Ravula Subba Rao and Another vs The Commissioner of Income Tax, Madras

Rewritten Version Notice: This is a rewritten version of the original judgment.

Court: Supreme Court of India

Case Number: Not extracted

Decision Date: 9 May, 1956

Coram: Natwarlal H. Bhagwati, T.L. Venkatram Ayyar, Sudhi Ranjan

In this case the petitioners, Ravula Subba Rao and another, challenged the order of the Commissioner of Income‑Tax, Madras. The judgment was delivered on 9 May 1956 by a bench consisting of Justice Natwarlal H. Bhagwati, Justice Aiyyar T. L. Venkatarama, Justice Das, Sudhi Ranjan (Chief Justice) and Justice Bhagwati, Natwar‑lal H. The petition was cited as 1956 AIR 604 and 1956 SCR 577. The dispute centered on the application of the Indian Income‑Tax Act, 1922 (Act XI of 1922), in particular sections 26‑A and 59 together with Rules 2 and 6 framed under section 59. The issue before the Court was whether the word “personally” used in those Rules excluded a duly authorised agent of a partner from signing an application on the partner’s behalf under section 26‑A, and whether the Rules were ultra vires the rule‑making authority vested by the Act.

The Court held that the term “personally” in Rules 2 and 6 of the Income‑Tax Rules, as framed under section 59, indeed required the actual partner to sign the application and therefore barred a duly authorised agent from signing on the partner’s behalf. The Court further concluded that the Rules were not ultra vires the rule‑making power because, on a true construction, the Act intended that an application under section 26‑A must be signed by the partner himself. In reaching this conclusion the Court examined not only the language of section 26‑A but also the character of the legislation, the overall scheme of the Act, and the nature of the right conferred by the provision. The Court emphasized that the Indian Income‑Tax Act is a self‑contained code that exhaustively covers the matters within its ambit and that its provisions demonstrate an intention to depart from the common law principle “qui facit per alium tacit per se.” Accordingly, the benefit under section 23(5)(a) was to be available to a firm only if the firm was registered under section 26‑A in compliance with the conditions laid down in that section and the accompanying Rules, which expressly require the partner’s personal signature. Consequently, a signature by an agent on the partner’s behalf was held to be invalid. The Court relied upon the earlier decisions in Commissioner of Agricultural Income‑Tax v. Keshab Chandra Mandal, [1950] S.C.R. 435 and Commissioner of Income‑Tax v. Subba Rao, [1947] I.L.R. Mad. 167, among other authorities.

The appeal arose under the civil appellate jurisdiction, being Civil Appeals Nos. 56 and 57 of 1954. These appeals were filed against the judgment and order dated 25 March 1951 of the Madras High Court in cases referred to as No. 32 of 1948 and No. 31 of 1950. Counsel appearing for the appellants included K.S. Krishnaswami Iyengar, while counsel for the respondent was also listed in the record. The Court proceeded to examine the factual background of the partnership, the registration under section 26‑A, the subsequent renewal application signed by an agent, and the legal implications of the statutory and regulatory requirements, ultimately delivering its holding as described above.

The counsel for the appellants was K. R. Choudhry together with an associate, while the counsel for the respondent comprised G. N. Joshi and P. G. Gokhale. The judgment was delivered on 9 May 1956 by Justice Venkatarama Ayyar. The appellant before the Court was a partnership firm that had been created by a deed of partnership dated 10 February 1941. The firm consisted of two partners, namely Subba Rao and Hariprasada Rao. The partnership applied for registration under section 26‑A of the Indian Income‑Tax Act No. XI of 1922 and obtained registration on 21 March 1942 for the assessment year 1942. Subsequent to that registration, Subba Rao is reported to have undertaken a long pilgrimage, during which the business affairs of the firm were administered by Hariprasada Rao, who acted as his agent under a general power of attorney executed on 1 July 1940. Acting in that capacity, Hariprasada Rao filed an application under rules 2 and 6 framed under section 59 of the Act for renewal of the registration certificate for the year 1942‑43. In the application he signed in his own name and also signed as the attorney of Subba Rao. The relevant rules state that an application for registration of a firm under section 26‑A, and for renewal of the registration certificate, “shall be signed personally by all the partners”. The Income‑Tax Officer rejected the renewal application on the ground that it had not been signed personally by Subba Rao and that the signature of Hariprasada Rao as an agent was not valid. The rejection was taken in appeal and ultimately referred under section 66(1) of the Act to the High Court of Madras. The High Court held that the term “personally” in rule 6 required the partner himself to sign the application and that general agency principles could not be invoked to satisfy that requirement, relying on the decision in Commissioner of Income‑Tax v. Subba Rao.

While the reference was pending, Hariprasada Rao filed two further applications for renewal of the registration certificate for the assessment years 1943‑44 and 1944‑45, the latter applications forming the basis of the present appeals. In each case the applications were signed by Hariprasada Rao both in his own capacity and as attorney for Subba Rao. During the hearing of these petitions the appellant contended that rules 2 and 6, when properly construed, did not exclude a partner’s authorised agent from signing on the partner’s behalf. The appellant also raised the additional argument that the rules themselves exceeded the authority of the rule‑making body and were therefore ultra vires. The Income‑Tax Officer rejected both contentions and dismissed the applications. Those dismissals were affirmed on appeal by the Appellate Assistant Commissioner and subsequently by the Appellate Tribunal. The appellant then sought a reference to the High Court, and the Tribunal framed two questions for determination: (1) whether the word “personally” in the Income‑Tax Rules framed under section 59 of the Act excludes a duly authorised agent of a partner from signing an application on the partner’s behalf under section 26‑A; and (2) if the answer to the first question is affirmative, whether rules 2 and 6 are ultra vires the rule‑making authority.

In the reference, the two judges Satyanarayana Rao and Viswanatha Sastry heard the matters. After considering the earlier decision in Commissioner of Income‑tax v. Subba Rao, they answered the first of the two certified questions affirmatively, holding that the term “personally” in the rules excluded a signature by an authorised agent of a partner. On the second question, the judges were not of the same opinion. Justice Satyanarayana Rao concluded that Rules 2 and 6 were beyond the authority of the rule‑making body, and therefore the applications filed under those rules were proper and should have been allowed. Justice Viswanatha Sastry expressed the opposite view, holding that the same rules were within the power of the authority that framed them and that the applications had been correctly rejected for not complying with the rules. Despite their disagreement on the merits, both judges granted a certificate under section 66‑A of the Act, and on that basis the appeals were brought before this Court. The Madras High Court, relying on Commissioner of Income‑tax v. Subba Rao, had previously answered the first question in the affirmative. That decision was later cited with approval in Commissioner of Agricultural Income‑tax v. Keshab Chandra Mandal, where the Court considered a rule under the Bengal Agricultural Income‑tax Act that required the individual making a return to sign it himself. The Court in that case held that the rule indeed required a personal signature. In the present appeal, counsel for the appellant, Sri K. S. Krishnaswami Ayyangar, did not raise any argument that would depart from the earlier holdings, and consequently this Court is bound to agree that the signature required by the rules must be that of the partner himself and cannot be satisfied by an agent signing on the partner’s behalf. The central issue that remains for determination is the second certified question, namely whether Rules 2 and 6 are ultra vires the authority that made them. The appellant argues that, under English common law, a person may act through an agent in exactly the same manner as he could act personally, and that this principle is incorporated in Indian law by section 2 of the Powers‑of‑Attorney Act VII of 1882. Section 2 provides that the holder of a power of attorney may, if he thinks fit, execute any assurance, instrument or other thing in his own name, with his own signature and seal, and that such execution shall have the same legal effect as if it were done by the donor of the power. The provision applies to powers of attorney created by instruments executed either

In this case, the Court observed that Section 26‑A of the Income‑Tax Act grants each partner the right to apply for the firm’s registration, and that right may be exercised both under the common law and under section 2 of the Powers‑of‑Attorney Act through an authorised agent. The Court noted that the legislature could, if it chose, nullify the common‑law rule, repeal section 2 of the Powers‑of‑Attorney Act and prescribe that an application under Section 26‑A must be signed only by the partner himself and not by any other person. However, the Court held that Parliament has not done so either expressly or by necessary implication. Consequently, the Court concluded that the application signed by Hariprasada Rao was legally equivalent to an application signed by Subba Rao. The Court further pointed out that the Rules require the signature of the partner and not that of his agent, but that, by prohibiting a act which is lawful under the statute, the Rules exceed the authority conferred on the rule‑making body by Section 26‑A, whose mandate is limited to framing rules that give effect to the statutory principles. Accordingly, the Court declared those Rules ultra vires. Assuming, alternatively, that the mandate given to the rule‑making authority under Section 26‑A was sufficiently broad to authorise the contested Rules, the Court held that they would still be ultra vires because they effectively abrogate the common‑law rule and repeal section 2 of the Powers‑of‑Attorney Act, which confers on a person the right to act through an agent. The Court emphasized that such a legislative function cannot be delegated to a rule‑making authority, and that if Section 26‑A is interpreted as conferring such power on an external body, it must be struck down as an unconstitutional delegation of legislative power. The Court therefore identified the correctness of these contentions as the issue that required determination.

The Court then turned to the underlying principle of agency law, stating that under English law, which is also reflected in the Indian Contract Act 1872, every person who is sui juris has the right to appoint an agent for any purpose and may do so when exercising a statutory right no less than when exercising any other right. The Court cited the observation of Justice Stirling in Jackson and Co. v. Napper and in Re Schmidts’ Trade‑Mark, noting that this rule is subject to well‑known exceptions, such as when the act to be performed is of a personal character, is attached to a public office, or involves fiduciary obligations. Apart from those exceptions, the Court affirmed that the law is well settled that whatever a person can do himself can be done through an agent. The Court further referenced the judgment of Justice Blackburn in The Queen v. Justices of Kent, which held that when a person authorises another to sign on his behalf, the signature of the authorised person is deemed to be the signature of the authorising person. On the basis of these established principles, the Court concluded that the appellant’s position was accordingly supported.

The appellant argued that, unless the statute expressly provides otherwise, an application that a partner is required to sign may validly be executed by a person who has been duly authorized as the partner’s agent. Accordingly, the court first examined whether the Income‑Tax Act contains any requirement that an application filed under section 26‑A must be signed by the partner himself. Section 26‑A reads as follows: “(1) An application may be made to the Income‑Tax Officer on behalf of any firm, constituted under an instrument of partnership that specifies the individual shares of the partners, for registration for the purposes of this Act and of any other enactment then in force relating to income‑tax or super‑tax. (2) The application shall be made by such person or persons, at such times, and shall contain such particulars, be in such form, and be verified in such manner as may be prescribed; and it shall be dealt with by the Income‑Tax Officer in such manner as may be prescribed.” The provision, it was observed, makes no mention that the registration application must bear the personal signature of the partner. This omission formed the basis of the appellant’s contention that the general right of a person to act through an agent—recognised both at common law and under section 2 of the Powers‑of‑Attorney Act—has not been displaced or limited by section 26‑A. To support this position, the appellant relied on several well‑established rules of statutory construction. First, statutes that infringe on established rights of individuals should, wherever possible, be interpreted in a way that preserves those rights; authorities such as Maxwell on the Interpretation of Statutes (10th ed., p. 285) and Craies on Statute Law (5th ed., pp. III‑114) were cited. Lord Justice Bowen, speaking in In re Cuno: Mansfield v. Mansfield, expressed the principle that “in the construction of statutes, one must not construe the words so as to take away rights which already existed before the statute was passed, unless the legislature’s intention to do so is expressed in plain language.” Secondly, the appellant argued that in the absence of clear and unambiguous wording, the legislature’s intention to alter existing law should not be inferred; this view is also supported by the commentary in Craies on Statute Law (pages 114‑115). Thirdly, the appellant submitted that the law disfavors the implied repeal of an earlier enactment; consequently, a later statute should not be read as repealing an earlier one unless the later provision contains express words or necessarily implies such a repeal. This principle is articulated in Maxwell on the Interpretation of Statutes (10th ed., p. 170) and Craies on Statute Law (5th ed., p. 337). As Justice Farwell observed, “if it is possible, it is my duty to read the section so as not to effect an implied repeal of the earlier Act.” Similarly, Justice A. L. Smith, in Kutner v. Phillips, warned that unless two statutes are so plainly irreconcilable that both cannot operate simultaneously, an implied repeal will not be found. The appellant therefore submitted that, applying these rules, section 26‑A should be read as granting a partner the right to register the firm, while leaving the method of doing so—to be governed by the existing law on agency—undisturbed.

The Court noted that a repeal of a statute could not be implied unless the language made such intention unmistakable, quoting Per A. L. Smith, J. in Kutner v. Phillips (3) and referencing the authorities cited as (1) [1890] 43 Ch. D. 12, 17; (2) [1936] 1 Ch. 266, 270; (8) [1891] 2 Q.B. 267, 272. In applying these principles, the Court explained that the appellant contended the true meaning of section 26‑A is to grant a partner a right to register the firm while leaving the method of exercising that right to be governed by existing law. Accordingly, the appellant argued that the provision does not aim to alter the general law concerning a person’s authority to act through an agent, nor does it intend to repeal section 2 of the Powers‑of‑Attorney Act; rather, the operation of section 26‑A depends on the continued existence of those earlier statutes. The Court observed that while the rules of statutory construction cited by the appellant are well established, they must be remembered merely as tools for discerning the legislature’s true intent as expressed in the enactment, and the fundamental question remains what the words of the provision signify in their context. The Court then referred to a passage from Crawford on “The Construction of Statutes”, 1940 edition, page 454, quoted by the appellant: “Why should a statute be subjected to a strict or a liberal construction, as the case may be? The only answer that can possibly be correct is because the type of construction utilized gives effect to the legislative intent. Sometimes a liberal construction must be used in order to make the legislative intent effective, and sometimes such a construction will defeat the intent of the legislature. If this is the proper conception concerning the rule of construction to be adhered to, then a strict or a liberal construction is simply a means by which the scope of a statute is extended or restricted in order to convey the legislative meaning. If this is the proper position to be accorded strict and liberal constructions, it would make no difference whether the statute involved was penal, criminal, remedial or in derogation of common right, as a distinction based upon this classification would then mean nothing.” Accepting this view, the Court asked whether, on proper interpretation, the statute requires that an application under section 26‑A be signed personally by the partner or whether it may be signed by an agent on the partner’s behalf. To resolve this issue, the Court stated that consideration must be given not only to the language of section 26‑A but also to the character of the legislation, the overall scheme of the Act, and the nature of the right conferred. The Act, as described in its preamble, is intended to consolidate and amend the law relating to income tax. Consequently, the appropriate rule of construction for such a statute is articulated by Lord Herschell in Bank of England v. Vagliano (1): “I think the proper course is in the first instance to examine the language of the statute, and to ask what is its natural meaning, uninfluenced…”

The Court explained that the provisions of the Indian Income‑tax Act had to be read as a self‑contained code that exhaustively covered every matter within its scope, and that the true meaning of each provision was to be determined by examining its language and purpose. In doing so, the Court referred to its earlier decision in Commissioner of Agricultural Income‑tax v. Keshab Chandra Mandal, where the issue concerned the interpretation of Rule 11 under the Bengal Agricultural Income‑tax Act, 1944, together with Form No 5. That rule required that the declaration in a return be signed “in the case of an individual, by the individual himself.” The Court had held, after reviewing the language of the statute, that the legislature intended to overturn the common‑law principle qui facit per alium facit per se, and that a valid declaration therefore had to bear the personal signature of the assessee. The appellant argued that the earlier case dealt only with the construction of Rule 11 and not with its constitutional validity, and that the question of whether the rule was ultra vires had not been raised. Nevertheless, the Court found that the significance of that precedent for the present dispute lay in the reasoning that led to the requirement of a personal signature. That reasoning was based on an analysis of the relevant provisions of the Bengal Act, which demonstrated that the legislature deliberately excluded the common‑law rule. Since the provisions of the Bengal Agricultural Income‑tax Act that were examined in the Keshab Chandra Mandal case are word‑for‑word identical to the corresponding provisions in the Indian Income‑tax Act, the Court concluded that it was logical to interpret the Indian Act in the same way, namely, as expressing an intention to discard the common‑law rule. For example, Section 25(1) of the Bengal Act stated that if the Income‑tax Officer was not satisfied that a return was correct and complete, he could issue a notice requiring the assessee either to appear personally at the tax office or to produce any evidence on which the officer might rely. This provision mirrors Section 23(2) of the Indian Income‑tax Act and explicitly allows the assessee to produce evidence through an agent. The Court observed that such an express provision for agent‑produced evidence would be unnecessary if the common‑law rule on personal signatures were intended to apply.

The Court observed that Sections 35 and 36 of the Bengal Act prescribe who may represent the assessee and in which proceedings such representation may occur, and that these provisions mirror Section 61 of the Indian Income‑tax Act, thereby forming a self‑contained code. Both the Bengal Act and the Indian Income‑tax Act also state that certain provisions of the Civil Procedure Code apply to proceedings under the Acts; however, the provisions of Order 3 of the Civil Procedure Code, which allow parties to appear and act through recognised agents, are expressly excluded. To summarise the effect of the Bengal Act, the Court cited Commissioner of Agricultural Income‑tax v. Keshab Chandra Mandal (1), noting that “the omission of a definition of the word ‘sign’ as including a signature by an agent, the permission under section 25 for production of evidence by an agent and under sections 35 and 58 for attendance by an agent and the omission of any provision in the Act applying the provisions of the Code of Civil Procedure relating to the signing and verification of pleadings to the signing and verification of the return while expressly adopting the provisions of that Code relating to the attendance and examination of witnesses, production of documents and issuing of commission for examination and for service of notices under sections 41 and 60 respectively, cannot be regarded as wholly without significance.” The Court held that this reasoning applies equally to the provisions of the Indian Income‑tax Act and strongly supports the respondent’s contention that the common‑law rule is not intended to govern proceedings under the Act. The Court further considered the nature of the right conferred by section 26‑A. Under English common law a firm is not a juristic person, the firm’s name merely designating the partners who comprise it. Yet, as this Court previously observed in Dulichand Laxminarayan v. Commissioner of Income‑tax, Nagpur (2), statutes have altered that conception, treating firms as distinct entities for statutory purposes. The Indian Income‑tax Act is one such statute; it regards a firm as a unit for taxation, imposing the charge on the total income of the firm under section 3, with assessment of the firm’s total profits under section 23. Section 23(5) creates an exception for firms registered under the Act, providing that the sum payable by the firm shall not be determined; instead, the total income of each partner, including that partner’s share of the firm’s income, profits and gains of the previous year, shall be assessed and the sum payable by the partner determined. Consequently, registration changes the tax treatment from a unit‑based charge to individual assessment, allowing partners to benefit from lower individual tax rates, a right that arises solely from the statute and may be claimed only in accordance with the statutory provisions that created it.

In this case, the Court observed that under the statutory provision, the income, profits and gains of the previous year of each partner are assessed and the tax payable by the partner is determined on that basis. Consequently, when a partnership is registered under the Act, it ceases to be treated as a single taxable unit, and the profits of the firm are deemed, according to the general law of partnership, to have been earned by the individual partners in proportion to their shares, so that each partner is taxed on his personal income together with his share of the firm’s profits. The Court noted that the advantage of this provision is evident, because the tax rate applicable to the partners' income will be the lower rate applicable to individual earnings rather than the higher rates that apply to larger aggregate incomes. Accordingly, registration confers on the partners a benefit that would not be available but for section 26‑A, and the Court stressed that this benefit, being created by the statute, can be claimed only in accordance with the terms of that statute; a person seeking relief under section 26‑A must strictly satisfy those statutory conditions before he may enjoy the benefit. In other words, the right created by the statute is governed solely by its own provisions, and it would be inconsistent with the nature of that right to enlarge its scope by reference to other laws. The Court further held that the statute must be interpreted as exhaustive concerning the conditions for claiming the benefit, and that, when the character of the legislation, its scheme and the nature of the right under section 26‑A are considered, it is unavoidable to conclude that common‑law rules were not intended to be preserved and that the entitlement to apply for registration is to be determined exclusively by the statutory prescriptions. Upon accepting this construction, the Court explained that when Parliament authorised the rule‑making authority to frame rules regarding who may apply for registration, the manner of application and the timing, the statute merely directed that authority to fill in details within the legislative field it had created, and there is no dispute that Rules a and 6 fall within the mandate conferred by the section. The Court then pointed out that section 59(5) of the Act provides that “Rules made under this section shall be published in the official Gazette, and shall thereupon have effect as if enacted in this Act,” and therefore the validity of the Rules is beyond question. The Court referred to the observations of Lord Herschell in Institute of Patent Agents v. Lockwood(1) in support of this view. The Court also recorded the appellant’s contention that the Rules in question are repugnant to section 2 of the Powers‑of‑Attorney Act VII of 1882 and therefore ultra vires. In addition to the reasons already given supporting the conclusion that common‑law rules were not intended to operate in the field occupied by section 26‑A, the Court indicated that there is a further consideration that weighs against accepting the appellant’s contention.

The Court observed that there is a stronger and more compelling reason for rejecting the appellant’s contention that the two statutory provisions were in conflict. In fact, the Court found that no conflict existed between section 2 of the Powers‑of‑Attorney Act and the provisions of the Income‑Tax Act. To appreciate the scope of section 2, the Court referred to the historical development of the legislation. Under the common law of England, an agent who possessed authority to execute an instrument was required to sign in the name of the principal if the agent was to be bound by the instrument. When an agent signed a deed in his own name, even though he indicated that he was acting as an agent, the law treated the agent as the party to the document and not the principal. Consequently, only the agent could enforce the deed and the agent alone would be liable under it. The Court cited the authorities In re International Contract Company, Schack v. Antony, Halsbury’s Laws of England (3rd edition, volume 1, page 217), and Bowstead on Agency (10th edition, page 93) to illustrate this common‑law rule.

To alleviate the difficulties created by that rule, the Court explained that the English Parliament enacted section 46 of the Conveyancing and Law of Property Act 1881 (44 and 45 Vict., chapter 41). Section 46 provides that a donee of a power of attorney may, at his discretion, execute any assurance, instrument or thing in his own name, with his own signature and seal, if such execution is authorised by the donor of the power. The provision further declares that every assurance, instrument or thing so executed shall be as effective in law as if it had been executed by the donee in the name, signature and seal of the donor. The Court reproduced the text of the provision and noted that it applies to powers of attorney created by instruments executed before or after the commencement of the Act.

The Court then observed that the Indian Legislature adopted the same principle when it enacted the Powers‑of‑Attorney Act VII of 1882, incorporating in its section 2 the wording of section 46 of the English statute verbatim. The purpose of this Indian provision, the Court said, is to give effect to instruments executed by an agent in a manner that does not comply with the strict common‑law rule, and that the provision is mainly procedural rather than substantive. The provision does not create a new right for a person to act through agents; rather, it presumes that the agent already has authority from the principal and it protects acts performed by the agent in his own name when such authority exists. However, the Court stressed that when the issue is whether the authority itself exists or is valid, section 2 of the Powers‑of‑Attorney Act has no operative effect.

From this analysis, the Court concluded that the fields occupied by the two statutes are entirely separate. Section 26‑A of the Income‑Tax Act expressly forbids a partner from delegating the exercise of his rights under that section to an agent. By contrast, section 2 of the Powers‑of‑Attorney Act merely provides the manner in which a delegation, if it exists, may be effected. Thus, there is no overlap or repugnancy between the two provisions, and the appellant’s argument that the rules were ultra vires on the ground of inconsistency was rejected.

The Court observed that there was no conflict between the two statutory provisions in question, and consequently no issue of one provision repealing the other arose. It then summarized that the Indian Income‑tax Act constituted a self‑contained code that exhaustively covered the matters within its scope, and that the language of the Act demonstrated an intention to depart from the traditional common‑law principle qui facit per alium facit per se. The Court explained that the purpose of the statute was to allow a firm to obtain the benefit provided in section 23(5)(a) only when the firm was duly registered under section 26‑A and complied with all the conditions prescribed in that section as well as the rules made under it. Because the rules required the application to bear the personal signature of the partner, any signature placed by an agent on the partner’s behalf was held to be invalid. Regarding the additional question raised by the appellant – namely, that the power to repeal a law is a legislative function that may be exercised solely by a properly constituted legislature and that delegating such a power to an external authority would be unconstitutional – the Court noted that this issue did not need to be decided. Finally, the Court agreed with the earlier judgment of Justice Viswanatha Sastry that rules 2 and 6 fell within the authority of the rule‑making body, and consequently dismissed the appeals and ordered the appellants to pay costs.