This case was tried before a special judge in the Circuit Court of Payette county, and resulted in a verdict and judgment for the plaintiff of ten thousand dollars in an action brought by Ella McKelvey, administratrix of Theodore P. McKelvey, deceased, against the Chesapeake and Ohio Bailway Company, to recover damages for the death of Theodore P. McKelvey caused by an explosion of the boiler of a locomotive, of which said McKelvey was engineer. The defendant brought the case here.
The first question is, Did the Court err in admitting certain evidence of John Calliney ? He stated that he was a locomotive engineer, and had been running about five years, and before that had fired over three years; that stay-bolts were used to strengthen the boiler according to his idea; that he did not understand the construction of a boiler thoroughly ; that he had never had any experience in boiler-making. ITe then stated that' stay-bolts were used on the boiler sheet on each side of the fire box and all around the fire box. Then the plaintiff asked him, “ Ho you know what the effects of broken stay-bolts would be ? ” and he answered, “ I think it would have a tendency to weaken the boiler. If the stay-bolts are to strengthen the boiler, of course broken ones would have'a tendency to weaken it.”
He was then asked, “ Suppose an engine had twenty five broken stay-bolts on each side, front, corner and top, what would be the effect on the boiler ? ” and answered, “ The *503only effect I know it would have on tbe boiler would be to weaken it.”
The locomotive, that wonder of our days, is the embodiment of great mechanical genius, study, science, skill and experience. It daily stands before us who have no learning and skill in that department of human knowledge as an insoluble enigma. None but those whose eyes and brains are trained in its principles and construction can pierce its secrets and as experts adequately or reliably tell of its elements and their office. Every one may, it is true, have an opinion from observation, but it is an untrustworthy opinion, not ranking in reliability as that of one proficient in the art of its construction. This witness had simply used engines as an engineer, and had perhaps become acquainted with the practical working of some of their parts, but that is all. He shows himself that he is not expert. If this witness is to justify his evidence because he is an expert, he can not do so; for an expert must be specially skilled in his line.
The 11011-expert testifies as to conclusions which may be verified by the Court or jury; the expert to conclusions which can not be. The 11011-expert gives results of a process of reasoning familiar to every day life; the expert gives the results of a process of reasoning which can be mastered only by special scientists. 1 "Wharf. Ev. § 434, 439. A mill-wright may give his opinion as to skillfulness of work done on a mill, but not a miller. Walker v. Fields, 28 Ga. 237. Cooley, J., in Sisson v. R. R. Co., 14 Mich., 497, said, “None but an expert can give his opinion from appearance of an engine whether it could draw a train. A person having a mere vague superficial knowledge of a profession ought not to be permitted to lay down its laws.” He must have “ a special practical acquaintance with the immediate line of inquiry.” "Whart. Ev. § 439. -
Reference is made to Bird’s Case, 21 Gratt. 800, for the proposition that “all persons who practice a business or profession which requires them to possess a certain knowledge of the matter in hand, are experts, so far as expertness is required.” This is an exceptable statement, but this witness does not fall within it. He has no knowledge *504derived from study or experience in tire construction or repair or locomotives. In tire case cited a priest was held competent to speak of the law relating to marriage where he celebrated maniages, because it was within his particular line of duty and experience. And the English case relied on in Bird’s Case for the doctrine was where a hotel keeper in London, a native of Belgium, who there had carried on the business of merchant and commissioner of stocks, was allowed to prove the law of Belgium as to presentment of a note made there payable at a particular place, because he had been situated so as to call him to become acquainted with the law on that particular subject. And this is justified by the rule that it does not require a lawyer to prove foreign law. 1 Bob. Prac. 232 for full discussion of subject.
I note in Rogers on Expert Test. 237 the citation of Chicago &c. R. R. Co. v. Shannon, 43 Ill. 339, for the proposition that “the opinions of a locomotive engineer are admissible on the question whether the boiler of an engine was safe.” Reference to that case shows in the syllabus and opinion that such evidence was held admissible to prove, not that the engine was in fact unsafe, a matter that was proven by makers of boilers, but to prove that among the employes the engine was regarded unsafe and had abad reputation, for the purpose, as an item of evidence, of- bringing home to the company knowledge that such engine was unsafe, or putting them on inquiry as to its condition. The case of Sheldon v. Booth, 50 Iowa, 209, cited by same author, 260, for the proposition that it is not necessary in all cases that .a witness should be a machinist to give an opinion as to what work a machine can perform, expressly shows the witness a machinist.
What has been said as to Cnlliney’s evidence is applicable to Bartlett’s. The evidence of both which was objected to was not admissible as if they were experts. But though a witness may not express an opinion as an expert, yet he may express an opinion as to some matters, because as to those matters opinion is admissible, as for instance distance, duration and value 1 Whart. Ev. § 509 ; 26 W. Va. 798 ; but as to other matters opinion evi*505denced is not admissible, but tbe facts and circumstances must be stated, and the jury are to thence form their own opinion, because these matters are of such nature as that a jury having the facts can form an opinion just as well as the witness. "Whart. Ev. § 513 ; opinion in Taylor v. Co. 33 W. Va. 54, 16 W. Va. 260. In Welch v. Ins. Co. 23 W. Va. 288 witnesses’ opinions as to whether a quantity of wool burned could have been completely destroyed in the burning of a building of a certain size were not allowed, the court saying the jury ought to form its own opinion from the facts disclosed.
The next question is, Did the court err in withdrawing special questions ?
The defendant asked the court to require the jury, should they find a general verdict, to return answers to the following particular questions:
“First. Was the locomotive engine which exploded and killed Theodore McKelvey out of repair and defective at the time of the explosion?
“Second. If it was out of repair and defective, how long had it been so ?
“Third. Did the defendant know or could it have known by reasonable care prior to the explosion, that it was out of repair and defective?
“Eourth. Did it explode by reason of the want of repair or defects existing prior to the explosion, which the defendant knew or could have known by reasonable inspection ?
“Fifth. Did Theodore McKelvey know of the defects in said locomotive which caused the explosion ?
“Sixth. Were the defects in said locomotive engine such as made it dangerous to ran said locomotive without having it repaired, and did said McKelvey know it was dangerous?
“Seventh.. If the said locomotive engine exploded by reason of defects therein, what were the defects that caused such explosion ?”
That these particular questions were material is evident on their face, because upon them hung the liability or non-liability of the company. Their materiality is not denied. The court submitted them to the jury to be answered, and the jury retired to their room, and after remaining out over *506two hours, the jury was called in for a recess at dinner, and had not agreed in a verdict. After recess, before the jury again retired to their room, on motion of the plaintiff, the court withdrew said special questions, unanswered, against the objection of the defendant, and in a short time the general verdict was returned. The plaintiff objected to the questions when the defendant proposed them, and would justify their withdrawal on the ground that they were not proposed until the argument had been closed and instructions given to the jury, and that they were offered so late that the court had discretion to refuse them; and as it could have refused them, so their withdrawal is not error. Even if the court would have been justified in refusing them, it does not follow that after the court had once exercised its discretion by submitting the questions it could retrace it steps and withdraw them. Their very withdrawal from the consideration of the jury after it had had them before it for two hours might, likely would, prejudice the defendant’s cause, because the jury might thence infer that in the opinion of the court they were unimportant and irrelevant to the case, and that the matters, or some of them, involved in the questions constituted no impediment in the way of a general verdict. Had they been immaterial, there would be no error in withdrawing them; but beiug material, they can not be withdrawn without consent of both parties. R. R. Co. v. Posite, 14 Kan. 37; Thomp. Trials, § 2688. In Duesterburg v. State, 116 Ind. 144, it is held that relevant and material interrogatories to a jury can not be withdrawn by the party proposing them, against his adversary’s protest, and the opinion strongly asserts the position that when once submitted they must be answered.
But were these questions offered too late ? Some Indiana decisions seem to be mainly relied upon to support an affirmative answer to this question. Malady v. McEnery, 30 Ind. 273, holds that after evidence and argument, and while instructions are beiug given, it is too late to offer questions. The opinion there refers to Ollam v. Shaw, 27 Ind. 388, and there the court said that there was a rule of court requiring special instructions and interrogatories to be presented before argument, and that it was in conformity to a statute *507winch required that “when the evidence is concluded, and either party desires special instructions,” they should be written and delivered to the court, and interrogatories were assimilated to special instructions ; aadin the case of Malady v. McEnery, the former case was followed. The court said it was important, that interrogatories should be proposed in reasonable time to allow the court and the opposite party time to investigate them before the retirement of the jury.
In Fleetwood v. Dorsey etc., 95 Ind. 492, held “the court may, but need not, in its discretion, receive interrogatories to be put to the jury after the argument has begun.” In R. R. Co. v. R. R. Co., 96 Ind. 450, held that court had properly refused to submit interrogatories which opposite counsel had had no opportunity to see until he had closed his argument. In Glasgow v. Hobbs, 52 Ind. 239 interrogatories were held to have been properly refused because presented after commencement of argument before the jury. Now, it is apparent to me that this is a practice prevailing for many years in Indiana, originating with the case of Malady v. McEnery, supra, which was based on rule of court and statute as I have stated.
McLean v. Burbank, 12 Minn. 530, and some other cases are referred to as holding that the statutes involved in them confer a discretion to submit or refuse to submit questions, and that it is not reversible error to refuse them, unless the discretion is abused. These cases.do not bear on the point of time at which they must be offered. Burleson v. Burleson, 28 Texas 383, justifying a refusal to submit interrogatories after the jury had returned into court with a verdict, and Hargrove v. Millington, 8 Kans. 480, holding that they were offered too late when the verdict was ready but not announced, are referred to but are not apposite in this case. Even in Indiana the court may, after argument, submit interrogatories, though it could without committing error refuse them because coming too late; but in case it once exercises a discretion in favor of the party asking them, and submits them, treating it as an exceptional case out of a mere rule of practice, would it be supposed they could be withdrawn, without giving any reason, they being proper and material questions ?
*508Our statute fixes no time when suclr questions may be presented; no rule of the Circuit Court of Fayette on the subject appears in the record ; and this Court in Kerr v. Lunsford, 31 W. Va. 659, has said that the discretion of the Court in the submission of such questions is a reviewable discretion, not one of absolute discretion as it.is under decisions cited from Oregon, Minnesota and perhaps New York. Doubtless these interrogatories would better have been submitted before argument; doubtless a rule requiring it would be advisable. Then counsel could present to the jury the evidence bearing on each question. The same may be said of instructions; if given before argument, counsel could comment upon the evidence with special reference to the propositions of law settled by the Court in the instructions ; indeed, it may be said there is more need that instructions'should be given early, that the counsel before a discussion of the evidence may know what is the settled law of the case. But in most, if not all, the trial-courts of. this State it is common, usual practice to ask instructions after argument. This Court or the Virginia Court has never said they came too late after argument and before the retirement of the jury, unless there be a court rule. They are often given after the retirement of the jury and before verdict. Bules and practice of courts are designed to further the ends of justice and can not be inflexible, even when there is a rule or practice upon a given subject. There is as much reason for allowing particular questions to be presented after argument as instructions. No foresight can meet all contingencies. Perhaps — in fact, it is often the case that — the arguments suggest and necessitate instructions or questions which until then have not been thought of or are not necessary.
When these questions in the present case were proposed, objection was made to them, general in character, specifying no particular ground ; the Court chose to submit them; afterwards, without any reason given, they were withdrawn. We think this is error. (Luoas, Pres., thinks the questions were proposed too late and might have been refused without error, but that it was error to withdraw them.)
*509Next as to instructions. Is plaintiffs instruction two good?
“2. The court instructs tlie jury that if it believe from the evidence that the boiler of the engine of the defendant, No. 259, was in a defective and dangerous condition, and that the servant of the defendant to whom the defendant had delegated its duty of examining said boiler and finding out and repairing the defects therein had knowledge of such ■defective and dangerous condition, and having such knowledge did by his conduct and actions or words lull said Mc-Kelvey into a sense of security as to the condition of said boiler, and that by reason thereof said McKelvey was killed by the explosion of said boiler, then the defendant is answerable to the plaintiff in damages.”
In Hoffman v. Dickinson, 31 W. Va. 142, point 10 of syll. is as follows : “If a master, having knowledge of a danger of which the servant is ignorant, by his conduct, actions or words lulls the servant into a sense of security, in consequence of which he is injured, the master is liable in damages.” It will be observed that this instruction omits the word “of which the servant is ignorant.” There was evidence tending to show that the engine involved in this case was defective and in bad condition from fifty broken stay-bolts and mud which had been suffered to accumulate and become packed in the boiler, and that the agents of the company were fully informed of this condition of the engine for several months before the boiler exploded, and that the deceased engineer also knew of its condition. The company contends that if he knew of the danger, he could not be lulled into a feeling of security so as to exempt him from the opei'ation of the doctrine of contributory negligence, but that he must In the interest of his own life act as a sensible, prudent man upon his own knowledge of the dangers and serve no further, and if he did continue to run the engine, he assumed the risk.
Shearman & Bedfield on Negligence,§ 61, thus defines contributory negligence : “One who, through the mere negligence of another, suffers an injury, to the blunging about of which the want of ordinary care or the wilful wrong of himself or any person for whose act or neglect in that mat*510ter be was responsible, so far proximately contributed a>s that, but for such concurring and co-operating fault, the injury would not have happened, can not recover in any court (other than admiralty) any compensation for such injury, unless the more proximate cause thereof is the omission of the other party, after having notice of the danger to which the contributory fault of the former party has exposed him, to use a proper degree of care for the purpose of avoiding the injury.”
But the question pointedly presented by instruction two is, whether when an engine is in bad condition and the master and servant are both aware of its condition, and the master not by compulsion, but by action, words and conduct lulls the.servant into a feeling of security, and he goes on and is injured, the master is liable. It seems to me he would not be liable, for the servant has failed to act on his own knowledge, has followed the mere opinion of another instead of his own, by continuing in the service and exposing himself to danger. This Court has put the ignorance of the servant as an element of its proposition, and instruction two omits it, and tells the jury that if the engine was dangerous and the company knew it, and did by the conduct, action or words of its agents lull McKelvey into a feeling of security, the company is liable, no matter that McKelvey well knew of the danger, because it ignores his knowledge in the hypothesis. The instruction is defective because it does not conform in the respect stated to point 10 in the syllabus of Hoffman v. Dickinson.
The principle involved in this instruction two is not the principle applicable where both parties knowing the dangerous condition of a machine, the master promises repair. That is governed by legal principle applicable specially to it, and below referred to. Outside of evidence touching repair, I see no evidence of a tendency to show that defendant lulled the deceased into a feeling of security.
Is plaintiff’s instruction three good ?
“3. If the jury believe from the evidence that the death of T. F. McKelvey was caused by the explosion of the of the boiler of the defendant’s engine number 259, and that such explosion was caused by the defective condition of said boiler *511or of its stay-bolts, and that said McKelvey was notin fault and that the servants of the defendant to whom it had delegated its duty to examine and repair said boiler and stay-bolts knew of such defective condition and could have repaired the same, and even if they further believe that said McKel-vey also knew of such defective condition of said boiler and stay-bolts, yet if they believe from the evidence that he also had at the time of said explosion reasonable grounds to believe that the defendant would immediately cure said defects, then they must find for the plaintiff.”
There was evidence tending to show, that the engine was defective in the respect indicated in the instruction, and was so reported to the proper agents of the company about 28th May, 1888, and that the proper agent for the purpose promised that he would take it into the repair shop and overhaul and repair it as soon as another engine was repaired and ready to go on. the road, which occurred about the last of May, and that McKelvey knew of such bad condition for several months before the explosion which killed him on 31st August, 1888, and had requested that the engine be repaired and called attention to defects. There was, however, evidence tending to show that Me-Kelvey called attention, not to defects from which the explosion is supposed to have come, but to other defects, and it was claimed there was no promise to repair as to the former, but only as to the latter.
This instruction Ko. three contains no element of a promise express or otherwise by the defendant to repair. If the belief of the deceased that the defects in the boiler would be remedied was a mere surmise or expectation of his own based on no such promise, why, if he had knowledge of the defect causing the explosion, as the instruction assumes, would he not fall under the general rule that an employe continuing work knowing of defects in machinery and the danger to him can not recover ? (Wood, Mas. & Serv. §§ 374, 375 ; Whart. Keg. § 214; 14 Am. & Eng. Ency. Law, 845). Why should not something be required emanating from the employer to induce a belief that repair would be made, and justify the employe in risking known danger?
It was contended by the parties under the evidence that *512there was and was not a promise to repair the particular defects from which it was claimed the explosion occurred. The instruction is that under the facts in it assumed the defendant was liable, as a conclusion of law, simply from these facts — no matter whether there was or was not a promise to repair.
But if said instruction contained the element of a promise of repair on the part of the company, it would be further bad in failing to submit to the jury the question whether the defects of the engine were of such a nature and so great as obviously so constitute a danger imminent and always preseut before McKelvey such as no prudent man would risk. There is a defective machine ; the servant knows of its defects and dangerous character; the master promises to repair it, but does not; the servant knowing that it has not been repaired, continues in service using this defective machine, and is injured by reason of its defective condition. Is the master liable ? That depends upon circumstances to be passed upon by a jury. If the nature of the defects in the machine is such as to create an open imminent danger, such as no prudent man, careful of his life and limb, and none but a reckless man would risk, then the injury results from his own negligence; he has stepped into the pit open before his eyes; but if the nature of the defect is not such as to impress a prudent man with a feeling or consciousness of imminent danger, then the master is liable. This doctrine is, I think, sound in reason, and is supported by eminent authority.
Wood, Law of Master & Servant, § 378 says : “Where the servant, being aware of the dangers of the service, complains to the master, and he promises to remedy the defect, the master is liable for injuries resulting to him therefrom, and he is not chargeable with contributory negligence by remaining in the service, unless the danger is so great that a man of ordinary prudence would not remain.” And in § 380 Wood says : “But if the master promises to repair the defect, the presumption that the servant assumed the risk is thereby rebutted, and the master is liable for any injuries resulting to the servant from such defects, unless the danger was so obvious and imminent that negligence can fairly be *513imputed to the servant for exposing himself thereto ; and in all cases in which this question has been raised, even where a promise to remedg the defect is shoion, the right to recover has been made to depend upon the question whether the dan_ ger was so obvious and inevitable that gross uegligence was fairly imputable to the servant in remaining.”
In the U. S. Supreme Court, in Hough v. Co., 100 U. S. 213, the syllabus is : “If the servant of such a company who has knowledge of defects in machinery gives notice thereof to the proper officer, and is promised that they shall be remedied, his subsequent use of it in the well grounded belief that it will be put in proper condition within a reasonable time, does not necessarily or as a matter of law, make him guilty of contributory negligence. It is a question for the jury whether, in relying upon such promise, and using the machinery after he knew its defective or insufficient condition, he was in the exercise of due care. The burden of proof, in such case, is upon the company to show contributory negligence.” Justice HaRlan said in the opinion : “We may add, that it was for the jury to say whether the defect in the cow catcher or pilot was such that none but a reckless engineer, utterly careless of his safety, would have used the engine without its being removed. If, under all the circumstances, and in view of the promises to remedy the defect, the engineer was not wanting in due care in continuing to use the engine, then the company will not be excused for the omission to supply proper machinery upon the ground of contributory negligence. That the engineer knew of the alleged defect was not, under the circumstances, and as a matter of law, absolutely conclusive of want of due care on his part.”
In Dist. Columbia v. McEligott, 117 U. S. 632, the principle enunciated in the language quoted above from Justice HaRlan is approved and emphasized, and, what is applicable to this case, a charge of the lower court was held bad because it told the jury that though the employe knew of the danger from a sand bank liable to fall yet as he informed his superior of this danger and asked for a man to watch it, and was promised one, he would be relieved of the imputation of negligence in working on during the time necessa*514ry to provide a watchman. The fault was that the charge said that if the jury found that the employe did not work longer than the time necessary to provide a watchman, they could find for him, failing to submit to the jury the question •whether a man of prudence would have entered upon and continued in the work for any time, however brief, without a watchman. Assuming that the district might be liable under some circumstances, “it certainly would not be liable, if the danger which the plaintiff apprehended from the beginning was so imminent or manifest as to prevent a reasonably prudent man from risking it upon a promise or assurance by the proper authority that the cause from which the peril arose would be removed. The plaintiff had had experience in the kind of business in which he was engaged at the time of his injury. He recognized, from the beginning, the peril to which he was exposed,” says the opinion. ■
In Patterson v. Co., 76 Pa. St. 389 (18 Am. R. 412) where both parties know the defect and the master requires the seiwant to use the machinery, it is stated that if the defect in the machinery is not so great as to threaten immediate inj ury, or where it is reasonably probable it may be safely used by extraordinary caution and skill, it may be used and the master is liable; but though the servant notified the master of the defect, and he was told to use it, observing care, yet if the machineiy “is so obviously and immediately dangerous that a man of common prudence would refuse to use it,” the master is not liable, the seiwant being guilty of negligence in using it. The Court said : “ These are, however, questions which necessarily grow out of the facts of the particular case, and must be therefore referred to the jury.” See Conroy v. Vul. I. Works, 62 Mo. 35; 14 Am. and Eng. Ency. Law, 857. Upon these authorities I repeat that it should have been left with the jury to say whether the continuance in service by the deceased in running the locomotive, with knowledge of its dangerous condition, was or ivas not, under all the circumstances, an imprudent, negligent act, constituting contributory negligence on his part and forbidding recovery. But this instruction declares, as a conclusion of law, in effect, *515that if the defendant'promised rep'airs, it is liable without regard to the character of the defects or the probability or improbability of danger, or whether, all things considered, the plaintiff was or was not so negligent in continuing to run the engine that he ought not tó. ■ recover — thus withr drawing from the jury a question peculiarly proper for it'.
The words in instruction 3 “that said McKelvey was not infiaiilt,” have reference to the manner of lids operation of the locomotive, that he did not mismanage it, not to his continuing to run it. ' . " ••
Plaintiff’s instructions numbered -9 'and 10 are bad for reasons given as to No. 3. • ■ '
The feature of No. 10 that if the deceased had reason to believe that the defendant had repaired the defects has no relevancy, as there is no evidence tending to show that it had been repaired or that McKelvey had reason to so believe, and it is united with the promise to repair by the disjunctive “ or,” so that either would answer the instruction, one of which is, as above stated, objectionable, though the other-if standing alone would be good. Circumstances may be imagined where in compliance with a duty owing to the public to save the lives of passengers or other imperious necessity an engineer might be justified in remaining upon his engine.
Another objection to said instruction No. 3 is its-inconsistency with defendant’s No. 5, which was given.
(INSTRUCTION NO. 5.)
“ And the Court further instructs the jury that if Theodore McKelvey knew while he was running locomotive engine No. 259 that it was so defective and out of repair as to be liable to explode at any time while he was so using it, then no promise on the part of the defendant or its agents authorized to make such promise to repair said locomotive engine in the future would justify Theodore Mc-Kelvey in continuing to run said boiler and engine while it was still in that dangerous condition, aiid if after such knowledge he continued to run it and'it blew up'by reason of the defects so known to him and he was killed, the plaintiff' is not entitled'to recover in this case.”
Instruction 3 above given tells the jury that though Me-*516Kelvey knew of the dangerous condition of the locomotive, yet if he believed that it would be repaired, they must find for the plaintiff, while defendant’s instruction No. 5, tells the jury that if he knew that it was so defective as to be liable to explode at any time, then no promise to repair justified him .to continue running it, and if with such knowledge he continued to run it and he was killed from its explosion, he could not recover. Here is irreeoncileable inconsistency — opposite verdicts coming from substantially the same state of facts. "Which instruction must the jury follow? In McMechen v. McMechen, 17 W. Va. 683, this Court said : “ It is error to give inconsistent instructions to the jury, for it is calculated to confuse and mislead them ; it leaves the jury at liberty to decide according to the correct rule of law or the contrary, and renders it impossible for the Court to determine upon what legal principle the verdict was founded.”
In Pound v. Turk, 5 Otto. 461, Mr. Justice Miller said : “ The first and the third propositions distinctly enough declare that if the piers and booms materially obstructed navigation, the act of the legislature was no protection ; while the second as distinctly affirms that if built in the manner provided by the act giving authority, they are not liable for any injury from them. As they appear to us, these propositions given each as an independent one on that subject are necessarily contradictory, and we can not tell which of them the jury accepted as the foundation of their verdict.” See R. R. Co. v. Sawyer, 15 Gratt. 230, and Sears v. Loy, 19 Wis. 96, and p. 240 of 20 W. Va.
Even if one instruction states the law correctly, it would not cure the vice of said instructions 3, 9 and 10 ; for I find it stated in R. R. Co. v. Maffit, 67 Ill. 431, that the fact that the law is accurately stated on one side will not obviate errors in instruction on the other side. “Erroneous instructions given for one party are not in general sufficiently corrected by inconsistent explanatory instructions for the other party, but should be withdrawn.” Imhoff v. Co., 20 Wis. 362. “Error in instructions is not cured by the court afterwards instructing directly to the contrary, and so leaving the jury to digest the contradiction; the error'should *517be retracted.” Clay v. Miller, 3 T. B. Mon. (Ky.) 146. “An erroneous instruction can not be corrected by another instruction, which may state the law accurately, unless the erroneous instruction be thereby plainly withdrawn.” (Kinger v. State, 45 Ind. 518.) See R. R. Co. v. Kenrick, 40 Miss. 374; Pendleton Co. v. Stallman,) 22 O. St. 2.
There is no more difficult or important function performed by a trial-court than acting upon instructions, and frequently with but a few minutes time for deliberation. It is much easier to criticize them afterwards than to analyze them when offered. But instructions given are always rigidly scrutinized by appellate courts, because coming from the bench as expositions of the'law of the case, they wield decisive influence upon the jury. This Court has repeatedly said that an erroneous instruction is presumed to prejudice a party, andwill cause reversal, unless it clearly appears that the exceptor could not have been prej udieecl. Douglass’s Case, 28 W. Va. 297; opinion by Judge Green, Hall v. Lyon, on p. 420 of 29 W. Va.
Instruction No. 4 asked by defendant is bad, because it singles out certain facts for its hypothesis, and tells the jury that if they are found, the plaintiff’ can not recover, ignoring other facts which the evidence tended to show, namely, that defendant promised repairs to the engine.
The court added at its close the words “unless the jury believe that McKelvey had reasonable grounds to believe that such defects had been repaired, or would be repaired immediately, then they will find for the plaintiff.”
This addition is bad for reasons given above in reference to plaintiff’s instruction No. 3,.not leavingitfor the jury to say whether or not the danger was so imminent as to forbid a prudent man from continuing to ruu the engine, and having no element of promise to repair. Defendant’s No. 12 is bad and was properly rejected because it supposes only certain facts, and makes the case turn on them, ignoring the feature of the case arising out of the evidence tending to show a promise to repair. Point 14 Syll. McMechen v. McMechen, 17 W. Va. 684.
I do not see, with counsel for appellant, that point 11 of Syllabus in the Hoffman v. Dickinson case supports it.
*518Therefore the judgment is reversed, the verdict sét aside, and the cause remanded for another trial.
Reversed. Remanded.