Spencer, J.

1. I have had the advantage of seeing the judgment which my learned brother is about to deliver. On the questions of fact I have nothing to add to what he will state beyond expressing my entire agreement with his conclusions. On the two questions of law arising out of the 5th and 6th points I wish to express my concurrence with his opinion. His opinion on the 5th point agrees also with that of Ramesam, J., in Anne Brahmayya v. Chelasami Rattayya (1922) 20 L.W. 503. I agree with them in respectfully dissenting from the view expressed by Oldfield, J., in Viswasundara Rao v. Somasundara Rao (1920) I.L.R. 43 M. 876 to the effect that the class of sapindas, whose consent should be obtained by a widow in order to supply the authority lacking in consequence of her husband having died without authorising her to adopt a son, is limited to agnates. In the judgment of Seshagiri Aiyar, J., in Kristnayya v. Lakshmipathi (1920) I.L.R. 43 M. 650 : 30 M.L.J. 265 (P.C.) the Smrithi of Yajnavalkya, which is the basis for the theory that a woman, who has no husband, father or sons, needs the advice of those who are her natural protectors, is quoted.

2. Let her father protect a maiden, her lord a married woman, sons in old age; if none of these, other gnatis; she is not fit for independence.

Here the word used is "gnatis."

3. But temporal considerations cannot be left out of account. Proximity to the deceased with respect to rights of property is also an important factor in considering who should be consulted, for the reason that those persons who would succeed to the property, if no adoption were made, are most interested in the protection of the inheritance and in restraining the widow concerned from acting capriciously (vide Sri Raghu-nadha v. Sri Brozo Kishoro (1876) L.R. 3 I.A. 154 : I.L.R. 1 M. 69 (P.C.) and Veera Basavaraju v. Balasurya Prasada Rao (1918) L.R. 45 I.A. 265 : I.L.R. 41 M. 998 (1004) : 36 M.L.J. 40(42) (P.C.). As regards the devolution of property the original text of Manu is.

(Manu IX. 187) which may be translated literally:

Whoever is unremote from the sapinda to him let his wealth belong.

The word used here is " sapinda".

4. Now neither the word gnati nor the word sapinda is synonymous with the English word " agnate", whatever may be the suggestion conveyed by the common root " gna". For sapindas undoubtedly include bhinnagotra sapindas who are cognates, and in Professor Monier Williams Dictionary "gnati " is defined as a " paternal relation, a kinsman in general, a father, a brother, a distant kinsman, one who does not participate in the oblations offered to deceased ancestors (according to the School, gnatis is a paternal and sambhandh in maternal relation; the original meaning of gnati may Be intimately acquainted). Cf. Greek (Gustoo Gnotee).

5. I am satisfied that in the passage where Mr. Ameer Ali observed in Veera Basavaraju v. Balasurya Prasada Rao (1918) L.R. 45 I.A. 265 : I.L.R. 41 M. 998 (1004) : 36 M.L.J. 40(42) (P.C.) " The Ramnad case established the proposition that in the Dravida country under the Dravidian branch of the Mitak shara Law there is force in the absence of authority from her deceased husband, a widow may adopt a son with the assent of his male agnates", and in the other passage in Krist-nayya v. Lakshmipathi (1920) I.L.R. 43 M. 650 : 30 M.L.J. 265(P.C.) where Viscount Cave observed at p. 654: " The consent required is that of a substantial majority of those agnates nearest in relationship who are capable of forming an intelligent and honest judgment on the matter", their Lordships had no intention of laying down the propositions (1914) 20 L.W. 503 that in the absence of agnates available for being consulted no adoption could be made by a widow who had failed to obtain the authority of her husband, and (1920) I.L.R. 43 M. 876 that in no circumstances should cognates be consulted. No doubt agnates should as a rule be consulted first, if there are agnates, because the principal agnat