Matthews v Williams [1941] HCA 32; (1941) 65 CLR 639 (7 November 1941)

Matthews Defendant, Appellant; and Williams and Others Plaintiffs and Defendants, Respondents.

Mayo K.C. and Piper, for the appellant and for a respondent great grandchild and a respondent great great grandchild.

H. N. Tucker, for representatives of nephews and nieces entitled under the gift over.

Ligertwood K.C. (with him Culshaw), for respondent grandchildren and representatives of next-of-kin.

In this appeal we are called upon to interpret and apply an ambiguously expressed trust in a will which came into effect in 1867. The testator, who died in that year, left an estate which at the time of his death was valued at not more than £5,000. The period of distribution was fixed by the will at the death of the last surviving child of the testator, an event which did not occur until 5th April 1940. In the intervening seventy-three years the value of the estate grew, until it is now estimated at £82,500. The trust which must be interpreted is that defining the class among whom, and the manner in which, the trust estate is to be distributed. The testator left him surviving seven children, whose ages went from fourteen to thirty-five years. His general plan of disposing of his residuary real and personal estate, called his "trust estate," provided, until the death of the last surviving child, for the division of the income amongst his children, substituting, upon the death of each up to the last survivor, the children of the deceased child as the recipients of the share of income which their parent otherwise would have received. The will then goes on to declare that when and so soon as the whole of his children should be dead the trustees should call in investments and sell the trust estate. Then comes the trust in question, which is expressed as follows:—

All seven children of the testator had children living after his death. But before the period of distribution some of their children, the testator's grandchildren, had died. Of those who died some left children, that is, great grandchildren of the testator. Of these some died before the period of distribution leaving children in their turn; great great grandchildren of the testator. These great great grandchildren were living when the period of distribution came.

Richards J., from whose decision the appeal is brought, held that the objects of the gift are the grandchildren of the testator, and that remoter issue are not included. He further held that the stocks according to which the distribution is to be made are constituted by the seven children of the testator, and that the trust estate is to be divided into seven equal parts, one of which is to be distributed equally among the children living on 5th April 1940 of each of the testator's children. It appeared that in the case of one of the testator's children, the eldest child, all his children died before that date. Richards J. held that there was an intestacy as to that one-seventh share. The intestacy appears to have been brought about by a construction which his Honour placed on the words "the issue of deceased children taking by substitution ... the respective shares only which their deceased parent would if living have taken."

These words were treated as limiting the share which the children of any of the testator's children could take to one-seventh. Accordingly, although the gift is to a class consisting "of issue then living" and is a gift of the whole trust estate, it was considered that as there was "issue," in the sense of children, then living of only six of the testator's children, one-seventh of the trust estate was undisposed of. Why the words in question should be treated as limiting the interest to be taken by each family of grandchildren to one-seventh we have not found it easy to understand. Doubtless the expression, "the respective shares only which their deceased parent would if living have taken," defines the share to be taken by each family of grandchildren (or remoter issue). It defines it by reference to an hypothesis, namely, that the deceased parent remained alive. On that hypothesis you are to find what share he would have taken. Then his offspring or issue get that and no more. But the construction producing the intestacy seems to have added another hypothesis, namely, that all six of that parent's brothers and sisters also lived, or at all events left issue who survived to the period of distribution.

Only on that added hypothesis would the trust estate be divided in seven parts or shares. We can see no justification at all for importing this further hypothesis or condition into the definition of the share to be taken by each family of children or remoter progeny.

It appears to us that all the words mean is that, in the case of each substitution in turn, you are to take the events as they actually existed, with one exception, namely, that you are to suppose that the particular parent of the children to be substituted was still alive and then for the purpose of ascertaining their share inquire what, in the events as they have actually turned out in all other respects, would have been the interest that parent would have taken.

In our opinion, whether the word "issue" means "children" or includes remoter descendants, the limitation to "issue then living" is an ordinary class gift. It is a gift of the whole trust estate to the persons who qualify as members of the class, and there can be no lapse, that is to say, unless there is no-one at all who qualifies as a member of the class and unless the gift over also fails.

We are unable, therefore, to see how the fact that the word "issue" is construed as meaning "children" can result in an intestacy. If "issue" means "children," then in the events that happened the consequence would be a division of the estate into six equal parts, a part for every one of the six families of grandchildren surviving the period of distribution for equal distribution among the members of that family then living.

But we have come to the conclusion that the gift to "issue" is not limited to children of the testator's children, but includes remoter descendants. "Issue" is a word with a clear prima-facie legal meaning. It means descendants or progeny. No doubt "issue" is a flexible word and its prima-facie application may be restricted by any sufficient indications appearing in the documents: Cf. In re Birks[1]. But "the essence of the word issue, which primarily means all descendants, is totality rather than succession" (In re Cust; Glasgow v. Campbell[2], per Cussen J., distinguishing "issue" from "heirs of the body").

"The best rule of construction is that which takes the words to comprehend a subject that falls within their usual sense, unless there is something like declaration plain to the contrary" (Church v. Mundy[3], per Lord Eldon, applied to "issue" by Knight-Bruce V.C. in Head v. Randall[4], and by Isaacs J. in Campbell v. Glasgow[5]).

The word "issue" therefore ought not to receive a secondary or restricted meaning unless, upon a consideration of the whole document, it satisfactorily appears that it was so intended.

While the will does contain some indications which, by themselves, might support an inference that "issue" is used for "children," none of these indications is irreconcilable with the natural or wider meaning and, on the other side, one part of the trust is consistent only with that meaning, or at all events with a meaning which lets in issue no less remote than grandchildren of the testator's children. That part is comprised in the words, "the issue of deceased children taking by substitution as tenants in common the respective shares only which their deceased parent would if living have taken." This passage clearly implies that a "deceased child" would, but for his death before the period of distribution, have taken a share under the trust, and on that footing it directs that a substitution shall be made and the issue of a parent shall take his share. Now it is undeniable that no-one nearer to the testator than his grandchildren could take under the trust. For them no-one closer can be substituted than his great grandchildren, that is, children of his deceased grandchildren. It follows, therefore, that great grandchildren of the testator are comprehended in the gift. But they are not brought in under a distinct limitation amounting to a grant or gift to them. They are brought in by force of part of the description of the "course of distribution" among the "issue" to whom the trust estate is limited. They are not substituted as a new class outside the main class and introduced to take by representation in lieu of a prospective member of that class who failed to attain membership. They are dealt with as part of the only class to which the gift is made, namely "issue," and are mentioned in directing the mode of division among the class. Indeed, "a course of distribution" is an expression which almost imports successive generations as donees.

We are unable to regard this part of the trust as a mistakenly expressed reference to the shares which the testator's children took in income. It appears to us to be the one part of the trust which is incapable of more than one application. It is true that the word "issue" in the expression "issue of deceased children" might be read down to "children," and that the word "parent" might be relied upon for the purpose. But that doubt or uncertainty of meaning is beside the point that the direction in which it occurs is unambiguous in its irreconcilability with an interpretation of the trust which would confine to grandchildren of the testator the class to whom, under the name of "issue" of his children, the trust estate is limited.

The consideration which appears to us to have most weight amongst those which may be said to tend in favour of restricting "issue" to "children" in the description of the donees is the form of the gift over. It must be conceded that according to its more natural meaning it would amount to a gift over of the whole trust estate if at the period of distribution there were no grandchildren of the testator living.

If the gift over were incapable of any other meaning the trust would then present two opposing indications of the meaning of the word "issue"; one, that portion providing for the representation of deceased grandchildren, logically and verbally inconsistent with a restriction of "issue" to "children," and the other, the gift over, forming what must be felt to be an incongruity with any wider meaning. Perhaps, even in such an opposition, it would be right to allow the prima-facie general meaning of the word "issue" to prevail. But the gift over is not incapable of another meaning. Upon examination it appears to be not unambiguously expressed. It is, we think, susceptible of a construction consistent with the wider meaning of "issue." The words "child or children issue of any or either of my children" are not the same as "no child of any of my children," and they do not necessarily have the same meaning.

The words "child or children" may be used not as importing immediate parentage, and the word "issue" may be inserted to show that a remoter degree of descent than child is included. As will appear, the words "sons" and "daughters" are used in an earlier part of the trust in a sense which does not in our opinion import the parentage of any particular generation of issue, and throughout the trust words of lineal relationship are used, as we think, with what has been called a sliding application, that is, so that they apply between any two successive generations toties quoties. In such a context we think that the gift over should be understood as speaking of the failure of any final class of children descended from the testator, and not as referring to the failure of his grandchildren.

It is, however, convenient, before dealing with the further considerations which on one side or the other affect the interpretation to be placed upon the trust, to state in full the meaning which, in the result, we place upon it. We construe it as a stirpital limitation to lineal descendants of the testator, unrestricted in point of degree of propinquity, but limited to those living at the period of distribution who attain full age or, being females, marry. We think that the relative clause "who being a son or sons" &c. has for its antecedent the words "such issue": the "such" and the "who" correspond. The "who" might correspond with the "any" in "any child or children of mine," but though formally it might be better to connect them, yet in substance the probabilities of sense are against that construction.

The words "being a son or sons" and "being a daughter or daughters" we do not read as referring to a filial relation with the testator's children to the exclusion of remoter issue of the testator. The reference is to individuals, members of the class denominated "issue," and they are spoken of as sons and daughters because the whole gift is based upon the idea of representation in successive generations by children of deceased parents who would have taken if living.

We interpret the "course of distribution" as referring to the successive substitution of children for a parent whose death excluded him from the class of donees.

It is well settled that a course of distribution according to the stocks implies that descendants of a lower degree are not let in to compete with any ancestor of theirs, parent, grandparent, et cetera, but only by way of representation of a deceased ancestor.

"According to the stocks" means a stirpital distribution going through the stocks, beginning with the testator's children as the stirpes according to which the first division into shares is made. This must be so unless "issue" means "children."

The words "issue of deceased children" we regard as an expression of general application, and not confined to any particular generations. It is possible in this phrase to limit issue to "children," and in one sense that must really be its meaning. For the object of this part of the trust is to substitute children for parent toties quoties. But we do not think that "issue" should be read as "children" in such a sense as to restrict the application of the provision to children of the testator's children or to children of his grandchildren. Its application is what has been described as "sliding." In the same way "deceased parent" refers to every given case of the decease of an earlier possible taker leaving children surviving the period of distribution.

The gift over we regard as taking effect only on a total failure of issue at the period of distribution. The subject of the gift over is described by the relative "same." We think the antecedent of this word is "my trust estate," not "respective shares."

In support of the construction by which the word "issue" was restricted to "children," so that only grandchildren of the testator would be objects of the trust, the main reliance was placed upon four matters: (1) the gift over; (2) the words "their deceased parent"; (3) the words "being a son or sons" and "being a daughter or daughters"; (4) the manner in which the word "issue" is used elsewhere in the will.

(1)

To what we have already said about the gift over it is unnecessary to add, except perhaps to say that though the strength of the argument it provides in favour of a restricted meaning of "issue" is great it depends upon the prima-facie or natural effect which the words of the gift over produce upon the mind when it is read independently and without careful study of the gift which precedes it. That effect weakens in face of a close scrutiny and consideration of the preceding trust, and the countervailing argument provided by the passage beginning "in a course of distribution" is seen to be much stronger, because it shows unequivocally that issue at least extends to great grandchildren of the testator. At the same time closer consideration of the gift over shows that its language is not incapable of a meaning reconciling the gift over with the full connotation of "issue," the meaning we have already discussed.

(2)

The occurrence of the expression "their deceased parent" in relation to "issue" led counsel to invoke the rule in Sibley v. Perry[6].

This rule is briefly stated by Mr. Hawkins, as follows:—"Where the parent of issue is spoken of, the word issue is prima facie restricted to children of the parent" (Hawkins on Wills, 3rd ed. (1925), p. 114). Whether there is such a rule, whether it is a sound or an unsound rule, and whether it should be considered as arising from Sibley v. Perry[7] or Pruen v. Osborne[8] are matters on which judicial expressions of opinion have varied: Cf. Ralph v. Carrick[9] and In re Timson; Smiles v. Timson[10].

Perhaps the best discussion of the rule will be found in the judgment of Skerrett C.J. in Guardian, Trust, and Executors Co. of New Zealand Ltd. v. Ramage[11].

In Whiting v. Moffatt[12] Holroyd J. made a clear statement of the principles, which it is convenient to set out, the more so because it is illustrated by reference to Ross v. Ross[13]. It is as follows:—"The word issue in its sense of progeny is one of ambiguous import. It may be restricted to children, or it may be extended to all the descendants of a common ancestor. According to the ordinary rules for the construction of wills, a gift to the issue of any person is a gift to all the living descendants of that person (Davenport v. Hanbury4(1796) [1796] EngR 2511; 3 Ves. 257 [30 E.R. 999].). But if property is given to such of several persons as may be living at the testator's death, or other stated period, with a direction that the issue of any of them who may die before that period shall take the shares which their parents respectively would have taken, if living, or words to that effect, or even if the gift be to such of the several persons as may be living at the period fixed, and to the issue of such of them as may be then dead, with a similar direction, in either of these cases the general rule is, that issue must be translated children (Sibley v. Perry5(1802) [1802] EngR 323; 7 Ves. 522 [32 E.R. 211].; Pruen v. Osborne6(1840) [1840] EngR 829; 11 Sim. 132 [59 E.R. 824].; Bradshaw v. Melling7(1853) [1853] EngR 790; 19 Beav. 417 [52 E.R. 412].; Smith v. Horsfall8(1858) [1858] EngR 917; 25 Beav. 628 [53 E.R. 776].; Maynard v. Wright9(1858) [1858] EngR 1237; 26 Beav. 285 [53 E.R. 908].). These rules, however, are not rigid, and will yield to slight expressions of the testator's intention, to be collected from other parts of his will. For instance, where the gift to the issue is distinct, and followed by the direction that the issue shall take the parent's share, the question may arise whether the word parent is limited to the specified persons who would, if living, have taken to the exclusion of all issue. Thus in Ross v. Ross3(1855) [1855] EngR 671; 20 Beav. 645 [52 E.R. 753]. there was a bequest to Christian Ross for life, with remainder to all her children who should be living at the time of her decease, and the issue then living of such of her children as might have died in her lifetime each of her surviving children to take an equal share; and the issue, if more than one, of such of her children as might have died in her lifetime, to take equally amongst them the share which their parent would have been entitled to if he or she had survived the said Christian Ross, and if but one, then to take the child's share. There was a gift over on the death of Christian Ross, without leaving a child, or issue of a child. One of Christian Ross's children predeceased her, leaving no children, but a grandchild, who survived Christian Ross; and it was held that the grandchild took a child's share along with Christian Ross's children. The reason of the decision was, that the word parent was not brought into such correlation with the word issue as to indicate that the meaning of issue was restricted, and that there was nothing to limit the generality of the word issue in the gift over"[21].

The foundation of the rule is the relationship implied by the use of the word "parent" between a first or earlier taker or person otherwise referred to and a second or subsequent class of takers called "issue." It is evident that in applying the rule the first step to take is to identify the earlier taker or takers to whom the word "parent" is applied. When this is done and it appears, as of course in most instances it naturally does, that ascertained or ascertainable persons are regarded as standing in the relation of parent and child to the issue, an inference arises that by issue the testator meant children.

But, adopting the language of Cotton L.J., in Ralph v. Carrick[22], if there is nothing to show that the person referred to as a parent must of necessity be the ascertained person the foundation of Lord Eldon's decision upon the third clause of the will in Sibley v. Perry[23] has no application. This point is brought out by Sargant J. in In re Burnham[24]. He says: "In Pruen v. Osborne5(1840) [1840] EngR 829; 11 Sim. 132 [59 E.R. 824]., Smith v. Horsfall6(1858) [1858] EngR 917; 25 Beav. 628 [53 E.R. 776]., and other cases of the kind, the only persons expressly mentioned as donees under the original gift, and for whose decease provision is expressly made are the first generation. ... It is therefore natural to limit provisions for representation to the case of this generation, and when this is done, the force of the word parent almost necessarily limits the class of introduced representatives to the next succeeding generation."

Skerrett C.J.[27] was even more definite. "The so-called rule applies only where the word issue can be clearly referred to the issue of a parent who if living would be the first taker under the gift"; and again, "I think that to apply the principle of Sibley v. Perry3(1802) [1802] EngR 323; 7 Ves. 522 [32 E.R. 211]. you must be satisfied that the word parent when used in relation to issue must point to the first taker under the trust."

Now in the present case the donees are a class living at a specified event described as issue, that is, prima facie, successive generations then living. There is to be a stirpital distribution among them. It is quite clear that when a possible taker of an earlier generation, because he predeceases the event, does not qualify for membership of the class, the next generation is substituted to represent him, and it is equally clear that the relationship between him and the next generation is that of parent and child. Upon the question whether the earlier generation is confined to a particular degree of propinquity to the testator the use of the word "parent" can throw no light. But it can and does show that there must be at least two generations contemplated by the word "issue," and by doing so provides a strong reason against the construction adopted by Richards J. confining the class of donees to the grandchildren of the testator.

(3)

The expressions "son or sons" and "daughter or daughters" no doubt suggest a relationship of parent and child, and as the testator's children are the generation last referred to it is at first a natural association of ideas to treat that generation as the parent. But again, as the clause proceeds it becomes clear that under the term "issue" more than one generation is comprehended. There is therefore less ground than might at first appear for supposing that it is between the testator's children and their children and no other successive generations that the filial relation is contemplated. It is clear that within the class called "issue" the relation of parent and child is contemplated and there is no occasion for thinking that such a relation with a generation antecedent to and outside that class was in the mind of the draftsman.

Farrant v. Nichols[29] was decided on a very different limitation, one in which Lord Langdale found in the expression "whether son or daughters" a restrictive intention. No general rule was laid down, and Lord Langdale observed that "each case depends on the peculiar expressions used, and the structure of the sentences"[30].

The will contains a number of other instances in which the word "issue" is used. In one case it is used to cover remoter descendants. In one or two it is doubtful whether it has or has not a restricted meaning. In others the context shows that "children" only were intended.

The fact appears to be that the draftsman employed the word wherever the dominant idea in his mind was lineal descent. Sometimes he either by design or by chance qualified the degree of proximity; sometimes he did not.

In our opinion there is no guidance to be obtained from the other instances in the will of the use of the word "issue."

The trusts under consideration are like those of the will before Eve J. in In re Sutcliffe; Alison v. Alison[31], and it appears from the report[32] that he placed the same construction upon "issue" as we have adopted. Unfortunately his reasons are not reported.

A question has arisen of a peculiar nature in reference to three of the testator's great grandchildren. Their parents, who are both dead, were cousins, the mother being a daughter of a deceased daughter of the testator and the father a son of a deceased son of the testator. The question is whether they are entitled to share by way of representation or substitution both in the interest which, if she had lived, their mother would have taken, and in that which, if he had lived, their father would have taken. In our opinion the answer must be yes. The purpose of the trust is to distribute the trust estate according to the stocks, not equally throughout. Equality is the standard only as between stocks and as between children of a deceased parent in dividing his share.

Here the three great grandchildren in question have separate titles, so to speak, to share in the division of different interests. This is the view taken by Sargant J. in In re Burnham[33]. The limitation in Pruen v. Osborne[34] which was relied on was quite different and the decision appears to us to have little or no bearing upon this particular question.

In our opinion the appeal should be allowed and the order of the Supreme Court discharged except as to costs and in lieu thereof it should be ordered that the questions in the originating summons should be answered as follows:—

The costs of all parties of and incidental to the appeal should come out of the estate, those of the trustees as between solicitor and client.

Appeal allowed. Order of the Supreme Court discharged except as to costs, and in lieu thereof order that the questions in the originating summons be answered as follows: 1. (a) The word "issue" includes remoter issue in both cases. (b) The word "children" in the expression "issue of deceased children" applies whether the deceased child was a grandchild or remoter issue of the testator. (bb) The distribution is to begin by taking the seven children of the testator who had issue living at the period of distribution, viz., 5th April 1940. (c) The descendants of the testator living on that date but so that children or remoter issue of any descendants of the testator living on that date shall not compete with their parent or ancestor and shall be excluded. (cc) They are entitled to take one share as issue of Lavinia Mary Matthews and one share as issue of Ernest William Matthews. The costs of all parties of and incidental to the appeal to be paid out of the estate; those of the trustees as between solicitor and client.

Solicitors for the respondents, L. von Doussa; Thomson, Buttrose, Ross & Lewis; Genders, Wilson & Pellew; Piper, Bakewell & Piper; Baker, McEwin, Ligertwood & Millhouse; J. F. Astley.

[1] (1900) 1 Ch., at pp. 418, 420.

[2] (1919) V.L.R 221, at pp. 254, 255.

[3] (1808) 15 Ves. 396, at p. 406 [1808] EngR 428; [33 E.R. 804, at p. 808].

[4] [1843] EngR 378; (1843) 2 Y. & C.C.C. 231, at p. 235 [1843] EngR 378; [63 E.R. 101, at p. 103].

[5] [1919] HCA 57; (1919) 27 C.L.R. 31, at p. 51.

[6] [1802] EngR 323; (1802) 7 Ves. 522 [32 E.R. 211].

[7] [1802] EngR 323; (1802) 7 Ves. 522 [32 E.R. 211].

[8] [1840] EngR 829; (1840) 11 Sim. 132 [59 E.R. 824].

[9] (1873) 11 Ch. D. 873.

[10] (1916) 1 Ch. 293; (1916) 2 Ch. 362.

[11] (1927) N.Z.L.R. 288.

[12] (1898) 20 A.L.T. 113; 4 A.L.R. 246.

[13] [1855] EngR 671; (1855) 20 Beav. 645 [52 E.R. 753].

[14] [1796] EngR 2511; (1796) 3 Ves. 257 [30 E.R. 999].

[15] [1802] EngR 323; (1802) 7 Ves. 522 [32 E.R. 211].

[16] [1840] EngR 829; (1840) 11 Sim. 132 [59 E.R. 824].

[17] [1853] EngR 790; (1853) 19 Beav. 417 [52 E.R. 412].

[18] [1858] EngR 917; (1858) 25 Beav. 628 [53 E.R. 776].

[19] [1858] EngR 1237; (1858) 26 Beav. 285 [53 E.R. 908].

[20] [1855] EngR 671; (1855) 20 Beav. 645 [52 E.R. 753].

[21] (1898) 20 A.L.T., at pp. 113, 114; 4 A.L.R., at p. 248.

[22] (1879) 11 Ch. D., at p. 886.

[23] [1802] EngR 323; (1802) 7 Ves. 522 [32 E.R. 211].

[24] (1918) 2 Ch., at p. 204.

[25] [1840] EngR 829; (1840) 11 Sim. 132 [59 E.R. 824].

[26] [1858] EngR 917; (1858) 25 Beav. 628 [53 E.R. 776].

[27] (1927) N.Z.L.R., at pp. 291, 293.

[28] [1802] EngR 323; (1802) 7 Ves. 522 [32 E.R. 211].

[29] [1846] EngR 501; (1846) 9 Beav. 327 [50 E.R. 370].

[30] (1846) 9 Beav., at p. 330 [50 E.R., at p. 371].

[31] (1934) Ch. 219.

[32] (1934) Ch., at p. 221.

[33] (1918) 2 Ch., at p. 206.

[34] [1840] EngR 829; (1840) 11 Sim. 132 [59 E.R. 824].

High Court of Australia

Matthews v Williams [1941] HCA 32; (1941) 65 CLR 639 (7 November 1941)

HIGH COURT OF AUSTRALIA