King v Perpetual Trustee Co

Dec. 15 Dixon C.J., Williams, Webb, Fullagar and Taylor JJ.

The only question that arises on this appeal is the true meaning of the residuary gift in the will of John O'Flaherty who died on 26th July 1937. The gift is in these terms: - "All the rest and residue of my real and personal estate I give devise and bequeath to my trustee upon trust to divide the same equally amongst the children of my sister Bridget King the children of my brother Patrick Flaherty the children of my niece Irene Sheehy and the children of my nephew James Patrick Ford in equal shares absolutely." Admittedly it is divisible among thirteen beneficiaries. They comprise the six children of the sister of the testator, Bridget King (the appellants here), the two children of his brother Patrick, the three children of his niece, Irene Sheehy, and the two children of his nephew, James Patrick Ford. These were all the children of these four praepositi living at the death of the testator. One of this class, Thomas Flaherty, has since died and Perpetual Trustee Co. (Ltd.) is the executor of his estate. The contest is whether these thirteen beneficiaries participate in the gift per capita or per stirpes. Myers J. held that they take per stirpes. It is contended for the appellants that they take per capita.

The plan of the will which is dated 15th March 1935 is simple and the beneficiaries easy to identify. There is first a direction to the trustee of the will to set apart and invest a sum sufficient to produce an income to meet an annuity of five pounds per week which the testator directed his trustee to pay to his widow during her life or widowhood (later increased on an application under the Testator's Family Maintenance and Guardianship of Infants Act 1916 to seven pounds per week during her life or widowhood) and after the death or remarriage of his wife a direction that the sum so set apart shall fall into and form part of his residuary estate. The widow did not remarry and died on 16th October 1947. The testator further directed his trustee to set apart and invest a sum of money sufficient to produce an annuity of one pound per week which he directed his trustee to pay to his brother Patrick during his life and directed that from and after his death the sum so set apart should fall into and become part of his residuary estate. He further directed his trustee to pay to the two sons and to the daughter of his brother Patrick the sum of £1,000 each. He then directed his trustee to settle £3,000 upon trust to pay the income arising therefrom to his sister, Bridget King, during her life and after her death to pay this sum to such of her children as should survive her if more than one in equal shares absolutely. He also directed his trustee to set aside a further sum of £3,000 and pay the income to his niece, Irene Sheehy, during her life and after her death to pay this sum to her children if more than one in equal shares absolutely. He also directed his trustee as to a further sum of £4,000 to pay the income arising therefrom to his nephew, James Patrick Ford, during his life and after his death to hold this sum upon trust for his children if more than one in equal shares absolutely. Up to this point the testator has made provision for his widow, his sister and brother, his niece and nephew and the children of his sister, brother, niece and nephew. The gifts to the sister, brother, niece and nephew, and their children are made to them as distinct families and vary somewhat in the amount and manner of their enjoyment. Then comes the residuary gift the text of which has already been set out.

The gift of residue is a gift to the children of four named persons in equal shares and it is a settled rule of construction that in the case of such a gift, in the absence of a sufficient indication of intention to the contrary, the persons comprising the class take per capita. The rule was first established by King L.C. in Blackler v. Webb [1] and has therefore had a long life. The headnote to that case is that "One having had five children, A., B., C., D. and E.; B. is dead leaving several children, and by will the testator devises the residue of his personal estate to his son A. and to B.'s children, and to his daughter C. and D.'s children, and to his daughter E.; D. is living and has children; decreed the children of B. and the children of D. shall take per capita, and not per stirpes, as if all named" [1] . The Lord Chancellor at first seemed "inclinable" that the grandchildren should take per stirpes only, yet at length he decided that the testator's son James, and the children of the deceased son Peter and his daughter Traverse, and the children of his daughter Webb, and his daughter Man (being in all fourteen in number), should each of them take per capita, as if all the grandchildren had been named by their respective names. He decided that to determine that the grandchildren should take per stirpes would be to go too much out of the will, and contrary to the words, when the meaning of the testator might be according to his words, and that meaning a reasonable and sensible one. The existence of this rule of construction was conceded in the present case. It was a concession from which there was no escape. But it was contended that there were sufficient indications of intention in the will to exclude the rule and that the division should be per stirpes. This contention found favour with the learned judge below. Before us it was sought to support it on three grounds: (1) that the trust to divide the residue equally amongst the children of the four named relatives indicates that the division is to be between them by stocks; (2) that the rule is excluded where the beneficiaries are not in the same generation but of different generations to the testator and to each other and here two of the stocks are the children of a brother and sister and the other two are the children of a niece and nephew; (3) that the repetition of the initial direction for equal division at the commencement of the gift occurring at its conclusion indicates that there is to be a double division, first an equal division between the stocks and then an equal division of each sub-division between the members of each family. None of these grounds can find any support in the language of the will and it is from the words of the will that the intention of the testator must be ascertained, aided only by such facts as existed and were known to the testator at the date of the will which it is permissible to take into account in interpreting that language. This is trite law. In Towns v. Wentworth [2] it was said in the Privy Council: "In order to determine the meaning of a will, the court must read the language of the testator in the sense which it appears he himself attached to the expressions which he has used, with this qualification, that when a rule of law has affixed a certain determinate meaning to technical expressions, that meaning must be given to them, unless the testator has by his will excluded, beyond all doubt, such construction" [1] . In Charter v. Charter [2] the Lord Chancellor (Lord Cairns) said: "But, my Lords, there is a class of evidence which in this case, as in all cases of testamentary dispositions, is clearly receivable. The court has a right to ascertain all the facts which were known to the testator at the time he made his will, and thus to place itself in the testator's position, in order to ascertain the bearing and application of the language which he uses, and in order to ascertain whether there exists any person or thing to which the whole description given in the will can be, reasonably and with sufficient certainty, applied" [3] . If these principles are applied to the present case, all these grounds disappear. The directions for equal division at the commencement and conclusion of the residuary gift, on their ordinary natural and grammatical construction, relate to one division and one only, that is to the division of the whole of the residuary estate in equal shares amongst one class consisting of all the children of the praepositi living at the death of the testator, and the reference to their parents provides a means and nothing more of identifying these beneficiaries. In Sumpton v. Downing [4] Dixon J. (as he then was) said: "Prima facie, under a gift to the children of named persons as a class, the children take per capita and not per stirpes. It has been said that no man who was guided only by a knowledge of English speech would suppose that a direction to distribute money between the children of A and of B equally could mean anything but a division in which each child took a share equal with that of every other child, whether his parent was A or B. However this may be, it is enough that at least the prima-facie legal meaning of such a direction is that the distribution should be per capita. This is so whether in point of expression the class is described as the children of A and (of) B or as the children of A and the children of B" [5] . Here the class is described as the children of A, of B, of C, and of D which is as clear an example as can be had of a gift to a class per capita. At most the repetition of the direction for equal division of the residue gives emphasis to the intention of the testator that it should be equally divided among the members of the class. The double direction has no more significance than the commonplace expression which so often occurs in wills "equally between them share and share alike" (and in some cases with the addition "and in equal proportions"). It would be a misuse of language to construe the residuary words in the present case as meaning that there is first to be a division of the residue into four equal parts and that these four parts are then each to be sub-divided into a number of equal parts corresponding to the number of children of each of the named relatives of the testator. The language is concerned and concerned only with the division of residue into equal shares not between the four named relatives and their children, which might provide sufficient context for a stirpital construction, but between all the children of all these relatives as a composite class.

1. (1726) 2 P. Wms. 383 [24 E.R. 777]. 2. (1726) 2 P. Wms. 383 [24 E.R. 777]. 3. (1858) 11 Moo. P.C. 526 [14 E.R. 794]. 4. (1858) 11 Moo. P.C., at p. 543 [14 E.R., at p. 800]. 5. (1874) L.R. 7 H.L. 364. 6. (1874) L.R. 7 H.L., at p. 377. 7. (1947) 75 C.L.R. 76. 8. (1947) 75 C.L.R., at pp. 87, 88.

It is a mistake to attempt to ascertain the meaning of one will from the meaning attributed to another. But that does not mean that where there is a settled rule of construction the same start cannot be made in the construction of all wills to which the rule is prima facie applicable. To do otherwise is to open wide the field to pure conjecture. In Lady Lincoln v. Pelham [1] , the ultimate bequest in the will of the testatrix of a settled legacy was that it should be equally divided among the younger children of the Duke of Newcastle by her late daughter Catherine and the younger children of another daughter, Lady Sondes. Lord Eldon said: - "Upon the next question, whether the distribution is to be per stirpes or per capita, I am not quite sure, that my opinion is not against the intention. If there is a settled construction, founded upon cases decided, applying to the terms used, it is better to adhere to that settled construction, though I may entertain some doubt, whether it is according to the intention, than upon grounds, on which I cannot rest in every view of the case, to come to a decision, having a tendency to shake that, which forms a rule of construction; and which may in practice have been acted upon in many cases. It is clear, that if this had been a bequest to the younger children of two persons, equally to be divided between and among them, the division would be per capita The particular circumstances are very strong to raise conjecture and doubt as to the intention: but do they, by the inference arising from them, overpower the settled construction of the words? Whatever the actual intention may have been, the legal effect is a distribution per capita; and I cannot safely draw an inference from the other part of the will; introducing distinctions, tending to shake the settled doctrine The distribution must be per capita " [1] . In In re Stone; Baker v. Stone [2] a testator gave real and personal estate to his wife for life and directed that after her death the income should be equally divided between his brother and sisters therein named, "at the decease of either of my before-named brother or sisters their interest herein to be equally divided amongst their children, and after the decease of all I desire the whole of my property to be sold, moneys called in &c. &c., and to be equally divided between the children of the aforesaid share and share alike" [3] . It was held by the Court of Appeal (overruling Stirling J.) that the ultimate gift to the nephews and nieces was a clear gift per capita, and could not be controlled by the fact that so long as any brother or sister of the testator was living the income was divisible per stirpes. Lindley L.J. said: - "Why are we to take this to mean that the distribution is to be per stirpes? The obvious meaning of the words is, that the division is to be per capita, and the language is not open to ambiguity I do not enter into an examination of the cases: when I see an intention clearly expressed in a will, and find no rule of law opposed to giving effect to it, I disregard previous cases" [4] . Kay L.J. said: - "No one contends that if these words stood alone the division would not be per capita. But it is said you can see from the context that this was not the testator's intention. A context ought to be very strong to alter the effect of such plain words We ought to abide by the language of a testator, and not alter it on conjecture. Stirling J. seems to have felt himself bound by the decisions; but I am against construing one will by another where the language of the two is not identical" [5] . A slightly later case is Capes v. Dalton [6] (before Farwell J. and the Court of Appeal) (sub nom. Kekewich v. Barker [7] in the House of Lords). There the gift was in trust for George Barker, his sister, Mary Barker, and the children now living of Richard Hollings who being male shall live to attain the age of twenty-one years, or being female shall live to attain that age or marry, and if more than one in equal shares, the share or shares of any of them being female to be for her or their sole and separate use. There were four children of Richard Hollings living at the date of the death of the testator all of whom attained the age of twenty-one years. It was held by Farwell J. in the first instance that the gift was divisible into equal sixths between George and Mary Barker and the four children of Richard Hollings. He was overruled in the Court of Appeal by a majority, Stirling L.J. dissenting [1] . But the House of Lords reversed the order of the Court of Appeal and agreed with the dissenting judgment of Stirling L.J. The importance of the case lies chiefly in the remarks of Stirling L.J. about the case of Davis v. Bennett [2] . After applying what he called the rule established by Blackler v. Webb [3] and constantly followed ever since, that in the absence of a sufficient indication of contrary intention the six beneficiaries would take per capita, he said, referring to Davis v. Bennett [2] : - "There the fund was directed to "be equally divided between my sisters Jane and Mary, and the lawful issue of my deceased sisters Elizabeth and Anne in equal shares if more than one of such respective lawful issue" [4] . Lord Romilly, then Master of the Rolls, held that the fund ought to be divided per capita, and Lord Westbury said "that construction would have been correct if the bequest had ended with the words "if more than one" " [5] ; and although he came to a different conclusion he did so by reason of the weight which he considered ought to be attached to the word "respective". I am unable to find any expression in the present will which affords ground for coming to such a conclusion" [6] . In the House of Lords Lord Halsbury [7] completely agreed with these remarks of Stirling L.J. about Davis v. Bennett [2] . Lord Davey said that: "a gift of this kind is prima facie a gift per capita to the persons who are named either nominatim or by reference, and that there is not sufficient context, in my opinion, to prevent the application of the ordinary rule here" [7] . Lord Lindley said: - "I think that the view taken by Stirling L.J. was correct" [7] . It will be seen that in the passage from Davis v. Bennett [2] cited by Stirling L.J. there was, as there is in the present case, an initial provision for equal division repeated at the end of the gift. A similar duplication of this provision occurred in Re Harper; Plowman v. Harper [8] . There the words were "the other moiety to be divided equally between the unmarried daughters of my brother-in-law Dr. H. and Dr. G. equally". Sargant J. held that the moiety was divisible per capita in equal fourth shares between the three unmarried daughters of Dr. H. and Dr. G. He said: - "I was for some time impressed by that word (equally) as possibly meaning that Dr. Grant was to take something which was equal to the whole amount given to the unmarried daughters taken together; but on the whole I am of opinion that I should be attributing too much meaning to the word by that construction" [1] . In line with these cases are certain decisions in the Australian courts and in particular the decision of Cussen J. in In re Jones; Harris v. Jones [2] ; Macfarlan J. in In re McInnes; Trustees Executors & Agency Co. Ltd. v. McInnes [3] ; Harvey J. in Gibson v. Abernethy [4] and Sugerman J. in Perpetual Trustee Co. (Ltd.) v. Pryde [5] . Later English cases where the distribution was held to be per capita include In re Dale; Mayer v. Wood [6] ; In re Cossentine; Philp v. Wesleyan Preachers' Association [7] (a decision of Maugham J. as he then was) and In re Alcock; Bonser v. Alcock [8] (a decision of Evershed J. as he then was). On the other hand in In re Walbran; Milner v. Walbran [9] (criticized by Sargant J. in In re Harper; Plowman v. Harper [10] and by Maugham J. in In re Cossentine; Philp v. Wesleyan Preachers' Association [7] ); Re Daniel; Jones v. Michael [11] ; In re Hall (dec'd.); Parker v. Knight [12] ; In re Jeeves; Morris-Williams v. Haylett [13] and In re Birkett (dec'd); Holland v. Duncan [14] the division was held to be per stirpes. In these cases, apart perhaps from Re Daniel; Jones v. Michael [11] where assistance could be derived from the context of the will, what appeared to be very chimerical circumstances were held sufficient to displace the prima facie rule of construction. For instance, in In re Hall (dec'd.); Parker v. Knight [12] Harman J. expressed the opinion that "through the authorities runs a reconciling principle that cases of capital distribution are cases of distribution between strangers or persons of no corresponding relationship; and that cases of stirpital distribution are cases of family distribution" [15] . He had already said that he would expect the stirpital basis in family gifts. With respect it is impossible for us to find any such reconciling basis in the authorities and we are unable to expect the stirpital basis in family gifts. All this is pure conjecture. In In re Jeeves; Morris-Williams v. Haylett [13] , Vaisey J. was convinced that it was a matter of guesswork and equally convinced that it was his duty in those circumstances to guess as best he could. In re Birkett [1] is perhaps the highwater mark of these cases for there Danckwerts J. guessed that the testatrix would have wanted to reward a friend with whom she had gone to live and who had looked after her, not merely on an equal footing with two children of a deceased sister, but on a footing that she took one-half of the gift and they shared the other half between them.

1. (1804) 10 Ves. Jun. 166 [32 E.R. 808]. 2. (1804) 10 Ves. Jun., at pp. 175-177 [32 E.R., at pp. 811, 812]. 3. (1895) 2 Ch. 196. 4. (1895) 2 Ch. 196, at p. 199. 5. (1895) 2 Ch., at p. 200. 6. (1895) 2 Ch., at p. 201. 7. (1902) 86 L.T. 129. 8. (1903) 88 L.T. 130. 9. (1902) 86 L.T. 129. 10. (1862) 4 De G. F. & J. 327 [45 E.R. 1209]. 11. (1726) 2 P. Wms. 383 [24 E.R. 777]. 12. (1862) 4 De G. F. & J. 327 [45 E.R. 1209]. 13. (1862) 4 De G. F. & J., at p. 328 [45 E.R., at p. 1210]. 14. (1862) 4 De G. F. & J., at pp. 328, 329 [45 E.R., at p. 1210]. 15. (1902) 86 L.T., at p. 131. 16. (1903) 88 L.T., at p. 131. 17. (1862) 4 De G. F. & J. 327 [45 E.R. 1209]. 18. (1903) 88 L.T., at p. 131. 19. (1903) 88 L.T., at p. 131. 20. (1862) 4 De G. F. & J. 327 [45 E.R. 1209]. 21. (1914) 1 Ch. 70. 22. (1914) 1 Ch., at p. 76. 23. (1910) V.L.R. 306. 24. (1925) V.L.R. 496. 25. (1918) 18 S.R. (N.S.W.) 122; 35 W.N. 43. 26. (1949) 49 S.R. (N.S.W.) 203; 66 W.N. 70. 27. (1931) 1 Ch. 357. 28. (1933) 1 Ch. 119. 29. (1945) 1 Ch. 264. 30. (1906) 1 Ch. 64. 31. (1914) 1 Ch. 70. 32. (1933) 1 Ch. 119. 33. (1945) 2 All E.R. 101. 34. (1948) Ch. 437. 35. (1949) Ch. 49. 36. (1950) 1 Ch. 330. 37. (1945) 2 All E.R. 101. 38. (1948) Ch. 437. 39. (1948) Ch., at p. 440. 40. (1949) Ch. 49. 41. (1950) 1 Ch. 330.

We are not bound by these decisions or the reasons on which they are based and it is better to keep our imagination in abeyance and adhere to the settled rule of construction, and not to depart from it, unless there is in the context of the will, or in the admissible evidence in the light of which the language of the will can be interpreted, a sufficient indication of intention to the contrary. Some examples of what is a sufficient indication of a contrary intention are referred to in Neil v. McDonnell [2] ; McDonnell v. Neil [3] . In any event these cases throw no light on the present case. Here the gift of residue is, on its ordinary grammatical construction, a gift to a single class identified by the members of the class being the children of any one of four named persons. The prima facie rule is that such a class takes per capita and it is a rule that should not be lightly departed from. There is nothing in the language of the present will from which a contrary intention could be implied. The structure of the will as a whole supports the prima facie presumption. The testator has, in the first instance, provided for his brother, sister, niece and nephew and their children as separate stocks or families. Having done so he has directed that the residue should be divided among all the children of these four persons in equal shares. He has not directed that there should be a primary and second division. He has directed one and only one division.

1. (1949) 79 C.L.R. 177, at pp. 196-199. 2. (1951) A.C. 342; (1951) 82 C.L.R. 275.

For these reasons the appeal should be allowed. The declaration in the decretal order below should be set aside and in lieu thereof it should be declared that the residuary estate is divisible amongst the twelve living beneficiaries and the personal representatives of Thomas O'Flaherty deceased per capita. The costs of all parties of the appeal should be paid out of the estate as between solicitor and client: Neil v. McDonnell [4] .