Scottish Court of Session Decisions

At advising on 13th July 1933, the Lord President intimated the decision of the Court, and stated that the opinions of the judges would be circulated,—

LORD PRESIDENT (Clyde).—The scheme of the settlement is as follows:—After payment of debts (purpose 1), the settlor's widow is given the liferent of his residential estate or the proceeds thereof if sold (purpose 2), and also the liferent of half his general estate (purpose 3). Then the residue of his entire estate (that is, everything after satisfying the liferent interests provided to the widow) is directed to be held "for behoof of my children" (purpose 4); and by the same purpose, his trustees are directed to "divide" the half of the general estate not subject to the widow's liferent—hereinafter referred to as the first moiety of the estate—"equally among my children share and share alike," sons' shares being payable at the age of twenty-five, and daughters' shares at that age or on marriage. Next, on the death of the widow (in the event of the settlor predeceasing her), the trustees are directed to "divide" the remainder of the settlor's entire estate—hereinafter referred to as the second moiety—"into equal shares for behoof of my children," which shares are to be paid to and to vest in sons on their attaining the age of twenty-five, and to be settled on daughters and their issue on such daughters' marriage or on their attaining the age of twenty-five (purpose 5). It is to be observed that, according to the grammatical construction of purpose 4 read as a whole, the direction therein contained to divide equally among the settlor's children, share and share alike, would apply to half the residential estate as well as to half the general estate. This reading would, however, be repugnant to the scheme of the settlement; and the trustees have (without objection) interpreted the latter part of purpose 4 as referring only to the first moiety of the estate. The settlement also contains two declarations. The first of these provides that "the said provisions in favour of my said children shall not vest until the respective terms of payment of the same." The second is a provision in favour of the issue of children who predecease the settlor or his widow, as the case may be; and its terms will be the subject of consideration hereafter.

The settlement was executed less than a year before the settlor died. At the date of the execution of the settlement he had seven children living. No further children were born to him between the date of the settlement and the date of his death. All seven survived him, and all attained the age of twenty-five, or (in the case of daughters) married. The settlor's widow died thirty years later. Only two out of the seven children survived her; they are the second and third parties. Of the five children who predeceased her, one left a son who is the fifth party; and another left four children (two of whom are in minority)—they are the sixth parties.

It is common ground that the interests in the second moiety of the estate to which the second, third, fifth, and sixth parties are respectively entitled became vested in them on the death of the widow. The whole dispute is as to the computation of the amounts of those interests. The second and third parties maintain that the computation must be based on the division of the second moiety into one-seventh parts; while the fifth and sixth parties contend that it should be based on a division into one-fourth parts.

Computations of this kind are apt to be complicated; and it may serve a useful purpose to take a simple illustration.

Suppose a family settlement "in favour of the testator's children A, B, C, and D, share and share alike, or in equal shares; provided that if any of the said children should die before the testator, the share of the child so dying shall accresce to such of the others as may survive the testator." If A and B predecease the testator without issue, it is obvious that each of C and D will take in two capacities—first, as institute under the will in his own right; and second, as conditional institute under the will in right of each of the predeceasers. In the first capacity, each will take one-fourth; in the second capacity each will take one-half of each of the one-fourth parts in which the respective predeceasers were institutes under the will.

The case supposed is so simple that the ascertainment of the amounts of the total shares taken by C and D at one-half each hardly seems to require a computation on the basis of one-fourth parts. But, that the true basis of the computation is a division into one-fourth parts becomes manifest if it be supposed that—the facts being otherwise the same—A had predeceased the testator leaving issue. In that event, the conditional institution under the will in favour of C and D in the share in which A was the institute suffers defeat; because the conditio si institutus sine liberis decesserit is implied as a condition of the right of C and D to take under that conditional institution. The institutus (A) has not predeceased without issue, but on the contrary leaving issue; and the result is that the issue, in the capacity of implied conditional institutes in the share in which their parent was the institute, must be preferred to C and D. While this is so, the conditional institution of C and D in the share in which B was the institute remains unaffected. The computation, therefore, would be impossible unless on the basis of one-fourth parts, thus:—

1. (a) To each of C and D, as institute, one-fourth part.

2. (b) To A's issue, being implied conditional institutes (under the conditio si institutus) in place of A as institute—one-fourth part.

3. (c) To each of C and D, conditionally instituted under the will in place of B as institute—one-half of one-fourth part.

In this way, the computation of the total shares taken (respectively) by A's issue and by C and D together works out at one-fourth and three-fourths. The position would be precisely the same if the conditional institution implied in the conditio si institutuswere expressed in a declaration in the will.

The reason of all this is simply that a conditional institution always is—and must be—the pendant of an institution; and that, where there is (1) a plurality of institutes provided by the will in equal interests or shares in the estate, and (2) a conditional institution in place of only one of them—or several conditional institutions in place of different institutes—the computation can only be based on the equal interest or share originally provided by the will to each of the institutes. Incidentally, the illustration just given explains why it is that the conditio si institutus (or an express clause of like import) gives the predeceaser's issue only their parent's "original" share, and is incapable of applying to "accrescing" shares—See Lord Chancellor Westbury's judgment in Young v. Robertson (second case).

The case for the second and third parties rests on three propositions. First, that the class-gift in the settlement in favour of "my children" instituted each member of the class as heir in an equal share in both moieties of the estate, payable (in the case of the first moiety) after the death of the settlor, and (in the case of the second moiety) after the death of his widow. So stated, the proposition takes no account of the post-ponement of vesting until marriage or the attainment of the age of twenty-five; but, as all the children married or attained that age, this is of no moment. Second, that by the right of accretion in favour of survivors which is inherent in a class-gift to children, the second and third parties were (respectively) conditionally instituted in place of each of the other children who died childless before the settlor (as regards the first moiety), or before the widow (as regards the second moiety). Third, that the declaration alluded to above gives the issue of children who might die before the settlor or his widow (as the case might be) just what the conditio si institutus would have given them in any case; and that, accordingly, the fifth and sixth parties are entitled only to what their respective parents would have been entitled to as institutes under the settlement.

Each of these propositions must be separately examined. Whatever controversy there may be with regard to the first two of them evidently turns on the legal qualities of a class-gift to children; the third is a matter of construction.

First. It cannot, I think, be disputed that a testamentary gift to "my children," as a class, is a valid mode of instituting the children as the testator's heirs. It would be meaningless otherwise. And, if the gift is to be shared equally among them, it cannot, I think, be gainsaid that each member of the class is instituted as heir in an equal share with the others. No doubt, it is a condition of any of the children taking under the gift that he, or she, shall survive to take. This is so with regard to all heirs. But heirs are instituted when they are named in the will—that is, at the date when the testator signs his will—not when (if ever) they survive to take benefit by their institution. Suppose that, in the present case, all the settlor's seven children had survived the widow (as they did in fact survive the settlor) they would each have taken an equal share in both moieties (as they actually did in the first moiety); and the capacity in which each would have taken an equal share would have been that of an institute under the settlement. There is no other capacity in which they could have taken it. Again, if (as happened) only two of the children survived to participate in the second moiety, they can only take in respect (1) of their institution under the will to an equal share with the other children, and (2) as conditionally instituted by survivorship in place of each of the other children who predeceased without issue. Survivors cannot be instituted otherwise than conditionally, for the simple reason that survivorship necessarily implies the institution of each member of the group to which the survivors belong. The argument on behalf of the fifth and sixth parties was that there was no institution of the members of the class in the settlement; and that the survivors of the children at the date of the widow's death were truly the heirs instituted by the fifth purpose of the settlement in the second moiety. This line of argument was coupled with a construction of the declaration above alluded to, which (it was contended) entitled the issue of predeceasing children to participate equally with surviving children. This was the sole foundation for the claim that the computation in the present case should be made on the basis of one-fourth parts. But (apart from the difficulties which attend the construction of the declaration) this contention involves something very like a legal impossibility. For reasons already indicated, it is plain that, when the survivors of a group are called, they can only take in one or other, or both, of two capacities: (1) as surviving institutes in right of their own shares, (2) as conditional institutes in right of the shares in which others were institutes. Suppose a gift in the form of the following direction:—

"to divide for behoof of my children surviving at my death (or at the death of my wife) in equal shares"

; and suppose that all the testator's children survive the testator (or his wife, as the case may be). Each child would take an equal share with the others, and would do so as an institute under the will in an equal share with the others. They could not take a share at all, if they were not all equally institutes. Again, if only two out of (say) seven children survive, and if none of the predeceasers has left issue, the two survivors would take their own shares which they have duly survived to take—nothing has occurred to deprive them of these—and they would also take (equally among them) the shares of each of the predeceasers. But it is inevitable that they would take the share of each of the predeceasers in the capacity of conditional institutes. There is no other capacity in which they would take them. All testamentary gifts are, of course, subject to the condition that the donee lives to take; but the benefit of survivorship within a group of donees is meaningless and impracticable without invoking the machinery of conditional institution. In a word, as the benefit of survivorship among a group—no matter in what form it is conceived—necessarily implies both institution and conditional institution, the computation of the total share of the estate ultimately falling to each surviving member of the group necessarily starts with the share in which each member of the group was originally instituted. The legal quality of a provision in favour of a class is defined by Lord Curriehill (delivering the judgment of the Court at p. 1014) in Douglas v. Douglas thus:—

"It is true that the provision is in favour of a class, but this means, not that it was in favour of a corporate or aggregate body as a separate person in law, but that an equal pro indiviso share of the residue was provided to each individual included in that class; and the vesting of the right to such a share in each of these individuals at the death of the testator was quite consistent with their number being either diminished by deaths or increased by births before the arrival of the term of payment."

This is precisely the doctrine on which the foregoing observations are founded. It is right, however, to advert to the one specialty of a class-gift to children—which is, that if, between the date of the will and the date of the testator's death, further children are born to him alive, each such child becomes a member of the class on its birth, and therefore an institute equally with the others. The only result is that the equal shares provided to the members of the class as existing at the date of the will are liable to a possible defeasance in amount during the testator's life. Nothing of that kind happened in the present case; but, in any event, I do not see anything in this specialty to qualify the soundness of the principles I have endeavoured to state and apply.

Second. I have to some extent anticipated in the preceding paragraph what I have to say on this proposition. It has never been doubted that accretion by survivorship is inherent in a class-gift to children; in other words, that those members of the class who survive the date of distribution (be it the date of the death of the testator, or the date of the death of his widow, or some other postponed date) are conditionally instituted to each of those who predecease it without leaving issue. The general doctrine is stated without qualification in M'Laren on Wills and Succession, (3rd ed.) vol. i., pp, 631–632, paragraphs 1143–1145. The benefit of survivorship is rarely expressed in a class-gift to children, simply because it is inherent in it; and nothing is better settled than that the benefits of a class-gift (that is, the vested rights therein) are confined to those members of the class who survive the period of distribution, simply because it is only then that the survivors who are actually to take the gift can be ascertained. The survivors, in short, are the final and definite objects of the gift. There is nothing in this peculiar to class-gifts, except that the benefit of survivorship (or jus accrescendi) is inherently implied in them, and is not therefore usually expressed; while, in gifts to members of a class separately identified, it must be expressed if it is to apply. Even in a bequest "to A, but, if A predecease the testator (or the date of distribution), then to B," the object of the gift can only be ascertained at the testator's death (or other date of distribution); but nevertheless A is institute, and B conditional institute under the will from the date when it is signed. What appears to be a fatal objection to the suggestion, on which (as I have explained) the argument for the fifth and sixth parties is founded, is that, in the case of a gift such as is contained in the fifth purpose of the settlement (as they interpret it), the conditio si institutus could never apply. For, if a gift in the form of a direction to "divide on my death or on my wife's death (as the case may be) for behoof of my children in equal shares" could be read as an institution of the children then surviving, it is plain that the issue of a predeceasing child would appeal to the conditio si institutus in vain. The surviving children would take in the capacity of institutes, and not otherwise; and the parent of the issue would never have been an institute at all. I should be driven to such a conclusion only with the utmost reluctance. It has never been doubted that the conditio si institutus applies in the event of the predecease of one of the members of a class leaving issue. But the conditio si institutus would be incapable of application unless the members of the class were, each of them, institutes under the will.

Third. It remains to deal with the declaration alluded to above. It is in the following terms:—

"Should any of my said children die before or after [me] or predecease my said wife leaving lawful issue such issue shall be entitled to the share that would have fallen to their parent by survivance."

The words, "die before or after [me] or predecease my said wife," mean, I think, just what might have been more neatly expressed by the words, "predecease me or my said wife"—the intention being to make the declaration apply equally to both moieties of the settlor's estate. Further, it seems clear that the word "survivance" (like the word "predecease") refers to the date of the death of the settlor or his wife, as the case may be. The share that would have fallen to the parent who predeceases the settlor or his wife is therefore the share that would have fallen to such parent by survivance of the settlor or his wife. So far at least as form goes, the declaration is nothing but a conditional institution in favour of the issue in place of the parent (as institute) in the interest or share provided to the parent in that capacity by the settlement (in either of the two moieties of the estate). That is the only interest or share that could fall to the parent by the simple fact of his surviving the settlor or the widow. The declaration does not bear to make the issue conditional institutes, along with the children who survive the settlor or his wife, in place of other children who have predeceased the settlor or his wife without leaving issue. Neither the event of the parent having survived such other children, nor the event of such other children having died without issue, is so much as hinted at in the declaration. There is a long series of cases in which similar declarations have been examined to see whether they could be construed so as to include, not merely the parent's "original" share, but also "accrescing shares." I can see no ground for giving to "the share that would have fallen to their parent by survivance" any wider meaning than was given to "the share of their mother, as if she had been in life," in M'Nish v. Donald's Trustees; or "the share which would have fallen to their parent had he or she survived me," in Henderson v. Henderson's Trustees; or "to succeed to their parent's share equally between them in the same manner and as fully as if such parent had survived," in Cumming's Trustees v. White . In none of these cases were the phrases used found sufficient, either expressly or by implication, to go beyond the effect of the conditio si institutus; and they may be contrasted with the phrase which, in Laing v. Barclay, was held to include accrescing shares:—

"In the event of any of my children predeceasing the said term of division, leaving lawful issue, it is my desire that such issue shall represent and be entitled to the proportion which would have been payable to their parent."

The ordinary way of including "accrescing" as well as "original" shares is to do so in so many words. It is not without significance that the testator, or his draftsman, was familiar with the use of such a style, as appears from a later clause in the settlement. It was held in Laing v. Barclay that the general right given to the issue to "represent" their parents in the final distribution, and to receive the "proportion" of the distributable fund which the parents they represented would have received in that distribution, was more than the conditio si institutus would have given them. There is nothing of that kind here. The fifth and sixth parties were unable to point to anything in the words of the declaration in the present case to give it a wider meaning than the practically identical clauses construed in the long series of cases referred to, other than Laing v. Barclay .They pointed out that, in most of the cases, right of accretion was expressly conferred, while in the present case it is only implied from the nature of the gift. This is a distinction of form only, which cannot affect the substance of the matter. Moreover, they conceded that, if the declaration included both "original" and "accrescing" shares, they could not avoid a computation based on a division of the moieties into one-seventh parts, of which their respective parents were provided by the settlement in one each, and of which each of the children who died without issue was similarly provided in one each.

The construction of the declaration on which they founded their case was entirely based on the argument, which I have already found myself unable to accept—namely, that there was no institution by the settlement of the members of the class of "my children," but that the two children who survived the widow were the first heirs of the settlor to be called in the capacity of institutes. Each of the two thus took, under the opening words of purpose 5, a half of the second moiety on the death of the widow. Starting from that point, the fifth and sixth parties construed the declaration as referring to the shares in which their respective parents would have been instituted (so the argument runs) if they had survived the widow—that is, one-fourth each; and they argued that consequentially the shares of one-half each, which vested in the two surviving children in terms of purpose 5, must be deemed to be cut down to shares of one-fourth each—with the result that the computation of the total shares of all four parties must proceed on a division of the second moiety into one-fourth parts. Being unable to accept the foundation of this argument, I cannot accept its superstructure. But, apart from that, I distrust the whole argument as artificial and strained. It seems to me to put a great deal more into the settlement, and into the declaration in particular, than the words will bear; and it is difficult to imagine that either the settlor, or the professional conveyancer who put his wishes into legal form, ever really thought of adopting so novel a mode of testamentary disposition.

I prefer the normal method contended for by the second and third parties, according to which the computation is as follows:—(a) To each of the second and third parties, as institute under the settlement along with the other members of the class of children (seven in all)—one-seventh part; (b) to each of the fifth and sixth parties, as conditional institute in the share in which their respective parents were institutes under the declaration in the settlement—one-seventh part; (c) to each of the second and third parties, as conditional institutes (by virtue of the class-gift) in each of the one-seventh parts in which the three children who predeceased the widow without issue were respectively instituted by the settlement—one-half of three-seventh parts. The result of this computation is that the second and third parties' total shares amount to five-sevenths, and the fifth and sixth parties' total shares amount to two-sevenths.

[His Lordship then dealt with other questions, with which this report is not concerned.]

LORD JUSTICE-CLERK (Alness).—Mr Alexander Miller died on 23rd January 1901, survived by seven children, all of whom attained the age of twenty-five. He left a trust-disposition and settlement, dated 20th April 1900. The trustees original, surviving, and assumed under that deed are the first parties to the case. Mr Miller's widow survived him, and died in April 1931. Only two children of the marriage survived her. They are the second and third parties to the case. Of the five children who predeceased her, two only left issue. Their issue are the fifth and sixth parties to the case. The shares of the other three children of the testator in the portion of the estate in question here lapsed; they take nothing.

The truster conveyed his whole estate to his trustees, and, by the first purpose of the trust-disposition and settlement, he directed them to pay his debts and the expenses of the trust. By the second purpose he bequeathed to his widow the liferent use of certain heritage which he owned, and the free annual income of certain other heritages of which also he was the proprietor. No question arises with regard to these provisions.

By the third purpose of his settlement the testator directed his trustees to pay to his widow one-half of the income of the remainder of his estate during her life. By the fourth purpose the testator provided as follows:—

"That my trustees shall hold the whole residue and remainder of my estate for behoof of my children and divide one half thereof equally among my children share and share alike and shall pay the share thereof falling to each of them on in the case of sons his attaining the age of twenty-five years and in the case of daughters on her attaining that age or on her marriage whichever of these events shall first happen."

The testator's widow and all his children having survived him, and, all his children having attained the age of twenty-five, the first parties duly paid to these children their shares in one-half of the residue as it fell due, in terms of the fourth purpose. These payments exhaust one-half of the residue of the testator's means and estate.

As regards the other half of the residue, the truster provided in the fifth purpose as follows:—

"On the death of my said wife my trustees shall realise the whole remainder of my means and estate and shall divide the same into equal shares for behoof of my children and shall pay to each of my sons the share falling to him on his attaining the foresaid age of twenty-five years but shall pay to each of my daughters on her attaining that age or on her marriage as aforesaid only one-half of the share falling to her from the proceeds of the liferented provision in favour of my said wife in said second purpose of this trust and shall hold and retain the remainder of the share of the residue falling to each of my daughters for her liferent use allenarly and to her issue in fee Declaring that the said provisions in favour of my said children shall not vest until the respective terms of payment of the same and that should any of my said children die before or after my death or predecease my said wife leaving lawful issue such issue shall be entitled to the share that would have fallen to their parent by survivance."

That clause, and, in particular, the latter portion of it, is crucial in the determination of this case.

Questions have arisen as to (1) the number of parts into which the residue of the estate of the testator is divisible, and (2) the parties entitled to receive these. The second and third parties contend that the residue is divisible into seven parts, of which they, as the surviving children of the testator, are entitled equally between them to five parts, the remaining parts being divisible equally per stirpesbetween the fifth and sixth parties. The fifth and sixth parties contend that the residue is divisible into four equal parts, and that the second and third parties take no greater interest in it than do the issue of each predeceasing child. Is the division of the residue then, in the circumstances, to be equal or unequal? That is the problem which we have to solve. Its solution depends on the true interpretation of the fifth clause of the settlement. Now, there can be no doubt that the pivotal consideration in solving such a problem is—What is the enixa voluntas of the testator? If it is clear, then cadit quæstio. I propose to consider the question, first, as if I had to decide it on the terms of the settlement alone, unaffected by any previous decision; and then to inquire whether the conclusion at which I arrive is prohibited by anything which the authorities contain.

The problem truly is—What becomes of the three lapsed shares of residue? (a) Are they given to the survivors exclusively? or (b) Do the issue of predeceasing children share in them equally with the survivors? To the first of these questions I reply—the testator, so far as I can see, has nowhere said that these lapsed shares shall belong exclusively to the survivors. To the second of these questions I reply that the testator clearly has provided for equality of distribution. Had the parents of the fifth and sixth parties been alive at the date of distribution, they would have taken an equal share of the residue, including a proportion of the lapsed shares. Their issue, I think, stand in the shoes of their parents, and, in the absence of any restriction of their right to the original share of their parents, they take the shares which would have been taken by their parents had they survived. That would seem to be the common sense of the matter. Viewing the problem then as a question of the intention evinced by the testator, and neglecting authority for the moment, I should without hesitation accept the contention of the fifth and sixth parties, and reject the contention of the second and third parties.

I proceed, however, to inquire whether that conclusion is excluded by any artificial rule which the authorities disclose. Now, the governing rule which is yielded by all the cases cited in debate, in my opinion, is that the intention of the testator, if clearly expressed, must always prevail. All other rules are subordinate to that rule. Do the cases cited affirm that, in circumstances such as exist in this case, where the testator has provided that the issue shall be entitled to the share that would have fallen to their parents by survivance, the claim of the issue is confined to their parents' original share? I can discover no such rule.

In Young v. Robertson, Clelland, M'Nish, Henderson's Trustees, and Cumming's Trustees, as Mr Russell pointed out, there was an express clause of survivorship. In other words, the lapsed shares were disposed of, in a particular manner, by the testator himself. That is but another illustration of the rule that the will of the testator, if clearly expressed, must prevail. Here there is no express clause of survivorship. That, it seems to me, makes all the difference in the legal result. As Lord Cowan said in Laing(at p. 1150):

"It is impossible to hold that principle applicable"

—i.e., the principle which excludes the issue of predeceasers from benefiting by the doctrine of accretion—"to cases where there is no clause of survivorship, and where, as in the present deed, there is an express declaration of the extent of interest in the succession to be taken by the issue of predeceasing children." Similar views were expressed by Lord M'Laren (at p. 779) in M'Culloch's Trustees. The presence of a clause of survivorship in the cases cited, and its absence in this case, to my mind, involve a different legal result.

The second and third parties contended that, in this case, a clause of survivorship must be held to be implied. That view is based upon a dictum of Lord Trayner (at p. 460) in Cumming's Trustees—in which, be it observed, there was a survivorship clause—to the effect that it is immaterial whether a survivorship clause is express or implied. Lord Trayner's view was obiter, and it is out of alignment, I respectfully think, with the train of authority cited and the principle on which these authorities were based.

The cases affirm, as I read them, that, if it can be shown, by the terms of any particular deed of settlement, that it was the intention of the testator to give the issue of a legatee predeceasing the term of vesting not only their parent's proper share, but also a portion of the accrued share of legatees who died without issue, there is no rule known to the law which precludes the Court from giving effect to that intention; for "the ultimate principle in all cases of testate succession is to carry out the true and expressed voluntas testatoris." In this case I find the voluntas testatoris expressed in the manner contended for by the fifth and sixth parties. The Lord Justice-Clerk, who doubted the decision in M'Nish, to which he was a party, said (at p. 102):

"What is left to the issue of a deceasing sister is not the share specifically left to the deceased, but something quite different, namely, the share which the mother would have taken if she had survived, which would have included both her aliquot share and the proportion of that which would have fallen to the childless deceaser."

These words might have been penned regarding this case.

Giving effect then to what I conceive to be the dominant and indeed determinative consideration, viz., the intention of the testator, finding the instrument under review intractable on that point, and further finding myself undeterred from that course by any rule to be deduced from the authorities to which we were referred, I am for rejecting the contention of the second and third parties, and for affirming the contention of the fifth and sixth parties.

LORD HUNTER .—The difficulty of properly interpreting the language used by a testator is often intensified by a preliminary discussion on alleged settled principles of construction and an examination of cases in which different testators in different deeds have used more or less similar language. The cardinal rule of construction is to find out, if possible, from the language used what must be assumed to have been the testator's intention. As the Lord Chancellor (Lord Westbury) said (at p. 325) in Young v. Robertson :

"The primary duty of a court of construction, in the interpretation of wills, is to give to each word employed, if it can with propriety receive it, the natural ordinary meaning which it has in the vocabulary of ordinary life, and not to give to words employed in the vocabulary of ordinary life an artificial, secondary, and technical meaning."

Accepting this duty, I think that, if the language used by a testator is clear and unambiguous, and an inference as to his intention is unavoidable, the examination of preceding decisions is merely a work of supererogation.

By the fourth purpose of the settlement which we have to construe the testator has directed his trustees to "hold the whole residue and remainder of my estate for behoof of my children and divide one half thereof equally among my children share and share alike and [to] pay the share thereof falling to each of them on in the case of sons his attaining the age of twenty-five years and in the case of daughters on her attaining that age or on her marriage whichever of these events shall first happen."

The fifth purpose is in these terms:—

"On the death of my said wife my trustees shall realise the whole remainder of my means and estate and shall divide the same into equal shares for behoof of my children and shall pay to each of my sons the share falling to him on his attaining the foresaid age of twenty-five years."

As regards the daughters' shares special directions are given which it is unnecessary to quote.

The settlement contains a general declaration "that the said provisions in favour of my said children shall not vest until the respective terms of payment of the same and that should any of my said children die before or after my death or predecease my said wife leaving lawful issue such issue shall be entitled to the share that would have fallen to their parent by survivance."

The testator was survived by seven children, all of whom survived the age of twenty-five. They received payment of their respective shares of the half of residue dealt with in the fourth purpose of the settlement. Only two of the children survived the testator's widow. Of the five predeceasing children, two left issue. On behalf of the surviving children it is maintained that they are entitled to five-sevenths of the residue which falls now to be divided, the issue of the two predeceasing children being entitled to the remaining two-sevenths. For the issue, however, it is maintained that the division should be into four parts, the surviving children taking no greater interest in the residue than the issue of each predeceasing child.

When the fourth and fifth clauses are read together I think that, while the testator intended the whole of his residue to be held by his trustees, he has used language that shows that a child's share in the first half of his residue is or may be a different proportion of the fund to be divided from what it is as regards the second half of residue. Survivorship of the testator is a condition of a child having a share in the first half of residue. Survivorship of the testator's widow is a condition of a child having a share in the second half of residue. Until the term of division no child has a share in the testator's estate. Until the term of payment no child has a vested interest in his share. In a footnote to the case of Young, I observe that during the argument the Lord Chancellor asked, "What is share? Must it not be in possession? Is it not from ‘shear,’ to cut off, or divide?"

A share being determined at one or other of the two events, there can only be a lapsed share if a child survives one or both of these events but dies before reaching the age of twenty-five. In the event which has occurred the trustees are directed to realise the whole remainder of the testator's estate and to divide the same into equal shares for behoof of his children, and pay to them their shares on their attaining the age of twenty-five. The declaration in favour of children's issue brings them in as representing their parent. They are to get that share which would have fallen to their parent by survivance. That share would have been a fourth to each of the children who predeceased the widow leaving issue. I think, therefore, that the testator's intention, in the event which has occurred, clearly was that the issue of predeceasing children should represent their parent and take the share, no more and no less, which would have been taken by that parent.

The argument for the surviving children proceeds on the assumption that you must regard the residue of the testator's estate as divisible into seven parts, one part being the share of each child who survived the testator. I have already indicated my view that the language used by the testator, while admitting this view so far as the first half of residue is concerned, excludes it in the case of the second half of residue. If, however, I am wrong in this view, and if the residue is to be treated as divisible into seven parts, I am still unable to accept the reasoning which gives the surviving children greater rights than the issue of predeceasing children take. It is true that a division into seven shares gives you three lapsed shares in the case of the children who died without leaving issue, and that these lapsed shares will accresce for the benefit of the other beneficiaries. The testator has not, however, given accrescing shares to the survivors of the children, and I do not think that, in such circumstances, there is any rule of construction which compels the Court to disregard a testator's express instructions that there is to be an equal division of his estate among his children and that the issue of a predeceasing child are to take the share that would have fallen to their parent by survivance.

It is well settled by the law of Scotland that, where a child takes a provision in virtue of the rule si institutus sine liberis decesserit, the right so taken is limited to the right which the parent would have taken on the assumption that distribution took place at the time of death. Any additional right to a share of the estate accrescing subsequent to that date is excluded. Similar reasoning has been held to exclude the issue of a predeceasing beneficiary from participating in an accrescing share where that is given in express terms to the surviving members of a class who are to benefit. The leading illustration of this rule of law is to be found in the second appeal of Young v. Robertson, decided in the House of Lords. In the first appeal it was held that, where the residue of the testator's estate was destined on his wife's death to certain named beneficiaries, with a declaration that, if any of the residuary legatees should die without leaving lawful issue before his or her share vested in the party or parties so deceasing, the same should belong to and be divided equally among the survivors, no vesting took place until the period of distribution. In the second appeal it was decided that the issue of a predeceasing child, although entitled to his father's share, was not entitled to participate in the lapsed shares of other residuary legatees who had predeceased the liferentrix without issue. In the course of his speech the Lord Chancellor said (at p. 341):—

"The testator … contemplating the possibility of the death of one of the residuary legatees before the period of distribution, has introduced a clause of conditional institution … that in the event of any one or more of the legatees dying without leaving lawful issue during the life of the liferentrix … the share of the individual or individuals so dying shall go to the survivor or survivors."

This passage shows that that eminent judge held the claim of the issue excluded by the very terms of the settlement. He added (at p. 342):

"I think it is impossible to hold that any principle of surrogation or substitution would give to the child that which the father by no possibility could have taken. The very principle of surrogation is merely to place the child in the room of the father; but it would be contrary to all principle to make the surrogation extend to give to the child a right which the father by no possibility could have."

It appears to me that the doctrine expounded in the House of Lords in the case of Young can apply only where there is an express gift to survivors of an accrescing share. There is no room for applying it where, as here, there is no survivorship clause, and where an independent gift is made to the issue the extent of which is to be measured by assuming that the parent is alive at the date of distribution. In the case of Laing v. Barclay a testator provided that two of his sons should enjoy an alimentary liferent of certain heritable subjects. Upon the death of the longer liver the subjects were to be sold and the proceeds divided equally among the testator's children, "and in the event of any of my children predeceasing the said term of division, leaving lawful issue, it is my desire that such issue shall represent and be entitled to the proportion which would have been payable to their parent." Upon the death of the longer liver of the liferenters all his brothers had predeceased him, only one of them leaving issue. It was held that the issue were entitled to the whole fund. In dealing with the cases where issue were held to be excluded from participating in accrescing shares the Lord Ordinary (Lord Kinloch) said (at p. 1148):

"In all these cases there was an accretion expressly declared in favour of survivors. And it was because the Court could not get over the plain meaning of this word ‘survivors’ that they held the accretion only available to the actually surviving children, and no part of it to belong to the children of a predeceaser."

In the Inner House Lord Cowan, who gave the opinion of the Court, said (at p. 1150):

"Questions regarding the right of accretion have occurred under settlements expressly providing that the share of the child or children predeceasing shall accresce to the survivors equally; and in such cases, whatever the form of expression may be, provided only it be clear that the surviving children are to have right of accretion, the issue of predeceasing children to whom their parent's share is either provided in express words or has been held to vest in them under the implied condition, can only take the parent's share, the immediate issue alone of the testator having right of accretion through any of the other children predeceasing without issue. This principle may be held as now quite fixed. The case of Clelland v. Gray, in this Court in 1839, with the prior cases there cited, and the recent judgment of the House of Lords in Young v. Robertson, &c. (Donaldson's Trustees) put an end to all dubiety on this point. But it is impossible to hold that principle applicable to cases where there is no clause of survivorship, and where, as in the present deed, there is an express declaration of the extent of interest in the succession to be taken by the issue of predeceasing children."

No good purpose would be served by a detailed examination of the different cases cited. The only support for the view that issue of a predeceasing child are excluded from participation in accrescing shares, in the absence of express institution of the survivors to such shares, is to be found in the opinion of Lord Trayner (at p. 460) in Cumming's Trustees. As, however, there was a destination in that case to survivors, and the opinion was therefore unnecessary to the decision in the case, I do not think it can be accepted as an authoritative exposition of the law. It is in conflict with the view of Lord Cowan and of other judges who have had to consider the circumstances under which the rule expressed in Young has to be applied. If, therefore, the case is, contrary to my view, to be treated as one where there is an initial division of residue into seven shares, of which three shares have lapsed owing to three of the beneficiaries having predeceased the death of the testator's widow without issue, I think the absence of a clause instituting the survivors to the accrescing shares relieves us from the necessity of applying a rule of construction which in this case would manifestly do violence to the declared intention of the testator as expressed in the language used by him. I am therefore for answering the questions put in such a way that the issue of the predeceasing children will share equally per stirpeswith the surviving children.

LORD SANDS .—Apart from certain details which do not concern this case the truster's scheme of settlement of the residue of his estate was simple. The residue was to be divided into two equal parts. The one part was to be divided among his children on their attaining twenty-five years of age, or marriage in case of daughters, with no vesting until payment. The other half was to be held for the widow in liferent and for behoof of the children as regards the capital, and fell to be divided upon the death of the widow "in equal shares among the children." Here, again, there was to be no vesting until payment. It is common ground in the case that, apart from any question of issue of children, the division "in equal shares among my children" meant in equal shares among the children then surviving. Seven children survived the testator, but only two survived the widow. Had none of the predeceasing children left issue, the division which fell to be made would plainly have been bipartite—one-half to each surviving child. It happened, however, that two of the predeceasing children had left issue. This was a contingency which the testator had foreseen and provided for. The clause in which he does so covers both branches of his residue. It runs as follows, after providing that there shall be no vesting until payment:—

"Should any of my said children die before or after my death or predecease my said wife leaving issue such issue shall be entitled to the share that would have fallen to their parent by survivance."

The words "before or after my death" are directed doubtless primarily to the first half of the residue not liferented by the widow, which fell to be divided among the children on the attainment of twenty-five years or marriage. It is not necessary to consider the possibility of their application as regards the other half in case the wife predeceased the testator, as that event did not happen. The words "predecease my wife" refer plainly to the second part of the estate and to the event which did happen. Accordingly, eliminating what is in the circumstances irrelevant, the direction reads:—

"Should any of my said children predecease my said wife leaving lawful issue such issue shall be entitled to the share that would have fallen to their parent by survivance."

As "survivance" is here correlated with "predecease," the survivance must be survivance of the widow.

As I have stated, seven children survived the testator. I shall call them A, B, C, D, E, F, and G. E, F, and G predeceased the widow without issue. They therefore take nothing. As I have already pointed out, if C and D likewise had predeceased the widow without issue, the division would have been bipartite. On the other hand, if A, B, C, and D had all survived the widow, the division would have been into four equal shares. One-fourth would have been the share that would have fallen to C and to D respectively by survivance of the widow. Accordingly, in my view, giving its ordinary meaning to language, the issue of C and D seem now to be entitled each to one-fourth share. Authority may forbid us to give its apparently plain prima facie meaning to language. I shall deal with that later, but in the meantime I am regarding the matter apart from the cases. Now, so regarding it, what is there that can be suggested against the view I have presented? I understand it to be as follows. Although the initial gift was to the children as a class, and although there was to be no vesting and a child predeceasing the widow without issue took nothing, still, to satisfy legal principle as to the institution of heirs, each child must be regarded as being as from the death of its father the institute in one-seventh of this part of the residue, seeing that there were seven children then alive. I confess I find this doctrine of a sevenfold division, because there happened to be seven children alive at the date of the testator's death, somewhat difficult of reconciliation with the doctrine enunciated by Lord M'Laren, that a gift to a class is a gift to the members of the class alive at the date of vesting—Wills and Succession, (3rd ed.) vol. i., p. 635, paragraphs 1150–1151. Lord Curriehill's dictum (at p. 1014) in Douglas,to the effect that a gift to a class is a gift to each individual member of the class, applied to a case where he held that there was vesting, and is therefore quite consistent with Lord M'Laren's view. But, accepting this theory of a sevenfold division, I do not see how it in any way affects the construction of the testator's direction. The testator does not direct that the issue of a predeceaser are, on the death of the widow, to take the share in which, according to legal theory, the parent was the institute, at his, the testator's, death. It is the share which the child would have taken if he had survived the widow. To warrant reference to the original share in which it is said the deceased child is to be regarded as the institute, one must add after "would have fallen to the parent by survivance" the words "provided all his brothers and sisters who survived me had also survived my widow." I confess I cannot find reason for such a bold reading in. No doubt, for reasons which have been explained, the law limits what can be taken under the conditio to the original share. But the conditio is an equitable innovation upon the rule that one cannot make a will for a testator even though he appears to have overlooked something, and, accordingly, it is restricted to the narrowest bounds of conjectural intention but for the overlook. Lord M'Laren clearly draws the distinction between the application of the conditio and the interpretation of a will in Neville v. Shepherd (at p. 357):—

"It is settled as matter of legal implication, as distinguished from construction, that a child under the conditio never takes more than what is described as the parent's original share, that is, the share the parent would take supposing everyone in the destination to survive the period of payment … When, on the contrary, the substitution is contained in a will or deed, the language used in the deed determines the extent of the right of the substituted legatee."

The reference to "the language used in the deed" in the foregoing quotation from Lord M'Laren leads directly to the second branch of my argument. I have shown that, in my opinion, the language used in the deed is unambiguous if given its prima facie interpretation, and this must prevail unless we are constrained by authority to give to the language an artificial meaning.

We were reminded in the course of the argument of certain recent weighty dicta to the effect that it is the duty of the Court to interpret a testament according to the opinion it forms of the intention of the testator as derived from the language which he employs, without any necessity to adopt the interpretation which may have been put by the Court in some former case upon somewhat similar language in some other person's will. But this comfortable doctrine is undoubtedly subject to some qualification. Where by a series of decisions a definite technical or artificial meaning has been attached to a certain expression, not in itself antecedently technical, we cannot ignore this meaning in the absence of any clear context which requires us to do so. But the circumstance that, in a former case, the Court took a different view of the intention underlying language somewhat similar to that with which we are concerned from what we are disposed to take, although it is a consideration entitled to the greatest respect and may influence our decision, is not conclusive, unless it can be shown that the interpretation has by decision been made a rule of law. Accordingly, if we are satisfied that, on a reasonable construction of a testator's settlement such as the present, the testator must be held to have meant what he literally directs—that the children of a deceaser are to take the share the parent would have taken of the whole estate had he been alive at the period of distribution—in order to disturb this conclusion it is necessary to show by a course of decisions that it is a rule of law that "the share the predeceaser would have taken had he survived until the period of distribution," or words to the same effect, means, not what the phrase literally imports, but the parent's original share only, viz., the share he would have taken had he and all the other possible beneficiaries survived.

In Laing v. Barclay the direction was that the issue of a predeceaser should "represent and be entitled to the proportion which would have been payable to [the] parent." It was held that the issue had a right to an equal share with the survivors, although there had been predeceasers without issue. Now doubtless the word "represent" occurs there. But, in order to interpret that word, it was necessary to explain it, and it is explained, as meaning, "take what the parent would have taken had he survived the term of division." In other words, in order to interpret "represent," one must use very much the same language as the testator employs in the present deed. The suggestion seems to be that, if the testator says "represent," he means "take the share the parent would have taken if he had been alive"; but, if he does not veil his meaning in the word "represent," but simply contents himself by saying "take the share the parent would have taken had he been alive" he must be held to have meant something else. But, even if the case be regarded as distinguishable, it is of importance for the clear enunciation by Lord Kinloch and Lord Cowan that the rule of Young v. Robertson, under which issue do not share in a bequest of lapsed shares to survivors, is limited in its application. After reviewing a number of cases Lord Kinloch says (at p. 1148):—

"In all these cases there was an accretion expressly declared in favour of survivors. And it was because the Court could not get over the plain meaning of this word ‘survivors’ that they held the accretion only available to the actually surviving children, and no part of it to belong to the children of a predeceaser."

So Lord Cowan says (at p. 1150):—

"It is impossible to hold [the principle of Young v. Robertson ] applicable to cases where there is no clause of survivorship, and where, as in the present deed, there is an express declaration of the extent of interest in the succession to be taken by the issue of predeceasing children."

I am aware that it is suggested that there is in the present case an implied clause of survivorship, because the distribution was to be among the children in life at a certain date. But so was it in Laing v. Barclay, in which the distinction is pointed out, and so will it be in every similar case. The real basis of the distinction between express and implied survivorship I take to be that, when a testator speaks of predeceasers and their shares and how these are to be dealt with, he is thinking of, and pointing to, a share as effeiring to each individual from the date of his own decease, whereas, when he directs that a fund shall be divided among the members of a class alive at a certain date, he is not directing his own attention or calling his trustees' attention to any particular shares as in existence before the date of division.

There is only one case to which we have been referred in which I have found what seems to be out of line with the explanation or qualification of Young v. Robertson enunciated in the case of Laing. That case is usually referred to as Cumming's Trustees. But that case is not out of line, for in that case there was a survivorship clause and a gift to survivors. A dictum of Lord Trayner's in that case (at p. 460) is, however, out of line, and that dictum was given effect to in the case of White's Trustees, where there was no express survivorship clause, which was heard and disposed of along with Cumming's Trustees. With all respect, however, I am unable to accept this isolated judgment as displacing the doctrine that the undoubted rule of law in regard to successors which rests on Young v. Robertson is limited as explained by Lord Kinloch and Lord Cowan.

Two other cases which may seem to create difficulty are M'Nish and Henderson. In both these cases, however, there was a survivorship clause and a gift to survivors. Apart from this I do not think that M'Nish carries one very far. Such expressions as "the share of the mother" following upon an initial direction for a division share and share alike have been held as suggestive of an original share. In the case of Henderson the bequest was of "the share which would have fallen to their parent had he or she survived" the testator. In the case of Beveridge's Trustees I tried to distinguish Henderson's case on the ground that, in that case, "share" had been given a definite meaning in the settlement. At I pointed out (at p. 589):

"The Lord President says that share must mean one-sixth; ‘there is no other share spoken of in this deed.’"

One might perhaps apply this reverse-wise to the present case. "Share" must mean an equal share in the division on the death of the widow, there is no other share spoken of in the deed with reference to this part of the estate.

The case of Neville v. Shepherd is worthy of examination, because it is subsequent to Henderson'scase, and Lord Adam and Lord M'Laren were parties to both judgments. There was a clause as follows:—

"In the event of any of my said grandchildren predeceasing me leaving a lawful child or lawful children, I hereby appoint and declare that such child or children shall succeed equally to the share of my said estates which would have fallen to his or her deceased parent under these presents, if he or she had been alive."

"Predeceasing me" was presumably a blunder for "predeceasing the term of distribution," but the Court held that it could not be so read. It seems clear, however, from the opinions of both Lord Adam and Lord M'Laren that, but for the blunder, this clause would have been a short cut to a result which they were able to reach on other grounds, viz., that the children took all that would have fallen to the parent if actually in life.

In the case of M'Culloch's Trustees Lord M'Laren, in referring to M'Nish's case, says (at p. 779):—

"I should not be disposed to assent to the proposition that there is any artificial rule of construction which obliges us to hold where a residue is disposed of among different members of a family that the children of one of the residuary legatees who may die leaving lawful issue are cut out from what their parent would have taken by accretion."

In the view I take the present case is not strictly one of accretion, but my general conclusion is similar to that of Lord M'Laren, viz., that there is no artificial rule of law which limits the right of issue, to whom is bequeathed the share which their parent would have taken "on survivance" or "had he been alive" at the period of distribution, to the amount of the share which he would have got had the estate fallen to be divided on the testator's death, or, in other words, a share determined by the number of contingent ultimate beneficiaries who happened to be alive at the testator's death.

When a testator directs that he desires the children of a deceased child to take what their parent would have taken if then alive, I think that presumably he means what he says—that the children shall stand in their parent's shoes, or, in other words, to use a phrase which was decisive in one case, viz., that they shall "represent" their parent. I concede, of course, that it might be shown by some implication that the testator, although he used this language, really meant to refer to the parent's original share. But the presumption appears to me to be decidedly to the contrary. Accordingly, when a testator directs that the issue of a predeceaser of a date of division shall take the share that his parent would take had he then survived, I shall infer that the testator meant what he said. I find no artificial rule of law which constrains me to hold that he meant what he did not say, viz., that the children shall take, not what the parent would have taken had he survived as matters then stood, but only the share which the parent would have taken on the hypothesis that there was no other predeceaser.

Stripped of what is irrelevant to the present question, the direction of the testator as regards the final disposal of the half of the estate liferented by his wife runs as follows:—

"On the death of my wife my trustees shall realise the whole remainder of my means and estate and shall divide the same into equal shares for behoof of my children—declaring that the said provision shall not vest until the term of payment, and that should any of my said children predecease my said wife leaving lawful issue such issue shall be entitled to the share that would have fallen to their parent by survivance."

It appears to me to be impossible to attach a different meaning to the word "share" in the second limb of the direction from that which it necessarily must bear in the first limb.

[His Lordship then proceeded to deal with other questions with which this report is not concerned].

LORD BLACKBURN .—The residue of the estate which the testator directs his trustees to hold for "my children" consisted, roughly speaking, of two halves, one of which was available for immediate distribution on the testator's death while the other would only become available for distribution on the death of his widow. By the fourth purpose of the deed he directs his trustees to divide the first half "equally among my children share and share alike," and by the fifth purpose he directs them, on the death of his wife, to realise the other half "and divide the same into equal shares for behoof of my children." With regard to both halves of the estate the date of the payment to the children was postponed until they had attained twenty-five years of age, or in the case of daughters until they should be married, and with regard to the second half the date of payment was further postponed to the date of the widow's death. By a subsequent passage in the deed the date of vesting of his children's shares was declared to be the date of payment.

In my opinion these directions fall to be construed as a bequest of his whole estate to his children as a class, and, as he was survived by seven children, the original bequest to each child in the event of his survivance of the respective dates of vesting amounted to no more than a one-seventh part of each half share of the estate. Any further sum which might accrue to the surviving children in respect of the predecease of one of their number without issue would not be a part of their original shares but an accretion thereto.

As I read the deed, there is nothing to justify the suggestion that the testator intended that there should be a fresh division of the second half equally among those of his children who might be alive at the death of the widow. Any such intention could have been expressed quite easily, and the effect would have been to increase the amount of the original shares of such children as might survive the widow. But, so far from expressing any such intention, the language used in the fifth purpose is almost identical with that used in the fourth purpose, and the conclusion appears to me to be unavoidable that the original share of the testator's estate to which each child would become entitled by survivance of the respective dates of payment was one-seventh of each half of the estate.

That the testator contemplated that the original division of each half of his estate should be among his whole children is suggested by the terms of the declaratory clause which follows the fifth purpose, and which gives rise to the question which we have to decide. He there directs "that should any of my said children die before or after my death or predecease my said wife leaving lawful issue such issue shall be entitled to the share that would have fallen to their parent by survivance," and this direction applies to the provision under the fifth purpose as well as to the provision under the fourth purpose of the deed, and refers generally to "my said children" as being the same beneficiaries under each purpose.

The question which is raised by this clause is as to what the testator meant by "the share that would have fallen to their parent by survivance." I agree that these words are open to construction, and that, if anything can be found in the deed to indicate the intention of the testator that the interest to be taken by the issue of a predeceasing child should include what the child might have succeeded to by accretion had he survived the date of payment, that intention must be given effect to. There is no difficulty in making such an intention perfectly clear, either by a reference to accrescing shares or by instituting the issue in the place of their deceased parent. Failing either of these methods, the testator's intention that the issue of a predeceasing child should take not only the original share of that child but also any sums which might have accresced to that child by the predecease of other children may be implied from language which he has used in the deed or from the terms of other provisions which he has made. But I thought it was well settled that a mere declaration that the issue of a deceased beneficiary should be entitled to take "the share that would have fallen to their parent by survivance" must be presumed to refer to the parent's original share only unless clear evidence can be found to indicate that the testator intended that the words should bear the wider meaning. The restriction of the issue's right to the parent's original share only is in effect to give them exactly what they would have got under the conditio si sine had it been applicable to their case.

It was argued that this presumption only prevails in cases where there is an express survivorship clause; but, in endeavouring to ascertain what may have been the true intention of a testator, it appears to me to be impossible to draw any distinction in principle between a gift to a class with a survivorship clause and a gift to the same class under the condition that no member of the class who may predecease the date of payment shall be entitled to a share of the gift.

In my opinion, therefore, unless we can find any evidence in this deed which supports the view that the testator intended that the share which the issue of a predeceasing child should take was to include all that might have accresced to the parent had he survived, we are bound to restrict the meaning to the original share which the child would have taken had he survived.

The only passage in the deed which appears to have any bearing on this question is the last declaratory clause at the end of the fifth purpose of the deed, which has reference to the one half of the share falling to each of the daughters on the widow's death. The testator, having restricted the daughters' interest in this half of their shares to a liferent and having given the fee to their issue, directs that, should his daughters die unmarried or without issue, they should have power to dispose of the share by will, and he then goes on to provide that, in the event of a daughter dying without issue and without having disposed of this portion of her share by will, this half share should "fall and accresce to my children or their issue per stirpesamong them." There can be no doubt as to the meaning of this clause, for it amounts to the conditional institution of the issue of a deceased child of the testator along with such of his children as might be surviving. This clause deals only with a comparatively small portion of the testator's estate, and relates to an event which might not occur till long after the date when the bulk of it would have vested. But it indicates that, where stirpital succession to a deceased parent's share was intended, the language in which the intention could be given effect to was familiar. One may speculate as to why the testator should have provided expressly that the succession to this portion of his estate should be stirpital, and should have failed to do so for the bulk of his estate, but I am very far from feeling that the fact that he has done so affords any evidence of his intention that the expression "parent's share" as used in the declaratory clause should be construed as an institution of the issue in their parent's place. Indeed, I think that, if any inference is to be drawn from the direction as to this one-half of the daughters' shares, it is adverse to any such intention, suggesting as it does that the difference between instituting the children in their parents' places and giving them a right to their "parent's share" was fully realised. In my opinion, accordingly, there is no sufficient evidence in this case to upset the ordinary presumption that the issue of predeceasing children take right only to their parent's original share and no right to shares which might have accresced to their parent had the parent survived.

I therefore concur in the opinion of the Lord President as to how the case should be disposed of.

LORD MORISON .—This special case raises questions regarding the disposal of one-half of the residue of the estate of the late Alexander Ronald Miller. It is not clear how this residue arose, but this seems to me to be immaterial, because counsel for the parties and also for the trustees stated that it now consisted of that moiety of the deceased's estate of which his widow enjoyed the liferent until her death in April 1931. The trustees, in the execution of the fifth purpose of the settlement, had realised the balance of the estate after making the interim payments to each of the testator's seven sons and daughters. At the date of Mrs Miller's death they also held certain shares in engineering companies. The residue in question consists of trust investments—£14,804, 6s. 6d.—and the value of the non-trustee investments appears to be in the neighbourhood of £13,000. Under the fourth and fifth purposes of the settlement the trustees were directed by the testator to hold this portion of the residue for behoof of his children until the death of his widow—the liferentrix. The question then arises, Among whom has the testator directed that this residue should be divided?

The answer to the question depends upon the construction of the fifth purpose of the settlement, which alone determines the division. The operative part of the clause is as follows:—

"On the death of my said wife my trustees … shall divide the same into equal shares for behoof of my children … declaring that the said provisions in favour of my said children shall not vest until the respective terms of payment of the same and that should any of my said children die before or after my death or predecease my said wife leaving lawful issue such issue shall be entitled to the share that would have fallen to their parent by survivance."

Three things are, I think, clear from this destination. (First) there cannot be vesting in any share of this portion of the residue prior to the widow's death. (Second) the keynote of the division is equality among the children who acquire a vested right and the issue of any of them who die either before or after the testator or before his wife. And (Third) the number of the shares in the division of the residue is to be determined, not by the number of his children at the date of his death, but by the number of his surviving children plus the number of predeceasing children who have left issue. In the present case three of the testator's sons predeceased the term of payment without issue. Necessarily, therefore, they took no share in this residue. A son and daughter—the second and third parties to the case—and issue of a predeceasing son and daughter—the sixth and fifth parties to the case—all survived the widow. In my opinion the residue was, therefore, necessarily divided into four parts—an equal portion going to each of the second, third, fifth, and sixth parties. In no other mode of division could effect be given to the imperative words in the destination, that the issue of predeceasing children "shall be entitled to the share which would have fallen to their parent by survivance." If Mrs Miller or Brown and Mr John K. Miller had survived their mother, they would have succeeded to an equal fourth of the residue, and this share, under the express words of the destination, passes to their issue. I am quite unable to read the clause as meaning that the issue of the two predeceasing children are only to take one seventh shares of the residue. That would have been the result if five of the testator's family had survived their mother. Nor am I able to read the deed as directing that the contingent interests of the testator's three predeceasing sons passed to his children who survive. I think this view contravenes the equal division among his children and the issue of predeceasing children which the testator directs his trustees to make on the death of his wife.

It was scarcely argued that the language of the settlement itself was susceptible of any other construction. The contention of the second and third parties was that a series of decisions required us to construe the deed as if each child took a one-seventh prospective share in the residue, that a clause of survivorship among the children was implied, and that the one-seventh shares of the three predeceasing children became part of the shares of the two surviving children. The argument brought to my mind Lord Halsbury's words of protest in his speech in the case of Bowman (at p. 71):

"I have always protested, and still protest, against reading one man's will in the light of another man's will when they are in construction, in design, and in language often extremely different. I do not deny that what are called ‘canons of construction’ are sometimes very useful for the purpose of making known the meaning which the law would attach to particular phrases and words, but I am very much disposed to say that canons of construction are what has been popularly said of fire—very good servants but very bad masters."

In my opinion the canons of construction upon which the series of cases (of which Henderson's Trustees is a typical example) was decided have no application to the instrument here. In all the cases quoted in support of the argument there was both a survivorship clause and a definite division of the residue into shares. It is settled that, where a testator has left a specific and definite share of his estate to each of his children with a clause of survivorship among them, then, although the child of a predeceaser is entitled to the share of the parent so defined, he is not entitled to participate in the shares of other children who had predeceased the term of vesting without issue. That is because lapsed shares are destined to the survivors under the survivorship clause. And, of course, the expression of the conditio in a deed of this character adds nothing to the bequest if it is one in which the conditio would be implied by law. In this deed, however, there is no division into one-seventh shares, and there is no survivorship clause. The only division of this portion of the residue directed in the instrument makes it clear that such clauses are quite inappropriate. The testator's intention, as revealed by the deed, is to give to the issue of a deceased child, not a defined share of residue, but an equal share in the division of a residue the amount of which may have been affected by the demise of certain possible legatees. In my view, the testator's three predeceasing sons held no specific share of his estate. They failed to take any share. They had a mere expectancy to succeed to some indefinite share in an equal division of the residue at the death of their mother. I think that the residue liferented by the widow must now be divided equally among the two surviving children and the issue left by the two predeceasers—that is to say, into four equal parts. I think it is unnecessary for me to discuss the authorities. I agree respectfully with the views on them expressed by Lord Sands.

[His Lordship then dealt with other questions with which this report is not concerned.]

LORD MURRAY .—This special case raises a question as to the sound construction of the fifth purpose of the trust-disposition and settlement of the late Mr Miller. He died in 1901 survived by a widow and seven children, both sons and daughters.

By his settlement he liferented his widow in part of his estate, and thereafter, by the fourth purpose, directed his trustees to hold the whole residue of his estate for behoof of his children, and to divide one-half thereof equally among them, the shares being payable on attaining majority (twenty-five years) or marriage, in the case of sons and daughters respectively. By a subsequent declaration the vesting of these shares was expressly postponed till the term of payment. All the children attained majority or marriage, and each, on satisfying this condition, received payment of one-seventh share of half the residue. By the fifth purpose, which is now under construction, the truster gave directions as to the remaining half of his residue. Shortly stated, he directed his trustees, on the death of his wife, to divide the remainder of his estate into equal shares for behoof of his children, payable, as before, on majority or marriage. Again, vesting in the children was postponed to the date of payment. The declaration already referred to provided that, in the event of any child predeceasing the widow leaving issue, "such issue shall be entitled to the share that would have fallen to their parent by survivance." There is no expressed conditional institution of surviving children to the share of a child predeceasing without issue. If, however, the original gift to children be a "class-gift," the law may, no doubt, presume or imply a survivorship clause. What happened was that two children only (the second and third parties) survived the widow; two children predeceased the widow leaving issue (their respective stirpes being the fifth and sixth parties); three children predeceased the widow without issue, and these three shares thus lapsed. The question is whether, in terms of the above clause, the issue of a predeceasing child take the original share of a parent only, or the share of a parent original and accrescing in consequence of a share or shares lapsing.

The answer, of course, depends on the ascertainment of the intention of the testator as this is evidenced by the terms of the deed. Taking the terms of the deed as they stand, and assuming that one is untrammelled by previous decision or authority, I think the intention of the truster is quite clear. The total fund is to suffer equal division among the beneficiaries, whoever they may prove to be. The beneficiaries who ultimately take are to be either (1) the testator's own children who satisfy the condition precedent of attaining majority and are alive at the date of distribution, which is the date of vesting, or (2) the issue, if any, of a child who has predeceased the date of distribution, such issue taking the share of the total fund which would have fallen to their parent had such parent been alive at the date of distribution. Had the parent been alive at the date of distribution he would have taken an equal share with his then living brothers and sisters of the total fund, including therein his just proportion of any lapsed share. In short, such issue just step into the shoes of—or "represent" (to use the term which appears in the destination in the case of Laing v. Barclay, already referred to by your Lordships)—a parent, who is deemed to be still alive and therefore to be one of the beneficiaries. This result appears to me to be in strict accord with the intention of the truster, whose primary direction was that there should be equal division among the beneficiaries, who are just those who ultimately participate in the gift. The contention of the second and third parties results in unequal division of the fund, and on this ground alone should, I think, be rejected. If this be the true construction of the clause in question, it is clear that the contention of the fifth and sixth parties is well founded, and that a one-fourth share of the total fund falls to be paid over to each of the two surviving children of the truster, the remaining two-fourths being divisible equally between the two stirpes of predeceasing children.

It was argued for the second and third parties that we are constrained to apply in this case what has been termed the "rule in Young v. Robertson " (second appeal).This would limit the rights of issue of predeceasers to a parent's original share only. In my opinion this case is not ruled to this effect by Young v. Robertson; indeed, if that case be examined, the present case is precisely its converse. In Young v. Robertson the position was that there was an express conditional institution of surviving children to the lapsed share of any child predeceasing the period of distribution; the conditio si sine which the grandchildren invoked was merely implied. Here there is an express conditional institution of issue to the predeceasing parent; the rights of surviving children in or to lapsed shares stand only on the presumption or implication which arises from the original gift being a class-gift. The legal position as between children and grandchildren is accordingly reversed.

The principle upon which Young v. Robertson was decided is quite clear, and has been frequently explained in the cases which have since followed. It is simply that an express gift to A must prevail over any presumed or implied gift to B—Lord Westbury, L.C., arguendo, at p. 338, or, as he put it in the course of his speech, the surviving children in the case of such express gift-over of a lapsed share take to the exclusion of the grandchildren "by force of the words contained in the settlement."—See also M'Laren's Wills and Succession, (3rd ed.) vol. i., p. 641, paragraph 1159, and M'Culloch's Trustees, at p. 779. If the above principle be here applied, the express conditional institution of issue should prevail over any implication from class-gift, and admit the issue to share in accrescing shares. The principle of Young v. Robertson in this view is really an authority in favour of the fifth and sixth parties.

I am not aware of any case in which the rule of Young v. Robertson has been applied in the absence of an express gift-over of lapsed shares to surviving children alone. All the cases cited to us exemplified this. In the case of Cumming's Trustees there was such an express gift-over, but Lord Trayner expressed the opinion obiter that the rule of Young v. Robertson extended to cases in which the gift-over was merely implied. To this opinion the Second Division of this Court gave effect in the case of White's Trustees, which was argued along with Cumming's Trustees and decided the same day. In neither case, however, was the point of distinction with which I am dealing argued. With all respect to that eminent judge, this opinion appears to me to disregard the whole foundation of the judgment in Young v. Robertson . I may add that the reluctance with which a number of our judges have followed and applied the rule in Young v. Robertson is itself, in any event, a good reason for not extending it to cases which are not fully and expressly covered by the rule.

The case of Laing v. Barclay is a significant illustration to the contrary. There, as here, there was no express gift-over to children, and an express conditional institution of issue. The latter were held entitled to participate in accrescing shares. No doubt the precise terms of conditional institution were, in that case, somewhat different from the present case, but in substance there is, in my opinion, no material distinction.

The whole foundation of the argument for the second and third parties rests upon the assumption that there is no difference in result between an express and an implied "clause of survivorship." I do not accept any such assumption. It is true that, in general, what is known as a "class-gift" does imply "survivorship" as between the members of the class called. Such a gift is read as in effect equivalent to an express joint-gift (which is the typical case); but it rests, however, solely upon the presumption that a testator, purporting to dispose of his whole estate, prefers the member or members of the class instituted to his heirs-at-law as in intestacy. A class-gift may thus have the same result as an express clause of survivorship; to hold that it must is, in my opinion, a non sequitur. For this, as Lord M'Laren pointed out in Neville's case, is to confuse matters of implication and matters of construction—see also Farquharson v. Kelly . An express gift-over to survivors, not being ambiguous, is not open to construction; an implied gift, resting upon a legal presumption, is always so open. And it must always yield to that which the Court hold to be the expressed intention of the testator. The difference between these two is therefore more than "a question of form." It may, on a question of construction, just constitute the essential difference. And it is just this difference which lies at the root of decisions such as Laing v. Barclay and M'Culloch's Trustees.

I am accordingly of opinion that there is nothing in the rule of Young v. Robertson which prevents this Court from giving effect to the natural construction of the terms of the deed.

The remit to the Court of seven judges is limited to the subject-matter of the first three questions in the case. In my opinion the contentions of the fifth and sixth parties should be sustained.

If the construction I adopt be sound, this result will follow whether in virtue of the original gift the primary division of the fund be into seven shares, or whether the division into shares be regarded as indeterminate and suspended until the conditions of the gift as to vesting and so forth prove to be purified. In this latter view the only effective division would be into four shares, one to each of the ultimate "takers," the two children and the two stirpes. For myself I regard the first view as more in accord with the principles of our law in the construction of destinations with conditional institutions. The second view was rejected by Lord Gifford in M'Nish's case, although it would appear to be the accepted rule of construction under similar destinations in England—see Eyre v. Marsden, (1838) 2 Keen, 564, per Lord Langdale, M.R., at p. 572, and (1839) 4 Myl. & Cr. 231, per Lord Cottenham, L.C., at p. 239. But, however this may be, the result of the effective division is the same in either case, viz., into four equal shares.