Page, J.

I. (1) James E. Wilbur, employed by the defendant, identified entries made by him contemporaneously in a book of records, showing tests and inspections of the automatic signalling apparatus at Barberry Crossing from December 10 to 30, 1935. In the same book were entries made prior to December 10 by another inspector, who did not testify. These other entries were not identified. Wilbur testified, after refreshing his recollection, to the complete scope of every one of his entries. His entries were then received in evidence. The whole book went to the jury, but with instructions to “pay no attention to any entries before the date of December 10th. Confine yourselves to the entries beginning December 10th, beginning with that date, on that date and after that date.” There were no entries after December 30.

The plaintiffs excepted on the grounds (a) that the entries were not admissible, since the entrant had testified to the subject-matter of his own knowledge and memory, and (b) that the book contained unidentified entries prior to December 10.

The admissibility of regular entries generally depends on the unavailability of the entrant as a witness, the testimony of the entrant being usually regarded as primary evidence and the entries themselves as secondary. Wigmore, Evidence (3d ed.), s. 1521. The reason for the limitation is that the entries must be tested by cross-examination whenever that is possible. 76., s. 1522. Wilbur was present and actually cross-examined. The real reason for the exclusion of the entries as a matter of law disappeared, even though they refreshed his recollection. If the court in discretion might have ruled the entries out as superfluous and cumulative, as in a sense “secondary evidence” of facts already shown by “primary evidence” of refreshed memory, it cannot be said as a matter of law that it was improper to admit them. There could be no possible prejudice, perhaps there might be advantage, in letting the jury see the original, contemporaneous records. They were, in fact, the sole basis of Wilbur’s testimony; in a real sense, since he was present to testify to them and was cross-examined concerning them, they became primary evidence, and his refreshed recollection of his doings, wholly dependent on the entries he made, was in a sense secondary. *152 No reason is perceived for supposing that the jury, contrary to the clear and definite instructions, were prejudiced by other entries than those made by Wilbur. The rule that the offering party using a memorandum to refresh recollection may not introduce it, while the contrary party may, and the jury may demand it as a matter of right (Wig. Ev. (3d ed.), s. 763) cannot rest upon the theory that it is always harmful for the jury to see the original entry. What the jury may see on their own motion or the motion of the cross-examiner cannot be harmful. This is one of those rules that are not “ dogmas of inherent efficiency. They are merely crude rules-of-thumb, . . . The trial Court’s discretion should control.” 3 Wigmore (3d ed.), s. 765.

(2) The reading by the same witness of a carbon copy of a report he made to the defendant, showing that within an hour after the accident he tested the automatic crossing signals, was excepted to by the plaintiffs. This report was marked for identification, but was later withdrawn and never went to the jury. The fact that the test had been made had already appeared, and there was no possible, harm in the reiteration of that fact. State v. Slocinski, 89 N. H. 262, 266.

II. The plaintiffs excepted to the failure of the court to grant certain requests for instructions.

(1) They sought instructions to sustain their claim that the engineer should have sounded his whistle after he saw the decedents approaching the crossing. There was no last clear chance. The uncontradicted evidence of the engineer was that he first saw the decedents’ car when he was about seventy-five or one hundred feet from the crossing and the automobile was about fifteen feet from it. He immediately applied the emergency brake and the sand. These operations required the throwing of two levers by hand. He operated the brake with his right hand. The sand valve was somewhat to the left, and the whistle was operated with his left hand. The engineer was unable to remember whether he used his left hand to throw the sand valve lever, or whether he used his right hand successively on the brake valve and the sand valve. Whether he disregarded the whistle and operated the sand lever with his left hand (thus accounting for the possible ceasing of the whistle before the train reached the crossing) can make no difference. In the emergency, which involved no more than a second or two for thought and saving action, whatever the engineer did was bound to be instinctive only. Moreover, it could not be found that the sounding of the *153 whistle would be more effective in avoiding collision than the application of sand. Paulette v. Railroad, 88 N. H. 10; Gaudette v. McLaughlin, 88 N. H. 368, 373; Lavigne v. Nelson, 91 N. H. 304. The plaintiffs’ assertion that the engineer had over three seconds for action does not rest upon any evidence of distances and speeds found in the record, but upon the mere supposition that the speed of the automobile was less than four miles an hour.

(2) The plaintiffs requested instructions that the defendant was bound to exercise reasonable care to provide an automatic signal that would give an efficient warning and that if the failure so to do caused the deaths of the decedents, the defendant would be liable, provided the decedents were not themselves negligent.

Though the instruction was not given in this precise form, the jury were told what reasonable care is, and that in deciding whether the defendant was causally negligent they should “consider what care the ordinary person of average prudence engaged in a similar business would have taken with respect to the following matters: The installation of the crossing signals and the length of its circuits, the maintenance of the strength of the batteries and the tightness of contacts of the signal system.” To this charge there was an exception because of claimed insufficiency.

The particular element that the plaintiffs say was omitted from the charge, though called for by the requests, was the alleged inefficiency of the automatic signals resulting from the arrangement of the electric circuits northeast of the crossing. Southwest of the crossing, there was a single circuit 2,700 feet long, designed to give impulse to the signals when a train approached from that direction, as did the one that collided with the decedents’ car. In the other direction there were several circuits, which in conjunction were designed to give notice of the approach of trains from the northeast. The latter were so arranged that the flasher and bell operated while shifting operations were going on at a factory located some 1,200 or 1,500 feet northeast of the crossing.

It was argued that warning to the decedents was inefficient, since the lights might flash and the bell ring when no train was about to cross Barberry Lane in either direction. The driver of the automobile, it was said, might be misled. This whole issue was tried out at much length. There can be no doubt that the jury knew precisely what the claim was. They must have understood that the words “installation” and “length of circuits,” as used by the court, referred to this issue of claimed inefficiency. The requests were fully *154 covered in substance. The suggestion that Ingram, the decedent driver, would be misled by the situation overlooks entirely his perfect familiarity with the crossing, the unquestioned fact that one approaching the crossing at the time the decedents did had an unobstructed view of any shifting operations there might be at the right, and the further fact that on prior occasions Ingram knew about shifting operations and took them into account. If the signals were actually operating when the decedents approached, as the jury must have found, the driver, if not Davis, had warning that a train was approaching the crossing, for no shifting operations were then in progress, and since no train was in view at the right (where there was an unobstructed view of over 2,000 feet at fifty feet from the rail; at twenty feet, a full mile), the warning was bound to be of a train coming from the left. The supposed “defect” was not causal.

But this “ineffectiveness,” as the plaintiffs call it, they say is not “a defect in the crossing signal,” as the court called it, but inherent in the design, which was defective in not providing a single circuit of 2,700 feet northeast of the crossing. However, the very thought of the plaintiffs was well conveyed to the jury by the reference to “length of circuits.” It is impossible to believe that the jury were in any wise confused.

III. (1) The jury were fully and correctly instructed with reference to the statute relating to whistling for highway grade crossings. They were informed by the court that this applied in case they found that Barberry Lane was a public highway. Then they were told: “If you fail to find that Barberry Lane was a highway . . . the only issue then so far as whistling ... is concerned is whether the defendant used reasonable care by this means in giving notice of the approach of the train.”

It is now argued that the effect of this instruction “was to preclude consideration of the defendant’s duty to exercise reasonable care in general, and of the defendant’s rule [for prolonging the whistle to the crossing] as evidence of what reasonable care required in particular.” The charge as a whole does not bear the construction claimed. The trial court, after the portion of the charge relating to the statute, gave instructions as to negligence in general, and under this head told the jury to consider the question of the “giving of reasonable notice of the approach of the train by whistle from the train,” and further told them that if they found “the defendant through its agents and servants was negligent in any respect” that they found causal, the plaintiffs could recover, and that such recovery could not *155 be had if there was “no fault either of statutory violation or of negligence.”

(2) Concerning the decedent Ingram, the jury were given numerous instructions relative to contributory negligence to which the plaintiffs excepted. It would be useless to consider them. If erroneous, which we are far from holding them to be, they were nevertheless harmless. Both verdicts were for the defendant. Since the jury were clearly instructed that the decedent Davis could not be found guilty of contributory negligence or fault, and that Ingram’s negligence or fault could not be imputed to Davis, the verdict in the Davis case clearly implies a finding that the defendant was not causally at fault or negligent. The issues raised by the exceptions now being considered thus became moot, and the exceptions raise no question of law. Smith v. Bailey, 91 N. H. 507.

IV. The plaintiffs excepted to the withdrawal of three issues from the jury.

(1) The attention of the engineer to traffic on Barberry Lane. It is argued that the engineer, who testified that he saw the automobile as soon as it was visible, could have seen the top of it sooner, in spite of the intervening embankment. The engineer said that generally he could see to the line of the fence on the railroad right of way, twenty-one feet from the nearest rail. A photograph in evidence indicates that at one hundred and twenty-five feet south of the crossing and four feet ten inches above the rail, the top of a Pontiac car would be visible above the bank, provided the car was within ten feet of the nearest rail. At that point, the back of the top would just clear the fence. The engineer testified that his eyes, eight feet above .the rail, first saw the decedents’ Ford car when it was fifteen feet from the crossing. Comparative measurements of Pontiac and Ford cars do not appear. His testimony was perfectly consistent with the photograph. If he could in fact have seen the top of the car any sooner than he did, it is purely conjectural that he could have seen it enough sooner to have done anything about it.

(2) The issue whether in the exercise of due care the brakes should have been applied sooner was properly withdrawn. The argument that an application of the brakes one hundred feet further from the crossing would have avoided a collision is supported only by conjecture. It presumes that the brakes might have been applied at one hundred and fifty feet from the crossing (instead of fifty), and that the engineer should have seen the automobile when it was more than fifteen feet from the nearest rail. If the head of the car be as *156 sumed to be twenty feet from the nearest rail, the distance it must have traveled in order to pass unharmed in front of the train was twenty feet, plus fifteen feet (the length of the car), plus the distance between the rails, plus the overhang of the locomotive, over forty-one feet in all. The time for traveling that distance would have been over a second if the automobile was going twenty-five miles an hour, as the engineer estimated, or over seven seconds if it was going less than four miles an hour, as the plaintiffs claim. The engineer stopped the train as soon as possible,. some twelve hundred feet northeast of the crossing. He could have avoided the collision only by delaying the arrival of the train at the crossing by at least a second and a quarter. The argument that such a delay could be found possible overlooks the shortness of the time between the assumed application of the brakes and the collision. That a train going seventy-five feet a second (fifty miles an hour) could by application of the brakes have been retarded even a second and a quarter in a space normally passed in two seconds or a little more is beyond credence. In any event, there is no evidence at all as to possible rates of retardation, and the argued conclusion could be reached only by pure guess.

(3) The issue whether there should have been a crossing-tender involved special protection beyond the special protection already provided by the automatic signals. The only argument made for further protection is that the risk of failure of the automatic signals was so great that it might be found reasonable to employ a crossing-tender. Such risk of failure was negatived by the verdicts, for the jury were told in substance that a defect in the crossing signals would be evidence of negligence if the railroad should have known of it in time to make necessary repairs. This instruction was sufficiently favorable. Kingsbury v. Railroad, 79 N. H. 203, 204. On general grounds there was no evidence to warrant a finding that a tender was reasonably required. Though the visual conditions to the southwest were bad, the decedents knew them as well as anybody. Barberry Lane is a dead-end way, and the crossing was used for access only to the Davis farm, to a slaughter-house, and to two dwellings. The total of vehicle-crossings was thirty or forty a day, of train-crossings twenty. It could not be reasonably found that the protection afforded was so ineffective as to require further special protection.

Judgments on the verdicts.

All concurred.