The sole question for determination on this petition for review of the decision of the Tax Court of the United States is whether a loss to a dwelling house caused by termites is deductible from income as a "casualty" within the meaning of Section 23(e) (3) of the Internal Revenue Code, 26 U.S.C.1946 ed. § 23. The pertinent portions of that section are:

"§ 23. Deductions from gross income.

"In computing net income there shall be allowed as deductions:

* * * * * *

"(e) Losses by individuals. In the case of an individual, losses sustained during the taxable year and not compensated for by insurance or otherwise —

* * * * * *

"(3) of property not connected with the trade or business, if the loss arises from fires, storms, shipwreck, or other casualty, or from theft. * * *" The facts are not disputed. The findings of fact of the Tax Court are:

"Petitioner's individual income tax return for the taxable year 1947 was filed with the collector for the first district of Missouri at St. Louis, Missouri. He resides with his wife at No. 4 Prado Drive, Ladue 5, St. Louis County, Missouri. He purchased the property in April, 1946, for $38,500. Prior to purchase he had the property inspected to determine whether there were termites by one Schlesinger, a builder and architect of over 15 years' experience. The architect checked the joists with a sharp-pointed instrument and by pounding with a hammer and informed petitioner that he was satisfied that the construction was sound. Petitioner thereupon purchased the property, moved into it in September, 1946, and has resided therein since that time.

"In April, 1947, termites were discovered and after Schlesinger had been called and agreed there were termites an examination was made by an entomologist, a specialist on termites, and it was found by him that damage done by termites was confined to a small area. A joist in the basement, measuring 2 inches by 8 inches, had been affected to a length of about 10 or 12 feet. The rest of the joist was solid and it was only necessary to nail to it a small piece of wood where the joist had possibly been weakened. Also, the sills and jambs of a picture window were affected by the termites to the extent of about 8 or 10 inches. That part of the wood was sawed off and a new piece of wood put in to replace it. Because of damage to the sill of a plate glass window it was necessary to take the plate glass out and replace the subsill and the finish sill and put in a new jamb and then replace the plate glass. The termite expert treated the entire building for termites. The total amount expended for treatment and the repairs made was $1800.74 and that amount was claimed by the petitioner as a deduction in his income tax return for 1947 as `Damage to No. 4 Prado Drive caused by Termites.' The claim was denied.

"Termites are social insects and the infestation by the termites in petitioner's home was a colony project rather than the work of a few individual insects. The known presence of termites in a house makes it difficult to sell it."

We need only add that the area affected (other than the floor joist), described as a "small area", consisted of an area of five or six feet adjacent to the point of entry of the termites through a small crack in the foundation and that the colony of termites was of recent origin at the date of the inspection in April, 1947. Since the Tax Court found that whatever loss or damage was occasioned by the termites was not deductible, it did not determine the amount of the loss.

The question presented is not new. The Ninth Circuit held in United States v. Rogers, 120 F.2d 244, 246, that the word "casualty" connoted the sense of suddenness, and that as used in Sec. 23, Webster's definition applied:

"`1. Chance; accident; contingency; also, that which comes without design or without being foreseen; an accident * * *

"`2. An unfortunate occurrence; a mischance; a mishap; a serious or fatal accident; a disaster * * *.'"

"It can be seen that `casualty' may properly be used in the sense of `accident'. The latter word is defined by the same source as `An event that takes place without one's foresight or expectation; an undesigned, sudden, and unexpected event'. Showing that casualty may have the sense of suddenness is the definition in 1 Bouv.Law Dict., Rawle's 3d Rev., p. 430, as follows: `Inevitable accident. Unforeseen circumstances not to be guarded against by human agency, and in which man takes no part'. Such definition is in accord with that given in Matheson v. Commissioner of Internal Revenue, 2 Cir., 54 F.2d 537, 539.

"Since damage by termites or dry rot is not a sudden occurrence but is a development over a longer period of time we think the deduction was improper."

On rehearing, United States v. Rogers, 9 Cir., 122 F.2d 485, that court applied the doctrine of ejusdem generis and held that the statute should be construed as though it read "`loss by fires, storms, shipwrecks, or other casualty of the same kind'", and held that the class was distinguished by the suddenness of the loss, saying:

"The similar quality of loss by fire, storm or shipwreck is in the suddenness of the loss, so that the doctrine requires us to interpret the statute as though it read `fires, storms, shipwrecks or other sudden casualty'."

The Second Circuit in Fay v. Helvering, 120 F.2d 253, again held that loss from termite destruction was not a "casualty", saying:

"It is not necessary to say whether or not the word `casualty' should be limited by its context under the doctrine, ejusdem generis. Even though it had been used alone we should not have held that it covered such a loss as this; we agree with the Ninth Circuit which held that exactly this kind of destruction was not a `casualty,' United States v. Rogers, 9 Cir., 120 F.2d 244. That word denotes an accident, a mishap, some sudden invasion by a hostile agency; it excludes the progressive deterioration of property through a steadily operating cause. Our decision in Matheson v. Commissioner, 2 Cir., 54 F.2d 537, is very close aboard. There we held that the injury done to the taxpayer's house was not a `casualty' when worms had eaten the piles on which it stood, and water had rusted steel reinforcing bars. Both piers and bars were protected by concrete, which water washed away from the piers, allowing worms to attack the wood; and which water also penetrated to corrode the steel. All this had, however, been a gradual process and on that account we held that the loss was not a `casualty.' If there be any general principle applicable at all, the cases are indistinguishable."

In Matheson v. Commissioner, 2 Cir., 54 F.2d 537, 539, cited in Fay v. Helvering, we note that the conclusion reached when that case was decided appears to have been based upon the fact that the building was improperly constructed, resulting in more than normally rapid depreciation rather than from any effect from pilings being eaten away by worms. In the Matheson case the court said:

"In our opinion, the word `casualty' as used in section 214(a) (6) is an event due to some sudden, unexpected, or unusual cause. Anything less than this renders it hardly distinguishable from depreciation from ordinary wear and tear which cannot be deducted by a taxpayer in the case of property that is not used in trade or business. Section 214(a) (8). While it is contended that the concrete casing of the piles was washed away by storms and tides so that the piles became exposed, eaten by worms, and weakened, the evidence that there were no unusual winds and that there was nothing to cause extraordinary tides afforded a substantial basis for the conclusion of the Board of Tax Appeals that the loss was due only to the ordinary action of the elements upon a poorly constructed building. In such circumstances we are bound by its findings and hold that the damage did not arise from a casualty resembling `fires, storms or shipwreck.' The disintegration was due to encasement of the steel framework in walls of concrete so thin as to let in moisture and give rise to corrosion, and to inadequate sheathing of the piles. This conclusion seems inevitable when the proof of any serious storm or cataclysmic operation of the elements failed. No time is satisfactorily established when any serious injury to the foundation occurred. It was never overwhelmed by any unusual thing. There was simply a steady labefaction from wind and weather more rapid than usual because of structural defects. Such a deterioration cannot reasonably be termed a casualty. Crystal Spring Distillery Co. v. Cox, (C.C. A.), [6 Cir.,] 49 F. 555; Forsdick v. Board of Supervisors of Quitman County, (Miss.) 25 So. 294; Eaton v. Glindeman, 33 Idaho 389, 195 P. 90. (Italics ours.)

* * * * * *

"We do not suggest that the destruction of a building by an unusual storm would not be a `casualty' if the edifice was so badly constructed as to be readily destroyed. We simply hold that a loss due, as here, to progressive decay or corrosion, occurring without any unusual action by the elements, does not arise from a `casualty' within the meaning of the statute." (Italics supplied.)

While both the Rogers and Fay cases involved loss caused by termites, the facts involved are in the respect that the deterioration was very slow in developing different in those cases from the facts in this case. In the Rogers case the house was purchased in 1920 and improved in the following year. Not until 1929 was extensive damage "from some cause of the nature of termites or dry rot" discovered. And then the damage was so extensive that upon the advice of an architect the entire building was razed. In the Fay case the house was built in 1913 and in 1935 it was discovered that termites had eaten away the entire strength of the wooden framework of the columns and horizontal beams that supported the roofs of a porch at either end of the house. The court observed that — "The insects had obviously been at work for a long time, and the loss had therefore in fact taken place gradually although it was not discovered until it was complete."

In Crystal Springs Distillery Co. v. Cox, 6 Cir., 49 F. 555, 559, a "casualty", as used in a statute relating to "`accidental fire or other casualty'" was defined as "an accidental destruction by some cause of like character and operation as fire; such as lightning, floods, cyclones, storms, or other uncontrollable force," as distinguished from casualties which ordinary foresight and prudence could guard against or prevent. But in Shearer v. Anderson, 16 F.2d 995, 51 A.L.R. 534, the Second Circuit rejected the theory that in order to constitute a "casualty" it must be some act of nature analogous to storms or fire due to natural causes, and not to some casualty due to negligence.

We deem it appropriate to note that from our examination of the authorities that through the cases, and particularly the Rogers and Fay cases, runs the dominant thought that the primary characteristic of casualties which will fall within the meaning of Sec. 23 is that of the suddenness of the occurrence. As said in Matheson v. Commissioner, supra — "the word `casualty' * * * is an event due to some sudden, unexpected, or unusual cause. Anything less than this renders it hardly distinguishable from depreciation from ordinary wear and tear which cannot be deducted by a taxpayer in the case of property that is not used in trade or business." (Italics ours.)

We note that the Tax Court in applying the rule of ejusdem generis has treated as "casualties" and deductible, occurrences which were found to be (a) accidents, or (b) sudden occurrences. Loss from the freezing of water pipes in a residence during the absence of the occupant, Seward City Mills v. Commissioner, 44 B.T.A. 173, loss occasioned by the bursting of a boiler in a heating plant of a residence (I.T. 2231, IV-2 Cum.Bull. 53 (1925), and loss due to an extensive gradual vertical sinking of land caused by a subterranean disturbance, Grant v. Commissioner, 30 B.T.A. 1028, have been held to fall within the class of "other casualties" by the Tax Court. And in Whipple v. United States, 25 F.2d 520, the District Court in Massachusetts held that damage to trees on residential property caused by a sleet and ice storm was a casualty and deductible. In Appeal of Leland Webb, 1 B.T.A. 759, a deduction was allowed for damage to household goods and an automobile from storms encountered while the articles were being shipped by sea. And in Burkett v. Commissioner of Internal Revenue, CCH Dec. 18,581 (M), Dkt. 27,344, a loss occasioned by the washing away of sand from wooden foundation posts under a house in a hurricane, causing the posts to gradually sink over a period of two years, eventually requiring the replacement of many of the posts in order to level the floors, was held to be a loss by a casualty.

We agree with the definition of "other casualty" found in the Rogers and Fay cases that the occurrence must be sudden. We agree that the doctrine of ejusdem generis should be applied and that the characteristic of suddenness common to fires, storms, and shipwrecks must exist in order that the occurrence causing the loss be a casualty within the meaning of Sec. 23. But it is inherent in the application of the rule of ejusdem generis that the class of occurrences which will constitute "other casualties" will not be confined to fires, storms, or shipwrecks alone. To do so would violate the intent of the rule and also the intent of Congress in inserting the words "or other casualty" in the Act in 1916. We do not assume that the courts which have held that losses caused by termites were not deductible were intending to hold that loss or damage from an invasion of termites was under no circumstance deductible. It is clear from those opinions that the lack of demonstrated suddenness of the loss was the controlling factor in those decisions. If the loss from termites had been shown to have been sudden in those cases and the criterion of a sudden invasion of a hostile agency had been applied, it seems reasonably clear that the result would have been different. We understand the doctrine enunciated by those cases to be that if the operating forces which give rise to a loss are shown to be a sudden invasion of a hostile agency, the loss will be deductible as a casualty. Naturally, we are not now concerned with the question existing in the Fay and Rogers cases of whether the evidence in a particular case showed that an operating force, capable of sudden destruction, actually did or did not act with suddenness, for as we shall demonstrate, the proof in this case does show that the operating force did act with suddenness and was a hostile agency.

Suddenness is a comparative term. That is demonstrated not only by the ordinarily accepted meaning of the word, but also by the fact that the courts have held that the element of suddenness existed in a great variety of occurrences where the degree of suddenness varied widely. And being comparative, it is necessary that in each case the class of occurrences which is to be excluded as not being sudden be ascertained in order to identify the opposite element for making the comparison. In that manner the segregation of those things which should fall into the intended meaning of things sudden from those which were intended to be excluded can best be accomplished.

The loss under consideration is the deterioration of a building. We know that some kinds of deterioration of residence buildings are intended to be deductible because for many years losses from deterioration of such buildings arising from fires and storms have been specifically made deductible by statute. We know that by the insertion of the words "or other casualty" in the statute after the words "fires," "storms," "shipwrecks" in 1916, Congress intended to enlarge the kinds of deterioration of buildings which were to be deductible losses. We know that by the juxtaposition of the inserted words "or other casualty", the application of the rule of ejusdem generis and from the ordinary meaning of the word "casualty", that all kinds of deterioration of residence buildings were not intended to be deductible losses. What character of losses from deterioration was Congress intending to exclude from deductible losses by the use of the words "or other casualty" in the manner those words were used in the statute? We believe it reasonably clear that losses resulting from deterioration caused by normal depreciation, as distinguished from sudden deterioration, constituted the class of losses which Congress intended should not be deductible, and that losses from normal depreciation constitute the other element of comparison for segregating sudden or deductible losses from those which are not sudden or deductible.

It is commonly known that residence buildings in this climate, properly constructed and maintained, have an ordinary life of from 25 to 50 years, or even longer, the variation between those periods resulting from the type of materials used. Comparatively speaking, an invasion of a colony of termites which destroys the timbers of a building in a month, three months, or a year, is a sudden destruction, when from natural depreciation it would have required from 25 to 50 years or longer for them to have been substantially injured. The damage done by termites in a relatively short time may not have the same dramatic characteristics as destruction reasonably to be anticipated from the invasion of a bull into a china closet, but the damage from termites would be as effective and possibly more so, and when the time necessary to accomplish substantial loss in both instances is compared with the time necessary to result in damage from natural depreciation, both would be relatively very sudden.

The facts in this case may be and probably are unusual in that the time within which the damage or loss occurred could be and was definitely fixed within a relatively short time. We do not criticize the principle applied in United States v. Rogers or Fay v. Helvering, if as we understand those cases, they adopt as the principle for classification of losses deductible that the loss be occasioned by the sudden invasion of a hostile operating force. We adopt and follow that principle. The facts shown in the case at bar, unlike those in Matheson v. Commissioner, supra, and the Fay and Rogers cases, clearly distinguish the loss from ordinary wear and tear, and demonstrate that the invasion of the colony of termites and the resulting damage were both accidental and sudden. We are forced to the conclusion that the Tax Court erred in holding that the loss was not deductible. Since that court did not determine the amount of deductible loss, the cause is remanded for further proceedings not inconsistent with this opinion.

JOHNSEN, Circuit Judge (dissenting).

The majority opinion recognizes the principle of the Fay case, 2 Cir., 120 F.2d 253, and the Rogers case, 9 Cir., 120 F.2d 244; Id., 9 Cir., 122 F.2d 485, that general termite damage is not a loss arising from "other casualty", within the meaning of section 23(e) (3) of the Internal Revenue Code, because it represents mere "progressive deterioration of property through a steadily operating cause", but at the same time the opinion removes from the application of this principle such termite damage as occurs and is discovered during the tax year that the infestation has taken place.

My difficulty with this view is that it seems to me inconsistent with the underlying concept and scheme of the statute. To me, section 23(e) (3) is designed to cover character or kinds of operating forces, which will give rise to a deductible loss, without regard to the degree of their play. Thus, manifestly, "fires, storms, shipwreck" represent types of causes with whose extent of operation the statute has no concern. The term "other casualty", in its associationship, must equally be read, I think, to embrace only such types of causes.

In other words, an operating force is, I believe, either wholly within or wholly without the provisions of the statute, as a basis for loss deduction. It is difficult for me to see how under the language of the statute the consequences of termites or any other operating force can be said to constitute a casualty loss for a certain period of time and a non-casualty loss thereafter. To my mind, a cause, operating throughout in the same manner, either constitutes a casualty force in its entire operation or else it does not do so at all. Time of discovery in such a situation may affect the scope of the injury done, but it does not give rise to any metamorphosis in the nature of the force, the mode of its operation, or the character of the result. This is true as to "fires, storms, shipwreck," and it seems to me too that it equally must be regarded as characterizing any "other casualty" force under the statute.

I am unable to read the Fay and Rogers cases, supra, to imply any distinction between the termite damage done to the first foot of timber in a structure and that done to the one-hundredth foot. It is the operation of the termites, after their entry and multiplication, which causes the damage, and that operation is the same upon the one-hundredth foot as upon the first. Throughout, the whole thing is, as the Fay and Rogers cases recognize, "a gradual process," and the damage done is "not a sudden occurrence," within the concept of what constitutes, from its nature, mode of operation and character of result, a "casualty" force. The only difference in suddenness which can be said to exist between a situation where the operation of the termites is discovered after only a few feet of timber have been damaged, as here, and one where a hundred feet or more have been so damaged, is in the relativeness of the time elapsed before detection. That same element of suddenness would exist in relation to a difference in the time of discovery of "dry rot." It is not this kind of suddenness to which the Fay and Rogers cases refer as characterizing losses "from fires, storms, shipwreck, or other casualty".