(dissenting):
For me, Johns v. Cobb,1 decided just two years ago, furnishes the necessary guidance toward resolution of cases like the one at bar. There, by an approach sharply divergent from that which my colleagues now pursue, we held that a child adopted in 1924 took under each of two wills bequeathing income from a trust to the “issue” of his adoptive mother. One of wills was executed in 1922 by the adoptive mother’s grandmother, who died in 1928; the other in 1944 by the adoptive mother’s aunt, who died in 1946. Nothing within the wills indicated whether the word “issue” was intended to include adoptees, nor did the evidence extrinsic of the wills assist on that score; nor, we found, did “issue” have an established meaning proprio vi-gore barring relatives by adoption.2 Lacking direction from these sources, we reasoned:
In such a situation, where the record and the inferences to be drawn therefrom do not suggest what the actual intent of the testator might have been, it is our duty to supply a reasonable intent by implication. Our guide in this undertaking is necessarily public policy, in accordance with which we assume the testators would have wished their wills to be interpreted.
The 1963 adoption statute of the District of Columbia 3 provides a clear indication of this jurisdiction’s general policy toward adopted children. That statute, like the adoption statutes of many other jurisdictions, ac*213cords adopted children all the rights of natural offspring. In keeping with this policy, and in view of the unmistakable trend of judicial decisions in other jurisdictions, we held that, absent any contrary indication of the testators’ actual intent, appellant should be deemed to be included within the terms of the testamentary gifts to “issue” in the wills now before us.4
I am no more able to discern here an actual testorial intent than was the court in Johns v. Cobb. I would, accordingly, abide the canon of construction that case lays down. My colleagues, however, conclude that “the language of the will, read in light of the surrounding circumstances, establishes that the testatrix’s use of the word ‘issue’ was not intended to include children adopted by the grandsons.” 5 On that premise, “[they] do not find it necessary as a last resort to look at ‘public policy’ to determine the testatrix’s intent, as we did in Johns.” 6 In my view, “the surrounding circumstances” they cite to distinguish Johns do not suffice for that purpose, and their reading of Johns as a decision resting on a “look at ‘public policy’ to determine [testorial] intent” misses what Johns really holds. For these reasons, now to be elaborated, I respectfully dissent.
I
The majority offers several broad considerations in the effort to differentiate Johns from the situation at hand. The first is that the draftsman of the will currently under examination was expert in local estate law, and so was familiar with preexisting legal definitions of “issue.” This I have no reason to doubt; the will attests both the expertise and meticulous care that went into its formulation. But the draftsman in Johns, it appears, also was competent7 and, more importantly, the point detracts from the argument. As Johns signifies, the more expert the draftsman, the less likely he would be to depend on an ambiguous word like “issue” to exclude adoptees.8 The decisions in this jurisdiction leave its meaning, as respects adoptees, imprecise;9 those elsewhere come out both ways.10 I cannot assume that a careful, knowledgeable draftsman would deliberately choose “issue” in an effort to restrict residuary beneficiaries to the natural-born; rather, I would expect him to select terminology setting the clarity of his objective beyond cavil.11 *214That the draftsman of the instant will was inexplicit indicates to me that at best the matter was not considered, and that a worst there was no purpose to distinguish adopted from natural children.
The second consideration upon which the majority relies is twofold. One branch of the thesis here is that the cross-remainders upon death of named life tenants without issue ran initially to named persons in the testatrix’ own blood line. The other is that, since the duration of the trust is partly measured by the lives of grandsons’ “issues” living at the testatrix’ death, either grandson could extend the trust, if “issue” included adoptees, by adoption of persons in being during the testatrix’ lifetime. These facts, my colleagues say, indicate a testorial intent to limit “issue” to those born issue.
I think these circumstances are much too weak to evidence an intent to confine the trust benefits to issue by birth. A testator normally picks as the objects of his bounty those who are closest to him; more frequently than not, those picked eo nomine just happen to be relatives by blood. So also when the gift is to a class, it is usually a class of relatives. But I am unable to extract from these commonplace, recurring facts of life an intention to discriminate among those within a designated class purely on a biological basis.12
Indeed, our testatrix provided that any surviving widow of either of the two grandsons was to receive, until remarriage, one-third of the net trust income to which her spouse, if living, would have been entitled either as initial life tenant or as a cross-remainderman from another beneficiary. The testatrix provided, too, that in the event of a total failure of grandsons’ issue upon termination of the trust, the entire corpus of her estate is to be distributed to another stranger to the blood — The Washington Foundation — in fee simple. Less clear to me than to my colleagues, then, is an overriding testorial purpose to maintain the trust exclusively to relatives by blood.
Nor does an intent to exclude adoptees emerge from the fact that lives measuring the duration of the trust — the lives of issue in being at the time of the testatrix’ death — could be unascertainable at that time because a living child might later be adopted. Trust duration is commonly congruent with the life spans of beneficiaries, who may or may not be blood relatives and whose identity may be completely unknown during the testatrix’ lifetime — future spouses, for example. Any real concern is an adoption done in bad faith depriving other issue of their just shares of the trust. That is clearly not the case here. Appellee was the daughter of John Vandergrift Summerlin’s third wife, six years old when the marriage occurred, and so a normal candidate for adoption.13 In addition, she was born almost five years after the death of the testatrix, so her inclusion as “issue” cannot extend the trust.14 Furthermore, where the adop*215tion is in bad faith, there is adequate precedent to support judicial protection of the proper recipients.15 Since an adoption in bad faith is unlikely,16 did not occur here, and the court can intervene when it does, there is no reason to use its remote possibility to establish a general inference adverse to all adoptees.
II
The majority relies, perhaps most heavily, upon the circumstance that not until long after the will was made did District of Columbia adoptees become eligible to take by intestate succession through as well as from an adoptive parent17 That circumstance, my colleagues say, illuminates a testorial intent to exclude adoptees from benefits under the trust. That conclusion, I submit, departs completely from Johns v. Cobb. There we expressed our disbelief “that the question is controlled by the adoption statute which was in effect at the time [the] will was drafted; ‘it is the will that is here being construed, and not the statute.’ ”18 There, in this manner, we cleaved problems of intestate succession by adoptees from interpretative problems under wills possibly favoring adoptees. There, in like manner, we also rejected the theory, basic to the majority position, that legislation existing at execution of the will or at the testator’s death has automatic relevance to testorial intent. There, the legislation aside, we could find no trustworthy evidence of intent; we pointed out that “[i]t may well be that [the testator] never considered the matter at all.”19 There, in this way, we admonished that a statute fixing the intestate succession rights of adoptees is an unreliable resource as long as testators are apt not to focus on adoptees when they make their wills.
I see nothing in the situation now before the court distinguishing it, in terms of the pertinence of prior legislation, from the situation with which we were confronted in Johns. The wills in Johns were executed in 1922 and 1944; the will here in 1929. There and here, each testatrix died prior to the District’s “from and through” statute.20 We have no greater call now to look to previous legislation for aid.
The majority view in this case, I think, misconceives the role that the “from and through” statute plays in the solution of the problem before us. In Johns, we did not resort to the current *216statute out of a feeling that it could bear significantly on pre-enactment testorial intention; quite obviously, it could not. We could find no satisfactory indication of intent as to whether adoptees were to be included within the term “issue,” and we recognized the probability that the question had not even been considered. Thus faced with a record barren of the indicia of actual intent, we were sensitive to a “duty to supply a reasonable intent by implication.”21 And “[o]ur guide in this undertaking,” we continued, “is necessarily public policy, in accordance with which we assume the testators would have wished their wills to be interpreted.”22 So it was that we laid down a canon of interpretation: “[A]bsent any contrary indication of the testator[’s] actual intent, [an adopted child] should be deemed to be included within the terms of * * * testamentary gifts to ‘issue’ * * 23 We explained the rationale for this canon as follows:
This result, we think, is consistent with the probable preference of the great majority of testators. We have no reason to assume that these testators would have wished to discriminate against appellant “upon a biological basis.” There is no suggestion that his adoption was motivated by a desire to cut off other bequests under the wills. Had the present testators considered the problem of adopted children and decided to exclude them from their wills, they could have done so explicitly. In the absence of such a statement of intention, or any other evidence of testators’ actual intent, we can perceive no reasonable basis for excluding appellant from sharing, as “issue” * * * in the proceeds of the testamentary trusts.24
Johns did not look to the public policy of the adoption statute with a view to determining actual testorial intent. For that purpose, the statute was explicitly held to be irrelevant, with good reason. Unlike those allowing their property to pass by intestate succession, who may consciously rely on the statutory scheme of devolution and forego wills because they cannot improve upon that scheme, those writing wills have as their objective the replacement of the scheme with a personal expression of donative intentions. Thus courts construing wills cannot presume that words otherwise equivocal were intended to derive clarification from a statute which the will was drafted to avoid.
Once it is understood that it is attributed, rather than interpreted, testorial intent which public policy is to determine, it is clear that no purpose whatsoever is served by resort to past public policy. The public policy responsible for the present adoption legislation, concretizing the “desires of the many”25 by treating adopted and natural relatives on an equal plane, is the only appropriate *217reference for the judiciary. Johns so held in prescribing its canon of construction for ambiguous wills, as did the District Judge, following Johns when he construed the provisions in controversy. Measured by current policy, appellee is “issue” of her adoptive father, and in my view is entitled to share in the trust estate.
In final analysis, Johns stands for the proposition that unless it can be concluded with relative safety that the testator actually entertained an intent on the subject and that the intent is ascertainable, the judicial function properly is one of attribution to the testator of an intent realistically conceived, rather than a construction supported by doubtful vestiges of intent distilled from the particular combination of language and circumstances under scrutiny. This is but to say, as judges constantly say, that when a court lacks persuasive direction from either within or without the will, it is thrown back on a canon of construction. I see no part that bygone statutes can legitimately play in the formulation of a sound canon.
Today’s decision marks what in my view is a retrogression to the uncertainties characterizing the era predating Johns v. Cobb. It means that testamentary provisions in favor of “issue” may, as respects adopted children, depend upon an assessment of testorial intent drawn from factors the true import of which is a mystery. With such an approach, we may expect all the shades of difference in opinion inevitable in any human evaluative process. We may look forward to fine decisional distinctions to which testators never gave significance. We may anticipate, too, rulings bottomed on inferences many people might deem unreal. Johns was designed to put an end to difficulties of this sort by a canon predicated on popular expectations and sound social policy. In departing from Johns, my colleagues, I fear, open the door wide for pre-Johns frustrations to again plague litigants, lawyers and courts alike.
. 131 U.S.App.D.C. 85, 402 F.2d 636 (1968), cert. denied, 393 U.S. 1087, 89 S.Ct. 876, 21 L.Ed.2d 781 (1969).
. Id. at 86-87, 402 F.2d 637-638.
. 77 Stat. 537 (1963), D.C.Code § 16-301 et seq. (1967). The “form and through” provision reenacted therein first entered the law of the District of Columbia in 1954. 68 Stat. 245 (1954), D.C.Code § 16-312 (1967).
. Johns v. Cobb, supra note 1, 131 U.S. App.D.C. at 87, 402 F.2d at 638 (footnotes omitted). See also text infra, at note 24.
. Supra p. 204.
. Supra p. 204.
. Johns v. Cobb, supra note 1, 131 U.S. App.D.C. at 86, 402 F.2d at 637. See note 8, infra.
. Id. at 86-87, 402 F.2d at 637-638 (footnotes omitted). There we said:
Our cases have seldom discussed the term in relation to adoption, and enough uncertainty surrounds its usage that we doubt a competent draftsman who wanted to exclude adopted children would rely on the word “issue” alone to express that purpose. “Terms such as ‘issue’ * * * may or may not include adopted children; read alone, and apart from context, those words are ambiguous.” In re Upjohn’s Will, 304 N.Y. 366, 374, 107 N.E.2d 492, 495 (1952).
. Id. at 86-87, 402 F.2d at 637-638, quoted supra note 8. My colleagues concede that, for imprecision, “issue” is not “decisive” or “determinative,” although they find it “persuasive” in context. Supra p. 205.
. See the cases collected in Annot., 86 A.L.R.2d 12, 69-84 (1962).
. “It is not too much to ask the draftsman to address himself specifically and unambiguously to this question before finding that the exclusion of an adoptee was intended.” Halbacli, The Rights of Adopted Children Under Class Gifts, 50 Iowa L.Rev. 971, 980 n. 43 (1965).
The same consideration applies with equal force, I submit, to the majority’s suggestion that the fact that appellee was adopted after the testatrix’ death negates testorial intent to include her as a beneficiary. Supra p. 207. The point, once again, is that the meticulous draftsman of the will in suit would not have rested on that circumstance. Those adopted after the testator’s death have been per*214mitted to take, see In re Heard’s Estate, 49 Cal.2d 514, 319 P.2d 637, 643 (1957) ; In re Coe, 42 N.J. 485, 201 A.2d 571, 577 (1964) ; In re Park’s Estate, 15 N.Y.2d 413, 260 N.Y.S.2d 169, 207 N.E. 2d 859, 860-862 (1965), and this court lias said that it does not “attach great importance” to the time of the adoption except as it may have evidentiary bearing on good or bad faith. Johns v. Cobb, supra note 1, 131 U.S.App.D.C. at 86 n. 2, 402 F.2d at 637 n. 2.
. There was a time in our history when limitation of land ownership to blood relatives was a policy in vogue. The ancient doctrines governing the devolution of landed estate in England, particularly fee tail, together with the doctrine of primogeniture, fitted well in a way of life grounded politically, economically and socially in a system of fuedalism, to which statutory adoption was unknown. Today, we live in a society in which adoptions are everyday occurrences, and natural children share the limelight with adoptees.
. Compare Halbach, supra note 11, at 990-992.
. Only issue of the grandsons “living at the time of [the testatrix’] death” could do so.
. Compare Johns v. Cobb, supra note 1, 131 U.S.App.D.C. at 86 n. 2, 402 F.2d at 637 n. 2; Halbach, supra note 11, at 987-993. And see note 25, infra.
. See Noreen v. Sparks, 92 U.S.App.D.C. 164, 204 F.2d 56 (1953). See also Bedinger v. Graybill, 302 S.W.2d 594 (Ky.1957). And see note 25, infra.
. See note 3, supra.
. Johns v. Cobb, supra note 1, 131 U.S. App.D.C. at 86, 402 F.2d at 637, quoting Noreen v. Sparks, 92 U.S.App.D.C. 164, 166, 204 F.2d 56, 58 (1953).
. Johns v. Cobb, supra note 1, 131 U.S. App.D.C. at 87, 402 F.2d at 638.
. Our testatrix died in 1930. The two involved in Johns died in 1928 and 1946. Joint Appendix, Johns v. Cobb, supra note 1, at 203, 131 U.S.App.D.C. 87, 402 F.2d 638.
My colleagues endeavor to bolster their position by reference to a statement in the legislative history of the “from and through” statute and the nonretroactivity provision related to it. Supra pp. 208-209. Both antedated Johns v. Cobb, but the court there made no mention whatever of them although either, I think, would have altered the Johns holding if given the construction my colleagues espouse. For me, the court’s silence indicates a contrary view — one which I share. To the extent that the statement reflected anything at all on the relation of succession law to wills interpretation, it was that those disliking the new “from and through” provision must change their wills to avoid it — that the new legislative policy would supplant the old in the interpretation. The nonretroactivity provision, D.C.Code § 16-315 (1967), while of course applicable to intestate succession by an adoptee, is irrelevant to the problem at hand. As is elucidated herein, the consideration important to testorial intent is not the “from and through” provision in substantive operation, but the public policy that gave birth to it.
. Johns v. Cobb, supra note 1, 131 U.S. App.D.C. at 87, 402 F.2d at 638.
. Id.
. Id.
. Id., quoting In re Estate of Coe, supra note 11, 201 A.2d at 575 (footnotes omitted).
. “The proposed [‘from and through’] law * * * brings to the foreground the growing concept that adopted children are in all respects the actual children of the adopting family. Sometimes the argument is made that in this manner a child will be sneaked in on unsuspecting grandparents but the committee believes that this situation seldom occurs. One of the members of the adoption committee of the Bar Association of the District of Columbia conducted a survey among adopting grandparents a few years ago and there was unanimity among such grandparents that they much preferred a ‘from and through’ law. It was therefore the thought of the committee that the law should be brought into conformity with the desires of the many and the few that objected to it could specifically change their wills to the contrary. Furthermore, the proposed change is in keeping with the idea that the greatest possible protection should be given the adopted child.” H.R.Rep.No. 1347, 83rd Cong., 2d Sess. 7 (1954). See, to the same effect, S.Rep. No. 1379, 83rd Cong., 2d Sess. 6 (1954). See also Halbaeh, supra note 11, at 985 n. 66.