Knowlton, J.
To maintain his case the plaintiff was obliged to show that the flag hung across the street was an object which was so likely to frighten horses as to render driving upon the street unsafe, and that in its position there it was a public nuisance. The fundamental question in the case was whether ordinarily safe and gentle horses would be frightened by it. The inquiry was in regard to the effect of an inanimate object upon an animal acting from instinct. The only way in which knowledge on this subject could ever be acquired is by observation of the effect of the object, or of similar objects, upon the animal. Inasmuch as no two flags hung in different places with different surroundings could ever present precisely the same appearance in different aspects to an unreasoning animal, the most satisfactory way of ascertaining the fact would be by observing the effect of this particular flag upon different horses. In all the observations and experiments, one factor in the problem, the swinging flag, would always be the same. The other factor, the horse, would always truly exhibit his real feelings, and the only possible difference in the results of different observations would arise from the difference in the horses. The question of fact whether a particular horse comes within the class of ordinarily safe and gentle horses is not difficult or complicated, and witnesses could easily give the results of their *344observations of the conduct of horses which they considered ordinarily safe and gentle. We are of opinion- that the best way to decide the main question in dispute is to show whether ordinary horses have manifested fear of the flag as it hung over the street. The question is not whether the results of experiments with other ordinary horses might be introduced upon the question whether the flag frightened the plaintiff’s horse, although there is much authority for holding that, where the elements entering into the experiments are so nearly the same, the results may be shown to establish a fact of this kind; but the question is how a certain kind of animal is commonly affected by the sight of a particular object. To ascertain the truth, the jury must either use such knowledge as they happen to have on the subject without the aid of testimony, or experts must be called to give their opinions if the subject is one in regard to which experts can be found, or witnesses must be permitted to state particular facts which they have observed, each one of which is an illustration and example of the general fact in dispute. The only objection to testimony of the last kind in such a case is that in testing it collateral issues may.be raised. Such an objection in many cases is a sufficient reason for excluding the testimony. Whenever a line of inquiry will give rise to collateral issues of such number or difficulty that they will be likely to confuse and distract the jury, and unreasonably protract the trial, it should not be permitted. But the mere fact that a collateral issue may be raised is not of itself enough to justify the exclusion of evidence which bears upon the issue on trial. Most circumstantial evidence introduces collateral issues, and ordinarily it is a practical question, depending upon its relation to the other facts and circumstances in the case, whether it should be received. It may be remote from the real issue, or closely connected with it, and in many cases its competency depends upon the decision of questions of fact affecting the practical administration of justice in the particular case such that a court of law will refuse to revise the ruling of the presiding judge, but will treat his ruling as a matter of discretion.
In the present case the only collateral inquiry which could arise is whether a horse called by a witness an ordinarily safe and gentle horse comes within that class. Such an inquiry is *345certainly simple. We think there would be no practical difficulty in receiving and weighing testimony in regard to the conduct of horses which seem to be like ordinary horses in common use.
This precise question has been decided in favor of the plaintiff’s contention by many courts of the highest respectability, and we have been referred to no decisions to the contrary. In Brown v. Eastern & Midlands Railway, 22 Q. B. D. 391, 393, which was an action for an injury caused by the shying of the plaintiff’s horse at a heap of dirt, the Court of Queen’s Bench held that the plaintiff was rightly permitted to show that various other horses had previously shied at the same place, and all the judges of the Court of Appeal “ were clearly of opinion that the evidence was admissible, and affirmed the decision of the Queen’s Bench Division.” Crocker v. McGregor, 76 Maine, 282, is to the same effect. House v. Metcalf, 27 Conn. 631, was a suit for maintaining a wheel which frightened the plaintiff’s horse. The court says the plaintiff “ had a right, not only to show the facts regarding its size, form, location, exposure to view, and mode of operation, from which the jury might infer what effects it would naturally, necessarily, or probably produce, but also to prove what effects it had produced in fact. . . . The inquiry in every such case is, not whether the evidence offered is sufficient to prove the fact claimed, but whether it tends to prove it.” In Darling v. Westmoreland, 52 N. H. 401, a suit for damages caused by the fright of a horse at a pile of lumber, evidence was received that other horses had been frightened by the same pile. The justices of the Supreme Court of New York who sat in Champlin v. Penn Yan, 34 Hun, 33, 37, unanimously sustained the admission of evidence “ that on another occasion, prior to this accident, a flag similar to this in appearance, suspended over the same street and in a similar manner, did frighten other horses when driven along the street under the same.” The Court of Appeals of New York takes a similar view of the law. Quinlan v. Utica, 11 Hun, 217; S. C. 74 N. Y. 603. Wooley v. Grand Street & Newtown Railroad, 83 N. Y. 121.
The defendant relies upon a line of cases in this Common, wealth, brought against cities or towns to recover for accident» *346received while travelling on highways, in which it has been held that a plaintiff cannot introduce evidence of other similar accidents occurring at the place where he was hurt for the purpose of proving that the way was defective. Collins v. Dorchester, 6 Cush. 396. Hall v. Lowell, 10 Cush. 260. Aldrich v. Pelham, 1 Gray, 510. Kidder v. Dunstable, 11 Gray, 342. Hinckley v. Barnstable, 109 Mass. 126. Schoonmaker v. Wilbraham, 110 Mass. 134. Merrill v. Bradford, 110 Mass. 505. The ground on which these cases were decided is, that such collateral inquiries would be opened before the evidence could be properly received as would multiply issues for the trial of which the parties had had no opportunity to prepare, and, would lead away from the main issue and tend to confuse the jury. In most of these cases the facts and circumstances of other accidents were so diverse and complicated that the decisions rest on grounds which are generally deemed satisfactory. In others, if they were to be considered apart from authority, it may be that an effect of an attempt to pass on another occasion was so closely connected with the alleged defect, and so free from other possible contributing causes, that, as a simple experiment, it might well have been proved. Such evidence has sometimes been received in other jurisdictions. District of Columbia v. Armes, 107 U. S. 519, 524. Morse v. Richmond, 41 Vt. 435. Darling v. Westmoreland, 52 N. H. 401. Calkins v. Hartford, 33 Conn. 57. Quinlan v. Utica, 11 Hun, 217; S. C. 74 N. Y. 603. Delphi v. Lowery, 74 Ind. 520. Chicago v. Powers, 42 Ill. 169. Moore v. Burlington, 49 Iowa, 136. Augusta v. Hafers, 61 Ga. 48. In this case we have no occasion to consider whether the strict construction put by this court in former years upon the statute giving damages for accidents caused by defects in highways, and the disinclination to look with favor upon claims brought under it, have led in some such cases to too great an extension of the rule excluding testimony involving collateral inquiries. This is not an action against a city or town. It presents a simple question in the law of evidence. It should be decided in accordance with sound principles, and this court has established precedents in favor of the plaintiff’s contention that accord with those which we have already cited from other courts. In Reeve v. Dennett, 145 Mass. 23, upon the *347question of the effect of the use of a certain medicine in dentistry, evidence was received that dental operations performed by a certain dentist who used the medicine were less painful than those performed by other dentists who did not use it. In Brierly v. Davol Mills, 128 Mass. 291, to prove that an attachment would be effective on a certain loom, it was held competent to show that it worked successfully on another loom of similar construction. See also Gahagan v. Boston & Lowell Railroad, 1 Allen, 187; Hunt v. Lowell Gas Light Co. 8 Allen, 169; Commonwealth v. Goodman, 97 Mass. 117, 119; Fay v. Whitman, 100 Mass. 76; Hodgkins v. Chappell, 128 Mass. 197; Baxter v. Doe, 142 Mass. 558; Commonwealth v. Leach, 156 Mass. 99; Shea v. Glendale Elastic Fabrics Co., post, 463.
A majority of the court are of opinion that the evidence offered should have been admitted. Exceptions sustained.
