The petitioner, Charles A. Shepard, a major in the medical corps of the United States Army, has been convicted of the murder of his wife, Zenana Shepard, at Fort Riley, Kan., a United States military reservation. The jury having qualified their verdict by adding thereto the words 'without capital punishment' (18 U.S.C. § 567 (18 USCA § 567)), the defendant was sentenced to imprisonment for life. The judgment of the United States District Court has been affirmed by the Circuit Court of Appeals for the Tenth Circuit, one of the judges of that court dissenting. 62 F.(2d) 683; 64 F.(2d) 641. A writ of certiorari brings the case here.

The crime is charged to have been committed by poisoning the victim with bichloride of mercury. The defendant was in love with another woman, and wished to make her his wife. There is circumstantial evidence to sustain a finding by the jury that to win himself his freedom he turned to poison and murder. Even so, guilt was contested, and conflicting inferences are possible. The defendant asks us to hold that by the acceptance of incompetent evidence the scales were weighted to his prejudice and in the end to his undoing.

Shepard v. United States/Opinion of the Court

The evidence complained of was offered by the government in rebuttal when the trial was nearly over. On May 22, 1929, there was a conversation in the absence of the defendant between Mrs. Shepard, then ill in bed, and Clara Brown, her nurse. The patient asked the nurse to go to the closet in the defendant's room and bring a bottle of whisky that would be found upon a shelf. When the bottle was produced, she said that this was the liquor she had taken just before collapsing. She asked whether enough was left to make a test for the presence of poison, insisting that the smell and taste were strange. And then she added the words, 'Dr. Shepard has poisoned me.'

Shepard v. United States/Opinion of the Court

The conversation was proved twice. After the first proof of it, the government asked to strike it out, being doubtful of its competence, and this request was granted. A little later, however, the offer was renewed; the nurse having then testified to statements by Mrs. Shepard as to the prospect of recovery. 'She said she was not going to get well; she was going to die.' With the aid of this new evidence, the conversation already summarized was proved a second time. There was a timely challenge of the ruling.

Shepard v. United States/Opinion of the Court

She said, 'Dr. Shepard has poisoned me.' The admission of this declaration, if erroneous, was more than unsubstantial error. As to that the parties are agreed. The voice of the dead wife was heard in accusation of her husband, and the accusation was accepted as evidence of guilt. If the evidence was incompetent, the verdict may not stand.

1. Upon the hearing in this court the government finds its main prop in the position that what was said by Mrs. Shepard was admissible as a dying declaration. This is manifestly the theory upon which it was offered and received. The prop, however, is a broken reed. To make out a dying declaration, the declarant must have spoken without hope of recovery and in the shadow of impending death. The record furnishes no proof of that indispensable condition. So, indeed, it was ruled by all the judges of the court below, though the majority held the view that the testimony was competent for quite another purpose, which will be considered later on.

We have said that the declarant was not shown to have spoken without hope of recovery and in the shadow of impending death. Her illness began on May 20. She was found in a state of collapse, delirious, in pain, the pupils of her eyes dilated, and the retina suffused with blood. The conversation with the nurse occurred two days later. At that time her mind had cleared up, and her speech was rational and orderly. There was as yet no thought by any of her physicians that she was dangerously ill, still less that her case was hopeless. To all seeming she had greatly improved, and was moving forward to recovery. There had been no diagnosis of poison as the cause of her distress. Not till about a week afterwards was there a relapse, accompanied by an infection of the mouth, renewed congestion of the eyes, and later hemorrhages of the bowels. Death followed on June 15.