delivered the opinion of the court.
Appellant owns and operates a canning factory. For the purpose of cooking its products, it used what is designated as a “ cooker,” known as “ Baker’s process kettle,” a cylindrical apparatus made of iron, in which the vegetables previously canned are placed for the purpose of subjecting them to a sufficient degree of heat to complete the process of canning. This is accomplished by closing the cooker perfectly tight and turning in steam which is supplied from boilers through a pipe connecting them with it. The steam thus confined supplies the required degree of heat. Appellee’s son, a minor, was an employé of the appellant, engaged in operating the cooker, which, without fault upon his part, exploded, inflicting injuries which resulted in his death. In the first count of her complaint, she bases a right to a recovery upon the ground that the neglect of the appellant in failing to equip the cooker with proper safety appliances was negligence upon its part, which was the proximate cause of the explosion. This issue was submitted to a jury.
*68Appellee claims that the explosion was dne to the excessive pressure of steam confined in the cooker, which would have been avoided had it been provided with a safety valve, so regulated as to allow the escape of the steam upon reaching a pressure within the limits of safety, or the pipe connecting the cooker with the boiler had been equipped with a pressure regulator which would have automatically prevented the pressure exerted by the steam reaching beyond a fixed point. Appellant contends that the evidence fails to establish that it was negligence upon its part not to use either of these appliances; that it does not appear that excessive pressure of steam was the proximate cause of the accident, and that to have used any other appliances than it did, would result in injuring its product.
Various reasons, based upon the evidence, are assigned by counsel for appellant in support of its contention, and they can all be disposed of by a brief review and analysis of the testimony bearing upon the alleged negligence of appellant in the respect noticed, and the cause of the accident.
The boilers in which the steam used in the cooker was generated were capable of withstanding a pressure of from ninety to 100 pounds to the square inch. At the time of the accident this was about the force exerted by the steam within these boilers. The cooker was capable, as variously stated, of resisting a pressure of from twenty-five to fifty pounds to the square inch, although, .according to the catalogues of the manufacturers, its resistance was thirty pounds. It was sold with a safety valve, and when first installed was equipped with one which was subsequently discarded, because it did not work well. The cuts issued by the manufacturers show that these kettles are also provided with a steam gauge and thermometer. It was operated by appellant with a gauge and thermometer, a combination appliance which indicated the pressure exerted by the steam in the cooker, and the degrees of heat generated, by a scale indicating the heat and its equivalent in pressure. The mercury by which these conditions were shown was enclosed in a receptacle on the inside of the cooker, directly *69connected with the appliance on the outside. The flow of steam from the boilers was regulated by an ordinary globe valve operated by hand. The cooker when filled with cans was closed, and the steam turned on with full force, suddenly, until the temperature reached about 240 degrees, when it was closed off, the exhaust valves on the cooker opened, and the steam allowed to escape until the temperature ran down to about 212 degrees, when these valves were closed and steam again turned on with full force until the temperature reached 240 degrees. The valve regulating the flow of steam from the boilers was then closed down, so as to allow only sufficient to pass through to overcome the exhaust, radiation, and condensation, and thus hold the temperature at about 240. It is claimed that with this degree of heat the steam within the cooker would not exert a pressure greater than fifteen pounds, although there is evidence to the effect that by this method of operating, a direct and full pressure from the boiler might be exerted upon the cooker, and if the globe valve was disturbed when the pressure was equal to fifteen pounds, steam sufficient would pass to the cooker within the space of about one minute to raise the pressure beyond the safety point. The thermometer and gauge would indicate the heat and pressure, but would not respond instantly, for the reason that the mercury would not rise until the heat had been communicated through the receptacle containing it; neither would it relieve any pressure, while a safety valve would act automatically by releasing the steam the moment a pressure was exerted beyond that at which it was set to open, and a pressure regulator on the pipe connecting the cooker with the boiler would prevent the pressure beyond a given point. In some canning factories safety valves upon the cookers are used, in others not, the evidence being that the thermometer and gauge was the appliance most generally employed. There was evidence to the effect that a safety valve would not operate on the cooker of the kind in question, because sugar and other substances would be taken up by the steam and deposited upon its mechanism, which would *70cause it to stick; but there was also evidence to show that if these substances were removed, it would work. It was also stated that such an appliance would injure the product being prepared in the cooker, because it would reduce the temperature by releasing the steam, or if set to release the steam at a point above the desired degree of heat, the product would be injured by overcooking; but there was also evidence that when a high temperature was used, the cooking was accomplished more rapidly. On the afternoon of the accident, the work of operating the cooker was being hurried, as delays had occurred during the forenoon. On behalf of appellant there was evidence that about a month before the accident, the cooker had been tested by hydraulic pressure, and showed a resistance of at least forty pounds, but there is a dispute as to whether or not, as a matter of fact, this test was made. It appears that the employé whose duty it was to turn the steam off and on the cooker, had two others to which he also attended, and in addition had other duties to perform which necessitated his being away some feet from the cooker; that just before the explosion he noticed that the thermometer, according to the degrees of heat, only registered fifteen pounds of pressure. The explosion shattered the door of the cooker into several pieces, and the kettle, weighing over a ton, was thrown backwards by the recoil a distance of ten or twelve feet, striking a pine post 6X8, with sufficient force to break it. Some defects, not noticable from the surface, were observed in the material of the broken door, but no witness pretends to state that these imperfections would reduce the resistance of the door, to less than fifteen pounds. The thermometer was not broken by the force of the explosion. Some of the product in the cooker at the time of the accident was afterwards prepared in another, and when opened, was found to be uninjured and marketable.
On behalf of appellant it is contended that this evidence does not establish negligence upon its part, because it appears that such safety appliances as are sanctioned as safe, or those generally employed by persons engaged in similar *71business, were in use upon the exploded cooker, and that the explosion was due to latent defects, and not overpressure.
When the determination of the question of negligence depends upon inferences, which may be drawn from facts and circumstances of a character that different minds may honestly draw different conclusions therefrom, the question should be left to the jury. Lord v. Pueblo S. & R. Co., 12 Colo. 390; 2 Thompson on Negligence, 1236; Colo. Cent. R. Co. v. Martin, 7 Colo. 592; Sherman & Redfield on Negligence, § 11.
In operating its factory appellant, through those whose duty it was to manage and superintend it, was required to exercise that care and prudence in supplying the cooker with such appliances to avoid injury to its employés, that ordinarily prudent persons, entrusted with like duties, with due regard to the dangers to be guarded against, would have employed for that purpose. N. Y. & Colo. M. S. Co. v. Rogers, 11 Colo. 6; Cooley on Torts, (2d ed.), 657 ; Wash. & G. R. Co. v. McDade, 135 U. S. 554; 2 Thompson on Negligence, 982.
Its failure to do so would be negligence. The jury found that it had not exercised the requisite care with respect to equipping its cooker; so the first question presented, is, does the evidence fairly warrant the inference that appellant was negligent in this particular ?
In some establishments safety valves were used; in others only the thermometer and gauge, used by appellant. The use of impounded steam is not confined to canning factories. Which of these appliances was the one which the exercise of reasonable care and prudence would adopt ? Steam has ever been recognized as a dangerous agency, the use of which results in accidents even where the most approved safety appliances are utilized. The cooker was connected directly with the boilers, and was only capable of withstanding a much less degree of pressure than the steam generated in these receptacles usually exerted. It was designed to be •used with a safety valve. The thermometer and gauge on *72the cooker would not instantly indicate the pressure of the steam within it. The steam was only controlled by a globe valve on the pipe through which it was conducted to the cooker. Before any warning of danger pressure would be indicated by the thermometer and gauge, an explosion might occur. So that from this testimony, viewed in the light of the difference between the resisting capacity of the cooker and the boiler, the manner of using the steam, the inefficiency of the thermometer and gauge to instantly indicate a dangerous pressure, and the absence of any appliance to relieve such pressure, if there was room for an honest difference of opinion between intelligent and upright men as to whether or not appellant had exercised that eare, caution, and prudence in equipping its machinery for the purpose of preventing injury to its employés that it should, it at least is sufficient from which to deduce the inference that it had not and the jury was clearly justified in so concluding.
The contention of appellant that the use of a safety valve on the cooker would injure the product is without foundation; on the contrary, such an appliance would prevent its being subject to an excessive heat by allowing the steam to escape before that degree was reached. To release the steam at a pressure slightly above that which was required to do the best work would not prevent the heat from reaching the required degree, but would automatically relieve the product from overheating by releasing the agency through which this heat was communicated. That it would not work successfully upon the cooker is, also, without foundation, because the evidence establishes that it would require more care to keep it in order, but with the exercise of that care, it would perform the function for which it was intended.
The next question relates to the sufficiency of the evidence to establish that the explosion was caused by overpressure. At the time of the accident, the pressure in the boiler was at least ninety pounds. The globe valve was not entirely closed. The work of operating the cooker was being hurried. According to the claim of appellant, it had been tested about *73one month before, by subjecting it to a hydraulic pressure nearly three times as great as the steam exerted when it was in use. The violence of the explosion indicated that it was caused by a force of great energy. Latent defects in the door did not appear to exist to a sufficient extent to cause it to burst under ordinary pressure. This state of facts is ample to support the conclusion of the jury, that the explosion was caused by an excess of pressure within the coolcer. In this connection it is urged that because the thermometer was not broken, which would have been the result had the mercury been subjected to an unusual degree of heat, and because some of the product in the cooker at the time of the accident was not injured by overheating, that these facts, establish that the explosion was not caused by an extraordinary pressure. These are circumstances proper to consider in endeavoring to ascertain the proximate cause of the accident, but are not conclusive. A sudden rush of steam into the cooker could readily cause it to explode before the heat it contained would be communicated to either the bulb of the thermometer, or the contents of the cans.
The errors assigned upon the giving and refusal of instructions will now be considered. Appellant requested the court to instruct the jury that if the accident was caused by the negligence of a coemployé, appellee could not recover unless the duty of such coemployé was to superintend the works, or keep the machinery in proper condition. Our attention is not directed to any evidence tending to prove that the coemployé named in this request was negligent in the discharge of the duties imposed upon him by appellant; on the contrary, its counsel contend there is not. An instruction should not be predicated on the existence of a fact which there is no evidence to establish, and it was, therefore, not error for the court to deny this request. Gibbs v. Wall, 10 Colo. 153; Wood v. Chapman, 24 Colo. 134; Robinson v. D. & R. G. R. Co., 24 Colo. 98; Heartt v. Rhodes, 66 Ill. 351; American et al. v. Rimport, 75 Ill. 228.
*74For the same reason, the eleventh instruction requested by appellant was properly refused.
The purport of instruction six given by the court, was that if the duties imposed upon the employé intrusted with the operation of the valve by which the steam was turned on or off the cooker were such that, in the performance of such duties, he was compelled to neglect the valve in question, and that this was, or by the exercise of ordinary skill and diligence, could have been known by appellant, it was guilty of negligence. The objection urged against this instruction is, that no such issue was presented by the pleadings, the claim of counsel for appellant being, that the negligence charged consisted in not providing appliances to prevent the explosion of the cooker. It is unnecessary to determine whether the charge of negligence stated in the complaint is limited, as claimed, or not. Appellant by its answer seems to have treated the question submitted by this instruction as an issue in the case, for it averred, in substance, that in conducting its business it had furnished all necessary apparatus for regulating the passage of steam from the boiler to the cookers, and kept a competent and experienced man to operate the apparatus so furnished, who, at the time of the explosion, was so engaged. It also appears that the evidence regarding the duties of this employé was admitted without objection on the part of appellant, and. that it introduced evidence on this subject. This evidence discloses that in the performance of these duties, this employé was frequently compelled to be away from the immediate vicinity of the globe valve which it was his duty to operate. On the submission of the case to the jury, counsel for appellant requested the court to give an instruction predicated upon his negligence, which, as we have already noticed, was refused. In the motion for a new trial no question appears to have been made regarding the proposition that the instruction now being considered submitted an issue not properly made by the pleadings. In the briefs filed by counsel for appellant in this court, no such proposition is suggested. It was raised for *75the first time on oral argument, so that it seems from the conduct of the case, that in'the court below counsel for both parties treated the question submitted by this instruction as being within the issues, and for that reason it cannot now be raised on appeal.
In its motion for a new trial, appellant sought to establish that the amount of the verdict returned by the jury was the result of a resort to chance. In support of this ground of the motion, the affidavits of two of the jurors were obtained, to the effect that it was agreed by the jurors that each should mark down the sum which, in his opinion, was the measure of damages, and that the quotient resulting from dividing the aggregate of such several sums by six, should be the verdict ; that this agreement was carried out, and a verdict returned accordingly. Subsequently, the affidavits of all the jurors were obtained, the purport of which was, that they all agreed that the plaintiff was entitled to recover, but experienced great difficulty in reaching an agreement as to the amount; that it was then proposed to take an average of the amount of damages to which each thought plaintiff was entitled, which was done; that there was no agreement in advance that the members of the jury should be bound by the result; that after this average was obtained in the manner indicated, it was discussed by the jurors, and conceded by all to be a fair and proper sum; that another ballot was then taken upon the amount of damages to be awarded, and each voted for the average, which was then agreed upon and returned into court. One of the jurors who made the affidavit filed by appellant states in the one subsequently filed, the substance of which is above given, that he did so for the purpose of explaining those statements in his first which appear to be inconsistent with his last. The other juror in his second affidavit also stated that he did not wish to be understood by his first, that the jurors had agreed in advance to return a verdict for the average amount of damages ascertained in the manner indicated, or that they did not after-wards deliberate and agree upon the amount. He also states *76that at the time he made his first affidavit, he informed counsel for appellant that after the average was ascertained, another ballot was taken. From these statements it appears there was no agreement in advance that the jurors should be bound to return a verdict for the amount indicated by an average of the respective sums which, in the individual judgment of each, was the sum to which plaintiff was entitled, but that after ascertaining this average it was, upon deliberation and discussion relative to the amount of damages, agreed upon, and so returned.
The fact that the jurors ascertained their average judgment as to the amount of damages to which plaintiff was entitled by constituting the sum total of the opinion of each on this subject as a dividend, their own number as a divisor, and the quotient as the average which was subsequently agreed upon as their verdict does not vitiate it in the absence of a preliminary agreement to be bound by the result thus obtained, especially where it appears that such average was subsequently discussed and finally agreed upon as the amount of damages which they would return in favor.of plaintiff. Ill. Cen. v. Able, 59 Ill. 131; City of Pekin v. Winkle, 77 Ill. 56; Village of Ponca v. Crawford, 23 Neb. 662; 8 A. S. R. 144; Barton v. Holmes, 16 Iowa, 252; Dana v. Tucker, 4 Johns. 487; Harvey v. Jones, 3 Humph. 157; Dunn v. Hall, 8 Blackf. 32; Wilson v. Berryman, 5 Cal. 44; 63 Am. Dec. 78; Schoolfield v. Brunton, 20 Colo. 139.
Other errors assigned are argued, but in our opinion they are either not tenable, or not of sufficient moment to warrant a reversal of the case, and for this reason we do not notice them in detail. The judgment of the district court is affirmed.
Affirmed.