The only question presented is, as to the charge upon the statute of limitations. We think-the defendants were entitled to the instructions they asked, and that his Honor erred in refusing them. In proportion to the atrocity of conduct imputed to any one, ought to be the care, with which we should guard ourselves against the feelings it naturally excited. The crime,-imputed to the defendants, for it is a criminal act, punishable by indictment, is little less than that of stealing, and only less, because the law does not make it a felony. The slave had been sold to pay the debts of the defendants, and for seven years, the plaintiff had been by them fraudulently deprived of his services. The feelings naturally excited by such conduct should not be permitted to obscure our judgment. Criminal as the defendants certainly were, and deserving of all the punishment, which the law affixed to their actions, they are still entitled to the protection, which the law throws around their per*204sons. The statute, which makes the harboring of a slave an indictable offence, also give to the owner an action for the damages he may have sustained. The time, within which such action shall be brought, is regulated by the general statute of limitations. By it, it is enacted, that actions upon the case must be brought within “three years next after the cause of such action or suit occurred, and not after.” Every act of harboring a slave is a fresh cause of action, but an action when brought covers all such acts for three years next before the bringing of the suit, for the law does not countenance the splitting up of actions. In this case, the writ issued on the 15th of September 1846, and embraced all the time between that period and the 15th of September 1843, and could not, by force of the statute, extend any further back. This was the opinion of the presiding Judge, and he so charged, but he proceeded “ that they might look to the antecedent acts of the defendants, and if, from them, they believed, that the harboring continued after the 15th of September 1843, the plaintiff would be entitled to recover for the loss of the services of the slave after that time.” This was stating to them, that the acts of harboring down to the Fall of 1842, were evidence of a harboring after the 15th of September 1843. In this opinion we do not concur. It is, in effect, holding, that the defendants, having been fixed with acts of harboring, at any time before the action was brought, were to be considered as still harboring, not only down to the time, when the statute of limitations would begin to run, but to the bringing of the action. Such a construction would virtually repeal the statute in every case, where there was more than one act of harboring. Nor can we see, why, if correct, it should not have that effect in every case. A dozen acts of harboring in 1842 is no more evidence of a harboring after the 15th of September, 1843, than would be one such. If, in measuring the time, which the statute *205covers, we pass the 15th of September, 1843, where sha i we stop, in ’42, ’41, ’40, or where else ? It is impossible, that a transgression of the law, in either of those years, can be evidence of snch transgression in ’43. A., in ’42, steals the negro of B., who regains the possession of him; in ’43, it is alleged he stole the slave again ; could it be pretended, that, on an indictment for the latter offence, the first taking could be given iir evidence to convict? Again, in an action for usury, evidence of other usurious contracts of the defendant, either with the same or another person, would not be evidence. 7 East, 103. Spencely v. Willott. The general replication in this case to the plea of the statute, is, that within three years next before the bringing of the action, the defendants did harbor the £ lave, is the replication supported by evidence, that he harbored him four years before the action was brought ? In an action of fraud in the sale of personal property, if the defendant plead the statute of limitations, it is no sufficient replication, that, within three years next before the bringing of the action, the defendant acknowl-ed the fraud. Oathout v. Thompson, 20th John. R. 277. That was an action of tort for fraud in the sale of a slave ■ — this is an action for a tort in harboring a slave. In the former, the confession referred to the time of the sale, when the fraud was committed, which was not within the time limited by law for bringing such actions. If, in this case, the defendants had confessed, after the 15th of September, 1843, that they had harbored the slave, previous to September 1842, it would, according to the case cited above, be no evidence to bring the case within the statute. Much less, could the fact of harboring, previous to the Fall of’42, be any evidence of a harboring after the 15th of September, ’43, more than a year thereafter. The plaintiff) by his replication, undertook to show, that the defendants had committed the offence within the three years next before the bringing of the action : and he must *206prove an act of harboring,’ within that time, either by positive evidence, or by proof of such circumstances occurring within that lime, as could justify a jury in so finding. It may be that the law ought not to shield from punishment, the perpetrators of such offences. Yet it has been the pleasure of the Legislature to limit to two years the prosecution of all misdemeanors-, with a few exceptions. If this, instead of being a civii suit, were a criminal prosecution for this offence, would the proof of an act of harboring in the Fail of’42 support an indictment, commenced in ’45; or would it be correct to refer the jury to the act in ’42, as justifying a conclusion by them, that the offence had been perpetrated within two years next before the preferring of the indictment 1 It is true, these are certain facts which are of such a nature, that the law presumes their continued existence, until it Is shown that they do not exist. Thus a man, who is shown to be in life at a certain time, is presumed to.be still in life, until it is shown that he is not in lile, either by proving his death, or such a lapse of time, since he was last heard of, as raises a counteracting presumption of death. So, also, sanity, when established previous to doing a particular act, is presumed to continue at the time the act is done, until the contrary is proved. But harboring a slave is not a continuous act, and no presumption arises from the proof of one such act, that another at a subsequent time, has taken place. If the slave had been seen on or near the premises of the defendants after the 15th of September 1S43, the acts of harboringprevioua thereto would be evidence-, not to increase the damages, but to show the nature of the act of being at the place at that time.
The declaration was for harboring the slave from the year 1835, to September 15, 1846 ; and the prayer of the defendants’ counsel was for instructions *207upon two points. The first, that for the harboring which occurred before September 15th, 1843, that is, three years before suit brought, the action was barred by the statute of limitations ; and the second, that there was no evidence of any harboring after that day. The presiding Judge declined giving the whole instruction, as prayed for, inducting both propositions. But he gave it, as asked, in respect to the first point; and then he left the case to the jury on the second point with instructions, that they might look to the antecedent acts of the defendants, and if they believed, from them, that the harboring was continued after the 15th of September, 1843, the plaintiff would be entitled to recover for such harboring as took place after that day. Such is my understanding of the case stated in the exception ; and therefore it strikes me as not being precisely correct to say, that his Honor had refused to apply the statuteoflimitations to the plaintiff’s demand. For I consider that he did so in direct terms, to the full extent required by the party and the law ; so that the jury were obliged to understand, that they could give damages for only such harboring, if any, as actually took place after September 15, 1813. The question, then, as it strikes me, is, whether there was evidence of a harboring within that time. His Honor thought there was, because it might be inferred by the jury, that the defendants had harbored the slave from Sept ember 15th, 1843, to September 15th, 1S4G, inasmuch as they had harbored him from 1S35 to the Autumn of 1842. So, the point is, as it seems to me, simply this : whether the previous harboring constitutes such eireumstantial proof of a subsequent harboring as ought to have been left to the jury as competent to establish it. If it was competent, the verdict ought to stand ; because it is the province of the jay to weigh and determine the effect of evidence, as enabling them to infer one fact from another, with this proviso, however: that the fact, from which the other is inferred, *208would be such as affords a fair presumption of the required fact. It is that proviso, which raises the question of the competency of the circumstantial evidence in any ease ; for it isa question of just reasoning, what inferences may be made from an admitted or established fact. Therefore, if the fact sought cannot be rationally deduced from the circumstances relied on, the circumstance itself ought not to go to the jury, because its consideration cannot serve the justice between the parties, but may mislead the jury. Suppose, in this case, the declaration to have been only for the harboring from September 15th, 1843, so as to make the statute of limitations altogether inapplicable — which, indeed, is the state of the case under the opinion given on that point to the jury; the question then would be a naked one, whether the harboring laid would or could be established by the previous harboring. It is often a delicate point to determine what mayor may not be justly inferred from particular premises ; and persons will frequently differ upon it ’When I see that a learned Judge thought, that a harboring up to the time of the action might be presumed, and that twelve gentlemen were able to affirm on their oaths its actual existence, because it had been practised for several years before and down to 1842,1 cannot but be somewhat diffident of my own conclusions to the contrary, and reluctant to disturb the verdict on that ground. But, as the law does not allow a question tobe submitted to a jury without evidence, which means also, with such evidence as, taken in the whole, will not fairly authorise a verdict in favor of the party offering the proof — in other words, evidence, on which a Judge must say, he could not find a verdict — it seems to be the imperative duty of the Court here to reverse the judgment, when a case in fhat situation has been left to the jury. Now, it seems to me, notwithstanding the previous connection between the slave and the defendants, that one cannot justly and with any rea*209sonable confidence affirm, that the connection continues for four years after the last visible trace of its existence. And after the expiration of that period it may be fairly contended, I think, that, if the harboring did not continue duing the whole period, it did not exist at all within it. For, it is a much stronger presumption now, that it did not exist, for instance, in 1843, than it was in that year, because, then it might be said, there were not opportunities, and that time had not been allowed to discover direct proof. But. now, the negro has been taken and could point out the means of establishing the fact, if it had occurred, and there has been full scope for enquiry in other quarters for nearly six years, in all, since the latest day to which the direct proof brought down the harboring ; and the whole period affords not the slightest vestige of a harboring or correspondence of any kind after the autumn of 1842. If it appeared, that the caves and other places of secret resort, once used by the runaway, had been still in use by some one, or that this negro had been seen in the neighborhood of the defendants, although personal intercourse might not be directly shown, there would be something for the mind to act on, and, possibly, the case might have been fit for a jury’s deliberation. But, with nothing further than the naked facts, that ten years before the negro had belonged to the defendants, and that four years before they had entertained and concealed him, the case is too bare of proof to go to the jury There is nothing within the time ; and the previous circumstances, thus solitary and antiquated, af-' ford a presumption too remote and inconclusive to be the ground of judicial determination.
I concur in holding, that the judgment must be reversed and a venire de novo awarded.
Per Curiam. Judgment reversed and venire de novo.