Appellant’s firm had a private bank at Belvidere, of which Irving Terwilliger was acting cashier and W. R. Dunton bookkeeper. One of its members, James S. Terwilliger, was cashier of the First National Bank of the same place, and had occasionally acted as such for appellant also, during short periods, in the absence of Irving. Appellee, who was a farmer in good circumstances, residing at a distance of ten miles, in the county of DeKalb, kept an account and had a pass book with each of these banks, though his transactions with them were not frequent.
On Nov. 17, 1881, he presented at appellant’s bank his check thereon for $1,500, and received the money. They claimed that he then had there only $500 and brought this suit to recover the alleged over-draft.
The journal and ledger of the bank, showing his account, and the testimony of Irving Terwilliger, who stated the balance to liis credit and the payment to him of the amount claimed *491to have been overdrawn as above mentioned, made the plaintiff’s case in chief.
Defendant testified that on Monday, the 27th day of September, 18S0, he deposited there the sum of §1,000 (which was not entered on their books), and was apparently corroborated by a credit for that amount as of that date on his pass book, which was produced in evidence.
In rebutt il James S. Terwilliger testified that this entry was in his handwriting, which was admitted; that the money therein mentioned was in fact deposited in the First National and there received by him as its cashier; that this pass book was the one there presented, and that he made the entry therein without observing that it was not the book of that bank. He claimed to remember distinctly that he received but one deposit of that amount from the defendant, and that one in the back room of the First National Bank in the presence of Wm. S. Dunton, its president. He had no independent recollection of the date, but fixed it with absolute confidence not only by the entry in the pass book but by the deposit slip or ticket which he then made out and spindled, bearing date the 27th day of September, 1889, which was produced, and after repeated rejection, finally admitted in evidence.
Defendant admitted that he gave the money mentioned in the pass book to James S. Terwilliger, but reiterated that it was at appellant’s bank and on the 27th, as therein entered. He further stated that the deposit referred to by Terwilliger was another and different one made on the next day, September 28th; and in this also was apparently corroborated by a credit for the same amount, as of that day, in the handwriting of said Terwilliger, on his pass book from the First National, which was put in evidence. To those transactions, including the dates, he also professed to swear from personal recollection independent of the entries in the books.
Terwilliger then accounted for the one last above mentioned as follows: It was the uniform custom in the bank to make the entries in the journal from the deposit tickets for two days together, that is, under the date of two days written at the head of the page, and then to post them all in the ledger *492as of the latest only. Thus the business of the 27th and 2Sth was written up in the journal under the heading of “ Monday and Tuesday, September 27th and 28th, 1880,” and all posted in the ledger under the date of September 28th. He said that customers often make deposits without presenting their pass books; that for this reason, when they do bring them the course is, before entering therein the deposit then made, to see from the ledger if any previously made had been omitted, and if so to credit it then in its proper order according to the date in the ledger; that on the 26th of October, 1880, defendant next made another deposit in the First National, bringing the proper pass book, and witness then entered in it the preceding one also, but under date of September-28th as it appeared in the ledger.
He further stated that he did no business at and for appellant’s bank after August 10th of that year, fixing the time, however, only from the books and papers of that bank. Irving also testified that James had never done any business there when he was present, and that he knew, though from the books only, that he was present generally throughout the 27th of September, 1880, but could not swear that he was not out of the bank at all during the business hours of that day. W. R. Duntuu, appellant’s bookkeeper, testified to the same effect. The manner of writing up the journal and posting at the First National, as stated by James Terwilliger, was also testified to by Hr. Frink, its bookkeeper.
The single ultimate question of fact in the case was whether or not appellee deposited §1,000 with appellant’s bank on Sept. 27,1850. As bearing upon that question the date of the deposit of §1,000 admitted to have been made in the First National was important if not conclusive, for he made but one such in that bank about that time and but one anywhere on the 27th; so that if the deposit of that date was in the First National his testimony is false, the evidence of the pass book turned against, him and the case of appellants fairly established.
In reference to this date James S. Terwilliger and the appellee were the principal and the only positive witnesses. *493Though one claimed to fix it from writings alone and the other from independent recollection as well, they were equally confident and their statements absolutely irreconcilable-To establish a preponderance, every item of evidence fairly tending to corroborate or to contradict either of them was more or less material. On both sides there were circumstances of this character not necessary to be here stated.
Appellants claim that Terwilliger would have been further strengthened and appellee correspondingly weakened by the books of the First National Bank and certain testimony of Mr. Frink in connection with them, which was offered but not admitted. Although the journal and ledger would not have contradicted the statement of appellee or the entry in his pass book, yet they would have verified Terwilliger’s account of the manner of keeping them and tended to explain the entry in the pass book under date of the 28th as not of itself inconsistent with the claim that the deposit so credited was really made on the 27th. These are the only books of the bank in which the depositor’s name appears; but others were offered which it is claimed would have tended to identity the transaction and fix the day of its occurrence, viz., the cash book, and two figuring books kept by Terwilliger and Frink, respectively, — “ for figuring up cash, and figuring up items during the day to see that they agree with the cash on hand,” as Tenviiliger described them. These would show the transactions of each day separately, so far as they would show them at all.
It is said by counsel that the page offered, which was of the book kept by-Frink, contains only an unintelligible mass of figures and not a credit of $1,000 among them.
It may be quite unintelligible to others without explanation by the person who made it, and yet enable him to swear positively that it does show, and to explain satisfactorily how it shows, the receipt of $1,000 in one item on that day and none such on the next.
Mr. Frink having stated that he knew appellee by sight, but knew of his being at the bank that day only from the books, identified the figuring on the page for the 27th as made *494by him in the regular course of business, and was then asked to “ state whether there is included in the figures of that date a deposit of a thousand dollars from S. II. Stiles,” and also the following: “ After refreshing your recollection by reference to that book can you state to the jury whether or not a deposit was made in the name of Samuel II. Stiles in the First Hational Bank on the 27th day of September, 1880;” to each of which questions a general objection was sustained. In the further course of his examination, however, he did state that “ Samuel II. Stiles received a credit of $1,000 at the First Hational Bank of Belvidere on September 27,1880,” which statement, though ruled out upon like objection, served to show at least that the witness thought he understood the figures, and was prepared upon his faith in them to swear positively to the fact so understood, without any independent recollection of it.
The materiality of these books and of his testimony being sufficiently clear, the question is whether they were competent.
And first, as to the books. In the argument' for appellee the question of their admissibility as original evidence is not discussed. They are there treated solely as memoranda for refreshing the recollection of witnesses, and their incompetency as evidence of themselves is simply asserted as if it were too clear to require a reference to authority or the assignment of any reason. This may be dne to the fact that so little prominence or distinctness was given to it by the counsel for appellant; but we regret it not a little, and none the less because our first impression was as it remains, against him.
It is well known that memoranda and entiles made at or about the time of the transactions to which they relate, in the regular and usual course of business and of the employment and duty of the person who made them, have long been admitted as part of the ros gestae. Price v. The Earl of Torringford, 1 Smith’s Lead. Cas. with Hare and Wallace’s Notes (6th Am. Ed.) marg. p. 390, top 499 to 515; 1 Greenl. on Ev. (Redfield’s Ed.), Secs. 115-120, inclusive, and notes; also incidentally stated in §§ 147, 151.
*495Being held to presuppose the existence of better evidence in the oath of the person who made them, such entries were at first received only after his death. Price v. Torringford, supra Doe v. Turford, 3 B & Ad. 898; Pool v. Dicas, 1 Bing. N. C. 649; Welch v. Barrett, 15 Mass. 380; Brewster v. Doane, 2 Hill (N. Y.), 557; Nicholas v. Webb, 8 Wheat. 326, and other cases in the notes above referred to. In Massachusetts, insanity, and in Pennsylvania, Maryland and South Carolina, absence beyond the jurisdiction, have been held tantamount in this regard, to death; and the latter seems to have been so treated in this State. C. & N. W. R. R. Co. v. Ingersoll, 65 Ill. 399. In such cases they are admitted upon proof of his handwriting.
They are original legal evidence, and not hearsay; else the death of the person who made them would not render them competent. If, therefore, being alive and produced, he can not recollect the fact so recorded independently of the entry; after referring to it, the way would seem to be opened for its admission as secondary, just as though he were dead, insane or beyond the jurisdiction. Upon this reasoning the American authorities have extended the principle and admitted the writing in that ease during his life, upon the single-additional item of preliminary proof that it was truly made. 1 Greenl on Ev., Sec. 115, note 4, and Sec. 120, note 2; Merrill v. The Ithaca and Owego R. R. Co., 16 Wend. 587; Bank of Monroe v. Culver, 2 Hill, 532; Cole v. Jessup, 10 N. Y. (6 Sheld.) 96; Bunker v. Shed, 8 Metc. 150; Farmers and Mechanics Bank v. Boreaf, 1 Rawle, 152; Smith v. Lane, 12 Serg. and R. 84; Redden v. Spruance, 4 Harr. 265, 269; Bullard v. Wilson, 5 Martin (N. J.) 196 (3 Cond. La. 505); Spann v. Baltzell, 1 Florida, 302, 321; Underwood v. Parrott, 2 Texas, 168, 176. So in this State. In Humphreys v. Spear, 15 Ill. 275, Mr. Justice Treat delivering the opinion of the court said: “ It is well settled in this country that entries made by a clerk in the regular and usual course of business are admissible in evidence after his deatli on proof of his handwriting; and during his life if authenticated by him” — citing Have and Wallace’s Flotes and Greenleaf on Evidence, ubi supra.
*496Writings of this character are not only parts of the res gestae but, practically speaking, are often “ the best evidence of which the case in its nature is susceptible.” Those whose business it is to take part daily in similar transactions and make minutes of them at the time, may not long remember the particular transactions themselves, and certainly can not recollect such particulars thereof as date, amount, quantity or price, independently of their minutes. As was said in the opinion just referred to, “In the case of a running account at a store, consisting of numerous items, better evidence is not generally attainable.” And in the Bank of Monroe v. Culver^ 2 Hill, 532, in reference to bank books, the admissibility of which was in question, the court, per Bronson, J., said: “The memoranda and entries were made in the usual course of business and as a partof the proper employment of the witness. Ido not see how it is possible to doubt that such evidence ought to be received. * * * There are a multitude of transactions occurring every day in banks, the offices of insurance companies, merchants’ stores and other places, which, after the lapse of a very brief period, can not be proved any other way.”
In the case at bar the bank whose books were excluded had no interest in the proof sought to be made by them. The fact and amount of the deposit in question, as claimed by appellee, were never disputed. It is shown to have been paid back to him in full when called for, and there is no controversy between them. The fact to be proved, and which has come to be material in a controversy between him and another party, is simply the time when it was made — whether on the 27th or the 28th of September, 1880. The journal and ledger of the b mk are important only for the purpose of identifying the particular deposit credited oii the pass book, for neither of them purports to fix the date. In the journal the transactions of the two days are not distinguished. The ledger shows it of the 28 th; but it shows no deposit whatever as of the 27th) nor any for years except on Tuesdays, Thursdays and Saturdays, though the bank was all the time receiving them on other days as well — thus importing on its face that deposits of two days have been uniformly credited as of the last only. *497Now if the figuring book contained an entry of the character above described, which did clearly, though only incidentally, tend to fix the day, and the witness who made it was unable, of his independent recollection after referring to it, to swear to the time, we are of opinion that upon the principle held and illustrated by the authorities above cited it ought to have been received in connection with his testimony. This book was kept and preserved as much in the course of business as was the journal or ledger. Mr. Frink was a disinterested witness. He had been employed by the bank as bookkeeper from September, 1879. The figures offered were made by him on the 27th day of September, 1889, in the usual and regular course of business, as part of his employment and duty, and before any question as to the date of this deposit arose or was anticipated. He had no motive to make them incorrectly. Their truth was tested by the amount of cash on hand. Nothing appears to cast the slightest suspicion upon them. If he did not swear in direct terms that they were truly made, he did in effect, and that is sufficient. Presbyterian Church v. Emerson, 66 Ill. 269 (271). The objection to them and their exclusion were not based upon any defect of preliminary proof, but upon their supposed incurable incompetency. In the Merchants and Mechanics Bank v. Boreaf, 1 Rawle, 152, the appéllee sought to recover an alleged deposit of $800 which was credited on his bank book and by a witness sworn to have been made by him for the plaintiff as so credited. The bank claimed that the amount deposited was only $80 and that $800 was set down in his book by mistake. To prove this amount the bank offered the entry in its own book, supported by the oatli of its clerk that he received the deposit and made the entry a^ the time but had no independent recollection of the fact in question. The book was rejected; which the Supreme Court held to be error, saying: “ It is assumed that the clerk was able to swear that his entry in the book was true, to the best of his knowledge and belief.” With no less reason we may assume it here.
Just what these figures, if explained, would have tended to show we do not know with certainty, but for reasons hereafter *498state! presume it would have been pertinent; and as was observed by Mr. Justice McAllister, in C. & N. W. Ry. Co. v. Ingersoll, 65 Ill. 405, Many such entries are capable of explanation by applying to them evidence of the usages and customs of the business.”
As to the other books it does not appear that there was any attempt or formal offer to make preliminary proof with technical strictness, nor that this was the ground of their exclusion. From the manner in which they were identified, offered,objected to and rejected, we suppose the court and counsel understood it was at hand; and upon such proof, made or waived, we think they would have been competent.
It would be unnecessary to say that the deposit slip or ticket was properly admitted, except with reference to another trial and in view of the fact that counsel for appellee resisted it and still insists it was error. The cashier testified that he made it at the time of the deposit and truly, as part of his employment and duty — the first record of the transaction and its concomitant, according to the uniform course of business — and yet independently of it he could not recollect the date. All the required conditions of admissibility seem to have been fulfilled by this proof.
In this connection we may also notice a point made by appellee upon the bearing, as evidence, of the books of appellant. Tiiese were received without objection, but it is said they were not competent “ to prove negatively that an alleged fact did not exist because they had made no memorandum of it in the books of Yourt, Lawrence & Co.;” citing Lawhorn v Carter, 11 Bush. 7; Keim v. Bush, 5 Watts and Serg. 377; Morse v. Potter, 4 Gray, 292.
In the first cited the court uses the expression that “ mercantile books are never admissible to establish a negative proposition; ” but this is broader than was required by the case Or sustained by the authority relied on, which is the one in 4th Gray, where it was held that the time book of the employer, showing certain days on which the laborer worked, was not competent to show that he did not work on other days for which he also claimed, and the reason given was that it *499was a book of credit only and not of charges. The evidence in the cases in 11th Bush, and 5th Watts and Serg. consisted of books of a like character, the party in each attempting to avoid a charge for goods delivered by showing that his delivery book did not contain an entry of their receipt. We are not aware that it has ever been so held in "reference to books like those under consideration. In Furness v. Cope, 5 Bing. 114, bank books were admitted to show that a party had no funds there. In McLean Co. Bank v. Mitchell, 88 Ill. 54, the deposit blotter of original entries and the cash book were treated as competent evidence tending to show that an alleged deposit claimed by Mitchell was not made, and that its entry on the individual deposit ledger and transfer thence to his pass-book were by mistake. So was the effect of the book declared competent in the Farmers and Mechanics Bank v. Boreaf, supra. In the Bank of Monroe v. Culver, 2 Hill, 552, where it was claimed that a certain note had been discounted by the bank at usurious rates, and the books were offered to show among other things that such a note was never discounted by it, which was objected to, the court said: “ If the §2,000 note had been discounted by the plaintiffs that fact would, in the ordinary course of business, have appeared upon the books of the bank. The fact that there was no such entry in the books was, I think, proper evidence for the consideration of the jury.” In Mourse and wife v. McGay and another, 2 Rawle, 70, which was an action of ejectment, for the purpose of proving that an instrument, a copy of which had been put in evidence, was not genuine, the account book of the officer before whom it purported to have been acknowledged was admitted as competent and pertinent, because it showed charges for taking the acknowledgments of three other deeds for the same party, on the day that the one in question purported to have been taken, but none for taking that of the latter. And the reason for admitting negative evidence of this character is well stated in the American Hotes upon Smith’s Leading Cases, as follows: “It is obvious that the ground on which entries in the course of business are admitted is that as res gestee they afford a presump*500tion as to the other facts; and an omission, in a course of usual entry, is often as strong a fact as an entry.” Vol. I, (top) p. 506, (6th Am. Ed.)
If, then, the books are admissible at all, no objection is perceived to their use for what they may be worth for this purpose.
We have seen that Mr. Frink was prepared to explain his figuring book, and from his own entries made at the time, without actual recollection of the fact, to swear positively that appellee received credit for §1,000 at the First National Bank, on the 27th of September, 188), but was not 'allowed to do so. Was the proposed testimony competent?
Although as to matter of fact, a witness can testify only from his own knowledge and recollection, and not from hearsay, yet he may refresh and assist his memory by the use of a written instrument, memorandum or entry in a book. And the authorities recognize two kinds of recollection by aid of the writing, from which he may testify — the one being actual and the other imputed, one absolute and independent of the writing after reference to it, the other still dependent upon and inferred from it. When he finally testifies from the kind first mentioned, being the first class described by Gfreenleaf in Sec. 437, the character of the writing by which it is refreshed is immaterial. It may be one not made by witness, nor in the regular course of business, nor contemporaneous with the fact, nor even an original, nor need it be produced on the trial*. .And it is because it may be of such a character— not of itself capable of inspiring absolute confidence, but only of bringing to mind what he had known and forgotten — that the witness is required, after using it for that purpose, to speak from a present, actual, and independent recollection. This is the common case and the general rule. 1 G-reenl. Ev. Sec. 436.
But there are others which are exceptional, where such a recollection is not required, its place being supplied by what is deemed equivalent as a ground of assurance and accepted from necessity to prevent a failure of justice. Thus where the writing is an original, made at or about the time of the occurrence, *501and the witness recollects that he has seen it before, and that when he saw it he knew the contents to be correct, which is the second class described by Greenleaf, his testimony to the fact itself is received as if it was based on actual recollection; for 1ns knowledge of it is thereby shown to be actual and certain, as well, though inferentially, as if it were so based. Here, however, since he can not be cross-examined in relation to the fact or circumstances because he has no actual recollection of them, the writing on which he relies must be produced, in order that the other party may cross-examine in relation to it and have the benefit of every part of it. With this security against falsehood and error it is deemed safe, as well as necessary, to admit the testimony. 1 Greenleaf, Ev., Secs. 436-7 and note 3 to each. Our own Supreme Court has so declared it in relation to names, dates and amounts. C. & A. R. R. Co. v. Adler, 56 Ill. 348; Chicago and Wilmington Coal Co. v. Liddell, 69 Id. 641; Wolcott v. Heath, 78 Id. 434-5. Neither of the two cases cited by counsel as to the contrary, were of this character. In Elston v. Kennicott, 46 Ill. 187 (206), the material question was not merely whether the taxes had been paid, but whether they had been paid by Kennicott or under his title. The memorandum which was entered opposite the description of the land in the collector’s warrant, was ¿simply, “paid February 2d,” and of course could not enable the witness to say by whom paid, lie would not undertake to swear as to that, either from a direct recollection or from the memorandum. In C. & A. R. R. Co. v. Adler., su^tra, the writing used was a copy made two years after the original. The report of the case does not show when, with reference to the time of the occurrence, or by whom the original itself was made, nor its contents, nor what the witness testified. It is evident that in each of them the court regarded the writing as of the kind governed by the general rule, and that the language relied on by counsel was not intended to apply to the exceptional kind here considered. In the one last cited, the competency or testimony from contemporaneous memoranda, without any independent recollection, as to dates, was expressly declared. Authority in support *502of it from other States is abundant. Bartlett v. Hoyt, 33 N. H. 151; Haven v. Wendell, 11 Id. 112; Dower v. Rowell, 24 Vt. 346; Mattocks v. Lyman, 16 Id. 117; Dugan v. Mahoney, 11 Allen, 592; Crittenden v. Rogers, 8 Gray, 452; New Haven Co. Bank v. Mitchell, 15 Conn. 206 (224); The State v. Colwell, 3 R. I. 132; Halsey v. Linsbaugh, 15 N. Y. 485 (overruling Lawrence v. Baker, 5 Wend. 301 and Foster v. Heath, 11 Id. 485); Russell v. Hudson River R. R. Co., 17 Id. 134 (139); Peck v. Valentine, 94 Id. 569 (571); Martin v. Good, 14 Md. 409; Spring Garden Mut. Ins. Co. v. Evans, 15 Id. 62; The State v. Rawls, 2 Nott and McCord, 334; O’Neale v. Walton, 1 Richardson, 234; Bank of Tennessee v Cowan, 7 Humphrey, 70.
Whether Mr. Frink remembered that he had seen these figures before and then understood them and knew them to be correct, was not shown. He was not directly asked as to that. Having disclaimed any present and actual recollection of the deposit by appellee as revived by reference to them on the trial, all further examination in regard to them seems to have been cutoff. But if the case is not within the second class, as it probably is, it appears to have come fairly within the third described by Greetileaf (Sec. 347), as follows: “Where the writing in question neither is recognized by the witness as one which he remembers to have before seen, nor awakens his memory to the recollection of anything contained in it, but, nevertheless, knowing the writing to be genuine, his mind is so convinced that he is on that ground enabled to swear positively as to the fact;” oras more generally stated by Hare and Wallace in 1 Smith Lead. Cases, topp. 507, “ where the witness, after referring to the paper, undertakes to swear positively to the fact, yet not because he remembers it, but because of the confidence he has in the paper,” which seems to include both the second and third classes described by Greenleaf. In the latter, it seems that the memorandum must be not only an original, and contemporaneous with the fact, but also one made by the witness, or at least with his knowledge at the time. It was so in the examples given, and at the close of Sec. 346 he says: “But where the witness neither recol*503lects the fact, nor remembers to have recognized the written statement as true, ancL the writing was not made by him, his testimony, so far as it is founded on the written paper, is but hearsay.” Where the nature of the business carried on requires many entries of the same general character every day, as is the case with banks, it must often be that the person who makes them and at the time knows them to be correct, can not, after the lapse of even a brief period, actually remember a particular transaction so entered, or the making of the entry, nor any occasion when he afterward saw it and then recognized it as being true. Neither the transaction itself nor the entry of it, after being made, may be ever again brought to his attention until he is called to testify; and yet from the book, the order of dates and entries, the position of the one in question, the course of business, and his own habits in respect to such transactions, he may, upon seeing the entry in his own handwriting, be more certain both of the fact and the date of its occurrence than he would be of the particulars of almost any similar transaction so long past, of which he would testify only from what he believed to be actual recollection. The same is true of isolated transactions. Men sometimes give receipts for money paid to them, and sometimes they do not. When they do, the receipt passes at once out of their possession and thoughts. Afterward they may forget the payment, and when shown the receipt may have no recollection of having given it, but, knowing its genuineness, entertain no doubt of the fact expressed by it. So one may attest the execution of a deed by signing his name as a witness, and afterward, on being shown his name so signed, recall nothing of the fact of execution or of his signing his own name as a witness, but have no doubt of the fact from the known genuineness of his signature shown. In such cases the testimony of the witness is received. Thus, in Dugan v. Mahoney, 31 Allen, 572, which was an action for goods sold and delivered, two witnesses were called to prove the delivery of numerous parcels. Neither had any recollection of the facts, but were able to swear to them positively from certain entries and checks made by them in memorandum books in the regular *504and usual course of business. It does not appear that they remembered having seen these entries before and that when they saw them they knew them to be correct, but they knew they were made by them and were positive that they would not have been made if the fact had not been known to have occurred. That seems to be the kind and ground of the witnesses’ knowledge in the third class of cases described by Greenleaf, as distinguished from that in the second, if indeed ¡there is any substantial distinction between them. Here the .Supreme Court of Massachusetts said: “It is obvious that this species of evidence must be admissible in regard to numbers, dates, sales and deliveries of goods, payments and receipts of money, accounts and the like, in respect to which no memory could be expected to be sufficiently retentive without depending upon memoranda; and even memoranda would not bring transactions to present recollection. In,such cases, if the witness, on looking at the writing, is able to testify that .lie knows the transaction took place, though he has no present memory of it, his testimony is admissible. A similar rule is applied to attesting witnesses who, if they know from seeing their handwriting that they witnessed the execution of the instrument, may so testify, although the sight of the handwriting does not bring to their recollection even the fact of attestation.” Citing Rex v. St. Martins, 2 Ad. and El. 210; Crittenden v. Rogers, 8 Gray, 452. In Martin v. Good, 14 Md. 398 (409 et seq.), the subject is carefully considered, and upon many authorities cited the statement of the law by Greenleaf in the third class and herein above quoted, is held to be correct. In England the same doctrine in substance was settled by the case of Mangham v. Hubbard et al., 8 Barn, and Cress. 14; Powell on Evidence, 308.
We are of opinion that whether the figuring and cash book and ledger and journal of the First National Bank were or were not admissible, the testimony of Frink should have been received. 1 Greenl. Ev., Sec. 115; Dugan v. Mahoney and Mangham v. Hubbard, supra. Though based on the writing and without actual recollection, it was not in the legal sense matter of inference merely. The same may be said of much *505of the testimony given by Irving and Jam es Ter wi 11 i ger, based on memoranda. It was positive evidence, and therefore it was improper to give the 4th and 5tli instructions asked by the defendant. There is no doubt about the correctness of the principle that the mere inference of witnesses are not evidence, but it was here misapplied. Wolcott v. Heath, 78 Ill. 433. These instructions must have discredited this testimony if they did not wholly exclude it from the consideration of the jury. For the errors above indicated the judgment of the circuit court will he reversed and the cause remanded.
¡Reversed and remanded.