Three errors are alleged: first, that the intent to rape was not proven; second, that the evidence of *498the assault on Mrs: Casper was inadmissible; third, that the charge of the court was incorrect.
1. We cannot consider the first assignment of error, because the bill of exceptions is not certified to contain all of the evidence. It is true that a paper has been filed with the clerk a few days since, purporting to be an additional or amended certificate to the bill of exceptions, bearing the signature of the trial judge, which states that the bill of exceptions contains all of the testimony in the case, and that the failure to so certify in the original certificate was an oversight; but we cannot consider it because it is not attached to the bill of exceptions and does not even come here with the certificate of the clerk of the trial court. The bill of exceptions is to be made and fully completed in the court below, and then becomes a part of the record of that court, and comes to this court (if at all) in response to an appeal under the certificate of the clerk of the trial court. It cannot be made or changed in this court, except, at least, by agreement of the parties.
2. Was the evidence of the separate assault committed on Mrs. Casper an hour before the assault upon Mrs. Montgomery competent evidence ? It is freely admitted by the state as a general rule that upon a prosecution for 'one offense evidence of the commission of another and separate offense is not admissible, but the claim is made that the evidence was admissible in this case for the purpose of proving intent. The rule that, where intent must be proven, other crimes of like nature, which are so intimately related to the act in question as to show a common purpose ora continuity of purpose in all, may be shown upon the question of motive or intent, or to repel the inference of accident, is well recognized. State v. Miller, 47 Wis. 530; 1 Jones, Ev. §§ 143, 144; Zoldoske v. State, 82 Wis. 580. The rule is one which is not always easy to apply, and it is manifestly one which needs to be most carefully applied and guarded, or it is *499likely to result in many convictions based largely upon proof of the commission of crimes not charged in the information,— a result which our criminal law does not contemplate. In the case of Proper v. State, 85 Wis. 615, which was a prosecution for rape upon a girl, proof that the accused had previously got into bed with the prosecutrix and another girl named Emma, and had sexual intercourse with the other girl, was held proper on the sole ground that such an act was an indecent assault upon loth girls; but it was said in the opinion by the late Mr. Justice Pinney:
“We do not suppose that evidence that the defendant had committed adultery or been guilty of acts of improper familiarity with the girl Emma at another time and • place would be competent evidence on the trial of the present issue.”
While this remark was oliter in that case, it is believed that it expresses the rule which has been generally approved by the authorities, namely, that in prosecutions for crimes •of this nature evidence of previous attempts by the accused to commit the crime upon the same person is admissible on the question of intent, but that evidence of attempts to commit the crime upon other persons is not admissible. Mr. Wharton says: “Evidence of prior sexual assaults on the prosecutrix is admissible on an indictment for rape, though not of rapes on other persons.” Wharton, Cr. Ev. (9th ed.), § 46. If this be the rule as to rape actually committed, it would seem to apply to mere unsuccessful assaults, where the purpose does not clearly appear, with equal, if not greater, force. There may be a number of motives for 1/he commission of an assault besides rape,— such as robbery, revenge, and the like,— and it can hardly be logically argued that because a man has assaulted two women, although the assaults were both on the same day, the same motive impelled both assaults. We are therefore of opinion that the evidence in question was erroneously received.
*5003. The trial judge charged the jury upon the subject of reasonable doubt as follows:
“The court instructs the jury as a matter of law that in considering the case they are not to go beyond the evidence to hunt up doubts, nor must they entertain such doubts as are merely chimerical, — matters of conjecture. The doubt, to justify acquitted, must be reasonable and must arise, either negatively or positively, from a candid and impartial investigation of all the evidence in the case; and unless it is such that, were the same kind of doubt interposed in the graver transactions of life, it would cause a reasonable'and prudent man to hesitate and pause, it is insufficient to authorize a verdict of not guilty. If, after considering all the evidence, you can say you have an abiding conviction of the truth of the charge, you are satisfied beyond a reasonable doubt.”
The accused duly excepted to all of this instruction except the last sentence, and he now claims that the part so excepted to is misleading, and really places upon the defendant the burden of raising a reasonable doubt. This charge in almost exactly the words here used has been approved by the courts of last resort of at least four states: Miller v. People, 39 Ill. 457; Spies v. People, 122 Ill. 1; Polin v. State, 14 Neb. 540; Maxfield v. State, 54 Neb. 44; Boulden v. State, 102 Ala. 78; State v. Pierce, 65 Iowa, 89. A very similar instruction was approved in Minich v. People, 8 Colo. 440. The proposition is stated also in 2 Thomp. Trials, § 2475, and we have found no authorities which can be said to directly disapprove it. In the face of such an array of very respectable authority it seems somewhat presumptuous to declare the instruction erroneous, and yet we feel compelled to do so for reasons to be stated.
In the first place, the instruction seems very unfortunately framed and worded, in this, that it apparently assumes that the jury are to start upon their deliberations upon the basis that conviction is to be the result unless a reasonable doubt has been proven. It reverses the customary and approved manner of putting the question before *501the jury. Ordinarily, a jury is told that, in order to justify conviction, the evidence must prove guilt beyond a reasonable doubt. This form of instruction calls attention to and emphasizes the leading principle of our criminal jurisprudence, namely, that a man is deemed to be innocent until his guilt is proven; that the jury must find evidence to justify the conviction. But the form of instruction used in the present case comes very near intimating to the jury that guilt is the natural presumption and that they must find a doubt from the evidence in order to justify the acquittal. If it be calculated to, leave such an impression on the minds of the jury, it is certainly misleading and prejudicial.
There is another ground, however, upon which we think it is very clear that the instruction was erroneous. This court held in Anderson v. State, 41 Wis. 430, that in passing on the evidence in a criminal case the jury should be instructed (if instructed at all on that line) that in performing their duty they should “scrutinize the testimony with the utmost caution and care, bringing to that duty the reason and prudence which they would exercise in the most important affairs of life,— in fact all the judgment, caution, and discrimination they possessed”; and that it was reversible error to charge that they “ should be guided by that reason and prudence which govern you in the ordinary conduct of your affairs.” In so deciding this court followed the great current of authority. See, in addition to the cases cited in in the Anderson Case, 2 Thomp. Trials, §§ 2470 et seq. This doctrine was approved in Emery v. State, 92 Wis. 146, where the test given to the jury was that the judgment and conscience must be convinced to an extent such as would lead “a careful and prudent man to act affirmatively in important matters of his own,” and it was held that it was distinct error to use the words “ important matters of his own ” instead of of the words “most important affairs of life.” The rule was recognized as correct .in the case of Buel v. State, *502104 Wis. 132, and it was applied to tbe definition of reasonable doubt in the case of Butler v. State, 102 Wis. 364. In this latter case the trial court had defined reasonable doubt as being such a doubt as “would govern and control a reasonably prudent man, and deter him from acting in Ms own most important affairs and concerns of life,” and it was objected that the expression “ Ms own most important affairs ” was not equivalent to “the most important affairs of life but it was held that the words were fairly legal equivalents in the connection in which they were used. It was assumed in that case (and with entire propriety) that a juryman must use the same degree of care, judgment, and prudence in determining whether there was a reasonable doubt as in determining whether any given fact was proven; in other words, that his mental attitude in the consideration of every question presented to him as a juryman throughout the case must be the same, namely, that of meeting and solving it with all the care, judgment, and prudence of which he is capable.
These cases have settled this principle fully and completely in this state, and, as we believe, upon sound, logical, and just lines. Were we to approve of an instruction which authorizes the jury to use a less degree of care and caution in passing upon one of the questions which arise in every criminal case, we should confuse and make uncertain legal principles which have been supposed to be certain and definite,— a thing which should not be done except for very good cause. The words “ graver transactions of life ” may mean many things, but it cannot be argued that they are the legal equivalent of the words “ most important affairs ” of life. A man may not be able to decide satisfactorily as to what matters constitute his “graver transactions,” but there is little or no doubt in any man’s mind as to what the “ most important affairs” of his life consist of.
We must, therefore, in view of the settled position taken *503by this court, as above stated, hold that the instruction given by the court in this case, which uses the words “ graver transactions of life,” instead of the words “ most important affairs of life,” or their equivalent, is erroneous on that account.
Although the correctness of the definition of reasonable doubt which was approved in Butler v. State, supra (102 Wis. 364), is not now before us for decision, we deem it proper to call attention to the fact, as previously stated in this opinion, that the only ground on which the definition was assailed in that case was the ground that the words “ his own most important affairs ” were not equivalent to “ the most important affairs of life,” and that the decision is authority on that question only. Attention is now called to this for the reason that it may be a matter of serious doubt whether in other respects the definition is correct. The definition, as given, is:
“ Such a doubt as would govern and control a reasonably prudent man, and deter him from acting or deciding in his own most important affairs and concerns of life.”
There is much reason to say that this definition is too strong a definition to be accurate. A doubt which would govern and control a prudent man and deter him from acting must be quite an able-bodied doubt, and amount practically to a mental conviction. As said by the supreme court of Pennsylvania in Comm. v. Miller, 139 Pa. St. 77:
“ Our actions are determined by a preponderance of considerations. "We doubt, hesitate, examine, balance the argument for and against the given 'action, and act as the preponderance indicates. A doubt that would control our actions in the important transactions of life would be one that was so strong as not to be overcome by the balancing process. Such a doubt would be practically an unconquerable one. It would lead us not simply to refrain from acting, but to act.”
This was said in condemning as erroneous an instruction to the effect that reasonable doubt “is such a doubt as *504would influence or control yon in your actions in any of the important transactions of life,” and it was, in substance, suggested that the proper test would be whether the doubt would cause a man to “ pause or hesitate ” rather than “control his action.” The reasoning seems to’ us very cogent. If a doubt, to be reasonable, must be such a one as to control action, it seems that it must have passed beyond the realms of doubt and become a belief. We all know that we frequently act in our most important matters when we still have doubts of the expediency or wisdom of such action, but we should be likely to repel the intimation that such doubts were unreasonable. In a word, it seems clear that there may be a doubt which will cause the mind to pause and hesitate, but will not necessarily control our actions in our most important affairs, and yet it may be a reasonable doubt and one which if interposed in a criminal case would call for the acquittal of the accused. It would certainly be far safer for the trial courts, in charging on the subject of reasonable doubt, to define it as a doubt which would cause a reasonable and prudent man to “ pause or hesitate ” than as a doubt which would “ control or govern ” his actions, or “ deter ” him from acting, if any definition on this line is attempted.
By the Court.— Judgment reversed, and action remanded for a new trial. The warden of the state prison will deliver the plaintiff in error to the sheriff of Douglas county, who is directed to keep the said McAllister in his custody until he is duly discharged therefrom or until otherwise ordered according to law.