The opinion of the court was delivered by
Hoch, J.A brief opinion was filed in this case on July 14, 1947, reversing the judgment because of error in the admission of' incompetent evidence, granting a new trial and announcing that a fur*623ther opinion would be filed latér. At that time we directed the release of the appellant, upon bond, pending such new trial.
Appellant was convicted and sentenced under section 42 of the crimes act (G. S. 1935, 21-435). He appeals from an order denying his motion for a new trial, asserting that the trial court erred in admitting incompetent evidence, in unduly restricting the cross-examination of witnesses for the state, and in failing to instruct the jury on the elements- of section 42 of the crimes act. The contention principally stressed is that the appellant’s rights were prejudiced by the admission of testimony as to the results of “lie-detector” tests upon both the defendant and the complaining witness.
Appellant was tried upon two counts, the first being for felonious assault (G. S. 1935, 21-431) ,■ and the second for kidnapping as defined in section 21-449, G. S. 1935. At the first trial the jury was unable to agree and was discharged. Upon the second trial, the defendant was found guilty of felonious assault under section 42 of the crimes act (G. S. 1935, 21-435) and sentenced to imprisonment for a term not exceeding five years. Motion for a new trial was overruled, and this appeal followed.
The factual background of -the questions here presented may be briefly stated. Appellant Lowry and the complaining witness lived about four miles apart on farms in Linn county. Lowry had asserted that one of his horses had been stolen about a year prior to the incident here involved.
Richards’ story was that on the evening of April 24, 1946, when he -was at-his barn with a lantern to milk cows and do other chores, Lowry suddenly appeared carrying a shotgun, flashed a light upon him; and under threats upon his life forced him to walk across fields' and pastures-to the Lowry farm, continually .threatening to shoot him, and that when they reached the Lowry place Lowry said, “Now what did you do with my mare?”; that he answered, “I nevér stole or never saw your mare”; that after renewed threats, Lowry shot him in the foot; that in order to avoid further injury he said he “was passing out”; that after slapping him repeatedly, Lowry got a rope and tied his hands; then got a horse and helped him get on the horse, took him to the barn and forced him to sign two statements confessing that he had stolen the Lowry horse. . The many other details of Richards’ testimony need not be recited.
Lowry, of course, told an entirely different story. He testified that, upon the night in question he went to his bam, after dark, to *624shoot at rats; spent about an hour there, then took his shotgun with a flashlight attached to the barrel and started back to the house; that he thought he heard a voice saying “whoa,” that he walked to where some of his horses were about a half-mile from the barn; that he found some of his horses and then walked farther on and saw something white moving; that he turned his flashlight on and then saw someone leading one of his mares and yelled to him; that he then shot towards the ground; that Richards begged him not to call the sheriff and said that if he would not do it, he would sign a statement confessing that he was attempting to steal the horse when he was shot in the foot. The testimony of numerous witnesses called by both sides need not be narrated.
We first note appellant’s contention that the trial court unduly limited cross-examination of some of the state’s witnesses. That question is not here for review. If such cross-examination was unduly restricted to the prejudice of appellant, it was a trial error, and wasi not specified in the motion for a new trial. It is true that in the motion the defendant included as grounds for a new trial “erroneous rulings of the court.” But he specifically enumerated the alleged erroneous rulings and did not include therein any reference to limitation of cross-examination. Not being called to the attention of the trial court on the motion for new trial, it is not reviewable on appeal.
The next contention of appellant is that the trial court failed to instruct the jury upon essential elements included in the crime defined in section 21-435, G. S. 1935 (Crimes Act, § 42) for which he was convicted. The trial court did instruct the jury, at some length, that the defendant might be found guilty of the lesser offense defined in section 21-435, stating, inter alia, that under the statute, proof of intent to injure is not a necessary element of the offense. The defendant did not object to the instruction when given, did not ask any modification or clarification of it, and requested no substitute instruction as to the elements necessary to be shown to justify conviction under section 21-435. Appellant contends that such failure does not preclude review here, citing in support State v. Phelps, 151 Kan. 199, 97 P. 2d 1105, and State v. Carr, 151 Kan. 36, 98 P. 2d 393. The cases are not persuasive here. In the Phelps case, the trial court failed to give any instruction whatever as to a lesser offense — one which was most clearly indicated under the evidence. That is not the situation here, and the jury *625here did find the defendant guilty only of the lesser offense. A like distinction exists as to the Carr case. Appellant’s contention on this point cannot be sustained.
We come to appellant’s principal contention, relating to admission of evidence with reference to the “lie-detector” tests. Following the first trial, the trial court suggested to both defendant and complaining witness that they submit to “lie-detector” tests before the second trial. In compliance with the court’s expressed desire, they submitted to such a test, but there was no stipulation nor agreement that testimony might be admitted concerning the results of such tests. Prior to the opening statement for the state, and in the absence of the jury, counsel for the defendant — aware that such testimony was to be offered by the state — argued that it was inadmissable, but the court stated that it considered the results of these tests to be of a scientific nature, to be competent “to a certain extent,” and that the state’s witness would be permitted to testify as to the tests made upon both the defendant and the complaining witness. Defendant’s right to review on this question was further protected by timely objection when the evidence was offered and upon motion for a new trial.
The tests were made upon the same day upon both the defendant Lowry and the complaining witness Richards, by Phil Hoyt, captain of police in Kansas City, Mo. Hoyt testified that he had been conducting tests with the lie-detector since 1937, having taken training under Keeler of Northwestern university, who is largely credited with devoloping the instrument. As to his experience as an expert witness, he testified:
“I never did testify before as an expert in this particular branch of police work as to the result of a machine that w^s used to bolster up the testimony of a prosecuting witness in a ease. In all these 2,400 cases that I gave I have testified merely on rebuttal and preliminary. The reason 1 went on rebuttal, it concerned confessions that persons made charged with crime after they had been subjected to this test.” (Italics supplied.)
The witness described the “lie-detector” and its operation as follows:
“This instrument is an ordinary blood pressure cuff similar to that used by a physician used in taking blood pressure. You place it around the right arm and the pressure from the artery is carried to an attachment that registers with a pen based on a pivot which leaves a graph on paper, being wound by a motor at the rate of six inches per minute. Another attachment is an ordinary rubber tube one end of which is a smaller hose which goes to an at*626taehment that records the respiration on the graph paper by a pen. By respiration I mean, breathing of the person taking the test. The object is that theoretically every person who lies deliberately will show some physical reaction, and with most people it is registered by a rising blood pressure and a subconscious block in breathing. The difference in breathing is ordinarily seventeen to nineteen inhalations a minute, which is usually increased during a five second period.”
The witness Hoyt was then permitted to testify not merely as to whether, in his opinion, the tests indicated, generally, truthfulness or falsification on the part of Richards and'Lowry, but was permitted to give his interpretation of their answers on the essential issue of the .case — his answers bearing directly upon the guilt or innocence of the defendant. From Hoyt’s extended testimony, we take brief excerpts. As to the test upon Richards he testified:
“To the question ‘Did Melvin Lowry force you to go with him from your farm to his farm on the night of April 24th’ he answered ‘yes’ and we didn’t get any reaction, indicating that he is telling the truth. To the question, ‘did you have some coffee today’, we got no reaction indicating he was not telling a lie. To the question, ‘did Melvin Lowry catch you leading a horse from his pasture on- the night of April 24th’, he said, ‘no’, and we didn’t get any unusual reaction, indicating that he was or he believed he was telling the truth. To the question, ‘did you try to steal a horse from the Lowry farm on the night of April 24th’, he said ‘no’, and there is no reaction. To the question, ‘did Lowry shoot you because he found you trying to take one of his horses’, he said' ‘no’. There is no reaction, indicating he is telling the truth.” (Italics supplied.),
As to the test upon the defendant Lowry, the witness testified:
“To the question, ‘did you force Rex Richards to go with you to your farm on the night of April 24th, this year?’, he shows' definite reaction, and that is when he said, no. The reaction indicated by the graph is similar to that of persons not telling the truth. The physical reaction indicated is a rise in'- blood pressure and a hesitancy in breathing. There is a typical reaction of that to a person not telling the truth. . . . ‘Did you find Rex Richards leading a horse from your farm on the night of April 24th?’ He said, ‘Yes’ and showed definite reaction to not telling the truth. . . . ‘Did you go to Rex Richards’ farm on the night of April 24th?’ ‘No’, showed positive reaction indicating not telling the truth. ‘Did you shoot Rex Richards because you found him stealing one of your horses?’ ‘Yes’, same reaction, indicating not telling the truth.” (Italics supplied.)
All this testimony went directly to the issue of whether Richards’ story or Lowry’s story was the true one; whether Richards was kidnapped, deliberately shot in the foot and under threats to his life forced to sign the written confessions, or whether Richards went to the Lowry farm after dark and Lowry there discovered him *627attempting to steal a horse, and úpon making such discovery, shot towards the ground with resultant injury to Richards.
We need not set out the court’s instructions to the jury with reference to this testimony except to note that the court did not limit its province to a general question of credibility but put it in the same class as all other evidence to be weighed by the jury in determining the guilt or innocence of the accused.
The practical effect of the admission of this testimony was to constitute a mechanical device — as reported by the operator- — a sort of witness in absentia on the question of the defendant’s guilt or innocence.
In no case cited by appellee or found in our own research has a court of last resort sanctioned the admission of such testimony. Nor has any trial court, as far as we are aware, admitted testimony, over objection, as to the result of such tests upon a complaining witness. The ultimate, logical result of doing so would be to have such tests made upon all witnesses for the purpose of helping the jury determine their credibility.
Cogent reasons in support of this attitude of the courts readily suggest themselves. In the' first place, the vital function of cross-examination would be impaired. The operator, appearing as a witness to report and interpret the results of the test, might be questioned as to his. qualifications, experience, his methods, and on similar matters, and that is about all. But the machine itself— conceding the comparatively high percentage record as to accuracy and reliability claimed for it — escapes all cross-examination. There is no persuasive analogy here with such tests as fingerprinting which have a strictly physical basis, clearly demonstrable. It is not contended that the lie-detector measures or weighs the important psychological factors. Many innocent but -highly sensitive persons would undoubtedly show unfavorable physical reactions, while many guilty persons, of hardened or less sensitive spirit, would register no physical indication of falsification. This the trained o'perators of course understand, and proceed upon the basis of a large percentage of error. But it seems quite too subtle a task of evaluation to impose upon an untrained jury.
Consider the situation in the instant case. Two men were involved. One wap a defendant on trial. The other was merely a witness and under no such emotional strain. Can it be said that with such wholly different mental states existing, the tests would *628be equally fair? Must the jury be asked to consider and weigh such intangible and elusive elements?
We are not ready to say that the lie-detector has attained such scientific and psychological accuracy, nor its operators such sureness of interpretation of figures on a dial that the testimony here in question was competent, over objection, for submission to a jury holding the fate of the defendant in its hands. It must be remembered that we are not here considering a case where there was a prior agreement that the results of the test might be admitted in evidence. While the two men agreed to take the tests, there was no such stipulation. In fact, the witness Hoyt agreed upon cross-examination that in the conversation between counsel about taking the tests, the defendant’s counsel stated that it “wasn’t for the purpose of placing before a jury.” Moreover, the defendant might well have hesitated to refuse to take a test proposed by the judge himself before whom he was about to be tried.
All this is not to discredit the lie-detector as an instrument of utility and value. Its usefulness has been amply demonstrated by detective agencies, police departments and other law-enforcement agencies conducting criminal investigations. It is also being frequently employed in matters, other than investigation of crimes. By its use admissions and confessions are frequently secured, and facts developed which assist in further discoveries. Such admissions and confessions, if otherwise competent, have generally been admitted, and no reason now appears why they should not be admitted. But we are not here dealing with such questions.
The conclusion here reached is in line with the almost unanimous holding of other courts and with the conclusions of law writers generally. About twenty years have elapsed since the admissibility of a lie-detector test was first passed upon in a reported case. The only reported decision of which we are aware holding such a test admissible is People v. Kenny, (1938) 167 Misc. 51, 3 N. Y. S. 2d 348, and in that case no appeal was taken. In all other reported cases the courts have rejected such tests, generally upon the ground that their reliability has not yet been sufficiently established. (139 A. L. R. 1174.) (See, also, annotations in 34 A. L. R. 147; 86 A. L. R. 616, and 119 A. L. R. 1200.) And even the Kenny case, in which an expert was permitted to testify concerning a test made upon the defendant, is no authority to support the admissibility of a test upon a complaining witness.
*629In Frye v. U. S., 54 D. C. App. 46, 293 Fed. 1013, which was a murder case, the accused had taken the test, with a result said by the expert witness to be favorable to him. His offer to submit to the test again, in the jury’s presence, was rejected on the ground that the test “has not yet gained such standing and recognition among physiological and psychological authorities” as to justify its use as evidence in court.
In State v. Bohner, 210 Wisc. 651, 246 N W. 314, which was a bank robbery case, the defendant sought admission of the result of such a test in support of an alibi. The offer was rejected.
In an article in the Wisconsin Law Review, May, 1943 (page 430), a number of unreported Wisconsin cases are discussed in which trial courts have admitted in evidence the results of such tests. But these were cases where either the confession only, which apparently had been in this manner secured, was admitted, or where there had been a prior written agreement that the testimony might be admitted.
In an article in 29 Cornell Law Quarterly, 535 (1944), the writer, who looks with favor upon a larger use of such tests in criminal procedure, discusses a number of cases in most if not all of which admission of such testimony was refused. While the writer thinks that the results that have been attained “would seem to justify the assertion that using the results of lie-detector tests to verify the testimony of important witnesses would be an inestimable advance in the efficiency of the legal process,” he none the less concedes that there are real difficulties attendant upon use of lie-detector tests by the courts, which “merit serious consideration.” He says:
“First, a party might produce only those tests which were favorable to himself, and his opponent would not be able to examine him under the lie-detector, nor to cross-examine him should he refuse to take the-stand. Second, it would be difficult for the opposing counsel to expose an incompetent or dishonest expert and to cross-examine him concerning the tests, because the lie-detector is not adequately standardized as to instrument, manner of conducting tests, qualifications of examiners, and interpretation of the records.” (p. 543.)
A leading authority on the lie-detector is Fred E. Inbau, professor of law at Northwestern university. As an associate of Doctor Keeler of Northwestern, who is credited with perfecting the instrument, he is a firm believer in the usefulness and reliability of the instrument and has had wide experience in its use. But Professor Inbau apparently does not share the view entertained by some en*630thusiasts that courts should readily admit in evidence testimony or the results of such tests. In an article in the Boston University Law Review (April, 1946) Professor Inbau says:
“A number of trial courts have admitted in evidence the testimony of lie-detector examiners as to their interpretation of test records in instances where, prior to the test, counsel for both litigants agreed and stipulated that the test results could be used in evidence without objection on the part of the party adversely affected thereby. Although no appellate court has thus far passed upon the legality of such agreements and stipulations, a prediction may be ventured that the procedure would be declared valid, with the proviso, however, that the approval or disapproval of the agreement and stipulation in any given case should rest within the trial court’s discretion. But in the absence of such pre-test agreements and stipulations, the test records and the examiner’s interpretation thereof have been held inadmissible as evidence by every appellate court which has had occasion to consider the matter, for the very understandable reason that the test had ‘not yet gained such standing and scientific recognition among physiological and psychological authorities as would justify the courts in admitting expert testimony deduced from the discovery, development, and experiments thus far made.’ ” (Italics supplied.) (p. 270.)
Lastly, it is contended that even if the trial court erred in admitting this evidence, the error cannot be said to have prejudiced appellant’s rights. The argument is that appellant was charged with felonious assault, with malice aforethought; and with kidnapping in the first degree; that the lie-detector evidence tended only to support Richards’ testimony and to contradict Lowry’s as to the kidnapping and the deliberate shooting, and that since the jury found the defendant guilty only on the lesser offense involving neither kidnapping nor malice, it cannot be said that such evidence prejudiced the defendant’s rights. The argument is not realistic. Going as it did to the heart of the controversy, we cannot safely say that the jury was not influenced to resolve a reasonable doubt as to defendant’s guilt on the offenses charged, by finding him guilty on the lesser offense.
Based upon the conclusion that the trial court erred in admitting the lie-detector evidence, we reversed the judgment, as heretofore noted, and directed that a new trial be granted. That conclusion and disposal of the appeal are here reaffirmed.