Plaintiff and appellant, a servant of defendant and respondent railway company, sought to recover for personal injuries. At the close of the evidence the court directed'a verdict for defendant. This appeal 'was taken from an order refusing to set aside that verdict and to grant plaintiff a new trial.
Plaintiff’s work, for a period of about three months, had been to shovel coal into the coal buckets of an elevator operated by a gasolene engine in a concrete "building twelve feet square. For present purposes it must also be regarded as a fact that he was in charge of the engine. On this occasion he had started it, taken his torch, and gone to work. The light was not one of the issues in the case as finally present-*262eel. While he was shoveling, he observed that the engine “didn’t sound right.” He returned to the engine house, and when he had taken one or two steps in from the door he “stepped on something and fell.” The belt and pulley were a short distance above the floor. The lower side of the belt moved from the top downward, and, passing around the driving wheel, went over the big wheel to the upper part and front of the building. What it did with him was “more than [he] could tell.” Pie was badly injured. Inter alia, part of the brain escaped through the right ear. He was found unconscious, bleeding, and dirty on a bench in the engine house. His ear was found under the lever, together with a lot of blood. The trial court directed a verdict for the defendant.
1. The initial question in the case is whether such actionable negligence was shown on the part of the defendant as to require submission to the jury. Plaintiff claims that the factory act (section 1813, R. R. 1905) required respondent to make an additional protection for this engine within the inclosure built for it. Defendant argues that this statute did not apply, and that under any circumstances the building complied with its provisions and rendered a second guarding unnecessary. A majority of the court are of the opinion that the terms of the statute are broad enough to apply to this case.
In Christianson v. Northwestern Compo-Board Co., 83 Minn. 25, 27, 85 N. W. 826, 827, 85 Am. St. 440, Start, C. J., said: The statute “was intended to protect from personal injury the workmen or employees in any shop, mill, or factory by reason of dangerous machinery therein, and it must be so construed as to give effect to such wise and humane purpose.” The writer of this opinion thinks that this was an engine room, and not a workshop, and that the statute did not apply. Pie is also of the opinion, however, in view of the positive duty, apart from the statute, which rests on the master to exercise due care in providing a reasonably safe, place for the servants’ work, that, although the master was not necessarily liable for the failure to guard, it was a question for the jury to determine what protection should have been afforded and whether a guard was reasonably necessary. Smith v. Baker, 1891 App. Cas. 325, at page 353; Lloyd v. Woolland Bros., 87 L. T. (N. S.) 73; Wheeler v. Wason, 135 Mass. 294, 296; Nadau v. White River, 76 Wis. 120, 126, 43 N. W. 1135, 20 Am. St. 29; Ander*263son v. C. N. Nelson Lumber Co., 67 Minn. 79, 81, 69 N. W. 630, per Mitchell, J.; 1 Labatt, Master & Servant, § 76, and cases cited.
It appears from the record that no railing or other device was provided. No particular difficulty has been suggested in the matter of providing an adequate safeguard against the occurrence of just such an accident as here took place. Evidence that it was practicable and feasible so to guard this machinery as to provide reasonable protection against such a fall as plaintiff experienced was admissible. The trial court erred in excluding it. In Myers v. Hudson, 150 Mass. 125, 128, 22 N. E. 631, 15 Am. St. 176; Allen, J., said: “In order to aid the jury in determining whether defendant had exercised reasonable care in providing and maintaining the machinery in actual use, it was competent to show what other kinds of machinery or appliances were used elsewhere, and might have been used here.” This general rule has been specifically applied to evidence as to guards of machinery in use. Wheeler v. Wason, 135 Mass. 294. And see Peterson v. Johnson-Wentworth Co., 70 Minn. 538, 541, 73 N. W. 510; Barbo v. Bassett, 35 Minn. 485, 29 N. W. 198; Poczerwinski v. C. A. Smith L. Co., 105 Minn. 305, 117 N. W. 486.
The four walls of the building were no more an adequate guard, because the building was small, than if the building had been large. That they protected men working on the outside did not tend to show that they would protect men working on or abput the engine. Indeed, the limited space within which the plaintiff worked was likely to increase the peril to which he was exposed.
It is urged, however, that defendant’s negligence in failing to guard the machine was not sufficiently shown to have been the proximate cause of the injury of which he complains. Circumstantial evidence tended directly -to show that plaintiff was injured by coming in contact with the moving belts and machinery. How else the accident could have occurred has not been suggested. That there is an unwritten page in this narrative is obvious. But whether it was negligence not to provide a guard, and whether, if that negligence was proved, it was the proximate cause of the injury, were for the jury on authority and on principle.
“Proof of proximate cause is subject to no more burdensome rule than is applied to the proof of any other essential fact in an ordinary *264law action. It must be established by a preponderance of the evidence, direct or circumstantial. * * * A cause being shown which might produce an accident, and it further appearing that an accident of that particular character did occur, it is a warrantable inference, in the absence of showing of other cause, that the one known was the operative agency in bringing about such result.” Weaver, J., in Lunde v. Cudahy (Iowa) 117 N. W. 1063, at page 1068. In Bolen-Darnell Coal Co. v. Williams (C. C. A.) 164 Fed. 665, 669, the servant walked along a narrow path, slipped, and was precipitated into a pit and on a wheel. The question of cause was submitted to the jury. In Swick v. Ætna, 147 Mich. 454, 111 N. W. 110, plaintiff was opening an air cock in a pipe leading into a vat. His wrench slipped, and he came in contact with an unguarded belt running on pulleys. In holding that the unguarded belt was the proximate cause of the accident, Hooker, J., said, at page 462: “Nor should it be said that the slipping of the wrench and not the unguarded belt was the proximate cause of the ‘injury. Manifestly, if this man had not fallen or slipped, or if something out of the ordinary had not happened, he would not have come in contact with the belt; but that is always the case where one is hurt by machinery. Something happens which causes him to get against or into it, and this case is not distinguishable, in that respect, from Swoboda v. Ward, 40 Mich. 420, where a plaintiff stepped into some unguarded cog wheels.” And see Miller v. Kimberly, 137 Wis. 138, 118 N. W. 536, 538.
' 2. The learned trial court after careful consideration held that as a matter of law plaintiff assumed the risk. As it will be subsequently pointed out, his conclusion was sustained and opposed by an almost equal number of decisions of this court on similar facts. This is one of the class of constantly recurring cases in which counsel are unable to advise their clients and courts unable to agree or decide consistently. It would be easy, but inadequate,, to dogmatically determine this particular question without considering the “veritable chaos of conflicting precedents” which have applied the doctrine. It may tend to clarification of the subject to consider at some length its ultimate basis, its character as a distinctive defense, its standards, and the extent to which assumption of risk is a question for the court or for the jury.
In the first place it is insisted at the one extreme, that its ultimate basis is economic, not juristic (1 Habatt, Master & Servant, §§ 62, *265156), and that sound public policy demands, if not its abolition, its restriction and modification to a considerable degree. The tendency of academic discussion in this direction is perhaps not more pronounced than of that of many decisions. In Yarmouth v. France, 19 Q. B. Div. 647, 653, for example, Lord Esher said: “It is to my mind a horrible way of stating the duty, to say that a master owes no duty to a servant who knows that there is a defect in machinery and, having pointed it out to one in authority, goes on using it. It seems cruel and unnatural, and in my view utterly abominable.” Among the many objections urged are these: The original rule was artificial in origin and is unjust in operation, because it imposes upon the servant the risks of employment with compensation for consequent injury with which the industry should be charged. The servant does not in any common-sense way exercise any volition. He must work where danger exists, or in some other place with corresponding peril, or remain idle. See Lindley, J., in Yarmouth v. France, supra. The reasoning by which it is sought to be justified, carried to its logical conclusion, tends to result in this paradox: The more grossly the master is negligent, the more certain is the assumption of risk by the servant and the master’s exoneration. If the master be more careful, then the more doubtful is the servant’s assumption of risk, and the more probable is the master’s liability. The employer who is successfully careful and he who is extremely careless are equally protected. The exercise of care is discouraged, and a premium is put on negligence.
At the other extreme many authorities regard the rule as founded in nature because the general risks inherent in the employment are necessarily assumed. There is a natural difference between the necessary dangers involved, for example, in working at a height, as on a steeple, and on the level or in a sand pit; in handling dynamite, or in driving a mule; in operating machinery which revolves or moves, or in selling dry goods behind a counter; in ordinary operations and in peculiar exigencies. The master has a natural right to conduct his business in his own way and the servant to refuse to work under its conditions. Moreover, the rule is so permanently fixed as to render iconoclastic attempts to modify or repeal it or to refuse to follow its reasoning to the logical extremity of holding assumption of risk to be a matter of law. Compare Schlemmer v. Buffalo, R. & P. Ry. Co., 207 Pa. St. 198, 56 *266Atl. 417, with the same case in 205 U. S. 1, 12, 15, 27 Sup. Ct. 407, 51 L. Ed. 681.
It is undoubtedly true that the doctrine is so well established that to abrogate it would be judicial legislation. The difficulty arises because the premises on which the rule, regarded as juristic, is based are so confused and indefinite that to deduce from them with metaphysical consistency is almost certain to result in practical absurdity. None the less a faithful effort, accompanied by neither hostility to nor advocacy of the abstract merits of the doctrine, must be made to discover and apply the law as the courts, taken as a whole, have determined it to be.
In the second place, the question naturally arises whether assumption and contributory negligence are separate defenses, and, if they are, what is the essential distinction between them. In the leading case of St. Louis Cordage Co. v. Miller, 126 Fed. 495, 515, 61 C. C. A. 477, 510, 63 L. R. A. 551, Sanborn, Circuit Judge, said of three cases decided by the supreme court of the United States and of other cases that they so confused the two separate defenses as to be without persuasive force. Thayer, Circuit Judge, dissenting, said: “I do not regard the question whether ‘contributory negligence’ and ‘assumption of risk' * * * are identical or are different defenses as of much practical importance. That is rather a'question for the schoolmen,” Without reasonable controversy, current opinion, judicial and academical, has insisted that the defenses must be regarded as distinct, and that an injured servant who has not been guilty of contributory negligence may be precluded as a matter of law from recovery of damages from his master because he had assumed the risk. For example, see Miller v. White (Iowa) 118 N. W. 518, at page 523; Chicago v. Heerey, 203 Ill. 492, 68 N. E. 74, collecting cases. It is true, however, that in some jurisdictions — for example, in Wisconsin — only “some degree of distinction is observed” (Campshure v. Standard, 137 Wis. 155, 118 N. W. 633); that “even the Massachusetts cases have occasionally wobbled” (8 Harvard L. Rev. 469 [5]); and that members of the federal supreme court are not agreed as to the exact nature of the distinction. See Schlemmer v. Buffalo, R. & P. Ry. Co., 205 U. S. 1, 6, 12, 16, 27 Sup. Ct. 407, 409, 410, 51 L. Ed. 681. “The extraordinary prevalence” of failure to observe this primary distinction has been one of the many causes of confusion in thought on this subject. In a particularly valúa*267ble note to be found in 49 L. R. A. 33, from page 46 to 62, inclusive, Mr. Rabatt considers the relations of contributory negligence and assumption of risk, demonstrates that courts and counsel have been guilty of inexcusable remissness in neglecting the natural distinctions between the doctrines, and puts in a strong light the deplorable consequences, sometimes amounting to a miscarriage of justice, which inaccurate terminology and logical laxity have contributed to produce.
The exact nature of the distinction, however, is not generally agreed upon. It is often said that assumption of risk is a matter of contract; contributory negligence of tort, or of conduct. This is fallacious. As will subsequently be pointed out, assumption of risk cannot rest on contract in this particular group of cases at least. Both defenses are equally peculiar to the law of tort. Both are usually implied from conduct. A view at least less indefensible is this: Failure to exercise due care, which the law has imposed as a duty, is the necessary differentia which distinguishes the species, negligence, from the genus, tort. It is equally an essential characteristic of contributory negligence. It has no logical connection with assumption of risk. Carelessness is not the same thing as intelligent choice. Bowen, L. J., in Thomas v. Quartermaine, 18 Q. B. Div. 685; In re Morgan, 26 Colo. 415, 58 Pac. 1071, 47 L. R. A. 59, 77 Am. St. 269. Voluntary assumption negatives the idea of even prima facie liability. Contributory negligence displaces liability prima facie established. The former is mere passive subjection by the servant to risk of injury inherent in known defective conditions. The latter is an act or omission on complainant’s own part tending to add new danger to his situation not necessarily incident to conditions, and bringing upon himself a harm caused not solely by them, but created in part, at least, by his own misconduct. Mr. Francis Bohlen, in 21 Harvard R. Rev. 246, 249; Mr. Chas. Warren, in 8 Harvard R. Rev. 458. Contributory negligence is a breach of legal duty to take due care, imposed by law upon the servant, however unwilling or protesting he may be. Assumption of risk is not a duty, but is purely voluntary upon the part of the servant. The doctrine of assumption of risk rests on intelligent acquiescence with knowledge of danger and appreciation of the risks. The distinction varies from being clear and vital at one extreme to being vague and insignificant at the other. This is equally true of the shadowings of many other legal principles. It *268must sometimes occur that the circumstances of a case affecting the defenses are as interwoven and the defenses themselves so interrelated that for purposes of convenience the two defenses may be conveniently considered at the same time.
In the third place, the question is whether the doctrine rests on contract or on the principle expressed by the civil-law maxim, “Volenti non fit injuria.” In Thomas v. Quartermaine, L. R. 18 Q. B. Div. 685, 698, Bowen, L,. J., said: “It is in most cases a barren and metaphysical inquiry to discuss whether such duties are best treated as arising by implication from the contract or from the general law outside.” While there is obviously much wisdom in this comment as a criticism of current legal thought, the fact remains that in a number of classes of cases (see note to O’Maley v. South Boston [158 Mass. 135] 47 L. R. A. 161, at page 164), and in the class of cases in particular to which the instant case belongs, the conclusion to be reached depends upon the decision on which of the two bases the doctrine is to be rested, as will subsequently appear.
The doctrine of estoppel is sometimes, but decreasingly, used to account for the rule. Beach, Cont. Neg. § 16; Busch v. Robinson, 46 Ore. 539, 81 Pac. 237. Estoppel, however, prevents the successful assertion of an accrued right. Assumption of risk prevents the accrual of the right. Moreover, it will be found to resolve itself into either or both the contract or the maxim theory. At best, the term enriches the vocabulary, but tends to confuse the reasoning, by introducing the use of an ambiguous middle.
Nothing is gained by treating assumption of risk as “part of the body of law incident to the status or relationship of master and servant.” 5 Columbia L. Rev. 159. The suggestion involves at the best an identical proposition. Moreover, if the doctrine should be regarded as based on the civil-law maxim, it is not peculiar to the relationship. See Lord Halsbury, in Smith v. Baker, 1891 App. Cas., at page 337.
The late and lamented Dean Huffcut adopted the view that the contract feature should be regarded as controlling defects and dangers which existed at the commencement of the service and the maxim those subsequently arising. Huffcut, Agency, 354. While there is some judicial sanction for the device, it has not been enforced with any considerable degree of unanimity or consistency, nor does it avail to *269solve the difficulty involved in the controversy over the master’s failure to comply with various factory acts, as will subsequently appear.
In this class of cases Judge Taft’s opinion in Narramore v. Cleveland, C. C. & St. L. Ry. Co., 96 Fed. 298, 37 C. C. A. 499, 48 L. R. A. 68, has demonstrated that assumption of risk by the servant is excluded as a defense if the hypothesis be that the doctrine rests on contract; for the agreement by the servant that his master violate a penal statute is necessarily illegal and void. Sipes v. Michigan, 137 Mich. 255, 100 N. W. 447. On this basis, it is not clear what answer can be made to the authorities holding that a servant cannot be held to have assumed the risk of the master’s negligence generally because of his inability to contract against his master’s wrong: Buswell, Personal Injuries, § 207; Brents v. Louisville, 31 Ky. L. R. 1216, 104 S. W. 961; Kirby v. Manufacturers, 127 Mo. App. 588, 106 S. W. 1069; Garaci v. Hill, 124 Mo. App. 709, 102 S. W. 594. See Limberg v. Glenwood, 127 Cal. 598, 60 Pac. 176, 49 L. R. A. 44; San Francisco & P. S. S. Co. v. Carlson (C. C. A.) 161 Fed. 851, 854; 8 Current Law, 908 (a fortiori against his wilful wrong); Roff v. Summit, 119 La. 571, 44 South. 302. See Lord Herschell, in Smith v. Baker, 1891 App. Cas., at page 362. In Choctaw O. & G. R. Co. v. McDade, 191 U. S. 64, 24 Sup. Ct. 24, 48 L. Ed. 96, the federal supreme court has recognized the rule last stated, but has in effect abrogated it by an exception. Cf. Texas & P. Ry. Co. v. Archibald, 170 U. S. 665, 18 Sup. Ct. 777, 42 L. Ed. 1188.
There are significant objections to resting the doctrine on the maxim, volenti non fit injuria. Why should the servant be permitted, by assenting to his master’s wrong, to connive with his master to evade statutory obligations, and bind himself to do what he could not bind himself to do by agreement for a consideration? See Baddeley v. Granville, 56 L. J. (N. S.) 501, per Willes, J Why should the master be held liable if he orally promises to correct the evil complained of by the servant and be exempted from an obligation imposed by positive statute? 61 Cent. L. J. 448. The only answer is that such is the well-established rule of law.
It is upon the maxim that later English cases consistently and expressly sustain the interposition of this defense by the master in cases under the factory acts. The earlier English cases and the American *270cases generally do not appear to have deliberately answered the question, and have based decisions indifferently upon either or both the contract and the maxim. Ordinarily this confusion has been harmless. It has, however, occurred where the distinction was important, as in cases under the various factory acts. For example, in St. Louis Cordage Co. v. Miller, 126 Fed. 495, 61 C. C. A. 477, 63 L. R. A. 551, the syllabus written by the circuit judge sets forth that exemption of risk rests on contract. Subsequently the opinion rests the doctrine both on the maxim, volenti non fit injuria, and upon contract. In the case immediately following, Glenmount L. Co. v. Roy, 126 Fed. 524, 528, 61 C. C. A. 506, 510, the same judge rests the doctrine on “the right to contract.”
The authorities of this country, however, are generally veering around to accord with the opinion that “the consent must be bottomed on the maxim ‘volenti non fit injuria’ — that to which the party assents is no wrong.” Per Deemer, J., in Miller v. White (Iowa) 118 N. W. 519, 523. The doctrine is no more based on contract than is the action of the servant resulting from a breach of duty ex contractu. Martin v. Chicago, 118 Iowa, 148, 98 N. W. 1034, 59 L. R. A. 698, 96 Am. St. 371, 376. The best-considered case on the subject (Dempsey v. Sawyer, 95 Me. 295, 49 Atl. 1035) thus summarizes the law: “The risk of injury to a servant from defective machinery is primarily upon the master, and remains upon him unless the servant voluntarily assumes it. The servant may voluntarily assume such risk, and relieve the master from it; but such assumption is his voluntary act, not his legal duty.” And see Birmingham v. Allen, 99 Ala. 359, 13 South. 8, 20 L. R. A. 457. In Denver & R. G. R. Co. v. Norgate, 141 Fed. 247, 253, 72 C. C. A. 365, 6 L. R. A. (N. S.) 981, a factory act case, Carland, J., regarded the Narramore case and Green v. Western, 30 Wash. 87, 70 Pac. 310, as arising from the false doctrine that the law of assumption of risk is “created by the contract between master and servant,” and said: “It is independent of the will of either. It is not a term of the contract of employment. * * * It is a principle of common law. * * * It is over and above the contract, and depends in no manner for its existence upon agreement of the parties. It is founded upon public policy, the status assumed by master and servant, and upon the maxim ‘volenti non fit injuria.’ ”
*271The fourth question is whether, in the absence of proof of actual knowledge of damages and appreciation of the risk, the doctrine of assumption of risk imputes to the servant the knowledge and appreciation resulting from the exercise of the care of the ordinarily prudent man. Obviously, if he is so held, the distinction between contributory negligence and assumption of risk breaks down, and assumption of risk may become a question of care. Some opinions by judges who recognize that the two defenses are separate have, however, held the defendant to the standard of that favorite of the law. Apart from the hopeless logical inconsistency, the practical effect of such a rule involves a reversal of undisputed and indisputable elements of the law of negligence of the master. The master and servant do not stand on the same footing. The master, not the servant, owes a duty of inspection, and the servant has a right to assume that the master has exercised due diligence in providing reasonably safe appliances and in inspecting his premises and instrumentalities. He is not bound to anticipate his master’s negligence. He is not, in theory, at least, required to exercise due care in investigating his master’s actual carelessness. He “is bound to use his eyes and to see that which is open and apparent to any person using his eyes.” Dillenberger v. Weingartner, 64 N. J. L. 292, 45 Atl. 638. “But he is not required to look for danger.” Bolen-Darnell Coal Co. v. Williams (C. C. A.) 164 Fed. 665; Riley v. Cudahy (Neb.) 117 N. W. 765, 788. It is the master’s duty to both observe and inspect; the servant’s, merely to observe. Bailey, Master’s Liability, 107, 108, 109; Illinois v. Mann, 100 Ill. App. 367, affirmed 197 Ill. 186, 64 N. E. 328.
By reversing this reasoning, certain courts appear in this particular class of cases to have achieved the dialectical feat of absolving both the master and the servant from the duty of being careful, and of abolishing the law of negligence and of contributory negligence. For admittedly the servant assumes risks which are obvious and open; on this theory he also assumes what reasonable care would disclose; and it has been held that he also assumes latent defects which the master’s exercise of due care would not have discovered. See Cryder v. Chicago, R. I. & P. Ry. Co., 152 Fed. 417, 81 C. C. A. 559. It is plain that in the last case initial negligence could not have been shown.
If knowledge which the servant would have acquired in the exercise *272of the care of an ordinarily prudent man is not the standard, by what criterion is that constructive knowledge to be determined? The old-fashioned formula as to the distinction between ordinary and extraordinary risks has by no means entirely disappeared, but is largely fading from vision, because in so many cases no one can tell where the one begins and the other ends. This is true, also, with respect to the distinction between usual and unusual risks. In Calloway v. Agar, 129 Iowa, 1, 104 N. W. 721, the rule is laid down that to make out assumption of risk the master must show either that the employee knew and appreciated the incidental danger or that the conditions were such that they rendered ignorance inexcusable. But in De Kallands. v. Washtenaw, 153 Mich. 25, 116 N. W. 564, it was held that the actual ignorance of the servant of the danger of his employment will not alone suffice to relieve him from his assumption of the risk but that ignorance must be excusable. Excusable ignorance, however, finally resolves itself into the extent to which the danger and risk obtrudes upon the attention of the employee. This is true generally of other suggestions for a substituted. standard.
In Laughy v. Bird, 136 Wis. 301, 117 N. W. 796, 797, Marshall, J., said of the consideration that, if the employee has as good or better opportunity to know of the defects as the master, he assumes the risk: “That is good law; but it does not suggest that defects which are discoverable by the exercise of such care as is required to be exercised by the master are necessarily discoverable by such attention to his surroundings as an employee is bound to exercise. Defects may be such as to be perfectly plain to the former and not attract the attention of the latter, since no duty of inspection rests with the servant to turn aside from the acts of operating in order to specially look or search for defects, in the absence of some circumstances reasonably efficient to suggest existence thereof. The employee is only required to see and comprehend those imperfections which are so significant as to ordinarily attract the attention of a person under the same or similar circumstances, in the exercise of ordinary care, and whether the servant in any particular instance is chargeable with knowledge of conditions imperiling his personal safety is, generally, although, of course, not always, a jury question.” With the exception of the confusion of assumption of risk with contributory negligence, this statement is a par*273ticularly clear variation of the natural and generally accepted standard. Of this, there are a number of equivalent formulae, as whether the danger or defect was so “clear, plain, obvious,” or “so readily observable” as to be apparent to a person of ordinary intelligence (Christiansen v. William, 7 Am. & Eng. An. Cas. 69), or “such as is apparent to observation” (Davidson v. Cornell, 132 N. Y. 228, 30 N. E. 573), or “in the exercise of common observation, any man of ordinary intelligence would completely understand, and appreciate the .situation” (Hermann v. Clark, 89 Minn. 132, 136, 94 N. W. 436, 438).
In Choctaw O. & G. R. Co. v. McDade, 191 U. S. 64, 24 Sup. Ct. 24, 48 L. Ed. 96, it was said “the true test is not in the exercise of care to discover dangers, but whether the defect is known [to] or plainly observable by the employee.” The plaintiff is not required to look for danger, but is held to assume the risks — only when the danger is so apparent that one who owes no duty to inspect was bound to discover it, but that is ordinarily a question of fact for the jury. As to appreciation of risk, as distinguished from knowledge of danger, the question is: Did the servant understand the risk, or by the exercise of ordinary observation ought he to have understood the risk, to which he was exposed by the dangerous situation? Mitchell, J., in Scharenbroich v. St. Cloud Fiber-Ware Co., 59 Minn. 116, 121, 60 N. W. 1093.
Most, if not all, of the opposed decisions, which have adopted indifferently either or both this standard and the exercise of reasonable care, are to be differentiated, because they were rendered when and where the distinction between contributory negligence and assumption of risk was ignored. Indeed, the misuse of ordinary care in this direction has usually been the result of inadvertence and has often been the sequence to mere casual inaccuracy in language. The test has rarely, if ever, been the result of deliberate consideration of principle or precedents. As Mr. Justice White pointed out in Texas & P. Ry. Co. v. Archibald, 170 U. S. 665, 18 Sup. Ct. 777, 779, 42 L. Ed. 1188, “no reason can be found for and no authority exists supporting the contention.” It would be a work of supererogation to add to the authorities there cited. They exist, however, in large numbers.
It is evident that there are objections to the adoption of this criterion. The traditional variations in the size of the chancellor’s foot are no more uncertain. Yet the question is universally regarded as one *274resting, not in the sound discretion of the court, but in positive law, unless it should be in the sense that the law should not be too exacting on one situated as the servant is. Earl, J., in Hawley v. Northern, 82 N. Y. 370. The decision must depend largely upon the circumstance: of each case and with due reference to the familiar rule that the jury' must determine where reasonable minds would draw different conclusions as to what the facts are and as to what inferences should be drawn from them.
In the fifth place, the gist of the controversy arises when courts are asked to give a peremptory charge to the jury. When are the. questions of fact, and when of law? It will conduce to avoidance of current confusion to consider the multitude of cases on this subject in this order: First, the English cases; second, the American cases generally; and, third, the previous decisions of this court.
First. In Thomas v. Quartermaine, 18 Q. B. Div. 685 (1887), plaintiff brought action under the employer’s liability act for injuries received by falling into a cooling vat which was not fenced in. It was held by a divided 'court that the defense arising from the maxim, volenti non fit injuria, had not been affected by the employer’s liability act and that therefore plaintiff was not entitled to recover. The pronounced difference of opinion between the judges who decided this case became the basis of much controversy. While the American cases have rarely observed the pronounced hostility to it on the part of the English judges, its career in its own country has been a stormy one. In Baddeley v. Granville, 56 L. J. (N. S.) 501 (1887), it was regarded as holding that the maxim did not apply at all to breaches of statutory ■obligation. In Thrussell v. Handyside, L. R. 20 Q. B. Div. 359 (1888), it was held that the servant working with knowledge of the danger and of the risk could recover for injuries inflicted by the servant of an independent contractor, which differentiates it from Thomas v. Quartermaine supra, notwithstanding the maxim. In Yarmouth v. France, 19 Q. B. Div. 647 (1887), it was pointed out that the Quartermaine case was decided upon imperfect argument (see page 654), and that its holding in effect was that “mere knowledge of the danger will not do. There must be an assent on the part of the workman to accept the risk, with full appreciation of its extent, to bring the workman-within the maxim ‘Volenti non fit injuria.’ If so, that is a question of fact.” *275Ford Esher, at page 657. Findley, F. J., said at page 659 : “The question whether in any particular case [the conduct of the servant] was [volenti et scienti] is a question of fact and not of law. * * * The mere fact that the plaintiff knew of the danger and yet incurred it is not conclusive. * * * The question in each case must be, not simply whether the plaintiff knew the risk, but whether circumstances are such as necessarily to lead to the conclusion that the whole risk was voluntarily incurred by the plaintiff. * * * The doctrine does not apply when what the servant does is induced by fear of dismissal.”. The opposed view of Ford Bramwell was “that where a man is not physically constrained, and where he can at his option do a thing or not, the maxim * * * applies.” Membery v. Great Western, 14 App. Cas. 179. According to Mr. Fabatt, whose opinions on this subject are entitled to peculiar consideration (1 Fabatt, Master & Servant, § 62), Ford Bramwell was “one of the straightest of the sect of those economic Pharisees whose Gamaliels were such writers as Ricardo and John Stuart Mill. It is not too much to say'that the opinion delivered by him in Dynen v. Leach (1857) N. L. J. Exch. (N. S.) 221, has contributed more largely than any other judicial utterance to establish the principle of assumption of risks, in the rigorous and unmerciful form in which we now have it.”
It is clear that the English courts do not accept Ford Bramwell’s views. In, Smith v. Baker, 1891 App. Cas. 325, plaintiff was employed by a railroad contractor to drill holes in a rock cutting near a crane worked by men in a different department and over which he had no control. The crane lifted stones and at times swung them over the plaintiff’s head without warning. The plaintiff was fully aware of the danger to which he was exposed by thus working near the crane without any warning being given, and had been thus employed for months. Complaint to the master had been made. A stone fell from the crane and injured him. It was held that the action was maintainable, and that the master’s systematic neglect to give warning did not preclude him from recovery. The general rule laid down is: “The mere fact that a servant undertakes or continues [in a dangerous] employment with full knowledge and understanding of the danger is not conclusive to show that he has. undertaken the risk so as to make the maxim * * * applicable in case of injury. The question * * * is one *276of fact, and not of law; and this is so both at common law and * * * under the employer’s liability act.” Emphasis was laid upon the fact that the maxim was not scienti, but volenti. It is significant that Lord Bramwell held that plaintiff assumed the risk as a matter of law. In Williams v. Birmingham (1899) 2 Q. B. 338, while a workman, in the course of his employment, was descending from an elevated tramway belonging to his employers, his foot slipped and he fell to the ground, receiving injuries which caused his death. The employers had provided no ladder or other safe means of ascending to or descending from the tramway. The jury found that it was dangerous to descend from the tramway without a ladder, that the deceased had the same means of knowing that it was dangerous as the attendants had, and that he did know that it was dangerous. None the less it was held that, in the absence of a finding by the jury that the deceased had agreed to undertake the risk of descending from the tramway without a ladder or other safe means of descent, the plaintiff was entitled to judgment upon the findings of the jury. Romer, L. J., thus summarizes the liability at common law of an employer of labor: “If the employment is of a dangerous nature, a duty lies on the employer to use all reasonable precautions for the protection of the servant. If, by reason of a breach of that duty, a servant suffers injury, the em^ ployer is prima facie liable, and it is no sufficient answer to the prima facie liability for the employer to show merely that the servant was aware of the risk and of the nonexistence of the precautions which should have been taken by the employer, and which, if taken, would or might have prevented the injury. In order to escape liability, the employer must establish that the servant has taken upon himself the risk without the precautions. Whether the servant has taken that upon himself is a question of fact to be decided on the circumstances of each case. In considering such a question the circumstance that the servant has entered into or continued in his employment, with knowledge of the risk and of the absence of precautions is important, but not necessarily conclusive against him.” Page 345. It is to be noted that in this case no proper appliances had ever been provided, and this negligence was the cause of the accident. Compare Griffiths v. London, 12 Q. B. D. 493, 13 Q. B. D. 259. To conclude, in England, “it would seem, however, that there is now a clear tendency *277on the part of the courts to treat the question of what inference is to be drawn from the fact of the plaintiff’s knowledge of the danger as always one of fact for the jury. * * * At any rate, the mere fact of a servant continuing to work, with knowledge of the danger and appreciation of the risks, will not necessarily raise the inference that he was volens to take the risk.” 17 Jurid. Rev. 45, 46. The best American review of the English authorities will be found in Mr. Labatt’s note to 47 L. R. A., at page 161. And see Kager v. Vickery, 61 Kan. 342, 59 Pac. 628, 49 L. R. A. 161, 78 Am. St. 318.
Second. ' In America the confusion of authorities has been aggravated and not diminished by the change in English theory. The English rule has received judicial and academical sanction. Marks v. Cotton, 138 N. C. 401, 50 S. E. 769, 3 Am. & Eng. An. Cas. 812-814; Lloyd v. Hanes, 126 N. C. 359, 35 S. E. 611. In 20 Harvard R. Rev., Mr. Frances H. Bohlen, discussing the liability of the master under the statute says (page 22): The servant must have encountered and known the danger “under such circumstances as to indicate that he voluntarily, willingly, and affirmatively assumed the risk; that he recognized the danger, knew his right to be protected from it, but still chose to encounter it, not because of the pressure or coercion put upon him by the defendant’s wrongdoing, but for some private reason of his own.” More frequently the English rule has been considered and given weight without such formal adoption as to render the result entirely inconsistent with the former general opinion. Choctaw v. Jones, 77 Ark. 367, 92 S. W. 244, 4 L. R. A. (N. S.) 837; Atlas v. National Biscuit Co., 100 Minn. 30, 110 N. W. 250; Dempsey v. Sawyer, 95 Me. 295, 49 Atl. 1035, at page 1038. In Mahoney v. Dore, 155 Mass. 513, 30 N. E. 366, the English rule was apparently adopted ; in O’Maley v. South, 158 Mass. 135, 32 N. E. 1119, 47 L. R. A. 161, which has been regarded as “of little practical effect” (8 Harvard R. Rev. 469), it was apparently rejected (47 L. R. A. 161; 1 Labatt, Master & Servant 376).
The English cases, taken as a whole, have tended directly, naturally, and logically to induce American courts to place emphasis, inter alia, upon the rules that the maxim is not “scienti,” but “volenti,” that in close cases the questions are for the jury, and that assumption of risk *278is not presumed; the burden of proof being upon the defendant to show assumption in fact, or such facts as lead to its natural inference. It is to be noted in this connection, however, that there is earlier authority as to this burden of proof. Blankavag v. Badger, 136 Wis. 380, 117 N. W. 852; Nadau v. White River, 76 Wis. 120-131, 43 N. W. 1135, 20 Am. St. 29; Swoboda v. Ward, 40 Mich. 420-424. But compare Dallemand v. Saalfeldt, 175 Ill. 310, 51 N. E. 645-647, with Chicago v. Heerey, 203 Ill. 492, 68 N. E. 75.
The statement is currently made that the American cases taken as a whole are inclined to regard the questions involved in assumption of risk — knowledge of dangerous conditions, appreciation of risk, and acquiescence therein — as for the jury. In connection with this opinion many hundreds of cases have been examined and that tendency verified. It is conspicuous in the great number of distinctions drawn by courts, some shadowy, some substantial, in the apparent effort to approximate what they have regarded • as substantial justice without abolishing the rule. None the less the tendency has often varied. Consistencies are innumerable. To enumerate or discuss the vast multitude of relevant precedents is out of the question.
A few of many illustrations will be referred to. In general, see Mahoney v. Dore, 155 Mass. 513, 30 N. E. 366 (per Knowlton, J.) ; Dempsey v. Sawyer, 95 Me. 295, 49 Atl. 1035; Fitzgerald v. Connecticut, 155 Mass. 155, 29 N. E. 464, 31 Am. St. 544; Choctaw, O. & G. R. v. McDade, supra. The tendency is clear in many cases which have considered facts more or less similar to these in the case at bar. In Washington, & G. R. Co. v. McDade, 135 U. S. 554, 10 Sup. Ct. 1044, 34 L. Ed. 235, plaintiff was injured while pulling a belt on a pulley by hand. He had for eighteen months performed the same duties as when injured. After stating the current formula as to contributory negligence and assumption of risk, Eamar, J., said, quoting from Jones v. East Tennessee, V. & G. R. Co., 128 U. S. 443, 9 Sup. Ct. 118, 32 L. R. A. 478: “We see no reason, so long as the jury system is the law of the land, and the jury is made the tribunal to decide disputed questions of fact, why it should not decide such questions as these as well as others.” ' And see Miller v. Kimberly, 137 Wis. 138, 118 N. W. 536; Cox v. Capital, 47 Wash. 148, 91 Pac. 555 (employee working on a rip saw injured by hand coming into con*279tact); Louisville v. Wilson, 124 Ky. 836, 99 S. W. 634, 30 Ky. L. R-734 (revolving machinery which plaintiff did not know could have been guarded); Heffernan v. Fall River, 197 Mass. 28, 83 N. E. 5 (cleaning machinery) ; Little v. Bousfield (Mich.) 117 N. W. 903 (servant’s clothing caught in revolving set screw while replacing pulley belt) ; Chopin v. Combined, 134 Wis. 35, 114 N. W. 95 (similiter); Kalker v. Hedden, 72 N. J. L. 239, 61 Atl. 395 (defective keys used to fasten a pulley to a revolving shaft, which caught in the edge of the belt).
In Busch v. Robinson, 46 Ore. 539, 81 Pac. 237, as in the case at bar, plaintiff slipped and fell, and was injured by being caught in machinery. In that case a board on a platform on which she was compelled to stand had been negligently allowed to become defective. It was held that the question of assumption of risk was for the jury. And see Mundhenke v. Oregon, 47 Ore. 127, 81 Pac. 977, 1 L. R. A. (N. S.) 278; Meade v. Ashland, 125 Ky. 114, 100 S. W. 821 (in which plaintiff on several occasions had nearly fallen into the cogs). See also Cristanelli v. Saginaw, 117 N. W. (Mich.) 910. A different element is introduced when the floor on which the plaintiff slipped is to Lis knowledge wet and slippery. Ordinarily tLe risk tLen is assumed. Kleinest v. Kunhardt, 160 Mass. 230, 35 N. E. 458; Feely v. Pearson, 161 Mass. 426, 37 N. E. 368; Scharenbroich v. St. Cloud, Fiber-Ware Co., 59 Minn. 116, 60 N. W. 1093.
Third. The decisions of this court involve the complexity and confusion and inconsistency of opinion on the subject which are to be found everywhere. The rule that both contributory negligence and assumption of risk may be utilized as defenses by a defendant sought to be charged with negligence because of violation of a statutory duty to safeguard dangerous appliances was early adopted and has been consistently followed. Anderson v. C. N. Nelson L. Co., 67 Minn. 79, 69 N. W. 630; Hermann v. Clark, 89 Minn. 132, 94 N. W. 436; Swenson v. Osgood & Blodgett Mnfg. Co., 91 Minn. 509, 98 N. W. 645; McGinty v. Waterman, 93 Minn. 242, 101 N. W. 300; Schutt v. Adair, 99 Minn. 7, 108 N. W. 811. Inevitably, however, uncertainty has arisen in applying the rule. Such cases constitute a class by themselves. See De Greif v. Northwestern Knitting Co., 106 Minn. 15, 118 *280N. W. 558. The distinction between knowledge of danger and appreciation of risk has been repeatedly recognized and affirmed.
The leading opinion on the subject in this jurisdiction, which has been followed times without number by other courts, and the one most nearly controlling the present controversy, is that of Mitchell, J., in Wuotilla v. Duluth L. Co., 37 Minn. 153, 33 N. W. 551, 5 Am. St. 832. There plaintiff’s clothing was caught while he was going past gearing where two wheels “mashed,” whereby he was injured. He knew the gearing was uncovered. The question of appreciation of risk was submitted to the jury. Among other decisions, Gray v. Commutator Co., 85 Minn. 463, 89 N. W. 322, Lund v. Woodworth & Co., 75 Minn. 501, 78 N. W. 81, Christianson v. Northwestern Compo-Board Co., 83 Minn. 25, 28, 85 N. W. 826, 85 Am. St. 440, Hendrickson v. Ash, 99 Minn. 417, 109 N. W. 830, Atlas v. National Biscuit Co., 100 Minn. 30, 110 N. W. 250, Samuelson v. Hennepin Paper Co., 101 Minn. 443, 112 N. W. 537, Lee v. Wild Rice L. Co., 102 Minn. 74, 112 N. W. 887, and Seely v. Tennant, 104 Minn. 354, 116 N. W. 648, also tend to require the submission to the jury of plaintiff’s assumption of risk.
The authorities upon which the trial court relied tend, on the contrary, to sustain his conclusion that the plaintiff should not recover. Hermann v. Clark, 89 Minn. 132, 94 N. W. 436; Anderson v. C. N. Nelson Lumber Co., 67 Minn. 79, 69 N. W. 630; Swenson v. Osgood & Blodgett Mnfg. Co., 91 Minn. 509, 98 N. W. 645; McGinty v. Waterman, 93 Minn. 242, 101 N. W. 300; Schutt v. Adair, 99 Minn. 7, 108 N. W. 811. It is apparent, however, that the tendency in this court conforms to the general opinion, viz., that it is proper to send to the jury close question^ as to which reasonable minds might reach different conclusions. See De Greif v. Northwestern Knitting Co., supra; Borchardt v. People’s Ice Co., 106 Minn. 134, 118 N. W. 359.
Finally, in the view of the law as it thus appears and in view of the facts of this particular case, we are of the opinion that the trial court was in error in holding plaintiff was precluded from recovery because of assumption of risk. There was no controversy as to any constraint upon his conduct. So far as the court held that plaintiff was as a matter of law volens to the risk, it may be conceded that no error appears. There was no direct testimony as to plaintiff’s express knowledge of the danger or appreciation of the risk. The question then resolves *281itself into whether or not the danger and the risk were so apparent that knowledge and appreciation should have been attributed to him as a matter of law.
Plaintiff, while engaged in shoveling coal, was in charge of an engine. He had some opportunity to observe that the belt and pulley were unguarded. His work was, however, done at night. The record is not ample as to how often he had opportunity to observe, nor as to the brightness of the light. On the present record we are somewhat in doubt whether plaintiff could have been held to a knowledge of the condition as a matter of law. However this may be, it is clear that his appreciation of the risk was for the jury. He had no special occasion to animadvert to the possible danger. He had done his usual work with safety under the same conditions. No peril necessarily confronted him. Was he bound to reason that if something should cause him to fall, and that if he should fall in the direction of the shaft, and that if it should at the time be in motion, he might be hurt ? So far as assumption of risk was concerned, he was bound by what would have been revealed by ordinary observation, not by exercise of reasonable care. We think that it was error to hold him to have assumed the risk as a matter of law, because, as Ryan, J., said, in Dorsey v. Phillips, 42 Wis. 583, “the consequence of acquiescence must rest upon positive knowledge * * * of the precise danger assumed, and not on vague surmise of the possibility of danger.”
3. It follows, from the reasoning as to assumption of risk, that the question of plaintiff’s contributory negligence was for the jury. It is obviously inconsistent for defendant to urge that no actionable negligence appeared on the master’s part in failing to anticipate possible injury, but that it was negligence for the servant not to have appreciated the danger. Inter alia, Swick v. Ætna, 147 Mich. 454, 111 N. W. 110, is apt. There a hand wrench slipped, and plaintiff, who was using it, slipped, whereby he fell into an unguarded pulley and was injured. It was held the defense of contributory negligence was for the jury.
Reversed.