TUTTLE, Circuit Judge.

This is an appeal from a district court judgment for the Government in an action to recover taxes allegedly improperly imposed. Upon agreed facts the only issue is whether under the provisions of Section 23(f) of the 1939 Internal Revenue Code, 26 U.S.C.A. § 23(f), the Commissioner was correct in permitting taxpayer, appellant's transferor, to deduct only that portion of hurricane damage to his depreciable business property which constituted the same percentage of his adjusted cost basis of the entire property that the loss was of the total pre-hurricane market value. Taxpayer contends that the entire actual loss should have been allowed, up to an amount equal to the adjusted cost basis of the entire property.

In 1945 taxpayer's citrus groves were partially destroyed by a hurricane. From the stipulation of the parties the following pertinent figures appear:

Fair market value of property before hurricane $1,571,575.00 Fair market value of property after hurricane 1,380,025.00 _____________ Hurricane damage $ 191,500.00

191,500 Percentage of loss: --------- = 12.185615% 1,571,575

Adjusted (depreciated) cost basis of property before hurricane $ 523,479.77

Loss allowed by the Commissioner: $523,479 × .12185615 = $ 63,789.23 Additional deduction taxpayer claims $ 127,710.77

The court below approved the Commissioner's formula and computation, holding that the partial destruction was in effect a pro tanto disposition of part of the property for which taxpayer could deduct losses not in excess of the proper fractional share of the adjusted basis. The court and the Commissioner rely primarily on rules of the Internal Revenue Service first published in 1929 and apparently in effect at the time of the loss. It is not asserted that the computation has any direct statutory source, nor is it based on any regulation.

No court cases approving or even applying, the Commissioner's formula are cited by either party, though it is recited with apparent approval in a dicta footnote in United States v. Koshland (9 Cir.), 208 F.2d 636, at page 639 footnote 3. The Commissioner relies on several Tax Court cases, Frazer v. C.I.R., 10 B.T.A. 409; Krome v. C.I.R., 19 P-H T.C.Mem.Dec. ¶ 50,064; and the recent Knapp v. C.I.R., 23 T.C. 716, all of which apply the Commissioner's formula in situations analogous to the one here, without, however, discussing it or showing that the validity of the formula had been placed in issue or even considered; taxpayer points to the decision in Hinman v. C.I.R., 12 T.C.M. 1347, in which the computation in a similar case was according to taxpayer's contention, but again without discussion of the formula.

If this were entirely a matter of first impression it could certainly be said that nothing in the pertinent statutory provisions directly supports the Commissioner's attempt to limit allowable loss in case of the partial destruction of business property to a fraction of the adjusted basis rather than to the full amount of the adjusted basis. Even granting that the allowance of any deductions is a matter of legislative grace, where the statute explicitly provides for one the Commissioner cannot cut it down without specific statutory authority. Nor, as discussed further below, are we impressed either by the arguments by analogy from the computations applicable to partial sales, or by the policy grounds suggested, that the Commissioner's formula follows from the statutory pattern of the revenue codes by logical implication.

However, this question has not been left entirely open. The Supreme Court's decision in Helvering v. Owens, 305 U.S. 468, 59 S.Ct. 260, 83 L.Ed. 292, has explicitly determined that allowable casualty loss is to be the actual decrease in the market value of the property, measured by the difference in market values immediately before and immediately after the casualty, but limited to the total adjusted basis of the property; in that case this formula was actually applied in a partial loss situation.

The Owens case involved property not used in trade or business and thus not depreciable (though casualty losses were deductible under § 23(e)(3)), and the Commissioner attempts to distinguish it on that ground. It is true that the Supreme Court explicitly stated that it was there dealing with non-business property but this evidently merely referred to the fact that the particular formula there contended for by the taxpayer, and rejected by the Court, was one really objectionable only where applied to non-depreciable property, for it would have allowed the deduction of all past, non-allowable depreciation as part of the casualty loss; the formula there advanced by the Commissioner and approved by the Court is one that, on the basis of the statutory language of the provisions on which it relies, is equally applicable to business as to nonbusiness property. Merely because the rejected formula was one suggested only for nonbusiness property does not mean that the approved formula is similarly limited.

There is no statutory basis for making the distinction asserted by the Commissioner. The types of allowable loss were set forth in §§ 23(e) (1-3), 23(f) of the 1939 Code; their separate statement was evidently necessary as an enumeration of all the situations that Congress wished to cover, but nothing in the statute suggests that the deductions allowed by these four categories are to be computed by formulae differing from one category to another. In fact the subsequent statutory references relating to the computation of the allowable loss, principally § 23(i), treated all four categories together and were equally applicable to all of them — keeping in mind that the "basis," which according to the Supreme Court is merely a maximum limit on the loss, would, by virtue of § 113(b), assume different values for depreciable than for non-depreciable property.

There are a number of other grounds on which the computation in Helvering v. Owens might be distinguished from that applicable here without relying on a distinction that is evidently avoided by the statute, but none is urged here by the Commissioner nor are his rules or regulations based on or consistent with any of them. First it should be noted that the Owens case did not deal explicitly with the implications of a partial loss situation — this appears clearly by considering the facts of the companion case. The issue in both was not the extent to which a partial loss should be allowed, but how the amount of loss of nondepreciable property was to be computed. Nowhere from the records of the case below, or from the briefs in the Supreme Court does it appear that the formula here urged by the Commissioner was even suggested to the Court as an alternative by either party. The Commissioner, however, has by now accepted the Owens case as defining the correct method of computing the allow-ability of a partial loss — but for non-business property only.

Secondly, the Owens case involved a situation in which the property had actually depreciated in value and thus the pre-casualty market value was lower than the adjusted (undepreciated) basis; here, however, the property has appreciated (or at least depreciated less rapidly than the write-off allowed) and its pre-casualty market value is in excess of the adjusted basis. The application of the formula here advanced by the Commissioner to the Owens type situation leads to the apparently incongruous and possibly unauthorized result of allowing a deduction in excess of out-of-pocket loss. But even though tax formulae are often not required to be absolutely symmetrical for gains and for losses the Commissioner does not urge this distinction either — in fact he accepts the possibility of his formula allowing deductions in excess of out-of-pocket loss.

Finally, the Owens case involved injury to a single automobile which cannot be destroyed piece by piece without affecting the utility of the whole, and which thus clearly has an undivisible "basis." The same is not necessarily true of a citrus grove, where the destruction of some of the trees throughout the grove, or perhaps of all the trees in a portion of the grove, leaving the rest of the trees productive, might allow for the matching of the destroyed property with particular portions of the "basis"; clearly for some kinds of property physical separability means that each portion has its own "basis." Again the Commissioner does not urge this distinction, and we will therefore not explore this possibility.

The first objective of the courts is to follow the statute, and Owens v. Helvering defines what the somewhat ambiguous words of the Code mean in a situation indistinguishable from that presented here on any grounds that the Commissioner has seen fit to urge. The Commissioner strongly argues that great weight should be given to an administrative construction established in 1929 and allegedly maintained ever since and recently affirmed twice, especially since Congress meanwhile has several times examined and readopted without substantial change the basic statutory provisions. We are by no means convinced that the position of the Internal Revenue Service has indeed been clear, consistent, and constant throughout this period, but in any event the Commissioner is merely relying on a rule of construction used only if the meaning of a statute cannot be more clearly discerned from its text or from other authoritative interpretations of the language. Such a rule cannot be used to create a distinction where the statute itself has made none.

Even aside from the perhaps vulnerable authority of the Owens case the Commissioner has not persuaded us that his treatment of partial losses of business property is so clearly reasonable, and the alternative formula is so unreasonable, that his rule follows by necessary implication from, or is even allowed by the statute, which does not explicitly establish this limitation for a partial loss situation. Nevertheless, we think the arguments merit separate consideration and discussion.

The Commissioner argues that for depreciable property it is intended that its cost be recovered through depreciation write-off over the expected life of the property — but if taxpayer's approach were accepted a loss of 34% of his property would be sufficient to permit "recovery" of his entire basis. Depreciation formulae are at best rules of thumb which attempt to approximate economic reality, and should not be used to distort it when they are clearly inapplicable. Where a partial loss of an indivisible business property is suffered it may well be necessary to restore the damage in full by immediate repairs and replacements before any portion of the property is again usable. It seems reasonable, and contrary to nothing called to our attention in the statute that Congress would allow one who is exposed to such an expense to take his deduction immediately, if he chooses, without prorating the recovery of part of the loss and expense over the remaining life of the property; it should be noted that the restoration of a casualty loss differs both economically and taxwise from a new capital investment. At most, taxpayer thus receives an opportunity to take his deduction earlier rather than later, without affecting the cumulative total of allowable depreciation. If this flexibility were not allowed, the owner of business property destroyed in part by a casualty might actually be worse off on two counts than the owner of similarly destroyed non-business property; the latter may not only take his entire deduction at once, but may also deduct his full loss up to an undepreciated cost basis which, in time of inflation, will almost always allow him a greater deduction than if the limit were depreciated cost.

The Commissioner additionally urges that the same rule be adopted for partial casualty losses of property as is used for partial sales, i.e. the Commissioner's formula — as for instance in Columbia Oil Gas Co. v. Commissioner (5 Cir.) 118 F.2d 459. For this proposition he relies on Pioneer Cooperage Co. v. Commissioner (8 Cir.) 53 F.2d 43, decided under the considerably different language of the 1918 Revenue Act; the more recent case of Helvering v. William Flaccus Oak Leather Co., 313 U.S. 247, 61 S.Ct. 878, 85 L.Ed. 1310, bids us to be wary of any such analogy. Moreover, the suggestion is not inherently persuasive. A partial sale indicates that there is at least an economic divisibility of the property, and it seems reasonable to apportion the basis of the property according to what is kept and what is disposed of. A partial casualty loss indicates no such divisibility for frequently the entire damage must be restored before the whole is to have any productive value. Furthermore a sale is generally voluntary and endows the seller with assets roughly equivalent to the value of the part sold, which he may productively reinvest after the payment of some taxes; an uninsured loss yields no such assets — rather there is a need to obtain funds to restore the property.

Finally the Commissioner objects that the taxpayer's formula in effect allows him to take losses against anticipated profits from the appreciation of his property, for which he has as yet paid no taxes. This argument is deceptive. The only real amount is the out-of-pocket loss suffered by the taxpayer; this loss might indeed be larger than otherwise because of the (as yet untaxed) appreciation in the value and cost of the property, but it is nevertheless a real loss. This loss can in any event be deducted only to the extent of the original investment reduced by the previously allowed depreciation. The appreciation can still be taxed when and if actually realized; if the property should later depreciate before it is sold it is the taxpayer who is ultimately injured by having suffered his loss at the time when prices and presumably replacement costs were high — he should not be penalized taxwise because he "realized" his "profits" through involuntary conversion at a time when prices happen to be high.

The decision below must therefore be reversed, with judgment to be entered for the appellant.

* * * * *

* * * * *

* * * * *

"LOSSES OF NONBUSINESS PROPERTY