SHAW, J.

This is an action for damages to certain goods of the plaintiffs while situated in a room on the first floor of a building in San Francisco, alleged to have been caused by the flooding of the premises with water which escaped from an engine operated by the defendants upon an upper floor in the same building. The defendant is charged with negligence in the selection, management, and control of said engine and the boiler and pipes connected therewith. The answer denies the damage and the negligence on the part of the defendant. The cause was tried by a jury, and a verdict rendered in favor of the plaintiffs in the sum of $850, upon which judgment was entered. The defendant moved for a new trial, and, the motion being denied, it appeals from the judgment and from the order denying its motion for a new trial. No question is made concerning the sufficiency of the complaint. The points arising in the case are presented upon the decision of the motion for a new trial.

The evidence is sufficient to justify the verdict. The contention is, that conceding that the occurrence of the accident was prima facie evidence of negligence, yet this prima facie case was rebutted by uncontradicted evidence satisfactorily explaining the occurrence, and showing that it was not through the negligence of the defendant. The primary cause of the leakage of water was the breakage or warping of a metal cap at the end of one of the water-pines projecting into the fire-box of the defendant's engine, the effect of which was, that the water pumped into the boiler escaped through the pipe and ran into the room occupied by the plaintiffs. The engine had attached thereto an automatic pump, so constructed that it would pump water into the boiler until it reached a certain height, whereupon the pump would automatically close, and remain closed until the height was decreased sufficiently to open it again automatically. Thus, so long as the boiler was tight, the water in the boiler would never get above a certain height or be entirely exhausted. The effect of the leakage, however, was to prevent the water from getting to such height as to automatically shut off the pump. The leakage occurred during the night-time, and the consequence was, that the pump, continuing in operation during the night, flooded the plaintiffs' room. There was a stop-cock upon the water-pipe which could have been turned off, and this would have prevented any water from entering the boiler. The explanation given by the defendant, which, it is claimed, sufficiently rebuts the prima facie case of negligence, was, that the instructions of the machinist who attached the automatic pump were, that the defendant should not turn off the stop-cock during the night; that there was more danger in having it turned off than there was in having it turned on.

If the failure to turn off the stop-cock was the only negligence which could have caused the leakage, there might be some force in this contention; but the evidence shows, as before stated, that the primary cause was the warping or breaking of the cap which caused the leak, and there is other evidence tending to show that this was caused by the action of fire, and that warping or breakage would more readily occur if there had been sediment accumulated at the lower end of the pipe just above the cap. The inference from this would naturally be, either that there had been a fire in the fire-box when there was no water in the boiler or that sediment had been allowed to accumulate over the cap, thus rendering the action of the fire more direct and violent. In view of these circumstances, it cannot be said that the explanation given destroyed the prima facie case of negligence. Besides, there was evidence of admissions by the managers of the defendant that the accident was caused by a negligent operation and control of the engine.

The defendant complains of an instruction of the court, that the mere fact of the overflow of water from the boiler, under the circumstances of the case, was prima facie evidence of negligence. The rule applicable to this question is thus stated in Judson v. Giant Powder Co., 107 Cal. 549, quoting from Shearman and Redfield on Negligence (sec. 60): "When a thing which causes injury is shown to be under the management of defendant, and the accident is such as in the ordinary course of things does not happen if those who have the management used proper care, it affords reasonable evidence, in the absence of explanation by the defendant, that the accident arose from a want of care." It appeared from the evidence that this engine had been operated without injury, under the same circumstances, for three or four years before the accident happened. This sufficiently demonstrated that in the ordinary course of things such accidents did not happen, and was a justification of the instruction complained of. A large number of cases are cited from other states, apparently holding the contrary to this doctrine. It is sufficient to say that the rule is settled the other way in this state. (Butcher v. Vaca Valley etc. R.L.R. Co. 67 Cal. 518.) The cases cited which refer to the rule concerning explosions of boilers are based upon the proposition that such explosions generally happen from obscure and undiscoverable causes which ordinary human prudence and care can neither discover nor guard against. This reason is absent from the present case. The cap which caused the injury was visible, and, if it melted from the action of fire, it was fairly to be inferred, as above stated, either that the defendant had carelessly allowed the fire to burn in the box when there was no water in the pipe or that it had been negligent in allowing sediment to settle in the pipe, thus causing the same to heat more readily.

There was no conflict in the instructions given by the court. The instructions which are claimed to be conflicting are as follows: "The burden to prove lack of care, and to prove negligence, which is the same thing, is on the plaintiffs throughout this case"; also, "Unless they, the plaintiffs, have shown that the boiler and apparatus which was used by the defendant were improper appliances to be used, or that they were maintained or used by the defendant in a careless or negligent manner, and unless they have so shown to your satisfaction and by a preponderance of the evidence, you are instructed that you must find for the defendant." It is contended that these instructions, and others of similar import, are in direct conflict with the instruction mentioned in the previous paragraph, to the effect that the happening of the accident was prima facie evidence of negligence. This contention, however, is founded upon a failure to perceive the effect of the presumption as evidence. The negligence could have been established by direct proof of some act which, in its nature, was negligent, or it could have been established, as it was established, prima facie, by the mere proof of the happening of the accident and the circumstances surrounding it, from which the presumption of negligence arose. This presumption is itself evidence in the case, and it does not change the rule as to the burden of proof. It is merely the evidence by which the plaintiffs undertook to establish the fact which they were bound to prove. It still remained the law that, upon the whole evidence, the plaintiffs must have the preponderance in order to succeed.

The plaintiff Lazare Klein had testified to the value of the goods destroyed. On cross-examination, he was asked the question: "How much did the goods cost, according to the invoice?" An objection to this question was sustained. The question was proper cross-examination, and the exclusion of the answer would have been error, but it was cured by the introduction of the invoice itself, which completely answered the question, and upon its face showed what the value was according to the invoice.

The damage occurred on the 5th of January, and the plaintiffs kept the goods in the store until the 27th. It was claimed by the defendant that this enhanced the damages. Julius Wolf, an employee of the plaintiffs, was testifying in their behalf, and was asked whether, during the time between January 5th and January 27th, the firm had been negotiating with the defendant with reference to a settlement. It was stated that this evidence was sought for the purpose of explaining why the goods were allowed to remain in the store during the interval, and upon that statement an objection of the defendant was overruled. The witness answered the question, but in so doing he went beyond the question itself, and proceeded to give the details of the negotiations for a settlement, stating how much was offered by each side. Upon this, the defendant interrupted, and moved to strike out the portions of the answer referring to the details of the efforts to compromise. The motion was denied. There was then some colloquy concerning the matter, and the witness gave some further details, claiming that he had not finished his answer, whereupon the plaintiffs' attorney, referring to the motion to strike out the objectionable part of the testimony, stated to the court that he desired to withdraw the testimony objected to. To this defendant's counsel objected, and upon his objection the court denied leave to the plaintiffs to withdraw the testimony. Under these circumstances, the defendant cannot claim that there was error in denying its motion to strike out. If it desired the testimony to be stricken out, it should have consented to its withdrawal. After having refused to consent to the withdrawal, and by means of its refusal caused the evidence to remain before the jury, it cannot be heard to urge that there was error in denying its motion to strike it out.

There were two inventories introduced in evidence, — one taken by three appraisers on the 6th of January, and one made by Wolf on the afternoon of January 5th, being the next day after the overflow. Mr. Wolf, while on the stand, had testified that there were five pieces in the appraisers' inventory which were not in the inventory made by him. On cross-examination, he was asked to point out these five pieces. He answered that he could select one, but that to find the others would require his going over the inventories, and that it would take some thirty minutes or more. The court thereupon directed the counsel for defendant to continue the cross-examination on some other point, and refused to allow the cross-examination to stop for the thirty minutes that would be required to enable the witness to find the four pieces not included in his inventory. This was a matter within the discretion of the court, and we cannot say that the discretion was abused in this instance. The cross-examination was resumed and continued at considerable length after this refusal. The defendant's counsel did not ask to have the witness recalled to testify to the missing pieces, nor did he ask that the witness be directed to look for the missing pieces and remain for further examination. The court did not refuse to allow him to testify on the subject at all, but merely directed the counsel to proceed with the cross-examination on some other point, and intimated that the counsel might afterwards renew the inquiry. Both inventories were introduced in evidence, and counsel could have discovered the discrepancy without taking up the time of the court and jury in waiting for the witness to make the discovery. We cannot say the defendant was prejudiced.

Complaint is made of the action of the court in sustaining objections to several questions put to the witness Gilman as an expert machinist. He had made repairs to the engine about two months before the accident, and on the morning after that occurrence he had examined the engine, found that the cap which caused the leak was broken, and had removed it from the pipe. The questions were in the same form and to the same point. The first and last only need be quoted. The first one was as follows: "Assuming the fire was put out the night before, and that there was no leakage, state whether you as an expert, or any other person, could have told that the thing that did happen was going to happen, by an examination?" and the last one was thus: "Could any person have told that that cap would leak or come off, by an examination five minutes before it occurred, or gave way?" The objection was, that it was not the proper subject of opinion evidence. The form of the question called for a simple negative or affirmative, and an answer would have involved the conclusions of the witness upon several questions of fact. He would have had to determine, — 1. The true cause of the break or leak; 2. Whether or not, from the construction and material of the pipe, cap, and surrounding parts of the engine, there would have been visible or perceptible to either of the five senses any previous indications of such cause; and 3. The ability or mental capacity of any person, or of an expert, to foresee, from such indications, that a leak or break would probably occur. It would have been impossible for the jury to know upon which of these propositions the answer was based. The third is clearly not the subject of opinion evidence, and the second may or may not have been so, accordinig to what his opinion was as to the cause and as to the character of the previous indications. Under these circumstances, there was no error in sustaining the objections.

The defendant asked the following instruction: "The defendant in this case has a perfect right to use such a boiler or combination as it sees fit. It is not bound to furnish the safest, newest mechanical contrivances, but is bound to use reasonable care and prudence in providing machinery reasonably safe and suitable, and also reasonable care and prudence in keeping the same in repair. It is only bound to use ordinary care in maintaining safe machinery and appliances." The court modified the instruction by eliminating portions of it, giving it as follows: "The defendant in this case has a perfect right to use such a combination or boiler as it sees fit, provided that it exercises reasonable care in its selection, maintenance, and use." The instruction as given stated the law; and while it is not as full and complete as the one asked, we think it has the same meaning and effect, and covers the same points.

The defendant asked the court to instruct the jury that "the fact that the boiler was purchased of reputable manufacturers, and that reputable mechanics were called to examine and repair it, is proper to be considered as tending to justify its use." The instruction was refused. Evidence of the fact referred to in the instruction had been given in the case, and it constituted part of the evidence which the jury were instructed to weigh, and was necessarily included in the evidence which they were directed to consider when the court instructed them that the plaintiffs were bound to prove the negligence alleged, by preponderance of the evidence. Furthermore, it was an instruction upon a matter of fact. It directed the jury that they should consider this fact as evidence tending to justify the use of the boiler. The court was therefore justified in refusing the instruction, as a violation of the constitutional mandate not to charge upon matters of fact.

One of the witnesses, in response to the question by a juror, had testified that, after the overflow occurred, the defendant had shut off the stop-cock at night, so as to prevent water from flowing into the boiler. The defendant asked the court to instruct the jury that they must not consider this fact in determining whether or not "the defendant exercised due care prior to the accident." The court refused to so instruct the jury, but, instead thereof, gave the instruction that this fact must not be considered "as a confession of negligence." The defendant was entitled to the instruction as asked. The care due from the defendant is not to be measured by its conduct after the overflow had demonstrated the necessity of further precautions, but by the conditions existing at and before the time of the occurrence. If both instructions had been before them, the jury might have been misled to suppose that, although not to be considered as a confession of negligence, it was evidence of some weight, to be considered by them for some purpose. But they knew nothing of the instruction refused, and were not required to attempt to distinguish between the two. They were correctly instructed that the subsequent precaution was not to be considered as a confession of previous neglect. If the fact could be considered as of any weight, it would be because of the inference arising therefrom that the decision to turn off the stop-cock in the future was made from the consciousness that the failure to do so in the past was neglect, and in that respect it would be a tacit confession, evidenced by conduct. A jury is not expected to indulge in metaphysical distinctions so minute as this. The instruction given would be understood by the ordinary mind as substantially a direction to disregard the fact in question. This is all that the court was required to do, and it is not material that it was not done in language as accurate and precise in the instruction given as in that asked.

There are some nine objections to evidence in addition to those above mentioned, concerning which the defendant makes the following argument: "In conclusion, we desire, without making any argument thereon, to call attention to some other matters wherein we believe error was committed; nor do we wish to be understood as not insisting upon these points because we have not discussed them at length. Transcript pages 150(b), 151(d), 153(f), 155(i), 156(k), 158(p), 159(q), 159(r), 159(t)." The pages referred to in the transcript are the pages of the statement where the errors are assigned upon the exceptions to which they refer, and there is no reference, either in the brief or in those pages of the transcript, to any particular portion of the statement where the matter is set forth at large. Upon this argument, we do not think the court is justified in discussing the errors alleged.

We find no error in the record substantially affecting the case upon its merits, and the judgment and order are therefore affirmed.

Angellotti, J., and Van Dyke, J., concurred.