Appellant was convicted on nine counts of an indictment charging violations of 68A Stat. 551 (1954), as amended, 26 U.S.C. § 4705(a) (1958), 68A Stat. 550 (1954), as amended, 26 U.S.C. § 4704(a) (1958), and 70 STAT. 570 (1956), 21 U.S.C. § 174 (1958), for participation in narcotics transactions on three separate occasions. The principal issue on appeal is whether the evidence required the District Court to instruct the jury on criminal responsibility as requested.1 The mental condition relied upon by appellant as the basis for the instruction arises out of his addiction to narcotic drugs.
To raise the issue of criminal responsibility, it must appear from the evidence that the accused, at the time of the alleged criminal act, suffered from some mental disease or defect, which we have defined as meaning an abnormal condition of the mind which substantially impairs capacity to control behavior. McDonald v. United States, 114 U.S.App.D.C. 120, 124, 312 F.2d 847, 851 (1962). This mental condition may be shown by observation and opinions of lay witnesses or experts; it does not depend upon psychiatric labels, or medical classifications and terms, although testimony of an expert giving a diagnosis of a mental disease or defect, in those terms, would raise the issue for jury determination. Lay testimony describing significantly bizarre, abnormal conduct also could be sufficient to raise the issue. We have said this must be “more than a scintilla,” of evidence. McDonald v. United States, supra. In the first instance the question whether the evidence in a particular case constitutes “some evidence” of insanity, like other evidentiary rulings, is a question of law for the court. Of necessity, it is based on the trial judge’s evaluation of the facts before him.2
We hold only that a mere showing of narcotics addiction, without more, does not constitute “some evidence” of mental disease or “insanity” so as to raise the issue of criminal responsibility. This is not to say that evidence that an accused is an addict is without probative value along with other evidence on the issue of responsibility but only that alone it is not sufficient to require giving the Durham-McDonald instruction. Cf. Hightower v. United States, 117 U.S.App.D.C. 43, 325 F.2d 616 (1963). Some mentally ill persons are addicts and some addicts are mentally ill; the two conditions can coincide but we give no more credence to the notion that all addicts are mentally ill than to the converse that all mentally ill persons are addicts. To so hold would make every addict’s case an “insanity” case.3
*45We have gone to great effort in our McDonald holding to restore the issue of criminal responsibility as one of fact for the jury and to make it clear that the expert’s label is relatively unimportant but that his description and explanation of capacity to control behavior are critical. We have frequently urged that trial counsel and their expert witnesses should seek to avoid being content with mere expert conclusions and should emphasize the reasons, the factors, the symptoms, and the medical reasoning which led to the conclusions so that from the experts the jury will have a psychological profile of the accused and not simply a collection of psychiatric labels and technical jargon. See, e. g., Carter v. United States, 102 U.S.App.D.C. 227, at 236-237, 252 F.2d 608, at 617-618; Blocker v. United States, concurring opinion 110 U.S.App.D.C. 41, at 51-52, 288 F.2d 853, at 863-864.
The narrow question presented is whether there was sufficient evidence to require the Durham-McDonald instruction. Three experts were called as witnesses in this case. Two testified categorically that appellant was without any mental disease or defect and had no abnormal condition of the mind at the time of the alleged acts. All three psychiatrists testified that deprivation of heroin could to varying degrees impair a heroin addict’s capacity to control his conduct.4 The psychiatric testimony upon which appellant chiefly relies concerning capacity for control was premised on hypothetical facts not supported by evidence in this record,5 i. e., evidence of depriva*46tion of heroin at the time of the offense. Had there been evidence that, at the time of the alleged unlawful acts appellant “need[ed] a fix,” as one expert put it, and that narcotics were then unavailable to him, the hypothetical conditions postulated would find support in the record. But the record contains no suggestion that at the time of the alleged criminal acts appellant did in fact “need a fix”; to the contrary undisputed affirmative evidence shows that at that very time he possessed large quantities of heroin available for sale. No rational thesis can demonstrate that an addict “in need of a fix” would be in a state of deprivation at the precise time he had an abundant supply of heroin in his possession.6 In short, the most favorable interpretation of the medical testimony is that with no heroin available and the need urgent, appellant’s behavior controls might have been affected; but the evidence on this record negates those interdependent factors.
The conclusion is inescapable, therefore, that with no claim or showing of need for and deprivation of narcotics —the hypothesis on which the most favorable medical testimony was premised — there was no evidence that appellant’s capacity to control his behavior was impaired. There being an absence of the requisite evidence that at the time of the act charged appellant had an abnormal condition of the mind which substantially impaired his behavior controls, the McDonald standard for submission of the criminal responsibility issue was not met, and the judgment therefore is
Affirmed.7
. Appellant also contends that the evidence necessitated an instruction on entrapment. On the circumstances in this case, we cannot conclude that failure to so instruct was reversible error. See Berry v. United States, 116 U.S.App.D.C. 375, 324 F.2d 407 (1963), cert. denied, 376 U.S. 959, 84 S.Ct. 972, 11 L.Ed.2d 977 (1964).
. The dissent argues that under McDonald only the jurors, not psychiatrists or judg-. es, should pass on criminal responsibility. This oversimplification of the issue overlooks the historic proposition that jurors are not permitted to speculate their way to a verdict; a verdict must rest on evidence and it is the responsibility of the trial judge to decide, first, whether there is sufficient evidence on the issue for submission to the jury. Here there is no evidence either of mental disease or of actual impairment of capacity to control behavior.
. Reference in the dissent to our ambiguous if not misleading statement in Brown v. United States, 118 U.S.App.D.C. 76, 331 F.2d 822, 823 (1964), requires that the quotation from Brown and the reli-*45anee on Robinson v. State of Cal., 370 U.S. 660, 667, 82 S.Ct. 1417, 8 L.Ed.2d 758 (1962), be scrutinized in context.
That portion of the Brown quotation from Robinson which states that narcotics addiction is a mental illness is not the language of the Supreme Court. Rather, that Court quoted from the brief of the appellee in Robinson by way of a footnote to its opinion, only to show that the appellee in Robinson conceded the state of addiction to be an illness. The footnote in Robinson, at page 667, 82 S. Ct. at page 1420, reads in pertinent part as follows:
“8 In its brief the appellee stated: ‘Of course it is generally conceded that a narcotic addict, particularly one addicted to the use of heroin, is in a state of mental and physical illness. So is an alcoholic.’ ”
Compare Castle v. United States, 1964, 120 U.S.App.D.C.-, 347 F.2d 492 (concurring opinion). The quotation from the per curiam opinion in Castle, supra, relied on by the dissent, was unnecessary to our affirmance of the conviction. Obiter dicta cannot be escalated into a holding by reciting it in one case and repeating it in another.
Thus the Supreme Court’s footnote in Robinson, which hardly rises to the level of obiter dictum, did not in any sense intimate a view that the Court considered narcotics addiction a mental illness in the criminal-responsibility context. Its holding was merely that the addiction status may not be punished as a crime. If we are to speculate about what that Court meant by quoting from appellee’s brief in Robinson, it would be well to note that the full quotation equated narcotic addiction with alcohol addiction.
. Dr. Dobbs, the expert whose testimony was the most favorable to appellant’s claim, testified as follows:
“Q. Doctor, in this condition [addiction described by appellant as use of three or four shots per day and up to eight or nine] would his behavior controls be substantially impaired ?
“A. As a result of the heroin or of the active process of the addiction in which case he might need a fix, to use the slang term, yes, they would be.
“Q. And would his mental or emotional processes be substantially affected?
“A. For the same reasons, yes.
“Q. And would you consider this condition to be an abnormal condition of the mind?
[Objections intervened.]
“The Court: I will let her answer. Was it abnormal in the sense it is a mental disease or mental defect?
“The Witness: Abnormal in that it is not normal. Abnormal in the sense it is a mental disease in and of itself, no, it is not, Your Honor.”
Record, p. 93.
. More often than not expert opinions are based upon assumed facts embraced in hypothetical questions but a hypothetical answer based on facts which are not shown to exist is not evidence. Here the expert said in effect that if appellant had *46been deprived of needed narcotics at the time of the offense then his conduct might have been compulsive. But the record is barren of evidence of the very condition upon which the hypothesis was dependent. Cf. Hightower v. United States, 117 U.S.App.D.C. 43, 325 F.2d 616 (1963).
. The dissent asserts that “addiction, with adequate exposition of cause and effect, is ‘some evidence’ of mental illness.” With this proposition in the abstract we do not disagree; we note merely that the posited evidence of causality is not present in this record: the hypothetical testimony concerning breakdowns in behavior controls caused by separation of an addict from his drug is meaningless without evidence that this appellant was so deprived.
. Since this opinion was filed the court voted to deny Appellant’s petition for rehearing en banc; under our practice individual judges are free to file their comments on the court’s opinion.
Judge Bazelon’s statement seems to add up to the very conclusion reached by the division drawn to hear this case — namely that submission of an insanity issue to the jury could have resulted in nothing but impermissible speculation. The majority of this court do not, however, share Judge Bazelon’s view that trial counsel’s performance was deficient.
It should be further noted that the journals and other literature cited in Judge Bazelon’s statement were not cited to the court by the litigants and hence have not been subjected to the useful screening afforded by the adversary process. Whether they are fairly representative or reflect simply one point of view we do not know. Discussion of addiction as a mental illness in a broad social context or for treatment purposes is quite a different matter from labeling addiction “mental disease” in the context of determining criminal responsibility.
We can all agree generally that society is doing too little with drug addiction but the solution does not lie in excusing all addicts for their crimes, nor in straining the law to acquit all narcotics peddlers, nor in sending all addicts to overcrowded mental institutions. The judicial process is not adapted to accomplish even the most urgently needed social reforms. Such solutions must come from the Legislative and Executive Branches and we do a disservice to the whole structure by trying to use judicial power to these ends. The late Justice Frankfurter often reminded us that courts cannot find judicial remedies for every social, economic or political problem which afflicts Mankind.