Scottish Court of Session Decisions

At advising on 13th July 1956,—

LORD PRESIDENT (Clyde).—This case raises an interesting question as to whether an adopted child is entitled to participate in the residue of a testator's estate under a settlement whereby he directed his trustees to divide his residue on the death of his wife, or on his own death should she predecease him, equally among all his then surviving children.

The testator was thrice married. He had two children by his first wife. He married his second wife in September 1028. She had then an illegitimate child, aged 3 years. The testator was not the father of this child. The testator had two children by his second wife. On 9th May 1941 the illegitimate child was formally adopted by the testator. Two years later, on 14th May 1943, the testator signed his testamentary settlement. His second wife died in 1947 and he married again. He died on 9th February 1953, survived by his third wife, by whom he had no issue. In 1944 the adopted child married. These facts are not disputed, and such facts can of course always be taken into account in considering the meaning and effect of a testamentary deed—Dickson on Evidence, vol. ii, par. 1067. For without them it would not be possible to tell who benefited under the deed.

But in the present case the pursuer (the adopted daughter) seeks to go further and to prove a whole series of averments relating, inter alia, to the equality of treatment by the testator of the pursuer and the testator's legitimate children—representations made by the testator that he had provided in his will for the pursuer, and would treat her in his will equally with his own legitimate children, and averments of the ignorance of the solicitor who prepared the will that the pursuer was other than a legitimate child of the testator. This evidence is all designed to show that the testator intended to treat her equally with his legitimate children, and that the word "children" in the residue clause therefore included the adopted child.

The Lord Ordinary with some hesitation has allowed a proof before answer of these averments, and the first question in the case is whether such a course is a proper one so as to entitle the Court to infer what the testator meant by the word "children" in the residue clause.

So far as adopted children are concerned, the matter depends primarily upon the terms of the Adoption Act, 1950. This Act consolidated the previous Acts, and in section 15 re-enacted the earlier provision in the same terms which was contained in the Adoption of Children (Scotland) Act, 1930. Under section 15 the English provisions on the matter (which are different) are excluded from applying in Scotland, and by virtue of subsection (2) it is provided that "an adoption order shall not deprive the adopted person of any legal rights competent to him in the estate of his parents or of any right to or interest in property to which, but for the order, he would have been entitled under any intestacy or disposition, whether occurring or made before or after the making of the adoption order, or confer on him any right to or interest in property as a child of the adopter; and the expressions ‘child,’ ‘children’ and ‘issue,’ where used in relation to any person in any disposition, shall not, unless the contrary intention appears, include a person or persons adopted by that person, or the issue of a person so adopted." By subsection (3) the expression "disposition" means a deed, instrument or writing whether inter vivos or mortis causa whereby property is conveyed or under which a succession arises.

The effect of these provisions is to define the succession rights of adopted children who by virtue of an adoption order authorised by the statute become to some extent children of the adopter as well as being children of their natural parent. The result of the Act is not to cut these children completely off from their natural parents so far as succession goes, but to leave them with all the rights of succession which they would have had against these parents if the adoption procedure had never been invoked. As a corollary to this, Parliament has further provided that, so far as their adopter is concerned, they are not to be equated in all respects automatically to the position of a legitimate child of the adopter, although, if he chooses, the adopter may so treat the adopted child in a disposition either inter vivos or mortis causa. Hence the adopted child will not be able to claim legitim in the adopter's intestate estate and will not be included in a bequest to a child in the adopter's settlement or marriage-contract "unless the contrary intention appears."

It was contended for the pursuer that these words "unless the contrary intention appears" were inserted by Parliament to enable an investigation into intention to be made, and the investigation was not meant to be in any way limited. I decline to accept this construction of the subsection. It would involve an investigation quite contrary to existing legal principles of interpretation and would convert the function of the Court from the ascertainment of what the testator has done in his settlement into a speculation as to what he must have meant to do. Under the section inter vivosdispositions are in the same position as mortis causa deeds, and, if the pursuer's contentions were sound, it would entitle a Court to hear evidence from the granter of what he meant to achieve in his inter vivos conveyance of property. This necessary consequence of the argument appears to me to demonstrate its unsoundness.

But if the words "unless the contrary intention appears" will not enable a claimant to prove what the testator intended, it inevitably follows that proof of surrounding circumstances designed to lead to an inference as to the testator's intention is equally excluded, for it would be absurd to permit indirect evidence of intention but to exclude direct evidence of it. The proof which the Lord Ordinary allowed is therefore not appropriate.

The question then arises as to what content is to be given to the words "unless the contrary intention appears." In my opinion, they relate to the recognised and competent method whereby a Court of law can ascertain what is the granter's intention—namely, a consideration of the language of the deed itself, read in the light of the survival or otherwise of the possible claimants thereunder. It is from a consideration of these facts and these alone that the Court must determine whether the contrary intention appears. If it does not, Parliament has provided that an adopted child shall not be included among the children benefited by the disposition. Had Parliament meant to introduce so novel and startling a method of ascertaining the meaning of the word "children" as that contended for by the pursuer, I should certainly have expected clear words to that effect. In my opinion, there is in the language used in the section nothing to warrant the proposition for which she contends.

It follows from what I have said that if the adopted child is specifically named in the clause in question, or if the language of the deed as a whole shows that the adopted child is meant to be included along with the legitimate children in a benefit conferred, the primary meaning of child is displaced and the adopted child would benefit along with the legitimate child. Moreover, if it transpired that in fact the testator never had any legitimate children, but had adopted a child who survived him, the Courts might well interpret the use of the word child in the testator's settlement as referable to the adopted child, since otherwise the provision would fail. As M'Laren, Wills and Succession, vol. i, par. 702, says. in reference to this canon of construction:

"The rule…is confined in its application to cases where the evidence, which is always admissible for the purpose of identification, fails to disclose any person or thing (as the case may be) answering to the words of the will in their strict and primary acceptation. In such a case the words as construed are said to be insensible with reference to extrinsic circumstances, there being no external object to which they can be applied. But as a testator is always supposed to contemplate real persons and things in the expression of his testamentary wishes, it is presumed, in the case under consideration, that the words are used in some popular or secondary sense of which they may be susceptible."

In par. 703 the learned author goes on to point out that "the converse of this rule of law, namely, that words which, in their primary meaning, represent real persons or things, cannot have a different meaning impressed upon them by extrinsic circumstances is a self evident proposition, and requires no authority to support it."

This rule has been applied in regard to illegitimate children in Scott's Trustees v. Smart . In that case, Lord Carmont said (at p. 17):

"A bequest by a testator, who is a bachelor in status, to his ‘children’ may infer a gift to his illegitimate children surviving him, if he dies a bachelor, because the bequest would otherwise be inept; but, if the testator had married after the date of the execution of the will, the same interpretation might not be put upon the bequest if there were issue in posse,and would not be put on the bequest if legitimate children were in esse."

In the present case, however, all this is of no avail to the pursuer. She is not named in the deed, there are no words in any other part of the settlement from which any inference can be drawn that the testator was using the word "children" in his residue clause in a wider sense than the law would otherwise infer, and the existence of legitimate children qualified to take gives a content to the word "children" which excludes the rule of law referred to in the passage quoted from M'Laren on Wills and Succession. In my opinion, it follows that the contrary intention does not therefore appear and the adopted child is not one of the testator's residuary legatees.

It was contended for the pursuer that the word "children" was ambiguous, and that by the law of Scotland in cases of ambiguity evidence of intention was competent. But this proposition is too widely stated. In Dickson on Evidence, vol. ii, par. 1080, it is said:

"…it has been laid down repeatedly, and with authority, that evidence of the granter's intention is inadmissible for the purpose of clearing up a patent ambiguity."

The leading of evidence of intention is only permissible where the ambiguity is latent. As the same learned author says at par. 1078, latent ambiguities are "those which arise where the words of the deed in themselves are neither ambiguous nor obscure, but are shown to be equally applicable to more than one person or thing." It is quite clear that there is no latent ambiguity about the word "children," for the Act of 1950 itself prevents that word being equally applicable to adopted and legitimate children. Evidence of intention therefore could not be admitted on the ground of latent ambiguity. This conclusion appears to me to confirm the view that in a question under section 15 of the Adoption Act, 1950, no evidence such as the pursuer seeks to lead here would be competent. For the ground in law upon which it might have been maintained that evidence of intention was competent (latent ambiguity) cannot arise in a question under that section as to the meaning of the word "children."

It may be that, in a case of alleged misnomer or misdescription of the person intended to be benefited, an inquiry may be allowed, not into intention, but into such facts and circumstances as may help to give a key to the meaning of the words which the testator used—see Lord Dunedin in Nasmyth's Trustees v. National Society for Prevention of Cruelty to Children, at p. 83. But Lord Dunedin is alone in formulating the proposition in this way; a similar approach was not adopted by the rest of the Court, and in any event, in my opinion, the principle does not apply in a case such as the present. Here the issue is not as to whether there has been a misdescription in the testator's identification of a particular beneficiary, but as to what is the content of the generic term "children" in the deed. Upon that word the 1950 statute has impressed a primary meaning, and no question of misnomer or misdescription can arise. On this view the absence of anything in the settlement in the present case to displace the primary meaning of "children," and to extend it beyond the legitimate children, excludes the adopted child.

In Purdie's Trustees v. Doolan Lord Murray heard evidence as regards the testator's state of mind and knowledge, as a result of which he held that a bequest to T whom failing to his children equally among them included T's illegitimate children (of whose existence the testator was aware) as well as T's legitimate children (of whose existence the testator was unaware). On the evidence Lord Murray held that the facts regarding the testator's knowledge were sufficient to displace the prima facie meaning of the word "children." In my opinion, such evidence could not be admitted in a question regarding adopted children where the statute has provided the test. Indeed I have grave doubts whether the individual members of a class who are to take under a bequest to that class can ever be determined by ascertaining whether or not the testator knew of their individual existence. The decision in Purdie's Trustees affords, in my view, an example of the dangers of opening the door to a wide range of surrounding circumstances, including what the testator knew and did not know, and the inevitable speculation which such an inquiry starts as to what the testator must have meant to do. It stands quite alone in the general trend of Scottish authority.

Two English decisions were referred to—In re Fletcher and In re Gilpin. The precise point raised in the present reclaiming motion was not, however, really in issue in these cases and I find it unnecessary to consider them further.

On the whole matter, therefore, I have come to the conclusion that the additional facts which the pursuer desires an opportunity to prove could not competently be proved. Moreover, it follows that, as there are legitimate children to take, and nothing in the language of the deed to indicate that the testator was using "children" in other than its primary meaning of legitimate children, the pursuer is not entitled to be treated as one of the children sharing in the residue of the estate. The interlocutor of the Lord Ordinary should therefore be recalled, the defenders' first plea in law sustained and the action dismissed.

LORD CARMONT .—The question raised in this reclaiming motion is whether an adopted child is entitled to participate in a bequest to children.

The primary meaning of the word "child" excludes such a claim, and the statutory provision of section 15 of the Adoption Act, 1950, also excludes the claim unless it "appears" from the deed in question that the testator's intention was to include the adopted child along with the other children. It is true that the section qualifies what it enacts by saying unless it "appears" without specifying where or how it has to be made to appear. The respondent, however, offers to prove the averments she sets out, which, she says, will make it appear that an intention contrary to the primary meaning stated in the statutory definition of children should be preferred.

The section of the statute plainly indicates that adopted children can in certain circumstances be embraced by the term "children" but the burden is upon the claimant to show that the word as used by the testator was intended to mean something other than the primary meaning of that word. The statute contemplates some state of matters in which it can somehow be made to appear that the primary meaning was displaced by the testator, and the question arises whether such a result can be made to appear except by reference to something outside the document itself or only by pointing to other words used by the testator in his testamentary writings which lead to the conclusion that the intention of the testator was to give a wider meaning to the word "children" when used by him than the primary meaning. But unless that intention is to be found in the testamentary writings the Court would be throwing open the door to parole proof of testamentary intention. This is contrary to the fundamental principles of legal construction as applied to testamentary writings. The difference between saying that the Court does not allow parole evidence to show what a testator intended, but that it does allow, in suitable cases, investigation by parole into what the words used by a testator were intended to mean, although fine, is a very real one. In the latter case a testamentary writing is being construed to find the testator's true testamentary intention from the written words he has used, whereas in the former case an individual's intention as expressed in terms of parole evidence would be set up as testamentary. This latter position would simply mean that an individual's testamentary intention might be expressed without a testament and that would be contrary to all our established principles.

There is no doubt that certain investigations are usually allowable, as to the state of the testator's family and the surrounding circumstances of the making of the will. This is often absolutely necessary for the proper understanding of the will, and it has been for generations accepted that proper exposition of a testament requires that the interpreting Court should be put, as it were, into the testator's chair and given an outlook based on his knowledge—his actual proven knowledge with its actual limitations so far as known. But in so doing the Courts have always been astute to guard against attempts to introduce parole evidence as to what the testator intended under guise of evidence of the surrounding circumstances. The most subtle way, I think, to pass from the proper investigation of a testator's meaning into the realm of his intention is by presenting the case as one of latent ambiguity so that advantage can be taken of the special rules which take investigation into the realm of proof of the testator's intention. To get to latent ambiguity it is suggested that the word "children" as used by the testator in this case means either on the one hand a group of the lawful children born to him of his marriage or, on the other hand a group composed not only of such lawful children but also of any adopted child; between these groups, it is said, there is a latent ambiguity which the law allows contestants to resolve not merely by evidence of surrounding circumstances but by evidence of the testator's intention expressed in writings not testamentary or even merely expressed verbally.

In applying what I may perhaps be permitted to call the desperate remedies appropriate to an ambiguity which appears in a testamentary writing, we must be sure that there is truly an ambiguity. There is, in my opinion, no ambiguity where the name and description of a beneficiary is adequate and accurate as applied to one person or body and not adequate and accurate when applied to another. To allow the latter to displace the former would be to contradict the terms of the deed by displacing the named by the misnamed. A gift to X's son James is never ambiguous if X has a lawful son James merely because X has a bastard son James whom he writes to and calls his "son"see In re Fish. Cases of inadequate description in name, address and designation, if doubt remains as to which of two persons or bodies both accurately answering the description is meant, present true ambiguity which the law will resolve by evidence of intention. Indeed it may be said that slightly inaccurate naming and/or description of one beneficiary may present ambiguity where the description also applies to another person but with similarly slight inaccuracy. In such a case too the Court might well allow parole evidence of the testator's intention.

There are, of course, cases in which, in order to give some effect to a testamentary writing, a gift to a son of X will pass to an illegitimate son of X if X has no legitimate son or could have none to the knowledge of the testator, but such cases do not normally present any aspect of ambiguous gift. When the situation is one created by the testator's misdescription of an object of his bounty, the Court sometimes allows this to be corrected with a view to avoiding intestacy which would otherwise result. I draw attention, however, to the fact that it is not enough that there are in fact on a testator's death no legitimate nephews to induce the Court to admit as beneficiary of a bequest to "the children of my sister" a sister's illegitimate son, even if the result is an intestacy. The reason is that, as at the date of the making of the will, the testator could only be held in law to have contemplated by using the language he did that the object of his bounty was his sister's legitimate child, whether existing or yet to be born to her. In the case of Dorin v. Dorin Lord Hatherley said (at p. 575) with reference to circumstances figured by him in which it was not possible that a testator could be contemplating legitimate children of X as objects of his bounty:

"The only mode in which the word ‘children’ can be made to bear a different sense from that which is its first legal and natural sense is this, that if you look to the outward circumstances as well as to the expressions contained in the will, and find that the outward circumstances of the case, combined with the expressions contained in the will, fail to give any adequate or intelligible sense to the will, then you have at once to arrive at the conclusion that the word ‘children’ has been used in some other or different sense; just as if the will had spoken of ‘the children of my late brother’ or ‘the children of my late sister’; and if neither of these persons had been married, and they were then dead, and there was no possibility that they should ever marry, you would be driven necessarily to the conclusion that the only children that could be meant must be those whom the law would not otherwise allow to fall into that class, namely, illegitimate children."

I can see no reason for declining to apply with equal definiteness the principles applied to legitimate and illegitimate claimants' right to take, or to share in, bequests given in terms to "children" to claims put forward by adopted children. The legal can never be in competition with what is not accepted as legal. It in no way conflicts with such a view that, when the legal must be excluded, the defectively legal can qualify as a beneficiary. And, just as illegitimate children can under certain circumstances take under a gift to "children," that is not because the word "children" is ambiguous, but because, failing any persons to which the word can ever properly be applied, the law prefers claimants who can put forward what I may call a base or secondary title to its reaching the result that the testator's will has failed of any effect.

I do not find the approach of Roxburgh, J., in the case of Fletcher entirely satisfactory. He puts it, in the case of an adopted child making a claim under a gift to children, that he is not prepared to treat an adopted child worse than the law treats an illegitimate child. I do not see that in the matter of the interpretation of the word "children" it is necessary to do anything other than to apply the same principle of construction as between the primary and any reasonable secondary meaning which can be put forward. It should not be left out of account that a testator can specifically benefit his adopted child even by merely letting it appear that he means to include the adopted with the normally lawful. Moreover, the Adoption Acts see to it that those adopted children who have possible rights from their real parents do not lose such rights on adoption. Where this reserved right is of some substance, there is good ground for the law to decline to let such adopted children also participate in an adopting testator's bequest to children.

To conclude, in the present case there is no ambiguity which entitles the Court to clear up the testator's intention by allowing parole evidence.

In the absence of such ambiguity the Court cannot allow a group of lawful children, who satisfy the primary meaning of the word the testator has used when making the gift, to have their rights impinged on by allowing any proof of averments tending to show that the word "children" was used in any other than the primary sense. To allow such evidence would contradict the language used by the testator and not construe it.

Finally, there is no need for any proof of surrounding circumstances. Everything is admitted to the extent the law allows, and anything additional among the respondent's averments is aimed at something which the law rejects in circumstances about which there is and can be no dispute. I would recall the allowance of proof and dismiss the action.

LORD RUSSELL .—For the reasons referred to and expounded by your Lordships, to which I cannot usefully add, I agree that the reclaiming motion succeeds, and that the action should be dismissed.

LORD SORN .—We were given to understand that it had never been decided at common law whether the word "children," occurring in a will, included adopted children. It has long been settled that the word, in its primary meaning, does not include illegitimate children, and my impression is that the word has so long been accepted as designating legitimate children, i.e., children of the body, that it would be held not to include children by adoption. That question, however, does not need to be pursued, or decided, because section 15 (2) of the Adoption Act, 1950,gives the word a statutory meaning by providing that the expression "children," occurring in any disposition, shall not include an adopted child "unless the contrary intention appears."

Two conflicting arguments were submitted as to the meaning of this provision in section 15 (2), both of which seem to me to be extreme and unacceptable. The defenders argued that what it meant was that the statutory meaning of the word is to prevail unless a contrary intention is found expressed upon the face of the deed. If that view were correct it would provide a short cut to the decision of this case, because it is a matter of admission that there is nothing on the face of the deed itself which shows a contrary intention. I cannot, however, read the subsection as having that meaning. The argument treats the subsection as if, having referred to the use of the word "children" in any disposition, it continued "unless the contrary intention appears therein." The word "therein" is notably absent, and it seems to me much more natural to read the subsection as meaning that the statutory meaning is to prevail unless the contrary intention appears from the disposition as construed according to the ordinary rules of construction. The other argument was at the opposite end of the pole. The pursuer maintained that, since the provision did not expressly define or circumscribe the way in which the contrary intention might be made to appear, it was open to prove the existence of a contrary intention in any way at all, regardless of the ordinary rules of construction and the limitation imposed by those rules upon the admission of extrinsic evidence. It seems to me enough to say that, if the Act had meant to prescribe a special and unprecedented way of ascertaining intention in this particular instance, it would have prescribed it expressly; and that it would be going much too far to read the subsection as impliedly sweeping away the ordinary rules of construction applicable to the ascertainment of intention. Both arguments, therefore, seem to me to be unsound, and, if I am right in so thinking, the true question raised by the case is whether this will, construed according to the ordinary rules, evinces a "contrary intention."

In construing a will it is always permissible to do so in the light of the circumstances surrounding its execution, and I use the expression "surrounding circumstances" to indicate that limited amount of evidence to which it is always permissible for the Court to have regard, so as to put itself in the position of the testator. I have looked through the record in order to see what facts, properly falling within the category of surrounding circumstances, are averred—and what do I find? It is averred that, when he made the will, the testator had children of his own and that he also had an adopted child in the person of the pursuer. It is said that the pursuer was treated as one of the family and that the testator regarded her as if she was one of his own children. Taking these averments pro veritate, as one must do in testing relevancy, what is the result? It seems to me impossible to say that, when the will is read in the light of these circumstances, it appears that the testator used the word "children" in the secondary and broader sense of including an adopted child. Speculation does not come into a question like this. The effect of construing the will in the light of surrounding circumstances must be such as to exclude the probability that the testator meant to use the word in its primary, or statutory, meaning, and the only conclusion which can be reached here is that he may, or may not, have intended to include an adopted child. If there is nothing more in the case than this, it is plainly irrelevant and should be dismissed.

It remains to consider the other averments made by the pursuer in condescendences 5 to 9 inclusive. These averments I consider to be wholly irrelevant, because they seek leave to prove extrinsic facts which, in my opinion, cannot competently be regarded in a question of construction such as we have here. The averments in the condescendences to which I have referred may be summarised by saying that they set out various declarations made by the testator, and various events and circumstances, from which, if proved, it would be possible to infer as a matter of fact that the intention which the testator had in his own mind was to include his adopted child in his gift to children. The pursuer's only argument in support of the admissibility of this kind of evidence was based on the proposition that we had here a case of latent ambiguity. The will taken by itself, it was said, showed no ambiguity, but, when the surrounding circumstances revealed that the testator had left both ordinary children and an adopted child, an ambiguity then arose as to whether he had used the word "children" in a primary or secondary sense. Persuasively as the argument was put forward, it is, I think, very wide of the mark. In a true case of latent ambiguity, no doubt, it is permissible to have direct proof of the testator's intention as a fact. Where, for example, it is found that the designation of the legatee applies equally, or indifferently, to two objects, you have a problem which cannot be solved by construction and you may pass from the realm of construction to a direct inquiry as to which of the two objects the testator had in mind. But here there is no latent ambiguity. All that the surrounding circumstances reveal is that a question of construction arises as to the sense in which the testator has used the word "children." That question has to be solved purely as a question of construction, and the only extrinsic facts which may be regarded are the surrounding circumstances in the sense in which I have used that expression. It was suggested that the case of Nasmyth's Trustees v. National Society for Prevention of Cruelty to Children sanctioned a wider admission of evidence than this, but I do not so understand that decision. It is true that the word "ambiguity" is to be found in some of the opinions delivered, both in the Court of Session and in the House of Lords, but in both Courts the question was treated as a question of construction. The description given in the will fitted one of the claimants precisely and the other less precisely. The decision, as I understand it, did no more than affirm that the effect of reading the will in the light of the surrounding circumstances, and the facts relating to the two rival institutions, did not go the length of revealing, with the necessary degree of certainty, an intention to designate the institution less precisely described. Some of the evidence actually admitted in the case, as Lord Parmoor seems to indicate, may have gone beyond what was strictly relevant; but I do not think any of the noble Lords proceeded upon, or countenanced, anything beyond evidence of surrounding circumstances.

I have said all that needs to be said to explain why I agree that the case should be dismissed as irrelevant but there was an aspect of the defenders' argument to which, perhaps, I ought to refer. The defenders submitted (correctly, as I have held) that evidence of the surrounding circumstances was the only evidence admissible and, on that footing, they argued that certain observations made by Lord Cairns in Hill v. Crook applied to the situation and foreclosed any argument that the word "children" was used in a secondary sense. Lord Cairns divided the cases in which the prima facie meaning of the word "children" could be departed from into two classes. In the first class he put those cases in which the will itself contained no contrary expression of intention and where it was simply a question of construing the word "children" in the light of the surrounding circumstances. In the second class he put those cases in which the will itself contained some expression of a contrary intention and it was a question of construing a will of that sort in the surrounding circumstances. In referring to the first class Lord Cairns, in effect, observed that the primary meaning could only be departed from in cases in which the testator must have known that it was impossible for any legitimate children to take. The example he chooses is of a bequest "to the children of my daughter Jane, Jane being dead, and having left illegitimate children, but having left no legitimate children." The case then before the Court was a case belonging to the second class and there is no doubt that everything that was said by the other noble Lords, and by Lord Cairns in the rest of his speech, is just as much the law of Scotland as the law of England; but the question which was raised was whether, in dealing with cases where the word "children" has simply to be construed in the surrounding circumstances, the observation of Lord Cairns to the effect that the word was open to construction only in the event of it being impossible for legitimate children to take was part of the law of Scotland. If it is, it would provide an additional reason for holding the present case irrelevant, because here, to the knowledge of the testator, it was possible for legitimate children to take. The point does not require to be decided and I do not purport to decide it. Nevertheless it does seem open to doubt whether the observation should be regarded as applicable in a Scottish case. What Lord Cairns lays down is not a general principle and he does not lay it down as a general principle but rather as a rule, or principle, derived from earlier case law. It is to be noted that he prefaces his observation by saying that he has looked carefully through the long line of cases dealing with children who are said to be illegitimate "for the purpose of extracting any principle which appeared to be the ratio decidendi of those various cases." Those cases would appear to have no counterpart in Scotland and it certainly would be more in accord with our system to treat the question, in all cases, simply as a question of construction to be decided in the surrounding circumstances of the particular case. Of course, where there are legitimate children, or the possibility of legitimate children, who could take, it may be difficult to imagine a set of surrounding circumstances in which an intention to use the word in its wider meaning could be affirmed with the necessary degree of certainty, but I would not like to commit myself ab ante to the proposition that such a set of circumstances would never arise. We were referred to an Outer House case, Purdie's Trustees v. Doolan, in which Lord Murray held that, notwithstanding the existence of legitimate children, the word had been used to include illegitimate children. It was a peculiar case, in which the testator had known the illegitimate children and was quite unaware of the fact that their father also had several legitimate children in existence. I am not expressing any opinion as to whether, as a matter of construction, the result reached by Lord Murray was the correct result, but, in so far as he approached the question as an open question of construction in the particular surrounding circumstances, I am not prepared to say here and now that he was wrong.