The opinion of this court was delivered by
Gibson, C. J.
So far as regards wills consummated by the testator’s death — and this is one of them — the act of 1848 is founded on no power known to the constitution, but on the assumption of a power appropriated exclusively to the judiciary. Every tyro or sciolist knows that it is the province of the legislature to enact, of the judiciary to expound, and of the executive to enforce. These functions may, if the people will it, be performed by a single organ; but the people of Pennsylvania have not so willed it. They have ordained that the judicial power of the Commonwealth be vested in a Supreme court, in County Courts of Common Pleas, Oyer and Terminer, and Quarter Sessions, in a Register’s Court, and an Orphans’ Court: and in such other courts as the legislature may from time to time establish. But the judicial power of the Commonwealth is its whole judicial power; and it is so distributed, that the legislature cannot exercise any part of it. Under the constitution, therefore, there is no mixed power — partly legislative and partly judicial — such as was recognised in Bradee v. Brownfield. Bid it exist, it could be exercised only in concert or in common; for it would give the judiciary as much right to legislate as it would give the legislature right to adjudicate. Thus blended, I know of no constitutional power, principle, or provision, that would give to either of them, as a co-ordinate branch, an exclusive right to the whole. What then did the legislature propose by the statute of 1848 ? This court had ruled in Asay v. Hoover directly, and in Barr v. Strobell, Cavett’s Appeal, and perhaps Hays v. Harden, incidentally, that a testator’s mark to his name, at the foot of a testamentary paper, but without proof that the name was written by his express direction, is not the signature required by the act of 1833; and the legislature has declared, in order to overrule it, that “ every last will and testament heretofore made, or hereafter to be made, except such as may have been finally adjudicated prior to the passage of this act, to which the testator’s name is subscribed by his direction, or to which the testator has made his mark or cross, shall be deemed and taken to be valid.” How this mandate to the courts, to establish a particular interpretation of a particular statute, can be taken for anything else than an exercise of judicial power in settling a question of interpretation, I know not. The judiciary had certainly recognised a legislative *495interpretation of a statute before it had itself acted, and consequently before a purchaser could be misled by its judgment; but he might have paid for a title on the unmistakeable meaning of plain words; and for the legislature subsequently to distort or pervert it, or to enact that white meant black, or that black meant white, would, in the same degree, be an exercise of arbitrary and unconstitutional power. All ex post facto laws are arbitrary; and it is to be regretted that the constitutional prohibition of them has been restricted to laws for penalties and punishments. In a moral or political aspect, an invasion of the right of property is as unjust as an invasion of the right of personal security. But retroactive legislation began and has been continued, because the judiciary has thought itself too weak to withstand; too weak, because it has neither the patronage nor the prestige necessary to sustain it against the antagonism of the legislature and the bar. Yet, had it taken its stand on the rampart of the constitution at the onset, there is some little reason to think it might have held its ground. Instead of that, it pursued a temporizing course till the mischief had become intolerable, and till it was compelled, in Norman v. Heist and Bolton v. Johns, to invalidate certain acts of legislation, or rather to reverse certain legislative decrees. Conceding the right of legislative interpretation in the first instance, because it has prevailed too long to be disputed, we can pronounce the act of 1848 to be exclusively prospective without disturbing titles.
It is destitute of retroactive force, not only because it was an act of judicial power, but because it contravenes the declaration in the ninth section of the ninth article of the constitution, that no person shall be deprived of life, liberty, or property, except by the judgment of his peers or the law of the land. Taking the proof of execution, at this stage of the argument, to be defective under the act of 1833, it would follow that the plaintiff had become the owner of a third of the property in contest, by the only assurance that any man can have for his property — the law. Yet the legislature attempted to divest it, by a general law it is true, but one impinging on particular rights. Still it is argued that the act may be sustained as a confirmation of conveyances by will, as a confirmation of conveyances by deed was sustained in Underwood v. Silly, Mercer v. Watson, and other cases of the class. It was remarked in Menges v. Wertman, that a party, who has received a benefit from a transaction, is under a moral obligation to convey, and that the legislature may add a legal one to it; and I still think that a distinction between a purchaser and a volunteer, is the only ground *496left us to found a practical limitation of special legislation. In this case, the devisee is a volunteer, and the heirs are hound by no obligation which did not bind the legitimate heirs in Norman v. Heist. But the great obstacle in Menges v. Wertman, was the number of titles that depended on legislation of the same stamp. I have doubted whether we ought not to have swept them all away; but we had a choice of evils set before us, and want of steadiness in the judiciary was thought to be the greater one. Decision, however, has not established a rule of property to control us in the present case; for judicial interpretation of the act of 1833 had preceded the act of 1848. The statutes in Hess v. Wertz and Satterlee v. Mathewson were enacted, not to create a contract or to confirm one, but to remove a disability; and the protection not of a party, but the public, was certainly within the constitutional range of legislative action.
The defendant’s next position is, that the testatrix’s name, written by another, without her express direction, but acknowledged by her mark, was within the true intent and meaning of the original act, which requires that “ every will shall be in writing, and unless the person making the same shall be prevented by the extremity of his last sickness, shall be signed by him at the end thereof, or by some person in his presence and by his express direction.” But this express direction, like signing by the testator in the presence of the attesting witnesses, and attestation by them in his presence and in the presence of each other, under the English statute of frauds, is a substantive part of the act of execution prescribed by the statute under consideration; and where it is not proved, expressly or presumptively, by the oaths or attestation of two witnesses, it is not proved at all. Why use emphatic words, if there was no design to distinguish between an express and an implied direction ? Though express direction may be proved by presumptive evidence, it follows not that a subsequent act of ratification by the mark, is presumptive evidence of it. A direction precedes the act to be done in obedience to it; and in this respect a direction expressed in words, differs from a direction implied from subsequent assent. There is a maxim that ratification is equivalent to command ; but it was held in Dunlop v. Dunlop, not to be express direction. The question at this stage of the inquiry is not, whether the attestation may not be evidence of previous direction, but whether the mark is. It is not disputed that it is evidence of ratification: but what magic is there in ratification by the testator’s pen, which is not found in ratification by his tongue, which was held not to be express direc*497tion in Dunlop v. Dunlop ? Does the mark constitute the signature ? or does the name ? or do both together ? Take it that the signature is only a sign; still, to import anything, it must be a sign of something. It must have an intrinsic or a conventional meaning without being interpreted by anything appended to it. A name without a mark to it, individuates the person, and may express his assent to an act done by him: a mark without a name to it, individuates or expresses nothing. A cross, or a scratch, or a scrawl, or a dot, or a dash, or a flourish, unassisted by the name, imports no more than would a blot, or a stain, or any other accidental discolouration of the paper at the foot of the instrument. When few could write, a seal was the more frequent symbol: it signified the name of the party as intelligibly as if it had been expressed by a combination of letters, which also is no more than a conventional sign of it. A naked mark is not a signature at the common law; and the statute was not designed to make it so. ' It requires the whole signature to be written by the hand of the testator, or by the hand of a bystander — not a part of it by the one and a part of it by the other — and for the bystander to write the name and the testator to make the mark is no more a fulfilment of its requisition than if their agency in the transaction was inverted, or the bystander had, without express direction, written the one and made the other. Yet if the mark were an integrant part of the signature, it would have to be made by somebody, though the testator were too feeble to put pen to paper. If it is not an integrant part, what office does it perform? As an act of verification, it is as insignificant when made by the testator as when made by a stranger. No one would pretend that a dot would be enough, and the statute consequently requires the name; and as it sanctions nothing which it does not enjoin, a mark is surplusage, and consequently ineffective. It substitutes for it the name written by the testator’s express direction, as a more rational act of authentication, in order to dispense with it. Even without a statute to help it, the seal of a party, affixed by another in his presence and with his assent, was held, in Hart v. Withers, 1 Penn. R. 285, on the authority of Ball v. Dunsterville, 4 T. R. 313, to be his immediate act. As signing by the testator’s assent would have been good at the common law, the statute was enacted not to authorize it, but to regulate the evidence of it by requiring more than a wink or a nod, or a word not less ambiguous, and therefore not less liable to misconstruction or misrepresentation. The purpose of it was to have a straightforward direction expressed in terms, which would leave no pretence *498for the touch of an insensible or dead man’s hand to give colour to an artful tale told by willing witnesses. In other transactions, the mark is sometimes used as a badge of assent; but the assent required by the statute is to be signified, not by a badge attached to the name, but by a direction to attach the name to the paper. Were we to receive the implication of a direction as an equivalent for the express direction demanded by the statute, we should overrule Dunlop v. Dunlop, in which the doctrine of equivalents was rejected, and Burwell v. Corbin, 1 Rand. 131, which is a strong authority for the principle. The judgment in the latter was founded on the construction of a statute in Virginia, enacted in the words of our own. The will was proved by one of the attesting witnesses, to have been signed with the testator’s name by his express direction ; and the other, who had not been present, but attested it the next day, proved that the testator acknowledged it to be his will; and, though this recognition by words was more explicit than recognition by a mark, the proof was held to be deficient. After all, therefore, it is the name written by the testator’s express direction which constitutes the signature ; and the question comes to this: was the positive testimony of one of the attesting witnesses, and the attestation of the other, in the absence of his recollection, proof of it ?
Where the oath cannot be had, it is settled that the attestation is a substitute for it, and proof of compliance with the requisitions of the statute. Such is the principle of Hands v. James, 2 Comyn, 531, Provis v. Reed, 5 Bingh. 435, and 5 Ves. 504. In Croft v. Pawlett, 2 Stra. 1109, the attestation was in the usual form, except that it was not said in it that the witnesses had signed in the testator’s presence; and an an objection that it stood for proof of no more than the facts to which they subscribed, was disregarded. Even where the witness is present, but his memory is not, his attestation stands for proof. In Dayrell v. Glasscock, Skin. 413, Lord Holt held it sufficient, in the first instance, to prove the attestation of a witness, who would not swear to the sealing and publication. Pretty much to the same effect is Turnipseed v. Hawkins, 1 M’Cord, 272; Brice v. Smith, Willis, 1; Fetherly v. Waggoner, 11 Wend. 599, and Fitzhubert v. Fitzhubert, 4 B. C. C. 231. These prove very satisfactorily that the attestation of a witness dead, blind, insane, subsequently incompetent, or out of the reach of process, is primd facie evidence of every necessary fact. But what avails it that the man is living, if his memory is dead ? If it were blotted out by paralysis, or worn out by decay, his attestation *499would stand for proof by a witness ; but it must be immaterial how, or by what means, it lost its tenacity. The law of double proof would place wills on ticklish, ground, did it leave them to depend on the incorruptibility of attesting witnesses, or on their exemption from growing infirmity. Even where they flinch from their attestation, it satisfies the demands of the statute in the first instance, and is evidence to confront them before a jury. After the testator has executed his will in confirmity to the requirements of the law, it would be a mockery of his right of testamentary disposal, to exclude presumptive or secondary evidence of a compliance with forms, that would be received in cases of another nature. I lay out of the case the supposed effect of the probate, which, though it might entitle the paper to be read in the first instance, cannot shift the burthen of proof in a question like the present. As a special case seems to be made for the opinion of the court, and not the jury, I am of opinion that the attestation was primd facie evidence, and sufficient.
Judgment affirmed.