1. Since Doehler was in default as to delivery of all the furniture covered by the contract, the government had the privilege, conferred by the reletting provision, to refuse to accept delivery in toto and to let a new contract for the 6,321 pieces, holding Doehler for the “excess cost.” Doehler contends that the government could not exercise that privilege as to reletting with respect to part of the furniture, while calling on Doehler to deliver the balance, and that therefore Doehler cannot be held liable pursuant to the reletting provision. In effect, the argument is that, by its conduct, the government abandoned its rights under that provision. That contention is untenable. The government’s voluntary relinquishment of a portion of its privilege was not an abandonment of the rest of it. Although we do not so decide, we may assume, arguendo, that Doehler might properly have refused to deliver the 677 pieces; but even if it were true that, in that respect, the government sought to *133deviate from the terms of the contract conferring the privilege, that fact would not invalidate the exercise of that privilege as to the 5,644 pieces.

We may arrive at the same conclusion by another route: If we assume that the attempted exercise of the privilege, because partial, was invalid, then the notice of April 17-18 was an offer to Doehler by the government to exercise that privilege in part and to receive delivery from Doehler of the 677 pieces; assuming further that Doehler could properly have refused that offer and that, had it done so, the government, by reletting a contract for only 5,644 pieces, would have lost its rights under the reletting provision, the fact is that Doehler did not refuse but, on May 31, some six weeks after the notice, delivered the 677 pieces; accordingly, whatever Doehler’s legal position might otherwise have been, it is surely now in no position to object to the enforcement of the reletting provision as to the 5,644 pieces.

2. Doehler contends that the inclusion in the Equipment contract of the liquidated damage clause wiped out Doehler’s liability for “excess cost” under the reletting provision, regardless of whether or not, in fact, that inclusion caused all or part of the increased price under the Equipment contract. For, argues Doehler, the addition of that clause was a material change which “affected the price as a matter of law.” This alleged doctrine (reminiscent, in its rigid formality, of the rule in Shelley’s case) finds no support in the cases; those cited by Doehler related to relet contracts calling for performance of work of a kind substantially different from that which the first contractor had agreed to perform or for the delivery of goods of a substantially different character.2

The addition, however, of the liquidated damage clause may in fact have accounted, to some extent or entirely, for the increased price named in the Equipment contract. Doehler argues that Equipment must have known that the government could never prove any precise amount of damages caused by delay; and that, therefore, Equipment, recognizing that the added clause increased its risk, should it delay in making delivery, must be “conclusively presumed” to have demanded that its contract included a promise of extra compensation for taking that risk. But that argument impliedly rests on the axiom that businessmen invariably recognize the risks they take and always nicely calculate them in dollars and cents. Doubtless at one time many economists, believing it to be self-evidently true that every man is .solely a fanatically rational “economic man,” accepted that axiom as virtually a law of nature; but, as the result of accurate observation of human behavior, that axiom is now generally treated as an assumption to be used most cautiously with full awareness that it is a “neglective fiction.”3 Courts use such (and other) fictions; but *134this particular fiction should surely not be the basis of a “conclusive presumption” or otherwise serve as a substitute for evidence'.

Moreover, we would indulge in the sheerest guessing if we were to say that Equipment must necessarily have regarded the added clause as appreciably augmenting its risk; for the contract, even without that clause, would have subjected Equipment, if it delayed, to the consequences of the relet provision with a possibility of its liability for increased costs, a possibility which, in wartime conditions, might easily have become a strong probability. Besides, as Doehler (somewhat inconsistently) says in its brief, Equipment, when it made its contract, “may have been so sanguine of ability to perform in time as to make any delay damage seem unimportant.”

Whether, then, and to what extent, the added clause affected the price named in the Equipment contract are questions of *135fact which cannot be answered on the basis of the data before us or before the trial judge when he gave summary judgment for the government. There must be a trial on those issues.

With respect to those issues, we think that the burden of proof must be on the government. Had the second contract, except as to price, been the same in terms as the first, the price named in the second contract would be assumed to be reasonable, absent any contrary evidence.4 But such an assumption is unwarranted in the light of the additional clause. We cannot say that there is a strong probability that that clause did not raise the price; nor is it true that the facts are more accessible to Doehler than to the government; nor are there any policy grounds for not letting the burden rest where it ordinarily would rest, i. e., on the government, which, under the counterclaim, is asserting a claim against Doehler.

If, at the trial it should be proved that the added clause explains the entire difference between the prices in the two contracts, then the government’s counterclaim must fail. Should it appear, however, that the added clause increased the price only in part, then the counterclaim will not wholly fail but will merely be reduced by the amount of such partial increase. For the situation here is not like that which exists when the government, in a purported reletting, contracts for work or goods substantially different from that covered by the original contract; in such a case, the government cannot recover under the re-letting provision for the reason that, the two contracts being incommensurable, the price named in the second contract “cannot be used for the measure of recovery for breach of the original contract.”5 The attempted comparison is then like one between pigs and apples. But where the goods to be delivered under the two contracts are precisely the same and the only difference is an added item of obligation which enhances the cost in an ascertainable amount, the comparison, once the cost of that item is deducted from the price in the second contract, is unimpeachable. Consider, for instance, a contract for Idaho potatoes to be delivered at St. Louis followed by a relet contract, at a greater price, which requires delivery at Cleveland, where part of the increase in price can be shown to be due to additional transportation charges.

We take this occasion to suggest that trial judges should exercise great care in granting motions for summary judgment. A litigant has a right to a trial where there is the slightest doubt as to the facts, and a denial of that right is reviewable; but refusal to grant a summary judgment is not reviewable. Such a judgment, wisely used, is a praiseworthy time-saving device. But, although prompt despatch of judicial business is a virtue, it is neither the sole nor the primary purpose for which courts have been established. Denial of a trial on disputed facts is worse than delay. Cf. Arenas v. United States, 322 U.S. 419, 429, 433, 64 S.Ct. 1090, 88 L.Ed. 1363. The district courts would do well to note that time has often been lost by reversals of summary judgments improperly entered.6 Sartor v. Arkansas Natural Gas Corporation, 321 U.S. 620, 624, 64 S.Ct. 724, 88 L.Ed. 967, *136tells us that, especially where, as here, damages are in question, the courts should go slow about making haste by a summary judgment; there the Supreme Court said that “at least a summary disposition of issues of damage should be on evidence which a jury would not be at liberty to disbelieve and which would require a directed verdict for the moving party.”7 Of course this court will respect, as a precedent, the recent ruling on that subject in Madeirense Do Brasil, S/A v. Stulman-Emrick Lumber Co., 2 Cir., 147 F.2d 399, 405; but the judges sitting in the instant case think that that ruling, because of the peculiar atypical facts of the Madeirense case, should not be generously applied.

3. Both the contract with Doehler and that with Equipment provided for a discount of 1% of the contract price for payment within ten days. The government paid Equipment for the 5,644 pieces within the ten days and deducted the 1% in the amount of $1,050. In computing the excess cost, the government reduced the purchase price of the 5,644 pieces under the Doehler contract, by deducting 1% of the purchase price, i. e., $834.42, and the trial judge accepted that method of computation. Doehler asserts that, in this respect, he erred, contending that the $834.42 was improperly deducted and that the counterclaim, if valid, should be reduced by that amount; for,.says Doehler, the discount was to be given for prompt payment and no payment has been made to it for those pieces. But that contention neglects the fact that the Equipment contract was, in part, a substitute for the Doehler contract, and that prompt payment to Equipment with its resultant discount, reduced the “excess cost” by $1,050, thus benefitting Doehler; not to reduce the Doehler contract price equivalent in effect would be to give Doehler that benefit twice.

4. Doehler, in its pleadings, set up payment as an alternative defense to the counterclaim. In support of this defense, it filed an affidavit stating that, under a distinct contract of May 24, 1940, relating to the sale of articles not covered by the December 1938 contract, the government owed it $11,531.67 which had not been paid, and that, if the counterclaim is valid, this sum, withheld by the government, constitutes payment of the counterclaim. The trial judge did not consider this issue which also presents a question of fact to be dealt with at the trial.

5. The government, on its appeal, complains that the trial judge held that the government’s counterclaim must be reduced by $9,119.75, i. e., the amount collected as liquidated damages from Equipment. We think the trial court erred, for the following reasons: Suppose that the contract with Equipment had contained no liquidated damage clause. Then the “excess cost” for which Doehler would have been liable to the government would have included not only (1) the increase in the contract price but also (2) the damages caused by Equipment’s delay in delivery; for the words “excess cost” are broader than “increase in contract price.”8 If, in those circumstances, the government by suit against Equipment had collected a sum for damages caused by its delay, it could not also have recovered that amount from Doehler. But that amount, representing such delay damages, could not properly have been applied in reduction of the other item of Doehler’s liability, i. e., its liability for the increase in the contract price. The situation here must yield a similar result: The amount collected as liquidated damages from Equipment precludes recovery of that amount from Doehler on account of Equipment’s delay.9 But that amount may not be deducted from that part of the claim *137against Doehler constituting the increase in the contract price. However, there should be such a deduction to the extent that the amount collected as liquidated damages does not represent actual damages caused by Equipment’s delay, since that amount, if any, in fact reduced the government’s cost.

An alternative approach to this problem leads to the same conclusion: Let us assume that the “excess cost” for which Doehler is liable does not include any actual loss from Equipment’s delay. Then, ex hypothesi, that loss is not part of the government’s “cost” (which must be taken into account in computing the “excess cost”) but is an additional item of loss for which the government had a right to be reimbursed by Equipment. However, to the extent that the liquidated damages collected from Equipment exceed that item of actual loss, the excess must be regarded as reducing the contract price paid to Equipment.

We are left then with this problem: Must Doehler prove that, in whole or part, the liquidated damages paid by Equipment were greater than the actual delay damages caused by Equipment or must the government prove that they were not? We think that, for the following reason, the task is Doehler’s: We may suppose, as an extreme case (but one which does not exhaust the category), an express understanding between the government and Equipment by which Equipment paid the liquidated damages as a concealed means of reducing its contract price. Such a supposititious case supplies us with the rationale: The reason for crediting Doehler with all or part of the liquidated damages must be that otherwise the government would be unjustly enriched, at Doehler’s expense, by receiving more than the “excess cost.” But, absent proof, such unjust enrichment should not be assumed to be a fact. We know that the government may be seriously damaged in incalculable ways through delay in receipt of goods in wartime; we have recently held that a liquidated damages clause in a government wartime contract is not to be easily defeated for the very reason that only with great difficulty, if at all, could the government show the quantum of its loss. United States v. Walkof, 2 Cir., 144 F.2d 75, 77. Such would have been our ruling if the government had sued Equipment for the liquidated damages. It would be most unfair, then, to say, that, in litigation with Doehler, the government must be deemed to have suffered an actual loss less than the liquidated damages unless it can prove the contrary. And here, where we have no precedent to guide us, we must decide on the basis of policy, governed by what seems to us to be fair.10

Reversed and remanded.

(dissenting in part).

I agree in all that my brothers ,say, except that I think that the defendant should be compelled in computing its damages, to deduct from what it paid to the Equipment Company so much of the amount which it set off as liquidated damages, as it cannot prove as actual damages for delay. I agree that the plaintiff is liable for all damages caused to the defendant by the delay of the *138Equipment Company; indeed, for all actual damages for delay in delivery of the goods after the time fixed for delivery in the plaintiff’s contract until delivery by the Equipment Company; all such are part of its “proximate” damages. I cannot, however, see the relevance upon that issue of the contract between the defendant and the Equipment Company, which compromised in advance, as between them, the defendant’s damages for delay. There can be no real doubt that the liquidated damages did not in fact correspond to anything that the defendant could have recovered from the plaintiff on that score. I do not forget that the defendant’s actual losses may have been in fact as great as the liquidated damages —indeed, that possibility alone justified the clause — but the relevant question here is, not what was the defendant’s actual loss, but what it could have recovered from the plaintiff in dollars and cents; an issue as to which the contract between the defendant and the Equipment Company is res inter alios acta. To allow the defendant to disregard what it set off is to allow it to charge the plaintiff with liquidated damages; and if the defendant wished so to increase the plaintiff’s liability, the place, and the only place, was in the contract between the two.