The Bretton Woods road on which the accident occurred runs northerly from its intersection at right angles 'With a highway called the Portland road. The railroad runs generally parallel with the Portland road and crosses the Bretton Woods road at a distance of 52 feet from its southerly rail to the northerly line of the Portland road. For such distance the Bretton Woods road is nearly level. On the east side of the Bretton Woods road and near its intersection with the Portland road was a sign on a post and bearing the words “Speed limit 15 miles per hour on straightaway, and 10 miles per hour on curves. Slow down at crossing.” Below the sign was another one with the single word “Caution” on it. The first sign was about on a level with the observation of a person in an automobile and directly faced the driver as he entered the Bretton *212Woods road after coming from the west on the Portland road. On the north side of the crossing and on the west side of the road a post carried a sign headed “private crossing” and with a warning in smaller letters following that persons used the crossing at their peril and must not rely upon train signals. The sign was from 6 to 8 feet high, faced travel from the south, and the heading could be plainly read for a distance of several feet south of the crossing. At a point 35 feet south of the crossing the track could be seen towards the east for about 1,400 feet. From the Portland road the view of the track to the east of the crossing was obscured and the view of the situation on the Bretton Woods road was not clear from the Portland road until their intersection was nearly reached.
The automobile after proceeding easterly on the Portland road turned to the left as it entered the Bretton Woods road. In making the turn its speed had been slowed to wait until a car from the other direction had passed. On the Bretton Woods road there were no other cars south of the crossing at the time. Beyond the crossing a car on the westerly side of the road was coming and came to a stop. The roadway was 20 feet wide at the south side of the crossing and increased in width to its southerly end where the width was over 70 feet.
The train came from the east and its speed was from 15 to 18 miles an hour until after the emergency brakes were applied. It was the first time the driver of the automobile and those with him had been in the White Mountains, and he testified that until the collision he was unaware of the crossing, that he proceeded on the Bretton Woods road at not over 5 miles an hour until he increased his speed when he thought the car beyond the track had stopped to let him pass it, that after entering the Bretton Woods road he first saw this car and then watched it all the time until the collision, and that his car was struck by the train near the front right wheel just as his speed was increasing and at about the center of the crossing. At a speed of 5 miles an hour the automobile could have been stopped in 4 or 5 feet. The overhang of the train was about 18 inches. The accident occurred on a clear day late in August an hour or two before sunset.
One ground of the defendant’s alleged negligence is that the crossing was insufficiently protected with warning signs. It is claimed that if the crossing was on a highway, there should have been a cross-arm sign such as is generally found at highway crossings, and that if the road was not a highway, even then the warning signs *213were inadequate to protect the travel which the defendant knew passed over the crossing.
This ground of negligence cannot sustain the verdict unless evidence of the causal effect of the careless act or omission to produce the injurious result is shown. If the evidence is conceded to be sufficient to find lack of care in such respect, yet unless the evidence also tends to show the force of the carelessness in producing the injury, the claim must fail (McGill v. Company, 70 N. H. 125, 129; Bennett v. Company, 76 N. H. 180), and a finding that an ordinary cross-arm warning sign or some more prominent signs than there were would have been seen and heeded with action averting the collision must be a reasonable inference from the evidence.
While the relation between cause and effect need not be established with absolute certainty (Upton v. Company, 81 N. H. 489, 492), yet if a cause is conceded or conclusively proved, then to find another cause as to which the probabilities are equal is conjecture. Boucher v. Larochelle, 74 N. H. 433. And while inferences within the limits of reason are of fact, the question whether a given conclusion is within such limits is one of law. The determination of the border line between conjectural and reasonable inferences usually presents a problem difficult of application to the facts of a concrete case. Inferences of the causal connection between known antecedent facts and known subsequent facts are in general based on stronger probabilities than inferences connecting assumed prior conditions with assumed subsequent conditions. The conclusion of what would have happened under a situation differing from the actual one is ordinarily more conjectural than the conclusion whether what did happen is due to the antecedent facts shown. By reason of the special circumstances and features usually attending a given set of facts, authority and precedent cannot be of much aid or value. The sufficiency of the evidence on any issue must from the nature of the question be determined by general principles of reason and logic in application to the evidence, and rules of law are little if any guide. In a large way each case must be judged on its own merits.
One issue was the public or private character of the road where the crossing was. If the crossing was on a highway and if it may be assumed that the cross-arm sign in general use would have been prescribed by the public service commission (P. L., c. 249, s. 14), its location on either side of the road and on either side of the crossing would appear to have been a full compliance with the statute. And *214if the crossing was on a private road, the usual highway sign would clearly constitute reasonable warning facilities for this crossing when a train approached it at no greater speed than 15 to 18 miles an hour and when, as hereafter appears, both a customary highway whistle at the statutory distance for highway crossings and a station whistle at about 200 feet from the crossing were blown, and with the view there was of the track in the direction from which the train came. The situation differs from that in the case of Collins v. Hustis, 79 N. H. 446, where the speed of the train was at least three times as great as the speed here.
In Folsom v. Railroad, 68 N. H. 454, negligence in not having a flagman at a crossing in the compact part of Manchester over which trains ran at a speed of from 35 to 40 miles an hour was conceded, and hence called for and received no consideration. And in Gage v. Railroad, 77 N. H. 289, what is said about a possible need of lights is only dictum since under the facts no lights were needed. But if there may be crossings or movements of trains over them calling for more than the statutory requirements, it cannot be maintained that under a public or private character of the road here considered more than a usual cross-arm sign wherever reasonably located was called for.
It is necessary for the plaintiffs to establish the propositions, first, that the driver would have seen a cross-arm sign had there been one, second, that seeing it, he would have anticipated that a train might be at hand, and third, that he would have acted upon such anticipation and in such manner and so seasonably as to have avoided the collision. The driver’s conduct was without question a cause of the collision, whether or not the conduct was negligent. His inattention to most of his surroundings was such as to be almost incredible. But assuming the truth of his testimony, without knowing what he was to encounter he turned into the Bretton Woods road not seeing the caution signs directly facing him and nearly on a level with a horizontal observation, at .a slow speed proceeded the straight distance of 50 feet to the planked crossing of standard width and with a width of roadway of 20 feet at the nearer rail without noticing the crossing or the warning sign beyond it, when the crossing was in the direct range of his vision and the warning sign nearly so as he kept his eyes on the car that stopped at the side of it, and did not see the train’s approach or realize its presence until after the impact although there was a. clear view of it all the way after he reached a point about 35 feet from the crossing. Some point is made of the fact that shortly before reaching the Bretton Woods road he had been told on inquiry the route *215to take for Ms destination, but was not told to look out for the crossing. As the inquiry related only to the route, it is not reasonably to be inferred that he was put off his guard and given to understand that no crossing would be met.
In extreme favor for the plaintiffs it is at least as probable that an ordinary cross-arm sign would not as would have seasonably awakened the driver to the presence of the crossing. While it is true that he testified in effect that he would have seen such a sign, this was no more than his own conclusion, and as such an inference could not reasonably be made, the fact that he made it did not change its quality from conjecture to probability, and hence did not justify the jury in making it. In his obliviousness of all save a few limited features of a very obvious and open situation, it cannot fairly be said that more probably than not he would have seen an ordinary crossing sign in such reasonable location as it might have.
Since it is wholly problematical and speculative to say whether an ordinary crossing sign would have been seen by the driver, the ground of the defendant’s negligence in failing to maintain protection at the crossing fails.
As has been suggested, other cases in which the sufficiency of evidence on an issue has been passed upon are of little aid. But as argument has been made that the situation in Jones v. Railroad, ante, 73, is closely parallel with that here, it may be said that in that case, while it was held that if a warning whistle had been given, it might be found that it would have aroused the inattentive driver and led to action on his part averting the collision, yet the driver there was aware of the crossing just ahead of him although not of the oncoming train, and notice of the latter fact in connection with his appreciation of the former made some likelihood that he would take seasonable action. Here the driver was aware of neither fact and notice of the crossing would not be enough to indicate that he would have acted differently and seasonably unless he also knew of the train’s approach. Furthermore, in the Jones case the argument that a locomotive whistle sounded within 112 feet of a traveler would have attracted his attention, was not, under the circumstances, unreasonable, whereas; here the contention that the driver would have taken notice of another motionless feature of the landscape if it had existed, is without logical foundation in view of the extraordinary blindness which he manifested towards those objects which in fact confronted him. The distinctions if close are yet substantial.
Regarding the sufficiency of the evidence to show that usual warn*216ing signals were not given by the train in coming to the crossing, in the doubt whether the plaintiffs have waived the issue or still insist upon it, it has been considered. The rule that the burden of proof is not met when there is no other evidence to support it except of witnesses not listening to hear and which is contradicted by evidence of a positive character (Morier v. Hines, 81 N. H. 48, 53, and cases cited; Phillips v. Railroad, 81 N. H. 483, 484), is here properly invoked. Although the driver and his sister who was in the car testified that they heard no whistle or bell, they were admittedly giving no heed in listening for a train. On the other hand three trainmen and three other witnesses testified that they heard both the crossing and station whistles, one trainman testified to hearing the crossing whistle but did not recall about the station whistle, and another witness testified to hearing “one whistle.” This state of the evidence leaves the plaintiffs without proof of negligence in any failure of the train to give the usual signals.
Another issue of negligence is whether a warning whistle which would have led the driver to stop his car before reaching the crossing should have been given after the fireman regarded the situation as dangerous. Since the driver did not realize there was a crossing directly ahead of him, it seems extremely doubtful if he would have thought such a warning was meant for him, but if it may be found that he would, the evidence does not support a reasonable finding that he would have had time seasonably to act upon it.
The fireman was at the left-hand side of the locomotive cab and the engineer at the right. The fireman testified that he first saw the car when it was about 25 feet from the crossing and traveling at 5 or fi miles an hour, that it was a little more than its length from the crossing when he warned the engineer to stop, that the train was then .about 125 feet from the crossing and proceeding at a rate of 15 to 18 miles an hour, and that until the car nearly reached the crossing he thought it was going to stop and apprehended no danger until it then increased its speed. The testimony of the driver, corroborated in a general way by that of his sister, as to the position of the automobile when its speed was increased was even less favorable to the plaintiffs than that of the fireman. The driver testified that when the car across the track stopped he increased his speed to go by it, but that he “hadn’t really gone ahead any” before the accident occurred, and in cross-examination he testified that the car “must have been almost on the tracks, because we were almost in front of this [other] automobile when we speeded up.”
*217If the train was 125 feet from the crossing when the car’s speed was increased, at the train’s speed of 18 miles it would take about 5 seconds for it to reach the crossing with no allowance for reduction of speed from putting on the emergency brakes. Such an estimate would call for the rejection of all the other pertinent testimony bearing on the usefulness of a warning whistle and would mean that the car’s speed was increased at a point well over 40 feet from the crossing or that if the car was then nearer the crossing, it would have cleared the crossing ahead of the train. Considering the necessary conflict of the estimate with the other positive evidence on the subject including that of the fireman himself and taking into account the uncertainty of judgment any person who is in motion, however experienced he may be, has in accurately measuring distances and speeds in connection with objects moving in another direction, the estimate of the distance of 125 feet must be rejected as so unreliable as to give only a conjectural basis for any conclusions therefrom. While a jury may ordinarily adopt any positive evidence it may see fit, it may not do so when the only conclusions from it are necessarily inconsistent with other facts which are conclusively established or which must be established to justify the verdict. Under all the evidence, whatever the train’s distance from the crossing, the car was not over 15 feet from the crossing when the fireman first thought there was danger.
He instantly warned the engineer to stop. As to the suggestion that he might have given instead a warning to whistle, negligence in not giving such an alternative warning would not appear to be capable of inference. While the issue of care in an emergency is one of fact the same as in situations generally, yet the facts that the call for taking action is sudden and that the excitement and lack of time to think naturally affect the normal action of judgment are to be given consideration as a matter of law. Carney v. Railway, 72 N. H. 364, 372. And wdien the action is taken so instantly that it is instinctive so that the action precedes thought, it cannot be said that any choice of action has been made. Instinctive action when there is no time at all to think and choose cannot be said to be careless unless the actor is shown to be unfit to act in an emergency calling for such action. Carelessness means wrong thinking or failure to think in connection with action, and instinctive action when there is no time to think cannot of itself be called negligence. When instant and instinctive action to meet danger is called for, human limitations are such as to make the action taken, whatever it may be, the only care that can be given, aside from taking care that the actor be a careful person.
*218In the same way the engineer was called upon to act instantly and instinctively. He did not have the benefit of the fireman’s view of the danger and the time taken to appreciate what the warning signified must have been simultaneous with the time taken in acting. There was no time for any orderly mental process making his action dependent upon his conclusion from the warning received, but action had to keep up with thought if not to precede it. In the absence of evidence that the engineer had careless habits, carelessness in the selection of methods of response to the danger is not to be found in view of the lack of time in which to make the selection. In Jones v. Railroad, supra, the engineer saw the danger and it was held that a sufficient although brief time might be found to have been given in which to choose between different ways of meeting the emergency.
But even assuming that it might reasonably be found that the fireman should have given a warning to whistle or that the engineer without such warning should have whistled instead of applying the brakes, it is only conjecture to say that a whistle would have prevented the collision. According to all the testimony on the subject, the driver first saw the other car when it was not over 30 feet from him and before it stopped, so that with all fair allowance for the distance it then went before it stopped, for its distance from the crossing when it stopped, and for the width of the track, he must have been not over 15 feet from the southerly and nearer rail of the track when it stopped and he increased the speed of his car. As already stated, the fireman testified to a distance not so great, while the driver and his sister testified that the car was practically at the crossing before there was any increase of speed. Assuming the speed was increased when the car was about its length from the track, the increase was the danger sign to the fireman. If the speed was not increased until the car reached the track, then its failure to slow down when not over 15 feet from the track was the danger sign. There is no evidence the fireman apprehended danger until then. With allowance of a second from his appreciation of the danger to the engineer’s action, the car was within 7H feet of the track or 6 feet of the line of the train’s overr hang when the engineer’s action was taken. If the action had been in blowing the whistle instead of applying the brakes, some time would have been required for action to be taken by the driver, thus suddenly warned and brought to a realizing sense not only that a track and train were near but also that the warning was intended for him. With the car’s speed until then being under increase no stop could have been seasonably made. Taking into account its speed and *219the distance within which it could be stopped, even so small an allowance as half a second for the interval from hearing the whistle to setting the brakes of the car would have had no saving effect. If a whistle under some views as to speeds and distances might have resulted in preventive action, yet it is only a possibility and falls short of probability as the test of proof. For the jury to have found that a warning whistle after discovery of the danger would have made an avoidable situation it was necessary to make assumptions no fair view of the evidence justifies.
As no causal negligence on the part of the defendant is to be inferred, no attention has been given to exceptions other than those to the denial of the motions for directed verdicts.
Judgments for the defendant.
All concurred.