delivered the opinion of the court ':

The decision of the Circuit Court, by which this evi*522deneé was rejected, is predicated on the rule of law which excludes parol evidence calculated to add to, vary or controul the provisions of a will, f Phillips, 479.^1 Of the correctness and the salutary effects of this rule there is no question, but 1 think its want, of application to the case under consideration is equally clear. The error appears to inc to have originated in applying to this paper the appellation of will. The law writers define a will to he a complete and legal declaration of a man’s intentions as to ■what he would have done with his estate after his death. Does this paper comport with this definition ? is it a complete and legal declaration ? To determine this question, when there is any positive rule, we must look to that as our guide ; as for instance, to a will of real estate, the statute requires that the name of the testator should be signed, and that it should lie attested by three witnesses- .Without these, it wants that complete and legal character required by the definition to effectuate the object of the testator. In the absence of any positive rule, we must resort to the usages of mankind to determine it. Thus, when we speak of a bond or promissory note, we know that usage has rendered a seal indispensable to the character of the first, and the signature of the drawer to the latter, and without these appendages, they would want the most essential constituent of (he appellation given to them ; and so of a will, it was never heard of; that a man of ordinary understanding sat down to make a will, professed-]y with an intention not to subscribe his name to it. Such a thing rarely, if ever, happened. When we speak, therefore, óf a will as containing a complete and legal declaration, we do it with reference to that, circumstance. His signature is the act by which he seals his sanction to the operation it is intended to have ; it precludes the idea that any thing more was intended; it is the factum of the thing itself; and in the language of the definition, it isa complete and legal declaration of his intention; and in such a case, the rub- would operate in full force; but not so with respect to that to which this finishing stroke is not *523given. We cannot know' bu¿ ¡Von: extrinsic circuir st. ■ ' c-s whether the testator did or did not intend to give oí¡ or legacies, or to charge those already given, and from thwr Very nature, they can only be proved by parol. It is not intended to lay down the position that a paper which wants the signature oía testator, can, in no case, operate as a will ; on the contrary, there are numberless cases in which they have been admitted to probate, but always on proof of extrinsic circumstances going to show the animus tesiandi ; as in the common case of the testator’s being prevented from signing by his sudden death ; of its being found amongst his valuable papers, and a variety of others; but in all these cases, it will be found to need the aid of those circumstances to give it effect. There may be cases which would constitute an exception to this rule, as when it was expressed, (if such a case over did happen) that the testator intended it should operate without his signature, or when it was apparent from the paper that he intended it should so operate; none of which appear to me, however, to apply to this case. From this view of the case, it is apparent that the rule which governed the judgment of the Circuit Court was not applicable ; and iu the absence of authorities, 1 have found some difficulty in referring the case itself to its proper principle. It appears to me, however, to fall within the rule which admits of parol evidence to explain doubts created by extrinsic circumstances, (Phillips, 467.) On this principle, in the case of Beaumont vs. Fell, (2 P. Wms. 140,) the court permitted parol evidence to shew that Gertrude Yardly was the person to whom a legacy was bequeathed by the name of Catherine Iiarnly. There, the circumstance that there was no such person as Catherine Earnly w7as extrinsic, and so in this case, the will being incomplete, the fact, whether he intended it to operate as such was also extrinsic, and could only be supplied by parol. I have not been able to discover that this question has ever before been made, and indeed the solution appears so palpable as ‡0 lead to the conclusion that it never has. The cases in *524which such evidence has been admitted, without opnosition, are almost numberless ; and so far as precedent can operate to establish a will, none is better established than that it is admissible. Indeed, almost ali the. contests about admitting wills to probate, have arisen on those presented in an unfinished state, where the factum of the will and the animus testandi were the principle grounds of contest, and which, in the absence of written evidence, could only be proved by parol. To mention one only, it was admitted in the case of Scott vs. Rhodes, (1 I hit. 12,) to show the animus testandi in relation to an unfinished will, although it was apparent on the face of the will itself, that the testator intended to do more, to give it effect. 1 think therefore that the evidence tendered was improperly i ejected, and that a new trial ought to b% granted.