Truly, J.,
delivered the opinion of the court.
On tbe night of the 29th of November, 1902,Mrs. Annie Boyd died of acute gastritis. An autopsy disclosed the fact, according to the testimony of the analytical chemist, that her death was superinduced by arsenical poisoning. Appellant, the husband of the deceased, was indicted for her murder, and upon trial was convicted, sentenced to the penitentiary for life, and from that judgment prosecutes this appeal.
The testimony adduced at the trial was purely circumstantial. The theory upon which the prosecution proceeded was that the appellant had poisoned his wife through a desire to effect her death so that he might procure the proceeds of a policy of insurance on her life and continue an illicit intimacy with one Susan Artman. The following state of facts was developed by the testimony: Boyd was a candymaker, conducting a candy and fruit stand. His household consisted of himself, his wife, and one small child, and a cousin of his wife, a schoolgirl some sixteen years old, named Nora McDaniel. The family occupied rooms adjoining Boyd’s fruit stand. Artman and his wife, Susan, lived in an adjoining building, where the husband conducted a bakery. The rear premises of each house was in a yard common to both. Boyd and Susan Artman soon became friendly, and in constant conversation, Boyd devoting considerable attention to her, to the exclusion to some extent of his wife, though the testimony failed to positively disclose any criminality in this intimacy. Mrs. Boyd was much distressed, and brooded over this intimacy, and intimated on several occasions that she contemplated suicide. Nora McDaniel testified that when she returned from school on Monday'preceding the Saturday night on which Mrs. Boyd died she found Mrs. Boyd sick from nausea, vomiting incessantly, and seemingly in in*420tense pain. A doctor was called in, prescribed several remedies,which were continued during the week and until her death. After this spell on Monday, Mrs. Boyd seemed to rally, and on Thursday was so much better that she was able to be out of bed. On Friday morning she suffered a relapse, and from that time until her death lingered in constant and excruciating pain. During the course of the treatment of the case the attendant physician prescribed several different medicines, which were administered generally by Miss McDaniel, though occasionally by the doctor and by Boyd himself. None of these prescriptions, if correctly compounded, contained poison in any form. For some time prior to the attack on Monday, Mrs. Boyd had been taking another medicine for dyspepsia, from which she was a sufferer. On Friday before Mrs. Boyd’s death, the appellant asked the attending physician if he did not think his wife had been poisoned; that he (appellant) suspected that she had taken poison on account of jealousy of Susan Artman. Shortly after the death, while appellant was in the room, sitting on the bed with his dead wife, he was arrested, and carried to another place, where he remained under guard. While under arrest he inquired of the officer who had him in charge for what he was being held. The officer replied for the poisoning of his wife, whereupon the appellant remarked that he suspected that his wife was poisoned even before the doctor had discovered it. It was testified to by a clerk in a drug store that appellant, the week before this death, had purchased a package of poisonous mixture containing arsenic and called “Bough on Bats,” and the contention of the state was that this was the poison administered to Mrs. Boyd. It was further in proof that no one had access to the premises occupied by the Boyd family except his own household and the Artmans, with the possible exception of a lady who occupied rooms in another portion of the same building. Appellant did not testify on the trial. During the progress of the trial many objections were made and *421exceptions reserved to tbe introduction of testimony on bebalf of tbe state.
1. Nora McDaniel and the attending physician, Dr. Dampeer, were each permitted to relate in their testimony the statements made by Mrs. Boyd in reference to what caused her attack on the Monday preceding her death. They stated that Mrs. Boyd told them that, after eating a hearty dinner on Monday, she bad taken a dose of her dyspepsia medicine, the last dose in the bottle, and that shortly thereafter she bad been taken with nausea and cramps. This was error, plain and palpable. It is well settled that the statements of a party describing the symptoms of the suffering being endured at the time of the statements are admissible, but the testimony now being reviewed does not come within the scope of that rule. Mrs. Boyd was not stating the symptoms which she was then suffering, but was narrating a past transaction, was detailing the circumstances under which she began to expprience the suffering, and giving her opinion of the causes which produced her condition. This comes expressly within the condemnation of this court in Field v. State, 57 Miss., 474 (34 Am. Rep., 476), where the whole subject is lucidly treated.
2. The doctor who bad treated Mrs. Boyd produced before the jury several bottles, which, according to bis testimony, contained, accurately compounded, the different prescriptions which be bad given to Mrs. Boyd, and he was permitted to show the effect that Bough on Bats would have on the liquids in each of the bottles — changing the color of the various medicines in proportion to the amount of Bough on Bats mixed therewith. In connection with this evidence Nora McDaniel was permitted to testify that on the Sunday preceding the death Boyd, in accordance with his usual custom, brought bis wife at the breakfast table an orange; that on this occasion the orange bad been punctured, and that the juice which oozed out of the bole was of such a peculiar color that Mrs. Boyd refused it, and that witness also declined to eat it, and was allowed to give as her *422reason that she suspected there was something wrong with it; that the orange was not eaten, bnt in some way mysteriously disappeared. The witness was also allowed to state that the juice which came from the orange corresponded in color with one of the medicines in which the poison had been mixed. This was error of the gravest kind. There was no evidence before the jury that the orange contained any poisonous substance. There was no testimony that any of the medicines taken by Mrs. Boyd during her last sickness had in any wise been changed in color. The effect of this incompetent testimony was to allow, if not to induce, the jury from the bare proof that the admixture of poison would cause a change of color, to leap to the-conclusion that poison had been mixed, and yet the witness for the state emphatically and repeatedly denied any knowledge of any change in any of the medicines taken by or* administered to the deceased.
3. The only evidence of poison discovered beyond that disclosed by the autopsy was found in a vial, which it was shown had originally contained the dyspepsia medicine, the last dose of which was taken on the Monday preceding the death, and which, according to the improperly admitted statement of Mrs. Boyd, was the cause of her first attack. This vial was kept on a shelf in the kitchen, to which room it was shown no one had access save Boyd, Mrs. Boyd, Nora McDaniel, and the Artmans. During the continuance of Mrs. Boyd’s illness, Nora McDaniel, at the request of the doctor, searched for this vial, but, being unable to find it, she was permitted to state that Mrs. Boyd told her where she (Mrs. Boyd) had placed the vial. She was further allowed to testify that the vial had been moved from the place; that she had not moved it; and that she after-wards found it on another shelf, surrounded and hidden by other bottles, still containing a small quantity of a clear liquid different in color from the original compound, and which it was shown contained arsenic. We are unable to comprehend upon what principle the court permitted the witness to repeat to the *423jury the statement made by Mrs. Boyd, not in the presence of the appellant, as to what disposition had been made of the bottle. The damning effect of the testimony becomes manifest upon a moment’s reflection. Here was a bottle which originally contained medicine, but in which poison was found, secreted in a room to which only the household of Boyd and the Artmans had access. The witness not only testified that she did not personally remove the bottle, but she was further allowed to repeat the statement of Mrs. Boyd — another of the persons having access to the room — as to where she had placed it. By this improper process of exclusion it was driven 'home to the jury that, as the bottle had been moved, and as poison had been placed therein, and as two of the persons were shown not to have moved it, no one remained who possibly could have been the active agent except the appellant or Mrs. Artman, who it was argued was his paramour, or her husband, who it is not contended was even suspected of complicity.
4. To accentuate this improper testimony the state was granted an instruction, by which the jury Avere told that they should convict the defendant if they believed beyond a reasonable doubt that he himself mixed the poison, or “procured it to be done by another,” with the intent that it. should be taken by his wife, and it was so taken, and caused her death. The effect of this instruction was to say to the jury that, even though the evidence might fail to prove defendant’s direct connection with the crime, they should nevertheless convict if they believed that he had procured Susan Artman to mix this poison, yet the record is absolutely devoid of a semblance of proof that Susan Artman was in any manner implicated in administering the poison. This instruction injected into the minds of the jury the unsupported suggestion that appellant had procured in some unexplained manner some other person to administer poison to his wife. It presented to the jury two theories — one wholly unsupported by proof; the' other predicated entirely upon evidence purely circumstantial in its character.
*4245. It was clearly admissible to show that Boyd 'disputed one of the witnesses at the preliminary investigation, and denied the statement that he had purchased Bough on Bats from the witness. But a majority of the court think it was error to admit proof that appellant did not testify'before the justice of the peace who conducted the preliminary investigation; that he was not put on the stand as a witness, and did not go on voluntarily. It is true that this action of the court was condemned as error in Bunckley v. State, 77 Miss., 540 (27 South., 638), but the writer hereof is not satisfied of the soundness of that decision, and would not, on this ground alone, consent to a reversal of the judgment herein.
6. The witness who testified that he had sold the Bough on Bats to appellant was permitted on the trial in the circuit court to state that he had testified to the same fact at the preliminary examination. This effort to corroborate the sole witness by whom it was shown that appellant had possession of poison by proving that he had testified to a similar statement upon a different occasion is so plainly erroneous as to scarce require citation of authority. It is never competent to attempt to corroborate a witness by proving affirmátively that ón other occasions he had made statements which harmonize with his testimony then related to the jury. The testimony is delivered under the sanctity of an oath, and carries such weight as the jury may attach to it. If intrinsically of doubtful worth or little force, proof that it is mere repetition increases neither its value nor its force. Wharton, Cr. Ev., par. 492; Williams v. State, 79 Miss., 555 (31 South., 197).
7. The'attending physician, over the objection of defendant, testified that on Saturday evening, a few hours before Mrs. Boyd’s death, she made to him a dying declaration, which was as follows, as the same appears in the record: “I told her she was very sick, and suffering from some poison, and she remarked, Wes, I am going to die.’ I told her her husband was under suspicion, and it was her duty to tell us if she had taken *425anything herself, and she promptly replied, ‘The Lord is my witness, I have taken nothing except what you gave mé.’ ” That portion of the statement attributed to Mrs. Boyd was properly admitted on this theory: One of the hypotheses deducible from the testimony was that Mrs. Boyd took'the poison knowingly, with the deliberate intention of committing suicide. As her language in the dying statement impliedly, if not expressly, contradicted the suggestion of suicide, it was competent, because it tended to eliminate one hypothesis from the case. This part of the statement meets all the tests which render a dying declaration admissible. It was made under a sense of impending dissolution. It was not the mere expression of an opinion, but was the statement of a fact, and could have been given in evidence had the party lived and the defendant been on trial for an attempt to poison. In either event the material inquiry must have been, “Who administered the poison?” If living, Mrs. Boyd would have been a competent witness to deny that she had knowingly taken it. Being dead, her dying declaration was admissible for the same purpose. But the admission should have béen restricted to the language used by Mrs. Boyd. In and of themselves by no rational interpretation can her words be distorted into an accusation of her husband. Her meaning, so far as can be gleaned, was simply this: “I ’am going to die, but the Lord is my witness, I have taken nothing- but what you, my physician, gave me.” This was not a charge of guilt against appellant. Lying then in the throes of a painful death, the suffering woman, soon to' be ushered into eternity, solemnly calls that Deity, in whose presence all must stand, to witness that she has not violated “His canon against self-slaughter.” It was error to admit that portion of the dying declaration where the witness, the attending physician, states: “I told her her husband was under suspicion, and it was her duty to tell us if she had taken anything herself.” This was incompetent. It was a mere expression of an unsworn opinion of the witness, based upon suspicion only; yet it was introduced before the *426jury, clothed with the sanctity of the solemn circumstances which render a dying declaration admissible. If Mrs. Boyd came to her death by poison, the sole remaining question for the consideration of the jury was, “Who is guilty V’ The identity being established, the guilt followed of course. The defense of appellant was, not that the party committing the act was not guilty, but that he was not the guilty actor. It was error to place on the defendant, circumstanced as he was, charged with a most atrocious crime, the burden of having to combat this expressed suspicion of the attendant physician upon the crucial question of his defense, the vital point in the case. The dying declaration should have been limited to the statement made by the dying person, upon whose mind the knowledge óf impending dissolution weighs with such gravity as to obviate the legal necessity of an oath. Restricted to the language used by Mrs. Boyd, the jury might reasonably have considered her statement as simply an earnest denial of her own guilt and a rejection of the suggestion made by the doctor. But when her words, uttered under such circumstances, are permitted to go to the jury linked and blended with the suspicion of guilt voiced by the physician who had 'ministered to her dying needs, they are no longer simply a protestation of her own innocence, but become an ■ accusation from the silent lips of a poisoned wife against a faithless husband. The far-reaching and damaging effect of this admission cannot be overestimated.
After repeated and protracted examination of this record, after the testimony has been carefully scanned many times, we are unable to say that the appellant was granted that fair and impartial trial guaranteed by our law, and we are not able to affirm with confidence that the testimony erroneously admitted may not have contributed to the result. In cases of this character, where the Very atrocity of the crime is often of itself sufficient to array public sentiment against the unfortunate accused and place him almost beyond the pale of human sympathy, courts should be only the more scrupulously careful *427in guarding in every just manner the rights and privileges to which he is entitled.
Reversed and remanded.