Survey evidence and account defects justified rejecting books and estimating turnover for later assessment years.

A survey in one assessment year was held relevant for other years only where the discovered material, with surrounding account defects, showed unreliability extending beyond the survey year; a single isolated suppression could not by itself support a presumption for later years. The court also noted that account books may be rejected when defects are substantial, including unvouched purchases and sales not commensurate with stock. Once the books are rejected, turnover may be estimated using commercial considerations, past record, accounting method, general reputation, and stock position. On that basis, rejection of the books and estimation for the relevant years were upheld.

---

---

Issues: (i) Whether a survey report made in one assessment year could be relevant for assessment years other than the year of survey; (ii) whether suppression detected in one year could, by itself, justify an inference of suppression in other years; and (iii) whether there was material to reject the assessee's account books for assessment years other than the year of survey.

Issue (i): Whether a survey report made in one assessment year could be relevant for assessment years other than the year of survey

Analysis: The relevant factor was not the date of the survey but the nature of the material discovered in it. If the material disclosed suppression relating only to the year of survey, it would not automatically be relevant to other years. If, however, the material and surrounding circumstances showed defects or suppression extending to other years, the survey could be taken into account for those years. On the facts, the survey was directly relevant to the assessment year in which the suppression was found, but not, by itself, for the other years.

Conclusion: The survey report was not relevant for the years other than the year in which it was made.

Issue (ii): Whether suppression detected in one year could, by itself, justify an inference of suppression in other years

Analysis: A single instance of suppression in one assessment year could not, standing alone, justify a presumption that the assessee had suppressed sales in all other years. Such an inference required additional circumstances showing that the defect was not isolated and that the accounts for the other years were also unreliable. In the present case, the court found other defects in the accounts for the remaining years, including unvouched purchases and sales not commensurate with stock, which made the earlier suppression relevant in combination with those defects.

Conclusion: Suppression detected in one year could not by itself justify the inference for other years, but it became relevant when read with the other defects found in the accounts.

Issue (iii): Whether there was material to reject the assessee's account books for assessment years other than the year of survey

Analysis: The account books were found defective on more than one ground. Purchases were unvouched in several years, the extent of unvouched purchases was substantial in some years, and the sales figures were not commensurate with the stock position. Once the books were found unreliable, rejection was justified and turnover had to be estimated on the basis of surrounding commercial considerations, including past record, method of accounting, general reputation, and stock position.

Conclusion: There was material to reject the assessee's account books for the years other than the year of survey.

Final Conclusion: The references were answered substantially against the assessee, and the findings sustaining rejection of the books and estimation of turnover for the relevant years were upheld.

Ratio Decidendi: A survey conducted in one year is relevant for other assessment years only when the material discovered, together with other defects in the accounts, shows unreliability extending beyond the year of survey; once account books are rejected, turnover may be estimated on commercial and stock-based considerations.

---