OPINION.
Ivins :The taxpayer from 1909 to 1920 was the owner of an office building in St. Louis and derived its income solely from the opera*357tion thereof. On its hooks and its returns under the Corporation Tax Act of 1909 and the income tax provisions of the Revenue Act of 1913 and subsequent revenue acts, it never made any provision or claimed any deduction for depreciation or for exhaustion, wear and tear, or obsolescence of the building. It sold the building in 1920, and it becomes incumbent upon us to determine the extent of the gain upon such sale that should be included in the taxpayer’s gross income for 1920. In order to make this determination we must decide (1) whether depreciation, exhaustion, wear and tear, and obsolescence of the building during the period of ownership should be considered in computing that gain, and (2) whether, if so, they should be based upon the cost of the building or upon its value at March 1, 1913.
Upon the first point the taxpayer contends that it is under no obligation to make any adjustment for depreciation, exhaustion, wear and tear or obsolescence not claimed by it as deductions in its excise tax and income tax returns; that it is taxable only on the difference between cost and sale price. The Commissioner contends that the taxpayer should have taken deductions for the items mentioned in all the years from 1909 or from 1913 to the time of sale; that the basis for computing gain should be reduced by the amounts which should have been taken as deductions, whether taken or not, and that only the unextinguished cost or 1913 value should be subtracted from the sale price in determining the gain upon the sale.
Section 202(a) of the Revenue Act of 1918 provided:
That for the purpose of ascertaining- the gain derived or loss sustained from the sale or other disposition of property, real, personal, or mixed, the basis shall be (1) in the case of property acquired before March 1, 1913, the fair market price or value of such property as of that date; and (2) in the case of property acquired on or after that date, the cost thereof; or the inventory value, if the inventory is made in accordance with section 203.
Leaving until later the question whether the computation should be based upon cost or upon the 1913 value, let us first consider the intent of Congress in using the language “ for the purpose of ascertaining the gain derived * * * the basis shall be * * * .” The taxpayer urges that this means that the gain is the difference between the basis and the sale price, and that no adjustment need be made for depreciation, exhaustion, wear and tear, or obsolescence, where no deduction has been claimed for such items in its tax returns for prior years. It suggests that this might be the rule even if such deductions had been claimed and allowed, citing Ward v. Hopkins, an unreported decision of the United States District Court for the Northern District of Texas, Dallas Division. That case, for some reason, went no higher than the district court, and we feel constrained, for reasons which will appear, to disagree with the conclusion of the learned court. There is no ground for believing that Congress, in using the word basis intended it to carry any other than its commonly accepted meaning. The New American Encyclopedic Dictionary defines basis as follows:
A. Ordinary language:
I. Lit. Of things wMeh are or are assumed to he material: That on which anything rests or is supposed to rest; the lowest part of anything, as the foundation of a building, etc.
II. Of things immaterial: The fundamental principle, groundwork, or support of anything.
*358Webster’s New International Dictionary says:
Basis: 1. The foundation of anything; that ;n which a thing rests; the base. * * * 4. The principal component part of a thing. 5. The groundwork ; the first or fundamental principle; that which supports or sustains.
We have no hesitation in holding that Congress in using the word basis meant nothing but starting 'point or primary -figure in the computation of gain or loss, and had no intention of restricting that computation to a simple subtraction of the basis from the selling price or vice versa. It expected the computation to include all adjustments necessary to a logical ascertainment of gain or loss. The only reason for using the word at all was to take care of the different situations arising when the property disposed of had been acquired (a) before and (5) on or after March 1, 1913. It fixed the starting point or primary figure of computation in the respective cases, but did not attempt to define every step of the computation under varying circumstances. In some cases, as when a taxpayer buys a security for one price and sells it for another, a simple subtraction is all that is necessary to determine his gain or loss. But, in other cases, either the basis or the sale price must be adjusted before mak-i;ng the subtraction in order to have the difference truly represent the gain or loss. For example: If a taxpayer owned property on March 1, 1913, then worth $10,000, thereafter made permanent improvements thereon at an expense of $5,000, and later sold it for $16,000, it is obvious that the difference between the $10,000 basis and the $16,000 sale price is not a proper measure of the gain from the transaction. If one bought land with timber upon it for $10,000 in 1914, cut down the timber, and later sold the land for $11,000, his gain could not properly be computed without reference to the value realized by him in cutting the timber — and this would be true whether or not he had sold the timber, whether or not he had taken account of it on his books or in his tax returns, and whether or not he had claimed a deduction in his tax returns for depletion.
The establishment of a basis for computing profit or loss did not constitute a formula for that computation. Under the Corporation Tax Act of 1909 and the Revenue Act of 1913, no basis was indicated, but that did not prevent a computation under general principles. As the Supreme Court said in Doyle v. Mitchell Bros. Co., 247 U. S. 179, 184:
There is no express provision that even allows a merchant to deduct the cost of the goods that he sells. Yet it is plain, we think, that by the true intent and meaning of the act the entire proceeds of a mere conversion of capital assets were not to be treated as income.
If adjustment for cost could be made in computing profit or loss without express provision of statute, there is no reason why other proper adjustments should not be made without such provision.
The contention of the Commissioner is that the gain of the taxpayer can not properly be ascertained by a mere subtraction of the basis from the sale price without adjustment for depreciation, exhaustion, wear and tear, or obsolescence. This brings us to consider the meaning and effect of the statutory provisions for the deduction, in computing income, of allowances for those items.
The Corporation Tax Act of 1909, imposing an excise tax on doing business as a corporation measured by net income, provided *359that in computing net income, among the deductions there should be allowed “a reasonable allowance for depreciation of property, if any.” The Revenue Acts of 1913, 1916, 1917, and 1918, taxing net. incomes directly, provided for deductions by corporations as follows:
Act of 1913:
A reasonable allowance for depreciation by use, wear and tear of property, if any.
Acts of 1916 and 1917:
A reasonable allowance for the exhaustion, wear and tear of property arising out of its use or employment in the business or trade.
Act of 1918:
A reasonable allowance for the exhaustion, wear and tear of property used in the trade or business, including a reasonable allowance for obsolescence.
The various acts provided for deductions in computing the taxable income from property held. It does not follow that if a taxpayer, in any year or as a regular practice, refrained from claiming them, he becomes entitled to have the facts, upon which they might have been based, ignored in computing gain or loss upon the sale or other disposition of the property. The very reasons for allowing the deductions necessitate the consideration, in computing gain or loss, of the facts upon which the deductions could be based.
When a manufacturer sells his product he is really selling a composite of four or five things: (1) The materials entering into the product; (2) labor that has been applied to them, including the effort of executives and salesmen; (3) that proportion of the value of the machinery and of (4) the buildings which is consumed in the manufacture of the product, and (5) in some cases a proportionate part of the monopoly granted by patents, exhausted during the process of production. Until the manufacturer has recovered the cost or basic value to him of the proportionate parts of these items attributable to a unit of his product sold, he can not properly be said to have received income upon its sale, and only the excess of the sale price realized over the sum of these items (whatever particular form they may take) is truly income. We feel that this would he true whether the statutes made any specific provision for their deduction or not — the computation would be just as necessary to the determination of true income as the computation of profit on a sale by reference to the cost or basic value of the article sold. As the Supreme Court said in Doyle v. Mitchell Bros. Co., supra, at p. 188:
It may be observed that it is a mere question of methods, not affecting the result, whether the amount necessary to be withdrawn in order to preserve capital intact should be deducted from gross receipts in the process of ascertaining gross income, or should be deducted from gross income in the form of a depreciation account in the process of determining net income. In either case the object is to distinguish capital previously existing from income taxable under the act.
Because of the difficulty of attributing to each article sold or to the sales in a taxable year an exact sum representing the proportion of basic value of machinery, buildings, patents, etc., entering into such sale, the statutes have consistently provided for a reasonable allowance. And on the same theory a reasonable allowance for the *360recoveries made against capital in the sale of products should be taken into consideration in determining the gain or loss realized upon the sale of such capital assets. If a manufacturer has operated machinery in the production of commodities and has recovered, through their sale, half of the cost or basic value of the machinery and then sells what is left of the machinery, he certainly should not be permitted to compute his gain or loss on the sale by reference only to the cost or basic value of the machinery and its sale price without adjustment for the intervening recoveries of capital.
Nor should his failure, on his books of account or in his tax returns, to take credit in the years of their occurrence for the mesne recoveries of capital, permit him to ignore them upon reporting the sale of capital assets in a later year. “ Such books of account are not more than evidential, being neither indispensable nor conclusive. The decision must rest upon the actual facts. * * Doyle v. Mitchell Bros., supra, at p. 187.
What is true of a manufacturer is equally true of the owner of an office building. The instant taxpayer bought a building in 1909, rented offices in it until 1920, and then sold. it. Presumably the value of the building was in part used up during that period, and the rents received were attributable, before the discovery of income, to the reimbursement of the owner for such capital exhaustion or loss. It is true that in many cases the loss in value is partially offset by repairs and replacements, but general experience has shown that wear and tear and obsolescence do reduce the value of a building or any other tangible asset actively used in commerce, in excess of such offset. In the absence of affirmative proof to the contrary and in the light of the Commissioner’s determination to that effect in the instant; case — the taxpayer being the moving party in this appeal and it being incumbent upon it to submit proof to sustain any claim that the Commissioner has been guilty of an error of fact — we must accept the finding of the Commissioner, which was necessary to his determination, that such an exhaustion or loss of value did occur.
But, says the taxpayer, the depreciation of the building was more than offset by appreciation in value during the period of ownership, as is evidenced by the excess of the sale price over the cost. There is no evidence before us to show that this appreciation was in the building rather than in the land which was bought and sold with the building. However, even if it were shown that the land had not appreciated one cent in value, we do not think that would change the result. In the first place depreciation is a broad term, not always limited to the reduction of value by reason of exhaustion, wear and tear, or obsolescence. Taken literally, it is the converse of appreciation and merely means decrease of value not necessarily due to use or lapse of time. The Circuit Court of Appeals for the second circuit, in New York Life Insurance Co. v. Anderson, 263 Fed. 527, held that securities owned by an insurance company were subject to depreciation. In the Corporation Tax Act of 1909 Congress provided for an allowance for depreciation if any in computing income, but income under the 1909 Act and under the later revenue acts is not necessarily the same thing any more than income under the revenue acts and under the Constitution. “As has been repeatedly remarked, the Corporation Tax Act of 1909 was not *361intended to be and is not in any proper sense an income tax law.’’ Stratton's Independence v. Howbert, 231 U. S. 399, 414. And the Supreme Court has also said:
* * * But it is not necessarily true that income means the same thing in the Constitution and the Act [of 1913]. A word is not a crystal, transparent and unchanged, it is the skin of a living thought and may vary greatly in color and content according to the circumstances and the time in which it is used. Towne v. Eisner, 245 U. S. 418, 425.
It may be that depreciation in the 1909 Act was a broad term, subject to application to nonwasting assets (as applied in the New York Life Insurance Co. case, supra), and that it was the intent of that act to permit it to be so construed in computing the particular kind of income thereby made the measure of an excise tax. It may also be that the term there used was so broad as to mean only the net change of value after allowance for offset by appreciation, as it was applied in Nashville, etc. Ry. Co. v. United States, 269 Fed. 351 (though we should not be understood as agreeing with the conclusion arrived at in that case). But it should be noted that the word depreciation has disappeared from the revenue acts. Presumably the change was made by Congress in an effort to perfect the language of the statute. The act of 1909 said “ a reasonable allowance for depreciation of property, if any.” The Revenue Act of 1913 restricted the deduction to an allowance for depreciation by use, wear, and tear. And the later revenue acts eliminated the word depreciation entirely (see p. 359, supra). There is nothing in any of the revenue acts subsequent to the Sixteenth Amendment which would have precluded the taxpayer from taking a reasonable deduction for wear and tear upon its building, even though the building itself might have appreciated ir. value at the same time.
Depreciation in its broad sense, like appreciation, may be due to extrinsic causes, but that is not true of wear and tear. There is no reason why wear and tear, purely intrinsic matters, need be tied up to appreciation resulting from extrinsic causes. The two can go on simultaneously and no provision of law requires the one to be offset against the other. The revenue acts do not contemplate the annual computation of appreciation for the purpose of taxing the increase of value represented thereby — they expect appreciation to be accounted for entirely and exclusively at the final disposition of the property. This is quite logical, for an appreciation due to extrinsic circumstances, such as general market changes, is very likely to be offset by a corresponding reduction in value because of similar extrinsic causes in a later period, while the property is still held by the same owner; and it would obviously be highly inconvenient if not impossible, as well as unfair, to require every taxpayer to appraise all his capital assets at the end of every year for the purpose of seeing whether there were any net appreciation in value during the year and of taxing him thereon. But exhaustion and wear and tear are due wholly to intrinsic causes — the nature of the asset. They are not offset by any other intrinsic causes. Machinery and buildings do not gain in value, like cheese or wine, merely through lapse of time, nor do they improve with use (except possibly for a negligible *362“breaking in” period in the case of some machines). On the contrary, use and lapse of time reduce their value, gradually, ordinarily at fairly constant rates, and this loss of value is recovered by their owners through sale of products or through rents received, not all at once on final disposition as appreciation is realized, but regularly through the period of their use. So Congress has provided that this loss of value shall be considered, as it occurs, and is recovered, through the life of the property.
Obsolescence, being sometimes due to extrinsic causes, may be offset by other extrinsic causes. To the extent that it is so offset it may fairly be said not to occur. But to the extent that it is not so offset and does in fact occur, it partakes in the nature of exhaustion and wear and tear and has been similarly treated by Congress.
As exhaustion, wear and tear, and obsolescence occur, and are recovered, so they should be treated in tax returns, that is, they should be accounted for annually. And they should not be offset by unrealized appreciations. Otherwise many absurd situations would result. Suppose, for example, that a taxpayer erects a building at a cost of $10,000 with a normal life of 20 years. Immediately afterward the general cost of materials and labor goes up so that by the end of the year it would cost $20,000 to replace it, and this condition continues for 10 years. During this period, although the taxpayer would have consumed one-half of the original value of the building and would have recovered it through sales or rents received, he would be allowed no deduction for wear and tear if appreciation were an offset thereto. Then suppose, in one year, changes in the neighborhood, fear of invasion, competition, or other extrinsic cause’ should reduce the market value of the building to $5,000— what would be the proper deduction for wear and tear in that year % If only $500, and the condition continued until the building exhausted its natural life and became worthless, the taxpayer would by the end of the period have recovered only half of his investment and would have paid taxes on the other half as if it were profit. If $5,000 were allowed as the deduction for the year, it could not with propi’iety be attributed to wear and tear in that period; it might so far exceed the profits of the year and the two succeeding years that fro tanto it would avail the taxpayer nothing; and the rule under which it was allowed would in effect require every taxpayer to appraise all capital assets every year to account for losses at least. Frequently they would be extrinsic losses which would be offset by extrinsic appreciation in other years, and this appreciation could not be taxed until realized by sale or other disposition, which might never occur, since the property might continue in use until wholly consumed, or destroyed by casualty.
At the hearing of this appeal, counsel for the taxpayer laid stress upon the language of the statute providing that the deduction shall be allowed. He insisted that allowance indicated an option to the taxpayer to take the deduction or not as he saw fit. Of course a taxpayer may neglect or refuse to make a correct computation of income in a given year and pay a greater tax than he owes — and nobody will force the excess tax back upon him. But that does not entitle him to adjust matters by an improper computation in a later year when *363the tax rate is higher or when the statute of limitations has precluded the correction of the earlier return. We do not think the word allow intended any option. As used, it merely means consider or subtract. One weighing goods in packages allows for the tare in figuring net weight. In computing profit on a sale one allows or makes allowance for cost when he deducts it from the sale price. Instead of saying “ the following amounts shall be subtracted,” Congress said “ there shall be allowed as deductions,” meaning the same thing. And where it said “ a reasonable allowance for the exhaustion ” it merely meant a reasonable amount. To hold that the use of the word allow and its derivatives indicated an option in the taxpayer to claim his deduction, or to treat the fact entitling him thereto as if it had not occurred and to make future returns on the basis of its nonoccurrence, would lead to many absurdities. We can not hold that Congress intended a distinction between the treatment of other deductions which it said “ shall be allowed ” and of this deduction, because it tautologically said “a reasonable allowance.” And how are the other allowed deductions treated? If a taxpayer owned a plot of land with two buildings on it, which he bought as a unit, and one of the buildings was destroyed by fire in 1916, but he claimed no deduction for the loss in his 1916 return, should he be permitted on selling the property in 1920 to compute his income without reference to the 1916 loss, thus in effect transferring his loss from a year of low tax rates to one of high rates? Per contra, if the fire loss occurred in 1918 and the sale in 1922, should the taxpayer be denied the right to amend his 1918 return and show the truth on the ground that he had exercised an option and must be bound by it, though to his detriment because of the reduction in tax rates under the 1921 act? The only alternatives to such a ruling would be (1) to put every taxpayer in a position to postpone final determination of his taxes until just before the expiration of the statute of limitations and then to throw his deductions into the year in which they would benefit him the most, or else, and we think this the correct rule, (2) to hold strictly that deductions must all be taken in the year in which their allowance is provided for by statute, and that failure to take them does not preclude the Commissioner, in computing taxes for subsequent years, from determining the taxes for those years upon the facts applicable thereto.
The taxpayer held its building, upon which presumably there was wear and tear — and it has not shown that there was not — for a period of years. It must be presumed to have recovered or to have lost this reduction in value in the respective years of ownership to which it is attributable according to approved methods of accounting. Such recoveries or losses may still be attributed to the proper years, except as barred by statutes of limitation. But whether or not the taxpayer may still make these adjustments, the computation of gain upon the sale of the property must be made with due allowance for the adjustments which should have been made during the period of ownership.
This brings us to the question of the proper basis or starting point for the computation of the gain. The Revenue Act of 1918 provides that, in case of property acquired before March 1, 1913, the *364basis shall be the value on that date. But the Supreme Court has held in Goodrich v. Edwards, 255 U. S. 527, and Walsh v. Brewster, 255 U. S. 536, that if the cost were greater than the value on March 1, 1913, the taxpayer is entitled to use cost as a basis. These cases were decided upon a concession by the Solicitor General that the Constitution necessitated the conclusion reached, and though the court did not say so in its opinions we feel that the conceded constitutional ground and not any rule of construction must have furnished the ratio decidendi of the court’s judgments. We are satisfied that Congress intended to make the March 1,1913, value the basis for computing ■'gain or loss. If the Constitution gives a taxpayer the right to recover the cost of his property, upon selling it, where that cost exceeds the March 1, 1913, value, before accounting for gain, it must follow that if he recovers the value of his property through deductions for exhaustion, wear and tear, and obsolescence, through Avriting off losses by casualty or theft, or otherwise, he must be entitled to the same capital recoA^ery before accounting for profit thereon. It Avould be anomalous to say that the Constitution entitles a man to take out of his gross receipts from the sale of his property its cost to him, before reporting income, but that he is entitled to take out of his gross receipts from the consumption of the same property in producing his income, only a lesser sum. (Cf. Appeal of Grosvenor Atterbury, 1 B. T. A. 169.)
The same considerations that lead us to the conclusion that adjustment for recoveries of capital by allowance for exhaustion, wear and tear, and obsolescence must be made in computing gain upon the sale of property, compel us to the belief that similar adjustments should be made to cost before comparing it with value on March 1, 1913, for the purpose of deciding Avhich of them should be the basis for that computation. If the taxpayer recovered a part of the cost of his property before March 1, 1913, only the balance of that cost can properly be recoAdorable thereafter. The Constitution certainly does not, entitle a taxpayer to recover any part of his cost more than once, before becoming accountable for taxes upon his gain. If, after proper adjustment for partial recoveries, it appears that the cost exceeds the value at March 1, 1913, that adjusted cost rather than the March 1, 1913 value should be taken as the basis for all subsequent computations; if it be less than the March 1, 1913 value the latter is the proper basis. Thus, if a taxpayer in 1908 buys a building with a normal life of 20 years for $10,000, and recovers in rents one-half of that cost by 1913, he is entitled to recover thereafter through deductions or upon the sale of the property either $5,000 or the market value at March 1, 1913, whichever is higher. To allow more would be permitting him a double recovery of part of his capital investment before accounting for profit, and certainly the Constitution does not compel that.
In the instant appeal it is apparent that the cost adjusted to March 1, 1913, is higher than the appraised value at that date. Therefore, that adjusted cost becomes the proper basis for the computation of gain at the sale of the property, which computaion would include proper allowance for exhaustion, wear and tear, and obsolescence occurring in the time between March 1, 1913, and the date of sale.
*365Whether we (1) adjust the original cost to 1913 and then again adjust the balance to 1920 on the basis of an expected life of the property shortened by the time elapsed between acquisition and March 1,1913, or (2) adjust the cost on the basis of an undiminished life directly from acquisition to date of sale, we shall arrive at the same figure. This is demonstrable by a simple example: assume cost of a building with a 20-year life to have been $10,000 in 1903, and that it sold for $5,000 in 1918. Cost adjusted by a recovery of 5 per cent per annum to 1913 would be $5,000, and the building would have a remaining life of 10 years. This $5,000 basis, adjusted to 1918 on account of a 10 per cent annual recovery, would be $2,500. If the cost of $10,000 were adjusted at 5 per cent per annum from 1903 to 1918, it would be $2,500. It makes no difference which method of computation is employed.
The cost of the land and building together in 1909, in the instant appeal, was $41,942.36, but we have no evidence of how much of this cost was attributable to the building and how much to the land. The parties should determine the part of the cost attributable to the building either by agreement or by the submission of evidence to this Board upon the settlement of final determination; this cost should be adjusted by proper allowance for exhaustion, wear and tear and obsolescence from date of acquisition to date of sale, and the adjusted cost subtracted from the sale price to determine the gain upon the sale. This gain should be included in the taxpayer’s gross income for 1920 in making a recomputation of its tax deficiency for that year.
If the parties are unable to agree upon the facts necessary to a computation and the calculation of the deficiency in accordance with this opinion, we will hear, upon motion for final settlement of our determination, evidence of the cost attributable to the building and of the probable useful life of the building.
On consideration by the Board, Smith and Sternhagen dissent.