delivered the opinion of the Court.
Bethel E. Brooke, Jr., was convicted in the Denver district court of murder in the first degree. The jury determined his sentence to be life imprisonment. To this verdict and sentence of the jury on which the court entered final judgment after hearing and denying a motion for new trial, Brooke assigns error.
After writ of error was issued and pending in this court and before the reporter’s transcript and record on error could be filed here, Brooke underwent two polygraph tests, familiarly known as “lie detector” tests. On the basis of the outcome of these tests, he filed in the trial court a motion to vacate the judgment and sentence, labeling the motion in the nature of “Writ of Error Coram Nobis.”
Evidence against Brooke was entirely circumstantial. Brooke’s defense was that the deceased had committed suicide in his (Brooke’s) trailer home. Strong links in the chain of circumstantial evidence tending to bind Brooke to the crime and tending to refute his story of suicide was the evidence of Lt. Moomaw of the Denver Police Department. A ballistic and firearms expert, the police lieutenant testified that a paraffin test (designed to reveal whether the person tested had within recent hours fired a gun) proved negative as to decedent. The test was made on the hands and forearms of decedent. Because the paraffin failed to show telltale nitrates and nitrites, the lieutenant gave his opinion that the decedent had not fired or discharged a revolver and therefore *390could not have shot herself. The lieutenant also testified that he asked Brooke to take the paraffin’lest, and that Brooke refused, giving ás his reason that, his attorney had advised him to refuse the test. ‘ ’¡'
Only two of the points of error raised by Brooke will be commented upon. Other points raised have no merit and will not be discussed. A determination of the questions involved here do not require detailing or summary of the evidence and the circumstances surrounding the death, so we omit those details.
First Question to be Determined:
Was it prejudicial error for the court to admit the testimony of a police officer that defendant had been requested to take a paraffin test and had refused to do so on advice of his counsel?
This question is answered in the affirmative.
Whether it is proper or improper for the jury to be told that the accused, while under arrest, had refused to take a test requested of him by the police is a question of first impression in Colorado. The arguments presented to uphold the ruling of the court follow a rather circuitous and devious route, and the logic employed to reach the conclusion that the admission of such evidence should be sustained is of doubtful validity.
The line of authorities employed in support of the case for the People approaches the problem on the proposition that a constitutional privilege against self-incrimination such as in Article II, section 18, of the Colorado Constitution, applies only to testimonial compulsion, and that therefore any physical evidence obtained from the defendant, either against his will or while unconscious, does not violate the privilege against self-incrimination. Tuttle v. People, 33 Colo. 243, 79 Pac. 1035; Block v. People. 125 Colo. 36, 240 P. (2d) 512 (certiorari denied 343 U.S. 978; 344 U.S. 848); Kallnbach v. People, 125 Colo. 144, 242 P. (2d) 222; Ingles v. People, 92 Colo. 518, 22 P. (2d) 1109; Vigil v. People, 134 Colo. 126, 300. P. (2d) 545.
*391The best known and most frequently cited tests for physical evidence deemed admissible in the trial of a defendant are for intoxication through means of the breath, of samples of blood or urine; blood tests to determine identity; examinations for sanity; tests of fingernail scrapings taken from defendant, fingerprints, palm-prints or footprints and photographs.
The rationale upon which the various physical tests enumerated are permitted is best expressed in 28 A.L.R. (2d) 1138, wherein the annotator discusses People v. Sallow, 100 Misc. 447, 165 NYS 915. Referring to the taking of fingerprints in particular and the obtaining of physical evidence from a defendant in general, the court reasoned that,
“ * * * No volition, that is no act of willing, on the part of the mind of the defendant is required. Fingerprints of an unconscious person or even of a dead person are as accurate as are those of the living. * * * By the requirement that the defendant’s fingerprints be taken, there is no danger that the defendant will be required to give false testimony. The witness does not testify — the physical facts speak for themselves; no fears, no hopes, no will of the prisoner to falsify or to exaggerate could produce or create a resemblance of her fingerprints or change them in one line, and, therefore, there is no danger of error being committed or untruth told. * * * ” (Emphasis supplied.)
Thus the one thing all of the tests have in common is that they have been proven irrefutably accurate. There is no room for doubt now that fingerprints, palmprints and footprints are positive means of identification. Chemical analysis of the blood or urine to determine the content of alcohol is deemed acceptably accurate.
In contrast, the paraffin test as described by Lt. Moomaw enjoys no such reputation for accuracy. Nevertheless the court, upon the urging of counsel for the People, classified the paraffin test in the same category as-fthe others. From this it is argued, and the trial court *392assumed, that if the results of the paraffin test be; admissible, the refusal of the defendant to take the test was likewise admissible. This assumption on the part of the court was error. Lt. Moomaw, somewhat contradictory of his conclusion, gave an inkling as to the unreliability of the test in the following question and answer:
“Q. I believe you stated, testified that the paraffine test is not a positive test. It isn’t always accurate in showing just what the substance is; is that correct? A. The nitrate test, itself, is not specific for powder bums, that is correct.”
We have pursued the question further and find in the Journal of Criminal Law and Criminology, vol. 46 (1955-1956), p. 283, a revealing discussion of the unreliability of the dermal nitrate test. Following are some of the pertinent comments:
“It is doubtful that anyone would have sufficient trust in the dermal nitrate test to bring a criminal charge or institute criminal proceedings on the strength of the findings of this test alone. Invariably, there is a mass of other evidence already available, and the dermal nitrate test is done with the hope of corroborating or strengthening the already known facts. Although the series done is admittedly a small one, there is the suggestion that even in experienced hands the test is subject to 13% gross error. For one who might only perform the test very occasionally, the error might be even greater. Such percentage of gross error as above stated makes the test totally unsafe for use where weight is given to the test in determining guilt and a conviction would result in severe penalty.
“The test itself can determine with reasonable accuracy, only whether or not nitrates were present, but this is not what we need know. We actually want to establish whether or not a gun was fired and from this viewpoint, inconclusive findings must be considered as failures. (E.g., if 100% of all these tests were inconclusive, then *393we' would know nothing of whether or not a gun was fired, and this test was 100% a failure.) Any lesser percentage of inconclusive findings merely means a smaller proportion of failure to serve its purpose.
“In the sum total of the cases studied 75% were inconclusive. Coupled with the 13% gross errors, the test becomes less than worthless. * * * ” (Emphasis supplied.)
“The paraffin glove test for dermal nitrates is neither sufficiently certain nor subject to such scientific accuracy as to justify its routine use in establishing whether a suspect or deceased did or did not fire a gun.
“This opinion is based on the evident inaccuracy, even in relatively good hands, and on the fact that the inconclusive findings noted in most of, the tests done can too easily be used to offset the weight of other well documented facts of a case at trial.
“The test is performed in the San Francisco Coroner’s Office only if an investigating homicide inspector requests it be made despite our advice and cautioning as to its failings.”
We hold, therefore, that the result of a paraffin test, rather than being placed in the category of the accepted tests, has the same reputation for unreliability as the lie detector test. The authorities, therefore, which deal with the inadmissibility of the results of the lie detector tests, or which reject evidence of a refusal of an accused person to take such a test, are more nearly in point. See Ernest Joseph Mills v. The People of the State of Colorado, No. 18,869, decided this date. Also 23 A.L.R. (2d) 1306.
To paraphrase the authorities which have considered the efficacy of the lie detector, we hold that the paraffin test has not gained that standing and scientific recognition or demonstrated that degree of reliability to justify courts in approving its use in criminal cases. Therefore, because of its unreliability it was error to admit the results of the paraffin test as conducted on the *394body of the decedent, and if Brooke had submitted to the test the result thereof also would not have been admissible in evidence. This being true, the fact that-the defendant refused to take the test was likewise inadmissible. The inference to be drawn from both the testimony of the result of the paraffin test on the decedent, as well as those which probably were drawn by the jury from the recital of Brooke’s refusal to take the test, were prejudicial. :T'
Second Question to be Determined:
Did the court err in refusing to vacate the judgment and sentence on the motion in the nature of Writ of Error Coram Nobis?
This question is answered in the negative.
In so holding we do not affirm the action of the court in ruling on its merits. Although the end result was the same, it should have dismissed the motion instanter. At the time of the filing of the “Motion to-Vacate” the trial court had accepted the verdict of the jury, had denied the motion for new trial, had rendered a judgment of guilty on the jury verdict, and had sentenced the prisoner to the state penitentiary for the remainder of his natural life. Because the court had acted finally and conclusively in the matter and a writ of error had issued out of, and was pending in this court, jurisdiction of the case was vested entirely in this court, and the trial court had been divested of all jurisdiction in the cause other than to prepare and certify the record to this court in accordance with the applicable statute. If when the prosecutor joined in the motion it was intended to stipulate that the trial court could hear it, we call attention to the authorities which hold that jurisdiction can never be conferred by consent or otherwise. Geisler v. People, 135 Colo. 121, 308 P. (2d) 1000; United States Fidelity and Guaranty Co. v. Industrial Commission, 99 Colo. 280, 61 P. (2d) 1033. The court, having no jurisdiction to consider the motion, should have dismissed it without hearing.
*395The judgment is reversed.
Mr. Chief Justice Knauss dissents.
Mr. Justice Sutton does not participate.