delivered the opinion of the Court.
G. Walter Bowman, late of Washington County, died intestate on March the fourth, nineteen hundred and three. The administrators of his personal estate filed in the Orphans’ Court a petition asking that a day be assigned for the distribution of his assets. Due notice of this was given and later on a person claiming that she was the widow of the deceased, to whom she asserted she had been married on July 12th, 1887, at Camden, New Jersey, and giving her name as Catherine E. Bowman, appeared to the proceedings and asked that the share of the estate rightfully belonging to a widow should be turned over to her. Subsequently a certain Lettie E. Bowman, also claiming to be the widow of the deceased to whom she was married on January 18th, 1900, set up a similar claim. After other proceedings were had issues were finally framed and transmitted to the Circuit Court for trial. In the order sending the issues to the law Court, Catherine E. Bowman was made plaintiff and Lettie E. Bowman and the administrators were made defendants. Upon the suggestion and affidavit of the plaintiff the record of the issues was re*286moved to the Circuit Court for Carroll County where a jury was impanelled and the questions were tried.
The issues were as follows: “1st. Was Catherine E. Bowman at the time of the death of G. Walter Bowman the lawful wife of G. Walter Bowman ? . 2nd. Was Lettie E. Bowman at the time of the death of G. Walter Bowman the lawful wife of G. Walter Bowman ?” During the progress of the trial eighteen-exceptions were reserved—of which seventeen relate to rulings on the admissibility of evidence, and the last concerns the action of the Court on the numerous prayers presented by both sides for instructions to the jury. The verdict was in favor of the defendants; or to be more precise, the jury answered the first issue in the negative; and the second in the affirmative. From the rulings set forth in the bills of exception the plaintiff has appealed.
It is obvious from this, outline of the case that the single question before the jury was; which of these two women is the lawful widow of the decedent ? Around that question all the subordinate inquiries presented by the record revolve. There is not the slightest reason to doubt that Lettie E. Bowman, formerly Lettie E. Eakel, was in a formal manner, married to G. Walter Bowman in January, 1900, by a regularly ordained minister of the gospel. That fact is beyond controversy. The fruit of that marriage is one child, Walter E. Bowman, who, by his guardian, is also a'party to these proceedings.
At the close of the evidence the defendants, amongst other prayers, presented the following :
1st. That the verdict of the jury must be against the plaintiff upon the first issue, and their answer to. said first issue must be “No,” because the plaintiff has offered no legally sufficient evidence to prove that George Walter Bowman referred to in the record of a marriage in Camden, New Jersey, offered in evidence, is the same George Walter Bowman upon whose estate letters of administration have been granted to the defendant administrators in this case. .
*2872nd. That the verdict of the jury must be against the plaintiff upon the first issue, and their answer to said first issue must be “No,” because the plaintiff has offered no legally sufficient evidence to prove that the Catharine McGranigan referred to in the record of a marriage in Camden, New jersey, offered in evidence is the same person as the plaintiff in this case.
They were rejected. If they ought to have been granted there is an end of'the plaintiff’s case, even though there may have been errors in other rulings found in the record. Proof of the marriage of G. Walter Bowman, the decedent, to the plaintiff at the time and place alleged by her was absolutely indispensable. No one can contravene that proposition. But what kind of evidence is necessary to establish that status or relation, and of what probative value should it be, when the consequences incident to the sustentation of the alleged marriage of July, 1887, must inevitably be the branding of the deceased with the crime of bigamy, and the bastardizing of the innocent off-spring of the marriage of 1900? Let us first see what measure of evidence the law requires in such circumstances and upon what presumptions it relies. In Taylor v. Taylor, 1 Lee, 571, 5 Eng. Ecc. Rep. 454, where two women severally claimed administration of the effects of a decedent as being his widow (which was twice before the Ecclesiastical Court in England) it was said there must be “strict proof’ of the alleged antecedent marriage “as an actual fact." And this was cited with approval by this Court in Jones v. Jones, 45 Md. 159, and in the same case, 48 Md. 398. The reason upon which the doctrine that there must be “strict proof” of the first marriage, rests is apparent. When the presumption of a lawful marriage is met by a counter presumption of innocence, the former must yield to the force of the latter. After it has been shown that there was an actual marriage solemnized in the method which the law prescribes and followed by the birth of issue; every inference is invoked in support ofits validity and against an alleged antecedent marriage, because the presumptions of the law are always in favor of in*288nocence and of legitimacy. “The law presumes morality and not immorality, marriage and not concubinage, ’ legitimacy and not bastardy.” Teter v. Teter, 101 Ind. 129, cited in note 3, p. 1202, 19 Am. & Eng. Ency. L. (2nd ed.); Rooney v. Rooney, 54 N. J. Eq. 246; Patterson v. Gainse, 6 How. 550. In King v. Inhabitants of Twyning, 2 Bar. & Ald. 386, a very strong illustration of the predominance of the presumption of innocence over other presumptions is furnished. The case involved merely the settlement of a pauper. A woman had married a soldier who soon afterwards left for the East Indies. Within twelve months she married again, and the question turned upon the validity of the second marriage, and it was upheld. ' Bailey, J., said : “The facts of the case are that there is a marriage of the pauper with Francis Burns, which is prima facie valid, but the year before that took place she was the wife of Richard Winter, and if he was alive at the time of the second marriage, it was illegal and she was guilty of bigamy. But are we to presume that Winter was then alive? If the pauper had been indicted for bigamy, it would clearly, not be sufficient. In that case Winter must have been proved to have been alive at the time of the second marriage. It is contended that his death ought to have been proved, but the answer is that the presumption of law is, that he was not alive when the consequence of his being so is, that another person has committed a criminal act.” This is quoted with approval and adopted in Jones v. Jones, 48 Md. 399, and the case is cited as an authority in Le Brun v. Le Brun, 55 Md. 504. In Piers v. Piers, 2 H. L. Cas. 331, it was held that the question of the validity of a marriage cannot be tried like any other question of fact which is independent of presumptions, for the law will presume in favor of marriage, and that this presumption must be met by strong, distinct and satisfactory disproof. It is not.denied that in this country there is some conflict of decisions and of judicial opinion on this subject as was admitted in Jones v. Jones, 48 Md. 399; "but it cannot, we think (observed this Court in that case) be said that the preponderance of authority is the other way.”’ The *289law in this State is undoubtedly as the Maryland cases above cited have announced it. Now let us see whether the evidence adduced by the plaintiff measured up to these requirements and overthrew the presumptions to which we have alluded.
The evidence admitted consists of three distinct items of proof, and that which was offered and rejected comprises two others. These will be separately dealt with. It must not be forgotten that the burden of proof is on the plaintiff to establish by “strictproof ” the prior marriage as a ground for the annulment of the second one, which otherwise is confessedly valid. To discharge that burden there were adduced a marriage certificate; the testimony of a witness to certain occurrences in Harrisburg, Pennsylvania; and the testimony of a witness as to some things which transpired in Hagerstown. As the certificate is the most important feature of the plaintiff’s evidence it will now be transcribed.
State of New Jersey.
Marriage Certificate.
Full name of husband, George W. Bowman. Place of res idence Haleystown, Md. (If in the City, give name street and Number; if in a township give name of County.)
Age 25 years, — months, Number of his marriage, First. Occupation, Confectio7ier, Country of birth, U S. A. Name of father, George R. Country of Birth U. S. A. Full maiden name of wife, CatheriTte McGranagan. Country of Birth, U. S. A. Place of-residence, E. Plarrisburg, Pa. (If.in a City give name, street number, if in a township give name and county.) Age nearest birthday, 2‡. If any trade or business so state.) Last name if a widow,--Number
of brides marriage, 1st. Name of father, John ; Country of birth, U S. A. Maiden name of mother, Sarah Weaver, Country of birth, U. S. A. Date (in full) July 12th, 1887. Place, Camden, N. J.
In presence of Mary Westwood.
John R. Westwood.
The certificate shows that George W. Bowman of Haleystown was married to Catharine McGranagan, of E, Hci7'risburg, Pennsylvania; but it does not show that the George W. Bowman therein named was the identical G. Walter Bowman of Ha*290gerstowu whose estate is involved in this controversy. The certificate shows that George W. Bowman was married to Catharine McGranagan of E. Harrisburg but it does not show that the Catharine McGranagan therein mentioned is the identical Catharine E. Bowman who is the plaintiff in this case. The Certificate does not and could not prove an extrinsic fact, and personal identity is an extrinsic fact. You .may say it is not probable that some one personated Bowman, the deceased, and that some one else personated the plaintiff in the alleged New Jersey marriage. Perhaps that is so; and perhaps, too, it is- not probable that the delineation given in the certificate of the persons mentioned therein would have corresponded so closely with the description of the deceased and of the plaintiff, as it apparently does, if the deceased and the plaintiff were not the real parties who were married in July, 1887, at Camden; but the sum of these probabilities, plus all other probabilities that may be suggested from the record, would still be only á probability. Or, to state the same thing in a more precise way: A conclusion logically drawn from premises which are themselves mere probabilities; must of necessity be only a mere probability also, because a conclusion, to be formal, must always be contained in the premises, and probabilities can never contain a certainty.. There" is not a shred of evidence to establish the identity of the parties named in the certificate. Suppose it be conceded that the certificate proves that some man giving the same name, the same occupation and the samé place of residence as George Walter ■Bowman had, was married to some woman giving the same name and the same residence as the plaintiff professes „to have had; still, it does not prove that the G. Walter Bowman, now deceased, was that man; or, that the plaintiff was that woman. To justify the conclusion from' the certificate alone that the plaintiff and G. Walter Bowman were then and there married, it must be inferred that G. Walter Bowman now dead was the identical man described in the certificate, although the Bowman mentioned was said to be twenty-five years old and the deceased was then, according to the evidence, twenty-seven *291years of age; and it must be inferred that the plaintiff is the identical woman named in the same certificate, although her residence was given as East Harrisburg when she actually lived in Hagerstown. Two such improbable inferences must thus take the place of “strict proof ” as to personal identity/ and to be availing those same inferences must overcome the counter presumptions of innocence and legitimacy arising from the subsequent marriage of G. Walter Bowman to Lettie E. Eakel—a thing they are powerless to do, because these latter presumptions yield only to evidence which establishes a mental conviction amounting to a moral certainty—and a probability can never produce a moral certainty.
The authorities are clear upon the proposition that there must be evidence of the identity of the parties beyond the mere statements contained in the certificate. “Where the evidence of marriage consists of the certificate, license, or marriage record, the identity of the parties must also be established by satisfactory evidence.” 19 Am. & Eng. Enc. of L.,, (2 ed.) 1200, and cases in note 3. A very apt illustration of the doctrine stated in the text is to be found in the case of Brooke v. Brooke, 60 Md. 524. It was shown there from the records of the church where the marriage was solemnized, that Henry Brooke and Margaret A. Ridgely were married on February 29th, 1872. After the death of Brooke a claim was made by Margaret that she was his widow. This was disputed by the executor of the decedent. It was shown by the latter that the parties never lived together as man and wife after the alleged ceremonial marriage, though Brooke did not die until 1879; and the contention was that some one else had personated Brooke at the ceremony. The case for the widow was not permitted to rest upon the mere identity or similarity of the names on the church register, nor upon the testimony of the clergyman that he had joined in wedlock two persons giving the names of the deceased and the claimant; but a photograph of Brooke, proved to be an accurate and genuine one, was produced and was identified by the persons who witnessed the marriage, as the likeness or picture of the man who was *292actually married to Margaret Ridgely. Upon that evidence it was decreed that she was the widow of Henry Brooke.
It is true, generally speaking, identity of names is prima facie evidence of identity of persons; but the names, of course must be identical, and this involves the identity of the Christian names—the identity of the initials thereof being insufficient. 15 Am. & Eng. Enc. L., 418, 919, and cases cited in notes. As already indicated George W. Bowman as named in the certificate is by no means identical with G. Walter Bowman the deceased, and no inference can be drawn that these two designations point out the same individual; especially when that dissimilarity is appealed to as evidencing an identity, which, if established in that way, would overthrow the strong presumptions in favor of innocence and legitimacy.
The next evidence relied on to prove identity is that given by Mrs. Zimmerman, the mother of the plaintiff. This witness testified as follows: “I met and knew George W. Bowman in July, 1887, at my house in Harrisburg, and my daughter, the plaintiff was with him. They stayed with me during the day and at night at the hotel for two or three days; and I afterwards visited them at Hagerstown in the fall of the same year at Walnut street in Hagerstown, where they were living together. No one else lived in the house with them, when they came to see me in Harrisburg they professed to come from Camden, New Jersey; and George W. Bowman said they had come from Camden, New Jersey, and he introduced himself as my son-in-law and he said he and my daughter were married in Camden, New Jersey.” It will be perceived, at a glance, that this testimony establishes no independent fact, other than that Bowman and the plaintiff stayed at the house of the witness in Harrisburg during the day and at an hotel during the night, for less than the space of a week; and that when the witness visited them at Walnut street in Hagerstown, the fall of the same year, they were living to ■ gether; but its chief significance lies in the recital of the declarations made by the supposed husband and wife as to their alleged marriage. If by this evidence as to the conduct of the *293parties it was designed to show that Bowman and the plaintiff cohabited as man and wife—that they were man and wife— and, therefore, were the identical persons who had been married in Camden, New Jersey, the conclusion thus sought to be deduced cannot be accepted as impugning the validity of the marriage to the defendant, Lettie E. Bowman, especially, as the conduct thus relied on was just as consistent with concubinage as with matrimony, when it is remembered that the house on Walnut street was a house of ill-fame and that the plaintiff was a prostitute. “A prior marriage will not be presumed from cohabitation, acknowledgment and reputation, where such presumption will render void a second marriage formally solemnized.” 19 Am. & Eng. Enc. L. (2 ed.) 1206, and cases in note 4. If, on the other hand, the declarations of the parties have been introduced and are relied on to establish personal identity, they are palpably unavailing and plainly inadmissible for that purpose. In the case of Le Brun v. Le Brun, supra, this Court said : “Marriage has been considered among all civilized nations, as the most important contract into which individuals can enter, as the parent, not the child of civil society. The great basis of human society throughout the civilized world, is founded on marriages and legitimate off-spring; and where an existing marriage is proved, it is not to be exposed to the danger of being set aside by any species of collusion, or by the mere declarations of either of the parties, and should only be brought in question upon the most undisputed proofs. Shelford on Marriage and Divorce, 573; Fornshill v. Murray, 1 Bland, 481; Piers v. Piers, 2 H. L. Cas. 331; Gaines v. Relf, 12 How. 472.” The association for the two or three days which Bowman and the plaintiff were together in Harrisburg and their occupancy of the same house in Hagerstown do not prove the fact of their marriage, nor tend to establish the identity of the individuals named in the certificate. Their association for those two or three days or their occupancy of an house of evil repute, is not “undisputed proof" of their marriage, when if such an effect is given to either or both of those facts, the issue of the second mar*294riage will be bastardized and the crime of bigamy will be imputed to and fastened on the decedent. Whilst this is not a suit for nullity of marriage, the degree of evidence as to identity demanded in this proceeding is no less rigid than in that. “In a suit for nullity of marriage, by reason of a former marriage, strict proof of the identity of the parties is requisite. It is a clear rule that the identity must be proved by other testimony than that of the parties themselves, that is, by witnesses who can speak of the facts from their own personal knowledge." Shelf ord, Mar. & Div., as quoted with approval in Le Brun v. Le Brun. The testimony of Mrs. Zimmerman has therefore no probative value.
The testimony of Dr. Wareham was as follows: “I knew the lady sitting there (indicating the plaintiff) at one time as the wife of George Walter Bowman. I was George Walter Bowman’s physician for about twenty years for his first and second wife. At the time I knew plaintiff she was living in Hagerstown on North Walnut street, in a house in which the plaintiff arid George Walter Bowman lived. At this time George Walter Bowman employed me to attend to her as his wife * . * *. I do not know who owned the house in which I attended her on Walnut street. All I know is that George Walter Bowman employed me to go there' and he was there. I do not know if whether or not he was living there. There was no one else living there to my knowledge, and I saw him there day and night. I cannot tell how often, and he paid me.” Leaving but of view the testimony which proved that the house on North Walnut street was a bawdy house and that the -plaintiff was an inmate thereof under the name of Edith Boyer, and that under that name she was arrested and fined for disorderly conduct in April, 1888—more than eight months after-her alleged marriage at Camden—leaving all this out of view, does the testimony of Dr. Ware-ham amount to anything more than his inference or opinion that Bowman and the plaintiff were man and wife—an inference or opinion drawn from or founded on the statements of Bowman ? As -Bowman’s own statements would be inadmis*295sible to prove identity the conclusions of a witness drawn from those same statements are equally incompetent. Hence this testimony furnishes no evidence of identity.
The plaintiff was offered as a' witness but was not allowed to testify. She was not a competent witness to prove the marriage either directly or indirectly, because Bowman was dead. Redgrave v. Redgrave, 38 Md. 96; Denison v. Denison, 35 Md. 361; 19 Am. & Eng. Ency. of Law, 1198.
The only other evidence proffered on this same subject was the testimony of Clyde B. Burst, a nephew of the deceased, and who would be entitled to one-half of the personal estate left by Bowman if the second marriage were annulled and the first one were upheld. In the first bill of exceptions he was asked to give a conversation had with his uncle, with a view of establishing by the uncle’s admissions his identity with the Bowman named in the marriage certificate. The Court was right in excluding the testimony. Identity cannot be proved by the admission of the parties, as has been already pointed out, “Identity must be proved by * * * witnesses who can speak to the facts from their own personal knowledge.” In the next exception the same witness was asked, “During this time was it or not known among the family that George Walter Bowman was married?” The question was not allowed to be answered. A marriage solemnized at a particular time and place was set up and relied on, and it was not competent to show by repute that there had been a marriage prior to 1900. It did not serve to identify the parties; and as a ceremonial marriage, alleged to have been performed at a designated time and place, was depended on, the proposed evidence, even if it tended to prove by repute that there had been a marriage at sometime, was irrelevant as reflecting on the one upon which the plaintiff relied. Where a marriage is set up as having been performed at a particular place by a particular form or ceremony, and the evidence fails to support the assertion, the party so alleging such a marriage will not be at liberty to rely on general repute to establish it. Jackson v. Jackson, 82 Md. 28; Barnum v. Barnum, 42 Md. 251. *296It was also inadmissible for the reason given in discussing the testimony of Mrs. Zimmerman, 19 Am. & Eng. Ency. L., 1206.
Inasmuch as there is not a shred of evidence in the record to identify the parties named in the New Jersey certificate, or to show that one was the decedent and the other was the plaintiff, the latter's contention must fall to the ground; and it becomes wholly immaterial what were the other rulings in the case.
The third, fourth, fifth, sixth, seventh, eighth and ninth exceptions were taken to the refusal of the Court to allow the plaintiff to testify in chief, and the fourteenth, fifteenth, sixteenth and seventeenth exceptions were reserved in consequence of the Court excluding the plaintiff as a witness in rebuttal. There was no error committed in any of those rulings. • The evidence sought to be introduced either tended to prove the identity of the plaintiff, or it did not. If it did, it was inadmissible because she was incompetent to prove that identity. If it did not tend to prove her identity it was irrelevant.
The ruling in the tenth exception was right. It merely permitted the- defendant to prove her marriage to Bowman. That was one'of the issues before the jury.
The rulings in the eleventh, twelfth and thirteenth exceptions admitted evidence adduced to show by what name the plaintiff was formerly known in Hagerstown; what course of life she had led there and upon what charges she had been arrested. If the evidence showed that the plaintiff was not the person named in the certificate, it was unnecessary as she had failed to prove that she was. If it did not do this, it was immaterial. In either event its admission did no injury'in view of the fact that the defendant’s first and second prayers ought to have been granted. There is no reversible error assigned in those exceptions,
The' prayers need not be discussed, because if those granted at the instance of the defendants were wrong, they did no injury, inasmuch as the plaintiff had wholly failed to prove the marriage upon which she relied, and as, in that view, she was *297not entitled to a verdict at all, she could not have been prejudiced by instructions which permitted a verdict to be rendered against her on other and different grounds. If the trial Court was abstractly wrong in rejecting the prayers of the plaintiff, no injury was done, inasmuch as by reason of her failure to furnish adequate proof of identity she had no standing before the jury; and she cannot complain because prayers, which proceeded upon the theory that such proof had been offered, were rejected, when they were rightly rejected for the reason that no such proof had been adduced.
(Decided June 21st, 1905.)Since the verdict was right as a result of the failure of proof on the point we have indicated, it will not be disturbed; and the rulings will be affirmed without deciding that they would have been, in all respects, free from error under a different state of facts as to the identity of the parties. We are not to be understood as committing ourselves to the accuracy of any of the rulings on the prayers, if the alleged first marriage and the identity of the parties had been satisfactorily proved, or if evidence tending to prove those requisites had been introduced. By affirming those rulings in this case we merely decide that, because the identity was not established, no reversible error has been shown to the prejudice or injury of the appellant. With this explanation the rulings are affirmed and the cause is remanded to the end that the findings of the jury may be certified to the Orphans’ Court of Washington County.
Rulings affirmed and cause remanded.