The master has found as a fact, in this proceeding to have a supposed or alleged marriage decreed null and void, that the respondent had been previously married to one Anthony Krul. This is admitted by respondent to be true. The master further found as a fact, against denial by respondent, that, prior to the supposed or alleged marriage with libellant, respondent had gone through a marriage ceremony with one Carl P. Newton.
It was also found as a fact and as a conclusion of law that Krul was living on May 25, 1923, when the parties to this action entered into the alleged marriage. The correctness or incorrectness of this finding is the nub of respondent’s exceptions to the master’s report, which recommends that a decree be granted.
It is a question of law that is raised, since the master has likewise found that “there is no evidence whatever that Anthony Krul was alive at the time of the marriage of the present parties on May 25, 1923. There is no evidence concerning the existence of Anthony Krul since Aug. 13, 1920. There can be no finding of fact that Krul was alive at the time of the present marriage, unless a presumption will supply the lack of positive evidence.” A similar finding was made regarding Newton, excepting that he had not been heard of since 1918.
The master has decided that a presumption does supply the lack of positive evidence, using the following language: “The evidence proves that Anthony Krul married this respondent in 1912; also, that said Anthony Krul was known to have been alive in 1920. This raises a presumption that said Anthony Krul was alive at the time of the supposed or alleged marriage between the parties hereto on May 25, 1923.” In his brief in support of the exceptions to the report, counsel for the respondent argues that there is no such “presumption of life” and that libellant has not, therefore, established his case, in that he has not proved the existence of Krul at the date of the second ceremony.
*370Perhaps the first inquiry should he as to whether the burden was on the libellant to prove that Krul was actually living at the time of the second marriage, or on the respondent to show that Krul was dead at the time of the second marriage. It is said in Stymiest v. Stymiest, 4 Dist. R. 305, and O’Keefe v. O’Keefe, 15 Pa. C. C. Reps. 88, that the burden is on the libellant to prove that the former spouse is living at the time of the subsequent marriage. Whatever the desirability might be of having the burden cast on the respondent in such cases to prove the death, once it is established that he or she was married before and no divorce is shown to have been procured, there would seem to be no legal justification for so holding. The second marriage, under the policy of the law, is presumed to be valid and the burden must necessarily rest on him who attacks its validity. This is general law. See Ann Cas., 1918 E, 1233, and 34 Am. Law Reps. 464.
Does the libellant satisfy the burden which is his by showing that the former spouse has been heard of within seven years? Is there a “presumption of life” which obtains until the presumption of death arises? In 5 Wigmore on Evidence (2nd ed., 1924, § 2531, it is stated: “It is not possible to say that there is a genuine presumption of life, with a uniform application. The state of the pleadings will show whose duty it is to prove life at a certain time; and upon his showing life at a preceding time, the court will usually leave it to the jury to say whether he has proved his case, but may sometimes apply a genuine presumption, shifting the duty of producing evidence upon the circumstances of the particular case.”
That same eminent authority states further, in speaking of the presumption of death which arises from a person’s continuous absence from home for seven years unheard of: “The rule of the presumption, however, extends merely to the fact of death from and after the end of the period; it is not understood to specify anything further; for example, the time of death within that period. . . .”
The rule in- Pennsylvania does not follow the above quotations. In the early case of Burr v. Sim, 4 Wharton, 150, Chief Justice Gibson said (page 170) : “But the presumption of death, as a limitation of the presumption of life, must be taken to run exclusively from the termination of the prescribed period; so that the person must be taken to have then been dead, and not before. Indeed, that is a necessary conclusion from viewing it, not merely as a limitation, but as a countervailing presumption, which, as it does not supplant its predecessor before the end of the period, assumes no more than that the individual and the period expired together; and the predecessor being still in force, to rule the case in» respect to the time covered by it is sufficient to sustain an inference of intermediate existence throughout. Thus the presumption of life continues till it is displaced by a more potent one, which, however, has no retroactive force; and, indeed, it would be of little use if it had, for to leave the time of the death still uncertain would leave a perplexity which it was its purpose to remove.”
It is clear from this case and from numerous cases which have followed it, although many of them are manifestly based on dicta, that there is in Pennsylvania a presumption that life continues for a period of seven years after a man is last heard of, and if there is no evidence of some special peril to him during that period, it is only at the expiration of such period that the presumption of death arises; the latter presumption is not “retroactive.” See cases collected by the present Chief Justice in Groner v. Supreme Tent, etc, 265 Pa. 129; Fanning v. The Equitable Life Assur. Society, 264 Pa. 333; Francis v. Francis, 180 Pa. 644; Petition of Mutual Benefit Co., 174 Pa. 1; *371Whiteside’s Appeal, 23 Pa. 114; Young v. Sweigart, 69 Pa. Superior Ct. 525; Jaskalski v. Catholic Union, 64 Pa. Superior Ct. 535; Baker v. Fidelity Title and Trust Co., 55 Pa. Superior Ct. 15.
We now come to a matter which has evidently been overlooked by the master and by counsel for both parties in their briefs. We have, stated two presumptions; one concerning the continuation of life, and another concerning the validity of a second marriage. It is manifest that in suits for the annulment of the second marriage, where the evidence is such as has been produced in this case, these two presumptions conflict, and we must decide which is the stronger of the two. As was said in a very old case (Jayne v. Price, 5 Taunt. 326) : “Nothing can be clearer than this, a presumption may be rebutted by a contrary and stronger presumption.” Although there does not appear to be any clear-cut annulment of marriage case in Pennsylvania in which the question has arisen, there is sufficient authority for the proposition that the presumption of continuance of life falls before the stronger presumption of the validity of a marriage. The latter presumption, in the case of a second marriage, is based upon one of the strongest presumptions known to the law, namely, the presumption of innocence, for, in order to uphold the presumption that a man once shown to he alive continues to live for seven years, the law-will not presume or conclude that the crime of bigamy has been committed or that the children of the second marriage are illegitimate. Lawson on Presumptive Evidence (2nd ed., 1899), at page 526, reads: “Where there are conflicting presumptions, the presumption of innocence will prevail against the presumption of the continuance of life.” In the early English case of King v. Inhabitants of Twyning in Gloucestershire, 2 Barn. & Ald. 386, 106 Eng. Rep. 407, it appeared that Mary B. married W., who afterwards enlisted and went on a foreign service and was never heard of afterwards; twelve months after his departure she married B. It was held that the issue of B. were presumed to be legitimate. “It is contended that his (W.’s) death ought to have been proved, but the answer is that the presumption of the law is that he was not alive when the consequence of his being so is that another person has committed a criminal act.” The opposite result was reached in King Inhabitants of Harborne, 2 Ad. & E. 540, because of the very short time which elapsed between the second marriage and the time when the first wife was last heard from. Other early cases in which a first spouse was presumed dead to validate a second marriage are Yates v. Houston, 3 Tex. 433; Lockhart v. White, 18 Tex. 102; Wilkie v. Collins, 48 Miss. 496; Sharp v. Johnson, 22 Ark. 79; Greensborough v. Underhill, 12 Vt. 604.
Wilcox v. Wilcox, 171 Cal. 770, 155 Pac. Repr. 95, was a proceeding for the annulment of a marriage. There was no affirmative evidence that B. was alive on June 29, 1910, the date of the second marriage, other than that to the effect that he was alive seventeen or eighteen years before, and was then thirty-three or thirty-five years of age and rugged and healthy. The court said: “The most that can be said as to this evidence, however, is that it might possibly be held sufficient to create the prima facie presumption that he was still alive, nothing to the contrary appearing. But the presumption of the continuation of life is overcome by another. It is presumed that a person is innocent of crime or wrong. There is also a presumption, and a very strong one, in favor of the legality of a marriage regularly solemnized. Rather than hold a second marriage invalid and that the parties have committed a crime or been guilty of immorality, the courts have often indulged in the presumption of death in less than seven years, or, where the absent party was shown to be alive, have allowed a presumption that the absent *372party has procured a divorce. A more correct statement perhaps would be that the burden is cast upon the party asserting guilt or immorality to prove the negative — that the first marriage had not ended before the second marriage.” ,
In Vreeland v. Vreeland, 78 N. J. Eq. 256, 79 Atl. Repr. 336, it was held that the presumption of the validity of a subsequent marriage is not overcome by proof that a woman’s former husband was alive at a time between three and four years prior to the second marriage; and in Michaels v. Michaels, 91 N. J. Eq. 408, 110 Atl. Repr. 573, a similar result was reached, where it appeared that the former spouse of one of the parties had been seen alive about a year before. Numerous other courts have decided in favor of a second marriage, although the former husband or wife had been heard of within the seven-year period: Greenwood v. Frick, 233 Fed. Repr. 629, 147 C. C. A. 437 (two years); Cooper v. Cooper, 86 Ind. 75 (four years); Smith v. Fuller, 138 Iowa, 91, 115 N. W. Repr. 912 (three years); Kelly v. Drew, 12 Allen (Mass.), 107 (four years); Wagoner v. Wagoner, 128 Mich. 635, 87 N. W. Repr. 898 (three years). These cases and others are commented upon in an excellent annotation entitled “Presumptions Flowing from Marriage,” in 34 Am. Law Reps. 464.
There are several helpful cases in Pennsylvania, although they do not involve a proceeding for the annulment of a marriage. Thus, in Breiden v. Paff, 12 S. & R. 430, which was an action of ejectment, a witness testified that a woman involved in plaintiff’s chain of title had three prior husbands, and it was contended that to establish the validity of a certain conveyance made by her and Paff, her husband, it was necesary to show that all of these three husbands were dead at the time of executing it. Chief Justice Gibson, in the course of his opinion in the Supreme Court, said: “I am of opinion the court were right in leaving the jury to presume that the persons to whom she had been married previously to her marriage with Paff were dead. In an old transaction like this, the fact of a second marriage is, of itself, some evidence of the death of the former husband. There are sometimes cases where it is unavoidably necessary to decide on the existence of facts without a particle of evidence on either side, and if a decision in a particular way would implicate a party to the transaction in the commission of a crime or any offence against good morals, it ought to be avoided; for the law will not gratuitously impute crime to any one, the presumption being in favor of innocence till guilt appear.”
In Wile’s Estate, 6 Pa. Superior Ct. 435, President Judge Rice wrote as follows: “When the existence of a valid marriage relation is once established by proof, it is to be presumed, ordinarily, that it continues to exist until the contrary is shown or until a different presumption is raised. Where this presumption comes in conflict with the presumption of the innocence of either of the parties in marrying a second time, and of the legitimacy of the offspring of such marriage, the question arises which shall yield. If one of the parties has been absent from his or her domicil, unheard of for seven years, there is no difficulty. In such a case death is to be presumed: Francis v. Francis, 180 Pa. 664; but this presumption is subject to be rebutted, as it was in the present case, by proof that he was alive at the time of the second marriage: Thomas v. Thomas, 124 Pa. 646. But proof that he was alive is not positive proof that he was still the lawful husband of the woman to whom he was first married. That fact rests alone on the presumption of the continuance of a relation which might have been dissolved by divorce as well as by death. Upon that bare presumption the appellants’ whole case rests. In *373other words, they say that Benjamin Andrews was married to Elizabeth in 1866, and was alive when she married John Shetzline in 1884; therefore, because of the presumed continuance of a proved relation, the latter marriage was void, the child born of it was a bastard and was incapable of inheriting from or through his father. Possibly, if there were nothing further in the case, this presumption would neutralize the prima facie presumption in favor of the validity of the marriage directly in issue, although that has been denied in more than one ease. But be that as it may, the proposition that the former presumption must always prevail, in the absence of full proof of the dissolution of the first marriage, is not sustained by principle or .the weight of authority.”
This has to do with a slightly different thought, but it does show the reluctance, if not refusal, of the courts to declare a second marriage invalid on the basis only of presumptive evidence.
The most important case is that of McCausland’s Estate, 213 Pa. 189, in which Elizabeth Evans, in 1882, married John Rodgers and lived with him until 1884. The last time Rodgers was seen or heard of was Aug. 3, 1885. In April of 1887 she entered into a common law marriage with McCausland and had a child by him; they lived together until McCausland’s death in 1903. The court decided that McCausland’s child was legitimate. The reasoning of the court in its opinion, while somewhat roundabout, is pertinent to the present case. Even if Elizabeth was under a disability to marry McCausland because of her prior union with Rodgers, that disability presumptively ended seven years after Aug. 3, 1885, and thereafter, in the absence of proof that he was still living, she could become the lawful wife of McCausland. The court then goes on to say: “Though seven years must elapse before the presumption of death arises; when this period does elapse, there is no presumption as to the time when, during the seven years, the death of the absent party actually occurred, and, therefore, to help the presumption of innocence or legitimacy, there is no presumption that it occurred after the second marriage, but rather that it occurred before. Semper pmesumitur pro matrimonio. This is, of course, but a presumption to be rebutted by proof of the actual date of the death; but in the absence of such proof, with the presumption in favor of legitimacy, the presumption is in favor of the validity of the second marriage as not having occurred prior to the death of the absent husband.”
Concerning McCausland’s Estate, it was said in Baker v. Fidelity Title and Trust Co., 55 Pa. Superior Ct. 15: “That was a case arising on the distribution of an estate in the Orphans’ Court where the validity of a second marriage was involved. The wife having contracted a second marriage within seven years after the disappearance of her first husband, and the question of the validity of the second marriage having arisen more than seven years after her first husband was last heard from, it was said by the court, in considering one of the questions raised, that though seven years must elapse before the presumption of death arises, when this period does elapse, there is no presumption as to the time when, during the seven years, the death of the absent party actually occurred, and, therefore, to help the presumption of innocence or legitimacy, there is no presumption that it occurred after the second marriage, but rather that it occurred before. This was a statement in another form of the rule that the presumption is in favor of legitimacy and of the validity of the second marriage as not having occurred prior to the death of the absent husband. One presumption encountered another; the presumption of the validity of the marriage overcame the presumption of the continuance of life of the man who had disappeared.”
*374It is true that in the present case seven years have not elapsed since Krul, the first husband, has been heard of, but we do not think that changes the situation. Nor do we think McCausland’s Estate not in point, merely because we are not here concerned with the rights of issue of the second marriage as legitimate offspring, although we have read with interest the opinion of Judge Gest, of the Orphans’ Court, in Wallower’s Estate, 6 D. & C. 211, to the effect that a marriage is invalid where it appears that the woman had been previously married and her husband had admittedly been living three years prior, there being no positive testimony one way or the other that he was living at the time of the second marriage. While the rights of issue of a second marriage as legitimate offspring are not here involved, it is to be considered that there is a child born of this marriage and still living; to grant the decree sought would, of course, be tantamount to declaring the child illegitimate.
It is our opinion, therefore, that the libellant does not make out a case for the granting of a decree under the Act of April 14, 1859, P. L. 647, by showing only that there was a previous marriage and that the first spouse was alive three years before the date of the second alliance.
For which reason, the order heretofore entered approving the report of the master is now vacated, and a new order is entered sustaining the exceptions subsequently filed to such report and dismissing the libel.