A jury, in the Noble Circuit Court, found “that the instrument in writing probated as the last will and testament of Charles Edward Compton, deceased, is invalid; that it is not his last will and testament and its probate should be set aside.” Judgment went accordingly. In reaching this verdict the jury, over the appellant’s objections, was permitted *705to consider an issue charging the appellant, Margaret McCartney, with having unduly influenced the said Charles Edward Compton in the execution of said will. This was error if, as the appellant contends, there is a total lack of evidence, direct or circumstantial, tending to prove such issue.

The law applicable to the question before us has long since been definiately established in Indiana. Application of the law, however, to various sets of facts has caused some of the decisions of our courts of appeal to seem inconsistent in their determination of what constitutes undue influence. Undue influence sufficient to void a will must be directly connected with and operate at the time of its execution with such force that the supposed will is in reality that of another and not of the testator. Wiley v. Gordon (1914), 181 Ind. 252, 104 N. E. 500; Workman v. Workman (1943), 113 Ind. App. 245, 46 N. E. 2d 718. However, the one having influence over the testator need not be actually present at the time and place of the preparation and execution of the will in order to vitiate it on such grounds if undue influence previously acquired still persists at such time so that, but for it, the will would have been different from that actually executed. Cooper v. Cooper (1943), 114 Ind. App. 261, 51 N. E. 2d 100. The influence that will vitiate a will must be such, as in some degree, to destroy the free agency of the testator and constrain him to do what is against his will, but what he is unable to refuse or too weak to resist. Rabb v. Graham (1873), 43 Ind. 1; Noyer, Exr. et al. v. Ecker et al. (1954), 125 Ind. App. 63, 119 N. E. 2d 902; Ludwick, Extr. et al. v. Banet et al. (1955), 125 Ind. App. 465, 124 N. E. 2d 214.

*706*705The Supreme Court of this State, in Davis, Exr. v. Babb (1921), 190 Ind. 173, 125 N. E. 403, and this *706court, in Love v. Harris (1957), 127 Ind. App. 505, 143 N. E. 2d 450, quoted with approval the following excerpt from the court’s decision in Blackman v. Edsall (1902), 17 Colo. App. 429, 68 Pac. 790: “‘It follows from the very nature of the thing that evidence to show undue influence must be largely in effect circumstantial. It is an intangible thing which only in the rarest instances is susceptible of what may be termed direct or positive proof. The difficulty is also enhanced by the fact universally recognized that he who seeks to use undue influence does so in privacy. He seldom uses brute force or open threats to terrorize his intended victim, and if he does he is careful that no witnesses are about to take note of and testify to the fact. He observes, too, the same precautions if he seeks by cajolery, flattery or other methods to obtain power and control over the will of another and direct it improperly to the accomplishment of the purpose which he desires.’ ” And again in the same case the court said: “ ‘Only general rules concerning the amount and character of evidence required to establish undue influence in the execution of a will can be laid down. As to what is sufficient must depend upon the facts and circumstances of each particular case.’ ”

Professor Page, in his distinguished work on Wills, Yol. 2, §811, says: “From the nature of undue influence, from the fact that it is frequently employed surreptitiously, that it is shown chiefly by its results, and that in any event, the question is usually one of the effect of a long course of conduct upon the mind or the individual in question, the evidence by which undue influence is established is usually circumstantial. Undue influence is essentially a question of fact. The party who contests the will is entitled to the benefit of all inferences of fact which may be deduced, fairly and reasonably from the *707direct evidence. Direct evidence is necessary only to establish the fact from which undue influence may reasonably be inferred. Although each fact by itself may be insufficient, the facts when taken together may justify, and even require, a finding of undue influence.”

Nevertheless it should be borne in mind that relations arising out of the affection and confidence that naturally exist between a testator and his daughter constitute no evidence that said daughter exerted an undue influence which induced him to make a will which he otherwise would not have made, since the influences which naturally arise from family relationships are not undue or unlawful and should not be confounded with positive dictation and control exercised over the mind of the testator. Cooper v. Cooper, supra.

Charles Edward Compton spent most of his life in Ligonier, Noble County, Indiana, where he accumulated a modest estate which was appraised after his death in the sum of $14,484.74. He was married and had two daughters, Margaret McCartney, the appellant, and Bernice Rex, the mother of the appellees, who was killed in an automobile accident in 1952. His wife died in 1951, after which he went to Chicago and lived with his daughter Margaret and her husband for a short time. Early in 1952, he and the McCartneys returned to Ligonier where they lived together in the old family home. The McCartneys moved back to Chicago in August, 1952, but Compton remained in Ligonier usually spending weekends in Chicago with his daughter Margaret. He had no help in his house in Ligonier and Margaret came down from Chicago every two or three weeks to clean the house, take care of the laundry and put her father’s clothes in order. In May, 1953, under Margaret’s supervision, all of Compton’s household goods were sold at public *708auction and his home permanently closed, a proceeding which greatly stirred his emotions but there is no evidence indicating that the event took place against his will. He was somewhat reluctant to go to Chicago to live after giving up housekeeping in Ligonier and did so only with Margaret’s assurance that he could come and go as he pleased. He lived with Margaret from May, 1953, until he died on January 8, 1955, at the age of 83. He left a last will and testament dated October 22, 1954, by the terms of which his daughter Margaret is given his entire estate to the total exclusion of the appellees herein, sons of his deceased daughter Bernice. There is evidence tending to prove that at the time he made said will he was a person of unsound mind and far from vigorous physically. On the record herein and the evidence disclosed thereby we think the most that could be said for the appellees’ case based on undue influence is that at the time Charles Edward Compton made his will he was weak mentally; that the appellant Margaret McCartney had ample opportunity to exercise undue influence over him and that she profited by his will at the expense of his other heirs at law. The record is devoid of evidence that the appellant did anything from which an exercise of undue influence over the testator could have been inferred. Therefore, in our opinion the issue of undue influence should have been withdrawn from the jury. The court’s refusal to do so constituted error.

Judgment reversed and cause remanded to the Noble Circuit Court with instructions to sustain appellant’s motion for a new trial.

Note. — Reported in 145 N. E. 2d 400.