The plaintiffs sue to recover of the defendant two mortgage bonds of the Toledo, Norwalk & Cleveland Railroad Company. To entitle them to recover, they were bound to establish two facts: 1st. That the two bonds found in possession of the defendant belonged to the plaintiffs. 2d. That the circumstances under which the defendant purchased the bonds were not such as to protect his title. On neither of these points is the evidence very satisfactory.
1. There were no marks about the plaintiffs’ bonds tending specially to identify them or to distinguish them from other bonds of the • same company, except the numbers. The plaintiffs lost five bonds, numbered respectively from 336 to 340, inclusive. These numbers did not agree with those found in possession of the defendant, which were *240225 and 238. This is the first circumstance to show a want of identity. To account for this discrepancy it is claimed that the numbers had been altered from some two of those owned and lost by the plaintiffs. And this alleged alteration is claimed for two purposes: 1st. To aid in the identification of them as the plaintiffs’ bonds. 2d. To show that the defendant received them under suspicious circumstances. It is perhaps not improbable that the coupons had been altered, but the proof of the alteration is not very strong or satisfactory. The evidences of alteration relied on were that on one of the bonds and both of the coupons pieces of paper were pasted behind the numbers. On being held up to the light they presented an appearance of thinness of the paper and brightness of the -ink where the numbers occurred, as if made on erasures and written anew. It is also claimed that the paper at these places appeared to have been rubbed. Whether these appearances were so suspicious as to deprive the defendant of the character of a honafide purchaser will be presently considered. The question now is how far, in connection with other circumstances, they tended to establish the identity of these bonds with those belonging to the plaintiffs. The principal ground for claiming this identity is _that when the payments for semi-annual interest on the bonds next succeeding the date of plaintiffs’ loss became due, no coupons were presented for payment answering to the numbers of the plaintiffs’ bonds, but duplicates of five other numbers, 224, 225, 234, 238 and 455 were presented and paid. Such duplicates had never before appeared. There were also produced duplicate bonds numbered 225, 238; that is, one of each of said numbers besides those held by defendant of corresponding numbers. There was also evidence tending strongly to show that no duplicate bonds or coupons were ever issued by the company. This was substantially the case made by the plaintiffs. But there were other circumstances to be considered. Altera*241tians in the bonds and coupons were not necessarily evidence of fraud. The bonds might have been altered as to their numbers by the officers of the company, before they were issued. There is no proof that they were not. There is no proof that the two bonds produced on the trial numbered 225, 238, corresponding with similar numbers held by the defendant had not been altered, or bore more the appearance of genuineness than the others. There is no proof that the bonds and coupons lost by the plaintiffs ivere themselves free from suspicious marks, or may not be still outstanding with their true and original numbers. The proof of alteration is not marked or decisive. Some of the bonds were issued without all the coupons attached to them being numbered. The officers of the company were not themselves certain of the fact of alteration. The New Ydrk director, to whom the matter was referred, paid the August coupons to the defendant at the next succeeding date after the first presentation of duplicate coupons, notwithstanding he examined the bonds and the company had been notified of the supposed fraud or larceny. There was no proof to show that the numbers inserted in the defendant’s bonds were not in the handwriting of the officers of the company, or of the particular officer who filled in the bonds originally. The judge, at the trial, inspected the bonds, and we may perhaps conclude that he was not satisfied they had been tampered with. The proof of identification, and the plaintiffs’ ownership of the bonds in suit, is by no means complete. Possibly there was enough to justify and require the submission to the jury of this question of fact. And this makes it necessary to consider the other question in the case.
2. Were there such marks of alteration and defacement upon the bonds and coupons as should discredit them in the market, and deprive the defendant of the protection of a tona fide purchaser of negotiable paper.
The character of the alleged alterations has been already *242referred to:—the thinness of the paper where the numbers were, its rubbed appearance, the brightness of the ink, the patches upon the back. .The court, after an inspection of the bonds, held that these appearances were not sufficiently decisive of fraudulent alteration, and dismissed the complaint. I think his decision was correct, or at least should not be reversed, for the following reasons:
(1.) The fact of alteration is consistent with the genuineness and validity of the bonds. It is not shown when the alterations were made, nor that they were not made before the bonds were issued, nor that they are not in the handwriting of the officers of .the company.
(2.) The alterations (if any) are in immaterial particulars. The numbers are not an indispensable part of the bonds or coupons, and from that circumstance would not be likely to attract the attention of practical men.
(3.) We are not in a situation to say that the evidences of alteration and erasure may not have been wholly indecisive and unsatisfactory. This question must turn very much upon the appearance of the bonds and the coupons. .They cannot well be adequately described by witnesses; at least they are not so. If the signs of alteration were very slight-r-if the paper was only slightly thinner or slightly rubbed, or the ink only a little brighter, and all this may have been so, they may not have been sufficient to attract the attention of even a cautious and prudent observer. As to the patches on the back, they might not have been likely to be discovered, o'r might appear to have been made to cover an accidental rent in the paper. The evidence, so far as it stands upon the oral or written testimony of the.witnesses, is not of that character or strength that enables us to say that the court erred in holding it insufficient to support a verdict.
(4.) But stronger than all, in my opinion, is the objection that we have not the benefit of an inspection of the papers, and are, therefore, without a portion of the means *243which the court below enjoyed in passing judgment upon the case. I think it a mistake to say that the bonds (if not the coupons also) were not before the court and jury.
The first witness, Birdsall, speaks of them as the two bonds now shown him. It is stated in his testimony that the bonds in question and the coupons annexed to them are as follows, (giving copies thereof). It is then stated “the following coupon was also read in evidence, the same having been detached from, the bond, but admitted to belong to it. The question is put to the witness, Hays: '•* Look at those numbers on the coupons and state whether or not they have been altered.” It is evident, also, from the testimony of Coe, that the bonds were submitted to his inspection. The plaintiffs’ counsel refers to the condition and appearance of the bonds, in his claim to go to tie jury, in- opposition to the motion for a non-suit. And the case further states that “the court, after an inspection of said bonds, refused to submit any question of fact to the jury.” .
I do not deem it necessary to consider the question whether the court or the jury, in passing upon the case, would have had a legal right to judge for themselves by inspection of the appearances of defacement and alteration as being sufficient or otherwise to put the purchaser upon inquiry. It sufficiently appears that the court, in granting the motion, did, without objection on either side, inspect the papers, and in all probability decided the case, in part, upon the result of such inspection upon his mind. These papers, as is obvious from the opinions of the judges of the superior court, were not before them, and they are not before us. We are, therefore, as was the general term of the superior court, deprived, in part, of the means and instruments of evidence which the trial court enjoyed of arriving at a correct conclusion on the matter in question, and with this partial view of the case cannot safely say, and are not perhaps entitled to say, that the trial court was in *244' error. It'was required of the plaintiffs alleging error to have occurred in the trial of the cause to have affirmatively shown it. It was competent for him, if necessary for the purposes of justice, to have procured the original papers to be sent up to the court above;, .and not having done so, we are not at liberty to say, upon this incomplete view of the case, that “ manifest error hath intervened.” The only remedy left for the plaintiffs, in my opinion, is a new action.
The judgment of the general term of the superior court should be reversed, and judgment ordered dismissing the complaint, with costs.
Wright, J.The exceptions taken at the triai were directed to be heard in -the first instance at a general term. The principal, and only one insisted ón, was to the decision of the judge dismissing the plaintiffs’ complaint. The general term ordered a new trial, on the ground that this ruling was erroneous—holding that the case should have gone to the jury. • The action was to recover the possession of two bonds (in terms payable to bearer transferable by deliv cry), for $1,000 each, issued by the Toledo, Norwalk & Cleveland Railroad Company, alleged to be detained by the defendant. The bonds had been purchased for the defend ant by his brokers, in December, 1860, in the usual course •of business, and without actual notice of any defect in them, or in the title of the seller. The plaintiffs claimed tho/ they were the owners of the bonds. If, therefore, ther< was evidence tending to prove the plaintiffs’ ownership anc the'defendant’s purchase of them in bad faith, it was erroj to withhold the case from the jury.
■ It seems to me that "when the case was rested, the testimony failed utterly to show title in the plaintiffs to the bonds in question; and if the jury had given them a verdict, it would have been the result of mere speculation and conjecture. They were the owners, in the fore part of November, 1860, of five bonds of $1,000 each of the Toledo, *245Norwalk k Cleveland Railroad Company, bearing the numbers 336, 337, 338, 339, and 340; which bonds were left at the Fulton Bank in a box, and the box was afterwards lost or stolen. The bonds in suit of a similar description, which were admitted to be genuine, bore the numbers 225 and 238. Of course, to make any case on the part of the plaintiffs, it was necessary for them to show that the latter were two of their bonds that had been lost or abstracted. The bonds issued by the company were alike in all respects except the numbering; but those in suit were numbered differently from those formerly belonging to the plaintiffs. The theory of the plaintiffs was, that the defendant’s bonds were two of their lost bonds, which had been altered in these numbers. Was there any evidence to go to the jury as to the identity of the bonds which were the subject of the action with those lost by the plaintiffs?
If the defendant’s bonds were the genuine and original tíos. 225 and' 238, there could, of course, have been no pretense tho,t they ever were the property of the 'plaintiffs. Et was then primarily essential to show that the numbers had been altered. This was a fact susceptible of direct proof, but none was given. There would have been no difficulty whatever in proving by the secretary of the company, who filled up and numbered the whole series of bonds, or by others familiar with his handwriting, that the numbers on the bonds had in fact been altered. The bonds were open for examination on the trial; they had been seen and examined before the trial by some of the officers of the company; the person who numbered them was one of the plaintiffs’ witnesses, yet no questions were asked or attempt made to prove by them that the numbers upon the bonds were not in the handwriting of the person originally numbering the series. All the evidence introduced, bearing in the remotest' degree on the question of alteration, was this: The series of bonds when issued were numbered consecutively, from No. 1 to No. 525; there was no duplicate *246numbers either of bonds or coupons issued; for some ten years after issue, no duplicate coupons had been seen by the company’s paying agent in New York; in February, 1861, duplicate coupons of the numbers 225 and 238 appeared and were paid by the Corn Exchange Bank, and returned to the company; and there were in existence two other bonds numbered 225 and 238. This did not tend to prove that the numbers on the defendant’s bonds had been altered, or that they were not the genuine and original Nos. 225 and 238. Had it been shown that the numbers upon the other two bonds, admitted to be in existence, were in the handwriting of the officers of the company—or any proof whatever given that they were not altered or spurious; or had it been shown in any way that the duplicate coupons, numbered 225 and 238, which had appeared and been paid, were genuine and not altered ones—there would have been something for the jury. But, in these respects, there was an entire absence of proof. Under the circumstances, it would have been a mere matter of conjecture with the jury, that the numbers on the defendant’s bonds had been altered and those upon the other two had not; and that the bonds in suit were not the genuine and original Nos. 225 and 238, but those produced on the trial simply to show their existence, and in relation to which no evidence was offered, were.
If, however, it were admitted that the numbers on the defendant’s- bonds had been altered, was there really any evidence that they had been so altered from any of the bonds belonging to the plaintiffs? The proof showed, as has been stated, that the bonds issued were numbered consecutively from 1 to 525, and that there was but one bond of each number. There was also evidence tending to show that for nearly ten years after such issue only one bond and accompanying coupons of the numbers 225 and 238 respectively, was ever seen by the company, or its agents, or paid upon; that when the coupons of 1st February, *2471861, were paid by their agents, and returned'to the company, they amounted to $105 (the amount of their coupons) more than the entire sum due for coupons maturing at that date, counting as outstanding the numbers 336, 337, 338, 339 and 340 (the numbers of the plaintiffs’ bonds), which were not returned as paid; that on examination by the company it was discovered that there were duplicate coupons returned of the numbers 225, 238 and 455; that in August, 1861, duplicates of numbers 224 and 234 were returned to the company’s office from their paying agent in New York, making five duplicate coupons, duplicating the numbers 224, 225, 234, 238 and 455; that the coupons numbered from 336 to 340 (being the numbers of the plaintiffs’ five bonds) maturing after August, 1860, had not, as’ late as December, 1862, been returned; and that bonds of the numbers 225 and 238 respectively, other than those held by the defendant, were in existence. This was all the. proof. It undoubtedly tended to show that there were not two bonds originally issued ¡bearing the number 225, and bonds bearing the number 238, and that either the defendant’s bonds or the two others of like number admitted to be in existence, had been altered; but as to which of them had been altered the evidence furnished no possible clue. The two bonds, numbered 225 and 238, other than the defendant’s, were not produced except for the purpose .of showing their existence, and as to their condition or appearance there was no proof whatever. Thé evidence also tended to the conclusion that the alteration of one or the other of the two classes of bonds were alterations of other original numbers into numbers 225 and 238; but there was no proof whatever from which the jury could legitimately infer that either were two of the bonds lost by the plaintiffs from which the original numbers had been erased and the numbers-225 and 238 substituted. That there were duplicate coupons of numbers 225 and 238 returned to the company, and no coupons of the numbers of the plain*248tiffs’ bonds returned up to December, 1862, did not tend to show it. Coupons from any other of the 525 bonds may have been altered. Mistakes may have been made in the numbering of the coupons, a thing not unlikely to happen. Mistakes were made in this matter of numbering, as coupons, having no numbers were proved to have been paid and returned to the company. So, also, the duplicate coupons may have been forged. Nor would the fact that the coupons of two of the numbers of the plaintiffs’ bonds were outstanding in December, 1862, not having been paid or returned to the company, have justified the jury in presuming that the original numbers had been erased from any of the plaintiffs’ bonds.
Had the question, therefore, of the plaintiffs’ ownership of the bonds which were the subject of the suit been left to the jury, they would have been compelled, I think, to decide it upon mere conjecture, and not upon testimony. As the plaintiffs’ bonds were lost or stolen in November, 1860, the defendant purchasing his in December following, and as what purported to be duplicate coupons of Nos. 225 and 238 (the numbers of the defendant’s bonds), were presented in February, 1861, paid and returned to the company's office, and no coupons bearing the numbers of the plaintiffs’ bonds had been paid or returned since their bonds were lost or stolen; and as two other bonds numbered 225 and 238 were shown to be in existence, the jury might have suspected or conjectured that the bonds in the defendant’s hands were the lost or stolen property of the plaintiffs; but it would have been mere conjecture. The proof was utterly defective on which to rest any legal or reasonable presumption, either that the defendant’s bonds were not the genuine and original Nos. 225 and 238, or that they ever belonged to the plaintiffs. The case, in my judgment, which the judge refused to submit to the jury, was not one of insufficient proof merely, but one of an absence of proof.
If the proof were defective and incomplete as to the plain*249tiffs’ ownership of the bonds, the question of the good or bad faith of the defendant’s purchase was of no importance. But I think the facts show there was no bad faith in the purchase. The defendant or his brokers, it is not pretended, had actual notice of any defect in the title of the party from whom the bonds were purchased; nor do I think upon the facts as proved, the law imputed notice to him. This question, however, was for the court and not the jury. The facts, as to the appearance and condition of the bonds, were undisputed; and the instruments were in court where the defects alleged might be seen on inspection. There was no conflict as to the facts to be settled by the jury, and the circumstances from which notice was to be inferred were not extraneous to the instruments themselves. Constructive notice is a legal inference from established facts; and when the facts are not controverted, or the alleged defect or infirmity appears on the face of the instrument, and is a matter of ocular inspection, the question is one for the court. Whether, under a conceded state of facts, the law will impute notice to the purchaser,, is not a question for the jury. Now, what were the circumstances of' the purchase which it is alleged affected the defendant with notice? There was a piece of paper pasted on the back of one of the bonds, and also of some of the coupons; and on being held up to the light, the paper where some of the numbers of the coupons were, appeared to be thin and worn. This was all. Was, there anything in this appearance of the bonds to suggest the idea that the seller had no title to them, or calculated to excite the suspicion or notice of a prudent purchaser of negotiable instruments ' offered in the market for sale? I think not. , It is not denied that in the body of the, instrument, including the signatúres and seal of the company and the certificate of the trustee, they were in all respects the genuine obligations of the company, and were in the usual and proper condition. All that can be claimed is, that if *250the defendant’s brokers had made a close and critical examination they might have discovered such marks upon them as would have indicated that the numbers had been altered. The purchaser of a bond or other negotiable instrument, in open'market and in the usual course of business, is not bound to make such an examination to escape the imputation of bad faith in the purchase. But, had the defendant’s brokers, by close inspection, been able to see marks indicating that the numbers of the bonds had been altered, it would not have been notice of any defect or irregularity in the bonds themselves, or in the seller’s' title to them. The numbers are no integral part of the bonds themselves; and it would be a violent and unjust presumption to assume that a change in the mere numbering of a bond, or other quasi negotiable instrument, offered for sale in the usual course, is, as 'a matter of law, notice, or any intimation even, of a larceny of it. The rights of a purchaser are not to be affected by constructive notice, unless it clearly appear that the inquiry suggested by the facts disclosed at the time of the purchase would, if fairly pursued, result in the discovery of the defect existing but hidden at the time. There must appear to be, in the nature of the case, such a connection between the facts discovered and the further facts to be discovered, that the former may be said to furnish a clue—a reasonable and natural clue—to the latter. The hidden facts to be discovered in this case, upon the plaintiffs’ theory, were the loss of the bonds in suit by 'the plaintiffs, and their title to recover them. The mere knowledge that the' numbers on the bonds had been changed (conceding that their appearance disclosed. the fact), gave the purchaser no clue whatever, which, if followed up, would necessarily, or even probably, lead to a discovery of these facts. There was no natural or logical connection between them; nor was there any evidence in the case that would warrant any such conclusion. The plaintiffs’ evidence clearly showed that, as matters of fact, *251the'mere alteration of the numbers, if any, could not and did not suggest the idea that the bonds purchased by the defendant’s brokers were the plaintiffs’ bonds, or that there was any defect in them. They had passed through the hands of other brokers and bankers before the defendant purchased them, who saw nothing suspicious in their appearance. . Both the treasurer of the company and its agent in Hew York, after their suspicions were aroused, made a critical and careful examination of them, and reported them in all respects regular, and the coupons for August, 1861, were paid. The treasurer was not even sure that there had been any alteration of the numbers. These were facts proved by the plaintiffs, as a part of their case. ■ The evidence disproved, as matter of fact, that there was anything in the appearance of the bonds, when the defendant purchased them., calculated to excite the suspicion or notice of a prudent purchaser.
The order of the general term, granting a new trial should be reversed, and the judgment of the' special term affirmed.
All the judges concurred except Judges Mulles? and Seldejs?. who were for affirming the judgment of the general term. Judgment reversed,