Summary
rejecting use of (unsubstantiated) statistical evidence demonstrating that defendant's typewriter was used to forge a document and stating that probabilistic testimony could not be used to establish a historical fact
Summary of this case from U.S. v. HanniganOpinion
Argued December 15, 1914
Decided February 5, 1915
A.M. Mills and Charles D. Thomas for appellant. William E. Farrell, District Attorney ( Frank A. Schmidt and James H. Greene of counsel), for respondents.
Defendant was convicted of a violation of section 810 of the Penal Law, with having on the 6th day of February, 1911, in the trial of an action in the Supreme Court, entitled Bennett v. Iron Clad Manufacturing Company, offered in evidence, as genuine, a certain document bearing date December 1st, 1904, knowing that the same had been forged and fraudulently altered by the insertion of the words "the same" therein.
The unanimous decision of affirmance of the judgment of conviction by the Appellate Division precludes this court from a consideration of the weight of evidence. We deem it essential, however, in view of the determination of this appeal to refer to the principal facts relied upon by the People.
The defendant had been for many years a member of the legal profession. He was senior member of the firm of Risley Love; engaged in the practice of law in the city of Utica. Risley Love were attorneys for one Lewis Bennett, who was the patentee of a mechanical patent for improvements in metal baskets, and also the patentee of a design for metal baskets. On or about October 14, 1896, Bennett entered into a contract with the Iron Clad Manufacturing Company by the terms of which he granted to that company the sole and exclusive right, license and liberty of making, using, selling and vending to others to be used and sold, baskets embodying the inventions described, and baskets embodying improvements covered by an application made by him. The company covenanted and agreed to pay to Bennett as a license fee a royalty on each basket embodying the inventions and designs made and sold by it during the continuance of said agreement; to keep full and correct books of account which would be open to the inspection of Bennett and which would-show the number of baskets made and the number sold by it, and to furnish to Bennett statements of accounts semi-annually showing the number of baskets made and sold for the preceding six months, and at that time to pay to Bennett the royalties due under said contract.
On or about December 1, 1901, an action was commenced in the Supreme Court by Risley Love, as attorneys for Bennett, to recover for baskets made and sold by the company; to have the contract declared forfeited, and for a return of certain tools and machinery. Judgment for the relief demanded, and a money judgment for royalties to September 30th, 1902, was awarded to plaintiff in that action. Upon appeal the judgment was affirmed. ( Bennett v. Iron Clad Manufacturing Company, 90 App. Div. 611.) An appeal was taken by the company to the Court of Appeals which was dismissed.
October 15th, 1904, Risley Love, as attorneys for Bennett, instituted a second action against the company, by the service of a summons. October 18th, 1904, Mr. Risley wrote Mr. Hardie, who had represented the defendant in the former action, advising him of the commencement of the action to recover the amount of royalty on all baskets manufactured including solid and pieced metal baskets up to the date of the entry of an order dismissing the appeal to the Court of Appeals, to which Mr. Hardie replied under date of October 20, 1904, declining to furnish a verified statement, but wrote: "We admit having made and sold both pieced metal and solid baskets from the date of the entry of the judgment in this case up to the time of the dismissal of the appeal in the Court of Appeals and since that time, and you are at liberty to use this letter in proof of such fact, and this information alone is sufficient upon which a complaint can be expressed." The company appeared by Mr. Robert W. Hardie, its attorney, October 27th. Thereafter, upon the petition of Bennett, the plaintiff therein, and the affidavit of Mr. Risley, application was made at Special Term for an order permitting the examination of the books of the defendant to enable plaintiff Bennett to frame a complaint. An order was granted requiring the company to make discovery of its books by depositing the same with the clerk of Herkimer county or, in lieu thereof, that the company should within ten days serve upon plaintiff's attorneys, Risley Love, a verified statement containing the whole number of solid sheet metal baskets manufactured and the whole number of piece metal baskets manufactured and the date of said manufacture between September 30th, 1902, and the 10th day of October, 1904, and also a statement of the number of baskets sold and the dates of said sale. Upon the hearing of the application at Special Term the company submitted an affidavit of its general manager, accompanied by a memorandum attached thereto signed by Mr. Hardie, its attorney, in which he stated "defendant's affidavits show that the defendant admits making the baskets referred to in plaintiff's affidavits. There is, therefore, no question in regard to that point, The only point upon which the plaintiff desires information is the number of baskets made during a specified time by the defendant. This clearly is not essential in order to enable the plaintiff to form his complaint; he can state upon information and belief any number that he desires and on the trial he may subpœna the defendant to produce its books and submit to an examination. * * *" The affidavit of Bennett refers to the former action, the disposition of the same, the contract of September 14th, 1896, and stated:
"Your petitioner alleges that the defendant has since the commencement of the action above set forth continued to manufacture and sell certain baskets embraced in the aforesaid agreement as adjudged and interpreted by the court and has continued to manufacture and sell these baskets up to June 9th, 1904 * * *."
"That between September 30th, 1902, and June 9th, 1904, the defendant manufactured and sold large numbers of seamless metal and large numbers of pieced metal baskets for which it has never accounted nor paid royalties. * * *"
The company served upon the attorneys for Bennett, Risley Love, a statement of metal baskets of all kinds made and sold by the company between September 30th, 1902, and October 10th, 1904, verified by its manager, in effect that it contained a true and correct statement of all the metal baskets of every kind and description manufactured and sold by the defendant between the specified dates.
The complaint served alleged the ownership by Bennett of the two patents, the contract of September 14th, 1896, the manufacture and sale of baskets thereunder by the defendant, the former judgment in favor of the plaintiff, the appeals therefrom and disposition of the same: That by the former judgment it was determined that the solid sheet metal baskets and pieced sheet metal baskets made and sold by defendant were embraced in and covered by the terms of the contract, and that the defendant, during the pendency of the appeals, and down to October 10th, 1904, continued to make and sell and offer for sale sheet metal baskets within the terms of said contract, and had sold 6,441 sheet metal baskets.
The company answered in the action about January 25th, 1905, denying that it continued to make, sell and offer for sale sheet metal baskets within the terms of said contract, or that it had on hand baskets covered by the contract.
Upon the trial of the action on the 17th day of May, 1905, the complaint was dismissed by direction of the court. Upon appeal by the plaintiff, Bennett, from the judgment so entered, the judgment was reversed and a a new trial ordered. ( Bennett v. Iron Clad Manufacturing Company, 110 App. Div. 443.) Upon a retrial of the case plaintiff recovered a judgment from which an appeal was taken; the judgment was reversed and a new trial was granted. ( Bennett v. Iron Clad Manufacturing Company, 121 App. Div. 133.) The opinion of the Appellate Division in part stated:
On the trial of the case, the plaintiff relied for proof of part of his cause of action upon the statement of defendant hereinbefore referred to, taken in connection with the order pursuant to which the statement was furnished as establishing admission by defendant of plaintiff's cause of action for royalties and further stated that the affidavits used on the application for the order were not offered in evidence and did not appear in the printed record.
Upon a retrial of the case in 1907, the affidavit alleged to have been forged or altered was offered in evidence, but was excluded by the trial justice. Additional evidence was offered by a witness called on behalf of the plaintiff. Judgment was had in favor of plaintiff; an appeal was again taken to the Appellate Division by the company, where the judgment was reversed and a new trial ordered upon the ground that the additional evidence received was lacking in probative value to justify the court on appeal sustaining the same. ( Bennett v. Iron Clad Manufacturing Company, 127 App. Div. 943.)
Upon the next trial in February, 1911, the affidavit was offered in evidence. Upon examination of the same, it was claimed by the defendant that the words "the same" had been inserted in the original affidavit of the manager of the company, and after an inspection of the paper a juror was withdrawn and a non-trial was ordered.
It is asserted by the People that the affidavit in question was offered in evidence by the defendant for the purpose of supplying proof necessary to meet the objection made by the Appellate Division: that the affidavit of the manager of the company had been altered or forged, whereas it originally read: "Deponent admits that between September 30th, 1902, and June 9th, 1904, the defendant manufactured and sold seamless and pieced metal baskets for which it has never accounted to the plaintiff nor paid royalties," and as changed read: "Deponent admits that between September 30th, 1902, and June 9th, 1904, the defendant manufactured and sold ` the same' seamless metal and pieced metal baskets for which it has never accounted to the plaintiff nor paid royalties." Evidence was offered by the People tending to show that the defendant was at the clerk's office of Herkimer county on Thursday, February 2d 1911, at about the noon hour and asked a clerk for the papers in the case of Bennett v. Iron Clad Manufacturing Company; that the papers were obtained from the vault and taken to a table in the easterly side of the room, where the defendant was engaged for a time in looking over the papers; that again on Friday at about the same hour the defendant was in the office of the clerk of Herkimer county, and while there again looked over the papers, selected some of them and gave them to the person in charge with the request that they be produced in court the following week. The People claimed that the inference was that the defendant abstracted the paper in question from the office of the clerk on his first visit, took the same to Utica and inserted therein the words "the same" and returned the papers to the clerk's office on the following day.
Samples of typewriting made upon two typewriting machines in the office of the defendant in the city of Utica were produced upon the trial. On behalf of the People it was claimed that the words "the same" had been inserted by the use of one of the typewriter machines in the office of defendant known as an "Underwood machine." It was argued by counsel for the appellant that two specimens of typewriting, Exhibits Numbers 17 and 18, made upon the machine in the office of the defendant two days subsequent to the alleged offense, were not such standards as the law recognizes and were improperly received in evidence. I am of the opinion that the question has been decided adverse to the contention of defendant. ( People v. Storrs, 207 N.Y. 147.)
Counsel for appellant urged error on the part of the trial justice in permitting a witness to testify as to how often certain assumed conditions in letters in a typewriting machine would occur in other machines based on the law of mathematical probabilities. It appeared by the testimony of an expert called by the People that he had inspected some 20,000 typewriter machines and never saw one that was perfect; that the alignment of the machine was practically the "heart" of the same, that the spacing, key, levers, rolling, space, carriage, all these center around the alignment, and a machine cannot be manufactured so the alignment will come perfect. Another expert witness testified that "alignment is how the letters approach to or recede from the base line. These letters sometime strike below the line and above the line, and that is common observation I think in looking at typewriting. A letter may be normal in character in the usual typewriting operation; it may be on the line practically all the time, but it can be made to strike above or below the line. Sometime you will find the letter above the line, next time you will see it is on the line."
Attention has been called to this evidence in view of the evidence offered by the People to show that it was highly improbable if not impossible that defects similar to those in the machine owned by defendant exist in any other typewriting machine. The witness, who was a professor of mathematics in one of the universities of the state, under repeated objections by counsel for defendant, was permitted to testify that by the application of the law of mathematical probabilities the chance of such defects being produced by another typewriting machine was so small as to be practically a negative quantity. The witness asserted that when the facts are ascertained, the application of the law of probabilities to them is a matter of pure mathematics and not of speculation or opinion. He defined the law of probabilities as "A proper fraction expressing the ratio of the number of ways an event may happen divided by the total number of ways in which it can happen." The various defects claimed to be visible and pointed out by the experts in the specimens of typewriting made upon the defendant's machine corresponding to like defects in the letters of the two words "the same" in the alleged altered document were called to the witness' attention by the district attorney, and he was asked to apply the law of mathematical probability thereto. For illustration he was asked: "If it be assumed that it is as probable that any given letter will slant as it is it will not slant, are you able to ascertain what is the probability that the letter `t' in the six letters, t, h, e, s, a, m, will slant and the others remain perpendicular?" He answered: "One in sixty-four." Practically the same form of question was put to him in regard to missing serifs on other letters and embracing various features pointed out by the experts, and by a process of compounding these results concerning each of the particular defects the witness was permitted to give his conclusion that the probability of these defects being reproduced by the work of a typewriting machine other than the machine of defendant was one in four thousand million, which was, of course, equivalent to a statement that it could never occur. The witness, so far as the record discloses, was not qualified as an expert in typewriting; he had never made a study of such work, or of the machine claimed to have been used for the purpose of interpolating words in the document offered in evidence, nor did he take into consideration in arriving at his result the effect of the human operation of such a machine, that is whether the same defects would always be discernible, no matter who was the individual operating the machine. These factors and many others which we cannot foresee and which, in all likelihood, are beyond the possibility of any human being to ascertain, would enter into any calculation of this character. The statement of the witness was not based upon actual observed data, but was simply speculative, and an attempt to make inferences deduced from a general theory in no way connected with the matter under consideration supply the usual method of proof.
Evidence of that character is unlike that of the North-hampton table of expectancy in use by insurance companies, and at one time receivable in evidence upon the trial of a cause in our courts, for the reason that such table is based upon the result of actual observation. Notwithstanding the use of that table for a long period of time, it is common knowledge that within a few years the Carlisle table of expectancy which in many instances increases the probability of life has been adopted and is in use in this state. (Rule 70, Supreme Court Rules.) A table of expectancy of life based on actual experience is the only practical and satisfactory rule for determining the expectancy of life of a particular individual. That rule is used from necessity when the fact to be proved is the probability of the happening of a future event. It would not be allowed, for illustration, if the fact to be established were whether A had in fact died, to prove by the Carlisle table he should still be alive. The fact to be established in this case was not the probability of a future event, but whether an occurrence asserted by the People to have happened had actually taken place.
Similar evidence seems to have been admitted in the trial of the famous Howland Will case in the United States Circuit Court in Boston, Massachusetts, in 1868, where one of the questions was whether the words of the testator's name in the contested writings exactly coincided with the same name in the uncontested will held by the plaintiff. Professor Benjamin Peirce of Harvard University, the most famous mathematician of his time, was permitted to testify that the chance of the genuine production of such coincidence as that of the three signatures was that of one to two thousand six hundred and sixty-six million of millions of millions of times, in other words, that a person would not write his signature exactly alike more than once in so many millions of times that it would transcend all human experience. This evidence is commented upon in a note to section 9, Wharton's Criminal Evidence (9th ed.), and the author there points out not only the unreliability of testimony in criminal cases based upon the law of probability, but numerous instances where authors of works upon that subject have acknowledged errors in the adoption of such a rule. The Howland Will case is reported quite fully in 4 American Law Review, page 625, but no judicial review of the admissibility of the evidence referred to was had, as the case was disposed of upon another point ( Robinson v. Mandell, 3 Cliff. [U.S. Circuit Rep.] 169), and thereafter the parties adjusted their differences.
The admission of the evidence was error, prejudicial to the interests of the defendant in this case, which had already been burdened by expert testimony, and was an attempt to draw a line between assumed fact and reasonable conclusion to an extent never recognized by this court. For this reason the judgment of conviction should be reversed and a new trial ordered.
I concur with my brother HOGAN, but desire to add a few observations on the question whether we should disregard what we all agree was an error of law in admitting the testimony of the witness Snyder.
Our sole function in this case is to decide questions of law with which the unanimous affirmance by the Appellate Division has nothing to do. Of course, we are to consider the facts to determine the bearing of the law points and whether any errors of law committed on the trial were prejudicial to the defendant; but under our system of jurisprudence the determination of the defendant's guilt or innocence is the sole province of the jury. It is the defendant's right to have that determination made on legal evidence wholly uninfluenced by matters which the law says may not be considered to his prejudice. For us to convict the defendant upon conflicting evidence and thus to say that he was not harmed by prejudicial error is to disregard the forms of law, to assume arbitrary power, and to deprive the accused of his liberty upon our varying individual judgments instead of the judgment of the law.
There was a sharp conflict at the trial on the important question of fact, whether the words claimed to have been forged were written upon the defendant's typewriter. The People's case in that respect depends almost wholly on expert evidence, offered to show that certain alleged defects in the defendant's typewriter corresponded with defects in the typewritten words; and that evidence, so far as it was visualized for the jury, consisted of exhibits prepared by the People's expert. Of the thirteen similar defects testified to by him, two were variable and concededly may be eliminated, one was the slant of the letter "t," which the People's witnesses admitted was a common occurrence, and ten were due to alleged defects in the type. The People's expert witnesses who examined the type of the defendant's machine with a microscope were able to name five defects, which their testimony on cross-examination tended to show were common on used machines. Those defects existed in three letters most generally used. The photographs, ten diameters enlarged, are claimed to show five other defects, consisting of the letter "s" being heavy in the upper and lighter in the lower parts, the letter "a" being heavy in the lower and lighter in the upper parts, and an almost imperceptible swerve in one of the down strokes of the letter "m." Each of the defects alleged varies in the disputed writing and the two standards of comparison; and there are other differences between the standards and the disputed writing, which together with the differences in the defects in dispute may have resulted from defects in different typewriters, from different ribbons, from the way in which the writing was done, and conceivably even from the manner in which the enlarged photographs were produced. I do not say that to criticise or in any way to reflect upon the People's experts, but only to point out the extreme importance of confining the expert testimony within proper limits in a case admitting of such opportunity to err, and upon a question in the decision of which the jury had to depend almost wholly on such testimony both for the particular facts and for most of the conclusions therefrom.
Even assuming that the alleged eleven defects, variable as they were in extent and character in the disputed writing and the standards, had some resemblance to defects in the defendant's typewriter, the likelihood of similar defects in type recurring in another typewriter would depend on the dies from which they were made, on the process of manufacture, on the greater likelihood of particular parts, such as serifs, being broken by use, on the material composing the type, on the way in which the machine had been used, and doubtless on many other things which do not now occur to me; and yet the problem was sought to be determined by a so-called law of mathematical probability regardless of actual experience, physical facts or the element of human agency, and on the assumption manifestly false that a given thing was as liable to happen as not to happen. Upon selected data, which may or may not have been the ones involved in the problem, it was computed that the same slant of the letter "t" would occur once in 256 times, like computations were made as to each of the alleged defects and by compounding the results it was computed that all would occur in combination once in 4,000,000,000 times. Thus the happening of a past event was sought to be established not by witnesses of the fact, not by proof of the circumstances surrounding it or the causes contributing to it, but by an abstract doctrine of chances, which was put before the jury as a demonstration to a mathematical certainty.
So far from being a trivial incident to which no importance was attached at the trial, the evidence covers many pages of the record and in fact consisted of the entire testimony of a witness learned in the higher mathematics, summoned from a great university to testify. To the credit of the district attorney it should be said that he does not make the argument, too often made in this court by counsel, who against strenuous objection have succeeded in getting improper evidence before the jury, that no attention was paid to it. He asserts that it was proper to be considered by the jury and states the reason for its introduction as follows: "A large part of the cross-examination by appellant's counsel of the practical typewriter men called by the People as witnesses was devoted to an effort to establish by them that the several defects shown to exist in the type on appellant's typewriter were frequently found in typewriter type. * * * It was obviously intended to argue from this, that exactly the same condition as existed in the typewriter in question might reasonably be expected to be found with some frequency in other typewriters. There was thus brought into the case by the appellant the very proposition which the People sought to rebut by the testimony of Professor Snyder, namely, that the fact that the particular defects which existed here occurred in typewriter type with some degree of frequency, indicated a fair probability that other typewriters would show the same condition as existed in the appellant's typewriter." The evidence then based on actual experience and observation, that the defects in the defendant's typewriter discoverable by a microscopic examination were common in such machines, was to be overthrown by the testimony of a witness, who knew nothing on the subject from observation and experience, that to a mathematical certainty they would not recur with certain other assumed defects in another machine once in 4,000,000,000 times.
Of course, the combination of particular defects in one machine, if it existed, was the damaging fact against the defendant and if it were established that the eleven assumed defects actually existed in the defendant's typewriter and corresponded with the defects in the disputed writing, I should be prepared to agree that the fact in issue was so conclusively established as to render the error harmless; but a careful examination of the exhibits convinces me that upon this branch of the case the jury must have relied largely upon the conclusions of the experts and that in all probability they were profoundly impressed by the conclusion of Prof. Snyder that the assumed defects would not recur in another machine once in 4,000,000,000 times. Whilst the existence of the defects was merely assumed in the question put to the witness, it is going too far to assume that the jury so carefully analyzed the evidence as to distinguish between defects assumed in the abstract and those actually found to exist and to note the precise relation between the defects in the defendant's typewriter and those appearing in the disputed writing and the standards of comparison. The vice of the testimony consisted in its being purely an abstract theory having no relation to actual experience. The minds of the jury were thus diverted from the actual facts in issue and led into a maze of abstract theory and speculation, and it is not to be supposed that they perceived, what escaped the notice of the court and the district attorney, that there was no relation between the facts and the theory.
It cannot fairly be charged that the defendant's counsel have attempted to magnify a trivial circumstance, or an unimportant incident of the trial. They vigorously protested at the trial against the introduction of the evidence concerning the so-called law of mathematical probability and persisted in their objection to the very verge of showing disrespect for the court's ruling, and the effect of the adverse ruling upon the jury was heightened by such colloquies as the following:
"Defendant's Counsel: Isn't this purely speculation without any basis?
"The Court: We have the testimony of this witness that it is a matter of mathematical accuracy." and later:
"Defendant's Counsel: I don't want to be too strenuous about this, but I do want to object to this testimony on the further ground that there is no foundation whatever for it; it is perfectly obvious that it rests entirely on speculation. If there was one machine in existence this rule would not apply at all; if there was more than one machine in operation, it might apply in some way; if there were ten thousand in operation —
"The Court: The suggestion is upon the very line I indicated a moment ago. It may appear it is too ridiculous for consideration, but the People claim everything from it. I think it is competent. As to the weight of the evidence, it is for the jury. I think I better hear it."
I am not certain what any intelligent person would have concluded without the aid of expert testimony or that the jury had such common sense and keen perception as to ignore the entire testimony of a witness of distinction introduced by the district attorney in the manner described for the purpose avowed, allowed over strenuous objection and characterized by a learned judge as competent but am of the opinion that the admission of this testimony was prejudicial to the defendant.
The trend of popular thought may be in favor of having the courts disregard the rules and forms of law in deciding appeals in criminal cases; but although we may seem to be far removed and secure from the arbitrary exercise of power, it should not be forgotten that it is the very essence of our boasted freedom that the citizen holds his life, liberty and property subject to the law only and not at the discretion of any individual, be he judge, executive or legislator.
I have not intended to suggest or countenance the view that erroneous rulings are per se cause for reversal or that new trials should be granted for technical, trivial or unsubstantial errors. Being convinced that the defendant has not had what the law accords him, a trial free from prejudicial errors of law, it is not for us to express any opinion as to his guilt or innocence, which must be determined by another jury.
I dissent. The defendant, an attorney at law, was convicted of a felony for having violated section 810 of the Penal Law. His crime consisted in offering in evidence in the trial of a civil action an affidavit of Edward R. Gilman which had been fraudulently "forged" or altered by inserting therein the words "the same," which words at the time he offered the paper in evidence he knew to have been forged or altered. The judgment of conviction has been unanimously affirmed by the Appellate Division, and no question of fact is presented which is reviewable in this court. The record of the case is voluminous. The trial was fair and the issues were submitted to the jury by the learned trial justice in a charge to which no exception was taken. Although the trial was protracted and every step in it was strenuously contested there is but a single circumstance that can be claimed to present an erroneous ruling. The nature of this ruling was such that, in my opinion, we are not warranted in reversing this judgment on account of it. The question at issue upon the trial was whether the defendant had inserted or caused to be inserted the words "the same" in the affidavit which he offered in evidence in the trial of a civil action. These words were of vital significance in the proof upon the trial of the civil action in which the forged affidavit was offered in evidence. It was, of course, competent for the prosecution to prove that the forged words, "the same," were written upon the defendant's typewriter. Evidence of the most convincing character was offered to prove this fact. The fact that the words "the same" in the forged or altered affidavit contained the identical peculiarities which appeared in the letters of these words written upon the defendant's typewriter was a strand in the cable of proof which connected this defendant with the commission of the crime. The defendant's counsel appreciated the importance of this evidence, and upon the cross-examination of the witnesses called by the prosecution attempted to show that the several defects shown to exist in the type of the defendant's typewriter were frequently found in typewriter type. The witnesses called by the prosecution had pointed out thirteen separate and distinct defects in the six letters, "t, h, e, s, a, m," which were included in the words "the same." Two of these defects were said to be variable and were disregarded by the district attorney in the question which he put to Professor Snyder. His question asked the witness to assume the existence of the eleven defects referred to and asked him to calculate the probability of the recurrence in any other typewriter of these identical eleven defects in the six letters specified. The question propounded assumed that the six typewritten letters, "t, h, e, s, a, m," possessed the defects testified to, and asked the witness whether he was able to state from mathematical calculation and by the law of mathematical probability what is the probability that each and all of the distinguishing characteristics pointed out would exist in another typewriter. The witness answered that he was able to so state and was then asked, "What were they?" and answered, "About one in four thousand millions." So far as the witness assumed merely to state what the law of mathematical probability is and the illustrations which he gave as to the operation of that law his testimony was not prejudicial to the defendant. The only respect in which it is possible to claim that this testimony could be regarded as prejudicial is that part of it which related to the answer of the witness that he could state what the probability was that these precise defects would be found to recur in another typewriter and that this probability was one in four thousand millions. This testimony seems to me to have been unnecessary. The district attorney could have argued to the jury as a matter of common knowledge that if they found that the eleven distinct defects did exist in the six letters referred to it was exceedingly improbable that these precise eleven defects would be found to be present in any other typewriter, and that inasmuch as those six letters in the Gilman affidavit did contain these eleven defects, and these precise defects were also found present in the standards written by the defendant's typewriter, it was exceedingly probable that these six letters in the affidavit of Gilman were written on the same typewriter that wrote the standards. Common sense at once recognizes how remote is the probability that all of these defects should recur in these six identical letters in any other typewriter. Indeed, if the district attorney, basing his argument upon matters of common knowledge and the general perception of all men, had pointed out that there was not one chance in four thousand millions that these identical defects would be found in these identical six letters of another typewriter, he would, I think, have been within his rights. Substantially the same statement did not become prejudicial because it is made by one learned in the higher mathematics. In common speech men would say that there was only a remote chance that these precise defects would recur in another typewriter. This is what any intelligent person would have concluded without the aid of expert testimony. A reference to one chance in four thousand millions would denote in the juror's mind simply extreme improbability. I think that this is all that the jurors could have understood from this testimony. The whole significance of the testimony was merely to indicate that the probability was remote and the testimony of the expert, if improperly received, was not, in my opinion, prejudicial error. While there are doubtless many respects in which the principles of science may, with advantage, be applied to judicial proof, as has been effectively shown in the progress made in effecting identification by means of finger prints, yet courts must be cautious in receiving such proof and cannot sanction it until its accuracy has been fully tested and vindicated by experience. Certainty can only be attained in the light of perfect knowledge which it is often impossible to attain, and the law allows in such cases a consideration of that which is probable. That calculations based upon sufficient data and correctly made will bring us to a more correct estimate of probability than conjecture or general perception is doubtless true, but it is equally clear that the calculations must be based upon the assumption that all the data of the problem have been stated. There have been many interesting discussions upon this subject by scientific writers and nowhere have the difficulties of the application of this theory been more clearly pointed out than by these writers. Thus Professor Jevons in his "Principles of Science" says: "The theory of probability, though undoubtedly true, requires very careful application. Not only is it a branch of mathematics in which errors are frequently committed, but it is a matter of great difficulty in many cases to be sure that the formula correctly states the data of the problem." To illustrate the liability to error in the application of this theory, the same author gives many instances pointing out how often the most acute and powerful intellects have gone astray in the calculation of probability. If, therefore, the evidence in this case had been of a material nature and such as would probably have served to mislead the jury, I should hesitate to cast a vote which, in the present state of mathematical science, would sanction the reception of such evidence as to past transactions. In the view I take of this case, it is not necessary to sanction this kind of evidence to sustain the judgment appealed from. The whole incident was a relatively unimportant feature of a long trial. While several pages of the record are devoted to the examination of the witness Snyder, a large part of these pages consists in the objections and arguments of counsel. The testimony itself was very brief. The proof that the words "the same" were forged did not at all depend upon the opinion of experts. With the exception of the witness Snyder, the trial court limited the experts upon typewriting to describing the conditions of the type and did not allow them to testify as to their opinions. In determining as to the materiality of the testimony of Professor Snyder, it must be borne in mind that the fact that the words "the same" in the forged affidavit were written upon defendant's typewriter was proved satisfactorily by other evidence than that of Professor Snyder. It was also shown that the body of the Gilman affidavit was written upon a Remington typewriter in small pica type, whereas the inserted words "the same" were written upon an Underwood typewriter and were in medium Roman type. It was also shown that defendant's typewriter was an Underwood machine with medium Roman type, and that the words "the same" when written upon that machine corresponded exactly with the forged words in the Gilman affidavit. In addition to this a carbon copy of the original affidavit was introduced which did not contain the forged words, and it was proved that when the original affidavit was signed the words "the same" were not contained in it. The defendant's opportunity and motive to commit the crime were clearly proved, as also was the fact that, after he had been charged with this crime, he endeavored to obtain a typewriter which would reproduce the same defects as the Underwood machine in his office upon which the standards had been written. It is not necessary to review all the other circumstances tending to establish the defendant's guilt. The unanimous affirmance by the Appellate Division of the judgment entered upon the verdict of the jury and the convincing character of this testimony are circumstances that indicate that the error committed in the reception of Professor Snyder's testimony was not prejudicial. While it is true as pointed out in the opinion of Judge MILLER that the defendant's counsel protested vigorously against the admission of the evidence of Professor Snyder, that fact, in no way, indicates the importance of the evidence offered, or that it was improper or prejudicial, because a review of the record will disclose that the defendant's counsel made vigorous objection to a very large part of all the testimony that was introduced in this case. If vigorous objection by defendant's counsel is to be regarded as indicative of importance and materiality, the trial court might as well have directed a verdict of acquittal at the outset. Indeed as one reads this record and considers the obstacles thrown in the way of the presentation of perfectly proper proof, it is remarkable that the facts were so fully developed. Nothing but the great care and patience of the learned trial justice made this possible. I agree with my brother MILLER that "in our system of jurisprudence the determination of the defendant's guilt or innocence is the sole province of the jury." In this case the jury, upon evidence conceded to be sufficient to justify their verdict — apart from the evidence of the witness Snyder — have found the defendant guilty, and the Appellate Division, who are charged with the duty of reviewing the facts, have unanimously concluded that the verdict was sustained by the evidence. Nor do I advocate the assumption by the courts of any "arbitrary power," in refusing to concur in the reversal of this just verdict for what seems to me to be the trivial cause assigned. There is a wide difference between the exercise of "arbitrary power" by the courts and the refusal of the courts to allow the great constitutional guarantees of life, liberty and property to be used as means by which criminals are permitted to cheat justice. I regret that judicial tribunals have not always observed the difference. There is no law of this state requiring the court to reverse a judgment obtained after a fair trial because of immaterial error. In disregarding such an error, we do not convict the defendant, but merely refrain from justifying the reproach which the reversal of judgments in criminal cases upon unsubstantial errors has cast upon the law of new trials. The view expressed in the opinion of Judge MILLER seems to me to involve the application of the old Exchequer rule that regarded the making of any error as conferring upon the defendant the "legal right" to a new trial. My own view as to this matter is expressed in the following quotation from Professor Wigmore: "As to this theory of legal right, it may be said in reply that no man has a legal right to have his cause wrongly decided, for that is what this `right' comes to. He has indeed a legal right to a jury trial; and he has a right to a fair trial in general. But these are ends in themselves, because the one by constitution and the other by common sense of justice becomes a paramount object. But none can justify the exaltation of the ordinary rules of evidence, which are mere instruments of investigation, into an end in themselves. As well might a gardener cut down a thriving vine because his henchman has used a hoe instead of a spade in planting it; or a farmer bring valuable bantams to the block because they were hatched by a meddlesome duck instead of by their lawful parent. A glance at common affairs will awaken us to the intrinsic absurdity of the theory of `legal right.'" (1 Wigmore on Ev. p. 72.) The suggestion that "the trend of popular thought" invites courts to assume arbitrary power seems to me to be as irrelevant as I believe it to be inaccurate. In any event it is as foreign to the merits of this case as are the references to the observance of "forms of law" in order to preserve "the very essence of our boasted freedom." The defendant has had a fair trial by a jury. All the forms of law have been observed with strictness and impartiality. He has had the benefit of an appeal which has been decided adversely to him. Notwithstanding all this it is now proposed to reverse the judgment because of an unsubstantial error in the admission of testimony. To this practice I am unwilling to give my consent. A review of the whole record shows that upon the trial the testimony of Professor Snyder amounted to very little, and it was not even deemed worthy of mention by the learned trial justice in his admirable charge to the jury. But now, as is often the case, when the defendant has been convicted, this trivial circumstance is seized upon, elaborated and embellished and presented to this court, as if it were the controlling feature of the whole trial instead of being, as I think it was, a comparatively unimportant incident which the trial justice properly regarded as unworthy of being referred to in his charge. We ought not to forget that in the last analysis the great purpose of a criminal trial is to determine whether the defendant is guilty or innocent. When the defendant's guilt is proved, he should not be allowed to escape the consequences of his act because of a technical error in the course of a long trial.
The evidence satisfactorily proves the guilt of the defendant, and, except for the admission of the evidence of Professor Snyder, the trial was free from legal error. The error involved in the reception of that evidence was not, in my opinion, of such a character as to justify us in reversing the judgment.
I vote in favor of the affirmance of the judgment.
WILLARD BARTLETT, Ch. J., CHASE, COLLIN and CARDOZO, JJ., concur with HOGAN and MILLER, JJ.; SEABURY, J., reads dissenting opinion.
Judgment reversed, etc.