delivered the opinion of the court.
The appliance or machine by the working or motion of which the plaintiff received the injury that is the occasion of this suit, is a common and necessary part of laundry machinery and is known as a “wringer” as well as an “extractor,” because it does the work of the older device of a wringing machine in drying or at least removing water from the washed clothes. It is a machine the outer part of which resembles a large bowl made of iron. It is attached to the floor by iron supports. This bowl above the supports is about 24 or 25 inches in depth and from 28 to 30 inches in horizontal diameter at the centre. It is contracted at the top and bottom. The top is about three and one-half feet above the floor, and at the top the bowl is about *36120 inches in diameter. Within this outer bowl there is an inner receptacle made of copper, which is perforated. This inner receptacle is called “the basket,” while the outer bowl is called “the shell.” The basket is balanced and runs on a pivot so geared that the basket revolves on the application of power. The method of using the machine is to fill the basket with the wet, freshly washed clothes, and then by the application of the power to set it revolving within the shell. The basket at full speed makes 1400 revolutions a minute, and the centrifugal force thereby generated throws the water out of the wet clothes in the basket through the perforations into the space between the basket and the shell, whence it flows away through a drain at the bottom. The basket comes up flush with the top of the shell or the fraction of an inch above it. The basket has at its top a rim turned over and making a smooth curved surface. The top of the shell is curved in towards the rim of the basket. There is only a small space at any point between the basket and the shell— about half an inch.
Steam power was used to run the wringer in question here. The power was supplied from a main shaft. A power belt ran from a pulley on the main shaft to a pair of pulleys on a jack shaft over the machine—one a tight or friction pulley and one a loose pulley. The power belt could be shifted from one to the other, and when running over the tight pulley it started a down belt running from that pulley to one under the bowl. This was so connected and geared as to communicate the power to the basket and cause it to revolve. It is immaterial how the transmission of power was effected at this point. That which is material about it is that the machine was started and stopped by shifting the power belt from the loose to the tight pulley at the jack shaft and vice versa. This shifting of the power belt was done by the hand operation of a lever or “shifter” which was near by and over the wringer, *362.the handle being in easy reach of the hand of one operating the wringer, at about the height of his shoulder.
The persons who may be said, through the circumstances involved and their respective personalities and duties to be directly connected with the accident which brought about this suit, are the defendants, Peter Deahl and Sam Deahl, the plaintiff, Josef Skulimowski, Joseph Stanislawski, the foreman for the Deahls, Mike Trundt, an employee of the Deahls, who, with the exception of the plaintiff, was the only eye witness of the accident, and Theodore Deahl and John Tost, also employees of the defendants, who were in the immediate vicinity of the machine when the accident occurred. There were others who testified at the trial, but they were either, with the exception of the attending physician of the plaintiff, called to testify as to the previous careless conduct of the plaintiff at another laundry, or as experts in the mechanism and operation of the machine or in their examination of it subsequent to the accident.
The defendants, Peter Deahl and Sam Deahl, doing business under the name of ‘ The Derby Steam Laundry,” ran a laundry establishment on East Ontario Street in Chicago. The laundry occupied a floor space of about 120 feet by 40 feet. It contained 13 washing machines, 4 mangles, 2 tumblers and 8 wringers or 'extractors of the kind hereinbefore described. At the time of the accident they had about seventy-five persons employed in the laundry. The manager of it was Sam Deahl, one of the proprietors and defendants.
The plaintiff, Josef Skulimowski, was one of the two men employed to run the wringers; Mike Trundt was the other. The plaintiff, Skulimowski, had been a wringer man” for the defendants for about a year before the accident. How far he was to be called " an experienced wringer man, ’ ’ counsel in their argument seem to differ about. But as to the facts concerning *363Ms experience they practically agree. The defendants say, “he entered their service as a wringer man and represented himself to them as an experienced wringer man, and had worked as a wringer man before he became employed by them, and worked as a wringer man for a short time in another laundry—the Saratoga Laundry—during an interval while the defendants’ laundry was shut down. His sole duty while working for the defendants was operating the extractors and he was experienced in the work and its duties. ’ ’
Plaintiff says:
“Plaintiff was not an experienced man when he was at first employed by defendants. He had done similar work for a short period at other laundries. Nearly all of his knowledge and experience in running extractors was gained while in the employ of defendants—a period of seven or eight months and under the instruction and supervision of defendants. In running extractors it was no part of his work to adjust, care for or attend to the belts in any way. Defendants had a foreman whose duty it was, and who did in fact at all times undertake, to look after the belts.”
At the noon hour of the day on which plaintiff’s arm was torn off in the basket of the wringer, a new belt made of canvas with a “belt-dressing,” tarry or sticky in its nature, was substituted for the leather belt previously used on the “down drive” from the jack or counter shaft to the pulleys under the wringer. The new belt was brought to the laundry by an employee of the firm furnishing it, and was laced to the shaft by the men who brought it. After the noon hour the foreman, Stanislawski, put this new belt on the pulleys with which it had to be geared. While doing tMs work the foreman, the plaintiff says, called him over to the wringer “to turn the basket,” to help get the belt on the pulleys, and afterward, before taking away a step ladder on which he had been standing, said “Joe, it is right,—start him in, but rinse it out.”
*364The foreman’s version is: “Skulimowski was present when the belt was put on and saw it put on. After the belt was put on I told him the extractor was all-right. I did not tell him to start the machine and rinse out the extractor. After I put on the belt I tested the machine. I started it up and ran it at full speed and stopped it; it ran all right. ’ ’
The foreman was then taking the step ladder away and had gone about'25 feet, when he heard a scream, looked and ran back, and found the plaintiff held by Trundt and with his arm torn.
Trundt, the plaintiff says, was “right behind him” and caught him or lifted him off the wringer as he was falling, saying “Now you have got it!” Trundt therefore must have seen the accident. As we have said, he' was the only one besides the plaintiff himself who did.
It is very unfortunate, therefore, that the testimony of this witness which, considering the diverse theories of the accident hereinafter mentioned, would probably have been important, could not be obtained. The only explanation in the evidence as to the failure of either side to produce him is found in the testimony of one of the witnesses for the plaintiff, a fellow workman at the time of the accident, who said that he “guessed Trundt was now over in Europe. ”
Tost and Theodore Deahl were in the room and had seen the plaintiff at the machine just before the accident. They did not see the accident itself, but had their attention called to the occurrence first by the screams of the injured man. Sam Deahl was in the office below, and hearing the screams ran up stairs and assisted others in taking care of the plaintiff and removing him to a hospital.
The testimony of the plaintiff, who evidently has only imperfectly mastered the English language, concerning the material matters leading up to the accident and the accident itself, was, in substance, as follows :
*365“I am 33 years old. I was hurt May 10, 1907, in the Derby Steam Laundry. I had worked there over seven months. Sam'Deahl hired me, took me to the foreman Stanislawski, who put me to work and told me what I had to do. Stanislawski gave us orders and himself repaired machines, laced belts and put on belts. Sam Deahl told me to do what Stanislawski told me to do. Shortly after I went to work in the laundry .1 was starting a machine (a wringer) and ‘grabbed the lever and pulled it slow over/ and Sam Deahl said ‘I don’t want you to start that way. If you start that way it burns the belts. * * * If you don’t take hold and pull with your hand on the rim, why the belt throws off.’ * * * lie takes my place and shows me how to start it, and I says ‘I will.’ He takes on the rim with his hand and gave it a push and same time moved the belt from loose pulley onto tight pulley and they start, and when it goes a few turns, at the same time-after it go full speed. * * * Stanislawski told me the same thing. I got to use my hand on the rim, at the same time throw belt, because that don’t pull it off from the pulley. After the instruction had been given I put the machine in motion in that way—‘thousand times—-thousand times a day. ’ Before I got hurt all the belts used were leather belts. I saw a dressing like grease on belts, but not the extractors, only the washing machines. The leather belt was exactly tight and loose—just was right way, it start right away to go. It started always, I started, could start slowly, but it go good speed after, when it got to go good speed, that is, not sudden with a jerk, but a little after it went on full speed. I never saw a canvas belt in my life. I did not know before I got hurt the effect, if any, that the tightness of the belt would have on the starting of the machine nor the effect the dressing would have. Just before I got hurt I saw the foreman Stanislawski. He called me, wanted to get me to turn the basket at the time he put the belt on the pulley. I was turning the basket and he got the belt on the pulley. The foreman at that time was on the step ladder. I went to him. He told me, ‘Joe, it is right, start him in but rinse it out. Start the ma*366chine and rinse her out with the water. ’ He took the step ladder and walked away. I was taking them with my right hand, the basket, the shell, and I give them a turn, at the same time I take belt over slowly because the belt should take that, the lever, and move the belt with left hand and started the machine, and after that why it started and threw me in—it caught my right hand. Just started the machine and that machine it jerked like a shot and just caught and pulled me right around and twisted my hand over and I lost my balance and fell right under. I could not get out. I fell in * * and was hollering and somebody working there was taking me by the shoulder and got me out from the machine. Then they take me to the hospital. I moved the lever slow. We got to get him started with the hand, but when it goes it wants to take my arm, and it twists right off, twists right over; it first take me and then throw it over and it broke the flesh off and it flop right down inside the basket. I could not tell how far the belt got on one pulley from the other. The basket was empty.”
On cross-examination the plaintiff said:
“I was instructed at the Goodrich Laundry to run the extractor same as by Derby Steam Laundry. Got to take with hands the rim and give a start, at same time move belt over from loose pulley on tight. * * *
Q. Now at the time of this accident did you put it on slowly or quick? A. I took it slowly.
Q. So that when you were hurt you had your hand on the rim of the basket ? A. Yes.
Q. That was your right hand on the rim of the basket? A. Yes.
Q. Did you take hold of the rim? A. I did not hold the rim I guess, only put my hand and gave him a shove.
Q. Just with your fingers? A. Yes, with hand to give it a shove over. * * *
Q. You just had hold of it with the fingers—you hooked your fingers on the rim? A. Oh, there is nothing to hook, only that I hold that way (indicating), and it goes like a shot and just caught and took me *367•that way (indicating) so quick. It just shot so quick. Every Monday I started an empty wringer; would take a pail of water and throw it in and start the wringer for the purpose of rinsing it. I had not put any water in (this time). I just was starting it and I got to go and get this water.”
We have given the plaintiff’s own story at length and largely in his own words, in order to present a fair summary of the evidence on which, supplemented by testimony hereinafter mentioned, concerning an unusual tightness of the belt on the pulleys and expert testimony on the difference between the action of a new canvas belt with belt dressing on it, tightly drawn over the pulleys, and the action of the leather belt theretofore used,—the plaintiff’s counsel relies for his theory of the case and of the liability of the defendants.
That theory is that the substitution of a new belt for one that had been used, a canvas belt for a leather belt, a canvas belt, moreover, with belt dressing on it, and the unusual tightness of this belt on the pulleys, the existence of which the plaintiff argues there is both direct and indirect evidence to support, created a new and more dangerous operating condition; that defendants knew or should have known of this additional risk to the plaintiff, but that plaintiff did not know or appreciate this additional risk and, under the circumstances, could not be expected to know it; that defendants were guilty of negligence in failing to warn plaintiff of the additional risk and in assuring him that it was all right, and without such warning directing him to use the machine; that the plaintiff was guilty of no negligence contributing to the accident, but proceeded to start the machine in the customary and proper way and in the way that he had been instructed to do; that this way was to place his right hand on the rim of the basket to give it a shove, and his left hand on the belt shifter lever to slowly shift the belt; that because the *368belt was a new canvas belt, tight on the pulleys, with sticky dressing on it, the movement of the lever, although slowly made, caused the belt to seize with unusual grip the pulleys, and to start the revolving-basket of the wringer with a jerk at great speed, pulling the hand of the plaintiff around with it, throwing him forward off his balance, with his forearm in the mouth of the basket and his elbow on the rim on one side in such a way that the force of the revolving- basket tore the arm off.
The defendants ’ theory of the accident is very different. It is that the plaintiff, representing himself as an experienced wringer man and actually being so experienced, sought and obtained .employment with the defendants; that thereby he assumed the ordinary risks of the employment; that as a matter of fact it was necessary and customary to start a full, but not an empty, wringer by giving the basket a shove when turning on the power; that in such an action there was no necessary danger; that although a new belt made of canvas and covered with dressing had been connected with the wringer just before the accident, this had no necessary connection with the accident; that the appliances after the substitution were in good order and condition and started the machinery properly and normally; that the plaintiff knew all the new conditions introduced, if any were introduced, by the change of belts, which he saw made; that as an experienced wringer man he had all the knowledge and means of knowledge of them that the defendants had; that, moreover, the evidence showed that whatever was done in substituting- the new belt for the old one, and ordering- the machine started, was done by one who was a fellow servant of the plaintiff within the meaning of the "fellow servant ” or “ common employment” rule, so-called; that the plaintiff was careless and had been habitually careless before, as the evidence showed, in the management of wringers; that in the present case he started the extractor *369and then attempted while it was in motion to wipe it out with a rag or cloth by holding the rag against the inner surface of the basket as it revolved, and in so doing lowered his elbow below the rim so that his hand and forearm were entirely within the basket; that while in that position his elbow was struck or came in contact with the inner surface of the revolving basket, which jammed his hand with the rag in it against the opposite side, and that thus the arm was torn off.
Between these differing theories so far as the facts are concerned the jury found for the plaintiff and the trial Judge refused to interfere with the verdict.
We are obliged to confess our inability to decide on them with any more assurance of certainty. The case is a perplexing one. The defendants insist that the laws of physics “relating to inertia, elasticity, velocity of revolving bodies, momentum and centrifugal force,” as shown by computations in their argument, negative the possibility of the accident occurring by the machine starting as the plaintiff says it did, and producing the effect on his balance and on the position of his hand and arm that he says^t did. Further they insist that the nature of the injuries and of the wound as shown by the testimony of the attending physician, also preclude the possibility of the accident having happened in that way, and that the dimensions and character of the basket and the position of the plaintiff require the same conclusion, and that the great preponderance of the expert evidence (as well that introduced by the plaintiff as that presented by defendants) shows the impossibility of the machine starting with the speed and the result on the plaintiff’s position that he alleges and that his theory of the accident requires. Further, they call attention to the testimony of Joe Stanislawski, the defendants’ foreman, that after putting on the belt he himself started and tested the machine, and afterward merely told plaintiff that the wringer was ‘ ‘ all right, ’ ’ but gave him no specific orders, and that *370about two hours after the accident on returning from the hospital where he had been with the plaintiff, and examining the basket of the wringer, he found within it a rag or cloth—part of a wringer cover—with blood on it, which was not in it when he, Stanislawski, “started the machine before the accident.” This testimony is corroborated as to the rag in the basket two hours after the accident by a manufacturer of laundry machinery, Nelson, called in by the defendants to look at and test the machine, and by a brother of the defendants (Theodore Deahl), who says he saw the rag, a piece of canvas cover, in the basket after the accident (without indicating how long after), and by U. S. Deahl, one of the defendants, who says he saw it there two or three minutes after the accident, and also by a witness for the plaintiff—Yost—who likewise saw it immediately after the accident. The defendants argue that this testimony shows such a probability of the truth of the theory of the accident which we have before set forth as theirs, as to amount to proof of it, and that this proof is strengthened by the showing made by other witnesses that the plaintiff was careless in a laundry where he was formerly employed, putting his hands in the basket while it was moving, and by one of these witnesses that he was guilty of doing there what the defendants insist that he did in this case—holding a rag against the inside of a revolving basket to clean or wipe it out.
So far as the scientific computations of physical forces go, we cannot give so great importance to the calculations worked out by counsel for defendants as they do. Accidents of this sort happen so instantaneously and unexpectedly, and slight changes in position and circumstances of the machinery and the unfortunate sufferer, so completely alter any given situation and introduce new factors into the problem, that we must consider such arguments at the best as but weak substitutes for actual experience and the testimony of *371eyesight. The laws of physics are invariable, but their application is subject to an infinity of variation.
It is very unfortunate that there was among the witnesses no eyewitness of the actual accident except the plaintiff himself; but it must be conceded that the testimony of Tost and of the defendants’ witnesses before named gives much color to the defendants’ theory and produces in our minds doubt and uncertainty as to how this accident happened.
But, on the other hand, it is true, as counsel for the plaintiff point out, that the plaintiff himself explicitly denied that he used the cloth in question at the time of the accident or had ever used one in that way in the laundry in which he had before been employed; that the cloth was not of a character most* likely to be used for the purpose supposed; that it might easily have been soiled with the blood of the plaintiff and dropped inadvertently into the basket immediately after the accident in the excitement and confusion of the time, through circumstances and for reasons unknown, unnoticed and unremembered; that the testimony of the witnesses as to the careless conduct of the plaintiff in another laundry about a year before so barely comes, if at all, within the rule which makes competent and material a “usual habit” or “general course of conduct” “on a particular matter complained of” “as bearing on the question of ordinary care,” as to deprive it of any serious import in this investigation; and finally and chiefly that the questions raised by the contradictions in the direct testimony and the inferences to be drawn from circumstantial evidence, were peculiarly matters for the jury, and if left to them under proper instructions it is not a case in which a court of review ought ordinarily to interfere. It could only be justified in doing so on the ground that it could and ought to overrule the jury in the credibility that they found in the direct testimony of the plaintiff. We think, after careful consideration, that as it inheres in *372the verdict of the jury that they believed that testimony, we should not be justified in setting the verdict aside on the ground that it is against the manifest preponderance of the evidence as to the method of the occurrence of the accident, notwithstanding our personal feeling of uncertainty.
The defendants, however, say that even if the theory of the plaintiff as to the accident is received, namely; that it happened by the plaintiff’s losing his balance and falling forward with his arm in the basket just as he started the machine in the usual way, with an almost simultaneous shifting of the belt by a lever with his left hand and a twirl of the basket with his right hand, there is no negligence to be imputed to the defendants in connection therewith. It was, they argue, either entirely the result of the negligence of the plaintiff in keeping his hand on the rim of the basket after he had shifted the belt—a thing they contend he was not instructed to do—or at the very utmost was an accident for which the negligence of no one was responsible.'
The plaintiff’s evidence certainly tended to establish the fact that it was customary for the operator of a wringer to put one hand on the rim of the basket and shift the lever with the other, twirling the basket contemporaneously and keeping his hand on the rim for a very brief time—say five -or six seconds—either following it round for a revolution or two, before too great a speed was attained to make it impossible, or to take a new hold after a half revolution and repeat the twirl' or jerk three or four times, which last manner of handling was that which the plaintiff said he ordinarily used. “But this time” he says, referring to the accident, “I just put my hand and went to give it the first jerk and it goes like a shot with me.” The plaintiff also maintains that his usual method of starting the machine was in accordance with instructions that he had received; but we think his testimony, somewhat, as might be expected, imprecise in expression, goes no *373farther at the best as to instructions than to tend to show that he was told to twirl the basket contemporaneously with pulling the lever over. He does not say that he was told to give more than one twirl in starting the basket, nor that he was told to allow his hand to be carried, around any part of the circle with the revolution. The evidence, however, does tend to establish that whatever method of starting the wringer the plaintiff customarily used after his instruction, the defendants must sometimes personally and through their foreman have observed, without ordering or advising another method.
The evidence for the defendants very strongly tends, however, to establish the contrary propositions to all this—first, that it is' not customary for operators of the wringers to accelerate or steady the rotation of the basket by successive twirls or jerks or keeping the hand on the rim during one or more revolutions, but only to start the basket rotating by one twirl and then, after removing the hand from the rim, slowly to move the lever which shifts the belt; secondly, that no instructions as to starting the wringer were given to the plaintiff, he being supposed to be an experienced man not needing them; that certainly no instructions were given to hold or replace the hand on the rim after the belt was shifted, or to attempt to twirl the basket except before such shifting of the belt, and then only when it was loaded; and, thirdly, that the defendants did not observe the plaintiff with his hand on or in the wringer after it was started more than once, and that he was then warned that his action was dangerous.
We cannot detail the evidence more particularly, but we have considered it all thoroughly and must, as to these contradictory propositions, also hold that they presented questions for the jury, to whose judgment, rather than to ours, they are committed. We must therefore assume in the further discussion of the matter, that the plaintiff was instructed to start the ma*374chine by contemporaneously twirling the basket by hand and shifting the lever; that it was customary and usual for others to give several twirls in succession and sometimes to allow the hand to be carried at least once around, contemporaneously with slowly shifting the lever to move the belt; that the plaintiff himself usually started the machine, even when empty, by twirling the basket with more than one successive “jerk” or twirl, placing his hand on the rim separately for each twirl, and had not theretofore suffered thereby; but that on this occasion, for some reason, although, so far as his uncontradicted testimony shows, he “put the belt over slowly, ’ ’ the bowl started its rotation with such speed that it threw him from his balance and carried his arm to destruction.
Even so, however, although this conclusion, unless the plaintiff knew of an increased danger by the change of belts, relieves the cause of one factor in the discussion-—that is, the alleged contributory negligence of the plaintiff,-—it leaves still open the question of the negligence of the defendants. Assuming that the plaintiff was attempting the customary and usual method of starting the machine, directed by the defendants, and practiced by him, as by others, for months with their knowledge and assent, which had proved during that time harmless, we could not hold from the evidence (whatever the pleadings) in the absence of the change of belts hereinafter considered and the alleged specific order without warning following that change, that the plaintiff had proven any negligence on the part of the defendants. No such negligence could be deduced from a failure to warn plaintiff of the danger of so starting the machine before the belt was changed, for whatever ground there may be for holding the plaintiff inexperienced in the use of machinery or ignorant of the differing action on the machine of canvas and leather belts, of new and old belts, or of dressed or undressed belts, there can be no doubt that his work in laundries *375had rendered him as capable as the defendants of observing and judging of the usual speed with which the belt that had been in use started and drove the wringer when the lever was shifted slowly and gradually. To that extent there could be no question of the ‘ ‘ assumption of risk” and no necessity for warning.
We are brought, therefore, to the actions on which the plaintiff’s case is based—the substitution, as it is asserted, of a new canvas dressed belt put tightly on the pulleys, for an old leather belt not so tight on the pulleys and without»“particular” dressing or “without much if any” dressing, as counsel in examining expert witnesses variously phrased it.
It was, we think, a question for the jury whether all these differences in the belt put on just before the accident and the one theretofore in use on the machine existed. That it was an entirely new belt substituted for a leather one that had been in use several months, and that it was of canvas is admitted; but its being tighter on the pulleys, or its being more thickly covered- with dressing and with a different kind of dressing, is nowhere made certain to us. It is conceded, however, that there was dressing on the new belt, put on at the factory, though the nature of it does not clearly appear. There was testimony by the foreman that the dressing made “a black coating,” and the witness Tost testified that it was a “tar coat,” that it “smelled of tar,” that it “was a tar paint.” U. S. Deahl testified that it was put on at the factory and he thought it was some kind of “an oil dressing.” In view of the importance given in the expert testimony at the trial to the effect of belt dressing and the different kinds used, it is unfortunate that a more precise description of its character was not forthcoming. Nor is there much definiteness in the evidence as to how much or what kind of dressing was usually on the old belt. Tost, in answer to the question “How much dressing or that tar coating was on the leather” (the old) “belt?” said that “when the belt *376got loose and the foreman had no time to cut it and make it shorter, he put a little belt dressing on it.” The plaintiff himself being asked to describe what kind of dressing there was on the old belt said he ‘ ‘ did not know,” that he “saw them sometimes put a little like grease, something looks to me like it, by the foreman, but not on the extractors, only on the washing machine”—obviously a, very blind statement. On cross-examination he was asked, “Did they ever use any dressing on those belts?” (i. e. the old belts). The answer was “I never saw.” Question: “Never saw them use dressing?” Answer: “No, I never saw that. I did not only on the washing machine, dressing. I never hot at all saw them put dressing on the extractor belt.”
The foreman, Stanislawski, was asked whether they used any belt dressing on the old belts, and said that they did ‘ sometimes, whenever a belt would get dry. ’ ’
Mr. Pullen, a witness who was in the machinery belt business and evidently well qualified as an expert, testified to the very reasonable proposition that other conditions being equal the question whether belt dressing would cause a belt to set a machine revolving rapidly more quickly and suddenly, depended on the kind of dressing, of which there were many sorts, and that an oil dressing would cause a belt to slip, thereby losing power, while a resin dressing would cause it to grip the pulleys tighter and thus start the machinery more sharply. It will be seen from this how advantageous in the consideration of this cause would be more satisfactory evidence of the dressing, if any, which was actually used on the old belts running the wringer during the plaintiff’s experience with them, and of the dressing which covered the new belt put on just before the accident.
But after full consideration we have arrived at the conclusion that it is implied in the verdict of the jury that the differences alleged by the plaintiff to exist be*377tween the old and new belt did exist; that the new belt, that is, was so dressed as to take a stronger and quicker grip on the pulleys than the old one, and that being newly placed thereon, to take the place of an old belt, was also tighter on them, and that the evidence on thesé points was sufficient to make it right to leave them to the jury to decide. The question was also mooted before them whether the difference in material in itself was not sufficient to make a difference in the speed with which the machine would start, and a good deal of expert testimony was taken on this point. We think that whatever may be said of the weight of testimony on this question, we must hold that on the question of fact whether the change of belts produced a condition of greater danger in the use of the machinery or appliances with which the plaintiff was concerned, the decision implicit in the verdict we ought not to disturb.
The weight of expert testimony was to the effect that assuming that the belt was new instead of old, tight instead of looser (this might be inferred from the fact that it was new and newly adjusted, and was somewhat weakly sworn to by Tost), with more dressing on it than the old one had, it would start off the machinery “more quickly and sharply.”
The jury evidently believed that the differences mentioned existed and that a condition of greater danger was thereby produced, and we cannot say that their décision was against the clear and manifest weight of the evidence.
The duty of the master under these conditions is a matter of law. Plaintiff’s counsel thus states it fairly:
“It is in general the primary duty of the master to warn the servant of any new danger connected with his work arising out of a change of conditions or surroundings of which the master has or should have knowledge, and of which the servant is not aware and which are not discoverable by him in the exercise of ordinary care.”
*378This rule of law is plain and familiar, and there is no dearth of authority to support it. It may be found thus expressed, so far as the question of changed machinery is concerned, in Pullman Palace Car Co. v. Laack, 143 Ill. 242, p. 254: “if * * * the change in appliances was not known to servants, or so open and visible that they by the exercise of ordinary care would see and know of it, the legal dnty rested npon the master to notify them of the increased danger to which they are thereby exposed.”
The difficulty is not with the rule of law, but with the propriety of the application of it to this case.
Was it a question fairly to be left to the jury whether or not, under the conceded facts of this case, the increased danger which their verdict shows they found existed, was discoverable by the plaintiff in the exercise of ordinary care ? Stated differently, the question is: Considering that which was shown by the uncontradicted evidence and that which the other evidence for the plaintiff tends to prove, was it for the jury or the court to say whether the plaintiff assumed the risk of the danger of operating the wringer, increased as it was by the change of belting?
If it was for the jury, then there is the further question for us, whether the decision on this matter inhering in the general verdict of the jury is clearly and manifestly against the weight of the evidence.
We have several times stated a rule which we think is established by the decisions of the Supreme Court; for example, we so stated it in Grace & Hyde Co. v. Sanborn, 124 Ill. App. 472:
“If conceding to be true all that the evidence tends to prove in favor of the plaintiff, it would be apparent to all reasonable minds that such dangers as the evidence thus tends to prove were incident to and connected with his employment, that they were not concealed or latent, but patent and obvious, that they were not extraordinary and unusual, but usual and incident *379to the business engaged in as conducted by the defendant, and existed continuously during the plaintiff’s employment, and if there is no evidence tending to show that plaintiff ever complained of them, or that the defendant had promised to remedy them, or that the plaintiff had been ordered to incur them by some particular and especial order at the time of the accident, but on the contrary it is proven that he was then engaged in the regular line of his duties and usual employment, and that he was of mature age and of ordinary strength and intelligence, then the question of whether he understood and appreciated the danger (such understanding and appreciation being recognised by the Supreme Court as essential elements in the assumption of risk) is answered by a conclusive presumption and the assumption of risk by the plaintiff becomes a thing for the court to assert as a matter of law—not a question of fact to be left to the jury. ’ ’
We have italicized certain clauses of this quotation merely to call attention to the matters wherein the plaintiff herein especially maintains that his case does not fall within the conditions which could make his ‘ ‘ assumption of risk” a matter of law for the court to assert.
It is of course evident that if it was merely the increased danger caused by the new belt which was responsible for the accident, it could not be said that the risk “existed continuously during the plaintiff’s employment.”
But the rule as we have stated it after all states the extreme limit of the conditions which, when they are all met, remove any doubt of the propriety of a judicial determination of the question involved, and there may well be a case in which all the conditions mentioned •are not fully met, in which there is doubt as to whether or not the question of “assumed risk” is not foreclosed. Such a case we think is the present, and it has caused us much hesitation. The present is not much like most of the cases cited by plaintiff, where the *380changes of conditions, which the various courts held required warning’ or at least justified the submission of the necessity of such a warning to a jury, were serious and of a character very different from that of substituting a new belt, tight and dressed, for an old one looser and not dressed, in plain sight of a man whose business it had been for a long time to run the machinery moved by the belt. It is maintained, and, it may be assumed, was believed by the jury that the foreman, after changing the belt just before the accident, and as he was moving away, told the plaintiff “to start the machine and rinse it out;” but this order after all related to the action to be performed, which was directly in the line of the plaintiff’s duties, for which he had been hired and in which there could be no doubt that he was experienced. The method of performance, the increased care in handling necessary because of increased speed and suddenness of motion resulting from the new belt, was not directed or forbidden. The machine might certainly have been started safely had the plaintiff not followed his usual method; and the question therefore still remains'—was the increased danger not discoverable by the plaintiff by the exercise of ordinary care ? The defendants urge forcibly that the answer must be in the affirmative, since the plaintiff was an experienced “wringer man,” had operated the machine for months, had known that belts were dressed, had seen the new belt put on, and had, in the language of appellant’s reply brief, “the same opportunity to know and understand all the conditions incident to operating extractors with both new and old belts as other operators, including the defendants;” while the matters of possible increased speed, etc., were “matters of common knowledge to all persons operating extractors or running laundry machinery.”
In one of the cases cited by plaintiff the Supreme Court of Connecticut, while holding that a serious change in a dangerous paper making machine, increas*381ing the danger, required notice to an operator called on to run the machine, despite the fact that he had been employed on an adjoining machine and had opportunity to know of the change, yet by way of illustration, arguendo, declared:
“If the change in machine No. 2 was merely an ordinary adaptation of the machine to the purpose for which it was made, which a skilled operator must be presumed to anticipate; such, for instance, as the slight change necessary to keep the felt bands at proper tension, the risk of injury was one assumed by the plaintiff in accepting his. employment. ”
The answer of the plaintiff to this defense of “Assumption of Bisk” is that not only did the facts that the belt was newly put on and dressed in a manner the effect of which was not known or familiar to him, increase the danger, but that the mere change of material—canvas for leather—created a greater danger in the usual and approved way of starting the machine, and that plaintiff had never seen canvas belts and was not to be expected to know anything about them; that in any event there was no requirement that a “wringer man” should have knowledge of the nature or action of belts; that the plaintiff only casually observed the installation of the new belt; that there is nothing to show that he “understood and appreciated” the changed conditions or the increased danger resulting therefrom.
A reading and re-reading of all the arguments and of the evidence leave us still without a very emphatic conviction that these positions of the plaintiff are a sufficient answer, on their merits, to the defendants’ contentions, or that, had the question been first presented to us, we should have found as did the jury; but we have only to decide now, so far as the submission of the question of assumed risk to the jury goes, whether the defendants were right in their contention that all the evidence taken together had left it a ques*382tion of law. We have concluded that we should not be justified in holding that they were; and the conclusion that the conceded or uncontradicted matters appearing in the evidence did not, in this case, render the question of assumption of risk one concerning which reasonable minds could not differ, carries with it the further decision that we should not set the verdict aside as against the manifest weight of the evidence. For doubtful as the case may be on the contested matters concerning which expert and other testimony was presented, we do not think there was any such clear preponderance as requires us to interfere with the decision of the primary tribunal by which it was decided.
It is urged further, however, that even though the causes of the accident were the increased risk and the failure to warn, the act of creating the increased risk and the default in the duty of warning were the act and default, respectively, of a fellow servant of the plaintiff. This position is not sound as a matter of fact or law. Stanislawski was fairly proven, we think, not a fellow servant, but a vice principal under the true test of that question; nor can the master escape liability for a neglect to warn of increased danger resulting from a change of appliances or working conditions by any use of the fellow servant doctrine. If the duty existed, it was not delegable.
In view of our conclusions on the evidence thus somewhat too prolixly, perhaps, set forth, and of our further conclusion that the verdict was not excessive if the liability existed, we are left only to consider whether there was reversible error in the rulings or instructions of the Court below.
As to the evidence, there are no causes of complaint that we deem at all serious. We think that the questions, “Were any directions given to the employes operating those extractors by the manager?” and “What were they?” were too general in the connection in which they were asked to make the exclusion of the *383last reversible error. It is quite plain that there was no intention during the trial of preventing the proof of any notices-or instructions concerning the starting of the machine which could be brought even inferentially within the knowledge of the plaintiff.
The striking out of -a portion of -an answer of a witness which was not responsive to the question asked, was plainly harmless and immaterial and was in any event proper.
We do not think the instructions were erroneous or misleading. Those complained of are Nos. 2 (especially when read in connection with No. 6), 8 and 9. Instruction 2 was as follows:
“The jury are instructed that while an employe assumes the ordinary and obvious risks of his employment, yet he does not assume any unusual risks not ordinarily incident to his employment and which make his ordinary work attended with unusual danger, and of which danger said employe does not know and would not know of by the exercise of ordinary care, and if the jury believe from the evidence that the starting of the extracting machine in question just before and at the time of the accident, if the plaintiff in question was attended by an unusual danger to the plaintiff, which the plaintiff did not know of and would not have known of in the exercise of ordinary care, and if you believe from the greater weight of the evidence that the plaintiff started said machine in obedience to an order of his foreman, if you believe from the greater weight of the evidence there was such order, then the jury are instructed that the plaintiff did not assume the risk within the meaning of such words as used in these instructions. ’ ’
We do not think this instruction, either read alone or in connection with instruction 6, defining ordinary care, is erroneous or misleading. As we have before indicated, we do not think the fellow servant doctrine enters into this case, and there is no contradiction that *384Stanislawski, who gave whatever special order was given, was foreman.
The elements of “unusual danger” mentioned in the instruction must have been referred by the jury to those set forth in the declaration and referred to in the evidence as resulting from the change of belts. There was nothing else to refer them to.
Instruction 8 was on the tests by which the preponderance of .the evidence was to be judged. While this identical instruction has been sometimes criticized by the Supreme Court—e. g., in E. J. & E. Ry. Co. v. Lawlor, 229 Ill. 621,—in which Mr. Justice Cartwright said that the court had not given its unqualified approval to the instruction nor decided that it might not be misleading in a case where the question of numbers was important; and Lyons v. Ryerson & Son, 242 Ill. 409, in which Judge Carter said that it might have misled the jury—it has certainly received a qualified if not an unqualified approval in other cases; e. g., in West Chicago R. R. Co. v. Liesorowitz, 197 Ill. 607, and we know of none where it has, by itself, brought about a reversal. We think it implies that the number of witnesses on each side of any proposition involved in the case is one element to be considered by the jury and, other things being equal, a controlling one, and that it cannot be called inaccurate, although, like other accurate statements of the law abstractly, it might be under some circumstances misleading, as the Supreme Court says. We do not think it could have been so in this case.
Complaint is also made of instruction 9, which was the general statement that if the jury believed from the evidence that the plaintiff had made out his case as alleged in the declaration, or in any count thereof, by a preponderance or greater weight of the evidence, then the jury should find the issues in his favor. We regard this instruction in this case, not as misleading, *385but as useless and unenligktening. We cannot say it is inaccurate. The Supreme Court has frequently approved it, although in Krieger v. Aurora, Elgin & Chicago R. R. Co., 242 Ill. 544, it reversed the judgment because, as applied to the circumstances of that case, it was erroneous for the reason that the declaration therein was not “a complete statement of a cause of action. ’ ’ But we do not think the reasoning in that case applies to the one at bar, nor, as we have said, that the instruction could have been called herein either erroneous or misleading. But we do not commend it in this or any other case.
The reading of an instruction to the jury, which was afterward withdrawn in writing by the Court, cannot be considered reversible error. C. & E. I. R. R. Co. v. Zapp, 209 Ill. 339.
Appellants say reading and, on withdrawal, re-reading the instruction may have fixed the impression on the minds of the jury more firmly that it stated the law. We think it much more likely that the impression made on the minds of the jury would be that the proposition had been condemned by the Court and was not the law.
There was no abuse of the discretion of the trial judge in refusing to submit certain specific questions on the evidence to the jury for special findings and substituting others of his own, on his own motion. Such a practice has been approved by the Supreme Court when the questions submitted were proper and involved the controlling issues of fact as distinguished from some tendered in the nature of cross-examinations of the jury on their conclusions on evidentiary facts. C. & N. W. Ry. Co. v. Dunleavy, 129 Ill. 132; Fireman’s Insurance Co. v. Appleton Paper Co., 161 Ill. 9; C. & A. R. R. Co. v. Harrington, 192 Ill. 9.
We think the special findings asked by the court were proper and sufficient in the present case.
*386For the reasons indicated in this opinion we affirm the judgment.
Affirmed.