delivered the opinion of the Court.

This is a suit for tax refund which the District Court allowed. 62 F. Supp. 338. The Circuit Court of Appeals affirmed. 155 F. 2d 577. We granted certiorari because of an apparent conflict with Century Electric Co. v. Commissioner, 144 F. 2d 983.

The respondent, Ogilvie Hardware Co., Inc., was incorporated in Louisiana in 1907 with a paid-in capital of $100,000. In 1924 it increased its capitalization to $200,000 by declaration of a $100,000 stock dividend out of past earnings. Depressed business conditions during the 1930’s brought heavy operating losses so that by 1937 the company’s assets were about $71,000 less than the $200,000 capitalization. The company books accordingly showed a deficit in this amount. By 1938 this deficit was reduced to about $61,000. In this financial posture the corporation could not declare dividends without impairing its then capital structure (which included capitalization of the $100,000 stock dividend) and Louisiana law prohibited payment of a dividend under such circumstances.1 Section 14 of the governing Rev*712enue Act of 1936 imposed a surtax on certain corporate net income earned during the tax year but not distributed as dividends.2 It provided no exemption from that surtax because a corporation had an accumulated deficit at the beginning of the tax year, or because state law prohibited payments of dividends.

Acting under this 1936 law, the Commissioner, on examination of respondent’s 1937 and 1938 tax returns, determined that respondent was subject to the undistributed profits tax, despite the deficit and the state prohibition against payment of dividends. The Commissioner’s interpretation and application of the 1936 Act was in accord with our holding in Helvering v. Northwest Steel Rolling Mills, 311 U. S. 46, and Crane-Johnson Co. v. Helvering, 311 U. S. 54. The taxpayers in those cases claimed exemption from the surtax on the ground that they could not distribute dividends “without violating a provision of a written contract executed by the corporation prior to May 1, 1936, which provision expressly deals with the payment of dividends.” Section 26 (c) (1) of the 1936 Act relieved corporations from the tax if such contracts existed. 49 Stat. 1648, 1664. The question we had to decide in those cases was whether a state constitution, corporate charter, or state statute, which prohibited payment of dividends, was a “written contract” within the meaning of the § 26 (c) (1) exemption provision. We held that we could not so expand the provision’s language, relying in part upon previous statements of this Court “that provisions granting special tax exemptions are to be strictly construed.” Helvering v. Northwest Steel Rolling Mills, supra, 49. Since the respondent here had no “written con*713tract” against payment of dividends, it had no exemption from the surtax imposed by the original 1936 Act.

But this suit is not brought to determine the company’s tax liability under the 1936 Act as it stood in the taxable years 1937 and 1938. It is an action for a refund under a 1942 relief amendment to the 1936 Act specifically designed to authorize corporations to obtain repayments of taxes they had been forced to pay under the 1936 Act as we had interpreted it. That amendment, as enacted, provided for complete or partial retroactive immunity from the 1936 undistributed profits tax under the following circumstances:

“Deficit corporations.—In the ease of a corporation having a deficit in accumulated earnings and profits as of the close of the preceding taxable year, the amount of such deficit, if the corporation is prohibited by a provision of a law or of an order of a public regulatory body from paying dividends during the existence of a deficit in accumulated earnings and profits, and if such provision was in effect prior to May 1, 1936.” § 501 (a) (3), Revenue Act of 1942, 56 Stat. 798, 954.

This amendment was designed to grant corporations a refund on account of payments of undistributed profits taxes for tax years in which they had an accumulated deficit, and where, for that reason, state law, federal law, or public regulatory orders of either prohibited distribution of dividends. It therefore authorized refunds to the very taxpayers who had been lawfully required to pay taxes by the 1936 Act as we had interpreted it in the two cases cited above. Furthermore, in order to make sure that taxpayers who had paid under our interpretation might recover refunds, § 501 (c) of the same amendment specifically authorized claims for repayment to be filed within one year after its passage, without regard to *714any statute of limitations or other designated statutory bars. 56 Stat. 798, 955.

The Government’s contention is that we should construe the word “deficit” and the phrase “accumulated earnings and profits” according to their established meaning under federal tax law; that so construed the $100,000 allotted for stock dividends remained a part of earnings and profits for tax purposes; therefore, there was no deficit in the federal tax sense, and consequently the tax payments should not have been refunded here despite the state prohibition against distribution. We may assume that the Government is correct in contending that if Congress intended in the 1942 amendment to use the words “deficit” and “earnings and profits” in this federal tax sense, the stock dividend did not reduce “earnings,” there was no “deficit,” and the refund should be denied. See § 115, Revenue Act of 1936, 49 Stat. 1648, 1687-1689; Commissioner v. Bedford, 325 U. S. 283, 292. This construction would greatly limit the scope of the relief granted by the 1942 amendment. To determine whether Congress intended so to limit the relief it granted, we must look to the whole 1942 amendment in its relationship to the 1936 Act and the legislative and judicial history intervening between the two.

The 1936 undistributed profits tax law was a novelty in the field of federal taxation. Its chief novel feature was that it was designed to compel corporations to distribute current earnings to shareholders by imposing a surtax on corporations which failed to make such distributions. It had detailed provisions for defining the net income which would be reached by this tax. Its application, therefore, raised new and sometimes wholly unexpected problems. Widespread opposition developed to the tax. Since 1938, only a token of it has survived. See Revenue Act of 1938, 52 Stat. 447. But even after the 1936 undistributed profits tax was no longer in effect, complaints about its prior *715application from corporations which had been required to pay an undistributed profits tax continued to reach and to concern Congress. Representatives of these corporations appeared before the House and Senate Committees in 1942, and Congress responded to their complaints by enacting the several provisions of § 501—the retroactive relief legislation now under consideration.

One subject of complaint was that under the income tax definitions only a fraction of capital losses was deductible from taxable net income. Corporations which had suffered large capital losses in a given year were required to pay undistributed profits taxes in that year as though they had made a profit. The 1942 amendment, as reported by the House Committee, met this complaint by recommending that refunds be authorized for corporations who had paid under this 1936 definition of net income.3 This authorization, subsequently approved by the Senate Committee,4 clearly shows that Congress intended to provide for this phase of the refund without regard to tax definitions, and did not intend its authorized refund to be restricted by the application of established tax terminology.

When the bill reached the Senate Committee, insistent complaints related to the fact that corporations with defi*716cits in accumulated earnings and profits had been compelled to pay taxes for non-distribution of dividends although state or federal law prohibited dividend payments. A deficit railroad corporation had been taxed over its objection that payment of dividends would have rendered its officers subject to punishment for a misdemeanor under federal law and a money penalty under state law. The Board of Tax Appeals had overruled objections on these grounds, relying on our decisions in the Crane-Johnson and Northwest Steel cases, supra. Paris & Mt. Pleasant R. Co. v. Comm’r, 47 B. T. A. 439.5 The counsel who had represented Crane-Johnson before this Court also appeared on their behalf before the Senate Committee and made a plea for relief for deficit corporations which had been compelled to pay the undistributed profits tax.6 He *717urged that such corporations had been “caught in a trap,” and that they were justly entitled to have a refund for that reason. It was apparently in response to the foregoing complaints that the relief provision before us, not part of the House bill as it came to the Senate,7 was introduced by the Senate Committee.8 We think Congress was moved to relieve those corporations which it considered to be “caught in a trap” whereby they were taxed by the Federal Government if they did not pay dividends and subject to prosecution and penalties by the Federal Government or the states if they did.

Some of the language Congress used, considered tax-wise only, provides plausible support for the interpretation urged by the Government which would give the relief amendment more limited scope. But the provision before us is not a general tax exemption to be interpreted in the framework of a currently operating general revenue law. It is a special retroactive relief measure to authorize repayment of taxes collected in previous years under a revenue law which had already been substantially abandoned. The language of this extraordinary relief measure and the circumstances which prompted its passage convince us that Congress intended to provide refunds to corporate taxpayers, with possible minor exceptions, who had paid undistributed profits taxes as a choice between conflicting state and federal compulsions.

Furthermore, the very mechanics of the 1942 amendment require that determination of rights to refund under it be based on consideration of something other than the tax meaning of the 1936 Act or other tax terminology. The right to recovery in every case depends ultimately upon whether federal law or federal regulatory bodies, or *718state law or state regulatory bodies, prohibit payments of dividends. In this case the ultimate right to refund depends upon state law. Cf. Lyeth v. Hoey, 305 U. S. 188, 193. Before that right can be finally established, courts must examine state law at least to the extent of determining (1) what is a “deficit”; (2) what are “accumulated earnings and profits”; (3) what was the state law on these questions prior to May 1, 1936; (4) whether payments of dividends under these circumstances were prohibited by state law. Acceptance of the Government’s contention would mean that courts administering the 1942 Act must first determine whether a deficit exists under federal law; if such a federal deficit exists, they must then turn to state law to decide whether under it a deficit exists such as prohibits the payment of dividends. We do not think that Congress intended the courts so to administer the 1942 amendment. The Government’s argument that it does relies heavily upon the Senate Committee Report.

We think the Senate Committee Report, as a whole, leans toward the view we have taken of the purpose of the law.9 But in one of the six illustrative examples of application of the new tax relief provisions of the amendment, *719and in the subsequent Treasury Regulations, it was indicated that no tax credit should be allowed where a tax deficit resulted from “prior capitalization of surplus in the course of a nontaxable reorganization.”10 Aside from the fact that corporate reorganizations and simple stock dividends are quite different things, we find this one illustrative example insufficient to outweigh the considerations which have governed our interpretation of the 1942 amendment.

We are persuaded that Congress at least intended by the amendment to refund taxes imposed on corporations which had failed to distribute dividends when distribution, in violation of state law, would have impaired long-existing state-approved corporate capitalizations. See United States v. Byron Sash & Door Co., 150 F. 2d 44, 46. In order that this purpose may be effected, the judgment of the Circuit Court of Appeals is

Affirmed.

with whom Mr. Justice Reed joins, dissenting.

The Revenue Act of 1936 imposed a surtax on undistributed corporate profits. Section 26 (c) (1) gave relief *720from this surtax under defined circumstances.1 In Helvering v. Northwest Steel Mills, 311 U. S. 46, it was held that although a restriction on the distribution of corporate profits was imposed by State law, a credit for such withheld profits was not authorized by § 26 (c) (1). In reaching this conclusion, the Court took into account that it “has been said many times that provisions granting special tax exemptions are to be strictly construed.” Helvering v. Northwest Steel Mills, supra, at 49. By way of relaxing the restricted scope which this Court gave to exemption from the undistributed profits tax, Congress, by the Revenue Act of 1942, substituted a new subdivision (3) to § 26 (c) of the Revenue Act of 1936. This section did not undo the Northwest Steel Mills doctrine. It did not allow a deduction for profits forbidden to be distributed by State law, as it had, in § 26 (c) (1), allowed credit for profits undistributed because of a “written contract.” Congress gave relief for earnings forbidden to be distributed by State law only “In the case of a corporation having a deficit in accumulated earnings and profits as of the close of the preceding taxable year . ...” 2

*721This is tax language and should be read in its tax sense. We must not disregard the illumination of an authoritative tax lexicon in reading tax legislation. The language of the 1942 amendment carries with it tax usage, tax practice, and the gloss of authoritative legislative history. All combine to make the condition under which State law prohibiting distribution of profits comes into play, that which Congress in words of art said was the condition, namely, the existence of “a deficit in accumulated earnings and profits.” Here there was no deficit in the controlling sense of the term. And nothing warrants the attribution of a non-technical meaning to so settled a technical term. Nothing, that is, except the suggestion that to give the 1942 amendment this established meaning might not afford the relief that, as a matter of abstract justice, should be afforded. But this is merely an attempt to invoke what has been called the “equity” of a statute. I am no friend of artificial canons of construction, and I would not strain language in order to construe tax exemptions strictly. On the other hand, Revenue Acts are not the kind of legislation which should be loosely construed in order to grant exemptions.

The legislative history of this enactment and the administrative practice only reenforce what seems to me to be the compelling requirement, to render technical terms used by Congress with their technical meaning. If it be suggested that counsel for taxpayers at a Congressional hearing urged the fairness of the construction which the Court now places upon what Congress has expressed, it would not be the first time that the final legislation of Congress did not satisfy the desire of some of its proponents. In *722any event, I do not think the argument of counsel for a taxpayer urging relief should carry more weight than the use by Congress of settled tax language, carrying a meaning which excludes that result, a meaning which is reenforced by the legislative, judicial and administrative history that led up to and followed the enactment. See Century Electric Co. v. Commissioner, 144 F. 2d 983, affirming the Tax Court, 3 T. C. 297; S. Rep. No. 1631, 77th Cong., 2d Sess., pp. 244-46; Treasury Regulations 94 and 101, Art. 115-11; Treasury Regulations 103, §19.115-11; Treasury Regulations 111, §29.115-11. The short of the matter is, that even though corporate profits here were withheld because Louisiana forbade their distribution, there can be no credit allowed for a deficit because in a federal tax sense there was no deficit.

No doubt Congress, to some extent, desired to relieve from the undistributed profits tax corporations forbidden by State law from declaring dividends. But neither what Congress enacted nor its legislative history indicates a purpose to disregard the limiting provisions of § 115 (h) of the Revenue Act of 1936.3 This section, which embodies the analysis of Commissioner v. Sansome, 60 F. 2d 931, *723see S. Rep. 2156, 74th Cong., 2d Sess., p. 19, requires that in respect to federal taxes, assets be treated as available for distribution as earnings regardless of stock dividends which capitalize earnings and profits. H. Rep. No. 2894, 76th Cong., 3d Sess., p. 41, cited in Commissioner v. Wheeler, 324 U. S. 542, 546. The specific example cited by the Senate Committee Report on § 501 of the Revenue Act of 1942 shows that Congress intended to limit the relief afforded by the amendment to cases where the deficit in question had not resulted from the capitalization of accumulated earnings and profits.4 The majority finds a difference between capitalization of earnings in a nontaxable reorganization and capitalization of earnings by a simple stock dividend. The circumstances are different but the difference is not significant for the legal effect of the stock dividend on earnings and profits. The example given is concerned with the effect of capitalizing earnings and profits, not with the method. If Congress meant to relieve undistributed earnings and profits even though those earnings and profits were considered available under § 115 (h), it should have said so.

We think the judgment of the Circuit Court of Appeals should be reversed.