The theory upon which the State sought conviction was that the defendant brought about the death of the deceased through asphyxiation, or carbon monoxide poisoning, from gas issuing from a partially opened jet of the kitchen stove. When the autopsy was performed by the experts who testified for the State, the blood had been removed from the body in the embalming process, and it was admittedly impossible to make from the tissues of the body a conclusive quantitative analysis as to the degree of saturation. The necessity of establishing the cause of death by expert evidence and the important bearing of the testimony of Dr. Forbus, an expert pathologist, that he had made a post-mortem examination of the body and had found no condition which might have caused death other than the indications of carbon monoxide poisoning are apparent on perusal of the record. This leads to a serious consideration of defendant’s objection to the testimony of Dr. Haywood M. Taylor, an expert toxicologist, who also gave his opinion that deceased came to her death from carbon monoxide poisoning, admittedly basing that opinion in substantial part upon the statement of Dr. Forbus that his examination disclosed no other cause of death. The objection is that this finding of Dr. Forbus had not been submitted to the witness in the hypothetical question addressed to him, and that the opinion was not predicated upon the assumption that the jury should find the evidence of Dr. Forbus to be true; and, further, that the opinion of one expert witness based upon the opinion of another is incompetent as evidence.
Dr. Forbus and Dr. Taylor had performed an autopsy and examined certain organs and tissues of the body some thirty days after it had been embalmed and interred, and both testified that the characteristic cherry red color of the tissues indicated carbon monoxide poisoning, but the finding that there was no other cause of death was peculiar to the examination and testimony of Dr. Forbus, Dr. Taylor not having made any examination or finding in that respect. The hypothetical question addressed to Dr. Taylor — which elicited his opinion as to the cause of death — made no mention of this finding or statement, directly or indirectly, or of the purported fact, or condition, thus put in evidence; and it did not transpire that Dr. Taylor had based his opinion on a state of facts not so presented to him until brought out on cross-examination. For convenience, we repeat these interrogatories and answers:
*254“Q. Tour opinion, in the absence of a sufficient quantity of blood to determine the degree of saturation, is based upon color and the presence of the poison in the tissues ?
“A. Upon that and one other factor; the absence of any other causes of death and the presence of carbon monoxide in the tissues and the cherry red color in the skin and organs.
“Q. You say in the absence of any other cause, you refer to the pathological conclusions arrived at by Dr. Eorbus ?
“A. Ido.”
• The defendant then moved the court to strike out the testimony of Dr. Taylor — or rather, that part of it in which he gave his opinion as to the cause of death — and the motion was declined. Exception had been made previously to the admission of the evidence.
There are two avenues through which expert opinion evidence may be presented to the jury: (a) Through testimony of the witness based on his personal knowledge or observation; and (b) through testimony of the witness based on a hypothetical question addressed to him, in which the pertinent facts are assumed to be true, or rather, assumed to be so found by the jury. That an expert witness may base his opinion partly on facts of his own observation and partly on factual (as opposed to opinion) evidence of other witnesses, hypothetically presented, is, of course, within the rule.
It is clear that if in his testimony Dr. Taylor had reference to information concerning the Forbus finding obtained extrajudicially — that is, in any other manner than from the evidence given in court — the testimony is objectionable as based on a hearsay statement. If it had reference to the testimony of Dr. Eorbus which immediately preceded his own, it is equally objectionable because it was not hypothetically presented — that is, was not predicated on an assumption that the jury should find the purported facts in the Eorbus statement to be true. Dempster v. Fite, 203 N. C., 697, 167 S. E., 33; Summerlin v. R. R., 133 N. C., 551, 45 S. E., 898; Martin v. Hanes Co. 189 N. C., 644, 646, 127 S. E., 688; Tates v. Chair Co., 211 N. C., 200, 189 S. E., 500.
Our practice and procedure does not permit an expert witness to sit in, overhear the evidence and give the jury his opinion or conclusions thereupon, without regard to what might be the attitude of the jury toward the credibility and weight of the evidence with which the witness is dealing and upon which his opinion is based. The assumption of its truth in the mind of the witness, however self-satisfying, cannot be substituted for the finding of the jury, and necessarily invades the province of the jury. It invades the province of the jury not because it gives an opinion as to the ultimate facts to be found by the jury, which is sometimes permissible, but because it permits the witness to determine for himself *255the weight and credibility of the evidence of these facts, which ought always to be left to the jury. Even those jurisdictions which permit expert witnesses who have overheard the evidence to give an opinion upon it without a formal summation and interrogatory are strict in observing two conditions as requisite to competency: First, it must affirmatively appear that all the evidence pertinent to the inquiry has been heard by the witness; and, second, the opinion itself must be hypothetical — i.e., based on the assumption that the jury shall find the evidence upon which the opinion is based to be true, (a) State v. Medlicott, 9 Kan., 257; Kempsey v. McGinnis, 21 Mich., 123; Sebrell v. Barrows, 36 W. Va., 212, 14 S. E., 996; Howland v. Oakland Consol. St. R. Co., 115 Cal., 487, 47 Pac., 255; Williams v. State, 37 Tex. Crim. Rep., 348, 39 S. W., 687. (b) Yardley v. Cuthbertson, 108 Pa., 395, 1 Atl., 765; Owings v. Dayhoff, 159 Md., 403, 151 Atl., 240; Scheller v. Schindel, 153 Md., 547, 138 Atl., 415; Ingles v. People, 90 Colo., 51, 6 Pac. (2d), 455. Neither of these conditions obtained in the case under review.
In many jurisdictions — and we find the rule expressed in considered opinions — a witness is not allowed to give an opinion based on the testimony of another witness where that testimony is not incorporated in a hypothetical question. The content of the question controls the range of the answer, and thus keeps opinion evidence within its proper function. Typical of these cases, which are, of course, too numerous for exhaustive mention, we may cite and quote from the following:
In Craig v. Noblesville & S. C. Gravel Boad Co., 98 Ind., 109, 82 A. L. R., 1487, the rule is stated: “The only safe rule in allowing an expert witness to give an opinion, based upon the testimony of others, is to require the assumed facts, upon which an opinion is desired, to be stated hypothetically; then the jury can judge whether the assumed facts, upon which the opinion is based, have been proved, and weigh the opinion as applicable to them.”
In Ditton v. Hart, 175 Ind., 181, 93 N. E., 961, the Court said: “It is settled that an expert witness will not be allowed to give his opinion upon his recollection and construction of the evidence in the case: He must base his opinion upon his own testimony or upon facts assumed to have been proven, which facts must be- given to him as the foundation upon which to base his opinion.”
In Guetig v. State, 66 Ind., 94, 32 Am. Rep., 99, it is said: “An-expert cannot give his opinion upon evidence; it must be done upon admitted, proved or assumed facts.”
In Marx v. Ontario Beach Hotel & Amusement Co., 211 N. Y., 33, 105 N. E., 97, we find: “An expert witness may not draw inferences or state conclusions from the testimony of other witnesses. His opinion *256must be based on facts wbicb are stated in tbe form of a hypothetical question.”
In Brace v. Bath & H. R. Co., 154 N. Y. Supp., 931, it appeared on the cross-examination of a medical expert that his conclusion was not based wholly on the hypothesis contained in the question propounded, but partly on the evidence given in his presence, on the trial. The Court held that it was reversible error for the trial court to refuse to strike out the answer. See, also, Ayres v. Water Comrs., 22 Hun. (N. Y.), 297.
In Richmond v. Wood, 109 Va., 75, 63 S. E., 449, the Court said: “Before the opinion of an expert, when it is based on facts which he has not himself testified to, can be admitted, he must fully understand the facts already proved, and his testimony must come in response to a hypothetical question which embodies the evidence.” To the same effect is Kerr v. Lunsford, 31 W. Va., 659, 8 S. E., 493.
In Bunagan v. Appalachian Bower Co., 33 Fed. (2d), 876, 68 A. L. R., 1393, the rule is thus stated: “It is well settled that in the examination of experts as to matters which they have not themselves observed, testimony as to their opinions should be based on hypothetical statements propounded in proper questions, not on the testimony of other witnesses whom they have heard testify.”
See 32 C. J. S., p. 347, sec. 551.
Our own decisions adopt this view.
In Summerlin v. R. R., supra, p. 554: “There is nothing better settled than that a witness can ordinarily speak only of facts within his own knowledge, unless he is an expert, having special scientific knowledge, in which case he may give his opinion, but only upon the facts as they may be found by the jury. It is usual, therefore, to formulate what is called a hypothetical question, which should contain a recital of such facts as may have been testified to by the other witnesses.”
In Dempster v. File, supra, p. 708, the rule as stated in N. C. Handbook (Lockhart) 2d Ed., p. 240, is adopted. (See. 204): “An expert may express an opinion, but he must base his opinion upon facts within his own knowledge, or upon the hypothesis of the finding by the jury of certain facts recited in the hypothetical question.” This is the rule established in Martin v. Hanes Co., supra: “While a medical expert may not express an opinion as to a controverted fact, he may, upon the assumption that the jury shall find certain facts to be as recited in a hypothetical question, express his scientific opinion as to the probable effect of such facts or conditions.” See Annotations, 82 A. L. R., 1468, et seq.
We think the matter is too well settled in this jurisdiction to need further citation.
*257Hitherto, for convenience of discussion, we have treated tbe statement of Dr. Forbus — that there were no apparent conditions other than the indications of carbon monoxide poisoning to which death might be attributed — as a simple statement of fact, but it must be regarded as more. Since it involved the application of scientific, technical and medical knowledge in examining the body and recognizing the presence or absence of pathological conditions therein upon which the statement must necessarily be based, in a field entirely beyond the knowledge of a non-expert witness, it must be classified as opinion evidence. The opinion of Dr. Taylor based thereupon is, therefore, objectionable, as it is uniformly held that the opinion of one expert based upon that of another is incompetent and inadmissible as-evidence. McComas v. Wiley, 134 Md., 572, 108 Atl., 196; People v. Bowen, 165 Mich., 231, 130 N. W., 706; State v. King, 158 S. C., 251, 155 S. E., 409. “It is generally agreed that the opinion of an expert, however qualified to speak, cannot be predicated either in whole or in part upon the opinions, inferences and conclusions of others, whether expert or lay witnesses.” 20 Am. Jur., 665, sec. 791.
It is not thought necessary to advert to other objections and exceptions noted in defendant’s brief.
For the error noted, the defendant is granted a new trial.
New trial.