Ladd, J.
In the administration of the law denouncing offenses against the sex of woman, it must always be borne *349in mind that, by the very nature of things, such crimes are ■committed in secret, and it ought not to be expected that the testimony of the alleged victim will be supported to any ■considerable extent by direct evidence. Of necessity, the prosecution must ordinarily rely upon proof of incriminating circumstances and admissions for corroboration when required, and it is often a delicate and difficult task to pass ■on their sufficiency for that purpose. On the one hand, the character of the charge precludes the hope of any •considerable amount of cumulative proof, while, on the other, its nature renders it peculiarly susceptible to the influences of passion and prejudice. The conditions surrounding a trial — the atmosphere, as it were, permeating the investigation — can never be transmited through the abstract or transcript, and, for this reason, appellate courts necessarily rely very largely upon the discernment and ■sound discretion of the nisi prius judge with respect to the •correctness of the jury’s findings as expressed in the judgment entered. Without reviewing the evidence in this case, it is enough to say that, in view of the conflict we ought not to interfere with the conclusion of the district judge refusing to set the verdict aside because not supported by the ■evidence.
II. The prosecutrix lived in the family of the defend.ant from June or July, 1898, until February, 1899, during which time she taught a country school. The theory of the state was that the accused began having intercourse with her about the first of September, when she became pregnant, and that about November 10th he inserted a speculum in her vagina, and attempted, by the use of an iron probe, to cause an abortion, and that about two weeks thereafter she lost the foetus. The defense was a denial of all these accusations, and it was insisted that, if pregnant, the prosecutrix became so by Oscar Fullerton, with whom she had been keeping company and to whom she had become engaged to marry, and who had waited on her up to the time of trial. *350Only two of the errors assigned in addition to the one noticed require attention, as the others are not likely to arise on another trial.
1 III. Fullerton was called as a witness, and, over defendant’s objection, permitted to detail a conversation with one White, not in the presence of the defendant. In response to an inquiry with respect to his knowledge of the miscarriage, he had said: “Yes, I had found out a few days before that there had been an abortion. So far as I am concerned, the sooner they push it the better. I do not want people to think I am interested in it, and I want to see it come out.” This evidence was wholly irrelevant to any issue. The state attempts to justify its admission on the ground that this conversation had been brought out in the cross-examination of White, and that seems to have been the ground of the judge’s ruling. This-was a mistake, as White testified concerning this conversation in his direct examination. Its detail by him did not justify the ruling permitting Fullerton to repeat the conversation in rebuttal. Allowing him to do so under the circumstances and in view of the theory of the defense was not only erroneous, but extremely prejudicial.
2 IV. The second instruction, otherwise correct, closed with this sentence: “A reasonable doubt does not mean a doubt from mere caprice or groundless conjecture, but it means a reasonable doubt, such as the jury are able to give a reason for.” .A similar statement contained in an instruction was condemned in State v. Cohen, 108 Iowa, 208" court="Iowa" date_filed="1899-04-08" href="https://app.midpage.ai/document/state-v-cohen-7108638?utm_source=webapp" opinion_id="7108638">108 Iowa, 208, decided after the ruling in this case. The authorities there cited fully sustain the decision, and it is, adhered to.- — Reversed.