after stating the case: There are many exceptions in the record, thirty-two, we believe. The number could easily be reduced to less formidable proportions, without any sacrifice to the plaintiff, if we desired to do so; but as only two or three of them will be examined, we will not undertake the task of reduction, but may be permitted to suggest that counsel, in preparing assignments of error, would greatly simplify and facilitate the work of this Court if after having had the time and the opportunity to carefully examine their exceptions reserved during the hurry of the trial, some of which are necessarily made inadvisedly *51and not upon proper or sufficient study and due deliberation, they would cull out those by a process of intelligent selection or elimination, as the case may require, and thus leave only those of real or supposed merit. This method would not only be of decided advantage to the Court by excluding immaterial matter calculated to divert attention from the main questions and relieve it of useless labor, but it would also greatly conserve the interests of the appellant by presenting his case in a more solid and compact form. We respectfully commend this admonition to our brethren of the bar, in the confident hope, that they will heed it in the future preparation of appeals.
The three exceptions we will consider are these:
1. The alleged expression of opinion by his Honor, when asking the plaintiff’s counsel why they did not call J. Sprunt Newton.
2. The testimony of the witness O. A. Lester as to imitations of the genuine signatures of Adam McArthur made by him, he being an expert engraver, which were used and submitted to the jury, with his explanation and illustration of them, to show that the signature of Adam McArthur was easily simulated, and also similar imitations of Mrs. McArthur’s genuine signature, which were permitted to be used for the purpose of disproving the genuineness of her signature to the notes in dispute. Certain of these imitations' by the engraver were handed to some of plaintiff’s witnesses, among others, A. L. McGowan and S. W. Cooper, D. L. Fort and R. M. Nixon, who had testified to the signatures of the two McArthurs as being genuine. They were shown to the witnesses in an envelope, with a section of the same cut out in the lower right-hand corner, sufficient only for the purpose of exhibiting the signature itself, and not the remainder of the paper. The witnesses were then asked for their opinions as to the genuineness of those signatures, and the court allowed them to be cross-examined in regard thereto, with a view of contradicting or at least weakening their former testimony.
3. The introduction of certain enlarged photographs of the disputed signatures — known as photographic-microscopic reproductions of the same, magnified 154 times by the process of photography — for the purpose of enabling David N. Carvalho and 0. A. Lester to compare or contrast them with the admittedly genuine signatures, which had not been so photographed and enlarged, and thereby show the discrepancies between the two, and otherwise to explain and illustrate their testimony as handwriting experts.
There was a vast deal of testimony in the case, and, as we have stated, numerous other exceptions, some of merit, and some having none, but the foregoing synopsis of three points will suffice for a clear apprehension of the case, so far as we will discuss it.
*52First. We are of the opinion that the remark of the learned and unusually careful judge, in regard to calling J. Sprunt Newton, should not have been made, and was calculated, as an intimation, if not a direct expression, of opinion upon the facts, to prejudice the plaintiff, and is forbidden by the statute, which provides: “No judge, in giving a charge to the petit jury, either in a civil or criminal action, shall give an opinion as to whether a fact is fully or sufficiently proven, such matter being the true office and province of the jury; but he shall state in a plain and correct manner the evidence given in the case and declare and explain the law arising thereon.” There have been numerous decisions upon this statute, and this Court has shown a fixed purpose to enforce it rigidly as it is written. There must be no indication of the judge’s opinion upon the facts, to the hurt of either party, either directly or indirectly, by words or conduct. The judges should be punctilious to avoid it, and to obey the statutory injunction strictly. We are absolutely sure that they fully desire to do so, and their occasional expressions which have come before this Court for review and held to be violations of the statute have evidently been inadvertent, but none the less harmful. The evil impression when once made upon the jury becomes well-nigh ineradicable. Judge. Manly, who was one of the most eminent and just of our judges, said in S. v. Dick, 60 N. C., 440: “He (the presiding judge) endeavored to obviate the effect of his opinion by announcing in distinct terms the jury’s independence of him; hut this was not practicable for him to do. The opinion had been expressed and Avas incapable of being recalled. The object (of the statute) is not to inform the jury of their province, but to guard them against any invasion of it. The dNision of our courts of record into two departments — the one for the judging of the law, the other for the judging of the facts — is a matter lying on the surface of our judicature, and is known to everybody. It was not information on this subject the Legislature intended to furnish, but their purpose was to lay down an inflexible rule of practice, that the judge of the law should not undertake to decide the facts. If he cannot do so directly, he cannot indirectly; if not explicitly, he cannot by inuendo. What we take to be the inadvertence of the judge, therefore, Avas not cured of its illicit character by the information which he immediately conveyed. The error is one of the casualties Avhich may happen to the most circumspect in the progress of a trial on the circuit. When once committed, however, it Avas irrevocable, and the prisoner was entitled to have his case tried by another jury.” And to the same effect did Justice Hoke speak in S. v. Cook, 162 N. C., 586, citing and approving S. v. Dick: “The learned and usually careful judge was evidently conscious that he had probably and by inadvertence prejudiced the prisoner’s case, for he added: Hut the court has no right to express an opinion about *53tbe ease,’ but tbe forbidden impression bad already been made, and as to tbe vital portion of tbe prisoner’s plea, and on authority, tbe attempted correction by bis Honor must be beld inefficient for tbe purpose.” So in S. v. Ownby, 146 N. C., at p. 678, we said: “Tbe slightest intimation from a judge as to tbe strength of tbe evidence or as to tbe credibility of a witness will always have great weight with tbe jury, and therefore we must be careful to see that neither party is unduly prejudiced by an expression from tbe bench which is likely to prevent a fair and impartial trial.” And again in tbe same case: “We know that bis Honor un-guardedly commented upon, tbe testimony of tbe witnesses, but when tbe prejudicial remark is made inadvertently, it invalidates tbe verdict as much so as if used intentionally. Tbe probable effect or influence upon tbe jury, and not tbe motive of tbe judge, determines whether tbe party whose right to a fair trial has thus been impaired is entitled to another trial.” Like views and cautionary requests to tbe judges were stated in Withers v. Lane, 144 N. C., 184: “Tbe learned and able judge who presided at tbe trial, inspired, no doubt, by a laudable motive and a profound sense of justice, was perhaps too zealous that what be conceived to be right should prevail; but just here tbe law, conscious of tbe frailty of human nature at its best, both on tbe bench and in tbe jury box, intervenes and imposes its restraint upon tbe judge, enjoining strictly that be shall not in any manner sway tbe jury by imparting to them tbe slightest knowledge of bis opinion of-the case.” Tbe case of Perry v. Perry, 144 N. C., 330, repeats this injunction to observe tbe mandate of tbe statute, for it is there said: “Any remarks by tbe presiding judge, made in tbe presence of tbe jury, which have a tendency to prejudice their minds against tbe unsuccessful party, will afford ground for a reversal of tbe judgment.” It is very strongly and urgently reiterated in Park v. Exum, 156 N. C., at p. 228, as follows: “Tbe Court has always been swift to enforce obedience to our law which - forbids a presiding judge to express an opinion on tbe disputed facts of tbe trial, and under numerous decisions construing tbe statute, we must bold this remark of bis Honor, in tbe presence of tbe jury and before tbe verdict, to be reversible error.”
We have not cited these cases for tbe purpose of adjudging that plaintiff can avail itself of what was said by the judge during tbe trial of this case, but to again emphasize tbe imperative necessity of keeping tbe. statute steadily in mind and freeing trials of an adverse or injurious intimation of opinion, to tbe end that there may be such a fair and impartial trial as is guaranteed by tbe Constitution and enforced by tbe statute. Many more cases could be cited to illustrate tbe great importance of this matter, but we omit any reference to them for tbe obvious reason that those mentioned are quite sufficient for tbe purpose. It is true, we have *54held that where, by the nature of the case, a party is called upon to prove or disprove a fact material to his success, and the witness who, if anybody, can testify to it, is accessible to him, the failure to produce and examine him is a proper, subject of comment before the jury (Powell v. Strickland, 163 N. C., 393; Goodman v. Sapp, 102 N. C., 477); but this, of course, meant comment of counsel and not of the judge, whose slightest intimation as to whether a fact has been found or not will have the greatest weight with the jury. It is not necessary to decide whether plaintiff was prompt and diligent enough in the protection of its rights to now take advantage of this slip of the judge, as we will order a new trial upon another point. It may be that he has, under the circumstances ; but we leave the question undecided.
Second. Our opinion is that there was error in permitting the witness of the plaintiff to be cross-examined in regard to the signatures which were written or engraved by Mr. Lester and exhibited to them through the aperture made in the envelope, without showing the rest of the paper in which the signature was written, it being called in this case, rather facetiously, though not inappropriately, the “cat-hole test.” These papers should not have been admitted at all. They tended to introduce collateral questions; to multiply the issues, in fact, though perhaps not in form; to divert the minds of the jurors from the real and only question to be decided; to confuse them in their deliberations and to put the witness to an unfair disadvantage and to entrap him unwarily, and also to take the plaintiff by surprise and deprive him of a fair opportunity to know the general nature of the evidence, so that he may prepare to meet it. It tends more to muddy the waters, like the cuttle-fish, than to advance the purpose for which all judicial procedure is adopted, and that is, to conduct the trial so as to establish the truth and to adjudicate rights according to the pertinent and determinative facts, and always to adhere closely to the issue upon which the decision should turn. It was well said in Hardy v. Harbin, 87 U. S. (22 L. Ed.), 378, S. C. Fed. Cases, No. 6059, that, taken at its best, evidence of experts by comparison of handwriting is very unreliable. And in Adams v. Field, 21 Vt., 256: “Those having much experience in the trial of questions depending upon the genuineness of handwriting will not require to be reminded that there is nothing in the whole range of the law of evidence more unreliable, or where courts .and juries are more liable to be imposed upon.” Expert testimony is permissible in such cases, but it is not always the best proof of which the matter is susceptible, and its efficiency and probative force is sometimes easily magnified, and by it the truth is deftly concealed. It, therefore, follows that all proper precautions should be taken to prevent impositions upon the court and jury. Hoag v. Wright, 174 N. Y., 36; U. S. v. Pendergast, 32 Fed., 198.
*55We may remark imprimis, and for the purpose of showing the application of the authorities hereinafter cited, that it was not competent to submit specimens of the admittedly genuine and the disputed papers to the jury for their independent examination, before the passage of our recent statute. The old and strict rule had been somewhat relaxed before then, by allowing the witness to hand the papers to the jury — the standards and the questioned documents — and explain the similarities and the dissimilarities to them, so as to illustrate his own testimony and the reasons for his opinion. Fuller v. Fox, 101 N. C., 119; Martin v. Knight, 147 N. C., 564; Nicholson v. Lumber Co., 156 N. C., 59; Thomas v. State, 18 Texas App., 213. But this was the extreme limit, beyond which the party was not allowed to go. The statute changes the rule, but is so carefully 'and explicitly worded as to exclude by clear implication the examination of any papers but those admitted to be genuine, as standards or models of the true handwriting, and the writings in dispute. It provides: “In all trials in this State, when it may otherwise be competent and relevant to compare handwritings, a comparison of a disputed writing with any writing proved to the satisfaction of the judge to be genuine shall be permitted to be made by witnesses, and such writings and the evidence of the witnesses respecting the same may be submitted to the court and jury as evidence of the genuineness or otherwise of the writing in dispute.” Before the enactment of this law, this Court held, in Tunstall v. Cobb, 109 N. C., 316, and other cases, that the investigation must be restricted to the writings that are genuine and those alleged to be spurious, and that other writings, not of this class, should be excluded in making the comparison. We are aware that there has been some conflict of authority upon this question, but in the midst of all the differing notions about it, we prefer, as we have the choice, not only to abide by the language of our statute as being restrictive in its nature, but to adopt what is said in Rogers on Expert Testimony (2 Ed.), p. 342, sec. 144, which is as follows: “The question has been raised whether it is competent on cross-examination to test the knowledge of the witness by showing him real and fictitious signatures and asking him to say which of them are genuine. In those States where a comparison is only allowed to be made with writings which are admitted to be genuine, it is evident that such a comparison should not be allowed except both parties are agreed which of the signatures are real and which false, for unless so agreed, side issues are raised which complicate the cáse. The rule which excludes writings not admittedly genuine applies with as much force to the cross-examination as to the direct examination. To admit such writings would lead, as well in the one case as in the other, to an indefinite number of collateral issues, and would operate as a surprise to the oppo*56site party, who would not know what writings were to be j)roduced, and therefore could not be prepared to meet them,” citing Howard v. Patrick, 43 Mich., 121, 128; Rose v. First National Bank of Springfield, 91 Mo., 399 ; Massey v. Bank, 104 Ill., 327; Hilsley v. Palmer, 32 Hun., 472 (1884); Van Wyck v. McIntosh, 14 N. Y., 439, which support the text. See, also, Dietz v. Fourth National Bank, 69 Mich., 287, 289. The case of Hilsley v. Palmer, supra, is an interesting one and decisive of this one, if we follow it as authority and adopt its conclusions, and we do not see why we should not, as it is of a very persuasive character, by reason of its clear statement and strong reasoning.
The New York Court of Appeals has considered this question in at least one famous and hotly contested case, People v. Albert T. Patrick, 182 N. Y., 175, in which, with its usual convincing logic, it demonstrates the unfairness of this kind of examination when applied to an opinion witness, that is, one who testifies from actual knowledge of the handwriting in question, or his familiarity with it acquired by business intercourse or other association with its supposed author, and concludes that “it was obviously an unfair test,” and then says: “It is to be observed that Harmon was called upon to testify, by reason of his competency to form an opinion from long acquaintance with the handwriting of the deceased, and not by reason of his being a professional expert in handwriting or penmanship. I think the question is distinguishable, upon the facts, from that passed upon in Hoag v. Wright, 174 N. Y., 36.” The case of Hoag v. Wright, supra, was distinguished from People v. Patrick because in the former the witness was an acknowledged expert in handwriting, and we suppose was, therefore, thought to be able to take care of himself, and to be better acquainted with the “tricks of the trade.” The law is well stated in Andrews v. Hayden's Admr., 88 Ky., 455, 459: “The admissions of these spurious signatures, prepared by an experienced expert, for the purpose of being presented to the witnesses for the plaintiff, was manifestly wrong. They were executed with such skill as to deceive any ordinary observer, or those having no other experience than their familiarity with their neighbor and his handwriting. Such writings should have been excluded because tending to obstruct the proper administration of the law, and deceiving, by the skill in their execution, the minds of honest men. It was neither a just nor legal test, and threw no light on the issue presented.” A strong case favoring this view is Washington Savings Bank v. Washington, 76 Vt., 331, 336, where the Court held that, in cross-examination, it was competent for the defendant, for the purpose of testing the correctness of the witness’s judgment, to show him signatures of the defendant conceded or proved to be genuine; but this was the.limit of comparison.
*57Tbe defendant examined the witness as though, he were an expert, when he was not. His testimony was directly upon the question whether or not the signature upon the note in suit was genuine. But if the witness had been an expert the rule required that a standard of comparison should be established before he could be examined by the use of signatures made for the purpose of the trial. It appears that the signature upon another note in evidence was conceded to be the defendant’s, but the exceptions do not show that it was used in the cross-examination. The witness was required to select from the three papers the genuine signatures of the defendant, and then the papers went to the jury. It was error to permit this course of examination and to allow the papers to be submitted to the jury. Sanderson v. Osgood, 52 Vt., 309; Rowell v. Fuller, 59 Vt., 6881; Costello v. Crowell, 133 Mass., 352; Abbott’s Trial Ev. (2 Ed.), 488, 489.
Speaking of collateral writings, the Court in S. v. Minton, 116 Mo., 605, 614, said': “They were no part of the record in the case, not admitted to be in the handwriting of either one of the defendants, and clearly inadmissible for the purpose of comparison.”
"While we have noted that there is some conflict, we yet think the weight of authority, and the rule of reason, is in favor of the plaintiff’s contention and against the court’s ruling. "W"e add to the authorities already cited the following: Thomas v. State, 18 Texas App., 213, approved by repeated decisions in that Court; Wigmore on Evidence, secs. 1996, 2001, and 2002; King v. Donahue, 110 Mass., 155; Sanderson v. Osgood, 52 Vt., 309; U. S. v. Chamberlain, 12 Blatchford, 390 (25 Fed. Cases, No. 14, 778); Gannt v. Harkness, 53 Kan., 405; Van Wyck v. McIntosh, 14 N. Y., 439; Bank v. Hyland, 53 Hun., 108 (6 N. Y. Sup., 78) ; Rose v. Bank, 91 Mo., 399. And analogous cases are: S. v. Griswold, 67 Conn., 290; Bacon v. Williams, 13 Gray (Mass.), 525; Kirksey v. Kirksey, 41 Ala., 636; Howard v. Patrick, 43 Mich., 128; People v. Murphy, 135 N. Y., 455; Fogg v. Dennis, 3 Humph., 48. In Rose v. Bank, supra, the Court said: “The rule which includes extrinsic papers and signatures is substantially the same in the direct and cross-examination. Papers not relevant as evidence to the other issues are excluded mainly on the ground that to admit such documents would lead to an indefinite number of collateral issues, and would operate as a surprise upon the other party, who would not know what documents were to be produced. The reason of the rule applies to the cross-examination with as much force as to the direct examination. The signatures should have been excluded, whether used to test the witness as an expert or to test his knowledge of the handwriting of the plaintiff.” Wigmore on Ev., at sections 1996, and 2001, 2002, subsec. 2, says: “The specimens, to afford a fairly trustworthy inference, must of course be genuine. But, *58furthermore, the process of proving their genuineness may result (as in the case of the expert’s use of them) in a multiplicity and confusion of issues. The specimens submitted to the jury must be genuine.” 2 Elliott on Ev., sec. 1105, says: “In those jurisdictions where there are no statutes regulating the admission of opinions as to a comparison of handwriting three distinct rules seem to prevail. In a few jurisdictions the rule is that the opinions of experts based on any comparison are improper; in other jurisdictions the rule is that opinions are admissible in case the writings to be compared are in evidence for another purpose and admitted to be genuine; and the third rule is that opinions of experts are admissible as in the rule immediately preceding and, in addition, on writings whose genuineness has been proved on the trial for the express purpose of comparison. The reason given for holding that the only papers that can be used in such an examination of an expert are those which have been brought into the case for another purpose is that such a limitation is necessary in order to avoid the evil of collateral issues, the danger of fraud in selecting specimens, and the danger of misleading the jury.” It is perfectly clear that evidence of this kind may be liable to abuse, and so manipulated or used as to give the jury a false impression as to the real value of a witness’s testimony. It is dangerous, to say the least of it, and may tend more to suppress than to disclose the truth. The imitation may be made so perfect as to mislead a trained expert, and it seems from the books that this has sometimes been done. How, then, can a mere opinion witness be expected to stand the ordeal of such an examination, without the benefit even of a tithe of the expert’s study and experience, if the latter fails? It shows the necessity of confining the examination to the genuine specimens and the disputed writings.
The above authorities, or some of them, also hold that it is incompetent to introduce such imitations of the genuine signature for the purpose of showing how easily it may be forged or counterfeited. Thomas v. State, 18 Texas App., 213. And in Hickory v. U. S., 151 U. S., at p. 306, the Court said: “In the absence of statute, papers irrelevant to the issues on the record were held not receivable in evidence at the trial for the mere purpose of enabling the jury or witnesses to.institute a comparison of hands. Bromage v. Rice, 1 Car. and P., 548; Doe v. Newton, 5 Ad. and El., 514; Griffiths v. Ivory, 11 Ad. and El., 322; 1 Greenleaf Ev., sec. 580. The danger of fraud or surprise and the multiplication of collateral issues were deemed insuperable objections, although not applicable to papers already in the cause, in respect of which, also, comparison by the jury could not be avoided.”
Third. The enlarged photographs of the disputed writings should not have been used, at least without proving, by the man who made them, how and under what conditions they were taken by him, so as to let the *59jury finally decide, from the facts, whether they are exact reproductions. It was held in Bank v. Wisdom, 11 Ky., 148, that such enlargements of the signatures should be authenticated by the photographer. In Tome v. Parkersburg R. R. Co., 39 Ind., 36, it is said, with reference to this subject : “The testimony of the photographer comes within the same principle as that of Paine. It was offered to establish the forgery of the certificates in controversy, by comparing them with copies (obtained by photographic processes, either magnified or of natural size) of certain signatures assumed or admitted to be genuine, and pointing out the differences between the supposed genuine and the disputed signatures. As a general rule, in proportion as the media of evidence are multiplied, the chances of error or mistake are increased. Photographers do not always produce exact facsimiles of the objects delineated, and however indebted we may be to that beautiful science for much that is useful as well as ornamental, it is at least a mimetic art, which furnishes only secondary impressions of the original, that vary according to the lights and shadows which prevail whilst being taken.” This principle has been sanctioned by the following authorities: Hynes v. McDermott, 82 N. Y., 41, 50; Oil Co. v. Bank, 34 Texas, 555; Railway Co. v. Bank, 56 Ohio St., 385; Bank v. F. S. S. and G. S. F. R. Co., 137 N. Y., 242; Houston v. Blythe, 60 Texas, 512; Eborn v. Zempleman, 47 Texas, 513; Howard v. Bank, 189 Ill., 577; Smith v. Martin, 135 Cal., 251; Healy v. Bank, 160 Ill. App., 637; Erb v. G. R. Railway Co., 42 U. C. Q. B., 42; B. and O. Railway Co. v. Wilkins, 44 Md., 37; Harrison and Brown, Trustees, v. A. and E. R. R. R. Co., 50 Md., 513; W. M. R. R. Co. v. Bank, 60 Md., 42, 48; Burrows v. Klunk, 70 Md., 460; Barabas v. Kabat, 86 Md., 34; Lamm v. Homestead Assn. Co., 49 Md., 241; Ins. Co. v. Swain, 100 Md., 575; Moses v. Bank, 111 U. S., 169.
The Court said in Hynes v. McDermott, supra: “It would be carrying the matter much farther to permit an expert to compare photographic copies of signatures, and therefrom to testify as to the genuineness of a disputed signature. Ye may recognize that the photographic process is ruled by general laws that are uniform in their operation, and that almost without- exception a likeness is brought forth of the object set before the camera. Still, somewhat for exact likeness will depend upon the adjustment of the machinery, upon the atmospheric conditions, and the skill of the manipulator. And in so delicate a matter as the reaching of judicial results by the comparison of writings through the testimony of experts, it ought to be required that the witness should exercise his acumen upon the thing itself which is to be the basis of his judgment; and still more, that the thing itself should be at hand, to be put under the eye of other witnesses for the trial upon it of their skill. The certainty of expert testimony in these cases is not so well assured as that we can *60afford to let in tbe hazard of errors or differences in copying, though it be done by howsoever a scientific process. Besides, as before said, there was no proof here of the manner and exactness of the photographic method used. It was right not to receive Loader’s evidence as that of an expert.” The courts that have allowed this kind of evidence have generally held that both the admittedly genuine signature and the one in dispute shall be alike photographed and the testimony of the photographer taken as to the accuracy of the method pursued by him and the results obtained. U. S. v. Ortez, 176 U. S., 422, and other cases supra.
It is not necessary now to say more upon this question, as the deficiencies in the proof may be easily supplied. All we decide is that the photographic copies were not admissible in the then state of the evidence, and no more.
There are other errors -assigned, which seem to have merit in them, but we will refrain from any further reference to them, as they may not occur again. For those already indicated, a new trial is ordered.
New trial.