The executors of Paul Garrett appeal from- petition to review- an order of the Tax Court, in banc, which included in whole or in part, the trust funds of two trusts, set up by Garrett, their testator: one in 1923, and the other in 1929. Although the settlor died in 1940, the Tax Court held that both trusts were made 'in contemplation of death' within the meaning of the statute,1 and the case hinges on that finding of fact. We shall consider the two trusts separately and in their sequence in time.

The 'Trust of 1923'

In 1923 Garrett had a wife, a son twenty years old, and three young unmarried daughters. For many years he had been a successful vintner; and, although Prohibition had much reduced the extent of his business, he had continued in it and remained its active manager. He was 59 years old and in entire good health, and so he remained until his death. The 1923 trusts were in two parts, set up by separate deeds and having different limitations. We shall disregard one of these called 'Trust No. 2,' for the issues at bar concern only the other, and, indeed, only a part of the trust fund in that. 'Trust No. 1' was made up in part of income producing securities, and, for the rest, of thirty policies upon Garrett's life, aggregating $250,000, of which $30,000 were fully paid. These policies he had taken out from time to time, beginning in 1890, and after 1901 at short intervals until the very eve of the creation of 'Trust No. 1.' The securities which he added to these policies to make up the trust fund, were worth $750,000 and produced an average annual income of over $30,000, during the interval between 1923 and his death. The deed divided the fund into four parts, one for each child, but directed the trustees to apply the income of all four parts to the settlor's wife, and upon her death to apply the income of each child's share to him or her for life, with remainders over to grandchildren, per stirpes. There were also provisions for cross remainders in case any child died before the wife, and for the payment of parts of the principal to a child during his or her life, which we need not consider.

The Tax Court held that all the policies and three-tenths of the securities had been transferred 'in contemplation of death,' because the beneficiaries of the trust could not have any enjoyment of the policies, living the settlor; and, because three-tenths of the income had been used in the payment of premiums. In reaching this result the court depended very largely upon Article Twelfth of the deed, which directed the trustees out of the income to pay such premiums as fell due upon the policies, to collect the principal when the settlor died, and to add the proceeds to the fund. However, in case it was found unlawful to pay the premiums out of the income, living the settlor, the trustees were directed to invade the principal, and for that purpose to sell securities or at their discretion to 'realize the cash surrender value of so much of said insurance policies as may be required (A) for the purpose of paying future premiums with said proceeds, and (B)' of refunding to the beneficiaries any sums theretofore erroneously deducted from their income, which had been used to pay premiums. It never became necessary for the trustees to use this power. Earlier in the deed Article Fourth had also given the trustees discretionary power to sell 'any part of the trust funds,' to realize the cash surrender value of any of the insurance policies,' and 'to invest' and reinvest the proceeds of the sales or of the surrenders without restriction. The Tax Court concluded that the preponderating purpose of the settlor was that the policies should be left in existence during his life, and that the income of the securities should be used for that purpose. Since for the seventeen years during which he lived after 1923, it had taken three-tenths of the income to pay the premiums, the court also held that it was fair to assume that that proportion of the securities had been devoted in limine to maintain the policies, and for that reason it included that much of them in Garrett's gross estate together with the policies. That is the subject of the appeal in the case of the 'Trust of 1923.'

We have twice recently considered under what circumstances the transfer of policies upon a settlor's life should be taken as made in contemplation of death;2 and we have said that it did not inevitably follow, even though the beneficiaries cannot receive any part of the proceeds, living the settlor, that the gift was so made. The settlor may have wished to part with control over the policies for motives which were to be satisfied while he lived; for example, he may have wished to assure himself against any temptation to use them in his business; or he may have feared possible bankruptcy, or have wished to give the beneficiaries a present certainty on which they might build while he lived. Nevertheless, although for these reasons it would go too far to say that a transfer of policies can never be other than testamentary, when the beneficiary cannot possibly profit by them, living the settlor, ordinarily such a transfer will be of that kind. A conveyance of property which the grantee can be no chance use until the grantor's death, will so commonly be in the main testamentary, that it is fair to infer that that was its preponderating, if not indeed its only, purpose, unless there be affirmative evidence of other contributory motives. At least, when, as here, the question does not come to use for original decision, but when we have only to say whether the finding of another court is 'clearly erroneous,' we should not be justified in reversing such a finding. Therefore, in the case at bar, if it appears from the deed that the settlor's prime purpose was that the trustees should keep the policies untouched during his life, our inquiry must end.

We are satisfied that that was his purpose, substantially his only purpose. If we look at the Twelfth Article alone, that conclusion is certain: it expressly directed the trustees to use the income to pay the premiums so far as they lawfully could do so. It was only in case that use turned out to be unlawful that policies were ever to be surrendered; and then the surrender was limited to procuring cash to keep alive the other policies. True, the trustees were free to sacrifice policies rather than securities for this purpose; but the only permitted purpose was nevertheless to preserve policies which the beneficiaries could not enjoy until the settlor died. It is only when we look to the Article Fourth that any doubt can arise. As matter of interpretation it is possible that that article conferred upon the trustees a power, independent of, and more extensive than, the powers granted in Article Twelfth; and arguendo, we shall so assume in spite of the fact that Article Twelfth was specific. Yet although on that hypothesis Article Fourth did therefore give the trustees the power to surrender policies for other purposes than to pay premiums, that does not establish that his testamentary motive was not predominant. Certainly the Tax Court was free to find, as we should ourselves have found, that the preservation of the policies was his chief concern, and that the discretion vested in the trustees by Article Fourth was subsidiary: intended perhaps to apply to cases of emergency, or at least against situations whose importance could not be measured in advance: as, for example, if an insurer's credit should become doubtful, or if a critical need for cash to save some security should suddenly arise. Such powers are often extremely important in a crisis; but their existence does not here overshadow the patently primary purpose of keeping the policies in existence. That purpose once found to exist, it extended to the transfer of so much of the securities as were necessary to furnish the means to accomplish it, and we see no reason to disturb the proportion- three-tenths- which the Tax Court fixed. At first blush it may indeed seem curious to hold that only a part of a single trust fund has been transferred in contemplation of death; usually, the question is whether, taking the gift as a whole, the testamentary motive predominates. However, it becomes plain on reflection that one motive may dispose the settlor to convey one kind of property, and another, another kind; and that it is quite irrelevant that he groups the two in one deed. One must distinguish between a conflict of motives, which might have to be resolved, in deciding whether Whiteacre, for instance, was conveyed in contemplation of death, and an unambiguous motive to convey Whiteacre coupled with an equally unambiguous, though contradictory, motive to convey Blackacre, though one deed conveyed both. The order, so far as it concerned the 'Trust of 1923,' will be affirmed.

The 'Trust of 1929'

The fund in this trust consisted of shares of stock in the Garrett Holding Corporation, which was organized substantially at the same time that the trust was set up, and as part of a single plan or scheme. The trustees- Garrett's wife, his son, Charles, and his son-in-law- were directed to divide the fund into four equal parts and to pay the income of each part to one of the settlor's four children for life, with remainders over to the issue of such child. The assets of the Holding Corporation were $479,235.26 in securities, contributed by Garrett himself, between April 15, 1929, and January 31, 1933, and $119,200 in vineyards in New York and Virginia, of which Garrett contributed $13,900, his wife $48,700 and his son and his two daughters, $56,600. These vineyards had been losing money for some time, and it was the hope and expectation of Garrett's son, Charles, that his father would assume the expense of their maintenance. Among the securities were 9,372 shares- valued at about $188,000- of common stock of Garrett & Company, another corporation, which were vested Garrett's remaining business- four wineries- of which he was the principal shareholder and which he entirely controlled. The Holding Corporation issued 10,000 shares- later reduced to 1,000- of which Garrett and his wife, each got one-sixth, and the remaining two-thirds went to the trustees. It does not appear what were the earnings of Garrett & Company, whether it ever declared any dividend, or what was the salary of Garrett, if any. Similarly, it does not appear that during the eleven years between 1929 and Garrett's death in 1940, the Holding Corporation ever declared a dividend (apparently it did not), or what was the income, if any, from the securities, or the loss upon the vineyards, or what were the salary and expenses of conducting the business. On the other hand the record does show that during the years 1930, 1931 and 1932 Garrett was debited on the books of the Holding Corporation with $333,626.80 and was credited with $819,718.01. Since it appears that before December 31, 1932, he had contributed all but $24,093.75 of the securities mentioned- $45,141.51- it follows that from sources, not disclosed, he must have contributed $364,576.50 ($819,718.01-$455,141.51) before that date in addition to the securities; and this, it will be observed, was very nearly the amount of his debits. Moreover, it is a fair inference from the evidence that he was in control of the Holding Corporation until his death. The Tax Court held that, so far as the assets of the Holding Corporation were made up of the vineyards conveyed by Garrett's wife and children, the trust was not made in contemplation of death, but that all that he contributed himself including his own vineyards, were so conveyed and should be included in his gross estate.

If the Tax Court's order had been based upon the evidence we have summarized, particularly in view of the absence of any explanation by the petitioners of the affairs of the Holding Corporation, we should hold that it was not 'clearly erroneous'; the trust may well have been made in contemplation of death, so far as the shares of the Holding Corporation represented the securities and the vineyards contributed by Garrett. The situation, as the record presents it, was at least susceptible of the interpretation that Garrett meant to retain control of the Holding Corporation while he lived and to use it for his own purposes, borrowing and repaying as he pleased. If the securities produced no income and if he anticipated that they would not do so, it was indeed an almost inevitable inference that the trust was not intended to benefit the children while he lived. If on the other hand they did produce income, the inference from his failure to declare any dividends is even stronger. True, it is possible that part of the income, if there was any, was used to maintain those vineyards which had been contributed by the wife and children, and, since that was presumably contemplated at the outset, pro tanto the securities would not have been conveyed in contemplation of death. That possibility we may, however, lay aside, for the petitioners did not produce any evidence by which the Tax Court could allocate any part of the securities necessary for that purpose, as appeared in the case of the policies in the 'Trust of 1923.' On the record as it stands we should therefore affirm the order, were it not that the Tax Court used as evidence of intent in 1929 the transactions in 1923 and the will of 1935.

We cannot agree that either of these were relevant to a decision of Garrett's primary intent in 1929; and, since we cannot be sure that they were not determinative of the order, we must reverse it. Our reasons for thinking neither consideration relevant are as follows. The trusts of 1923- 'Trust No. 1' and 'Trust No. 2'- were in part predominantly testamentary, and in part not, as has already appeared. 'Trust No. 2' was not predominantly testamentary in any part; 'Trust No. 1' was so only as to the policies and their ancillary securities. This being true, it appears to us as plausible, to say that the transactions in 1923, taken as a whole- because they were in such large part not predominantly testamentary- were evidence that the intent in 1929 was not predominantly testamentary, as to say the opposite. Indeed, the property which in 1923 was conveyed for a purpose, not predominantly testamentary, was more like the property conveyed in 1929, than were the policies. In short, it appears to us that the transactions of 1923 have no rational significance whatever upon the issue. So too of the will of 1935. It is true that Garrett used in the will the limitations of the deed of 1929 as the measure of the disposition to be made of what might remain of his property at his death. But no one doubted that the 'Trust of 1929' had some prospective testamentary purpose; the only question was whether that purpose predominated over any purposes, not testamentary, which could be found in the deed. It did not advance the solution of that question an inch that, so far as the deed was testamentary, Garrett made use of it in his will which was necessarily altogether testamentary.

For these reasons the order must be reversed as to the 'Trust of 1929' and the cause remanded for the Tax Court to make a finding without use of this irrelevant evidence. Since this must be done and since the evidence may be different upon the new hearing, it seems to us best not to deal with the proper computation of the amount to be included in the gross estate, if the Tax Court comes to the same conclusion that it did before. In any event the present record is most unsatisfactory upon that score, and it should be more complete, if the petitioners mean to challenge this part of any future computation.

A mere reading of the text satisfies us that the Act of 1949 was not intended to cover transfers in contemplation of death; indeed, this appears to us so plain that we shall not labor the point.

Order affirmed as to the 'Trust of 1923.'

Order reversed as to the 'Trust of 1929'; and cause remanded for further proceedings in accordance with the foregoing opinion.

GOODRICH, Circuit Judge (dissenting in part).

Since this case depends upon inference to be drawn from uncontested facts, elaboration of reasons for not agreeing with the majority opinion is not indicated. I should reverse as to the 1923 Trust. There seems to me to be no evidence from which a fair inference of testamentary intent can be drawn and I fear the reiterated suggestion made on behalf of the Commissioner that any dealing with life insurance is by nature testamentary.