LEWIS, J.

In the early morning of October 30, 1930, Frank LaScala was shot while standing at the corner of Lenox avenue and One Hundred and Twenty-first street in New York city. He died from bullet wounds six hours later. More than nine years thereafter the present defendant, Gabriel Bartelini, was indicted for the homicide and was convicted in the Court of General Sessions of murder in the first degree. Upon this appeal by the defendant the District Attorney, with commendable frankness, concedes that, without proof of dying declarations by the decedent, there could have been no judgment of conviction. Accordingly, our inquiry goes to the question whether the proof of dying declarations, when considered with relevant evidence, was sufficient to support the jury's finding of defendant's guilt beyond a reasonable doubt.

Reference should be made at the outset to the testimony of the People's witness, Arthur Donaldson, who was the only eye witness to the homicide. Shortly after midnight on October 30, 1930, Donaldson was walking southward on the west side of Lenox avenue. As he reached the northwest corner of Lenox avenue and One Hundred and Twenty-first street he saw across the street on the southwest corner a man facing westerly beckoning to someone in a candy store to come out. Immediately there emerged from the store a man who proved to be the decedent, Frank LaScala. He advanced to within eight feet of the man who had beckoned to him, when suddenly the latter fired two shots at LaScala and then ran to an automobile which was parked on One Hundred and Twenty-first street. As the assailant mounted the running board the car moved westerly toward Seventh avenue and disappeared.

At this point the fact should be stated that although Donaldson had seen the face of the man who fired the shots and fled, and although he was able to describe the assailant as a man of dark complexion, thin and of a height about five and one-half feet, he was not able, when called as a witness upon the trial, to point out the assailant, although we assume the defendant was present before him.

Continuing with Donaldson's testimony, when he heard the two reports and saw two flashes at the instant of the shooting, he hastened to the store where he found LaScala slumped down behind a counter bleeding profusely and in physical distress. Recognizing that his condition was grave, Donaldson, aided by the witness Hillick, a police officer, and by Mamikoff, a taxicab driver, rushed the victim to Harlem Hospital where he arrived at about twelve-thirty A.M. and was at once placed upon an operating table in the emergency room. There his first dying declaration was made which is presently to be considered.

Another of the People's witnesses, Irving Rosenthal, a taxicab driver, at the time of the murder was "cruising" in his cab northward on Lenox avenue. As he approached One Hundred and Twenty-first street he heard a report "like a gun shot" and turned "to where the report was coming from." As he did so he saw a man, ten feet from the southwest corner of Lenox avenue and One Hundred and Twenty-first street, running northward to the curb where he entered the driver's seat of a Ford sedan and the car moved away. Rosenthal did not see the face of the man who ran from the scene of the shooting.

In the testimony of each of these two men is the important item of proof that when the decedent came out from the store only one other person was there — the man who had beckoned to him — and that it was that one man alone who fled from the scene to the parked car.

From this narrative given by the only witnesses who were near the scene of the crime when it occurred, we pass to a consideration of the evidence by which the decedent's dying declarations were proved. Upon this branch of the People's evidence the defendant's challenge places emphasis upon vital inconsistencies between testimony received at the trial of the present case in November, 1940, and the sworn testimony upon the same subject and by the same witness given nearly ten years before — in 1930 — at a hearing before a magistrate. As the inconsistencies relate themselves to proof essential to the People's case, the testimony will be set forth at some length.

The first dying declaration made by the decedent and received in evidence was given in response to questions addressed to him by the People's witness O'Connor, a detective, who was assigned to investigate the case on the night of the murder. O'Connor arrived at the hospital at about twelve-forty-five A.M. and first questioned the decedent while the latter was lying upon the operating table in the emergency room: "Q. What was the first question you asked of the injured man lying on the table? * * * A. I asked the injured man did he feel that he was going to die from the injuries he received. Q. What was the answer, if any, which he made to you? A. He says, `Don't you see I am going out? I feel rotten.' Q. Did you ask him a question after that? A. I said to him, `What do you mean by "going out?"' Q. Did he answer that question? A. He says, `I mean I am going to die.' * * * `Don't you see the blood on my mouth and clothing?' * * * Q. What was the next question you asked of this injured man? * * * A. Now, when I said to him, `Now, you are shot and somebody shot you. Why don't you tell me who shot you and I can help you out?' * * * Q. Will you tell us what his answer was? A. He said, `Maxie, the Weasel, and Bobby shot me.' * * * Q. Did you ask him any further questions after that? A. I did. I asked him did he mean `The Weasel' from Harlem, and he says — he described `The Weasel' to me, and I said to him, `Who is Bobby? Do you know his last name?' He said `Bobby Bartelini.' * * * I asked him why did they shoot him, and he said he was away, and when he got out he went around looking for Bobby, who was a friend of his; and he went asking questions about Bobby, where he was, and he went talking about Bobby to other people, and Bobby never did send him anything to prison when he was in prison, and he was asking about that. Q. Now, after the answers made by this injured man — I am repeating the question, Mr. O'Connor — in which he told you in substance that a person called `The Weasel' and Bobby Bartelini had shot him, will you tell us what further questions you asked of him with reference to the shooting and what answers were made? A. I asked him how did it happen. He said he was in a candy store on 121st Street and Lenox Avenue, when Bobby come — a girl come to the store, come to the door, and asked him that somebody wanted him outside. * * * When he went to the door, he opened the door and looked out, and saw Bobby and `The Weasel,' and Bobby called him outside the store, and when he walked — took a few steps outside the door, Bobby shot him."

Following this declaration the decedent was removed from the emergency room to an X-ray room and thence to a bed in a ward where at one-thirty A.M. Detective O'Connor again questioned him. This latter declaration, however, was excluded upon the ground that the decedent had indicated he did not then believe he was about to die.

During the cross-examination of the witness O'Connor there was received in evidence a transcript of testimony given by O'Connor at a hearing in the Homicide Division of City Magistrates' Court, Borough of Manhattan, on December 12, 1930. The hearing involved Maxie Lowenstein, known as "The Weasel," who was charged with having killed the decedent LaScala. From the transcript of testimony thus introduced in the present record it appears that O'Connor had testified in 1930 that on the night of the homicide he had questioned the decedent in the emergency room at the hospital at twelve-thirty A.M. At that time he had asked the decedent, "Do you believe you are going to die?" — to which the decedent replied, "No, not now." That statement, made by O'Connor under oath and only six weeks after the dying declaration was made, varied on a vital point, as we shall see, from the sworn statement by the same witness, quoted above, made nearly ten years later on the trial the record of which we now review. If O'Connor's testimony given in 1930 was to be believed, viz., that at the time the decedent made the declaration in the emergency room of the hospital at twelve-thirty A.M. he did not believe he was about to die, it follows that the decedent's first dying declaration was not admissible as such.

O'Connor's explanation of the conceded inconsistency in his two statements is that in his testimony before the magistrate in 1930 he was referring to the decedent's declaration given at one-thirty A.M., the hour of which he mistakenly gave as twelve-thirty A.M. If that explanation is to be accepted it creates an added inconsistency because in 1940 he testified that the questioning at one-thirty A.M. was conducted by him after the decedent had been removed to a ward bed; in 1930 he testified that he questioned the decedent at twelve-thirty A.M. while he was on an operating table in the emergency room.

Evidence was also received of a dying declaration made by the decedent to his father, the witness Pietro LaScala, who arrived at the hospital at about two A.M. and went directly to the bedside of his son. He was asked: "Q. Did you, after you first came to the bedside where your son Frank was lying, ask him what was the matter? * * * A. Yes. * * * Q. What was his answer? A. (In English) He says, `Papa, I die.' Q. Did you ask him any question after that? A. I said, `What happened?' Q. And did he answer that? A. He said, `I got shot.' Q. Did you ask him any other questions? * * * A. I said, `Who shot you?' He said, `Bobby.' Q. Now, did he say anything about another person besides a Bobby? A. (In English) He says another fellow, but I don't remember the name. Q. Did you ask him any other questions as to how he came to be shot? * * * A. He said he was at 121st Street and Lenox Avenue by a candy store, and a girl went over to call him, saying that `Somebody wanted to speak to you.' Q. What else did your son Frank tell you? A. He went outside and he recognized a Bobby, and before he spoke to him he shot him."

Among exceptions to the conventional rule which prohibits the receipt in evidence of hearsay statements is the rule which governs the admission of dying declarations. Here the general rule yields to the exception as a matter of public necessity in aid of the administration of criminal law in the punishment of the guilty and the protection of the innocent. If we look for the basis upon which rests this exception, we find it in the assumption, born of experience, that "the approach of death produces a state of mind in which the utterances of the dying person are to be taken as free of ordinary motives to mis-state. * * * The situation supplies a circumstantial probability of accuracy equivalent to that of the tests of oath and cross-examination." (See 5 Wigmore on Evidence [3d ed.], § 1438, p. 230 et seq. and collated cases.) The fact that the exception has as its basis only the assumption mentioned above, and the further fact that the declarant is unsworn and free from harassment by cross-examination, prompted the statement made on behalf of this court that "Dying declarations are dangerous, because made with no fear of prosecution for perjury and without the test of cross-examination, which is the best method known to bring out the full and exact truth." ( People v. Falletto, 202 N.Y. 494, 499.) It is for those salutary reasons that extreme caution is required of the trial court before a dying declaration is received in evidence — the fundamental requirement being that "preliminary to the admission of such evidence * * * there must be clear proof of the certainty of speedy death, and that the declarant had no hope of recovery." ( People v. Falletto, 202 N.Y. 494, 500; People v. Smith, 104 N.Y. 491, 493, 495; People v. Kraft, 148 N.Y. 631, 635; People v. DelVermo, 192 N.Y. 470, 487, 488; People v. Sarzano, 212 N.Y. 231, 234; People v. Ludkowitz, 266 N.Y. 233, 238, 243.)

The admissibility of dying declarations does not depend upon any particular form of expression. "There is no unyielding ritual of words to be spoken by the dying." ( Shepard v. United States, 290 U.S. 96, 100.) If, however, it appears that the declarant "had any expectation or hope of recovery, however slight it may have been, and though death ensued in an hour afterwards, the declarations are inadmissible." ( Commonwealth v. Roberts, 108 Mass. 296, 301; and see 1 Greenleaf on the Law of Evidence [15th ed.], § 158.) "The declaration is kept out if the setting of the occasion satisfies the judge, or in reason ought to satisfy him, that the speaker is giving expression to suspicion or conjecture, and not to known facts." ( Shepard v. United States, supra, p. 101.)

These rules — the product of long experience by the courts — make it clear how important is the requirement that a dying declaration becomes competent as evidence only in the event there is clear proof that the declarant was conscious of impending death and had abandoned all hope of living. In the record at hand the assurance that such requirement was fulfilled is shaken by a conceded conflict in the testimony given by Detective O'Connor upon whose statement alone reliance must be placed to determine whether on the night of the homicide the declaration by the decedent in the hospital emergency room at twelve-thirty A.M. was made when he was in extremis and had abandoned hope of living. Six weeks after the declaration was made O'Connor had testified under oath that at that hour and place he asked the decedent, "Do you believe you are going to die?" To that question the decedent answered, "No, not now." Nearly ten years after the declaration was made the same witness testified upon the present trial that, at the same hour and place, viz., at twelve-thirty A.M. in the hospital emergency room, he asked the decedent "did he feel that he was going to die from the injuries he received" and that the decedent then said, "Don't you see I am going out? I feel rotten. * * * I mean I am going to die."

Both of these recitals of what was said by the decedent at a given time and place cannot be true. Upon the present record the inconsistency of these two statements of what was said by the decedent is not reconciled. Doubt is thus cast upon their admissibility. However, we have reached the conclusion that the preliminary testimony was sufficient to warrant the admission in evidence of O'Connor's recital of the dying declaration made by the decedent at twelve-thirty A.M. on October 30, 1930. But the evidentiary value of the declaration itself is weakened by the fact that at one point in his statement the decedent said, "Maxie, the Weasel, and Bobby shot me." As bearing upon the accuracy of that statement we have seen that, according to undisputed evidence by the only two witnesses who were near the scene of the homicide, only one man confronted the decedent as he emerged from the store and that it was that one man — unidentified upon the trial — from whose hand the flash of gun fire was seen.

Likewise, but with even greater doubt, we conclude that the second dying declaration, which was made by the decedent to his father and was prefaced only by the decedent's statement, "Papa, I die," was properly received in evidence. Without discussing the evidence in detail, it may be said that the convincing qualities of this second declaration were also weak when considered in the setting of circumstances disclosed by the record. In that connection it will be remembered that in his statement to his father, as in his prior declaration to O'Connor, the decedent, after naming the defendant as his assailant, mentioned "another fellow but I don't remember his name."

It is thus made clear that the judgment of conviction rests upon two dying declarations which, as accusatory statements, are not only of doubtful admissibility but are neither clear nor satisfactory.

With this state of the record in mind we reach a consideration of a ruling which we regard as erroneous and which so prejudiced the defendant's rights as to require a new trial in the interest of justice. At the close of the charge to the jury counsel for the defendant made the following request: "I ask your Honor to charge that the dying declaration, when admitted into evidence, is not to be regarded as having the same value and weight as the testimony of a witness given in open court, under the sanction of an oath and under the tests and safeguards which are there provided." In response to that request the trial judge ruled: "The Court: I decline to charge in the language requested. I think I have covered that subject in my charge to the jury, and I think they understand it."

We believe the language of the request to be a precise statement of the rule relating to the weight which should be given to a dying declaration. Indeed, the request appears to have been framed in the language of VANN, J., writing for this court in People v. Falletto ( supra, p. 498). It is true that the trial judge had made brief reference in his charge to dying declarations and the care with which they should be considered. We do not think, however, that he had "covered the subject," as his ruling suggests. (See People v. Ludkowitz, supra, pp. 242, 243; People v. Kraft, 148 N.Y. 631, 635.) In any event we cannot say, in this case, where a judgment of conviction of murder in the first degree rests upon two dying declarations of doubtful admissibility and weak probative value, that the ruling last mentioned above did not prejudice the defendant's rights.

The judgment of conviction should be reversed and a new trial ordered.

FINCH, CONWAY and DESMOND, JJ., concur with LEWIS, J.; LEHMAN, Ch. J., LOUGHRAN and RIPPEY, JJ., concur in result.

Judgment of conviction reversed, etc.