WILLARD BARTLETT, Ch. J.

On the 24th day of August, 1913, Willis Buffom, a farmer fifty-two years of age, died of arsenical poisoning at his home near Little Valley in the town of Mansfield, Cattaraugus county. I state the cause of death thus positively for the reason that counsel for the defendant in his opening address to the jury stated that the proof tended to show there was arsenical poisoning; and no serious attempt was made to dispute the evidence for the People to that effect. The defendant is charged by the indictment with having caused the death of Willis Buffom by means of arsenic administered to him first on the 7th day of July, 1913. She had been married to him when she was only fifteen years of age and she was thirty-nine years old at the time of his death. At that time the couple had six children living; five sons, aged respectively twenty three, twenty-one, nineteen, sixteen and thirteen, and one daughter, Laura, aged ten. The three younger sons and the daughter lived at home with their parents.

The evidence throughout the case shows that Willis Buffom and his wife lived a hard and laborious life, moving from one locality to another as tenants of different farms. Their relations toward one another were not harmonious and the husband had been in the habit of accusing his wife of marital infidelity with various persons, some of his accusations amounting to charges of incest. The theory of the prosecution was that the defendant had finally become infatuated with a young man who resided in the vicinity, named Ernest Frahm, and that it was her illicit passion for him which induced her to encompass the death of her husband. The theory of the defense was somewhat vague and it is not easy to gather precisely what it was from the addresses of counsel. There was undoubtedly, however, an attempt to prove in behalf of the defendant that Willis Buffom in consequence of the depression due to his disagreements with his wife and to the burden of debt which oppressed him, had contemplated taking his own life and the lives of his children to the end that his wife would thus be freed and his indebtedness satisfied. Thus in the opening address of defendant's counsel it was said: "In the haying season he again made threats which I expect to show that he had better do away with himself and the family and then his debts would be paid. He had a large accumulation of debts. He had acquired them here and there and they weighed upon his mind."

The conviction rests chiefly upon a confession obtained from the defendant at Buffalo on the 10th day of December, 1913. Earlier confessions had been secured, but they were excluded by the trial court upon the objection of counsel for the defendant and with the acquiescence of the district attorney who declared that he was not very insistent that they should be admitted. There is no doubt that all the confessions including that which was received in evidence were obtained as the result of a carefully planned scheme of falsehood, artifice and deception, organized and carried out at the instance of the public prosecutor. This is conceded by the learned district attorney who asserts, however, that the deception thus practiced was legitimate in order to induce the defendant to speak the truth and, therefore, that it did not vitiate the confession when obtained. "Courts of the highest authority," he declares, "have given approval in many cases to the practice of procuring confessions of guilt by deception." This statement, I think, goes too far. In People v. White ( 176 N.Y. 331) Judge VANN, who wrote the opinion of this court, said expressly that it did not sanction the deception practiced by one of the officers in charge of the defendant, although it could not exclude the confession made to him on that account. This is very far from approval. Nevertheless the law is well settled that the fact that a confession has been procured by deceptive practices is not sufficient to justify withholding it from the consideration of a jury nor will it operate to prevent them from basing a finding of guilty thereon if it is not open to any other objection and is corroborated as required by law.

If the statements attributed to the defendant in her confession to the district attorney on December 10th, 1913, are true there can be no doubt of the defendant's guilt. The annals of criminal jurisprudence, however, abound in cases of false confessions induced by the hope of escape from punishment or the mitigation of punishment or of some other benefit to be gained by the confessing party. Indeed, there have been instances of false confessions for which it was impossible to assign any reasonable motive whatever. Where, therefore, as in the present case, the confession upon which the conviction rests was concededly obtained as the result of a long course of fraud and deception, it becomes extremely important to protect the defendant against any judicial error which may have led the jury to attribute truth to the statements therein contained when otherwise they might have entertained doubt as to those matters. After a careful study of the record in the present case I am unable to resist the conclusion that a serious and harmful error was committed against the defendant in admitting the evidence which was received concerning the illness and death of the daughter Laura long after the death of Willis Buffom. Its reception violated the well-established general rule which forbids proof of any crime not alleged in the indictment. The disregard of that rule was the basis of the reversal of the judgment of death in People v. Molineux ( 168 N.Y. 264).

It appears that the girl had been somewhat ill at the time when other members of the family were sick prior to the husband's death; but she had fully recovered and was attending school on September 15, 1913, on the occasion of the onset of the illness which proved fatal some weeks later. She did not die until February 2, 1914, after the indictment against the defendant was found. Nevertheless the prosecution was allowed over the objection and exception of the defendant's counsel to prove all the facts connected with the symptoms and character of her illness and also to present for the consideration of the jury in technical detail the particulars and results of the autopsy performed upon her body. It seems to me that the inevitable effect of the admission of this evidence was to place the defendant upon trial for the murder of her daughter as well as the murder of her husband. The learned trial judge sought to limit the effect of this evidence as serving to illustrate the symptoms caused by arsenical poisoning, saying it was only important as indicating whether Willis Buffom and Laura Buffom died from a like cause. He said: "I think I should limit it for the bare purpose of an illustration as to how arsenic affects the human system, for the purpose of showing what the symptoms were that afflicted or troubled Laura Buffom and as indicating whether or not the symptoms displayed by Willis Buffom were those of arsenical poisoning."

I am utterly unable to see how this testimony was admissible from this point of view. I think the books will be searched in vain for a case in which it has ever been held that the symptoms produced by the administration of a specified sort of poison can be proved by recounting the circumstances of the illness and death of another patient who is alleged to have died from the effects of the given poison. But if there were such a case it would not be a precedent for the admission of such testimony when it carried with it the almost inevitable implication that the person on trial was responsible for the death of the other patient whose case was thus employed for illustrative purposes. Here, no matter what was said by the learned trial judge or counsel, it was certain that the reception of this grewsome testimony concerning the death of this little girl would impress the jurors with the idea that the defendant on trial was in some way or to some extent responsible for the fatality. That this was the impression produced on the jury is manifest from the question put by one of the jurors at the conclusion of the judge's charge. I quote from the record:

"A Juror: There is one point that has been in my mind; supposing this jury should find that Willis Buffom died from arsenical poisoning with intent to kill and that the same hand that gave the poison to Willis Buffom gave it to Laura Buffom, have we a right to consider that in connection with the case, as circumstantial evidence?

"The Court: No, I think not. That evidence as to the death of Laura Buffom as I stated to you was given for the sole purpose of affording you an opportunity of finding out how arsenic affected the body. You have a charge here of the death of Willis Buffom and that is all that is before you. You may retire, gentlemen.

"Mr. Cole: If your Honor please, I assume the jury might consider the circumstances of the death of Laura Buffom and the time of her attack as bearing upon this question of suicide and as to whether Willis Buffom's scheme of suicide comprehended the killing of his entire family. If they are satisfied that Laura Buffom's sickness did not begin until after the death of Willis Buffom then they might take that fact, bearing upon this theory of the defense that it was he who administered the poison.

"The Court: Yes, that is true.

"Mr. Collins: An exception."

In the first part of this narrative we have a distinct indication that one of the jurors, at all events, had been impressed with the idea that the death of the daughter Laura tended to prove the murder of the husband Willis Buffom. The latter part of the quotation in which the district attorney with the sanction of the court suggested that the jury might consider the death of the daughter upon the question of suicide and the husband's scheme of killing the entire family, seems to me indicative of a still further erroneous view as to the purpose for which this proof might properly be considered. How the subsequent illness and death of the little girl could throw any light on the question whether the husband in his lifetime intended to kill himself and his children I utterly fail to perceive. There does not seem to me to be any causal relation between the two sets of circumstances. An event occurring after a man's death and in no wise due to any agency exerted by him during his lifetime can have no relation to any intention formed by him or act done by him while he was alive.

In order to justify the reception of this evidence concerning the death of the daughter, the learned district attorney goes to the extent of asserting that it was admissible in any event as a part of the People's affirmative case under the authority of People v. Molineux ( supra). In the opinion in that case Judge WERNER classified the exceptions to the general rule which excludes evidence of other crimes with remarkable precision, showing that proof of another crime was competent to prove the particular crime charged only when it tends to establish either (1) motive; (2) intent; (3) the absence of mistake or accident; (4) a common scheme embracing the commission of two related crimes; or (5) the identity of the defendant.

In the Molineux case Chief Judge PARKER, although he concurred only in the result, had this to say in reference to the general rule of exclusion: "There is no controversy in this court — nor out of it — so far as I know, touching the general rule that evidence of the commission by him of other crimes is not admissible upon the trial of a defendant charged with crime. It is only on rare occasions that proof of the commission of another crime by the defendant is either necessary or helpful toward establishing the crime with which he is charged. Hence the evidence is ordinarily irrelevant, while at the same time its admission would necessarily operate to so prejudice the jury against a defendant as that in a doubtful case it might control the verdict. Therefore the courts long ago decided that a defendant should not be prejudiced by the admission of evidence of other crimes committed by him which in no wise tends to establish that he committed the crime for whose commission he is on trial." (p. 340.)

As to the five exceptions to the rule excluding proof of other crimes enumerated by Judge WERNER, it seems to me that only two can even plausibly be invoked as applicable in justification of the admission of such evidence in the present case — the admissibility of such evidence to establish motive or to establish a common scheme. I think it clear that the illness and death of the little girl cannot fairly be regarded as bearing upon the question of motive. As to this point, it should be observed that the learned trial judge expressly refused to receive this testimony as bearing on that question. He said when it was first sought to be introduced: "I will not receive it at all on the subject of motive. I do not think there is anything to show that it was the same motive at all." In this view I think the court was manifestly correct. The suggestion that this mother who had borne a family of seven children, toward whom she evidently entertained the warmest affection, should kill her daughter only ten years of age in order to enter upon and maintain illicit relations with Ernest Frahm with greater freedom, is to ignore every probability suggested by the record in the present case. The very confession upon which the prosecution relies and without which there could not possibly have been a conviction, shows again and again the defendant's love for her children. In it she repeatedly asserts that she refused to yield to the importunities of Frahm to go away with him because of her love for her children and her unwillingness to separate herself from them. If these statements represented her true feeling it is impossible for me to believe that she was actuated by any motive which could have induced her to desire to kill her daughter; nor can I spell out the existence of any common motive from which it could be inferred that because she tried to kill the daughter afterward she must previously have killed her husband. What I have already said applies to the evidence under discussion as tending to establish a common scheme as well as upon the question of motive. It was not receivable for either purpose.

The district attorney, in arguing that no error was committed in receiving this evidence of the illness and death of Laura Buffom makes four points which I will consider in order.

It is urged that this line of proof was abandoned before any substantial part thereof had been made, owing to the ruling of the court limiting its purpose and effect.

It is true that the district attorney said he would have to abandon that line of proof after the court had intimated that the history of Laura Buffom's illness must be limited simply to the question of her symptoms. Almost twenty-four pages of the objectionable testimony, however, had been taken before this point was reached. This evidence was never stricken from the record and the whole subject of Laura Buffom's illness and death was subsequently reopened in rebuttal when thirty pages more of this kind of proof were taken. This does not seem much like an abandonment of an effort on the part of the prosecution which, whatever its purpose, had a most prejudicial effect upon the defendant's case by subtly inducing the jury to believe that the defendant was responsible for the death of her daughter as well as the death of her husband.

The second proposition by which it is sought to justify the introduction of this evidence is the statement that the whole theory of the defense "fully opened up the question and made the evidence entirely competent in rebuttal." I quote from the brief in behalf of the People: "The whole defense and the whole theory thereof from the commencement to the end of the trial was that Willis Buffom killed himself as a part of a scheme involving not only his destruction, but the destruction of the children comprising the family living at home. While this theory was rather skillfully masked in the cross-examination of the People's witnesses, yet, in the opening of defendant's case the purpose of defendant's cross-examination became clearly apparent. This issue was not forced on the defendant by the fragments of proof already offered by the People regarding Laura's death. It was the defense. The defendant's proof was substantially all directed to this defense."

This is far too strong a characterization of the defense. As I have already intimated it is rather difficult to make out precisely what the defense was further than a denial that the defendant had intentionally poisoned her husband. It is true that counsel suggested the idea that Willis Buffom had planned the destruction of his own life and the lives of the other members of his family who were living at home; but no vigorous effort was made to impress any such theory upon the jury either by way of evidence or in the addresses of counsel. The district attorney, however, availed himself of the opportunity presented by the introduction of the weak testimony which the defendant's counsel elicited on this subject, to go into the circumstances of the poor little girl's illness and death in extenso by way of rebuttal; thus leaving the jury at the very end of the case deeply impressed with the wickedness of any one responsible for the innocent child's death, and certainly suggesting the idea that the defend ant was that person.

In the third place, the district attorney asserts that the court specifically restricted the evidence for the sole purpose of eliciting the symptoms and cause of the death of Willis Buffom both in connection with the admission of the testimony and in the charge.

This attempted restriction could not in the very nature of things have been effective with the jury or have prevented them from considering the evidence as tending to establish the commission of the crime charged in the indictment. This appears from the juror's question previously quoted. Furthermore, as I have tried to show, the evidence could not properly be received even for the limited purpose stated by the court.

Finally, we are told that the evidence in any event was entirely competent. This dogmatic statement is not supported by anything except the citation of the Molineux case; there is no attempt to sustain it by argument. It seems to me quite clear that its admission cannot be justified under the Molineux decision.

The suggestion has been made that counsel for the defendant relinquished his objection and exception to this line of evidence concerning the daughter's death by a concession which he made in the course of the trial. I am satisfied that no such effect should be given to what he did. After a long colloquy between court and counsel when the learned trial judge had come to the conclusion, as he said, to limit the evidence "for the bare purpose of an illustration as to how arsenic affects the human system," counsel for the defendant said: "I presume I may simply note an exception to the ruling." Further on in the case, as he says in his brief, to avoid the deadly inferences urged by the prosecutor in introducing this proof and in the hope thereby of eliminating the matter from the trial, defendant's counsel said: "I will make a statement. There is no question about the fact that Laura Buffom was ill at the time that this witness last sworn speaks of, that she was ill for a long time after that and eventually died. That is not questioned by the defendant." This concession was not accepted as sufficient either by the trial judge, who remarked that the People wanted to prove what the symptoms of the child's death were so as to aid the jury in coming to a conclusion as to what it was that killed Willis Buffom; or by the district attorney who said that he could not show that the girl was sick without showing the symptoms. The admission, therefore, failed to protect the defendant, as her counsel evidently hoped it would, or to prevent the further introduction of evidence on the subject. I do not see how it can be regarded as a waiver of his objection and exception to proof of this character. When the whole record of this branch of the case is considered it will be found that counsel resisted the introduction of the evidence under consideration as long as he thought it would do any good and, at all events, sufficiently to negative the idea of any waiver on his part.

To my mind the possibility — nay, the probability — that the jury may have been influenced to take a different view of the confession from that which they would have taken if the evidence in regard to the daughter's death had not been laid before them with so much emphasis and particularity convinces me that we cannot disregard the error which I think was committed in receiving that evidence.

For this reason I advise the reversal of the judgment of conviction and the granting of a new trial.

WERNER, HISCOCK, COLLIN and HOGAN, JJ., concur; CARDOZO and CUDDEBACK, JJ., dissent on the ground that although the evidence of the illness and death of Laura Buffom was inadmissible when first offered, it became admissible and the error was cured as a result of the course of the defense.

Judgment reversed, etc.