The finding of facts upon which judgment is founded, contains a statement in detail of inferences produced, in whole or in part, by weighing evidence and the credit to be given witnesses, and also of the conclusions drawn from these inferences. The former are called facts, as denoting adjudicated facts which can only be retried by an appellate court having jurisdiction in the trials of such facts. This court does not have appellate jurisdiction of that nature; the Superior Court is the court of last resort for that purpose, and its adjudication of such facts, in the exercise of original or appellate jurisdiction, is the end of litigation; unless in the process of adjudication it has violated some rule or principle of law. Styles v. Tyler, 64 Conn. 324; Thresher v. Dyer, 69 id. 404, 408. The alleged failure to determine such facts correctly, is improperly assigned in the appeal as error, and cannot be considered. The request of counsel, for the certification of testimony in support of such claimed errors, is an abuse of the provisions in respect to certifying evidence. Thresher v. Dyer, supra.
The latter, that is, conclusions drawn from such facts, are also called facts, but with a much broader signification, including all issues that the line separating the province of the jury from that of the judge in a jury trial practically leaves to the jury. The word fact, used in this broad sense, does not accurately denote matters not reviewable by this court. In defining facts as denoting those questions practically within the province of a jury, we are controlled not only by established practice, but by the constitutional pro*174vision forbidding any violation of the political “right of trial by jury.” In defining facts as denoting questions not reviewable by this court, we are controlled by “ the primary distinction drawn by the Constitution, between the jurisdiction original and appellate of courts for the full trial and adjudication of causes, and the jurisdiction of a court of last resort for correcting errors hr law which may have intervened in the course of a trial.” Atwater v. Morning News Co., 67 Conn. 504, 526. “ The true distinction as drawn under our system of jurisprudence, in connection with this provision of the Constitution, between facts that the trial court must find from the testimony, and the application of principles of law based upon such facts,” includes “ questions of law ” as distinguished from “ questions of fact ” in a jury trial, but is not fully expressed by that distinction. While the jurisdiction of this court may be affected in its practical operation by existing procedure or practice, the jurisdiction itself, “ is coextensive with the judicial power of the State in all matters wherein legal principles, that is, rules of law or principles of equity, appear to have been erroneously or mistakenly determined by a trial court.” Styles v. Tyler, supra, 454. The limitations of procedure formerly existing, in connection with the practice followed in view of that procedure, prevented in some instances the full exercise of our jurisdiction; and the conclusions of trial courts,—because not presented in such manner as to be reviewable under existing procedure and practice,—have been spoken of in language appropriate enough for that purpose, as questions of fact. But whenever the record before us has legally presented all the facts found by the trial court as the basis of its judgment, and the conclusion drawn from those facts has been plainly erroneous, and such error has been lawfully assigned, we have uniformly held such conclusion, although for some purposes it might be called a question of fact, to be, quo ad the jurisdiction of this court, a question of law; i. e., it is reviewable.
When the finding of facts states evidence, so that the conclusion must be reached by weighing evidence, the find*175ing is essentially a report of evidence and not a statement of facts adjudicated, and the question of legal inference from facts that may be involved, is irregularly presented. Corbin v. American Mills, 27 Conn. 274, 278. In Bloodgood v. Beecher, 35 Conn. 469, the judges were equally divided upon the question whether, upon the finding of facts in that case, the intention of a mortgagor to prefer the mortgagee to his other creditors, could be drawn by this court as a conclusion of law; Hinman, C. J. and Park, J., holding that it could not, and Butler and Carpenter, Js., holding that it could. The case was decided by the second vote of the Chief Justice. In Mead v. Noyes, 44 Conn. 487, the trial court, in an action of replevin, spread upon the record the facts from which it drew its conclusion that the plaintiff was the owner of and entitled to immediate possession of the property replevied; and, upon motion in error, this court reversed the judgment because the conclusion from the facts found was an error in law. In Hayden v. Allyn, 55 Conn. 280, 289, Judge Loomis, speaking for the court, laid down the broad principle that whenever in trials to the court the judge has fully weighed the testimony and passed upon the credit of witnesses, and specifically found as the basis of his judgment the inferences produced by the testimony, so that the evidence “ had exhausted itself in producing the facts thus found, nothing remained but for the court in the exercise of its legal judgment to draw its inferences from the facts; ” and “ in such a case the conclusion of the court can always be reviewed by the appellate court. An erroneous conclusion is an error of law and not an error in an inference of fact.” This principle was deliberately affirmed in Tyler v. Waddingham, 58 Conn. 375, 386, and applied to a special finding of facts from which was drawn the conclusion that the plaintiff, at the time of making a contract with a partnership, elected to give exclusive credit to a single partner. In Ward v. Ward, 59 Conn. 188, 197, the principle is recognized as established, although its application in that case is treated with some subtilty. , The principle may at times be misapplied, but a mistake of this kind cannot shake its au*176thority. It is not only supported by the true ratio decidendi of a long line of decisions, but is embedded in the very structure of our system of jurisprudence. Settling the credit of witnesses; weighing evidence; ascertaining the truth from conflicting testimony or incongruous evidential facts;—this is the peculiar province of, and under our system within the exclusive jurisdiction of, trial courts, and a mistake in the inference produced by such means is an error in fact; when such facts are adjudicated, a mistake in drawing the legal inference, i. e., in applying.the law to the facts found, is an error in law.
The application of this principle has been hampered and its meaning somewhat obscured, through inadequate and uncertain methods for bringing into action the jurisdiction of this court. Sometimes the conclusion of a trial court from conceded facts is so clearly right that practically no question is presented, and in such cases we have said that the conclusion is one of fact properly decided; yet if in such cases the conclusion, instead of being clearly right, had been a palpable non sequitur, we would have reviewed it as a question of law, unless the question were irregularly presented. But more frequently the alleged error in a conclusion from conceded facts has been irregularly presented; either through mistake in making up the record, or through defect hi methods for obtaining a record which should properly present the question; and in such cases we have said that the conclusion is one of fact not reviewable. Such indeterminate use of the phrase “question of fact,” or “conclusion of fact,” must be taken in connection with the circumstances of each case, and cannot be treated as precisely distinguishing those conclusions which are to be treated as facts in respect to the jurisdiction of this court, nor as modifying the settled principle that the unwarranted conclusion of a trial court in drawing its inference from the special facts which, in compliance with existing law of procedure it has found for the purpose of drawing that inference, is essentially an error in law.
For many years, and especially since 1879, the legislature *177has endeavored by means of various statutes to modify the law of procedure, so as to require a trial court to place upon record the special facts on which its judgment is founded, and to enable this court to exercise its full jurisdiction in reviewing the legal judgment of a trial court in drawing its inference from conceded facts. Such legislation has considerably enlarged the facilities for exercising the jurisdiction of this court. The effect of this legislation has been the subject of frequent consideration, and the main general conclusions reached are summarized in Thresher v. Dyer, supra, and Winsted Hosiery Co. v. New Britain Knitting Co., 69 Conn. 565, 575. In the latter case we say: “The judgment or ultimate conclusion of a court upon the special facts in issue, as ascertained from the evidence and settled by the trier, is a conclusion of law, and as such reviewable by this court; and this is true whether such facts are settled by a special verdict of a jury or by a special finding of a judge.” And referring to the changes in procedure: “We think that the result of this legislation is, that in cases tried to the court the jndge is now authorized, and upon request required, to find and state in a special finding the facts adjudicated by him in reaching his ultimate conclusion, including all specific facts which, when so adjudicated, must determine the ultimate conclusion and subordinate conclusions involved therein, by force of settled rules and principles of law. The judgment rendered on such an adjudication of facts is simply the voice of the law declaring the legal effect of the facts adjudicated.”
These considerations must control the disposition of the claim made in the ease before us, that the conclusions we are asked to review are conclusions of fact and not of law.
It appears that the defendant operated a single track railroad, that train 474 (an extra train) following train 1411 (a regular way freight) ran into the latter train, which had stopped to attach some freight cars standing on a siding, and in the collision one Jerry Nolan, the plaintiff’s intestate, was killed. Every special fact from which the trial court inferred the liability of the defendant for the injury is found, includ *178ing the rules and regulations adopted by the defendant for safely operating these trains. The finding shows the neglect of one Hall, the rear brakeman of train 1411, to obey the rules provided for the protection of his train under such circumstances, and that the collision might not have happened if the rules had been obeyed; but also shows that Nolan and Hall were fellow-workmen; and the trial court, therefore, holds that the defendant is not liable for this neglect. The question of liability is not otherwise affected by the relation of master and servant. The complaint does not allege that this relation existed between Nolan and the defendant, and claims nothing by reason of that relation; but charges the defendant with negligently conducting itself in the management of said trains; and the only allegations, so far as is material under the finding, of acts constituting such negligent conduct, arc: “ by failing to give proper telegraphic information to the conductor and engineer of each of such trains, relative to the position of the other of said trains,” and “ by failing to exercise proper supervision of the running of said trains,” so that train 474 ran into train 1411 and, as a result of the collision, said Nolan, who was rightfully on the former train, was killed. The trial court draws from the special facts found, the following inferences : that the rules and regulations of the defendant “ did not sufficiently provide for this emergency (i. e., the operation of the trains in the manner as found), and for the reasonably safe operation of trains 474 and 1411; . . . that the defendant was negligent in not providing for the emergency (¿. e., the operation of the trains as found) and for the safe operation of trains 474 and 1411, by special orders and instructions, in addition to the general rules; . . . that the operation of these trains under these general rules alone, under the circumstances of this case (i. <?., the special facts as found), was not a suitable and safe method to operate these trains; . . . that the traindespatcher did not exercise ordinary care in not issuing special orders for the reasonably safe movement of these trains; . . . that the injury to plaintiff’s intestate resulted from the combined negligence of rear brakeman Hall and of the defend*179ant; that the defendant did not exercise ordinary care in the movement and operation of trains 474 and 1411.”
These inferences are reviewahle as conclusions of law. It is true that in stating the inferences, the trial judge says, “ I find as a fact,” etc., but this is immaterial. It is the duty of the trial judge to find, as facts within the peculiar province of a trial court, those inferences which are controlled by the weighing of evidence and the credit given to witnesses ; it is also his duty to find his conclusions drawn from these inferences of fact, which in a certain sense are also findings of fact; but they are conclusions reviewahle by this court, and the name given them does not alter their intrinsic character of conclusions reviewahle for error in law.
We think the court below erred in reaching these conclusions. They are all drawn in support of the claim set up in the complaint, that the defendant violated the legal rights of the plaintiff’s intestate, because it failed to exercise proper supervision of the running of said trains, and because it failed to give telegraphic information to the conductor of each train, relative to the position of the other. The supervision of the trains (unless as involved in the failure to give telegraphic information) is to be found in the rules for the movement of the trains. These rules are before us ;■ they are substantially the same as those in use by about ninety per cent of the steam railways of this country, and the trial court finds that “ for the general movement and operation of trains these rules are the best and safest general rules yet devised by the best railroad talent of the country.” There is nothing in the record that calls for a review of this finding ; for the purposes of this case we assume the conclusion to be correct. These rules cover the movement of regular and extra trains ; they provide for special orders for starting extra trains; they require the train-despatcher to give telegraphic information of the meeting place of such trains, and of trains moving in the opposite direction, as well as of regular trains off the regular time; but they do not require him to inform by telegraph trains moving in the same direction, of their relative position, and for that purpose to keep in mind *180the position of all such trains so as to decide as they approach each station whether there is a likelihood that a rear train may overtake and, if the rules are not obeyed, run into the forward train. On the contrary they are drawn upon the theory that such telegraphic supervision of trains moving in the same direction, in view of all the conditions involved in operating a single track road, tends rather to lessen than increase the safety secured by the rules adopted relative to the movement of such trains. The soundness of this theory has received judicial sanction. Enright v. Railway Co., 93 Mich. 409, 413; Illinois Central R. Co. v. Neer, 31 Ill. App. 126, 134. In the latter case some self-evident reasons are given. We think in this respect the rules of the defendant do not violate its legal duty, and that their compliance or non-compliance with that duty, under a given state of facts,—notwithstanding many cases hold that such question must be submitted with instructions more or less precise, to a jury, by force of the law defining the province of a jury,—is essentially an inference of law; and when drawn by a trial court, is reviewable by this court. So far, therefore, as the proper supervision of trains 474 and 1411 depended on the adoption of adequate rules for the safe operation of these trains, the only legal inference from the facts found is that the defendant did not fail to exercise proper supervision of the running of these trains.
The failure to give trains 474 and 1411 telegraphic information of their relative positions, is found as a fact, and the conclusions of the trial court are simply an inference from this fact, in connection with other facts found, that the defendant by this means had violated the legal rights of the plaintiff’s intestate. In oth^r words, the inference is one of legal liability, and affirms that the law which defines the duties of railroad corporations and the rights of persons lawfully on their trains, imposed upon the defendant the duty of giving such telegraphic information, and gave to the plaintiff’s intestate the correlative right to have such information given. The validity of this inference is really determined in the disposal of the claim that the rules of the *181defendant did not fulfill its legal obligation in the supervision of these trains, under the circumstances of the case, i. e., the facts as found. The rules do not require such information to be given; because giving such information in all instances involving like conditions, would tend to danger rather than to safety. A railroad corporation in operating its road as a quasi public highway, is engaged in a business dangerous to human life, and is exercising for its private benefit a franchise granted by the State on condition that it transport such members of the public as have lawful occasion to use the highway, with every precaution for their safety that public policy, as fixed by legislation or recognized by adjudication, requires; for these reasons the law imposes upon the corporation a duty to use such precautions; that duty it owes to each person lawfully on its trains, and this independently of any special duty arising from a contract of carriage or employment. McAdam v. Central Ry. & Elec. Co., 67 Conn. 445, 447. Each person lawfully on a train has a right to the protection of such precaution, and is entitled to damages for any injury done him in violation of this right. In holding that the rules for the supervision of trains under conditions like those attending the trains in question, fulfilled the legal duty of the defendant, in this respect, and that public policy does not require these rules to prescribe giving telegraphic information in the manner claimed, we necessarily hold that giving such information to these particular trains on the day specified is not a precaution required by public policy and is not a duty imposed upon the defendant by law, and therefore a failure to give such information did not violate any legal right of the plaintiff’s intestate.
But the claim is made that the conditions attending trains 474 and 1411 differed essentially from the conditions in general attending trains moving in the same direction, and differed so essentially as to constitute an exceptional case or emergency unprovided for by the general rules, and of a character so peculiar to itself as to throw upon the defendant, or its vice-principal, the train-despatcher, the duty of acting in *182view of all these exceptional circumstances as a man of ordinary prudence should act; that the trial court has found that the train-despatcher did not so act; and that such a finding can never be reviewed by this court. This, in truth, is the real ground of the judgment, and the very gist of the question before us.
We think the conditions attending trains 474 and 1411 did not differ essentially from those in general attending trains moving in the same direction, and did not create an exceptional case or emergency unprovided for by the rules ; and that the finding of such emergency by the trial court is an inference from the special facts found, reviewable by this court. The conditions in general attending trains moving in the same direction under the rules, without telegraphic information of their relative position, include : all trains, regular and extra, made up in all ways, even to a single engine; trains off their regular time, way freights being commonly behind time ; stopping places for trains which are used only occasionally and not at regular intervals ; trains moving at all times of day and night, and in all conditions of weather and atmosphere; trains moving at various rates of relative speed. The special facts found, from which apparently the inference of an exceptional case or emergency is drawn, are the following: train 474 consisted of an engine pushing a snow plow; train 1411 was upwards of an hour behind its schedule timetrain 1411 stopped to attach three freight cars at Kent Furnace, which is merely a siding where freight trains stop occasionally and at irregular intervals ; the rear train when in motion moved at a faster rate of speed than the forward train; the day was very cold and the snow plow threw snow considerably, rendering it difficult for the lookout stationed on the snow plow to see ahead, but just before the accident the plow was not throwing much snow and the lookout could see. We think the conditions shown by these special facts, considered by themselves or in connection with all the special facts found, are within the conditions in general attending trains moving in the same direction; do not constitute an exceptional case or emergency unprovided for *183by tbe general rules; and did not throw upon the defendant, or its train-despatcher, the special duty of keeping the conductors of -those trains informed by telegraph of their rela^tive position. No other inference can be legally drawn from the facts. A conclusion from conceded facts drawn by a court in the exercise of its legal judgment, and controlled by the assumption that two and two make five, is just as truly an error in law as if it were controlled by the assumption that an ordinary breach of contract calls for exemplary damages. In the present case the error is not of this palpable land; as a matter of first impression there is room for hesitation and doubt; but upon full consideration it seems to us that the legal inference is clear, and that in reaching its conclusion the court below, unless influenced by the error involved in its inference that the rules adopted were inadequate, has failed to apply to the facts those settled principles of sound reasoning whose recognition and application are termed the soul of the law.
The plaintiff relies mainly upon the recent case of Sprague v. New York & N. E. R. Co., 68 Conn. 345. In that case the plaintiff’s intestate was a servant of the defendant, and was killed in a collision between his tram and the train moving in the opposite direction, caused by the misconduct of the conductor of the opposite train. The complaint charged the defendant with liability on account of its violation of its duty as master in employing the conductor at fault to run this train, knowing him to be incompetent. Any other violation of duty on the part of the defendant was inseparably entangled with this, the real ground of the action. The main question considered was one of law, as to the use of the admissions of a demurrer overruled upon a hearing in damages. Aside from this our decision turned wholly upon the question whether, upon the facts as stated, we could find error in the conclusions of the trial court, that the conductor was incompetent, that the defendant knew or ought to have known that he was incompetent, and with this knowledge placed him in charge of the train where his incompetency caused the injury. Here the inference of legal liability from *184the conclusions of fact as stated, was unquestioned. The finding shows that these conclusions were not inferred from subordinate specific facts found, but depended in part upon the weighing of evidence and credit given to witnesses. This sufficiently appears in the opinion, but more clearly in the record which is not printed in the report. We held that such conclusions were plainly conclusions of fact within the province of the trial court. No question was presented as to a conclusion from specific facts found, where such conclusion is clearlya non sequitur. The defendant claimed that the question of its violation of a legal duty by reason of the insufficiency of its railroad management, was a pure question of law. In referring to this claim the opinion holds the question to he irrelevant to the case in hand; and that if the principle claimed is sound, it cannot control emergencies which no system of rules can anticipate and provide for. It is in this connection that the language specially relied on, relative to emergencies, is used. The language must be read in this connection and in view of the circumstances of the case then before us. So read, the language does not establish any principle inconsistent with the views expressed in the present case.
After stating the specific facts found and the conclusions from those facts (which statement is strictly a part of the judgment specially setting forth the facts on which it is founded), the finding states that the defendant claimed, as a matter of law, “ that upon the facts in evidence the defendant, having operated its trains under suitable rules and regulations and having properly equipped said trains, had performed its entire duty toward plaintiff’s intestate, and the law imposed no higher degree of care than that exercised by it;” and that the court did not pass upon this claim, “ except to find as a matter of fact from the evidence submitted, that the defendant did not exercise ordinary care in the operation of these trains, and that ordinary care required additional precautions by way of special orders, to those provided in the rules, for the reasonably safe operation of these trains.”
*185The trial court did not err in refusing to sustain this claim of law, in the terms hi which it is phrased; but it did err in inferring from the specific facts found, that it was the legal duty of the defendant to give the telegraphic information mentioned; and for the purposes of review this inference is not a conclusion from the evidence, which, the finding shows, had exhausted itself in producing the conclusions of fact from which the inference was drawn. This error is not raised by the assignment specifying error in the refusal to sustain the claim of law, but is raised by the assignments specifying the reasons why the facts specially set forth do not support the judgment founded upon them. The only function in this respect, of that part of the finding reciting the claim of law made, is to show that the questions as to the legal inferences from the facts fomid were distinctly raised at the trial and decided adversely to the appellant’s claims. The real position of the plaintiff is that the trial court has found the defendant failed to exercise ordinary care under the circumstances of this case; that the legal liability of the defendant consists in negligence ; that the failure to exercise such care is negligence; that negligence is a question of fact entirely within the province of a trial court; and therefore the law determining the legal liability of the defendant, was properly found as a fact within the exclusive province of a trial court to determine.
"We have already indicated the mistake involved in this position; but it is a mistake so often made and so readily fallen into, through the use of words expressing different ideas without due attention to the particular idea the word as used is intended to express, that we deem it advisable to restate the ground of our decision with special reference to the confusion of ideas that leads to such mistakes.
We are dealing with a practical question of procedure, i. e., upon the process or record before us, what are the alleged errors that this court can review? The answer is briefly and broadly expressed in the saying: “ Errors in law can be reviewed; errors in fact cannot.” As we have seen, “ fact ” is a word of many meanings, and the saying is deceptive unless we keep in mind the particular meaning *186attached to the word as here used. We have explained that “ fact,” as here used, denotes those conclusions reached by the trier directly from sifting testimony, weighing evidence and passing on the credit of witnesses,—conclusions which are not within the jurisdiction of this court, and cannot be reviewed or retried on appeal, whatever the process may be; and that it does not denote those inferences drawn by the trial court from the facts ascertained and settled by it as described,—inferences which always involve to some degree the application of rules and principles of law to adjudicated facts, and which may be reviewed whenever the legal process properly presents an alleged error, and we can see that the inference as to which error is alleged is in truth one of this nature. The word, as here used, may possibly have a little broader meaning in some exceptional cases, but this is immaterial to the purpose in hand.
We have also explained that the word “fact,” as here used, must be distinguished from the same word when used to denote those matters within the province of a jury. In the latter sense it often denotes the whole question of legal liability which, by the law of jury trial, must in certain cases be settled by a general verdict; and so far as it may be used in connection with jury trials to denote inferences from evidence, as distinguished from inferences from adjudicated facts, it is of necessity used with an imperfect or uncertain meaning. The trial judge, in instructing a jury upon inferences of law, cannot ordinarily know what the inferences from evidence may be, and in all such cases must give his instructions hypothetically, and is limited in this by the consideration that the law of jury trial forbids his giving instructions upon the law in such manner as in truth to invade the province of the jury in drawing inferences from evidence. Whereas in a trial to the court, the judge first adjudicates all the facts, i. e., inferences drawn from the evidence, and puts upon record these facts. The inference from these facts, of legal liability or of the conclusions essential to legal liability, present to us, in a lawful process for that purpose, the questions of law free from the limitations *187which hamper a trial judge in instructing a jury upon the same questions. And it must be remembered that while there is in theory a distinction of a fundamental nature between the judicial function exercised in ascertaining facts from evidence, and that exercised in applying to conceded facts the rules of law for the purpose of inferring logical conclusions and legal liability, yet the two functions are essential to the one judicial act of passing judgment, and the precise line of distinction has little, if any, practical importance, except when the exercise of judicial power is divided between two branches of one court, as in a trial by jury, or is divided for the purpose of establishing an appellate court with a jurisdiction pertaining to the judicial function exercised in applying to conceded facts the rules of law, as in the correction of errors in law by this court. In such cases the necessary line of division, although based on a natural distinction in the character of judicial power exercised for the different purposes, may become, as in this case, a practical question of procedure.
Again, it must be remembered that the rule defining the erroneous conclusions of a trial court which this court can review, is and must be of universal application. There is not a different rule for each class of actions. Actions in tort as well as actions in contract, actions to recover damages resulting from intentional wrongs as well as to recover damages resulting from wrongs not intentional, are all subject to the same rule; and actions of tort or contract wherein questions of negligence arise, are subject to the same rule. The mistake in question is largely due to overlooking this, in the confusion caused by a failure to distinguish, when the word “negligence” is used, the precise one of its many meanings intended to be expressed through the particular use. Negligence is frequently used to express the cause of action where a party seeks redress for injury from an unintentional wrongful act. In the nomenclature of the common law this is called a cause of action enforceable by the action of trespass on the case; “ trespass,” as signifying a passing over or beyond our right, i. e., a transgression or wrongful *188act (Bao. Abr. ad verb.); “ trespass on the case,” as signifying a form of action devised to cover all cases where an actionable wrong is claimed under the particular circumstances of the case stated. Negligence, so used, is a comparatively modern term of aft, denoting a class of actions grouped under one head for the purpose of study and treatment. It covers the injuries received; the act done, positive or negative; the proximate relation of the act to the injury; the legal rule of liability applicable to the case stated, and the application of that rule. The numerous definitions of “ actionable negligence ” attempt to state briefly and comprehensively the conditions essential to all of this group of actions. So used, “ negligence ” has no more relation to the question before us than if the meaning denoted were expressed in the older language of the common law,—those causes of action where the appropriate remedy is an action of trespass on the case to recover damages for an unintentional injury. As related to this meaning and sometimes confused with it, “ negligence ” is used to denote the conception of moral blame or fault imputed to a person legally liable for the consequences of an unintentional act. This conception of the moral blame which should make one liable for the consequences of his acts, is the foundation of our law of tort, and, in a wider sense, of our whole system of jurisprudence. In the very early stages of English law this element of moral blame, in cases where the conception seemed too subtle for the times, was overlooked, and the innocent cause of a condition resulting in injury was held liable for damages to the innocent victim. But the true theory that moral blame is an essential element of legal wrong, was soon fully recognized, largely through the potent and unacknowledged influence of the Roman law. “ Dolus (i. e., injuria quam quis sciens volensque commisit) aut culpa (i. e., omnis protervitas, temeritas, inconsiderantia; desidia, negligentia, imperitia, quibus citra dolum, cui nocitum est) ”, as used in the Lex Aquilia, and as approximately expressed in our common law by the terms “ intention or negligence,” describe the two phases of the moral blame or fault essential to make the one causing damage legally liable. 2 Austin’s Juris. 109.
*189The body of our law has been developed through the application of this conception to an infinite variety of facts or events. It has its origin in the accepted law of ethics, and the unfolding of its meaning as a legal maxim is confided to the court. Indeed, the highest function of a judge is exercised in holding to a steady and definite course in the application of such a maxim to ascertained facts, whether their conjunction be novel or not. In the exercise of such a function our system of jurisprudence develops under the control of fixed maxims and prior authorities. It is this control of the element of discretion, inherent in the application of a general principle to each new state of facts, that distinguishes the justice administered by a common law judge from that of an oriental Cadi. Whether this primary conception of moral blame is vague or not in itself, the principle of legal liability which in theory it justifies, is certain; the legal liability is not wholly coincident with the moral blame, it depends upon the law (however derived); and that law consists mainly in rules of public policy defining and limiting the rights and duties pertaining to person and property. These rules are ordinarily intended to square with the highest theory of ethics; an intention not always realized. But it is . the laAv and not the theory which determines legal liability. No principle of legal liability which is in the legal sense uncertain,—-which is derived, not from settled maxims and authority that must govern every case, but from the unrestrained Avill of a single man or a dozen men which can govern no other case,—is a legal principle. Ubi jus incertum, ibi jus nullum. Negligence, Avith the meaning under discussion, is a convenient form for denoting the conception of legal blame in a certain class of causes of action, for the purpose of investigating the theory underlying the law which determines liability in such cases. It is a term more useful to the jurist in seeking the theoretical sources of the law, than to the judge in applying positive law to conceded facts. Used in this sense, negligence is largely synonymous Avith the law of liability in this class of actions (although it hardly covers the whole law of liability in such cases), and denotes a prin*190ciple which is essentially compound, i. e., one which includes a variety of maxims and rules, any or all of which may be involved in its application to particular cases. Tins use of the word might not' have produced the confusion it has led to, were it not that one of the rules covered by the term “ negligence,” and one most frequently invoked in actions of tins character, involves the use of negligence with the primary meaning of the word, and so the considerations peculiar to this single rule have been easily commingled with those applicable when the same word is used, not with its primary meaning, but as a term of art signifying the whole principle of legal liability. Negligence thus used is scarcely distinguishable from heedlessness, and may be shown in omitting to do an act or in doing an act. When the law says a person whose conduct is negligent (in this sense), the other conditions of liability existing, is liable for damages caused by his conduct,—the meaning of “negligent,” like the meaning of any material word used in stating a rule of law, is a question of law to be settled by the courts whose province it is to declare the law; and where the rule is intended to apply to an indefinite variety of events, the meaning of the rule in its application to each group of events is a question of law for the court. So it has been held in cases of “reasonable ” notice to take a deposition; Sharp v. Lockwood, 12 Conn. 155, 159; Sing Cheong Co. v. Wing, 59 id. 535, 543; “ reasonable ” time for presenting a note for payment; Lockwood v. Crawford, 18 Conn. 361, 372; “reasonable ” or “ diligent” search for a lost document; Kelsey v. Hanmer, 18 Conn. 311, 317; and the principle may be Illustrated in other ways. This is elementary law; were it not so, the growth of any stable system of jurisprudence would be far less juaeticable. No difficulty can arise where the group of events calling for the application is clear and certain. But where, as sometimes happens, the group of events in respect to which the meaning of a word must be declared, are of a nature insusceptible of clear statement, then, while the function of ascertaining facts from testimony and that of applying the law to the facts, should be separately exercised by the trial judge, *191yet any adequate separate statement by him in the record of the results, i. e., the facts as ascertained and the application of the law to those facts, may be impracticable. Dilemmas of a similar nature account for most of the troublesome questions relating to the instruction of a jury by the court. In a trial to the court, where the two functions are united, there is no difficulty so far as the disposition of that case is concerned ; but inasmuch as a statement of the facts is impracticable, the application of the law is not apparent, and the case cannot serve as a precedent, and caimot be reviewed for error in law by an appellate court.
In a large number of cases the material question is whether the ascertained conduct of a person is in law negligent, in the sense last indicated. Often the ascertained conduct cannot be formulated into a series of adjudicated facts; it consists of a single impression produced in the mind of the trier by the whole mass of relevant testimony. That impression cannot be stated in words; it is a mental view which language is not luminous enough to photograph. Is that conduct negligent in law ? Clearly the conclusion of the trier that it is or is not, must be, quo ad the trier, a conclusion of law; whether it be considered as an inference from many facts, or as the declaration of the meaning of a rule of law in relation to a single fact, it is the inference of a legal wrong from ascertained fact, and that is essentially a conclusion of law. Equally clearly the conclusion of the trier must be, quo ad the power of this court to review his inference, a conclusion of fact. Not because the inference is not one of law, but because the fact or facts from which the inference is drawn cannot by any process known to the law be transferred from the mind of the trier to the mind of the appellate court. When such a question comes before us we say we will not attempt to review the conclusion; it is one of fact; and this, notwithstanding the trial judge has conscientiously tried to give hi his finding a word picture of a mental impression that cannot be painted in words. He has not succeeded; he cannot succeed; the precise legal inference he has drawn must remain unknown; it cannot, there*192fore, be reviewed, and is practically a conclusion of fact. Now this condition may arise in any kind of action, and is not peculiar to actions of negligence, using the word as a term of art describing a certain class of actions.
The mistake of the plaintiff arises, mainly, from confusing an ascertained fact—where the fact is conduct claimed in law to be negligent,—incapable of transference from the mind of a trial court to that of an appellate court, with the legal inference drawn by the trial court from that fact, and which cannot be reviewed because the appellate court cannot have the fact from which the inference is drawn, before it, and then applying the practical result of such a condition to the inference of liability in all actions classed under the term actions of negligence. And this confusion arises mainly from the fact that the word negligence, as used in respect to the same general subject, has entirely distinct, although closely related, meanings, dependent on the particular purpose for which it is used. A similar confusion arises from the use of “ ordinary care,” which is used sometimes in reference to actual conduct under circumstances, as found by the trier, incapable of being so formulated in a finding as to state the facts really adjudicated, and show the inference actually drawn, and sometimes as indicating the rule of legal liability under any given state of facts; as used in the latter sense it may denote the ultimate ground of every cause of action.
In the present case we hold that the inference of legal liability drawn by the trial judge, is reviewable, because the events upon which the cause of action arises are of such a nature that they can all be fully and clearly found by a trial court as adjudicated facts, and have been so found and properly appear before us in the record; that the conclusion of the trial court of an “ emergency,” which required of the defendant a line of conduct appropriate only to the single transaction in question, is reviewable, because this subordinate conclusion is drawn from adjudicated facts fully and clearly found in compliance with existing law of procedure, and the trial court in reaching the conclusion—unless the law defining such an emergency was misapplied—-has plainly *193mistaken those rules of sound reasoning whose observance is essential to the validity of an inference from admitted facts.
In stating our conclusion we have not sought to lay down a binding rule, but have merely endeavored to explain, as far as the infirmities of language will permit, the test of our decision in this case. We believe that such a test if applied, with discrimination as to the circumstances of each case, to the whole line of our decisions where the question of review-ability of errors alleged in inferences of trial courts has arisen, will reconcile those cases as resting on a ground substantially the same in all. This question was very carefully considered in Farrell v. Waterbury Horse R. Co., 60 Conn. 239. The opinion of Judge Toeeance in that case touches cum acu the root of the difficulty; and has been our main guide in subsequent decisions. The underlying principle was there considered with special reference to contributory negligence, where the ascertained fact could not be stated in the finding, and so the inference of the trial court could not be reviewed, and remained for all practical purposes a conclusion of fact. In this case we consider the same principle with special reference to the question of legal liability, where the adjudicated facts can be and are fully set forth. And the task of testing the principle is now of a somewhat different nature and easier than it was then, by reason of further changes in procedure and our decisions upon the effect of such changes, in enlarging the facilities for the full exercise of our jurisdiction in correcting the erroneous inferences of a trial court.
We allude to some matters suggested by the record merely to avoid any implication from a failure to mention them. It is claimed that one or two of the facts stated, as found from the evidence, are in reality nothing but deductions from the rules or from other facts found, and as such erroneous. The facts referred to have too slight a relation to the controlling question to call for particular comment. As a whole the finding is exceptionally full and clear ; and furnishes proper facility for its correction in the matters complained of, were such correction necessary.
It is claimed that the court erred in its conclusion that the *194injury resulted from the combined negligence of the brakeman Hall and the defendant. The Massachusetts case of Hayes v. Western R. Corp., 3 Cush. 270, apparently supports this claim. As we hold that the court erred in finding the defendant negligent, otherwise than by reason of the fault of its brakeman, the question is not material and we do not pass upon it.
The court finds that “ the employees of the defendant occasionally violated this rule (the one whose violation by the brakeman caused the collision) and the conductor in charge of 1411 knew this.” If there were other facts in connection with this, clearly showing that the rule was not enforced through the neglect of the defendant, a different question would arise in respect to its liability. The duty imposed by law upon the defendant is not fulfilled by merely adopting adequate rules. The law imposes upon it a duty in respect to the enforcement of rules necessary for the protection of the public. This question was considered in Gerrish v. New Haven Ice Co., 63 Conn. 9, 16, and is discussed in Railway Co. v. Hammond, 58 Ark. 324, 332. It is not raised in this case.
The judgment xuider review is controlled by the conclusion that the collision in which Nolan was killed, was caused by the legal negligence of the brakeman Hall in conducting the business of the defendant; and that the defendant escapes liability solely on the ground that Nolan, the victim of the wrong, as well as Hall, through whom the injury was done, was its employee. The rule which produces such a result is too firmly established as law by a multitude of decisions to be now reversed or seriously modified by any exercise of the power vested in courts.
There is error; the judgment of the Superior Court is set aside and the case remanded for assessment of nominal damages.
In this opinion the other judges concurred.
Note:—The rule (of co-employee) is one deduced by a process of analogy from decisions rendered under a state of society very different from that of to-day; and the crystallization of such analogies into a *195binding rule is of comparatively modern origin. It first appeared in England in Priestley v. Fowler, 3 Mee. & Wels. 1; and in 1850 was applied to the employees of railroad companies, in Hutchinson v. York N. & B. Ry. Co., 5 Exch. 343. In 1841 it was formulated in South Carolina; Murray v. Railroad Company, 1 McMullan, 335; and in 1842 in Massachusetts, in the leading case of Farwell v. Boston & W. R. Corp., 4 Metc. 49. The very able opinion of Chief Justice Shaw in the last case, has largely dominated the law on this subject during the past fifty years, and contains the most plausible statement that can be given of the grounds supporting the public policy which compels a workman entering the service of a master to assume the whole risk of any injury that may be done him by the master through the misconduct of a fellow-servant. The vigorous language of this statement, however appropriate it may have been at that time, has a touch of grim irony when read in the light of existing conditions in the employment of labor. In England, in 1880, the rule was changed by the “ Employers’ Liability Act,” and is now practically abolished as to large classes of workmen, by the “Workmen’s Compensation Act” passed during the present (1897) year. The rule has been dealt with by legislation in several of our sister States. It was first formally recognized in this State in Burke v. Norwich & W. R. Co., 34 Conn. 474, 479, with a strong protest against the sufficiency of the grounds for a principle deemed too firmly settled in other jurisdictions to be differently treated here. In Darrigan v. New York & N. E. R. Co., 52 Conn. 285, the application of the rule was somewhat modified, and possibly cases may arise where the legitimate exercise of the duty of the court in applying established principles to novel conditions, may involve some limitations of its apparent reach. But the evil is too deep seated to be remedied by judicial action ; it needs radical treatment through wise legislation.
W. H.