Opinion by
Mr. Justice Teller.The plaintiff in error was convicted of the murder of William Dick in the County of Huerfano, and sentenced to death.
He alleges that there is error in the record in several matters, including a death sentence on a conviction on circumstantial evidence alone.
■ Section 1624, R. S. 1908, contains this provision:
“Nor shall any person suffer the death penalty who shall have been convicted on circumstantial evidence alone. ’ ’ '
Plaintiff in error alleges that his case falls within the terms .of this statute.
The State contends that the testimony of one Kacklis that the accused confessed to him that he and another *419committed the crime is direct, and not circumstantial evidence, and serves to take the case out of the statute cited.
It is not claimed that there was any other evidence which was not circumstantial. The question then is this: Is testimony of a confession circumstantial evidence within the meaning of that term as used in the statute in question?
In answering this question we are, of course, to keep in mind the purpose of the statute so far as it may be ascertained.
We must suppose that the law-makers were aware of the fact — as it is common knowledge — that not infrequently persons guiltless of crime have been convicted and punished; and that there is a much greater likelihood of such a result where a conviction is had’without direct and positive evidence of the criminal act.
It is true, as has often been said, that while witnesses may lie, facts do not; and that a fact may be established as clearly by inferences from circumstances proved, as by direct evidence; but that does not meet the objection that the more remote the evidence is from the matter to be established, the longer is the chain of inferences, and the greater is the possibility of mistake in the final conclusion.
“While facts do not lie, their narrators certainly may, and often do, and, further, inferences from facts may be false. * * * In quality of probative force, direct evidence necessarily has an inherent advantage, there being one less chance for error, and the' equation of human fallibility. In case of direct evidence, if the facts testified be true, the ultimate proposition is established. In case of circumstantial evidence, the facts testified to may be entirely accurate, and the consequent inference, which establishes the proposition in issue, be utterly fallacious. This is an essential difference in *420kind: Caeteris paribus, direct evidence is stronger.”— Chamberlayne’s Best on Evidence, sec. 293.
• Despite the remarks of judges and text-writers in defense of circumstantial evidence, it is generally regarded, by laymen and members of the legal profession, as less satisfactory than direct evidence to the precise matter to be proved. Doubtless this statute was intended to authorize the imposition of the death penalty in those cases only in which the guilt of the accused is established by such evidence as renders it least probable that a mistake has been made.
Kacklis, who testifies to the alleged confession, nowhere states that the accused said he had killed William Dick. His testimony is that the accused said he “killed the paymaster-, ” and the witness adds that “paymaster meant Dick.”
Does the testimony of Kacklis meet the above stated requirement? It is open to the objection that before the jurors could determine from it that the defendant was guilty, they must consider, First, whether or not the 'witness spoke the truth; whether or not, if truthful, he gave the alleged confession accurately and in full; and whether or not he was correct in his conclusion that the term “paymaster” meant the deceased. Second, they must then determine, if they believed the witness, whether or not the accused spoke the truth; this latter, because there are many instances of confessions which were wholly without foundation in fact. Extra judicial confessions are admitted in evidence upon the ground that a rational being will not make admissions prejudicial to his interests and safety, unless urged thereto by the promptings of conscience to tell the truth. They are exceptions to the rule which excludes hearsay testimony; and, being in fact hearsay testimony, it is universally held that they *421must be received with great caution. G-reenleaf on Evidence, Sec.. 214, 15th Ed.
Professor Wigmore points out in his discussion of this subject that the apparent contradictions among the authorities as to the value of confessions as evidence .is due to a failure to discriminate betwéen the probative effect of a confession when dearly proved, and the force of the evidence to prove it. The.latter is open to the objections to hearsay, and it is to the latter that the authorities refer when speaking of the dangers from this class of evidence. He says:
“ Assuming the confession as an undoubted fact, it carries a persuasion which nothing else does, because a fundamental instinct of human nature teaches each one of us its significance. But how do we get to believe in the fact of a confession having been made? Always and necessarily by somebody’s testimony. And what is our experience of that sort of testimony on which we are asked to believe that a confession was made? A varying and sometimes discouraging experience. Paid informers, treacherous associates, angry victims, and overzealous officers of the law, — these are the persons through whom an alleged confession is often, perhaps oftenest presented; and it is at this stage that our suspicions are aroused and our caution stimulated. Suppose a judge is offered, from the lips of a single witness, a detailed and complete avowal of guilt, attributed to the accused; and suppose the accused denies absolutely the fact of confession; suppose the judge now to think to himself, £Here is a confession which, if authentic, would make this man’s guilt clear beyond doubt. But do you expect us to take it as authentic, against his denial, on the word of this man alone, who has such and ■such strong motives for inventing it or for misinterpreting what was said? Must we not listen to him with the *422greatest doubt and suspicion?’ Then, would it not be natural for tbe judge, in commenting on such evidence to the jury, to say: ‘What you have heard here from this man about a supposed confession is to be taken with caution; for that is the weakest and most suspicious kind of evidence?’ This is a natural and proper attitude, and it is precisely that of the authorities above quoted. They were thinking, not of the confession as evidence of the act, but of the testimony to the alleged confession. Take, for instance, the phrase above of Mr. Justice Foster’s, which has been quoted again and again (with and without acknowledgment) in the records of the profession for a century and a half, in the mangled and misleading form that ‘confessions are the weakest and most suspicious of all evidence.’ Why did he so regard them? Not because of their own evidential weakness; but for the following reasons:
‘Proof may be too easily procured; words are often misreported — whether through ignorance, inattention, or malice, it mattereth not to the defendant, he is equally affected in either case; and they are extremely liable to misconstruction; and withal, this evidence is not in the ordinary course of things to be disproved by, that sort of negative evidence by which the proof of plain facts may be and often is confronted.’
In other words, the suspicion that he has found it necessary to entertain is directed entirely to the work of proving an alleged confession. This is the reason above suggested as the real cause of the distrust; we are ready enough to trust the confession if there really was one, but we are going to doubt and suspect for a long time before we accept it as a fact. * * *” Wig-more on Evidence, sec. 866.
This language is peculiarly applicable to the facts in this case, in view of the admissions of the witness *423Kacklis that he had heard that there was a large reward offered, and wanted to get some of it: and, further, from the fact that there was evidence of enmity between the witness and the defendant before they left Greece.
The State relies upon a number of cases in which it is claimed that it has been held that confessions are not circumstantial evidence; but an examination of them shows that they all consider the question as it relates to the necessity for an instruction on circumstantial evidence. By inference only do these cases hold as the State contends. In none of them is there any discussion of the question, and, in the few cases in which authorities are cited to the point, on examination, they,- too, are found barren of reasons for the holding. We do not, therefore,! regard the cases cited as precedents forming a rule of decision, and feel free to determine the question on principle.
There seems to be no ground for the statement that the testimony of a witness to the fact of a confession is direct or positive evidence of the thing which is said to have been confessed. It is direct testimony that there was a confession, but not that the facts confessed existed. This distinction is recognized by the most eminent writers on the subject.
Mr. Greenleaf says:
“In trials of fact it will generally be found that the factum probanchm is either directly attested by those who speak from their own actual and personal knowledge of its existence, or is to be inferred from other facts satisfactorily proved. * * * In the former case the proof applies immediately to the factum probanchm, without any intervening process, and it is, therefore, called direct or positive testimony.”
He adds that evidence of the second class is-called circumstantial.
*424Apply this language to the case at bar, and we find that there is direct evidence that a confession was made, but not to the factum probandum: i. e., that the defendant committed the homicide.
The defendant by confessing in open court would have given direct testimony to that fact, but the testimony of Kacklis was not that defendant killed the de- < ceased, but that he said he had killed the paymaster.
In Best on Presumption, sec. 12, it is said :•
“But, when any fact in dispute has neither fallen under the cognizance of the senses of any person, nor is stated in terms in any authentic writing, so that its truth or falsehood can only be deduced by special inference from facts proved directly, the evidence of that fact is said to be circumstantial.”
And again, in section 246 he says:
“When the existence of any fact is attested by witnesses, as having come under the cognizance of their senses, or is stated in documents, the genuineness and veracity of which there seems no reason to question, the evidence of that fact is said to be direct, or positive. By circumstantial evidence, on the contrary, is meant, that the existence of the principal fact is only inferred from one or more circumstances which have been established directly. ’ ’
The eminent Chief Justice Shaw said:
“Direct or positive evidence is when a witness can be called to testify to the precise fact which is the subject of the issue on trial; that is, in the case of homicide, that the party accused did cause the death of the deceased. Whatever may be the kind or force of the evidence, this is the fact to be proved. * * * The advantage of positive evidence is that it is the direct testimony of a witness to the fact to be proved, who, if he *425speaks the truth saw it done, and the only question is whether he is entitled to belief. * * * ”
He then speaks of circumstantial evidence as necessary where no witness can testify directly to the fact to be proved, its force being derived from a recognized connection between the facts proved and the main fact sought to be proved. Of this class of evidence he says:
4 4 The disadvantages are that a jury has not only to weigh the evidence of facts, but to draw just conclusions from them; in doing which, they may be led by prejudice or partiality, or by want of due deliberation and sobriety of judgment, to make hasty and false deductions: a source of error not existing in the consideration of positive evidence.” Commonwealth v. Webster, 5 Cushing 310-312, 52 Am. Dec. 711.
The distinction thus pointed out is so clear that it. is difficult to understand how it could have been overlooked, as it sometimes has been. With it in mind, we must conclude that the legislators in passing the law under consideration must have intended to include in the term 44circumstantial evidence” all that is not direct; all that carries with it those liabilities of error which the authorities have long recognized and discussed. Any other conclusion would render the law ineffective to accomplish its humane and evident purpose. We are, therefore, of the opinion that the testimony of a confession by the defendant is circumstantial evidence; and, there being in this case no direct evidence that the defendant killed the deceased, the jury was not authorized to fix the penalty at death. The court erred in not granting a new trial on that ground.
The special prosecutor put many questions to the witness Kacklis which clearly violated the rule against leading questions, in that they assumed the existence of material facts of which there was no evidence, and sug*426■gested to the witness further facts to which he then testified.
The objections to them should have been sustained; but, as these errors may not occur in another trial, we do not consider them in detail.
Upon the question of the overruling of the motions for a continuance, and for a change of venue, we do not find it necessary to express an opinion. The grounds for the prior motion no longer exist, and if a second motion for a change of venue be made, it must, of course, be determined on the facts as they exist at the time such motion is filed.
The judgment is reversed.
Judgment reversed.
Chief Justice Gabbert and Mr. Justice Garrigues dissent.