to it on principle. Reay v. Richardson, 2 C., M. & R. 422; Hall v. Davis, 36 N. H. 569; Emery v. Webster, 42 Maine, 204; Baldwin v. Carter, 17 Ct. 201; Drake v. Goree, 22 Ala. 409; Cowles v. Garrett, 30 Ala. 348; Waterman v. Johnson, 13 Pick. 261; Mechs.' Bank v. Bank of Columbia, 5 Wheat. 326; Jennings v. Sherwood, 8 Ct. 122; 1 Greenl. Ev. §§ 286, 287, and 288.

The general rule is well stated by the Supreme Court of New Hampshire, in Hall v. Davis, as follows:

"As all written instruments are to be interpreted according to their subject matter, and such construction given them as will carry out the intention of the parties, whenever it is legally possible to do so, consistently with the language of the instruments themselves, parol or verbal testimony may be resorted to, to ascertain the nature and qualities of the subject matter of those instruments, to explain the circumstances surrounding the parties, and to explain the instruments themselves by showing the situation of the parties in all their relations to persons and things around them. Thus if the language of the instrument is applicable to several persons, to several parcels of land, to several species of goods, to several monuments, boundaries or lines, to several writings, or the terms be vague and general, or have divers meanings, in all these and the like cases, parol evidence is admissible of any extrinsic circumstances tending to show what person or persons, or what things, were intended by the party, or to ascertain his meaning in any other respect; and this without any infringement of the general rule, which only excludes parol evidence of other language, declaring the meaning of the parties, than that which is contained in the instrument itself."'

If evidence of surrounding facts and circumstances is admitted to explain the sense in which the words were used, certainly proof of the declarations of the parties, made at the time of their understanding of them, ought not to be excluded. And so it was held in several of the cases above cited. 2 C., M. & R. 422; 42 Maine, 204; 13 Pick., 261. Such declarations, if satisfactorily established, would seem to be stronger and more conclusive evidence of the intention of the parties than proof of facts and circumstances, since they come more nearly to direct evidence than any to be obtained, whilst the other is but circumstantial. And though in general the construction of a written instrument is a matter of law. for the court-the meaning to be collected from the instrument itself; yet, where the meaning is to be judged of by extrinsic evidence, the construction is usually a question for the jury. Jennings v. Sherwood, and other cases above. The circuit judge was therefore right in receiving parol evidence to ascertain the sense in

which the word was used by the parties, and in submitting that question to the decision of the jury.

But he was clearly wrong in receiving evidence of the statements of the plaintiff's agent to the witness Gunn, at the time of making the contract with him. The occasions were different-the two contracts entirely disconnected, and though both concerned a machine of the same pattern and manufacture, yet what was said in the one case was not a part of the transaction in the other. It was no part of the res gesta. If the agent Chase, in negotiating with Gunn, had made an admission of his representations to the plaintiff, evidence of such admission could not have been received. Mil. and Miss. R. R. Co. v. Finney, 10 Wis. 388. It would be going much too far, were we to hold that it was proper to give the jury the agent's statement to Gunn, as evidence tending to prove that a similar statement was made to the plaintiff. If it has any such tendency, it is so remote that the law cannot lay hold of and apply it.

The question then comes up, must the judgment, for this reason, be reversed? The defendant's counsel insist not-that the evidence before the jury was sufficient without this, and if it had been rejected, the verdict must have been the same. We are inclined to take the same view. The defendant's testimony was clear and positive as to the kind of team-that the agent said "one span of horses" would work the machine up to the warranty. In this he was not contradicted, but rather corroborated by the agent, who was himself upon the stand. We would naturally expect, if the fact had been otherwise, the agent would have said so. On the other hand, he testifies very frankly that the defendant said he had but one team; that he told him one good team would work the machine. The admission of the improper evidence could not, therefore, have affected the finding of the jury upon this point; and consequently the plaintiffs were not prejudiced by it.

We can hardly believe that the argument of the plaintiff's counsel upon the construction of the warranty, that it referred to the capacity of the machine without regard to the kind of team employed, and was satisfied, if, under any circumstances, and with any number of horses, it could be made to perform as alleged, was urged with any real hope of success. Such a construction would be directly opposed to the manifest intention of the parties.

The jury, upon proper evidence, and under proper instructions having found that the machine delivered at Milwaukee was not

such as the contract called for, the judgment upon their verdict must be affirmed. Ordered accordingly.18

Evidence as to the usages of trade.

SOUTIER v. KELLERMAN.

18 MISSOURI, 509.-1853.

GAMBLE, J. The plaintiff alleges that he bought of the defendant (Kellerman) 4,000 shingles, and that he received eight bundles or packs, which only contained 2,500, and having paid for 4,000 brought this suit to recover the value of the number deficient. The defense made by Kellerman was, that by the custom of the lumber trade, two packs of a certain size are regarded as a thousand shingles, and are always bought and sold as such, without any count of the number, and that the eight packs delivered to Soutier were, according to such custom, properly reckoned as four thousand shingles.

1. The defendant asked the court below to declare the law in relation to the effect of the usage of the trade, and for that purpose presented two instructions, which the court refused. As this was a case brought into the court by appeal from a justice of the peace, the code of practice, which is not applicable to proceedings before a justice, is not applicable to the trial before the law commissioner on appeal. Such case is to be tried on the merits de novo, and the practice formerly prevailing, in trials by the court without a jury, of asking declarations of the law, is, in such cases, still to be pursued.

2. The defendant asked the court to declare the law as follows: (1) That if the shingles sold to the plaintiff were in ordinary sized packs, and that the price paid was a reasonable price for such kinds of packs, and that such packs are, by common custom, sold two for a thousand, then the plaintiff is not entitled to recover. (2) If the common custom of the lumber trade is to sell two bunches of shingles as a thousand, without regard to actual count, then the plaintiff must be presumed to have had notice of such general custom, and to have purchased accordingly. The court refused to make these declarations of the law, and, on the contrary, declared: 18 That parol evidence may be introduced to identify a person named in an instrument, see Andrews v. Dyer, 81 Me. 104.

"That if the contract was at so much per thousand, and not so much per bundle, and that no express agreement was entered into that two bundles should represent a thousand, then the defendant must deliver the four thousand, or else account to the plaintiff for their value."

The usage of a particular trade is evidence from which the intention and agreement of the parties may be implied; and, although it cannot control an express contract, made in such terms as to be entirely inconsistent with it, yet, in express contracts, the terms employed may have their true meaning and force best understood by reference to such usage. Evidence of such usage is admitted, not to vary the terms of an express contract or to change its obligation, but to determine the meaning and obligation of the contract as made. The usage must appear to be so general and well established, that knowledge of it may be presumed to exist among those dealing in the business to which it applies, so that the contract of the parties may be taken to have been made with reference to it. In this country, many articles which are in terms sold by the bushel (a dry measure, containing eight gallons) are, in fact, sold by weight; the bushel being understood to mean a certain number of pounds, and the number of pounds differing in different articles, as salt, wheat, etc. When such custom becomes general and well established, so as to be known to the community, it is obvious that a contract for a given number of bushels must mean the bushel as ascertained by weight, whether in fact the number of pounds of the article sold would measure more or less than the real bushel. The rule here stated is laid down with great distinctness, in 3 Starkie's Ev. 1033, and applied in Smith v. Wilson (3 Barn. & Adolph. 728) to a case where 1,000 rabbits was held to mean 1,200.

In the present case there was evidence that a general custom prevailed in the lumber trade of estimating two packs of shingles, of certain dimensions, as a thousand shingles, without reference to the number of pieces in the pack. If such was the usage of the trade, so general and well established that those buying and selling might be presumed to deal in reference to it, there does not appear to have been any such contract shown in this case as would prevent the usage from applying. The law commissioner seems to have thought that the defendant could not escape from liability "if the contract was at so much per thousand" unless there was "an express agreement that two bundles should represent a thousand." This was an incorrect statement of the law, in a case where evidence was

given of a general usage, that a thousand shingles meant two packs of certain dimensions. Whether there was as full evidence of the usage given as ought to have been given, is not a question upon which we pass, but there was evidence of the usage upon which the party was entitled to have the law differently declared, if the evidence proved the usage as general, well established, and known, so that contracts might be presumed to be made with reference to it. It was not necessary that the defendant should show an express agreement that two bundles should represent a thousand.

The judgment is reversed, with the concurrence of the other judges, and the cause remanded.19

ENGINEERING CASES: EXCERPTS FROM DECISIONS.

EVIDENCE.

Testimony as to information obtained from scientific text-books is admissible in court. Hardiman v. Brown, 39 N. E. Rep. 192 (Mass.).

Maps, profiles, cross-sections, etc., of engineering constructions may be allowed to go to the jury at the discretion of the court, if they have been properly introduced as evidence. Cunningham v. Massena R. Co., 18 N. Y. Supp. 600.

Scientific books cannot generally be used as evidence to prove the statements they contain. Johnston v. Richmond & D. R. Co., 22 S. E. Rep. 694.

The testimony of a witness that he had obtained rules from standard text-books for the excavation of cuts and filling of embankments was considered competent, and the rules were admitted as evidence. Central R. R. Co. v. Mitchel, 63 Ga. 173.

Testimony of facts must be from recollection. An engineer may refresh his memory from his notes, but he may not read his memoranda. Wilde v. Hexter, 50 Barbour 448.

Where an architect made measurements and took data while the

19 Sweeney v. Thomason, 9 Lea (Tenn.), 359. See also Walls v. Bailey, 49 N. Y. 464; Hubble v, Cole, 85 Va, 87,