In patent nomenclature what the infringer makes is “profits,” what the owner of the patent loses by such infringement is “damages.” In the case at bar no question as to the defendants’ profits arises, the complainant, apparently, being content to accept the defendants’ statement that no data can be produced upon which to base an estimate of profits. The appeal is here, therefore, to review an allowance of damages only.
The question whether a license fee was established is one of fact, *307and as the master has found that on November 1, 1901, a yearly license fee of $300 for each machine was established, and as the Circuit Court has overruled the defendants’ exceptions, we are not disposed to disturb the finding. It certainly cannot be said that it is unsupported by proof. Warren v. Keep, 155 U. S. 205, 267, 15 Sup. Ct. 83, 39 L. Ed. 141,
The defendants’ infringement began May 1, 1897, and continued until January 27, 1902. The license fee was established November 1, 1901. It is contended for the defendants that in no event can the proof have a retroactive effect and apply to infringements which occurred before November 1, 1901, when there was no license fee. In Rude v. Westcott, 130 U. S. 152, 9 Sup. Ct. 463, 32 L. Ed. 888, the court, at page 165 of 130 U. S., at page 468 of 9 Sup. Ct. (32 L. Ed. 888), says:
“In order that a royalty may be accepted as a measure of damages against an infringer, who is a stranger to the license establishing it, it must be paid or secured before the infringement complained of.”
The complainant argues that this quotation should be read as if the yords “the infringement complained of” refer not to the infringement in the suit before the court but to an imagined infringement of the party who, by paying the royalty, became a licensee. In other words, as we understand the complainant’s contention, the last paragraph of the quotation should read as follows, “it: must be paid or secured before the infringement of the party paying it.” It might, perhaps, he pertinent to inquire how a license fee can be paid before the “in friugement complained of” or, in oilier words, how an act can be called “an infringement” which occurs after the fee is paid and the payer has become a licensee. Ordinarily where a license fee is paid there is no infringement and certainly none after the license is granted.
It is conceded that the Supreme Court decided the question, as thus interpreted by the complainant, in the negative — namely that the fee paid by the licensee after infringement by him cannot be accepted as evidence of market value. As each one of the licensees referred to in the proof had been a past infringer and as there is not a particle of proof that a dollar was paid as license fee before these infringements, it would seem that the contení Ion of the complainant is even more fatal to a recovery than that of the defendants. There is no evidence of an established license fee until long after all of the infringements complained of. It is said that the opinion in Rude v. Westcott, if not interpreted as the complainant contends, is obiter.
We do not pause to consider whether it be obiter or not because the court has laid down a rule of law for the guidance of masters in chancery causes which has been followed ever since, which has never been recalled or modified and which, to our minds, is just and logical. Paraphrasing the rule of Rude v. Westcott to fit the circumstances of the case in hand it reads as follows:
“In order that the royalty found by the master herein may be accepted as a measure of damages against the defendants Brown and Miller, who were strangers to the licenses proved before the master, the royalty must be paid before the infringement of Brown and Miller complained of.”
*308We think there is nothing unfair or abnormal about this rule, it is simply applying the doctrine of the common law to patent causes. Even the thief indicted for grand larceny has a right to insist that the value of the property taken by him shall be fixed as of the date named in the indictment and to exclude evidence of its value five years aft-erwards.
An infringer is not an outlaw. If convinced that the patent is invalid or that he is not infringing he can go on and use the machine said to be covered by the claims, subject to the risk of an injunction and an accounting. If the patent be newly issued and generally infringed liis act should not be viewed in the same light as if the patent had been adjudicated and a license fee established. Ordinarily it is not until the validity of the patent has been established that parties are willing to pay for privileges under it. The validity of the patent in suit was upheld in the Dean Case (C. C.) Ill Fed. 380. This was in October, 1901. In November, 1901, for the first time a license fee was established.
The defendants’ infringement began in May, 1897. If this suit had been commenced shortly thereafter and if the accounting had been concluded prior to November, 1901, it must be conceded that no damages could be recovered for there would then have been an entire lack of proof of an established license fee. Because in 1901 a sufficient number of persons paid for the use of the patent to establish a license fee, it is contended that the amount so paid should relate back to acts done four years prior thereto for the purpose of proving that a right which, at the time, had no market value was in fact worth $360 per annum. If such testimony be admissible it may be used as a basis for damages occasioned by infringements committed not only 4, but 10, or even 15, years before.
It is argued that the double-cut mode of operation covered by the claims was an improvement, the value of which could not vary from year to year. We cannot assume this to be true. The addition of the patented pawl added to the capacity of the saw — it did more work. If, however, the conditions were such that little sawing was needed, the saw, without the pawl, would keep up with the other machinery in the yard. In a period of great activity in the building trade the proprietor of a stone yard might well afford to pay $360 per annum for the use of the improvement, whereas in a period of depression it might be of no value whatever. How could the master tell what was its value in 1897, for instance? There is no proof of uniformity of value, on the contrary it appears that in the settlement for past infringements, which preceded the granting of the uniform future licenses, the amounts paid differed in every instance.
It is impossible to formulate a rule which will do exact justice in every case, but where a license fee is relied on to establish the measure of damages we think it most unsafe to hold it applicable to infringements occurring before its establishment. The enforcement of a rule making such evidence incompetent to establish a measure of damages may work hardship in sporadic cases like the present, where it is said that profits cannot be proved; but much greater hardship and confusion will, in our judgment, result from a rule which permits damages *309to be computed on the basis of values fixed long afterwards. The rule which we sanction requires proof instead of conjecture and facts instead of theories.
The decree is affirmed without costs to either party.