L. HAND, Circuit Judge.

This case comes before us again, this time upon a final decree, entered upon the accounting which we directed before. 81 F.2d 49. The district court referred the account to a master, who heard the parties at great length and made a voluminous report which the district judge affirmed in most respects. The decree awards to the plaintiffs all the profits made by the defendants from exhibitions of the picture, "Letty Lynton", and the principal question is whether this was correct. The defendants insist that the profits should have been apportioned, and that the record contains evidence by which that can be done. Their reasoning is that the recovery of the author of a copyrighted work ought to be limited to those profits which result from its exploitation; and that since the value of the picture here depended only in very small measure upon those parts which the defendants have been found to have lifted, they should be accountable for only a correspondingly small part of the profits. Even if this be true, it is equally true that an infringer carries the burden of disentangling the contributions of the several factors which he has confused. The law requires him to resolve any doubts arising from his wrong; he is like any other constructive trustee. Callaghan v. Myers, 128 U.S. 617, 666, 9 S.Ct. 177, 32 L.Ed. 547. Unless, however, there is an absolute bar against his success, the only question is what evidence of separation courts will accept. Strictly and literally, it is true that the problem is insoluble. The profits from a picture consist of admission fees, which the playgoers pay because the picture attracts them with the hope of enjoyment. That enjoyment, which is one source of its further popularity, is made up of many factors: the actors, the work of the producer and director, the story, the scenery and costumes. The attraction and the hope which first draws them are principally aroused by advertisements, and the reputation of the stars and the producing company. These factors have no unit common to all, and are therefore incommensurable; in that, the situation is not different from the usual case of copyright infringement where the pirated material has been mixed with matter in the public demesne. The difficulties of separation have generally prevented infringers from attempting any apportionment; they have contented themselves with getting down the net profits as low as possible. That was the case in Callaghan v. Myers, supra; and in Belford, Clarke Co. v. Scribner, 144 U.S. 488, 508, 12 S.Ct. 734, 36 L.Ed. 514; and for this reason the general language there used is not to be taken as holding that the infringer must always be unsuccessful, no matter what evidence he may bring forward. They hold no more than that when he makes no effort to discharge the duty resting upon him, he will be cast for the whole profit. Dam v. Kirke La Shelle Co., 2 Cir., 175 F. 902, 41 L.R.A., N.S., 1002, 20 Ann.Cas. 1173, cannot, however, be so explained. The appeal came up from an interlocutory decree, which did not of itself involve the point, but the court, apparently sua sponte, undertook to declare that the plaintiff should recover all the profits. The defendant moved for a modification of this provision, so as to allow it to prove the value of its own contribution to the play. By affidavits it proposed to show what were the royalties usual in the industry, and other matters which it thought relevant. The court nevertheless adhered to its first decision, apparently not because the specific evidence offered was unacceptable, but because no evidence would serve. We understand it to have held that an infringer, at least if he is a deliberate plagiarist, must surrender his entire profits, regardless of the relative importance of the copyrighted and uncopyrighted material he has used. In Harold Lloyd Corp. v. Witwer, 9 Cir., 65 F.2d 1, the bill was dismissed for non-infringement, but Judge McCormick in a dissenting opinion declared that the profits might be and should be apportioned. He seems to have based his conclusion principally upon that clause in section 25 of the Copyright Act, 17 U.S.C.A. § 25, which gives the court power to award damages "in lieu of actual damages and profits". Hartford Printing Co. v. Hartford Directory Printing Co., C.C., 146 F. 332, is the only other decision on the point that we have found, and looks towards the possibility of apportionment. It must be owned, however, that in point of authority the plaintiffs have the advantage.

Essentially the same problem arises in patent accountings. It is true that the question most mooted there is in what circumstances the patentee succeeds in throwing upon the infringer the duty of separating that part of the profits which he has himself contributed; and upon that rock most patentees used to be wrecked. Westinghouse Electric Mfg. Co. v. Wagner Electric Mfg. Co., 225 U.S. 604, 32 S.Ct. 691, 56 L.Ed. 1222, 41 L.R.A., N.S., 653, was primarily intended to treat them more liberally, and, so far, it is not relevant to this appeal, for that burden is always on the copyright infringer. However, in those situations in which the burden does shift to the patent infringer, the same difficulty appears as in copyright suits, for in both it is nearly as unfair to cast the infringer for all the profits, as it would be to deny the patentee or author any recovery whatever, because he could not separate his contribution. This plight of the infringer was considered in Westinghouse Electric Mfg. Co. v. Wagner Electric Mfg. Co., supra, and in 225 U.S. on page 620, 32 S.Ct. on page 696, 56 L.Ed. 1222, 41 L.R.A., N.S., 653, the court plainly recognized that "by general evidence, expert testimony or otherwise" he might relieve himself; his privilege was confirmed in Dowagiac Mfg. Co. v. Minnesota Moline Plow Co., 235 U.S. 641, 647, 35 S.Ct. 221, 59 L.Ed. 398. In 1922, in rather delayed response to these decisions, Congress changed the statute by providing (section 70, Title 35, U.S.Code, 35 U.S.C.A. § 70) that "opinion or expert testimony" should be competent upon the issue, apparently without regard to where the burden of proof might for the moment lie. Since then such testimony has often been used in patent accountings, though it has by no means removed all the difficulties. It seems to us that we ought not to disregard the progress of the law in a field so close to that before us. While it remains true for the reasons we have already given, that, except in the most general way, the percentages of experts cannot be used to solve a problem in which there is no common measure; yet it would be a mistake to deny all weight to them. Men often make quantitative judgments and act upon them in matters which logically admit of them as little as this. If one says that he likes one kind of music twice as much as another, we do not charge him with talking nonsense. We should indeed do so, if he added that his liking for Wagner was ten per cent of his liking for Beethoven; but even then it would express, however pedantically, a different degree of preference from the first form, and it might well have different practical consequences. A court is justified in basing its decrees upon practices common in other human affairs; and we can no longer accept the doctrine of Dam v. Kirke La Shelle Company, supra, that by no hook or crook can an infringer be relieved of so manifestly unjust a result.

The evidence offered in the case at bar was of two kinds: the price of the picture rights in the play, as both plaintiffs and defendants had agreed to it, and the opinions of experts. As to the first, if we knew that the price represented the utmost that the defendants would have given for the rights, and if at that time they also knew how much the picture as a whole would cost, the proportion between that cost and the price of the rights might vaguely indicate the defendants' judgment of the play's proportional drawing power. These conditions were not, however, shown to have been fulfilled. The price resulted from a bargain, determined as well by the plaintiffs' necessities, as by the defendants' estimate of the play's probable contribution. Again, while the defendants knew whom they would make the stars, it does not appear that they had forecast the whole cost of the picture. Finally, even were all this known, the cost of the other factors was by no means a measure of their relative drawing powers. For these reasons we discard the agreed price. The expert testimony was of two kinds, that of producers and that of exhibitors. The questions put to each were substantially the same: what was the proportion of the gross receipts properly apportionable to the play? Their answers were in percentages that ran between five and twelve (one of them is perhaps to be understood as saying that the play contributed nothing at all). Very generally they professed to believe that the controlling factor in the success of a play was the popularity of the actors — the stars — and that in the case at bar the two "leads," Crawford and Montgomery, had stood very high in the public esteem in 1932. (Indeed, the infringing picture itself was sold to many exhibitors merely as "Production No. 208, Joan Crawford No. 2"). They agreed that in the case of a very well known book or play, the story might contribute more, but that the plaintiffs' play was not in that class. They thought that the other chief factors were the skill and reputation of the producer and the director, the scenery and costumes, the extent of the advertising and the reputation and standing of the producing company itself, which gave an assurance to exhibitors and to playgoers, on which both very largely relied. The plaintiffs called no witnesses to rebut this testimony, and if their failure to do so was because of the commanding position of the defendants in the industry, they did not prove it. We must therefore assume that the testimony represents the best opinion of the calling.

Before fixing a percentage upon the basis of this testimony we must consider two things peculiar to the case at bar. First, all the witnesses' estimates were based upon the contribution of the entire play; that is, as though it was completely the work of the plaintiffs. As we know, that was not the case; the plaintiffs worked over old material; the general skeleton was already in the public demesne. A wanton girl kills her lover to free herself for a better match; she is brought to trial for the murder and escapes. Nobody can say how far this basic plot is to be credited with whatever the play contributed to the drawing power of the picture. That consideration must therefore count towards reducing the percentage of profits recoverable. On the other hand the defendants may not count the effect of their standing and reputation in the industry; probably the most important factor of all, after the stars. They were not innocent offenders; they deliberately lifted the play; it would take far more than the denials in the old record to convince us to the contrary in the face of the step by step "tracking" which our comparison and analysis disclosed. The defendants appear to have understood our refusal on the first appeal to pass upon this issue as evidence of our acceptance of their disclaimers. We meant nothing of the kind; it was not necessary to decide the issue and we left it open against the chance of receiving new light. No new light has come, and we now hold that the borrowing was a deliberate plagiarism. It follows that they can be credited only with such factors as they bought and paid for; the actors, the scenery, the producers, the directors and the general overhead. Callaghan v. Myers, supra, 128 U.S. 617, at page 665, 9 S.Ct. 177, 32 L.Ed. 547; Flat Slab Patents Co. v. Turner, 8 Cir., 285 F. 257, 282, 283; Kansas City Hay Press Co. v. Devol, C.C., 127 F. 363, 369. Indeed a constructive trustee, who consciously misappropriates the property of another, is often refused allowance even of his actual expenses (Restatement of Restitution § 158(d)) and although this harsh rule, which would charge the defendants with the whole gross receipts, has been softened, a plagiarist may not charge for his labor in exploiting what he has taken. A fortiori he should not be allowed for the currency which his reputation may have given to the combined product.

We are aware that out of all this no real standard emerges, and that it would be absurd to treat the estimates of the experts as being more than expressions of very decided opinions that the play should count for very little. But we are resolved to avoid the one certainly unjust course of giving the plaintiffs everything, because the defendants cannot with certainty compute their own share. In cases where plaintiffs fail to prove their damages exactly, we often make the best estimate we can, even though it is really no more than a guess (Pieczonka v. Pullman Co., 2 Cir., 102 F.2d 432, 434), and under the guise of resolving all doubts against the defendants we will not deny the one fact that stands undoubted. Procedural duties are devised in aid of truth; and their unsparing use may defeat their whole purpose, as here it would. However, though we do not press the burden of proof so far, the defendants must be content to accept much of the embarrassment resulting from mingling the plaintiffs' property with their own. We will not accept the experts' testimony at its face value; we must make an award which by no possibility shall be too small. It is not our best guess that must prevail, but a figure which will favor the plaintiffs in every reasonable chance of error. With this in mind we fix their share of the net profits at one fifth.

The Accounting.

Both parties objected to a number of items in the accounts, as fixed by the master and confirmed by the judge. We shall first take up the defendants' objections, and then the plaintiffs', in the same order in which they appear in the briefs.

The Defendants' Objections

(1) The master refused to allow as a credit a certain part of the profits which had been paid to the Louis B. Mayer Company, a partnership made up of three persons — Mayer, Thalberg and Rubin — all important officers of one or another of the defendants. When they entered the defendants' employ, they were able, because of their strong position in the industry, to exact an agreement for a share of all the profits of the Metro-Goldwyn Pictures Company. This was evidenced by a contract which was later superseded by a contract with Loew's, Inc., whose profits were substituted for those of the Pictures Company. We do not see why payments made under these contracts should not be credited to the defendants. Even though we assume arguendo that the plaintiffs could recover them from the partners personally, in this suit they can reach only the defendants' profits. The payments were never profits of the defendants at all; the contracts effectively laid hold of them the moment they came into existence. Not to allow the credit would be in substance to introduce the partners as defendants into this suit, and yet to hold Loew's, Inc., liable for the recovery against them. We allow the credit.

(2) The next question is how to find the proper distribution cost of the infringing picture. Several ways might be theoretically proper; but we have to choose between only two: one, to divide the total by the number of pictures distributed by the defendants in 1932; the other, to divide it by the infringing picture's proportion of the total gross receipts. The master took the first; the defendants wish us to take the second; nobody suggests that the costs should be divided in accordance with the cost of production. The distribution cost is made up of a number of items, some of which it might be better to divide by one rule, and some by another. For example, the cost of the advertisements and other publicity might very well vary with the cost of production; the cost of carriage might perhaps best be divided by the number of theatres at which the picture was exhibited. But we cannot see why on any theory the cost of distribution should vary with gross receipts. As between the two we therefore accept the master's disposition of this item.

(3) The next point is whether to include the profits made from exhibiting the infringing picture outside the United States. At first blush it would indeed seem that these should be excluded. The plaintiffs made no proof of foreign law, and we cannot say that the exhibition of the positives abroad was a tort. However, exhibition is not the only act forbidden by the Copyright Act; Section 1(d), 17 U.S.C.A. § 1(d), gives to the author the exclusive right, not only to perform a dramatic work, but "to make * * * any transcription or record thereof * * * from which, in whole or in part, it may in any manner * * * be * * * reproduced." The Culver Company made the negatives in this country, or had them made here, and shipped them abroad, where the positives were produced and exhibited. The negatives were "records" from which the work could be "reproduced", and it was a tort to make them in this country. The plaintiffs acquired an equitable interest in them as soon as they were made, which attached to any profits from their exploitation, whether in the form of money remitted to the United States, or of increase in the value of shares of foreign companies held by the defendants. We need not decide whether the law of those countries where the negatives were exploited, recognized the plaintiffs' equitable interest; we can assume arguendo that it did not, for, as soon as any of the profits so realized took the form of property whose situs was in the United States, our law seized upon them and impressed them with a constructive trust, whatever their form. Compare Goulds Manufacturing Co. v. Cowing, 105 U.S. 253, 26 L.Ed. 987; Dowagiac Mfg. Co. v. Minnesota Moline Plow Co., supra, 235 U.S. 641, at page 650, 35 S.Ct. 221, 59 L.Ed. 398.

(4) The next item is the profits of the defendant Loew's theatre-subsidiaries, with which the defendants say that they should not be charged, except so far as they have received them in kind. If any subsidiary were insolvent, it would of course be true that its creditors would have a prior claim upon its profits; but there is no suggestion that any of them are. That being true, it makes no practical difference whether the profit comes to the defendants in the form of a dividend, or whether it merely enhances the value of their shares. (This we have already implied under the last point). The master in all cases prorated the subsidiary's profits to the proportion of its shares held by the defendants, and they can demand nothing more. It is true that the defendants were not formally the exhibitors of these positives, but that makes no difference. Indeed, the situation in this respect is precisely like that of the exported negatives which we have just considered. Moreover, as to these positives there are two other answers. The exhibitions were torts, even if they were not torts of the defendants, and the profits were benefits arising from them, which the defendants received as volunteers, for they paid nothing for them. Like other volunteers they are accountable for any benefits received from a constructive trustee. Restatement of Restitution § 160(g). Finally, the defendants were in fact the real tortfeasors, because, although the subsidiaries had a formal separate existence, the master found that Loew's, Inc. was the principal even in the act of exhibiting the pictures; their officers were in complete control of the subsidiaries, which were used merely as agents. Kingston Dry Dock Co. v. Lake Champlain Transportation Co., 2 Cir., 31 F.2d 265, 267; Pacific Can. Co. v. Hewes, 9 Cir., 95 F.2d 42, 45, 46; Berkey v. Third Avenue R. Co., 244 N.Y. 84, 95, 155 N.E. 58, 50 A.L.R. 599.

(5) Next, the defendants wish the "overhead expenses" allocated by the number of pictures, as the master allocated the distribution cost. The master, on the contrary, used the cost of production as the basis. Neither is theoretically accurate; to make a perfect allocation one would have to examine what part of the time of all the employees whose pay went into the "overhead", was given to each picture; and so of the other expenses. That was obviously impossible. It is on the whole more likely that a given picture required that proportion of the general services represented by its cost of production, than that each picture shared those services equally. The last is a very remote possibility indeed; the first is more likely to conform to the facts, if we could know them. The master's solution appears to us as nearly right as was practically possible.

(6) Next is an item disallowed by the master, which arose as follows. The defendants found by experience that of those "continuities" which they made from stories or from plays bought by them, or which were entirely composed by writers in their employ, a very substantial number turned out failures. To meet this, in 1932 for the first time, they set up on their books, as part of "studio overhead" an item of $500,000, representing what they estimated would be the loss on their existing stock of "continuities". Had this been an annual depreciation credit, based upon past experience, something might be said for treating it as an expense of the business, like the item of "rejected continuities" which will come up later. But it was not of that kind; it was an estimate of the proper deduction from the cost on the books of losses already suffered. We cannot understand by what reasoning this should be part of the "overhead" for the year 1932.

(7) The last of the defendants' objections was to the master's exclusion as a deduction of the interest upon a large loan made by Loew's, Inc., the "top company", to the Pictures Company, in the form of an open account going back twelve years. It is impossible to say how much of those things purchased by the proceeds of this loan continued to be used in 1932. The master included in the "overhead" an interest charge upon all the plant investment of the defendants used in making the picture. He computed this like the other "overhead" by taking that proportion of the total plant machinery and "properties", which the cost of production of the picture bore to the cost of production of all pictures. To this he added the interest for six months upon $300,000, taken as working capital. Certainly it would have been improper to allow both this and interest on the loan. If the Picture Company borrowed its capital and was allowed the interest as an expense, it could not charge anything for the value of those things used in production. But that aside, there was no warrant for allowing interest on the loan in any case. So far as its proceeds had been used up, the continued payment of interest upon the loan by which they had been procured, was not relevant to the production of the infringing picture. The master's method was in any event preferable. Moreover, the transaction seems to have been strictly intramural; we cannot understand why it should figure as an item in the accounts of the plaintiffs with the defendants collectively.

The Plaintiffs' Objections.

(1) The plaintiffs' first objection is to the master's allowance as a credit of the income taxes paid by the defendants. He allowed them because he thought that the defendants were not deliberate plagiarists; otherwise the case would have fallen within the ruling of the Supreme Court in L.P. Larson, Jr., Co. v. Wm. Wrigley, Jr., Co., 277 U.S. 97, 48 S.Ct. 449, 72 L.Ed. 800; Goodyear Tire Rubber Co. v. Overman Cushion Tire Co., 6 Cir., 95 F.2d 978, 985. See, also, Stromberg Motor Devices Co. v. Detroit Trust Co., 7 Cir., 44 F.2d 958; Stromberg Motor Devices Co. v. Zenith-Detroit Corp., 2 Cir., 73 F.2d 62. From what we have already said regarding the defendants' guilt, it appears that the master was wrong. It does indeed seem somewhat arbitrary to distinguish this from other expenses necessary to the business; yet on the other hand the distinction illustrates that in dealing with a conscious wrong-doer, courts do not feel obliged for consistency's sake to take one extreme or the other.

(2) The second objection is to the allowance by the master for what is known as "idle time". It was the frequent practice of the defendants to employ their more costly employees — directors, producers, stars and writers — at yearly salaries, rather than by the picture. This involved a certain waste of time since one picture cannot always be fitted closely upon another. The "idle time" of Crawford and Montgomery before making the infringing picture was allowed as a credit, and similar allowances were made in the "overhead". The plaintiffs must take the defendants' practices as they were; they are confined to the profits actually made, not to what the defendants might have made. If it was in the long run more economical to employ their assistants by the year, obviously the "idle time" must be paid for by the returns from pictures, and whether it was better to charge it to the next following picture, or to an earlier one, is not important. The allocation was fair.

(3) Next is the objection to deducting a payment made by the defendants to the Motion Pictures Producers and Distributors Association (the Will Hays Association) based upon the revenues derived from the infringing picture. The assumption is that this association will repay, or rebate, so much of this payment as represented the profits which must be restored under the decree in this suit. That is, however, speculative; the money was paid and created a credit; if the plaintiffs had wished to prove a corresponding charge, it rested with them to do so. They could have examined the officers of the association.

(4) Next is a challenge to any allowance for "overhead" at all, on the theory that the defendants did not show that it had been increased by the production of the infringing picture. The correct rule upon this point is stated in Levin Bros. v. Davis Mfg. Co., 8 Cir., 72 F.2d 163, and in substance it is this. "Overhead" which does not assist in the production of the infringement should not be credited to the infringer; that which does, should be; it is a question of fact in all cases. In the case at bar the infringing picture was one of over forty made by the defendants, using the same supervising staff and organization, which had to be maintained if the business was to go on at all. Without them no picture could have been produced; they were as much a condition upon the production of the infringing picture as the scenery, or the plaintiffs' play itself. Levin Bros. v. Davis Mfg. Co., supra, did not hold otherwise; nor did Haiss Mfg. Co. v. Link-Belt Co., 3 Cir., 63 F.2d 479.

(5) The next objection is to the inclusion in the "overhead" of an allowance for continuities scrapped, and for completed pictures never exhibited. The plaintiffs invoke the rule that an author, like a patentee, may select those infringements which prove profitable, and ignore the rest. Crosby Steam Gage Valve Co. v. Consolidated Safety Valve Co., 141 U.S. 441, 457, 12 S.Ct. 49, 35 L.Ed. 809. On the other hand there is in most industries a certain inevitable wastage, resulting from imperfect industrial technique and the like; and this, being a condition upon all production, is a part of the cost of production. Duplate Corp. v. Triplex Safety Glass Co., 298 U.S. 448, 56 S.Ct. 792, 80 L.Ed. 1274. The charge for wasted pictures and "continuities" was of this kind; owing to the imperfect forecast of what would prove a good "continuity", a number of false starts were inevitable; sometimes even a complete picture would also turn out to be valueless. The plaintiffs answer that they were not in partnership with the defendants, whose failures should not be charged to them. But the infringing picture owed its success in part to the fact that it was only one of a large number produced that year. Had defendants not had so large a capacity, the profits might never have been made at all; certainly they would not have been as large. Since therefore the plaintiffs profited by the fact that the defendants had developed this capacity, they must be content to take the breakage, so to say, which was its inevitable incident.

(6 7) The next objection we may pass, because it is covered by what we have said in section four. The next following relates to the way in which the interest item in the overhead was calculated; that is, to the "base" used. In computing it the realty was appraised, but the value of the personalty was taken at cost less depreciation. It was formally inconsistent not to appraise both, but the extravagant labor of doing so was an excuse, especially when one considers the small amount involved. It was better also to compute this item by assuming that the infringing picture used that proportion of the whole plant which its cost of production bore to the cost of production of all pictures made that year, than to attempt any allocation of buildings and other property according to their actual use for the picture. The second method would have been incredibly difficult in application, involving as it would a different proportional use of each bit of property concerned.

(8 9) The next objection is covered by what we said in section four, but the following one raises an interesting question. The plaintiffs say that they should not be charged with the losses resulting from the exhibition of the infringing picture in certain theatres, where the receipts did not equal the costs of maintenance during the period of the exhibition. We think that so far they are right. The exhibition of a picture at theatre X was a separate tort which the plaintiffs might elect to sue upon, ignoring any losses from the exhibition at theatre Y. As to exhibitions in the United States we therefore hold that the losses should not be credited to the defendants. Exhibitions in foreign countries stand on another basis. As we have said, the exhibition of the positives was not a tort, or at least it was not shown to have been a tort; therefore the plaintiffs had no choice but to sue for the next profits arising out of the manufacture of the negative. They could not therefore select the profitable exhibitions and discard those which resulted in a loss, but must take the bitter with the sweet. The ruling of the master is modified as to exhibitions in the United States.

(10) The last of the plaintiffs' objections is somewhat complicated, though it involves very little. It arose in this way. The Culver Company exported all the negatives manufactured by the defendants during 1932, the infringing picture among them. There were expenses of distribution for all these amounting to nearly $500,000, and the question is how much of this should be credited against profits from the infringing picture. We have already affirmed the master in his apportionment of distribution costs equally among the "feature pictures". In his apportionment of the United States costs, in order to get a commutation figure for "news reels" and "shorts" — of which a great number was sent out — he adopted the formula that one "feature" was equal to nine reels; and by applying this he found an equivalent of 60 "features" distributed in the United States, made up of true "features" together with commuted "news reels" and "shorts". Forty-three "features" were exported in 1932; and forty-three two reel "shorts" and forty-one one reel "shorts" along with them. Using the master's formula that was the equivalent of fifty-seven "features". He reduced this divisor to forty-five because not all of the "features" were sent to all the countries where the costs were incurred. On the other hand he made no allowance for the foreign films that went along with the American films, as the laws of many countries often require. The deduction, being a credit, the defendants had the burden of proving it; all they did prove was that the equivalent of fifty-seven films were exported. It seems to us that, for all we can tell, the reduction of the divisor to forty-five for those "features" which were not sent to all countries, may well have been balanced by the foreign films that went along with the American; and that there was really no warrant for any other divisor than fifty-seven, which we fix as the proper one. The item will be so computed.

Allowance to the Plaintiffs' Attorneys.

There remains nothing but the allowance to the plaintiffs' attorneys in all courts. This the judge fixed at $55,000, not including this appeal. He made up the figure as follows: $10,000 for preparing for trial, $5,000 for the trial, $10,000 for the appeal; $22,500 for the accounting; $7,500 for arguing the exceptions before himself. The great reduction in the recovery must result in a reduction of the allowances, though by no means in the same proportion. The amount of work was extremely large, and the computations were intricate and long; the plaintiffs had to go to California for part of the hearings, and the cause has been pending for more than seven years. We will make the following awards: for preparation for trial and for the trial $10,000; for the first appeal $5,000; for the hearings before the master $15,000; for the hearing on the exceptions $3,000; making $33,000 in all. We shall allow nothing for the argument on this appeal, since the defendants were successful. Section 40 of the Copyright Act, 17 U.S.C.A. § 40, awards "full costs" to the "prevailing party", and that will give the costs of the appeal to the plaintiffs; but it leaves the allowance to attorneys in the discretion of the court.

Decree reversed and cause remanded for further proceedings not inconsistent with the foregoing.