer both the questions in the affirmative and against the assessee. The assessee will pay the costs of the Department in R.C No. 100 of 1956. Counsel's fee Rs. 250.

32. Questions answered in the affirmative.