VAN OOSTERHOUT, Circuit Judge.

This is an appeal by the United States of America from final judgment of the district court in favor of taxpayer, Whitney Land Company, for $14,818.84, plus interest, for recovery of excess corporate income taxes paid for the year 1958. Taxpayer having paid the tax assessed and having filed timely claim for refund, brought this action pursuant to 28 U.S.C.A. § 1346(a)(1). This court has jurisdiction on appeal pursuant to 28 U.S.C.A. § 1291.

The facts are all stipulated and not in dispute. Taxpayer is a Minnesota corporation. It reports its income upon the cash basis method on a calendar year basis. Taxpayer sustained net operating losses for the years 1954 through 1957, inclusive, in the respective amounts of $18,363.14, $37,938.45, $5,063.40 and $3,148.19. In 1958 taxpayer had taxable income of $58,442.03 before application of any carry-over of any loss deduction. It is conceded that the 1956 and 1957 losses are available in full as off-sets against 1958 income. The problem arises with respect to the amount the 1954 and 1955 losses are available for offset against 1958 income.

Taxpayer had profits in 1952 and 1953. It is agreed that the 1954 loss must first be carried back to 1952 and that the 1955 loss must be carried back to 1953. It is clear that the 1939 Code provisions control with respect to the effect of the 1954 and 1955 loss carrybacks on the taxpayer's ultimate liability for 1952 and 1953 income and the amount of the tax due for such years is not here involved.

Taxpayer claims that the 1954 and 1955 net operating losses are to be reduced by the figures at line three in the following table. The Government contends that the net operating losses for the two years are to be reduced by the total of the figures appearing on lines one and four:

Items 1952 1953

1. Net Income ........................ $19,724.16 $26,384.87 2. Dividend Received Credit .......... 16,765.54 22,427.14 __________ __________ 3. Normal Tax and Surtax Income (Item 1 less Item 2) ............ $ 2,958.62 $ 3,957.73 4. Exempt Interest Excluded from above Figures ......................... $ 68.80 $ 68.80

The problem presented is whether the 1954 and 1955 losses when carried back to 1952 and 1953 are to be reduced by the economic income for 1952 and 1953 under the provisions of the 1939 Code as contended by the Government, or by taxable income as such term is defined in the 1954 Code as contended by the taxpayer, prior to being carried forward to the next profit year which is 1958. There is no dispute as to the figures involved in the computations.

The trial court held that the language of § 172(b)(2) of the Internal Revenue Code of 1954 is clear and unambiguous; that said section is not modified in any relevant context by the other provisions of the 1954 Code; that the amount of taxpayer's net operating losses for 1954 and 1955 must be determined under the 1954 Code net operating loss provisions; that under § 172(b)(2) such net operating losses must first be carried back to 1952 and 1953 respectively; that the amounts of the net operating loss deductions for 1952 and 1953 must be computed under the 1939 Code net operating loss provisions; that for the purpose of computing the amount of net operating loss deduction for 1958, the extent to which the net operating losses for 1954 and 1955 were absorbed by reason of the carryback to 1952 and 1953 respectively, must be determined in accordance with § 172(b)(2); that under § 172(b) (2), taxpayer's net operating losses for 1954 and 1955 must be reduced by the taxable income for 1952 and 1953 respectively, as computed by § 63 of the 1954 Code; that under § 172(b)(2), the excess of net operating losses for 1954 and 1955 over the taxable income for 1952 and 1953 constituted a net operating loss carry over to 1958, and that taxpayer was entitled to the refund it claimed.

The loss years and 1958, the year to which taxpayer seeks to have the unabsorbed losses applied, are all years governed by the provisions of the 1954 Code. Thus the problem presented relates to the interpretation of the 1954 Code provisions with respect to absorption of losses which had been carried back to pre-1954 Code years. The ultimate issue is whether the absorption of such losses on carry back to pre-1954 Code years is to be determined under 1954 Code standards or under 1939 Code standards.

The basis of the trial court's decision in favor of the taxpayer is found in the following excerpt from Judge Larson's opinion (not reported):

"The solution of the problem depends on whether the net operating losses for 1954 and 1955 should be reduced by the economic income for 1952 and 1953 or by the taxable income in those years.

"The question is whether the amount of a net operating loss sustained in a year in which the 1954 Code is applicable and carried back to a year in which the 1939 Code is applicable is governed by the 1954 Code as contended for by plaintiff or by the 1939 Code as contended for by defendant.

"The problem arises in part because of a change in concept in the two Codes. In the 1939 Code net operating losses were determined on the basis of an economic loss. By this concept net operating losses which were to be carried back were offset against the economic income of the taxpayer and not the income for tax purposes. Congress in the 1954 Code abandoned the economic loss concept and decided that net operating loss deductions were to be determined by the taxpayer's taxable income which excluded as one item intercorporate dividend income.

"Section 172(b)(2) of the 1954 Code provides as follows:

`(2) Amount of carrybacks and carryovers. — Except as provided in subsection (f), the entire amount of the net operating loss for any taxable year (hereinafter in this section referred to as the "loss year") shall be carried to the earliest of the 7 taxable years to which (by reason of subparagraphs (A) and (B) of paragraph (1)) such loss may be carried. The portion of such loss which shall be carried to each of the other 6 taxable years shall be the excess, if any, of the amount of such loss over the sum of the taxable income for each of the prior taxable years to which such loss may be carried. For purposes of the preceding sentence, the taxable income for any such prior taxable year shall be computed —

`(A) With the modifications specified in subsection (d) other than paragraphs (1), (4) and (6) thereof; and

`(B) by determining the amount of the net operating loss deduction without regard to the net operating loss for the loss year or for any taxable year thereafter, and the taxable income so computed shall not be considered to be less than zero.'

"Section 63 of the 1954 Code introduced the term `taxable income' and defines it as equal to gross income less specified deductions including the dividends received credit under Section 243. The term `taxable income' was apparently intended to eliminate such terms contained in the 1939 Code as `net income,' `adjusted net income,'`normal tax net income' and `corporation surtax net income'.

"Section 172(b)(2) of the 1954 Code refers specifically to the amounts of carrybacks and carryovers. The operating loss for any taxable year shall be carried to the earlier of the seven taxable years. The portion of such loss which shall be carried to each of the other six taxable years shall be the excess of the amount of such loss over the sum of the `taxable income' for each of the prior taxable years to which such loss may be carried.

"The term `taxable income' is clear and unambiguous and where this is true further reference to legislative discussion or background is not required. Congress may or may not have intended the result which will be reached in this decision. The Court must be guided by what Congress has actually provided in the statute."

We agree with Judge Larson's view that § 172(b)(2), which is quoted in the opinion, is the key statute bearing upon our interpretation problem and we also agree that the term "taxable income", which is defined by § 63 of the 1954 Code, has a clear and definite meaning and that Congress in using the term "taxable income" in § 172(b)(2) intended such words to have their meaning as set forth in the statutory definition.

Neither counsel nor the court has discovered any cases which deal specifically with the problem presented in this case. In our view, the cases of Pacific Rock Gravel Co. v. United States, 9 Cir., 297 F.2d 122; Kent v. Commissioner, 35 T.C. 30; and American Bank Trust Co. v. United States, E.D.La., 212 F. Supp. 148, cited and relied upon by the Government, are factually distinguishable from our present case, deal with a problem different than ours and afford no support for the Government's position. In Pacific Rock Gravel, the court held that where a 1939 Code year loss is carried to another 1939 Code year, the computation of net operating loss deductions and the amount by which the losses were absorbed in 1939 Code years are all governed by the 1939 Code. Neither party here would dispute the validity of such holding.

In Kent, the net operating loss in 1955 was carried back to 1953. The Tax Court there properly rejected taxpayer's contention that the loss deduction computation should be governed by 1954 Code provisions. The court held that the fact that the net operating loss for 1955 must be computed pursuant to the 1954 Code does not make the net operating loss deduction provision of the 1954 Code provisions applicable to years prior to 1954 with respect to determining the ultimate tax due for pre-1954 Code years. Such issue is not before us here as the parties agree that the taxpayer's net operating loss deductions for 1952 and 1953 must be computed under § 122(c) of the 1939 Code.

In American Bank Trust Co., the carryback of net operating losses for 1954 and 1955 was to 1952 and 1953, in each of which years the taxpayer had a substantial amount of exempt interest income. The taxpayer in addition to properly contending his 1954 and 1955 net losses were to be determined upon the basis of the 1954 Code, further urged that on carryback his net loss deductions for 1952 and 1953 were to be determined under the 1954 Code provisions. The court, following Kent, held that the 1939 Code controlled in determining the amount of net operating loss deductions for 1952 and 1953. The parties to this action would agree upon such determination up to this point as correct. Such holding is in no way inconsistent with Judge Larson's decision in the present case. The problem here presented does not relate to the question of the proper tax adjustment for a pre-1954 Code year but the separate and distinct problem of the extent that a 1954 Code year loss is absorbed by the carryback to a pre-1954 Code year. In our view, American Bank Trust Co.'s criticism of Judge Larson's opinion is at most dicta and is entitled to little weight as it is apparent that American Bank misconceives the scope of Judge Larson's opinion.

Two substantial changes were made by the Internal Revenue Code of 1954 which in our view have an important bearing upon the problem here presented. The first relates to the distinct shift made from the economic loss concept with respect to net operating losses manifested by § 122 of the Internal Revenue Code of 1939, which was designed to prevent a deduction if a taxpayer had not suffered a real non-tax loss by reason of the fact that included in its income are items such as exempt interest and intercorporate dividends for which a tax credit can be taken. When the 1954 Code was adopted the economic loss concept was largely abandoned. See S.Rep. No. 1622, 83d Cong., 2d Sess., pp. 211-212 (3 U.S.C.Cong. Adm. News (1954) 4621, 4847-4849).

Under the 1954 Code, a net operating loss deduction is taken pursuant to § 172(a). Section 172(c) provides in part: "[T]he term `net operating loss' means (for any taxable year ending after December 31, 1953) the excess of the deductions allowed by this chapter over the gross income." (Subject to certain modifications not here material.)

The extent to which a loss may be carried back or forward is prescribed by § 172(b)(2) heretofore quoted. The reason for the departure from the economic loss concept by the 1954 Code is expressed in a statement of the House Ways and Means Committee, reading:

"Your committee has also made changes in the method of computing the net operating loss deduction, in order to lessen the differences in tax treatment of firms with fluctuating and those with stable incomes. Under present law the loss is reduced for certain items with respect both to the loss year and the income year to which the loss is carried, before the loss can be offset against taxable income of the latter year. Thus under existing law taxpayers with loss carryovers are denied the use of tax benefits which are fully available to those with stable incomes. The items are tax-exempt interest, the excess of percentage depletion over cost depletion, the excess of nonbusiness deductions over gross income from other sources (but only for the purpose of determining the amount of the loss), the excess of capital losses over capital gains, the 50-percent deduction with respect to long-term capital gains for individual taxpayers, and the corporate dividends received credit." H.R. Rep. No. 1337, 83d Cong., 2d Sess., 3 U.S.Cong Adm.News 4017, 4052 (1954).

The other substantial 1954 change came in the adoption of "taxable income" as a substitute for "net income" and other terms used in earlier codes. Section 63 provides: "[T]he term `taxable income' means gross income, minus the deductions allowed by this chapter * * *."

The vital language in § 172(b)(2) bearing on this case is a sentence reading: "The portion of such loss which shall be carried to each of the other 6 taxable years shall be the excess, if any, of the amount of such loss over the sum of the taxable income for each of the prior taxable years to which such loss may be carried." (Emphasis added.)

We believe that Congress when using the words "taxable income" in the foregoing statute ascribed to such words the meaning specifically given to them by § 63. Thus it appears to us that § 172(b)(2) clearly and unambiguously provides that the loss to be carried forward in instances where the loss years and the tax years involved are 1954 Code years, is the excess of the loss in the earlier years over the taxable income as defined by the 1954 Code for such years.

Like the trial court, we find nothing in other parts of § 172 to modify the explicit language of § 172(b)(2). The Government's reliance upon § 172(e) is misplaced. While we believe there may be some ambiguity with respect to such statute, we believe that it is reasonably susceptible of a construction to the extent that the clause "the necessary computations involving any other taxable year shall be made under the law applicable to such other taxable year" means only that the net operating loss shall be determined under the law applicable to the loss year, and that it is not intended to deal with the absorption of losses that may be carried to other years. In any event, § 172(e) must be read in connection with § 172(b), and we are not persuaded that the ambiguous language of subdivision (e) overcomes the clear expression of legislative intent contained in subdivision (b).

The short answer to the Government's reliance upon § 172(f) and (g) as amended in 1958 is that such amendments have no bearing upon the problem here presented. They deal only with taxable years beginning in 1953 and ending in 1954 and short taxable years beginning in 1954 and ending before August 17, 1954. No such period is here involved. Any legislative intent in connection with the 1958 amendment could have little relevancy on the legislative intent asserted in 1954. Moreover, we find nothing in the 1958 amendments or their legislative history which is persuasive on the issue before us. We observe that the amendments reflect an intent to apply 1954 Code standards to all periods subsequent to December 31, 1953.

Finally, the Government relies upon Treasury Regulation 1.172-4 (a)(3), 1954 Code. Such regulation does support the Government's position as the regulation purports to make taxable income, as defined by § 172(b), the equivalent of net income under the 1939 Code. The trial court in disposing of this issue, states:

"The Commissioner by regulation paragraph 1.172-4 (a)(3) modifies the term `taxable income' by inserting `or net', thus converting the statutory language to `taxable (or net) income'. As stated previously, the statutory language is clear and unambiguous, and the Commissioner cannot by regulation change the law. The Commissioner's regulations endowed by considerable expertise are entitled to serious consideration but only where there is uncertainty in the statutory language."

We agree. It is apparent that the regulation is inconsistent with the statute upon which it is based as we have interpreted such statute. While Treasury Regulations are entitled to consideration and respect and may often be persuasive in resolving ambiguities, it is clear that the regulations which are inconsistent with the statutes upon which they are based are invalid. It is well-established law that the statute is the primary authority and that regulations to the extent they are in conflict with the statute are void. Trust of Bingham v. Commissioner, 325 U.S. 365, 377, 65 S.Ct. 861, 89 L. Ed. 1400; Helvering v. Credit Alliance Corp., 316 U.S. 107, 113, 62 S.Ct. 989, 86 L.Ed. 1307; Manhattan Gen. Equip. Co. v. Commissioner, 297 U.S. 129, 134, 56 S.Ct. 397, 80 L.Ed. 528; Boykin v. Commissioner, 8 Cir., 260 F.2d 249, 254.

The Government has failed to establish that the trial court committed error in interpreting the applicable statutes.

The judgment appealed from is affirmed.