action in which it was claimed the testator had made a bad or foolish bargain, or any instance in which the testator had foolishly spent or squandered his money. Neither of these offers was accepted by the contestant. The record fairly discloses that the testator was possessed of a considerable amount of Detroit real estate, which was in the main, if not wholly, unproductive, and that he had no fixed income, except for a few years a small amount received from the sale of gravel taken from one of the Detroit lots. This amount seems to have been inconsiderable. It is apparent that to preserve the body of his estate, pay taxes, and exist himself, it would become necessary for him from time to time to sell some portion of it. This he appears to have done. There is nothing in the record tending to show, nor is there any offer of proof to the effect, that the testator sold more of his estate than was necessary for his decent maintenance and for the preservation of what he retained.

While we have held that good business ability is evidence of mental capacity to be properly considered in will cases, we have never held that, because a testator had not in his possession at death the purchase price of property sold by him during his lifetime, that fact might be considered as tending to show incapacity to make a will. We are of opinion that the deeds were properly excluded.

It is strongly urged by counsel for contestant that reversible error was committed by the learned trial judge in withdrawing from the jury consideration of the question of undue influence upon the testator's mind exercised by Mr. and Mrs. Jones. In this connection it is proper to state that there is not a word of affirmative testimony in the record showing or tending to show that either Mr. or Mrs. Jones had at any time asked the testator to leave them, or either of them, any portion of his property. Both were placed

upon the stand by counsel for the proponents, and both categorically denied that they had ever sought to influence the testator either duly or unduly. An inference of such undue influence is sought to be drawn by counsel for contestant from several sources. They say that the testator was very nervous, did very little business, and that his will power was weak. They point to the relations between the testator and Mr. and Mrs. Jones, and argue that both had the opportunity to exercise undue influence over the mind of the testator, and that the result shows that they did exercise such influence. They urge that the clause in the will in which the testator distinctly negatived any such influence on the part of Mr. and Mrs. Jones, and recited that the relations existing between himself and them had always been of the most honorable character, "is capable of a construction that would lead to an inference of undue influence and furnish another circumstance to be weighed by the jury in determining that question."

We do not lose sight of the fact that undue influence need not be proven by direct evidence, but can be established by indirect and circumstantial evidence. This principle has been laid down in many Michigan cases, the last two of which are In re Loree's Estate, 158 Mich. 372 (122 N. W. 623), and In re Du Bois' Estate, 164 Mich. 8 (128 N. W. 1092). But we are of opinion that the contestant must introduce evidence from which inferences may fairly be drawn that such influence was exercised. The fact that Jones and his wife were made beneficiaries under the will is not alone sufficient nor is contestant aided by the fact that it is shown that testator and Jones and his wife were intimate friends for many years. The latter consideration serves rather to show why Jones and his wife became objects of the testator's bounty. Maynard v. Vinton, 59 Mich. 139 (26 N. W. 401, 60 Am. Rep.

276); Severance v. Severance, 90 Mich. 417 (52 N. W. 292); Lamb v. Lippincott, 115 Mich. 611 (73 N. W. 887). Opportunity alone cannot give rise to a valid inference that undue influence has been exercised. In re Shanahan's Estate, 176 Mich. 137 (142 N. W. 573); In re Foerster's Estate, 177 Mich. 574 (143 N. W. 616).

The record discloses the fact that the testator made an earlier will on January 27, 1900, almost exactly nine years prior to the date of the will in question. In that will he devised to Jones and his wife certain real estate, or in lieu thereof the sum of $30,000. The will itself was placed in the custody of Jones by the testator, with injunction to read it. This Jones did in the presence of his wife, and, so far as he knew, no other will had been made. Although not material, it seems that the provision made for Jones and his wife in the latter will is not so liberal as that in the earlier one. The conduct of Jones and his wife toward the testator during the last 12 years of his life must be viewed in the light of their knowledge of the fact that they were to be beneficiaries under the last will of the testator at his death. A very succinct and exhaustive discussion of the question of undue influence is expressed in the case of Ginter v. Ginter, 79 Kan. 721 (101 Pac. 634, 22 L. R. A. [N. S.] 1024), as follows:

"To vitiate a will there must be more than influence. It must be undue influence. To be classed as 'undue,' influence must place the testator in the attitude of saying: 'It is not my will, but I must do it.' He must act under such coercion, compulsion, or constraint that his own free agency is destroyed. The will or the provisions assailed does not truly proceed from him. He becomes the tutored instrument of a dominating mind, which dictates to him what he shall do, compels him to adopt its will instead of exercising his own, and, by overcoming his power of resistance, im

pels him to do what he would not have done had he been free from its control. A testator's favor expressed in a will may be won by devoted attachment, selfsacrificing kindness, and the beneficent ministrations of friendship and love. These influences are not undue. We expect partiality to attend them. They bring preferment as their natural reward, and they do not become unrighteous, although they establish a general ascendancy over the testator, leading him to find comfort and pleasure in gratifying the wishes and desires of the person exercising them. Other less worthy influences may make equally strong appeals and may result in the same general dominion and still be sufferable in contemplation of the law. Influences to induce testamentary disposition may be specific and direct without becoming undue. It is not improper to advise, to persuade, to solicit, to importune, to entreat, and to implore. Hopes and fears and even prejudices may be moved. Appeals may be made to vanity and to pride; to the sense of justice and to the obligations of duty; to ties of friendship, of affection, and of kindred; to the sentiment of gratitude; to pity for distress and destitution. It is not enough that the testator's convictions be brought into harmony with that of another by such means. His views may be radically changed, but so long as he is not overborne and rendered incapable of acting finally upon his own motives, so long as he remains a free agent, his choice of a course is his own choice, and the will is his will, and not that of another. * * 'It is not the influence of friendship or affection that can be complained of, nor the influence of argument or entreaty, nor the impression made by kindness or prudence, nor even the effect wrought by servile compliance or mean endurance of wrong. It must be something which destroys free agency. Motives of almost every conceivable kind may be offered, and if the mind of the agent, free to reject or adopt the motives, yields its assent, the act is the act of the agent.' Means v. Means, 5 Strob. (S. C.) 167, 192.

"In order to cause a will or deed to be set aside on the ground of fraud and undue influence, it must be established to the satisfaction of the court that the party making it had no free will, but stood in vinculis.'

Conley v. Nailor, 118 U. S. 127, 6 Sup. Ct. 1001. "Upon contest of [a] will for undue influence, the question is "whether the will is the will of the testator, or that of another." It is not influence that vitiates, but undue influence; and it must go to the extent of depriving the testator of his free agency, and amount to moral coercion which he is unable to resist.' Peery v. Peery, 94 Tenn. 328 (29 S. W. 1).

""The influence which the law denominates undue, and which vitiates a will executed under it, must amount to moral or physical coercion, destroying free agency, and constraining its subject to do that which, but for it, he would not do.' Westcott v. Sheppard, 51 N. J. Eq. 315 (25 Atl. 254, 30 Atl. 428).

"Undue influence, such as will invalidate a will, must be something which destroys the free agency of the testator at the time when the instrument is made, and which, in effect, substitutes the will of another for that of the testator. It may be exercised through threats, fraud, importunity, or by the silent, resistless power which the strong often exercise over the weak and infirm; but, however exercised, it must, in order to avoid a will, destroy the free agency of the testator at the time it was made, so that the instrument in fact expresses the mind and intent of some one else, and not his own.' Schmidt v. Schmidt, 47 Minn. 451 (50 N. W. 598)."

The record before us fails to establish the power of Jones and his wife to unduly influence the testator, which, even if shown, would be insufficient. Maynard v. Vinton, 59 Mich. 139 (26 N. W. 401, 60 Am. Rep. 276). The most that can be said of the evidence in the case at bar is that, through constant association and friendly acts, Jones and his wife had the opportunity to exercise an unlawful and undue influence over the testator. That they did so is not to be inferred from the evidence adduced. Indeed, the record discloses that, for many years prior to the death of the testator, Jones devoted a very considerable portion of his time to the interests of the testator without compensation, but under the assurance that he