Opinion by
Me. Justice Dean,The plaintiff, a brakeman on the Pennsylvania railroad, while coupling cars, on December 19,1895, on a side track of the company in the city of Pittsburg, had his right arm crushed. He *102alleged that there was an open hole or drain eight to ten inches deep between two ties at the point where he was injured and that, not knowing of the existence of the hole, in attempting to perform his duty of coupling, he fell, and in falling threw up his arm which was caught between the bumpers and crushed. Averring negligence on the part of the company in leaving thus exposed a dangerous hole," of which he was ignorant, he brought suit for damages. At the trial the plaintiff offered his own testimony and that of several other witnesses, which tended to show the existence of the hole or drain, and that it had been caused by a washout under the rails and between the ties; that although unknown to him some of the witnesses testified, it had been there for three or four months. The defendant offered evidence tending to show, there never was a hole there and that plaintiff was not injured at that point. The learned judge of the court below left it to the jury to determine whether the hole was there, whether plaintiff’s injury was caused by falling into it, and if so, whether it was negligence on the part of the company to leave it unbridged, unfilled or unrepaired. There was a verdict for plaintiff, and defendant now appeals, assigning for error, first, the refusal of the court to instruct for defendant; second, the admission of the Carlisle life tables, other life and annuity tables, and Scribner on Dower, for the purpose of showing plaintiff’s expectancy of life and the value of an annuity on his life; third, the court’s ruling as to the effect to be given such evidence; fourth, the instruction as to the measure of damages for the injury.
In discussing the assignments of error, the first question that presents itself is, what duty did defendant owe plaintiff as concerning the ballasting of its tracks ? For we must assume as facts, that the track between two ties at that point was not ballasted like unto the other parts of it, and that plaintiff’s injury was caused thereby; the jury, under the instruction given, must have so found, and that is ah end of the matter. ,
This is a case of master and servant; it was the duty of the master to furnish a reasonably safe place for the servant to work, in view of the duties imposed upon him. What is the test of “reasonably safe” when applied to the roadbed of a steam railroad? There was a point between two ties where there was no ballast, because it had been washed out; if an *103injury had been caused a passenger in consequence of this, he would, undoubtedly, have had a cause of action, for it was the company’s duty to maintain as to him a safe roadbed. But this was not a footway for foot travelers or employees; if the roadbed had been such a way, then ordinary care would have required planking on all the ties, for only by that method would the feet and ankles be free from sprains, and the person from injury; ballast is usually composed of loose material, such as broken stone or cinder, on which rest the ties, sometimes spaced or filled between with the same material, but always with more or less inequality of surface; sometimes there is a small constructed open box drain between the ties. Except at crossings the ties are not planked, nor so far as concerns their purpose is there any reason why they should be. No custom of common carriers is shown from which a reasonable inference of negligence could be drawn from its not ballasting full between the ties; the proof is the other way. If there was owing to plaintiff the very highest degree of care, the same as to the contract passenger, then the jury might have been at liberty to infer negligence, because they might have said that, between every two ties on its miles of road, the company was bound to maintain an even or planked surface. But it owed no such duty to its employees who at rare intervals must step or stand on the roadbed. Any such degree of care would warrant the jury in finding the company was bound to maintain guard rails on embankments and planking on all bridges, although not a single passenger on a train would be imperiled from the absence of either.
We concede, as demonstrated by appellee’s argument and citar tion of authorities, that certainly a modified, if not a different, rule holds in some of the states; but in our state it has been consistently held, that the railroad company owes no duty to the public or its employees to maintain a safe footway the length of its roadbed; that it is reasonably safe to the employee without it. In R. R. Co. v. Schertle, 97 Pa. 455, the deceased, a brakeman, lost his life by slipping into a hole between the ties where there was no ballast. This court said: “ There certainly was no duty to ballast the track for the safety of its employees, and except perhaps at a crossing no such duty to the public.” See also Costello v. R. R. Co., 32 W. N. C. 134. *104The same consistency in ruling obtains in New York state. In Finnell v. D., L. & W. R. Co., 129 N. Y. 669, the plaintiff was a brakeman who was injured while coupling cars on a side track used for storing cars and making up trains. His foot caught between two ties where there was no ballast. Earl, J., says, page 671: “ He claims that the defendant was negligent in not ballasting this branch track so as to make it safe for him to walk thereon and to discharge his duty as brakeman while standing thereon. But we do not understand that railroad tracks are ballasted for the purpose of making them safe for brakemen to walk upon, but for the purpose of making them firm and.safe for the passage of trains, and we do not think the defendant neglected any duty it owed the plaintiff in not ballasting the track at the place where the accident occurred.”
This is followed by other cases to the same effect. We prefer to follow our own and the New York rule, because we consider it sound in reason and that it opens up no wide field of guessing or conjecture on part of the jury. In Arkansas, Texas, and some other states, the opposite rule is held; the burden of proof in case of accident of this nature is on the company to negative negligence from the mere fact of an opening between the ties whereby the brakeman was injured. To this rule we do not assent.
- Nor does the fact that the accident was upon a side track or yard, at all change the rule. The learned judge of the court below, while seemingly conceding it to be correct when applied to an accident of this character on the main track, did not think it should prevail when the injury occurred in a yard or on a side track. It is not improbable that the danger from such cause is somewhat greater in the case of a side track; but the tracks and ballast are for the same purpose as on the main rail; they are constructed, not for employees, but for the running of cars, and the difference is one of mere degree of danger and not of kind. . We think on the undisputed facts the plaintiff had discloséd no negligence, as to him, which would support a verdict.
Although our ruling on this first assignment practically ends plaintiff’s case, nevertheless, the second and third assignments, as to the admissibility and effect of the life and annuity tables, *105call for notice. At the close of plaintiff’s testimony this offer was sustained:
“ Counsel for plaintiff offer in evidence 2 Scribner on Dower (ed. of 1888), pp. 811, 812, 818, containing the Carlisle and other tables, for the purpose of showing the expectancy of life of the plaintiff; also for the purpose of showing the value of an annuity on his life. Objected to by defendant’s counsel as incompetent and irrelevant.
“ The Court: The exception is overruled and bill of exception sealed to the defense, May 8, 1899.”
It will be noticed the offer was general in its character; it pointed to no particular life table as applicable to the special facts of this case; the purpose announced is to show the expectancy of life — it is not suggested whose lives, what class of persons, what the perils of their daily employment or avocation; then the further purpose of showing the value of an annuity on plaintiff’s life. This pellmell offer, with no specific disclosure of purpose showing its applicability to this evidence, is met by the most general objection that could be framed by defendant’s counsel; no attempt is made to pin the proposal down to a specific purpose which would throw light on the issue. The court, without the least aid from either counsel, admitted the offer, we do not doubt, inadvertently, and thus fell into error. If both counsel sought at that stage to get the case in shape for a reversal, they pursued the course most likely to result in one. As to the Carlisle tables, we have held them admissible only, as some evidence for the consideration of the jury in determining the expectancy of life of an individual where that question became material. They were considered of some general value, because, as nearly as could be ascertained, although made up a century ago, they were based on statistics of the general population of two parishes of England, covering a period of about nine years; they were not based on selected lives. In Steinbrunner v. Ry. Co., 146 Pa. 504, we first held them admissible. Any one can see at a glance from that opinion, that they were admitted because they were not based on selected lives; at the same time it was distinctly stated they were by no means conclusive ; that their value depended greatly on similarity of the life in question to the conditions and habits of those tabulated a century ago; and while they were de*106dared admissible, Pax,son, C. J., who rendered tbe opinion, at the close of it speaks thus: “ While we are unable to see how such evidence is to be excluded I must be allowed to express the fear that it may prove a dangerous element in this class of cases, unless the attention of the juries is pointedly called to the other questions which affect it.” Here the only proof tending to show Kerrigan’s expeetanc}'- of life is, that he was twenty-two years of age, and that he had been a railroad brakeman for about three years; his habits and health were not proved. Under such meager facts, the expectancy of life as fixed by any tables can have but little weight. Then, as to the tables made up by actuaries of reputable insurance companies, we have no doubt they furnish a fair expectation of life in the selected lives on which they are based.' There was scarcely any proof of facts which brought plaintiff within this class of selected lives. The court says to the jury, that the life of a young man of this age, according to the tables, is estimated at forty to forty-five years, and that they could take that as a means of estimation of what would compensate him for his loss. The fair inference that the jury might draw from this statement is, that the tables, not only were some evidence of plaintiff’s expectancy, but that they established it. The tables were not entitled to such weight.unless, by precedent proof he had brought himself clearly within the class of selected lives tabulated; and this he had failed to do. We call attention to the care that ought to be exercised by the court in dealing with this sort of evidence in any particular case. 'In two cases since Steinbrunner v. Ry. Co., supra, to wit: Kraut v. Ry. Co., 160 Pa. 327, and Campbell v. York, 172 Pa. 222, we have affirmed the admission of this kind of evidence., Experience has demonstrated however that what was merely apprehended by Paxson, C. J., has since been realized. Courts and juries as a rule give far more weight to this testimony than it is entitled to; they are apt to supply the place of proof of expectancy of the particular life by generalization from life tables. This is going further than was intended, or than is warranted; therefore a halt is called on the manifest tendency to give them undue weight.
As to the annuity tables, they were not admissible at all in a case of this character. Such a table is based on the average anticipation of death, without taking account of capacity to *107work, indolence, vicious habits, or a tendency thereto, or diminution of ability to earn. For example, these tables would answer this question: What is the present worth of 1500 per year to one who is now twenty-two years of age? We recur to the tables to see when he will probably die, and the lump sum or present worth is easily ascertained. This is purely mathematical. The health, earning power and industry of the particular individual have no place in the calculation; it is based on the assumption that the plaintiff is entitled to enough money down with which he may purchase an annuity for the remainder of his life, and enjoy the same annual sum at sixty or sixty-five as at twenty-two. We can see how, from such a table, the present worth of a widow’s dower can be computed; hers is a fixed annual payment, affected not by age, health or ability to earn, but only by her death; her age being known and her probable longevity being shown by the table, no other fact is necessary. We think that annuity tables throw no light on the essential facts necessary to making up a verdict in an issue of this character, and they should have been excluded.
There is nothing in the remaining assignment of error which calls for notice in this appeal. But for the reasons stated in the first assignment and the effect which the jury were permitted to give to the life tables admitted in evidence, as well as for admitting in evidence the annuity tables, the judgment is reversed.