SLOANE, J.

This is an appeal from a judgment in favor of the plaintiffs, heirs of Thomas Henry Wallis, in an action to recover damages for the latter’s death, caused from his being run into by an engine of one of defendant’s trains.

*663 The accident occurred on the evening of October 16, 1909, at the intersection of Oak Street and First Street, in the city of Oakland. The deceased was driving a lumber wagon, drawn by two horses, southerly on Oak Street. The train was approaching from the east on the tracks of defendant company on First Street. There was no load or bed on the wagon and decedent was seated on what is termed the cinch, about midway between the front and rear wheels. Just as the wagon reached a point where the driver was midway between the rails of the railroad track it was struck by the engine and the driver was killed.

The case was tried before a jury which returned a verdict for plaintiffs in the sum of three thousand five hundred dollars.

Plaintiffs’ case rests upon allegations of negligence of defendant in running its train at unlawful and excessive speed and in its failure to exercise due care in giving warning of its approach. Defendant, besides denying its own negligence, pleads contributory negligence of the deceased in failing to stop, look, and listen before attempting to cross the track.

That negligence of the defendant in exceeding the speed limit was shown by the evidence and is not seriously disputed. On the defense of contributory negligence there was no direct testimony as to whether or not the decedent took any precautions to ascertain whether the track was clear before driving upon it, but there were circumstances given in evidence which might justify an inference that the decedent did not exercise such precaution.

One of the grounds urged by defendant for a reversal, and the only one upheld by the district court of appeal, is alleged error of the trial court in admitting, over defendant’s objection, testimony of plaintiffs’ witnesses that the decedent on other occasions, not only at this crossing but elsewhere, was in the habit of stopping his team and, when necessary, going ahead to the railroad track to ascertain if any train was approaching.

The objection to this testimony was made generally on the grounds that it was incompetent, irrelevant, and immaterial, but the more specific contention of counsel for defendant was that such evidence of custom or habit to support a presumption of conduct in a "given case was permissible, if at all, only when there was no eye-witness to the fact. The language of counsel in supplementing his general objection was: “If there is evi *664 deuce of any eye-witness then it is not admissible. I think unless there is evidence that there is no eye-witness it would be very doubtful.”

After the court had overruled the objection, counsel for defendant stated: “I want the evidence to go in subject to my objection that it is incompetent, irrelevant, and immaterial and no sufficient foundation has been laid. . . . That is the only point. My theory is that it must first appear there was no eye-witness,. . . it is addressed to that only in this particular, that it has not been shown there was no eye-witness, and it is incompetent, irrelevant, and immaterial.”

We think it sufficiently appears that defendant’s objection to this evidence was thus limited to the condition that the evidence introduced was inadmissible if there was direct evidence of eye-witnesses to the fact. It was upon this assumption that the ruling of error wras made by the district court of appeal. Accepting this theory of the law, we doubt if the foundation appears in the evidence here to exclude this testimony. No eye-witness testified directly on the subject or made reference to it. Only two witnesses saw the decedent as he approached the crossing. Neither, of these saw him for more than a moment. Mrs. Borba, who lived on Oak Street, a block south of the crossing, opened her door and looked out just as what was probably the crossing whistle of the train sounded at a point shown to be some seven or eight hundred feet east of the crossing. At that time she saw the decedent driving on Oak Street toward the crossing at some point between her house and First Street. She re-entered the house, closed the door, and heard the continued whistle of the engine. J. F. Finn, the engineer of the locomotive, testified that just after he had finished blowing the crossing whistle he saw the heads of the horses coming into his line of vision at a point about ten feet from the north rail of the track at the crossing, and that he thereupon began to sound the warning whistles ; that the team continued toward the track until the collision happened. We do not agree with respondents’ interpretation of the testimony of Mrs. Borba, that she saw decedent just after he was past the intersection of Oak and First Streets. She distinctly states that she saw him on Oak Street between First and Second Streets “right by the red building, ’ ’ which was between these streets. That would *665 leave a considerable space where he was not seen by either witness. It is obvious that an interval elapsed between the moment when Mrs. Borba saw the decedent and when the moving team came into the engineer’s line of vision. There is room for inference, however, that decedent did not stop to look and listen in that brief interim, and yet he might have done so, and there was no eye-witness to testify on that point. The time and place for so stopping would have been as he came into First Street, where he could see up and down the railroad track. He might have just started his team- after an instant’s halt to glance up and down the track, when the engineer first saw the heads of the horses. If ever a cause was entitled to whatever presumption arises from habits of caution, to overcome an inference from adverse circumstantial evidence of negligent conduct, it arises under the facts of this case.

We do not understand that the authorities which uphold the admissibility of this class of testimony only in the absence of direct evidence base the condition of its admission upon an entire absence of other evidence as to collateral facts that may uphold an inference as to what happened, but upon the absence of direct testimony of any eye-witness that the thing did or did not occur. In this case there was no such direct testimony.

The law governing this class of evidence is perplexingly inharmonious. The weight of authority, however, seems to uphold its use under the conditions stated, that there is an absence of satisfactory testimony of eye'-witnesses as to the fact in controversy, while other decisions and authorities consider it legitimate evidence without such condition. This limitation upon the introduction of such testimony seems rather illogical. If the fact of the existence of habits of caution in a given particular has any legitimate evidentiary weight, the party benefited ought to have the advantage of it for whatever it is worth, even against adverse eye-witnesses; and if the testimony of the eye-witnesses is in his favor, it would be at least a harmless cumulation of evidence to permit testimony of his custom or habit.

The only authority in this state directly in point is that of Craven v. Central Pac. R. R. Co., 72 Cal. 345, [13 Pac. 878], In that case the plaintiff was suing for damages resulting from a fall from a train. The controversy was as to *666 whether she was thrown off by the jolting movement of the ear, or fell in an attempt to step off while the car was moving. Testimony was admitted as against testimony of eye-witnesses to the actual occurrence, that she was in the habit of alighting from cars while they were still in motion. The court in this connection says: “A sensible man called upon out of court to determine whether or not a certain person had on a certain occasion carelessly jumped off a moving train of cars and finding the direct testimony as to the matter conflicting would naturally and properly give some weight to the fact that the person was in the habit of alighting from ears in that manner, and the consideration of such a fact in cases resembling the one at bar has been frequently sanctioned in court. The evidence at least had some legal tendency to show that plaintiff’s conduct at the time of the injury was such as defendant ascribed to her.” If evidence of habit is to be recognized at all, this unconditional rule would seem to be the logical ground for its admission. It is contended by appellant that later decisions of this state, though not directly calling in question the case above cited, are in substantial conflict with it, citing Towle v. Pacific Imp. Co., 98 Cal. 342, [33 Pac. 207]; Cunningham v. Los Angeles Ry. Co., 115 Cal. 561, [47 Pac. 452] ; Langford v. San Diego Electric Ry. Co., 174 Cal. 729, [164 Pac. 398]; Carr v. Stern, 17 Cal. App. 397, [120 Pac. 35],

These are all cases in which negligence was charged against defendant corporations through acts of their servants or employees. In each instance it was sought to prove or disprove negligence, by the general character of the servant as to his skin or incompetence, prudence, or carelessness in the certain line of employment involved. Such evidence was excluded. This class of evidence is clearly distinguishable from that establishing a custom or habit of doing some particular thing in a particular way. Because one is a skillful workman in a given occupation does not tend to disprove negligence in some specific act, but if the question in controversy is whether he did the thing at all or his manner of doing it, his custom or habit regarding that particular matter would be significant. Most of the text-writers seem to recognize the competency of such testimony. Ruling Case Law (vol. 10, p. 955, sec. 127) thus states the doctrine: “A habit of doing a thing is naturally of probative value as indicating that on a particular *667 occasion the thing was done as usual, and if already shown as a definite course of action is constantly admitted in evidence, ’ ’ but recognizes the limitation upon such evidence by adding: “The weight of authority seems to be against admitting evidence of general conduct under proven circumstances, to show conduct of the same kind under similar circumstances on a particular occasion, when there were eyewitnesses of the occurrence.”

In Thompson on Negligence (vol. 6, sec. 7140), it is said: “Where there are no witnesses to the accident, evidence of the habitual care of the deceased is often admitted on the question of his exercise of due care at the time of the accident. ’ ’

Mr. Greenleaf says (1 Greenleaf on Evidence, 16th ed., secs. 14j and 14k) : “A habit of doing a thing is naturally of probative value as indicating that on a particular occasion the thing was done as usual and if clearly shown as a definite course of action is constantly admitted in evidence.”

In Elliott on Evidence (vol. 1, sec. 172) the author says that “evidence of personal habit is often of some probative value and is frequently admitted; but such evidence should not be admitted when to admit it would violate the character rule, and it cannot ordinarily be proved that a person did or did not do a certain thing at a particular time by showing that he acted in a certain way under similar circumstances at other times.”

We find in Wigmore on Evidence (vol. 1, see. 92) the following strong indorsement of habit evidence: “Of the probative value of a person’s habit or custom as showing the doing on a specific occasion of the act which is the subject of the habit or custom there can be no doubt. Every-day experience and reasoning make it clear enough. ’ ’ The author, however, very pertinently points out the distinction in the probative value of casual and infrequent acts along given lines of conduct and such a constant and repeated course of action as would indicate a fixed habit. He adds: “It is easy to see why in a given instance something which may be loosely called habit or custom should be rejected because it may not in fact have sufficient regularity to make it probable that it would be carried out in every instance or in most instances. Whether or not such sufficient regularity exists must depend largely upon the circumstances of each ease. ’ ’

*668 When we turn to the decided cases there is such a diversity of facts and confusion of rulings that it would be impracticable to analyze them or try to reconcile them other than in a general way.

In Parsons v. Syracuse etc. R. R., 133 App. Div. 461, [117 N. Y. Supp. 1058], it was held in a case similar to this that evidence was not admissible for the purpose of showing that a decedent who was killed while crossing a railroad track used care at the time of the accident, by introducing evidence of specific instances of care on other occasions, although no eye-witnesses were present. In the opinion, Mr. Justice Carter said: “A man may be careful on one occasion and may be careless on another. The circumstances at one time may ¡be such as to induce prudence, while they may not at another.”

In that case, however, it was merely offered to show that the decedent had been seen on several isolated occasions to stop and look and listen at this crossing. There may have been many intervening occasions when he failed to observe that precaution.

In the case at bar the testimony of the witness Glass was that he had worked in connection with the decedent, not only in hauling lumber over this particular crossing (which had only been for a few days), but in teaming elsewhere when they hauled hay and grain across a railroad track. The witness was asked if he was familiar with the practice of decedent when he approached a railroad- track, and if he knew decedent’s habit in that respect. In making objection to the testimony counsel for defendant stated that he did not make the point that the witness had not had sufficient observation but that his objection was to evidence of decedent’s custom in the matter. In testifying the witness states as to decedent’s habit “that he would get off and look to see if there was a train coming,” and that for the few days they had been at work over this crossing “he always stopped, looked to see if the train was coming.” So the evidence here was not to show casual acts, but a consistent habit.

In Smith v. Boston & Me. R. R. Co., 70 N. H. 53, [85 Am. St. Rep. 596, 47 Atl. 290], another case of a railroad crossing accident, evidence was upheld to the effect that the deceased “for many years always checked his horse and looked and' listened for the train at such crossing.” The supreme court *669 in the opinion says: “Cate’s uniform habit of slackening the speed of his horse to a walk at the Waukewan crossing and looking and listening for the approach of a train before attempting to pass the crossing tended to show that he did so on this fatal trip. It was substantial evidence of the exercise of care on that occasion,” citing a number of New Hampshire cases to the same effect.

This subject is quite fully discussed, with citation of numerous authorities, pro and con, by the court of appeals of New York in the case of Zucker v. Whitridge, 205 N. Y. 50, [Ann. Cas. 1913D, 1250, 41 L. R. A. (N. S.) 683,'98 N. B. 209]. In the trial of that case evidence had been introduced of the habitual exercise of caution on the part of the deceased killed by a street-car at a crossing, as tending to prove freedom from contributory negligence at the time of the accident. The admission of the testimony was held prejudicial error but the ruling was made on the ground that there were eye-witnesses of the occurrence, and expressly withheld any opinion as to the admissibility of such testimony where there was no direct evidence on the subject.

The following cases are cited in the foregoing opinion as disapproving this class of evidence. Attention is called to this list of decisions as indicating how generally the question arose on testimony of casual and disconnected acts or upon evidence of general reputation for carelessness or incompetence not going to a specific act or circumstance. “In Wisconsin, to show that the person injured ‘was an habitually careless man’ (Propson v. Leatham, 80 Wis. 608, 612, [50 N. W. 507]); in Pennsylvania, that the deceased ‘had made a practice of jumping from the elevator while in motion’ (Baker v. Irish, 172 Pa. 528, [33 Atl. 558]) ; in Connecticut, that the intestate ‘was a careful and prudent driver’ (Morris v, Easthaven, 51 Conn. 252); in Illinois, ‘that the deceased was in the habit of jumping on trains’ (Peoria & Pekin Co. v. Clayherg, 107 111. 644, 648) ; in Iowa, in a case where there was some evidence that the deceased was asleep in his buggy when he drove on the track, ‘that he had been found asleep in his buggy on other occasions’ (Dalton v. Chicago etc. R. R. Co., 114 Iowa, 257, 259, [86 N. W. 272, 273]); in Maine, ‘that, in the opinion of those who knew the deceased well, he was a cautious and careful man,’ no witness having seen the accident (Chase v. Maine Cent. R. R. Co., 77 Mo. 62, 65, *670 [52 Am. Rep. 744]); in Massachusetts, specific instances of want of care in the engineer in his business of running trains within three months of the injury, before or after (Robinson v. Fitchburg & Worcester R. R. Co., 7 Gray, 92, 95); also ‘previous specific acts of negligence on the part of defendant’s engineer known to its superintendent’ (Conners v. Morton, 160 Mass. 333, 335, [35-N. E. 860]).”

[1] We do not find enough in these authorities, or in the rule laid down in Towle v. Pacific Imp. Co., Cunningham v. Los Angeles Ry. Co., Langford v. San Diego Elec. Ry. Co., cited in this opinion, to discredit the ruling in the Craven case, supra, so far as it applied to evidence tending to show a consistent uniformity of action along given lines indicating a fixed habit, and applied to the question as to whether some specific act involving the same trait was done or was not done in a certain manner.

The probative force of evidence of habit, though not passed upon as a rule of evidence (that question being expressly withheld), is recognized by Mr. Justice Shaw in the case of Bresee v. Los Angeles Traction Co., 149 Cal. 131, [5 L. R. A. (N. S.) 1059, 85 Pac. 152], In that case the plaintiff was injured while riding as a passenger in a carriage which was run into by a street-car. Evidence of the driver’s habitual carelessness in crossing in front of moving cars, and of plaintiff’s knowledge of his customary want of care, was held admissible as tending to show contributory negligence on the part of plaintiff in failing to keep a lookout or to warn the driver to watchfulness. This, of course, could only be upon the theory that plaintiff, having knowledge that the driver was careless in this particular matter, was bound in common prudence to anticipate the probability of a repetition of his careless conduct on this occasion. The rule admitting evidence of habit to rebut or support conflicting testimony of eye-witnesses, as adopted in this state in Craven v. Central Pac. R. R. Co., 72 Cal. 345, [13 Pac. 878], is cited and approved by the appellate court of Indiana, under a similar state of facts, in Pittsburg etc. Ry. Co. v. McNeil, 34 Ind. App. 310, [69 N. E. 472]. To the same effect is Denver Tramway Co. v. Owens, 20 Colo. 107, [36 Pac. 848]. The supreme court of New Hampshire, in the case of Parkinson v. Nashua & L. R. R. Co., 61 N. H. 416, says: “Although it is quite generally held elsewhere that evidence of other specific in *671 stances of negligence on the part of either party is not competent because raising a collateral issue, yet in this state a different rule prevails and has become established in cases where the evidence is conflicting, and it is held to be competent to show that the party charged with negligence had performed or omitted the same act in the same way before, as tending to show that he did or omitted the act at the time in question, on the ground that a person is more likely to do a thing in a particular way as he is in the habit of doing it.” Gillett’s Indirect and Collateral Evidence, section 68, states; “The weight of authority favors the view that where the direct evidence shows that an act was done or omitted, it is competent to prove that a custom existed prior to the time to do or not to do such an act, as such evidence legitimately tends to uphold the theory of one of the contending parties. ’ ’

Other decisions upholding this character of evidence upon one ground or another are: State v. Manchester etc. R. R. 52 N. H. 528; International etc. R. R. Co. v. Kuehn, 2 Tex. Civ. App. 210, [21 S. W. 58] ; Shaber v. St. Paul M. & M. Ry. Co., 28 Minn. 103, [9 N. W. 575]; Suffer v. Dry Dock etc. R. R. Co., 53 Hun, 629, 5 N. Y. Supp. 700; Bower v. Chicago etc. R. R. Co., 61 Wis. 457, [21 N. W. 536]; Cleveland etc. R. R. Co. v. Moss, 89 Ill. App. 1; Cox v. Chicago & N. W. Ry. Co., 92 Ill. App. 15; Mayer v. Milwaukee St. Ry. Co., 90 Wis. 522, [63 N. W. 1048]; Meier v. Shrunk, 79 Iowa, 17, [44 N. W. 209].

Most of these cases involve habits of negligence, but, of course, habits of caution and prudence are of equal, if not greater, evidentiary weight. Professor Wigmore, in his work on Evidence (volume 1, section 97), even questions if acts of negligence can be anything more than casual or occasional. Experience, however, will not support that theory. It may be said, however, that prudence and caution in matters involving exposure to danger may more readily become habitual from the incentive always present to follow the path of safety.

The main argument which seems to prevail with the authorities excluding habit evidence is the fact that it introduces into the trial collateral issues of fact, of which the opposing party has had no notice. This condition, however, arises in numerous instances in nearly every trial on the introduction of probative facts, collateral in their nature, and of which the ultimate facts pleaded give no inkling.

*672 We think the evidence excepted to in this case was properly-admitted. The weight of this class of evidence, of course, depends upon the nature of the act and the fixity of the habit, •hut that is a question which can properly be left to the jury under proper instructions.

Men are accustomed to weighing, and acting upon, such evidence of conduct in their every-day affairs. Recognition of its force is seen in the adoption by the law of many presumptions based on common experience, such as the presumption that a letter properly deposited in the mails has reached its destination in due time. The presumption that the regular customs of business have been followed, and the express declaration of our code of a presumption “that things have happened according to the . . . ordinary habits of life (Code Civ. Proc., sec. 1963, subd. 28).”

It may be doubted if the jury in this case was seriously influenced in its verdict by evidence of decedent’s habit of caution, but we have felt justified in devoting so much attention to the question because of the general uncertainty of the law in other jurisdictions and the apparent confusion as to the rule in California arising out of the suggested conflict between the Craven case and other later decisions of this court herein cited to that point.

There are only two grounds of error presented. The first, which we have already ruled upon, is the admissibility of the evidence as to decedent’s habits of caution. The other is alleged error of the trial court in giving to the jury an instruction submitting an issue on the last clear chance.

[2] Without passing upon the objection, raised for the first ■time on motion for a hearing in this court, to the sufficiency in form of the instruction given, to fully present the doctrine of the last clear chance, we are compelled to the conclusion that there was not sufficient evidence to justify submitting this instruction to the jury, and for this reason the judgment must be reversed.

This instruction may have governed the jury in arriving at a verdict for the plaintiff, and, in the absence of evidence to support it, must be held fatally prejudicial.

The jury may, under the evidence, have found the decedent guilty of contributory negligence. A verdict for plaintiff after a finding of such negligence could not be supported, excepting under the doctrine of the last clear chance. [3] *673 The rule is well settled in this state that in order to make the defendant liable, notwithstanding contributory negligence of the plaintiff, the defendant must not only be aware of the danger in time to avert it, but must also know or have reason to believe that the plaintiff is oblivious of the danger and is in a position where he cannot extricate himself from it. (Arnold v. San Francisco-O. T. Rys., 175 Cal. 1, [164 Pac. 798]; Basham v. Southern Pac. Co., 176 Cal. 320, [168 Pac. 359]; Young v. Southern Pac. Co., 182 Cal. 369, [190 Pac. 36].) The only evidence on this point is that of the engineer himself. He testifies that when he rounded the curve some four or five hundred feet from the crossing he saw the heads of decedent’s horses some ten feet from the track. He immediately sounded the warning whistle. He doubtless believed, and had a right to believe, that the driver would at once stop his team and not attempt to cross the track. As soon as he saw that the team was still approaching and the driver not looking toward him, but in an opposite direction, he put on the emergency brakes with full force, and used every means shown by the evidence to be available to stop his train. He was justified in assuming that at the distance from which he sounded the warning, and with the headlight and noise of the approaching train, the driver would be warned of his danger. The team was moving at a walk and, under ordinary circumstances and so far as appeared here, could have been pulled up within a space of two or three feet. Even though the train had been running at a lawful rate of speed, it would have been a. new act of contributory negligence for the driver to attempt to cross ahead of it after the warning was sounded.

The remarks of Mr. Justice Shaw in Ms opinion, in Basham v. Southern Pac. Co., supra, apply aptly to the ease before us. Commenting on the last clear chance doctrine and the fact that the train whistle was sounding, the bell ringing, and the noise of the moving train considerable, he says: “The horses were in a slow walk. The lines were in Coffey’s hands. The horses were under his control. He could have come to a standstill in a second. Under all these circumstances it would be extremely unreasonable for anyone to suppose that he was unaware of the approaching train, and did not intend to stop before reaching the track as he could easily have done, and as is the frequent custom. When the fireman did, nevertheless, begin to entertain a fear that Coffey was inattentive, he *674 at once gave the urgent order to the engineer to stop. It cannot be said either that he did realize the danger sooner or that, in reason, with his knowledge, he had cause to realize it sooner. Hence the last clear chance doctrine never came into operation.” Such was the situation in this case.

The judgment is reversed.

OIney, J., Wilbur, J., Lennon, J., Angellotti, C. J., and Shaw, J., concurred.

Rehearing denied.

All the Justices concurred.