1. The suit out of which this second appeal arises was to recover possession of a house under a gift-deed executed by the mother (defendant 2) of the plaintiff's late husband, Subbayya, jointly to Subbayya and herself. The claim was resisted upon various grounds, the first with which I propose to deal being the question of attestation of the deed. It purports to have been attested by five persons, two being signatories and three marksmen. If the plaintiff fails to establish two valid attestations amongst these, there are the further contentions that the writer of the document and the Sub-Registrar who registered it may be held to rank as attestors.
2. Of the two signatories one has not been traced and there is no evidence which would allow the Court to conclude that he validly attested the document. The other has been examined as P.W. 7 and although he himself has stated in evidence that he did not see the executant sign, the learned District Judge has accepted the evidence of P. Ws. 9 and 10 who state that he attested the document. This assertion was not made the subject of cross-examination and I think that it must carry with it the implication that it was a valid attestation within the terms of the definition added to Section 3, T. P. Act, by Act 27 of 1926.
3. We turn then to the attestations by the three persons unable to sign their own names. On the same evidence, that of P. Ws. 9 and 10, the learned District Judge has found that the attestation of one of them, P. W. 8, is valid, but this attestation has been attacked upon the general ground that as the law now stands a person who cannot sign his name is not qualified to attest a document. Before the law was amended as stated above, there can be no doubt that a marksman could be accepted as an attestor. For purposes of non-testamentary documents there was no definition of 'attest' and Under Section 3(52), General Clauses Act, it was permissible to hold that sign' included 'mark.' I think however that the effect of Act 27 of 1926 in introducing a definition has altered the law in this respect. It is true that there is no case law directly in point but the definition is in precisely similar terms to that contained in the Succession Act as applying to the execution of wills. It is reasonable therefore to hold that, unless the contrary appears from the context in which each definition is to be found, words construed in a certain sense in the one must be construed in the same sense in the other. It has been held both in D. Fernandes v. R. Alves  3 Bom.382 and in Nityagopal Sircar v. Nagendranath Mitter 11 Cal. 429 that the direction contained in Section 50(3) Succession Act 1865 (now Section 63(c) of the present Succession Act), as to the signature of witnesses attesting a will is not satisfied by the witnesses affixing their marks. In the latter case it is pointed out that the legislature, while speaking of the action of the testator, uses expressions shall sign or affix his mark' signature or mark'. In speaking of the witnesses the section does not use any alternative expression but simply says 'the witnesses must sign.'
That being the case,' the learned Judges say: 'we cannot help thinking that the legislature advisedly drew a distinction between the action of the testator and that of the witnesses as regards the mode of their respective signatures.
4. This construction of the language used has no reference to the fact that the document in question is a will and I cannot find any ground in this circumstance to distinguish these cases. It will be apparent that the reasoning used is derived purely from the wording of the definition, which is the same in both cases; and since I cannot suppose that the legislature was unaware of these decisions when it extended the definition to the attestation of nontestamentary documents, I can only infer that the same construction was intended in the one case as in the other. Certainly insistence upon the attestation of such documents by persons who can sign their names would not lead to more hardship or inconvenience than in the case of wills, which often have to be executed at very short notice. There is an undoubted hardship in applying the principle to such a case as the present, where the document was executed as far back as 1912, long before the law was amended. But inasmuch as that amendment has expressly been made retrospective in effect it is not open to me to refuse to act upon it merely because illiterate persons were formerly accepted as capable of attesting. I must hold accordingly that none of the three so-called attestations by marksmen is valid.
5. We come then to the writer, Two objections have been raised to accepting him as an attestor. In the first place it is said not to have been shown that he intended to attest the execution. I should have been prepared to overrule this objection, because he signed after the attestors, and the evidence I think allows the presumption that he signed in the presence of the executant. But the other objection appears to me fatal. The executant of the gift-deed was herself unable to sign her name and it is clear from the evidence and from an examination of the document that her name was affixed to it by the writer himself. It has been held both in Sristidhar Ghosh v. Rakkya Kaly Dasi A.I.R.1922 Cal.168, and in a case of this Court, Radhakrishna Mudaliar v. Subbaraya Mudaliar 40 Mad.550 , that a person who signs on behalf of the executant of a document cannot also sign as an attestor of it. The former case related to a mortgage bond and the latter to a will, but there seems to be no difference in principle. The document must have been attested by
two or more witnesses each of whom has seen the executant sign or affix his mark to the instrument or has seen some other person sign the instrument in the presence and by the direction of the executant,
which means that it must be attested by two persons other than the person who signed the instrument in the presence and by the direction of the executant. We have accordingly to exclude the writer upon this ground. There remains the Sub-Registrar. The learned District Munsif was inclined to exclude him because he does not necessarily sign in the presence of the executant, as the rule requires and no proof was adduced that he in fact did so sign. In the absence of a decision of this Court he felt himself bound however to follow Radha Mohun Dutt v. Nripendranath Nandy : AIR1928Cal154 . There is now a case in point decided by this Court in Ramanathan v. Delhi Badha Tevar A.I.R.1931 Mad.835. It was there held by the learned Chief Justice and myself that proof must be adduced that this term in the requirement was complied with. There being no evidence upon this point I must hold that the Sub-Registrar also cannot be accepted as an attestor. The result accordingly is that only one person can be declared to have validly attested the document, which being a gift-deed must therefore fail for lack of due attestation. This conclusion no doubt involves some hardship to the plaintiff, but unless my application of the law as it stands at present is at fault I do not see how a different decision can be arrived at.
6. The gift deed has been further attacked on the ground that the gift made was not accepted by the donees. It was made to a husband and wife while they were minors by the husband's mother. The plaintiff has indeed contended that they were of age at the time, but this has been found against, although I do not think that the finding should bar the plea of acceptance if it may reasonably be inferred from the facts. The subject-matter was a house in which the parties continued to live with the husband's parents until the husband died some 11 years later. Under these circumstances it is difficult to see by whom, if not by the parents, acceptance could have been made at the time of the gift. The issue throws the burden of disproving acceptance upon the defendants and I think that in such a case as this the English rule should be followed that unless the contrary is proved acceptance of the gift should be inferred. The mere circumstance that the mother retained custody of the deed and kept the house in her name in the Municipal accounts and paid the taxes, does not show that she did not intend the gift to be acted upon. I do not agree that it was onerous in character. All that the donees had to do was to live together as husband and wife, with a provision for divestment should either fail to do this. No case has been cited to me in which it has been held that a gift made in such circumstances is invalid for failure to prove acceptance. The circumstances in Hariram Serowgee v. Madan Gopal Bagla A.I.R.1929 P.C.77, were quite different, the only evidence of an intention to make a gift being furnished by certain entries in an account. No doubt Debi Singh v. Bansidhar A.I.R.1922 All.44 and Dhannamull v. Parmeshari Das, A.I.R.1928Lah.9, cited to me on behalf of the plaintiff, raise the question rather of the completion of the gift by the donor than of its acceptance by the donee. But in Muhammad Abdul Nayeem v. Jhoti Mahton  41 I.C. 389, an inference of acceptance was drawn from the circumstances of the case, which I think it would be equally reasonable to draw here, I am not therefore disposed to interfere with the findings of the lower Courts upon this issue.
7. The only other point is that the learned District Judge while finding that the plaintiff had not been guilty of immorality during her husband's lifetime has not recorded a finding on the question whether she lived separately from him. It appears to me from the terms of issue 4 that these two questions were aspects of the same matter and that the finding with regard to the plaintiff's chastity covers the other point. In view of my finding on the question of attestation I must allow the second appeal, and set aside the decrees of the lower Courts so far as they relate to this property. The decree will stand as regards the payment of Rs. 7-2-0. In the circumstances the parties will pay their own costs throughout.