Gotham Silk Hosiery Co. v. Artcraft Silk Hosiery Mills, Inc.

BIGGS, Circuit Judge. This is an appeal from a judgment denying a recovery of profits to a patentee from an infringer. See 48 E.Supp. 131. The patent was held to be valid and infringed by the District Court of Delaware as reported in its opinion, 1 F.Supp. 643, affirmed, 3 Cir., 72 F.2d 47, certiorari denied 293 U.S. 595, 55 S.Ct. 109, 79 L.Ed. 688. On November 21, 1934, the case was referred to a Special Master for an accounting. 1 The plaintiff waived damages and sought to recover the profits which it *212 alleged the defendant has received by reason of the infringement. On December 15, 1938, the Master filed his first report He found that the defendant had received advantages from the use of the patent. He held, using the defendant’s price lists and circulars as evidence, that the defendant itself had regarded the infringing hosiery as having a market value of $1 more per dozen pairs than the defendant’s equivalent non-infringing stockings. He held that 'this $1 difference represented a fair measure of the profit received by the defendant from the infringement, subject however to a deduction of 5% estimated by the Master as the additional cost necessarily expended by the defendant in manufacturing the infringing feature of the hose. The Master decided that the defendant had sold 202,106 dozen pairs of stocking! during the accounting period. He thereupon awarded to the plaintiff the sum of $192,000 with 'Certain costs. Exceptions to the report were filed and argument was heard on the exceptions. The court filed an opinion holding that the Master had attempted to determine profit without ascertaining the cost of manufacture to the defendant and its receipts from sales. Since profit ordinarily is deemed to consist of the difference' between costs and receipts from sales, the court, citing Providence Rubber Co. v. Goodyear, 9 Wall. 788, 19 L.Ed. 566, remanded the case to the Master for the determination of the profit according to the formula prescribed. See 33 F.Supp. 344. After the remand to the Master, the parties introduced no further evidence. On September 19, 1941, the Master submitted a second or revised report, finding that the defendant’s profits amounted to $202,106. He recommended that this sum be recovered by the plaintiff from the defendant and recommended also an allowance of $12,979. to the plaintiff to cover its necessary fee and expenses. After exceptions had been filed to the report, argument again was had before the District Court. The court held that the plaintiff had failed to sustain the burden of proof placed upon it and allowed only nominal damages. See D.C., 48*F.Supp. 131, at page 140. Since we have reached the conclusion that the District Court was in error in its application of the law, it is necessary, the nature of the case being as it is, to embark upon an extensive statement of the facts, only a part of which can be deemed to be in dispute. The controversy between the parties lies primarily in questions of law relating to the burden of proof to be sustained by the plaintiff and the defendant under the extraordinary circumstances presented by the instant case. Rule 52 (a) of the Rules of Civil Procedure, 28 U.S. C.A. following section 723c, provides that findings of fact made by the District Court (in this case embodied in the court’s opinion) shall not be set aside unless clearly erroneous. Where findings of fact made by the District Court in the case at bar are set aside, we do so because in our opinion they are clearly erroneous or because the District Judge has not applied the correct rule of law thereto. 2 The patent held to be infringed by the defendant is Tilles No. 1,824,636, for a stocking with an adjustable welt or top consisting of three spaced rows of picots or hemlocks which enable the wearer neatly to turn down the stocking to the proper length and attach the garters to the folded top without danger of runs. The stocking, therefore, is adjustable to three different lengths. The plaintiff calls its stockings embodying the feature of the patent “Gotham Adjustables.” The defendant designates its infringing stockings as “Trilength.” We will so refer to them. The monopoly of the patent covers only the feature of the welt. The period of infringement was from September 22, 1931, through October, 1934. Ono December 6, 1934, an order was entered by the Master requiring the defendant to file an account in respect to the following matters: “(1) The number of dozen pairs of such *213 stockings manufactured by defendant from September 2, 1931, to the date of the account. “(2) The number of dozen pairs of such stockings sold by defendant during each calendar month from September 22, 1931, to the date of account. “(3) The price received for each such dozen pairs of stockings. “(4) The cost of making or acquiring the said stockings so made or sold, indicating the various items in detail which contribute to the cost of manufacture and/or acquisition and/or sale. “(5) The total profits, gains, savings, and advantages derived or made by defendant through the making and/or purchase and/or sale of such stockings. “(6) The number of such stockings on hand on the date of the report. “(7) Each item in detail claimed by defendant as deductible costs in arriving at its stated profits.” The defendant’s answer was as follows: “(1) See attached sheet. “(2) See attached sheet. “(3) See attached sheet. “(4) No cost figures are available. “(5) None. (No separate records kept on these styles). “(6) None. “(7) No detail records kept on these styles.” The attached sheet referred to in answers 1, 2, and 3 sets out tabulations of the total pairs of infringing hose produced by months, the total dozens of pairs, the numbers of pairs shipped by quality, the total dozens of pairs shipped and the range of selling prices by style and quality. The total dozens produced during the entire infringing period are listed as 27,652%. The total is net with the exception of 81-11/12 dozen pairs worn, found to be defective, returned and credited. The source from which these tabulations were derived is defendant’s exhibit 16, known as the “Tri-Length Book.” It is a columnar book from which the first and last few pages are missing. This book purports to be a complete record of all TriLength stockings produced in the defendant’s plant and in two other plants manufacturing stockings for the defendant and sold by defendant during the accounting period. The stockings purport to be classified according to style number and quality; i. e. first, irregulars, and so forth. The price range of each style rather than the exact prices at which definite numbers of stockings were sold is given. The book was kept on a net basis, that is to say, each day returns were deducted from sales. The genuineness, authenticity and accuracy of the book has been challenged. It was not kept in the regular course of business by the defendant’s bookkeepers and clerks. Instead it was kept by a Miss Kleinschrodt, secretary to the defendant’s president, Jack Kugelman. Kugelman was in complete charge of the defendant’s business. He owned at least four-fifths of its stock. Only Kugelman and Miss Klciuschrodt knew of the existence of the TriLength book. Kugelman testified that during the pendency of the suit he decided to keep a record of the Tri-Length hose; that he then instructed Miss Kleinschrodt as to the method to be employed in keeping the Tri-Length book and that it was maintained under his immediate supervision. 3 Kugelman subsequently testified that he was unfamiliar with the books of defendant’s business and paid no attention to such details. 4 Miss Kleinschrodt testified that entries in the Tri-Length book were first made by her in the latter part of October, 1931. It is the contention of the plaintiff that certain style numbers were used to designate Tri-Length hose and only TriLength hose. The defendant insists, however, that the style numbers were used to designate certain types of hose irrespective of whether or not they had the TriLength top. The defendant declares that the only correct records as to whether or not its hose were Tri-Length were contained in pencilled notations on payroll sheets and shipping vouchers. All these records were destroyed. Kugelman admits that the defendant’s records (with certain exceptions hereinafter referred to) from which information pertinent to the issues of the number of Tri-Length hose sold and the cost of manufacturing and selling them could have been derived were destroyed deliberately after the commencement of the suit. The appearance of the Tri-Length book is astonishingly uniform, the ink and *214 writing being unvaried from month to month. 5 The record was kept in pairs of stockings rather than in dozens, the usual unit in the hosiery business. It is conceded that the vouchers and slips from which the information for the Tri-Length book was drawn were kept in dozens. Moreover, the supersedeas order entered by the District Court on December 8, 1932, expressly provided for the filing of the defendant’s report in terms of pairs of stockings. The entries in the Tri-Length book purport to cover a period commencing on September 22, 1931, and there was testimony that the entries were commenced in October of that year.. There is other corroborating evidence as to the specious nature of the book. Both the District Court and the Master found the entries in the book to be unworthy of belief and this conclusion is fully supported by the evidence. Important issues of fact are presented in the instant accounting. The first is the number of pairs of hose embodying the infringing feature sold by the defendant, the prices at which it sold such hosiery, and the portion of the price therefor fairly to be attributed to the infringing feature. A second issue is the amount of the defendant’s costs in manufacturing and selling such hose. The defendant, as we have stated, destroyed most of the primary evidence. The defendant had records which would have disclosed the number of ■ hose with .the infringing feature sold by it and the prices obtained therefor. It also possessed records which would have disclosed facts from which its costs of manufacturing and selling the hose with the infringing feature could have been found. Bearing these facts in mind the primary question of law before us is the degree of proof required from the plaintiff under the circumstances. As Lord Mansfield declared in Blatch v. Archer (Cowp. 63-65), “All evidence is to be weighed according to the proof which it was in the power of one side to have produced and in the power of the other side to have contradicted.” 6 In apportionment cases situations have arisen analogous to that of the case at bar either because of the deliberate concealment of evidence by an infringer or because of an infringer’s failure to keep proper records. Such a tort-feasor usually seeks refuge in Garretson v. Clark, 111 U. S. 120, 4 S.Ct. 291, 28 L.Ed. 371, 7 which required the plaintiff to prove with exactness the precise amount of damages due to defendant’s admitted infringement or to forego all recovery. The rule was temporary. Compare the prior decisions in Providence Rubber Co. v. Goodyear, 1869, 9 Wall. 788, 19 L.Ed. 566, and Seymour v. McCormick, 1853, 16 How. 480, 14 L.Ed. 1024. Decisions subsequent to Garretson v. Clark mitigated the severity of the principle enunciated in that case and applied to the infringer in respect to proof of apportionment the general principles of equity ordinarily used to measure a recovery against a trustee de son tort. It is especially appropriate to do this in the instant case where the defendant is a wilful wrongdoer. In the leading case of Westinghouse Electric & Mfg. Co. v. Wagner Electric & Mfg. Co., 225 U.S. 604, 620-622, 32 S.Ct. 691, 697, 56 L.Ed. 1222, 41 L.R.A.,N.S., 653, the Supreme Court, dealing with the burden of proof to be imposed upon the parties on a question of the apportionment of profits between an infringing feature and the device as a whole, held that the burden first rests on the plaintiff to prove that profits resulted to the defendant from its sale of the infringing feature and that the plaintiff must exhaust all means available to it for the ascertainment of profits resulting *215 from the infringing feature. But, citing with approval Brennan & Co. v. Dowagiac Mfg. Co., 6 Cir., 162 F. 476, the Supreme Court went on to say: “ 'In the present case the infringer’s conduct has been such as to preclude the belief that it had derived no advantage from the use of the plaintiff’s invention. * * * In these circumstances, upon whom is the burden of loss to fall? We think the law answers this question by declaring that it shall rest upon the wrongdoer, who has so confused his own with that of another that neither can he distinguished.’ ” The Supreme Court then proceeded to shift the burden of proof as to apportionment to the defendant. 8 We think that the principle of proof enunciated in the Westinghouse case is applicable to the questions of proof presented in the instant case. 9 , 10 We find support for this view in many of the decisions. In Armstrong v. Belding Bros. & Co., D.C., 280 F. 895, affirmed 2 Cir., 297 F. 728, certiorari denied 265 U.S. 585, 44 S.Ct. 459, 68 L.Ed. 1192, a patent infringement action, the defendant bluntly stated that its books and records had been lost. The court held that under the circumstances, the plaintiffs evidence as to the defendant’s costs should be accepted and that the defendant was in no position to object to the findings of the Master based upon such evidence. See also Yesbera v. Hardesty Mfg. Co., 6 Cir., 166 F. 120. In United States Frumentum Co. v. Lauhoff, 6 Cir., 216 F. 610, 615, there was a wilful infringement of a patent. The plaintiff was unable to show the exact amount of damages which it had sustained. The court found, however, that defendant’s sales were so large that there could be no doubt that there had been substantial damage to the plaintiff. “Under such circumstances,” declared the court, “to have plaintiff recover nothing, because the difficulty of absolutely definite proof is insuperable, is a result so unfortunate that, if avoidable, it should not be permitted.” See the cases cited in Judge Denison’s opinion in which proofs of approximate damages were held to be sufficient. A similar ruling was enunciated by this court in respect to the recovery of profits or damages in Baseball Display v. Star Ballplayer Co., 3 Cir., 35 F.2d *216 1, a case of willful infringement. 11 In the case of Producers & Refiners Corporation v. Lehmann, 8 Cir., 18 F.2d 492, the defendant made a sworn statement that it had kept no separate accounting of its production of infringing oils. Recovery was given on all oils, infringing as well as non-infringing. It should also be pointed out that for many years courts have permitted the use of expert testimony to ascertain the amount of damages where documentary evidence is lacking. See Sauntry v. United States and United States Frumentum Co. v. Lauhoff, supra. This practice has been incorporated in the patent law by the Act of Feb. 18, 1922, c. 58 § 8, 42 Stat. 392, 35 U.S.C.A. § 70 which in pertinent part reads as follows : “If on the proofs it shall appear that the complainant has suffered damage from the infringement or that the defendant has realized profits therefrom to which the complainant is justly entitled, but that such damages or profits are not susceptible of calculation and determination with reasonable certainty, the court may, on evidence tending to establish the same, in its discretion, receive opinion or expert testimony, which is hereby declared to be competent and admissible, subject to the general rules of evidence applicable to this character of testimony; and upon such evidence and all other evidence in the record the court may adjudge and decree the payment by the defendant to the complainant of a reasonable sum as profits or general damages for the infringement. * * *” In the recent case of Sheldon v. Metro-Goldwyn Corporation, 309 U.S. 390, 402, 406, 409, 60 S.Ct. 681, 685, 84 L.Ed. 825, the Supreme Court, dealing with a question of apportionment of profits resulting -to a motion picture company from the deliberate plagiarism of a play, stated that the “analogy found in cases of patent infringement is persuasive.” The Supreme Court, affirming the decision of the lower court, discussed the patent law and declaréd that the use of expert testimony was appropriate for accounting for profits and that “only a reasonable approximation,” “not mathematical exactness” was required. In the cited case the Circuit Court of Appeals for the Second Circuit, 106 F.2d at page 51, had fixed the plaintiff’s share at one-fifth of the net profits from the picture, deeming that to be a figure “which will favor the plaintiffs in every reasonable chance of error.” This approach met with the approval of the Supreme Court. In Mishawaka Mfg. Co. v. Kresge Co., 316 U.S. 203, 207, 62 S.Ct. 1022, 1025, 86 L.Ed. 1381, in passing upon the profits to be awarded to the plaintiff because of a trade-mark infringement, Mr. Justice Frankfurter stated: “There may well be a windfall to the trade-mark owner where it is impossible to isolate the profits which are attributable to the use of the infringing mark. But to hold otherwise would give the windfall to the wrongdoer. In the absence of his proving to the contrary, it promotes honesty and comports with experience to assume that the wrongdoer who makes profits from the sales of goods bearing a mark belonging to another was enabled to do so because he was drawing upon the good will generated by that mark.” The Supreme Court was dealing in the cited case with the question of apportionment of profits under the trade-mark statute precisely as in the Sheldon case it was dealing with apportionment of profits under the copyright law. These situations are assimilable to apportionment under the patent law and, as we stated in discussing the Westinghouse case, we can perceive no reason why the principle of equity invoked in apportionment cases arising under the patent law should not be applied to the issues of proof presented by the case at bar. The defendant has exerted itself to put the plaintiff in a position where it cannot prove by ordinary means that the defendant received profit from the sale of hose with the infringing feature or the amount of that profit. In our opinion the showing which the plaintiff made was sufficient to require the defendant to demonstrate error in the plaintiff’s method of computation of the defendant’s profits. The defendant has not carried that burden and, therefore, we come to the question of the sufficiency of the plaintiff’s proof by the means available to it. 12 *217 Despite the fact that the plaintiff is unable to prove with exactness (1) the number of infringing stockings sold by the defendant, (2) the defendant’s costs in manufacturing and selling the hose, or (3) the defendant’s profits due to infringement, we conclude that the evidence offered by the plaintiff proves that the defendant made a profit from the sale of the infringing hose and the approximate amount of that profit. We will now analyze that evidence. It is, of course, essential to determine whether or not infringing and non-infringing hose were sold under the same style numbers. The defendant’s price lists or circulars for Fall of 1931, 13 Spring-Summer 1932, 14 Fall-Winter 1932-33 15 are divided into “Regular Styles” and “TriLength Styles.” The two earlier price circulars contain other lists designated respectively “Special Styles” and “Novelty Styles.” The Tri-Length styles are described clearly and these lists permit no doubt that these were infringing hose. The style numbers listed as Tri-Length are 30, 10, 80, 70, 90, 40 L, and No. 2 “Whiff Chiffon.” The circular for Spring-Summer 1933 16 is not divided into Tri-Length and other styles but lists styles 30, 10, 90, 70, 40 L and 2 (Whiff Chiffon), describing them as Tri-Length. Similarly the price list for Fall-Winter, 1933-34 17 describes Styles 80, 40 L arid 2 (Whiff Chiffon) as Tri-Length. Style 70 is listed but is not described as Tri -Length. The Spring-Summer circular, 1934 18 does not refer to Tri-Length but lists “Magic Fit Styles” No. 70, 4, 19 80, 40 L and Style No. 2 (Whiff Chiffon). We shall deal presently with the use of Magic Fit Style numbers by the defendant. Until the Spring-Summer circular of 1934 the TriLength numbers were listed always as TriLength. Although each circular throughout the accounting period docs not list every Tri-Length style number, the same style number is not designated in one circular as Tri-Length and in another circular as another style, with the exception of Style 70 in the 1933-34 price list as hereinbefore stated. Published advertisements bear out this apparent general consistency. Until late 1933 and 1934 the advertisements in the newspapers frequently list the defendant’s hose by style numbers and describe and illustrate the Tri-Length feature. 20 The defendant called as its witnesses hosiery buyers from stores which had purchased its hose. These buyers testified that they had bought both Tri-Length and non Tri-Length hose under the same number but this testimony was confusing and unsatisfactory. Buyers testified that they did not purchase any Tri-Length hose at times when newspaper advertisements run by their stores featured such stockings. 31 Some of these witnesses testified that they paid little or no attention to style numbers as they ordered from samples and generally purchased in person from the defendant’s sales manager. Lord & Taylor’s buyer testified that the defendant did not use the same number for plain welt and TriLength stockings at the same time. Miss Walker, the buyer for H. Braunstein in Wilmington also testified that the same *218 numbers were not used for Tri-Length and plain welt hose at the same time. Alfred Hahn, who was the defendant’s production manager from a date prior to the beginning of the accounting period until he left the defendant’s employment in 1933, testified that the same number was not used for plain and Tri-Length welts. He explained that there was a correspondence of numbers between plain and Tri-Length hose as follows : ' This testimony is substantiated by the evidence of several buyers and by the price lists. The lower court found that the Master was not justified in assuming that all hose manufactured under the so-called Tri-Length style numbers were infringing hose. Judge Forman referred to the Spring-Summer price list, 1933 advertising Magic-Fit Styles under style numbers which had formerly been used for TriLength styles. It is true that until some ■time in 1933 Tri-Length hose were differentiated clearly by style number. But it would appear that when the defendant began to diminish the number of Tri-Length hose manufactured by it, it saw fit, none the less, to retain the Tri-Length numbers in manufacturing hose equipped with Magic-Fit tops. Conversely, it appears that the defendant did not cease manufacturing and selling Tri-Length hose during the latter part of the accounting period for defendant listed in the Tri-Length book lists a number of Tri-Length stockings manufactured and sold by it under Magic-Fit style numbers for the latter portion of the accounting period. Consequently, using this portion of the Tri-Length book as an admission against interest, 22 there is evidence that some of the stockings sold by the defendant were Tri-Length hose despite the price list designation to the contrary. Furthermore, the defendant in its motion of November 29, 1932 to suspend the injunction against the manufacturing and sale of infringing hose declared that it had on hand a stock of infringing stockings and orders for them and also had the same styles in production. The motion went on to state that “ * * * if the defendant is required * * * to discontinue the manufacture of such style * .* * factory operations would cease. * * *” 23 We conclude that there was evidence from which the Master and the court below should have concluded that at least some of the hose designated as Magic-Fit included in the Tri-Length top. In view of the defendant’s actions from the time at which its president reached the view that the issue of infringement might be decided against the defendant, the conclusion is inevitable that Kugelman caused the defendant to commingle Magic-Fit and TriLength with the intention of rendering an accounting impossible. In this connection it is important to bear in mind that the defendant offered no means other than the Tri-Length book for segregating infringing from non-infringing hose. It should be observed also that the defendant’s price list for Spring-Summer 1931 24 mentions no Tri-Length hose. The defendant did not list any style number capable of being identified with Tri-Length stockings. The conclusion is justified, therefore, that the defendant used its style numbers in this period to designate stockings without regard for the Tri-Length top. Although the defendant’s style numbers in 1931 and 1932 did differentiate properly between Tri-Length and other hose, in view of the admissions of the Tri-Length book, to be construed against the defendant, it must be concluded that in 1933 and 1934 hose without the triple top and Tri-Length stockings were manufactured and sold by the defendant under the same style numbers. Despite the fact that the evidential value of the Tri-Length book is slight, the fact that the defendant’s president saw fit to include Tri-Length stockings in it for the period under discussion, furnishes proof entitled to great weight that Tri-Length stockings were sold during this period. The conclusion that they were sold under the guise of “Magic-Fit” is inescapable. A comparison of the invoices sent to seventeen of the defendant’s customers by the defendant 25 offers further proof that the *219 number of infringing stockings sold by the defendant to its customers is shockingly understated by the Tri-Length book. The number of hose of Style 40 L, an infringing style number, sold to these customers from March, 1932, to the end of the accounting period exceeds by nearly 130 dozen 26 the number sold to all customers as listed in the Tri-Length book. The defendant’s invoices for February, 1934, alone show the sale of 1838 dozen pairs of infringing hose whereas the Tri-Length book lists 35 dozen pairs. 27 As has been stated previously, Kugelman testified that all pertinent records were destroyed after the adjudication of infringement and that this destruction included shipping vouchers and invoices. Kugelman was wrong in one respect. Apparently through inadvertence, the defendant’s invoices for the month of February, 1934, and those covering the periods from September 29, 1934, to October 31, 1934, and from November 20, 1934, to December 31, 1934, were not destroyed. These invoices, the defendant’s general ledger, its cash book and its accounts receivable cards, were the only records made available to the plaintiff. The Master and the learned District Judge had before them the invoices relating to the seventeen accounts hereinbefore referred to. These showed the actual number of hose with infringing style numbers purchased by these customers for a limited period. By taking the proportion of infringing to non-infringing hose purchased on these accounts and by estimating the proportion of the sales covered by these accounts to the defendant’s total sales, the plaintiff’s accountants calculated that 202,-106 dozen pairs of infringing hose were sold during the accounting period. 28 The defendant asserts that the seventeen accounts taken were not representative. This assertion comes with very bad grace indeed from this defendant. The accounts taken were not selected by the plaintiff but were the only accounts received from those of the defendant’s customers which responded to the plaintiff’s request for invoices. Furthermore, the defendant introduced no evidence indicating that the purchases made by the accounts were unusual and did not reflect the general trend of business. The invoices received cover both large and small accounts and the customers were well dispersed geographically throughout the United States. We accept the plaintiff’s estimate of the number of pairs. The number 29 estimated is not unreasonable when compared with the defendant’s capacity and the amount of stockings produced in the outside mills engaged in manufacturing hose for the defendant. Hahn testified that the defendant produced 100-110 dozen pairs of hose per week throughout the accounting period by working double shifts. By the Autumn of 1931 all of the 44 to 54 machines in the defendant’s own mill were equipped to make TriLength hose. Full production averaged 3500-4000 pairs per week. Kugelman testified that with the exception of a cessation of operations caused by a strike in 1931 and a few slack periods production of hose was maintained steadily on double shifts throughout the accounting period. Approximately 85% to 90% of the Belber Mill’s production was for Artcraft. There is evidence that 90% of this production consisted of Tri-Length hose. The Marnet Mill manufactured hose for the defendant from November, 1931 to May, 1932 and from September, 1934 to the close of the accounting period. Many of the machines at Belber and Marnet were equipped to make Tri-Length hose. It seems improbable that these machines would have been equipped especially to make Tri-Length hose if, as indicated by the Tri-Length book, only a srhall fraction of the defendant’s total production consisted of infringing hose. Since profit, subject to certain adjustments, ordinarily is to be computed by *220 taking the difference between costs and sales prices, the plaintiff must prove as accurately as it can the cost of manufacturing the infringing hose but in view of the defendant’s deliberate actions to confuse the accounting the burden of proving that the plaintiff’s calculations are in error must be borne by the defendant. It must be remembered that the defendant submitted no figures and took the position that none was available. The defendant’s own accounting expert, Gotwald, admitted that it would be impossible to ascertain costs from the meager records which the defendant retained. Moreover, Kugelman admitted that he never checked his actual costs against his estimated costs. In this situation, pursuant to the patent statute, 30 the plaintiff was entitled to rely on expert testimony as to the defendant’s manufacturing costs. The Master, in his report, stated that because of the demonstrated falsity of the Tri-Length book, none of the defendant’s other books or ledgers was entitled to credence. The lower court disagreed with this ruling. We conclude that the District Court was in error in determining this rule of law contrary to the Master’s conclusion. Moreover, as Gotwald admitted, the use of the defendant’s books and records which remained in existence could not supply the necessary cost data. The expert testimony offered in this connection by the plaintiff is not, as the defendant maintains, based solely upon a theory as to what the defendant’s costs should have been. The plaintiff made use of Arthur F. Morton, an accountant experienced in the manufacture of hosiery in general and in the Philadelphia area in particular. 31 Morton was well qualified to testify in respect to these matters. Six pairs of defendant’s infringing hose were furnished to him and he in turn submitted them for analysis by the United States Testing Company. The findings of this company as to content and construction of the hose are not disputed. Both the plaintiff’s and the defendant’s expert witnesses, as well as Kugelman himself, agreed that it was impossible to ascertain exactly the grade of silk used in the house. The testing company could not furnish such information. Morton, from experience, knew the general range of quality of silk which would be used in the manufacture of the different portions of the stockings. In every instance he allowed for the highest grade of silk within the reasonable range and computed this at the open market price. Kugelman, in attempting to refute this testimony, read into the record bills for silk which contained handwritten notations as to the grade. He admitted that these were his notes and based upon recollection. During cross-examination he changed his testimony several times. We find nothing in the record to refute Morton’s generous allowance for costs of silk. In estimating costs of manufacturing Morton estimated labor costs at the union rate of pay. It is admitted that defendant was not a union plant and paid less than union wages. 32 As to fixed charges, Morton allowed a reasonable amount for the value of the plant to the defendant based, alternatively, on rental or ownership. He also made an allowance for taxes and carrying charges. The defendant rented the plant from Kugelman under an arrangement whereby it not only paid a high rental but also paid taxes. We think,' in view of Kugelman’s domination of the company, these items do not constitute bona fide elements of expense and had the defendant submitted a statement of costs including such items, the estimate would have been subjected to a drastic reduction. 33 Morton also allowed $25,000 a year for Kugelman’s salary which the defendant declared was $57,000. However, Sedransk, an accountant called by the defendant, testified that Kugelman did not draw the full amount of his salary from the defendant. We conclude that the salary allowance which Morton made for Kugelman is approximately correct. The defendant made objection to the amount of cost allowance for various technical operations such as wet-back winding, *221 silk boil-off and removal of snarls and slugs. Morton’s testimony clearly demonstrated, however, that he was cognizant of the fact that defendant used these methods and made due allowance for them. As to sales cost, Morton allowed 12% of the market price, such being the upper limit of the range of selling costs generally estimated for branded name hose. The defendant did not prove that its costs exceeded 12%. Morton also testified that his estimates were based upon his own experience and that of his clients, a number of which were hosiery manufacturers in Philadelphia. The defendant makes the objection that these firms or companies were not sufficiently like the defendant to offer a fair standard of comparison. But Morton’s calculations were not based on an average of the expenses and profits of these companies or on a comparison, of defendant’s operation with the operations of any one of them. He gave proper consideration to the exact nature of defendant’s business and products. His conclusions find adequate support in the record. We find them to meet the test required of evidence under the circumstances of the case at bar. All of the defendant’s accounts-receivable customers ledger cards were made available to Ernst & Ernst, a firm of certified public accountants employed by the plaintiff. From these cards the following summary of sales was made: 34 From Morton’s calculation of cost of production by style numbers and from the summary of total sales, these accountants computed the profit made on the total infringing sales as follows. 35 The accounts of the seventeen customers of the defendant from whom complete invoices were obtained by the plaintiff were used as a standard. The total sales of infringing styles were obtained as well as total sales of non-infringing styles in both dozens and dollars. The average selling price per dozen was determined. The percentage of each style to the total was worked out as to both price and quantity. By applying the percentage thus obtained to total sales, the estimated total sales of every given style of infringing hose was arrived at. The cost to the defendant, exclusive of selling expenses as determined by Morton’s report, was deducted from the amount of the total sales of infringing hose. There was then deducted from the net amount thus arrived at a 12% sales expense and a discount of 3.15% of total sales. 36 The entire net profit to the defendant on the infringing stockings from October 31, 1931, to October 31, 1934, was estimated at $589,331.21. The defendant attacks this calculation of its profit by attempting to show that its entire business operations were not so profitable. To this end the defendant introduced profit and loss statements prepared for the defendant by Sedransk. These statements list the annual net profits or loss as follows : 37 It should be pointed out that Betts, a certilled public accountant in the defendant’s employ, certified a profit and loss statement for the year ended April 30, 1932, which he had prepared from the defendant’s books. 38 He ascertained that the defendant had a profit of $192,470.03 for that year. Although this period is not coeval with the Sedransk accounting periods it reflects an astonishingly different picture. The Betts finding of a substantial profit is *222 confirmed by the statements which the defendant submitted to The Philadelphia National Bank. These were prepared by Sedransk or by members of his firm. As of April 30, 1932, according to a Sedransk statement, the defendant had current assets of $400,698.82, current liabilities of $160,872.81 and surplus of $734,548.65. 39 Another Sedransk statement showed that as of June 30, 1933, the defendant had current assets of $355,339.34, current liabilities of $76,265.02 and a surplus of $744,973.02. 40 A statement as of June 30, 1934, was prepared for the defendant by Balch, Funk & Co., certified public accountants, and by Sedransk, as was a statement as of June 30, 1935. The 1934 statement showed current assets in the amount of $461,889.13, current liabilities $79,816.30 and a surplus of $799,036.70. 41 The 1935 statement showed current assets totalled $542,730.82, current liabilities $100,820.10 and a surplus of $844,-014.95. 42 These statements present a picture of an increasingly prosperous business. This is a very cogent reason for not placing too great a reliance upon the defendant’s books or its protestations that it received no profits from the sale of infringing hose. The statements were presented to a local creditor bank throughout a period of years. We may assume that the bank scrutinized them with some care and are entitled to infer that if the statements had been grossly incorrect, at least as to earnings or net assets, the bank would not have extended or continued credit. It must be remembered also that the defendant’s factory worked twenty hours a day throughout most of the accounting period and had infringing stockings manufactured for it at two outside mills. These facts furnish evidence of successful and profitable operations entitled to much greater weight than the testimony of defendant’s books and witnesses. In this connection it is irrelevant for the defendant to point out that certain of its competitors were losing money during the accounting period. As the Supreme Court declared in Keystone Manufacturing Co. v. Adams, 151 U.S. 139, 148, 14 S.Ct. 295, 299, 38 L.Ed. 103, “ * * * nothing is more common than for one manufacturing concern to malee profits where another, with equal advantages, operates at a loss.” The Master’s conclusions as to the total net profit made by the defendant by the sale of the infringing stockings represents a correct application of the law to the evidence in the case. The District Court was in error in rejecting them. The defendant’s entire net profit from the sale of infringing hose as thus calculated must be apportioned in order to ascertain the profit which accrued to the defendant by reason of the hose’s infringing feature. The Master’s findings in respect to the amount of profit to be allocated to the infringing feature were incorrect. The learned District Judge made no findings in respect to them since he disposed of the case by applying incorrectly the rule of law as to the sufficiency of the proof or profits. It is not our function, nor do we possess the power to find facts. In view of the extraordinary circumstances of the case, however, and in as much as the case must be remanded to the District Court for further appropriate action, we -think it would not be amiss to indicate to the court below what we deem the facts of the case indicate in respect to the apportionment of the profit to the infringing feature. From the testimony of Hahn, the defendant’s former production manager, 43 from that of the defendant’s own employees, 44 from the testimony of a number of buyers, 45 and from the defendant’s own price lists, 46 the Master drew the conclusion that the defendant produced hose identical in most respects, save that some of them possessed the infringing welt and others were without it. The style numbers and prices of these two groups of stockings may be compiled as to dollars and style numbers as follows: *223 The prices stated in the compilation are obtained from the defendant’s price lists and are corroborated by the invoices for February, 1934. 47 Style 222 appears only once in any price list and is described as having “dainty open-work lace clock.” Although it is referred to in the testimony as being similar to style No. 2, which is a TriLength stocking, it does not seem to offer a fair standard of comparison with the infringing hose and therefore should not be included in an analysis of price differentials. Accordingly, is not included in the compilation set out above. The conclusion is hardly justified from the descriptions in the price lists that the defendant’s style 40 L (a Tri-Length stocking of extra length) should be compared with style No. 44 (a non-infringing stocking of regular length). Style No. 66 (a non-infringing stocking) and style No. 10 (a Tri-Length stocking) show little similarity and therefore have not been included in the compilation. Hahn testified that style No. 10 was made on both 42 and 45 gauge machines, that style No. 66 was made on a 42 gauge machine, and that style No. 44 was made on a 45 gauge machine. Since all of the hose under discussion, with the exception of styles 2 and 222, were made on either 42 or 45 gauge machines, the gauge cannot be used as the sole factor in determining comparability. On the contrary, according to the testimony of Hahn style No. 10 (a Tri-Length stocking) and style No. 44 (a non-infringing stocking) were comparable. Hahn’s testimony in respect to the similarity between style No. 10 and style No. 44 is substantiated by comparison with the price differentials between the other pairs of comparable infringing and non-infringing hose; viz. $1. There is a price differential of $3 between style No. 10 and style No. 66 which seems to militate against an inference that such hose are comparable in all respects except the infringing feature. For the first season included in the accounting period, fall, 1931, there is a differential of $1 per dozen on two comparable pairs of styles (30 infringing and 22 non-infringing, 80 infringing and 88 non-infringing) but there is no differential between the other two comparable pairs of styles (10 infringing and 44 non-infringing, 90 infringing and 99 non-infringing). For the second season included in the accounting period, summer, 1932, there are five *224 comparable sets of styles listed. There appears to be a $1 differential in 4 pairs of styles (30 infringing and 22 non-infringing, 10 infringing and 44 non-infringing, 70 infringing and 77 non-infringing, 80 infringing and 88 non-infringing) and none in the fifth pair of styles (90 infringing and 99 non-infringing). In the winter season, 1932-3, there are three comparable pairs of styles (30 infringing and 22 non-infringing, 10 infringing and 44 non-infringing, 90 infringing and -99 non-infringing) and no difference in price is indicated between infringing and non-infringing hose. The other seasons do not contain a sufficient number of comparable styles to form a fair basis of comparison. In attempting to determine the amount of profit to be apportioned by reason of the infringing feature, it will be necessary, as we have indicated, to deduct for the purposes of the accounting the 50 additional cost of manufacturing the infringing hose. How should this 50 additional cost be deducted and from what? The difference in price between pairs of infringing stockings and comparable non-infringing stockings was $1. In the first season cover'ed by the accounting period, as we have indicated, there was a differential of $1 as to two comparable pairs of styles of infringing and non-infringing hose but no difference in respect to the two other comparable pairs of styles of infringing and non-infringing hose. The difference, therefore, between infringing and non-infringing stockings would average 500 (50% of $1). Employing this percentage method, it would be appropriate also to take 50% of the additional cost of manufacturing the infringing hose (50% of 50) or 2%0. Subtracting the 2%0 from the 500 there would be a 47%0 average differential for the first season. Employing the same formula for the second season, there was a differential of $1 in four pairs of comparable styles of infringing and non-infringing hose and none in the fifth. The difference, therefore, between infringing and non-infringing hose would average 800 (80% of $1). Employing the percentage method it would be appropriate to take 80% of 50 or 40. Subtracting the 40 from the 800 there would be a differential of 760 for the second season included in the accounting. There is no difference demonstrated between the three comparable pairs of infringing and non-infringing styles for the third season. It may not be assumed that there was no difference in amounts between the sales prices of Tri-Length and other hosiery in the two seasons (summer, 1932 and winter, 1933-4). It would seem fair to assign to these respective seasons the average differential heretofore ascertained for the three seasons for which figures are available. Such a course appears appropriate since there is little doubt that most of the infringing hose were sold in the fall of 1931 and throughout the year 1932. Reference to the Tri-Length book, fabricated by the defendant to aid it in escaping liability, demonstrates that this method of estimating the defendant’s profits will not work an injustice to it. See Duplate Corporation v. Triplex Co., 298 U.S. 448, 56 S.Ct. 792, 80 L.Ed. 1274, sanctioning the use of average prices in an accounting for profits. In Computing Scale v. Toledo Scale Co., 7 Cir., 279 F. 648, certiorari denied 257 U.S. 657, 42 S.Ct. 184, 66 L.Ed. 420, the court took the average difference in selling prices between infringing and non-infringing scales as the measure of profits. The Master found that the difference of $1 per dozen was applicable throughout. This was clearly erroneous but it would be appropriate to strike a mean average among the three seasons. 48 Averaging the three seasons for which differential figures are available, a price differential of 41.160 per dozen pairs of hose is arrived at. This is the amount of the profit seemingly allocable to the infringing feature on the 202,106 dozen pairs of hose. The Master recommended the allowance of the fees and expenses for Morton in the sum of $2,665.50 and for Ernst & Ernst in the sum of $10,314.07, finding *225 such charges to he reasonable. 49 The lower court disallowed these allowances because o f its conclusion that only nominal damages should he allowed to the plaintiff. The use of the expert witness and the accounting firm was necessitated by the defendant’s deliberate destruction of evidence. The defendant, therefore, must bear these costs of its own wrongdoing. See Flat Slab Patents Co. v. Turner, 8 Cir., 285 F. 257 and Levin Bros. v. Davis Mfg. Co., 8 Cir., 72 F.2d 163. The judgment of the court below is reversed and the case is remanded for further proceedings. Any judgment rendered for the plaintiff should bear interest at the rate of 6% per annum from the date of the filing of the Master’s revised report. Cf. Duplate Corp. v. Triplex Co., supra, 298 U.S. 448 at page 459, 56 S.Ct. 792, 80 L.Ed. 1274 and the authorities cited in Walker on Patents, Deller’s Ed. § 864. The judgment when entered should bear interest at the rate of 6% per annum. Footnotes [1] Twenty-six hearings were had before the Master. Forty-three witnesses testified and their testimony covers more than five thousand pages of the transcript. One *212 hundred fifty-nine exhibits were introduced in evidence. The appellant has endeavored to compress those portions “of the record material to the questions presented” pursuant to our Rule 26(2) (e) in an appendix of five hundred fifty-five pages. In addition to this the appellee has printed an appendix of twenty-five pages. Counsel’s attempts at compression in the instant case were unsuccessful and it was necessary for the court to make use of the transcript filed in the Clerk’s office. The result has been to augment greatly our difficulties in disposing of the case at bar. [2] The District Court made no separate findings of fact as distinguished from those embodied in its opinion. [3] Miss Kleinschrodt corroborated this testimony. Transcript Vol. VII, p. 2610. [4] Cf. Kugelman’s testimony. Transcript, Vol. VII, pp. 2080-2089, in which he states that all bookkeeping was under his immediate supervision. [5] Even Miss Kleinschrodt testified upon comparing the appearance of the various pages of the book, “I would say they were written at the same time.” [6] Quoted in Sauntry v. United States, 8 Cir., 117 F. 132 at page 135. See also Wigmore on Evidence (Third Edition) Vol. IX § 2486. [7] In the cited case the Supreme Court enunciated the rule relied on by the court below, as follows: “1 The patentee,’ * * * ‘must in every case give evidence tending to separate or apportion the defendant’s profits and the patentee’s damages between the patent- ed feature and the unpatented features, and such evidence must be reliable and tangible, and not conjectural or speculative; or he must show, by equally reliable and satisfactory evidence, that the profits and damages are to be calculated on the whole machine, for the reason that the entire value of the whole machine, as a marketable article, is properly and legally attributable to the patented feature.’ ” See 111 U.S. 120 at p. 121, 4 S.Ct. 291, 28 L.Ed. 371. See also Computing Scale Co. v. Toledo Scale Co., 7 Cir., 279 F. 648, 673, certiorari denied 257 U.S. 657, 42 S.Ct. 184, 66 L.Ed. 420. [8] This principle has been adopted in the American Law Institute Restatement oí Torts § 912 and the American Law Institute Restatement of Contracts § 331 as follows: Torts. “Comment'd: Although * * * the burden is on the injured person to prove with a fair degree of certainty that the business or transaction was or would have been profitable, it is not fatal to the recovery of substantial damages that he is unable to prove with definiteness the amount of the profits he would have made or the amount of harm which the defendant has caused. It is only essential that he present such evidence as might reasonably be expected to be available under the ■circumstances.” Contracts. “Comment a: The requirement of reasonable certainty does not mean that the plaintiff can recover nothing unless he establishees the total amount of his harm; nor does it mean that he cannot get damages unless he proves the exact amount of his harm. The requirement merely excludes those elements of harm that cannot be evaluated with a reasonable degree of certainty. * * * Furthermore, there are cases in which the experience of mankind is convincing that a substantial pecuniary loss has occurred, while at the same time it is of such a character that the amount in money is incapable of proof. In these cases the defendant usually has reason to foresee this difficulty of proof and should not be allowed to profit by it. In such cases, it is reasonable to require a lesser degree of certainty as to the amount of loss, leaving a greater degree of discretion to the jury, subject to the usual supervisory power of the court.” [9] See Story Parchment Co. v. Paterson Co., 282 U.S. 555, 51 S.Ct. 248, 75 L.Ed. 544, an anti-trust suit in which the Supreme Court stated that the wrongdoer cannot complain that damages are only approximate. The court refused to pervert justice by denying all relief because of the difficulty of proof. See also Dowagiac Mfg. Co. v. Minnesota Moline Plow Co., 235 U.S. 641, 35 S.Ct. 221, 59 L.Ed. 398, in which the court held that a reasonable approximation of damages was all that was required. [10] See Macomber, Damages in Patent Cases (1910) 10 Col.L.Rev. 639 criticizing Garretson v. Clark and foreshadowing the Wagner opinion. The rule of the Wagner case shifting the burden of proof to defendant upon plaintiff’s showing that apportionment is impossible was adopted in Seeger Refrigerator Co. v. American Car & Foundry Co., 3 Cir., 219 F. 565; La Crosse Plow v. Van Brunt, 7 Cir., 220 F. 626; Yesbera v. Hardesty Mfg. Co., 6 Cir., 166 F. 120; Mishawaka Rubber v. S. S. Kresge, 316 U.S. 203, 62 S.Ct. 1022, 86 L.Ed. 1381; Stentor Electric Mfg. Co. v. Klaxon Co., 3 Cir., 115 F.2d 268. [11] See Armstrong v. Belding Bros. & Co., supra, allowing recovery on a like theory even where the defendant was not held to be a wilful infringer. [12] Compare the powers of the court to protect its decree where the plaintiff is the wrongdoer. In Asbestos Shingle, etc. v. H. W. Johns-Manville, C.C.N.Y., 189 F. 611, even though the plaintiff was not acting in bad faith, an injunction was granted. Cf. Keystone Driller Co. v. General Excavator Co., 290 U.S. 240, 54 S.Ct. 146, 78 L.Ed. 293. See 1934, 44 Yale L.J. 351 for a discussion of this problem. [13] Apparently this price list is for Fall, 1931, although no date appears o,n the printed price list. Plaintiffs exhibit No. 2. [14] Plaintiff's exhibit No. 3. [15] Plaintiff's exhibit No. 4. [16] Plaintiffs exhibit No. 5. [17] Plaintiff’s exhibit No. 00. [18] Plaintiff’s exhibit No. 10. [19] Defendant in its answer lists Style Nos. 3 and 4 as infringing styles during 1933 and 1934. Since plaintiff is unable to apportion the infringing and non-infringing styles manufactured under those numbers and since defendant has failed to do so (except by means of the “TriRength Book”), plaintiff is justified in basing its calculations upon the assumption that all hose of Styles 3 and 4 were infringing. [20] For example see Plaintiff’s exhibits Nos. 88, 94, 98, 99 and 100 and Defendant’s exhibits Nos. 32K; 330 (although the stocking illustrated does not have a Tri-Length welt) dated March 5, 1933; 33J dated Juno 11, 1933 (mentions Styles No. 40, 30, and 2 as having Tri-Rength welts); 33N dated April 21, 1933 advertises 5 thread, 3 thread and 2 thread Whiff Chiffon as T-R; 33co elated June 23, 1932 (Style 90 featured as Tri-Rength). [31] For example see the testimony of Miss Koch, buyer for Dewoes, Transcript Vol. VI, p. 1609 and plaintiff’s exhibits Nos. 75, 76 and testimony of Mr. Guertin, buyer for Lord & Taylor, Transcript Vol. VI p. 1476 and exhibit 49-A. Testimony of buyers to the effect that there was no demand for Tri-Length hose and that only a few pairs were sold, is suspected also in the light of the same buyers’ admissions that Gotham adjustables are still being sold in large quantities. [22] See Producers & Refiners Corporation v. Lehmann, supra, 18 F.2d 492 at page 603. [23] Plaintiff’s exhibit No. 92. [24] See Plaintiff’s exhibit No. 46. [25] Complete invoices were obtained by the plaintiff from these customers. See Plaintiff’s exhibits Nos. 17, 18, 19, 21, 22, 24, 25, 52, 53, 54, 55. [26] See Plaintiff’s exhibit No. 112. [27] See Plaintiffs exhibit No. 8. [28] Returns and credits were not included in the invoices of all 17 accounts, but suitable allowance was made based upon accounts which did submit such memoranda. [29] Much testimony was devoted to the number of pairs of infringing hose manufactured for defendant in outside mills. There is evidence that both the Belber and Marnet Mills manufactured Tri-Length hose for defendant. 'Schowald, a knitter and part owner of Belber, testified that approximately 90% of tlie production for Artcraft consisted of Tri-Length hose. This testimony was attempted to be refuted by Wilkins, a former employee of Belber, who, however, was reimbursed by Artcraft for his time spent in investigating other employees upon whose recollections he based his testimony. Since we accept the Ernst & Ernst calculation (subsequently referred to herein) it is immaterial how many pairs of infringing hose were manufactured in outside mills. [30] Act of Feb. 18, 1922, e. 58, § 8, 42 Stat. 392, 35 U.S.C.A. § 70. [31] Morton’s findings are embodied in plaintiff’s exhibits 61 and 62. They are. phrased as a response to a hypothetical .question. Defendant objects that Morton’s answer is incompetent- because he did not examine the Artcraft plant and books. This objection goes only to the weight of the evidence and not to the competency of the answer. Stuyvesant Insurance Co. v. Jacksonville Oil Mill, 6 Cir., 22 F.2d 515, certiorari denied 276 U.S. 634, 48 S.Ct. 420, 72 L.Ed. 743. [32] See testimony of Hahn. [33] See Providence Rubber Co. v. Goodyear, 9 Wall. 788, 19 L.Ed. 566. [34] Plaintiff's exhibit No. 56. Cf. defendant’s exhibit No. 13, the Sedransk balance sheet which shows total net sales for 1932 at $1,367,516.84. [35] Plaintiff's exhibit No. 70. [36] The discount is estimated from figures obtained from the accounts of the seventeen customers of the defendants which submitted them. [37] Defendant’s exhibits Nos. 12, 13, 14, 15. [38] Plaintiff’s exhibit No. 15. [39] Plaintiff’s exhibit No. 81. [40] Plaintiff’s exhibit No. 82. [41] piaintiff’s exhibit No. 83. [42] Plaintiff’s exhibit No. 84. [43] Vol. V, p. 614. [44] Vol. IV, p. 402, Vol. VI, p. 1667. [45] Vol. VI, pp. 1539, 1758. [46] Plaintiff’s exhibits Nos. 2, 3, 4, 5, 60, and defendant’s exhibits Nos. 10, 11. [47] Plaintiff’s exhibit No. 17. See Tr. Vol. VI, p. 1711. [48] Those cases in which the courts have awarded all profits to the plaintiff when a fairly accurate apportionment was impossible, involved basic patents, combination patents, or patents, the use of which was essential either to the manufacturing or to the sale of the product. Armstrong v. Belding Bros. & Co., supra; Yesbera v. Hardesty Mfg. Co., supra; La Crosse Flow v. Van Brunt, 7 Cir., 220 F. 626; Levin Bros. v. Davis Mfg. Co., 8 Cir., 72 F.2d 163. The circumstances of the case at bar will not justify the conclusion that all profits should be awarded to the plaintiff. [49] Bee Plaintiff’s exhibit No. 71, the itemized bill of Ernst & Ernst. 147 F.2d — 15