This is a will contest. After a trial resulted in a hung jury, the circuit judge granted judgment upholding the will under CR 50.02. On this appeal contestants contend they were entitled to another jury trial on three grounds.
The principal controversy concerns a holographic will apparently executed by the decedent, a middle aged widow, on July 1, 1946. She subsequently executed a codicil on April 26, 1952, properly attested by two witnesses. She died about five years later.
The beneficiaries of the will were á niece and nephew, both of whom had lived with her for several years. The nephew had resided with her since 1937, at which time he was 11 years of age, and he received the larger share of the estate. The codicil left $100 each to other nieces and nephews.
Before reaching the grounds upon which the will is attacked, it is necessary to consider a question not decided by the trial court but here raised by appellees. That question is whether the execution of the codicil six years after the purported will constituted a republication of the latter. If so (assuming the codicil was valid), it would be unnecessary to consider appellants’ contentions.
A defective will may be validated by a duly executed codicil. 57 Am.Jur. Wills § 626, page 428; 95 C.J.S. Wills § 303, page 95; Spradlin v. Adams, 182 Ky. 716, 207 S.W. 471; Farmers’ Bank & Trust Co. v. Harding, 209 Ky. 3, 272 S.W. 3; Hurley v. Blankinship, 313 Ky. 49, 229 S.W.2d 963, 21 A.L.R.2d 817. As pointed out in the foregoing authorities, perhaps the most important factor is identification of the earlier will. If a writing is to be validated as a will by a later formal document, the latter should clearly incorporate the former. The general principles of incorporation by reference apply. See Daniel v. Tyler’s Ex’r, 296 Ky. 808, 178 S.W.2d 411.
In the present case the codicil recited: “I wish to add this codicil to my will.” This recitation immediately raises the question: What will? There is no reference to the date of the will, or its terms, or its whereabouts. It has been held that such a reference by itself is not sufficient identification. In re Foulds’ Will, 21 Misc.2d 402, 196 N.Y.S.2d 816.
This conclusion is consistent with the principles set forth in the authorities heretofore cited. The uncertain character of this reference is more evident when we consider the fact that the will in question was executed six years prior to the codicil and there is a real controversy in this case as to whether the former was written and signed by the testatrix. She could possibly have written other wills.
It is true that a written reference to a former will is not the only means of identification. Other circumstances, such as a codicil written on the same paper as the will (Farmers’ Bank & Trust Co. v. Harding, 209 Ky. 3, 272 S.W. 3), or the physical attachment of the will to the codicil (Hurley v. Blankinship, 313 Ky. 49, 229 S.W.2d 963, *33121 A.L.R.2d 817), may serve to identify the-earlier testamentary paper.
An attempt was made in this case to prove the will and the codicil were put together in a sealed envelope, but assuming this would constitute sufficient incorporation, the evidence did not establish the fact. The only identification of the two instruments as a single testamentary act of the decedent is the reference in the codicil to “my will”. Under the circumstances shown, we do not think this validated the will by republication.
Returning to the will, appellants’ first ground of contest is that the decedent lacked mental capacity. Evidence offered on this point was that she lost by adverse possession a tract of real estate which her husband had deeded to her. The argument is made this proves she did not know the extent of her property. A rational explanation of the occurrence was given by appel-lees’ witnesses, but even if unexplained, we cannot comprehend how this title development had probative value that decedent lacked mental capacity.
An attempt was made to show that shortly before the execution of the codicil the decedent was ill and “didn’t know what she was doing”. Hospital records were offered in evidence (and refused) that showed after she executed the codicil she was “talking irrational”. These isolated events occurring almost six years later had no reasonable tendency to indicate decedent’s mental condition at the time she executed the will. See Pardue v. Pardue, 312 Ky. 370, 227 S.W.2d 403.
The next contention is that the nephew exercised undue influence and coercion. The principal basis of this claim is that the nephew on occasion would physically beat up his aunt. There is some evidence that the decedent was afraid of her nephew and that he wanted her to make over her property to him. However, the nephew’s brutal and peculiar acts were in no way tied up with the execution of the will. There is no showing he even knew of its execution.
Any slight probative value of this evidence on the issue of undue influence is dispelled by significant uncontradicted facts. The will itself divided decedent’s property between this nephew and a niece, both of whom were natural objects of her bounty. There was uncontradicted evidence the aunt had always intended to leave most of her property to these two. In addition, she made no apparent attempt to revoke this will for a period of over 10 years. Finally, the execution of the codicil six years later, wherein bequests were made to others, is entirely inconsistent with the theory that the nephew had coerced the disposition of her property. Under the circumstances shown, we are of the opinion that the proof of sporadic violence against decedent’s person cannot reasonably be converted into substantial evidence of undue influence.
We finally reach appellants’ contention that this holographic will was not wholly in • the handwriting of the decedent. On its face the paper shows what appears as two different types of handwriting, or at least the use of two different pens. Appellants introduced witnesses who compared the signature on this paper with other known signatures of the aunt and stated they were not the same. In addition, they offered to call some of the contestants to testify that they were familiar with the handwriting of the decedent and that the signature on the paper was not hers. The court would not permit appellants to introduce this evidence, apparently on the ground it violated KRS 421.210(2) (formerly Civil Code Section 606(2)).
We believe this testimony was clearly competent. Essentially it constituted opinions of the witnesses wholly unrelated to a specific transaction, act or omission of the decedent observed by or communicated to the witness. The purpose of the statute is to exclude testimony recounting specific *332conduct of the deceased which inures to the benefit of the witness.
In Kendall v. Hillsboro & P. P. Turnpike Road, 23 K.L.R. 2372, 67 S.W. 376, and Ferguson v. Billups, 244 Ky. 85, 50 S.W.2d 35, interested parties were permitted to testify to the contents of a lost document, their testimony being based on what they had observed independently of any act of the deceased. On the other hand, in Gibbs v. Terry, Ky., 281 S.W.2d 712, 713, and Whitcomb v. Whitcomb, Ky., 267 S.W.2d 400, it was held that a beneficiary could not testify concerning the contents of a will when the source of the information was an act of the testator in the presence of the witness. It is evident from these cases that the prohibited testimony must at least involve a fact arising out of the decedent’s conduct on a particular occasion and not a fact that may be established independently of it.
Even if the fact in issue concerns the general conduct of the deceased in the presence of the witness, opinion testimony as to the mental capacity of that person is admissible to the extent it is based upon observable conditions. Wilbur’s Ex’r v. Lemmon, 306 Ky. 675, 209 S.W.2d 82.
We have had one case involving the testator’s handwriting. In Nelson v. Nelson, 235 Ky. 189, 30 S.W.2d 893, it was held the widow of the testator could properly testify a will being contested was wholly in his handwriting. This decision, however, was based on the assumption that the statute did not apply to will contests. In Hale v. Hale, 242 Ky. 810, 47 S.W.2d 706, this assumption was repudiated. See also Whitcomb v. Whitcomb, Ky., 267 S.W.2d 400.
Nevertheless, we have no hesitancy in- concluding that testimony with respect to the handwriting of a deceased person based on observation and analysis of a document is not within the prohibition of the statute.Such conclusion is based upon two grounds: (1) the testimony does not recount or describe a specific transaction or act of the deceased, and (2) it constitutes an opinion of the witness based on physical evidence which does not purport to portray a particular event. The trial court therefore erroneously excluded proffered testimony of the contestants with respect to the apparent authenticity of the decedent’s handwriting.
This evidence, along with testimony that was admitted, clearly created a jury issue as to whether or not this will was wholly in the handwriting of the deceased. The judgment must be reversed for a new trial on this issue.
The judgment is reversed with directions to grant appellants a new trial on the sole issue of whether or not the holographic will tendered for probate is wholly in the handwriting of Iva Hendren Burton.