Magruder, Special J.,
delivered the following opinion, favoring a reversal of the judgment appealed from:
At the March term, 1897, of the circuit court of Kemper county, the appellant, Dr. W. H. Lipscomb, was tried, convicted, and sentenced to be hanged for the murder of Charles P. Stewart, from which judgment and sentence he prosecutes this appeal. Stewart died on the night of January 21, 1897. Just before going to bed for the night, he bathed his feet and took a capsule of medicine and laid down upon his bed. In a few minutes thereafter he became ill, and in twenty or thirty minutes died in convulsions, manifesting the symptoms usual in cases of strychnine poisoning. His wife, who was in the room, a Mr. Duran, and also a negro man, who came in, made endeavors at ministration for his relief. Duran went out to send to Scooba, a village about three miles away, for a physician. Stewart called upon the old negro to pray for him. In the interval between the third and last convulsions, and just before he died, he said to his wife: “lam going to die. I have been dead. The good Lord has sent me back to tell you that Dr. Lipscomb has killed me, has poisoned me with a capsule he gave me to-night; that Guy Jack had insured his life, and had hired Dr. Lipscomb to kill [575]*575him.” Only Mrs. Stewart and the negro were present. The statement was voluntarily made by the deceased, and without suggestion of any kind. He was a young man, who appeared to be strong and vigorous. On the day of his death he went to the village of Scooba on some business, and returned to his home about dark, fed his horse, and ate heartily at supper, apparently in good health and cheerful spirits. But it seems that he was being treated by Dr. Lipscomb for some disorder, and while in Scooba on that day Dr. Lipscomb prescribed for him. The prescription called for three capsules, to be compounded of quinine, antikamia, and strychnine, each capsule to contain one sixtieth of a grain of strychnine. After the prescription was written by Dr. Lipscomb at his office, which was at the drug store of Dr. Mohler, who was a professional partner of Lipscomb, it was taken in person by Lipscomb to the dl’ug store of Dr. Morney, about a square away, for the alleged reason that Dr. Mohler had no antikamia, Lipscomb telling Stewart he would get it for him. When the capsules had been prepared, manufactured tablets of strychnine of one-fiftieth of a grain each being used, they were placed in a box marked with directions, “Take one at night,” handed by the druggist to Dr. Lipscomb, who took the box and went out upon the street, found Stewart, and gave him the box, and instructed him to bathe his feet and to take one at bedtime. Stewart’s father was present when the box was handed to his son by Lipscomb, and, in a little while after parting from Dr. Lipscomb, Stewart, who had kept the box in his hand, opened it, and making some comment as to the size of the capsule, showed it to his father, who testified that there was but one capsule in the box. An analysis of the stomach of Stewart revealed the presence of one and one-half grains of pure strychnine, which had not been absorbed.
The foregoing is a brief statement of such of the facts as are deemed necessary to our understanding of the questions of law presented by this appeal. On the trial of the case, the wife of the deceased testified to the declaration above mentioned, made [576]*576by her husband before his death. This testimony was admitted as the dying declaration of the deceased, over the objection of the defendant, Which ruling is assigned for error.
An autopsy was held by several physicians, including the appellant. The result of the autopsy was reduced to writing. It described the appearance and condition of the body and the various organs. It was written out by one of the physicians, and read over, section by section, in the presence of all, and, as he testified, it was his understanding and recollection that all, including Dr. Lipscomb, assented to it, and he thereupon signed his own and the names of the other physicians. This paper was admitted in evidence, and read to the jury, over defendant’s objection. When the jury retired to consider their verdict, the defendant requested that the paper be delivered to the jury, which the court refused. Exceptions were also talqen to the action of the court in granting or refusing certain instructions. These rulings of the court, among others, are now, on this appeal, assigned for error. We will first consider the instructions.
The second instruction given for the state is an effort to define a “ reasonable doubt. ” It is as follows: “ The court charges the jury that by a reasonable doubt is meant, not a mere speculative doubt or vague conjecture, mere supposition or hypothesis, but such a doubt as reasonably arises out of the testimony in this case — a doubt for which a reason can be given, in view Of the testimony or want of satisfactory testimony.” This instruction is not erroneous, for it does not embody an incorrect definition. It is no definition at all. It is mere tautology, stated with awkward circumlocution. The terms of the expression, “reasonable doubt,” import the most exact idea of its meaning, and are incapable of simplification, and there is no equivalent in phrase more easily understood. All such endeavor is futile and foredoomed, the usual result being a maze of casuistry, tending to confuse rather than to enlighten, often evolving incorrect propositions, as shown in the recent cases of Powers v. State, 74 Miss., 779; Hammon v. State, 74 Miss., [577]*577214; Williams v. State, 73 Miss., 822; Burt v. State, 72 Miss., 408; and Brown v. State, 72 Miss., 95. In all of these cases, besides the specific errors of the particular instruction considered, the practice of attempting such definition at all is criticized and deprecated, if not condemned.
The seventh instruction for the state contains a hypothetical statement of facts, of which the jury are told, "if they believe, ” constitute guilt, omitting the word, ‘ ‘ beyond a reasonable doubt,” and is consequently erroneous; but the error seems to be avoided by the subsequent instruction, which is nearly identical, and by others.
The ninth instruction for the state is in these words: “ Circumstantial evidence has been received, in every age of common law as competent evidence, and it may rise so high in the scale of belief as to generate full conviction. When, after due caution, this result is reached, the law authorizes the jury to act on it. ” “ Full conviction ’ ’ is not the criterion of the degree of proof necessary to a conviction. It is a loose phrase. It has no distinct legal import, and is without accuracy to the common understanding. It is vague, indefinite, and inexact. It may be the equivalent of sincere or conscientious belief. It may mean that full conviction when the facts proven satisfy the judgment as to the truth of the charge. There is but one rule and one law in this state as to the measure and sufficiency of proof which will warrant conviction. It is that the evidence must engender a certainty of belief beyond a reasonable doubt. This rule has prevailed without abatement, not only in the ages of common law, but it embodies an everlasting human right, coeval with all society. It is not enough that “ the jury should be satisfied from the evidence, as fair, reasonable or conscientious men, of the guilt of the accused” (Powers v. State, supra), or ‘ ‘ that they conscientiously believe him guilty." Burt v. State, and Brown v. State, supra; Hammond v. State, 74 Miss., 214.
In Williams v. State, 73 Miss., 822, the jury were instructed that ‘ ‘ if, after a careful consideration of all the evidence in the [578]*578case, you can say and feel that you have an abiding conviction of the guilt of the defendant, and are fully satisfied of the truth of the charge, then you are satisfied beyond a reasonable doubt, and your verdict should be guilty;” as to -which the court in that case says: “The second instruction for the state is erroneous in attempting to define ‘ reasonable doubt. ’ . ' . The concluding part of the charge expressly defines reasonable doubt by telling the jury that ‘ abiding conviction of the guilt of the defendant, or full satisfaction of his guilt, is the equivalent of belief beyond a reasonable doubt. ’ This is another of the many vain attempts to compute "that which is not number, and measure that which is not space.” It was held one of the errors for which that case was reversed. The hypotheses of that instruction, it will be observed, are, however, stated con-junctively. The statement is, “If you have an ‘ abiding conviction ’ of the guilt. .’ . . /c and ’ are fully satisfied of the truth of the charge.” “Full conviction” cannot certainly mean more than £ £ fully satisfied, ’ ’ but if, conj unctively, £ £ an abiding conviction of guilt ’ ’ and £ £ full satisfaction of the truth of the charge” be erroneous, as not equivalent to a belief beyond a reasonable doubt, it must follow that £ £ full conviction ’ is in less degree such equivalent. It is error.
The fifteenth instruction asked by the defendant, and refused, is as follows: "The court instructs the jury that the dying declarations of the deceased, made to his wife, are not entitled to the same credit and force as if the deceased was still alive and testifying in the presence of the jury, under oath; that it is a species of hearsay evidence, and is intrinsically weaker than if the declarant was present and subject to cross-examination; and the jury alone is the judge of its weight and force.” We all concur that if it be the true purport of the instruction that it is intended to point out, by way of precaution, the inherent qualities, which, by law, pertain to all dying declarations, and be held to refer to the source, rather than the matter, of the testimony, it would, in that view, not be upon the weight of evi[579]*579dence, within the meaning of the statute. In cases of perjury, seduction, and the like, requiring corroborative evidence as to certain testimony, and in cases involving the testimony of accomplices and evidence of admissions made by a party against his interest, and the like, it is entirely proper to point out to the jury the circumstances affecting the source or character of the evidence, which, according to settled rules, operate to its disparagement, being careful to leave to the jury the untrammeled right to consider the testimony, and give it what weight they may deem it worthy. Abstractly, it is true, as a matter of law, that a dying declaration is a species of hearsay testimony, and, of itself, is not entitled to the same force as if the witness was living and testifying, and subject to a cross-examination. Lambeth v. State, 23 Miss., 322. The question presented by this instruction involves a consideration of the law applicable to the admission of dying declarations, and the reasons upon which they are founded, and which determine the character of such testimony from a legal standpoint. Dying-declarations constitute the only exception to the constitutional right of the accused to be confronted with the witnesses against him, and be afforded the right to cross-examine them. In all trials and on all issues the cross-examination is the most effective means of eliciting and ascertaining the truth. While the solemnity under which they are usually made is deemed, in some sense, an equivalent for the sanctity of an oath, yet their admissibility rests upon the grounds of necessity and public policy, and upon the presumption that, in the absence of other proof, crimes might go unpunished. The rules which govern the admission of such testimony are familiar and rudimental: (1) They must be made under the realization and solemn sense of impending death, when the motive for falsehood may be presumed to be lost in the despair of life; (2) they must be the utterance of a sane mind; (3) they are restricted to the act of killing, and the circumstances immediately attending it and forming a part of the res gestae; (4) no declaration, or any part [580]*580of it, is admissible, unless competent and relevant, if made by a living witness; (5) that great caution should be observed in the admission of such testimony, and the rules which restrict it be carefully guarded.
The authorities abound with discussion of the reasons and considerations upon which these rules are founded, looking to the conservation of truth and that justice might prevail. The circumstance that the declaration is hearsay, and is without the essential element of cross-examination, stands, facile princeps, the most important of these reasons, and incidentally and necessarily involves other considerations. 1 Greenl. Ev., sec. 162; 1 Phil. Ev., 300; People v. Sanchez, 24 Cal., 17. “ The admission of dying declarations as evidence, being in derogation of the general rule which subjects the testimony of witnesses to the two important tests of truth, an oath and a cross-examination, it is obvious that such evidence should be admitted only upon the grounds of necessity and public policy, and should be restricted to the act of killing and res gestee. ’ ’ Leiber v. Com., 9 Bush, 11; 2 Starkie, Ev., p. 366; Bell v. State, 72 Miss., 513.
There are other considerations which have been dwelt upon by law writers and judges. Statements made under the shadow of approaching death may come with the infirmity of inattention, when the mind is diverted to the thoughts of the future; the vigor of the mind may be impaired; facts may be but partially stated; inferences and opinions may be stated as facts; the passions of anger and revenge may linger, after all hope of life is fled, and affect the truth of the statement. It must come as the memory of those who heard it, subject to all the uncertainties of a-correct understanding of the speech as made, and of a correct reproduction by the memory of what was truly said. Mr. Roscoe says: ‘ ‘ Such considerations show the necessity of caution in receiving impressions from accounts given by persons in a dying state, especially when it is considered that they cannot be subjected to the power of cross-examination, a power quite as [581]*581necessary for securing the truth as the 'religious obligation of an oath can be.” Rose. Cr. Ev., p. 35. The foregoing principles are repeated, iterum iterumque, in varying phrase, in numerous authorities and cases, and are the well-settled law of this state. Bell v. State, supra; Lambeth v. State, 23 Miss., 350; Nelms v. State, 13 Smed. &M., 501; Brown v. State, 32 Miss., 433; Merrillv. State, 58 Miss., 66; Montgomery v. State, 80 Ind., 338; Moore v. State, 12 Ala., 764; Binns v. State, 46 Ind., 311. In State v. Vansant, 80 Mo., 78, it is said: “ Besides, such declarations are afflicted with the common infirmity which attaches to all oral statements or verbal admissions reduced to writing or repeated by another, and are liable to be colored or deflected by the medium through which they are transmitted to the jury.” And in Lambeth’s case, supra, Mr. Justice Yerger said that a dying declaration was not entitled to the same weight and force as if delivered by a living witness.
In Brown v. State, 32 Miss., 442, it is said, after commenting upon the nature of dying declarations, that "it is, therefore, the dictates of reason and common sense that declarations of this character, in all cases, and under any circumstances, should be admitted with caution and weighed by the jury with the greatest deliberation.” 1 Greenl. on Ev., sec. 162. We all concur that it is clear that as to dying declarations it would not be objectionable if the jury be charged that while they are the sole judges of the weight and effect to be given to a dying declaration, and that it is to be determined like any other evidence, in the light of all the evidence of the case, and to caution them, in determining its effect, that they should weigh it with great deliberation and care, and take into consideration the circumstances of its being hearsay; that it is the statement of one not subject to cross-examination, or such other relevant circumstances in that regard as may exist in any given case; and that it is the duty of the court to lay before the jury, by precautionary instructions, when asked, the inherent elements of weakness which the law recognizes in certain classes of evidence, but in [582]*582such form as not to invade the province of the jury. The majority of the court hold that this instruction is not upon the weight of evidence; that its true purport is cautionary, and refers rather to the source than to the effect of the testimony. For myself, I do not concur in this view. It embodies argumentative statements of the law abstractly correct, but so stated as to bear upon the weight of evidence, and, in my opinion, was properly refused. In Lewis v. Christie, 99 Ind., 377, an instruction which followed the text of the most authoritative writer in the language on evidence (Greenleaf) was condemned, and it was said that argumentative statements of the law, though correct; may not always be an accurate rule of guidance to a jury. 1 Thomp. on Trials, sec. 640. We find no error in the rulings of the court as to the other instructions.
We come next to consider the ruling of the court upon the admission of the dying declaration, in respect to which several propositions are contended for by the appellant, which may be formulated as follows: (1) The proof did not show that at the time of making the declaration Stewart realized that his days were numbered, and the solemn hour of his death was at hand; (2) that the statement itself, in its peculiar character, bore evidence of mental incapacity, of unbalanced reason, and that it was a mere hallucination of a disordered mind; (3) that the declaration was not the statement of a fact, as of his knowledge, but an opinion, a mere belief, a conclusion deduced and inferred from other collateral facts.
All of the above propositions involve questions of fact, to be ascertained and passed upon by the trial judge upon a preliminary investigation, and in that sense are commonly designated questions of law. 1 Rosc. Cr. Ev. (8th ed.), 61; 1 Greenl. Ev., 219; Whart. Cr. Ev., 689; 1 Phil. Ev., 3, 7, 573; Simmons v. State, 61 Miss., 243; State v. Burns, 33 Mo., 483; Kilgore v. State, 74 Ala., 1; Owens v. State, 59 Miss., 547; Ellis v. State, 65 Miss., 48; Bell v. State, 72 Miss., 510.
It is therefore pertinent to inquire what degree of proof in [583]*583law is required as to the facts which constitute what is ordinarily termed the foundation for the admission of such testimony — whether they should be established to the satisfaction of the court, or to that higher degree of certainty which excludes all reasonable doubt — and to what degree of certainty must it appear to be competent as the statement of a fact; and, second, what rules should prevail in determining whether it is an opinion or a statement of a fact — whether the literal form of a statement is to control, or whether its true character is to be gathered from the statement, considered in connection with the surrounding circumstances.
Taking up the first proposition, we find that the authorities in this state establish the rule that in a preliminary investigation by the court, for the ascertainment of facts precedent and necessary to the competency of proposed evidence, the degree of proof should be such as to exclude all reasonable doubt as to the facts of such foundation; and this is certainly true when the competency of any testimony rests upon collateral facts, independent of the main fact proposed to be proved; and in this case the rule applies to the preliminary questions (1) whether Stewart realized he was in extremis; (2) whether he was at the time sane and rational. Bell v. State, 72 Miss., 510; Owens v. State, 59 Miss., 549; Simmons v. State, 61 Miss., 257; Holly v. State, 55 Miss., 430; Ellis v. State, 65 Miss., 48.
But the competency of a statement, whether one of fact or opinion, is a question of judicial interpretation as to its meaning, as indicated by its terms, viewed in the light of surrounding circumstances. The interpretation must rest upon the ascertainment as to what was the true purport of the declaration. It is the conclusion to be reached as to this fact which must control. If it bo a reasonable construction that a statement is one of fact, it ought to be admitted.
Tested by the foregoing principles, we are of opinion that the judgment of the court below was right in holding that the foundation was sufficiently laid (1) as to the fact that Stewart [584]*584was in extremis when the statement was made, and that he realized that he was beyond the hope of recovery, and (2) that at the time he was not insane or delirious, but spoke with discernment, reason and intelligence. We think the evidence established both of these propositions beyond reasonable doubt. He was in the throes of death when the statement was made, and died shortly thereafter. He called upon a bystander to pray for him, and declared that he was going to die. As to his mental condition, there is nothing in the evidence to justify a doubt that he was rational, except the suddenness, violence and brevity of the attack — a condition not inconsistent with a sound mind. It is argued that the statement itself, in form and substance, res ipsa loguitw, is the utterance of a mind diseased — an illusion of a disordered imagination. We do not so view it. The words of the speech may be unusual, yet they are words of discernment and reason.
This brings us to the question as to the admissibility of the declaration itself as evidence. If, according to the rule announced, it is a reasonable and probable theory that the declaration, according to its just import, taken in connection with the surrounding and relevant circumstances, was a statement of a fact within the knowledge and observation of Stewart, then it is competent, and its credibility and value as evidence was properly submitted to the jury; but if such an interpretation be unreasonable, or involves an improbable theory as to any assumed fact, it was incompetent, and the ruling of the court erroneous. At the threshold of the question we are met with the contention that so much of the statement as refers to Huy Jack having insurance upon Stewart’s life, and. that he hired Lipscomb to kill him, are facts which are not of the res gestee, as immediately connected with the killing, and that part — “that he had been dead, and the Lord had sent him back to tell ” — is irrelevant, and consequently that the court erred in admitting the whole declaration which embraced these statements. That they were not a part of the res gestee, and are irrelevant, goes [585]*585without saying, and, as such, were inadmissible. But, inasmuch as no specific objection was made to these particular parts of the declaration, but was to the whole statement, it is contended, on the other hand, that such objection,, according to familiar rules, will not avail if any part of the declaration was competent or relevant. Inasmuch as the decisions in this state are not apparently uniform as to the practice in applying the 'rule, it is deemed proper that a statement of the law as to this very familiar rule, as we all understand it, be made.
It is well settled, by numerous adj udications, that a general objection raises no issue, except it is as to whether the evidence would, under any circumstances or for any purpose, be admitted, and that a specific objection raises no other issue than the particular one tendered. If, under any circumstances of the case, testimony be admissible, a general objection would be properly overruled; and, conversely, it would be error to sustain a general objection if, under any view of the case, the evidence might be admissible. This rule applies with particular force in cases where the objection may have a twofold aspect, and may be put upon different grounds. It might go to the competency of the witness, to the mode of proof, to the insufficiency of the foundation, or to the relevancy or competency of the matter of the testimony. Testimony may be material and relevant, and still incompetent by reason of the method of proof or lack of some precedent predicate required by law to be first established, and vice versa. In Heard v. State, 59 Miss., 546, it is said: “When an objection is made to evidence which in its nature is such as may be obviated, it must be specific, so as to allow the party offering it an opportunity to supply its place if the objection is sustained, and, where this is not done, it will not be noticed m the appellate court." Morris v. Henderson, 37 Miss., 492; Brown v. State, 72 Miss., 95. The reason of the rule is said to be twofold: (1) To enable the trial judge to understand the precise question upon which he has to rule, that he should not be required to search for objections [586]*586which counsel do not discover, or conceal; (2) to afford the opposite party an opportunity to obviate it if well taken. Decell v. Lewenthal, 57 Miss., 331; Heard v. State, 59 Miss., 545; 1 Thomp. Trials, sec. 693; Wesling v. Noonan, 31 Miss., 599; Morris v. Henderson, 37 Miss., 492; Brown v. State, 72 Miss., 95; Mills Co. v. Smith, 69 Miss., 299; Reynolds v. State, 68 Ala., 502; Archibald v. State, 122 Ind., 122 (23 N. E., 758); Dozier v. Jerman, 30 Mo., 216; Brown v. Weightman (Mich.), 29 N. W., 98; Moore v. Bank, 13 Pet., 302; Stone v. Oil Co., 41 Ill., 85; Gilbert v. Thompson, 14 Minn., 544 (Gil., 414); Rush v. French, 1 Ariz., 99 (25 Pac., 816).
A large number of cases to the same effect are collected in 1 Thomp. Trials, sec. 693. The rule also involves the particular proposition that a general objection is properly overruled to testimony a part of which is admissible and part not. But this particular rule has not been uniformly followed in this state. In Merrill v. State, 58 Miss., 66, and in Field v. State, 57 Miss., 474, statements were admitted in evidence parts of which were admissible and part were not. In both cases the ruling of the court was held to be erroneous, notwithstanding the objection was general. It is true the point was not made in either case invoking the rule that the objection was general. It has been well observed that a court might take cognizance of fatal error in proper cases to avoid injustice, notwithstanding, by inadvertence or accident, the objections were not made specific or properly framed. In view of the fact that the rule has not been strictly enforced in this state in all cases, by reason whereof some doubt may have prevailed as to the practice, it might be questioned whether it ought to be enforced in this case, especially as the record shows that the objection was argued by both sides, and presumably, of course, upon specific grounds. We do decide that Field’s case and Merrill’s case are not to be held as precedents for a different rule than that announced above.
Again, as preliminary to the question of the admissibility of [587]*587the statement, it is important to first ascertain, as a rule of interpretation, whether the form of the statement — the literal sense of the words used — is to govern, for, if this controls the interpretation, there is no room for construction, for leaving out the figurative reference to the Lord, as is proper to do, the words of the statement, in their ordinary significance, literally interpreted, fairly import knowledge, and not opinion. We appreciate the force of the argument so ably made by counsel, for the state, which is supported by some precedents, that the face of the statement alone must be looked to for its purport in that respect. The proposition is not founded in sound reason, and is contrary to the great weight of authority. The ends of justice are to be attained only by ascertaining, as far as practicable, the very truth. The truth as to the purport of any speech — its true meaning — is to be gathered, not from the literal form of the phrase and irrespective of the nature of the'act narrated. The situation of the speaker; what of mental impulse, motive, inducement or emotion was evinced, and all the surrounding circumstances; the words used, their literal signification, as well as that in which, in common parlance, they are often employed; and the character and nature of the fact stated, are all to be considered in determining the truth as to the meaning of any utterance. They aro ‘ ‘ the light unto the path ’ ’ which leads to a true solution of every such problem.
It is of common experience that matters of information, and of belief or opinion, are generally expressed in a form importing personal knowledge. Fixed convictions, regardless of' the mental processes involved, take the form of dogmatic assertion, and suspicion is couched in terms of direct accusation. Even the cautious and learned do not discriminate, and give emphasis to their settled conclusions by declaring, "I know. ’ ’ In this case, an expert physician, who had heard all the evidence pertaining to the symptoms of Stewart’s malady, and the evidence of the presence of strychnine disclosed by analysis, when asked if he could state from these facts the cause of death, replied [588]*588“ that the evidence makes me believe — I know; I believe I can say I know — that he died from strychnine poisoning. ” “1 know that my Redeemer liveth ’ ’ is the exclamation of an exalted faith, disclaiming all equivocation, and yet the thought traverses the realms of the unknown and the unknowable. Credo ut intelligam. The question is not one of the probability of the truth of the fact stated, or of the opportunities for knowledge, but what is the statement itself with reference to the fact? Is it an opinion or not? And the true rule, we have already announced, is that it should be given that construction in this regard which, under the circumstances, appears most reasonable. State v. Williams, 67 N. C., 12; Walker v. State, 39 Ark., 221; Jones v. State, 52 Ark., 345 (12 S. W., 704); Bell v. State, 72 Miss., 514; Nelms v. State, 13 Smed. & M., 500.
Cases could be multiplied all to like effect, that a statement is not to be held as one of fact because of the literal sense of its words, but is to be interpreted in the light of all the attendant circumstances. Applying these rules of interpretation to the dying declaration in this case, the majority of the court hold that a part of the declaration, to wit: that “Dr. Lipscomb has killed me, has poisoned me with a capsule he gave me to-night, ” is separable from the other parts of the statement, and is competent as a statement of fact. The other parts — the reference to “ the Lord sending him back to tell,” etc., “ that Guy Jack had his life insured, and hired Lipscomb to kill him ” — are, of course, inadmissible. Judges Whitfield and Thompson hold that no error can be predicated on the admission of the incompetent matter under a general objection. I dissent from the view that any part of the declaration was admissible. It seems to me that when the declaration, “ Dr. Lipscomb has killed me; he has poisoned me with a capsule he gave me to-night, ” is considered, not according to its literal terms, but in the light of all the circumstances and in connection with all parts of the statement and the nature of the act alleged and the mental processes necessa[589]*589rily involved, the conclusion is inevitable that the statement was not one of knowledge, but the expression of an opinion, a mere inference, a deduction, a conclusion. The very nature of the thing stated implies a process of reasoning by induction. Facts which the law recognizes as of knowledge must, be perceived through the senses. Cause and efficient agencies may be perceived and known as well as effect, but an agency or cause which is known only by subsequent effect, is not known in truth, but deduced, and the particular agency can only be inferred. Stewart knew he took a capsule. He did not know it was poison. He afterwards knew and felt certain sensations. They were subsequent effects. That they were the effects of poison he could not know, but only infer. He could not know any more, nor as much, as an expert physician, had one been present observing his condition, and what, in such case, would be observed by the physician and felt by Stewart, would be but symptoms to both, indications evidentiary in character, but which, from the nature of things, could only be the basis as facts from which other facts could be concluded. The facts known to Stewart pointed with unerring certainty to the truth. — that he was poisoned; yet, the truth still remains only an inference of fact. If he knew the capsule was poison, he knew it when he took it— an incredible theory. If he became convinced of it afterward, as well he may have done, then it was a deduction, logical and conclusive it may have been, but still a deduction, inferred from his symptoms, the taking of the capsule and the insurance upon his life. His assertion that Jack held insurance upon his life, and hired Lipscomb to kill him, though illegal evidence, throws a flood of light upon his meaning. It lays bear the processes of his mind. The capsule, the insurance and his dying condition, revealed to him to a moral certainty the motive and agencies of the foul murder, but the revelation came only through a process of reasoning by inference, and, at last, only as an opinion.
It is urged in argument that it is a “ fact ’ ’ that Stewart [590]*590was poisoned, and as the evidence tends to show it was administered by Lipscomb, that it is also a “fact” that Lipscomb poisoned him, and that the declaration, therefore, states them as ‘ ‘ facts. ’ ’
The argument is in a circle. Of course a ‘ ‘ fact ’ ’ may be stated, but not in the sense that the statement makes the ‘ ‘ fact, ” for the fact exists independent of the statement, but in the sense that it asserts knowledge that a thing was done or an act come to pass. The question is. not whether the matters of the statement are matters of fact, of things or acts done, but whether the statement made in reference thereto is an opinion concerning the facts or an assertion of personal knowledge of the facts. A “statement of a fact,” as distinguished from hearsay or opinion evidence, means the testimony by a witness of his recollection of things observed and perceived by him. It is that knowledge which is derived through impressions made upon the senses by external objects and through subjective sensations. Without reference to the metaphysical theories of the processes of sensation, experience, acquired perception, intuitions, reflection, ideation, and the like, it is enough for the practical purposes of the law that the knowledge of which a .witness may testify is the ordinary perception and understanding of things seen or heard, or otherwise perceived through the senses, or subjectively.experienced through sensation, and whatever other facts these import are inferences.
The argument based upon the illustration that when one is seen to fire a pistol toward another, and that other feels the shock and pain of the wound, and in such case knows that he is shot, though not seeing the flight of the missile, is without analogy, and its fallacy is apparent. For in that case the very agency or cause of the hurt is seen. The act of the shooting and design is seen before and simultaneously with the effect, and the instrument used, according to universal experience and common knowledge, is a deadly weapon, and when fired produces a wound instantaneously and simultaneously with the shot, [591]*591and a wound of a character as unique and universally known as are the outlines of a horse, by which we distinguish him from other animals.
Whenever poison is taken, and it be not known at the time, it cannot be known in a legal sense at all. It may be demonstrated as a certainty, and even that cannot be done absolutely, if the symptoms felt by the victim or seen by others be relied on. This is self-evident. It may be demonstrated beyond a reasonable doubt, but then only by inductive reasoning. If, besides the symptoms, an analysis is made, then the demonstration may become absolutely certain, but no one will say that the disclosure afforded by an analysis imparts knowledge of the cause of the death, as distinguished from conclusion.
It is argued that the nearness in point of time between the taking and the effect of poison, enables one to know that he is poisoned. The fact is, he knows nothing but the effect, and the cause, whether near or distant, stands in mental contemplation apart from it, and cannot be perceived by the senses, but must be reached by the reason. And this argument would concede that, if longer time intervened, it might not be ‘ ‘ knowledge, ’ ’ which demonstrates that it is only opinion in both cases, the difference being in the degree of certainty. Cause and effect are concepts of the relation of things. Both may be perceived and known, but where only one is known, the other can only be inferred. It is true that one may speak of his sensations and describe them as of his knowledge, and it would be competent if the statement meant only that he felt as if he was poisoned, for that would be but a description, however uncertain, vague and conjectural it might be. In my opinion no part of the statement is admissible. It is true the fact is stated by implication that he took a capsule which Lipscomb gave him, and if that stood by itself, it would be admissible. But it is not stated directly. It is stated by implication, sub modo, in the general charge that Lipscomb poisoned him with a capsule he gave him. The principal proposition of the statement is [592]*592that Lipscomb poisoned him, and this is inadmissible. It is a charge which sentiment and imagination invest with potential effect. In its chief significance and the quality of its most effective import it is mere conjecture, and legally false. It cannot be made to do duty as evidence of the incidental fact implied that Lipscomb gave him the capsule.
Nor can I concur with the majority of the court that the statement can be separated in its parts, and to hold a part admissible and a part not. Even conceding, as held by the majority of the court, that so much of the statement, to wit: ‘‘ That he poisoned me with a capsule he gave me, ’ ’ is admissible, taken by itself, yet, it seems to me that the declaration is so related in its parts that the true meaning of any part cannot be had, except when considered in its connection. It should stand or fall as a connected whole.
The statement cannot be even constructed grammatically into different sentences. It embodies a continuity of thought and expression relating to a single fact, the cause of his death. The parts proposed to be rejected stand in the relation of emphatic qualification of the meaning of that proposed to be admitted, even to the extent of casting upon the whole suspicion and doubt. Is not the defendant entitled to the benefit of the absurdity involved in the introduction: “I have been dead, but the good Lord has sent me back to tell, ’ ’ etc. ? It goes to the discredit of the whole statement. So the concluding part: “Guy Jack had my life insured, and hired Lipscomb to kill me,” is a component part of the statement. It is manifestly the expressed reason and ground for the statement that Lipscomb had poisoned him. It is rejected as inadmissible because it is a palpable conjecture, but that very construction supports the theory that the whole statement is mere opinion, and in that respect tends to qualify it. To vary the weight and significance of the statement by arbitrary and artificial adjustment is to juggle with its sense. It is a grave injustice, to be remedied only by admitting the whole statement — a greater injustice.
[593]*593An examination of all the cases cited by counsel, in support of their view of the dying declaration, and of many others, has been made, but none of them seem to me to support the proposition contended for. It is true that in Payne’s case, 61 Miss., which is followed in Powell’s case 74 Miss., the statement “shot me without cause,” and the case of Boyle, 105 Ind., where the statement was “he cut me without reason,” go to a further limit in holding such statements to be of fact than any other adjudications where the precise question is decided. For the decision in Payne’s case no reason is given. It was simply asserted that the statement was a fact and not an opinion — a petitio principii. But one authority is cited, and that is the case of Wroe v. State, 20 Ohio, which will be found upon examination to be doubtful authority for the principle announced. In the case of Boyle v. State, 105 Ind., 469, which follows Payne’s case, the dissenting opinion by one of the judges is an overwhelming demonstration of its error. Payne’s case rests upon Wroe’s case, and that upon Handy v. Commonwealth, 5 Crim. L. Mag. (Ky.), 47, and that, in turn, upon Rex v. Scaife, 1 M. & R., 551. Referring to these cases, we find that, in the case of Rex v. Scaife, “I don’t think he would have struck me if I had not provoked him, ’ ’ was admitted, after hesitation, by Coleridge, J., but upon the express ground that it was in favor of and not against the prisoner. And in Handy v. Commonwealth (Ky.), referring to Coleridge’s decision, supra, it is said: “That the general rule that declarations of deceased are admissible only when they relate to facts, and not to opinions, is subject to the exception that declarations of mere opinion of deceased are admissible when favorable to the accused, and explained the conduct of the deceased, ’ ’ and yet this case is cited in Wroe’s case, as authority for the proposition that the statement was one of fact. But that case of Handy v. Commonwealth is without any analogy to the cases of Wroe, Boyle, or Payne. The authority of that tribunal, however, in an exactly analogous case, condemns the cases mentioned. In Coll[594]*594ins v. Commonwealth, 12 Bush (Ky.), 271, the declaration was, “Michael Collins killed me, and killed me for nothing; ” as to which Chief Justice Lindsay says: ‘‘ The statement that Collins killed the deceased for nothing was but the expression of an opinion, and was clearly inadmissible.”
In the cases of People v. Farmer, 77 Cal., 41; Richards v. State, 87 Wis.; Robert v. State, 5 Tex. App., 141; People v. Green, 1 Denio; Weight v. State, 30 Tex. Co. App.; State v. Terrell, 12 Richards, 330; State v. Belton, 24 S. C., 188, the question as to whether the declaration was one of "fact" or “opinion” was not raised. In the case of Walker v. State, 39 Ark., 22, the statement was, “Nick Walker shot me,” and though the evidence shows it was through an auger hole at night, it was admitted upon the theory that it was in the form of a statement of fact, and if, by any possibility, it could be known to the declarant, it was admissible; but this case was overruled practically in Jones v. State, 52 Ark., where a person was shot through a crack, and it was held impossible that he could know. In Brotherton v. People, 75 N. Y., 159, the deceased stated he did not recognize the prisoner until he commenced his pranks. The prisoner was disguised as a tramp, was the son-in-law of and well known to the deceased, and the testimony was clearly admissible as indicating knowledge from his personal observation. In the case of State v. Arnold, 13 Iredell, the declaration was that “A. B. has shot me, and none other.” It is contended that, from the situation of the parties at the time, deceased did not have an opportunity of knowing the fact so as to enable him to express more than an opinion on the point, which was held to go to the credibility of the statement. In State v. Giles, 8 Wash., 12, the declaration was, “They butchered me.” The indictment was for manslaughter for an unnecessary surgical operation. The court held that it no more expressed an opinion than the word ‘‘ killed ’ ’ used without qualification. Hackett v. People, 54 Barbour, 370, merely distinguishes between what is and what is not res gestae. In Com[595]*595monwealth v. Matthews, 87 Ky., 287, the statement was that the accused and the deceased were engaged in playing, and that the shooting was an accident, and held admissible for the defendant. In State v. Clemons, 51 Iowa, 274, the declaration was, “Ed Clemons shot me; ain’t I right? ” The court says: ‘ ‘ The closing part is put in the way of an interrogatory, and may have been for the purpose of assuring himself, not that he was correct as a matter of opinion, but that his observation of the fact was correct.” In Sullivan v. State, 102 Ala., 142, “Jim Sullivan cut me; he cut me for nothing. I never did anything to him.” And to the objection made that the statement was one of opinion, the decision in the case made but a single reference to it, contained in two lines. It says, ‘ ‘ True, this statement was very general, but it is admissible as a collective fact,” citing Brickell’s Digest, 437. Certainly in this case the expression, “he cut me for nothing; I never did anything to him,” might well be held not an opinion, but a denial that the deceased made any overt act, in view of the circumstances shown for the state that, at the time he was stabbed, he was standing with both hands to his sides in a natural position, and made no movement whatever. But how the expression ‘ ‘ collective fact, ’ ’ whatever that may mean, tends to illustrate the difference between a fact and an opinion, is incomprehensible to me. In Jordan v. State, 81 Ala., 20, the statement was, "Jule shot me, and Handy cut me, and all for nothing, ’ ’ and the only objection urged to it in the appellate court was that no sufficient predicate had been laid, and this was the only question relating to the dying declaration decided by the court.
It will thus be seen that but few of the cases cited for the state are distinctly upon the question whether the dying declaration was a statement of fact or an opinion, or throw any light upon the precise question in this case. In but two of the cases where the question was a debatable one is any reason given for the conclusion that the statement was one of fact. In one (Sullivan v. State, 102 Ala., 142) it is put on the ground that [596]*596the statement was of a "collective fact, ’ ’ and the other is Boyle v. State, 105 Ind. Says the court in that case: ££ £He cut me without reason ’ is an inference of facts from observed facts, ’ ’ an exact definition of an opinion, and is its own refutation. Like the struck eagle, it may
View its own feather on the fatal dart
That winged the shaft that quivered in its heart.
Several of the cases above referred to have been challenged and condemned.
Rice on Evidence, vol. 3, p. 536, refers to Wroe’s case, 20 Ohio St., 460, to Roberts’ case, 5 Texas App., 141; Payne v. State, 61 Miss., 161; People v. Abbott, 4 Whart. Rep., 422; Brotherton v. People, 75 N. Y., 159, as being opposed to the weight of authority, as precedents for the admission of opinion evidence, and referring to the dissenting opinion in Boyle v. State, 105 Ind., 469, says: "It is seldom, indeed, that any opinion is so critical in its analysis, so exhaustive in its citation, or so logical in its conclusions. Any discussion of this subject which omits a careful consideration of this case must be regarded as grossly imperfect. The principal opinion was delivered by Mr. Justice Elliott. It is a very ingenious argument in favor of the prevailing view. But while perfectly aware that my function as a text writer will not tolerate the least attempt to make a law, I submit the dissenting opinion of this exceedingly able court contains the statement of the better view both on principle and authority. ’ ’
Turning- now to the other decision, I think it will appear that the declaration in this case, as admissible evidence, is absolutely condemned by precedent. Of all the adjudications examined, but one has been found directly in point, where the facts are almost identical. In the case of Berry v. State, 63 Ark., the circumstances were that the prisoner had given to the deceased a drink of whisky; that shortly after he was taken ill, and died from the effects of poison. To the physician who was called to at[597]*597tend him, he stated “ that he had drunk the whisky; that the prisoner had given it to him; that he was poisoned; that the prisoner had given him his dose; that it tasted nasty when he drank it. ’ ’ It will be observed how much stronger is this case than the one at bar in favor of the theory that it was a statement of fact, for he not only experienced the sensations which he thought might be those of poison, but he stated that the whisky “tasted ” nasty, a fact which he perceived through one of his senses and that he drank the whisky which the prisoner had given him, yet the court, in that case, held the testimony to be incompetent, because it was but the expression of an opinion, and, in my judgment, rightfully so held.
The same principle is illustrated in many analogous cases. In Whitley v. State, 38 Ga., the statement was, that “it was hard to be killed for telling the truth; that God knew he told the truth, and Ed knew it was the truth,” and was excluded as an opinion. In Willaims v. State, 67 N. C., the statement was excluded as an opinion which was to the effect that Williams shot him, but “I did not see him,” though it was contended that he might have heard the prisoner, and identified him in that way. A witness cannot be allowed to state that the shooting was intentional. Montgomery v. State, 80 Ind., 328. In McPherson v. State, 22 Ga., 478, the declaration was, “Did not believe that accused intended to hurt him. ’ ’ This was excluded as an opinion, although in favor of the accused. People v. Washington, 3 W. C. Rep.: “ I think this man Washington was the man that shot me, ’ ’ was excluded as an opinion. In the case of Shaw v. People, 3 Hunt (N. Y.), 272, it is said that “ it is more important to exclude an opinion deelaratio in articula mortis.” In Chambers v. State, 87 Mo., 408, declarant said he thought he ivas about to draw something from his pocket— a knife or pistol — and that he followed him so that if he did draw a knife or pistol he could catch or knock it out of his hand before he could hurt him, and this was excluded upon the ground that it was an opinion. Moore v. State, 33 Ala., 421, the de-[598]*598clarant, after detailing the circumstances of the homicide, said that defendant was the only slave on the plantation at enmity with him, was excluded. Of like effect is Jones v. State, 17 Ala.; Binn v. State, 37 Ala., 103; Loshbaugh v. Birdsell, 90 Ind., 466; Yost v. Conroy, 92 Ind., 464; Ferguson v. Hubbell, 97 N. Y., 507; Warren v. State, 9 Tex. Ct. App., 629; Moex v. State, 100 Ill., 240. In the light of these unchallenged precedents, on principle and justice, it is my conclusion that the dying declaration of Stewart was not'a statement of fact, but an expression of an opinion, and inadmissible as evidence.
It is assigned for error that the court refused the request of the defendant to deliver to the jury the paper containing a memorandum of the autopsy. Seeing that it was plainly required by the statute, it is not perceived upon what possible ground the request was refused. But as all the facts in the memorandum were testified to orally, and the memorandum read, the error is not deemed material. If the dying declaration was irrelevant and incompetent evidence, there can be no doubt, in view of its character and probable effect, as to the consequences of such error. It is conceded that such error must be fatal to any verdict, and for which, as well as for the error in the ninth instruction given for the state, I am of opinion that the judgment of the court below must be reversed.
The conclusions of the court are, all concurring:
1. That the court below erred in granting the ninth instruction given for the state.
2. Judges Whitfield and Thompson concurring, Magruder dissenting, that the court erred in refusing fifteenth instruction asked by the defendant.
.3. All concur that a part of the dying declaration is inadmissible. Judges Whitfield and Thomspon concur in holding a part admissible, to wit: “ Dr. Lipscomb has killed me; he has poisoned me with a capsule he gave me to-night,” and that this may be separated from the other parts of the statement. Judges Whitfield and Thompson hold that no error can be predicated of [599]*599the admission of the dying declaration on the ground a part of it is incompetent, for the reason that the objection to its admissibility is general, and not specific. Magruder holds that no part of the dying declaration is admissible, and, from the character of the statement, is incapable of separation without injustice to the defendant. Judges Thompson and Magruder concur that the case should be reversed, for the reasons indicated in their respective opinions, and Judge Whitfield, notwithstanding the errors conceded by him as to giving the ninth instruction for the state, and the refusal of the fifteenth instruction for the defendant, hold the case should be affirmed.
I fully concur in the view stated by Judge Thompson as to the law which should govern the reversal of cases for error. It is not within the province of the appellate court, under our constitution and laws, to be the triers of fact and of guilt. Our duty is to lay down, with unswerving purpose, the law as we find and understand it. Questions of public policy belong to another forum. A legal trial is a trial according to the law, and a legal conviction is a conviction by a jury according to recognized procedure and principles of law. If, in any trial, an error of law be made, the case should be reversed, if it be so material or of such character as is calculated to influence, and probably did influence, the verdict. The case is not to be looked at from a judicial standpoint, and the guilt of the accused be so ascertained, or the materiality and consequence of the error be so determined, the question in all cases being, not what the jury ought to do, or might do, according to our judgment, but whether the error is of such character.as that the jury itself might have been influenced by it.