This is an appeal from the assessment of a deficiency against the taxpayer in its “personal holding company su'rtax” for the year 1941. Two issues are raised: First, whether the taxpayer was a “person*451al holding company”; and, second, whether, if- so, ii was exempt from any surtax, because it distributed all its earnings of that year. The facts were stipulated, and were in substance as follows. The taxpayer is a Maryland corporation and did business in Brooklyn as a public investment company: its capital stock being offered for sale to the public generally from 1929 forward. At no time until the end of the calendar year, 1941, did it have less than three hundred shareholders; but during the year 1940 it purchased and retired so many of its shares that on December 31, 1940, five shareholders collectively owned a little more than fifty per cent; and this continued throughout the' year 1941. Realizing that this might expose the company to a “personal holding company surtax,” it was decided to dissolve and liquidate; and on the 24th of November, 1941, the shareholders passed a resolution to liquidate the company within a period of three years. A proper certificate was filed with the State Tax Commissioner of Maryland, and the dissolution was completed on December 15th of that year. The total assets realized upon liquidation were $401,940, and this sum was paid out in three equal dividends of $133,980 in each of the years 1941, 1942 and 1943. During the year 1941 it suffered “long term capital losses” of $43,083.93, and had “short term capital gains” of $6,198.07; it received dividends of $20,-096.16; and other income of $1,564.46. Its total deductions were $9,639.55. Therefore, omitting its net capital losses, its net income was $12,021.07. In computing its income tax it deducted $2000- — the only capital loss allowed by § 505(d), as amended by Revenue Act 1939, § 212(d), 26 U.S. C.A.Int.Rev.Code, § 505 note, — and charged itself with $10,021.07. However it claimed the right to offset this on the ground that it had distributed the whole of its earnings and was entitled to the “dividends paid credit” allowed by § 504(a), as amended by Act March 17, 1941, 26 U.S.C.A.Int. Rev.Code, § 504 note. This credit the Commissioner disallowed, and assessed against it a “personal holding corporation surtax,” based upon an income of $10,021.-07. The Tax Court — en banc, one judge dissenting — affirmed this assessment on two grounds: first, that the company was a “personal holding company”; and second, that our decision on which the company relied — Pembroke Realty & Securities Corp. v. Commissioner, 2 Cir., 122 F.2d 262— did not cover the'case; and if it did, “with all due respect to the Circuit Court of Appeals of the Second Circuit, to which an appeal in this case will lie, we decline to apply the Pembroke case here, because tve think its application would be contrary to the law as written by Congress.”

That the company was a “personal holding company” within the meaning of § 501(a) (2), 26 U.S.C.A.Int.Rev.Code, admits of no doubt. The only argument to the contrary is that it should not properly be deemed an “incorporated pocket-book,” because its shares had always been bought and sold freely on the market, because it was never closely held, and because it was only by chance that at the end five shareholders had come to own more than half the shares. This we cannot accept; so far . as we can see, there is no reason to suppose that, had Congress ha.d such a company in mind, it would have wished to except it. The majority of shareholders would be in control, it would be to their interest to avoid high personal surtaxes, and that interest might well be enough to impel them to spread over a number of years the distribution of its earnings: exactly what the statute was aimed to prevent. True, there is often no surer way to misconceive the meaning of a statute or any other writing than to construe it verbally; indeed, interpretation is the art of proliferating a purpose which is meant to cover many occasions so' that it shall be best realized upon the occasion in question. However, although there are no certain guides, the Colloquial meaning of the words is itself one of the best tests of purpose, and situations of which Markham v. Cabell, 326 U.S. 404, 66 S.Ct. 193, is the last example are the exception. We have no doubt that as to the first point, this is not a situation in which it would be proper not to follow the literal meaning of the words.

The second question raises the perplexing question of the limitations upon our power to review decisions of the Tax Court. *452In August 1941 we reversed the Board of Tax Appeals, and held that, when a “personal holding company” was in liquidation, the distribution of any earnings made in that year entitled it pro tanto to a “dividends paid credit” under § 504(a), in spite of § 115(c) which declared that such distributions should not be treated as dividends, but must be brought into hotch-pot in an equation of loss or gain. Pembroke Realty & Securities Corp. v. Commissioner, supra, 122 F.2d 252. In the case at bar the Tax Court has held that this was wrong and has flatly refused to follow it, although it suggested a possible distinction: _ i.e., that here there were no earnings to distribute at all in the year 1941, if one counted “net capital losses” for that year; while in Pembroke Realty & Securities Corp. v. Commissioner, supra, there were earnings, unless one counted the capital impairment as a loss. We cannot agree that this is a tenable difference, and the Tax Court itself obviously put it forward without conviction. Since § 505(d) is one of the specific directions by which to compute the “personal holding company surtax,” and since it particularly provides that “losses from sales * * * of capital assets shall be allowed only to the extent of $2,000,” it seems plain that no capital losses beyond $2000 should be allowed in computing the “Subchapter A Net Income” of § 505, from which the deductions (among them the “dividends paid credit”) allowed by § 504 are to be subtracted, in order to ascertain the “Undistributed Subchapter A Net Income.” There is nothing extraordinary in that; it is true that losses and gains are assumed to occur in the year in which they are “realized,” but that is merely a convention, which Congress was free to use or not at its pleasure. Dividends are received as much in years when old losses are “realized,” as dividends are received although capital has been previously impaired.

So much taken for granted, the question arises whether we should yield to the Tax Court’s ruling that we were wrong, and that it would not follow us. What it said, might of course change our opinion on the point, in which event there would no longer be any conflict; but in the case at bar it chances that this has not happened; and, if the case were an appeal from a district court, we should have no alternative but to reverse. But the Supreme Court has repeatedly admonished us (in so many decisions that it would be idle to repeat them), that our power to review a ruling of the Tax Court is very much more limited than in the case of a district court. As we understand it, before we may. substitute our own interpretation of a provision of the Revenue Act, not only must a naked question of law detach itself from the nexus of law and fact in the record as a whole; but we must conclude that the Tax Court has been indubitably wrong in its decision of the question which emerges: reasonable differences in ’ legal opinion we are to resolve in its favor. We have recently discussed the matter in Kirschenbaum v. Commissioner, 155 F.2d 23, and have now nothing to add to what we said. It seems to us that the right answer to whether § 115(c) is controlling here is not so certain that we should be justified in following ou'r own beliefs; and therefore, although personally we are of the same mind as before, we think that we should yield to the insistence of the Tax Court, which within these limits is really the court of last appeal. One question remains. It may be asked why the same reasoning did not apply, when in Pembroke Realty & Securities Corp. v. Commissioner, supra, 122 F.2d 252, we reversed the holding of the Board of Tax Appeals. It did; but in August, 1941, the Supreme Court had not yet made clear, as it now has, how straitly our jurisdiction is confined. Our decision was then wrong, not indeed because as an original question it was, but because, by assuming jurisdiction to consider the question at all, we invaded the prerogative of the Tax Court and disregarded the finality of its orders. That finality depends, as we understand, upon the added competency which inevitably follows from concentration in a special field. Why, if that be so, we — or indeed even the Supreme Court itself — should be competent to fix the measure of the Tax Court’s competence, and why we should ever declare that it is wrong, is *453indeed an interesting inquiry, which happily it is not necessary for us to pu'rsue.

Order affirmed.

(concurring).

I agree with the result and with all of the opinion except for the two conjoined statements that Pembroke Realty & Securities Corp. v. Commissioner, supra, is indistinguishable, and that personally we are of the same mind as when it was announced. Though I did not participate in that case, I should suppose it distinguishable. There at least the statutory purpose had been accomplished by a distribution of earnings which went to swell the taxable income of the distributees; here, on the taxpayer’s contention, the corporate earnings would be strictly nontaxable. But if I am wrong and the distinction is not sufficient, I do not see, for my part, how the Pembroke principle could have stood as applied to these facts, whatever the decision of the Tax Court, in view of the clear command of the statute.

(dissenting).

Under the prevailing opinion this court would seem to be given a purely technical power of review only in order to satisfy a conventional aspiration for such a procedural step rather than to afford relief to aggrieved litigants. I am not satisfied that the dice are so heavily loaded against any decision which differs from the Tax Court that our jurisdiction is rendered almost futile, nor do I believe that the opinions of the Supreme Court call for complete abdication on our part.

The determination of the Tax Court while disapproved of by the prevailing opinion is nevertheless affirmed for lack of power to give any relief. It purports to tax an accumulation of income — though there in fact was none — but, on the contrary, a deficit in capital during the taxable year. Such a result contravenes the purposes of the Congressional legislation while following a slavishly literal reading of the statute. In my opinion the Dobson decision (Dobson v. Com’r, 320 U.S. 489, 64 S.Ct, 239, 88 L.Ed. 248) does not preclude us from determining such a general question of law as whether the statute covers distributions in liquidation where no earnings or profits exist. Such lack of power would render appeals to this court meaningless and, if our view of the interpretation of the statute is correct, tragic for the taxpayer in the present case.

I think that the order of the Tax Court should be reversed.