to do either or both of those things. But where a patentee states in his specification that a particular part of his invention is to be constructed of a particular material, and states or implies that he does not contemplate any other material as being suitable for the purpose, it is not certain that any other material will be treated by a court as an equivalent of the one recommended in the patent.' Combination patents would generally be valueless in the absence of a right to equivalents, for few combinations now exist, or can hereafter be made, which do not contain at least one element, an efficient substitute for which could readily be suggested by any person skilled in the particular

art.2

§ 351. To define an equivalent is at present a weighty and difficult undertaking. It is weighty because many rights of property now depend, and always will depend, upon the definition. It is difficult because the deliverances of the Supreme Court upon the subject are inharmonious, and because none of those deliverances are accompanied by elementary reasoning on the merits of the question. The primary points are plain. The difficulties arise in the latter part of the problem; but the plain points require to be stated in order that the others may be methodically reached and intelligently discussed.

§ 352. One thing, to be the equivalent of another, must perform the same functions as that other.' If it performs the same function, the fact that it also performs another function is immaterial to any question of infringement.* Therefore, it sometimes happens that a junior device is an equivalent of a senior device in a sense that will constitute

1 Aiken v. Bemis, 3 Woodbury & Minot. 348, 1847; Harris v. Allen, 15 Fed. Rep. 106. 1883; Schillinger v. Cranford, 37 Off. Gaz. 1349, 1885. 2 Turrell . Spaeth, 3 Bann. & Ard. 458. 1878.

3 Machine Co. v. Murphy, 97 U. S. 125, 1877; Rowell v. Lindsay,

113 U. S. 103, 1884; Railway Mfg. Co. v. Railroad Co. 30 Fed. Rep. 238, 1887.

4 Foss v. Herbert, 2 Fisher, 31, 1856; Sarven v. Hall, 9 Blatch. 524, 1872; Wheeler v. Reaper Co. 10 Blatch. 195, 1872; Kendrick e. Emmons, 3 Bann. & Ard. 623, 1878.

it an infringement of a patent for the latter; at the same time that the senior device is not an equivalent of the junior device in a sense that will cause the former to negative invention or novelty in the latter.' One thing may accordingly be an equivalent of another, though it does more than that other, but it cannot be such an equivalent if it does less.' But the test of function is only the first of several tests of equivalency. The fact that one thing performs the same function as another, though necessary, is not sufficient to make it an equivalent thereof."

§ 353. Function must be performed in substantially the same way by an alleged equivalent, as by the thing of which it is alleged to be an equivalent, in order to constitute it such. This substantial sameness of way is not necessarily an identity of merit,' nor a theoretical scientific sameness. In a purely scientific sense, a screw always performs its function in a substantially different way from a lever, and in substantially the same way as a wedge. Screws and wedges are equally inclined planes, while a lever is an entirely different elementary power. But screws and levers can practically be substituted for each other in a larger number of machines, than screws and wedges can be similarly substituted. When a lever and a screw can be interchanged and still perform the same function with a result that is beneficially the same, they are said to perform the same function in substantially the same way. thing may be an equivalent of another in one environment,

But one

1877; Dryfoos v. Wiese, 124 U. S. 37, 1887; Forncrook v. Root, 127 U. S. 181, 1887; Sargent v. Burgess, 129 U. S. 19, 1889; Sickels v. Borden, 3 Blatch. 535, 1856; Peard v. Johnson, 23 Fed. Rep. 509, 1885; Tonduer v. Chambers, 37 Fed. Rep, 337, 1889.

5 Roosevelt v. Telegraph Co. 33 Fed. Rep. 510, 1887.

Turrell. Spaeth, 3 Bann., & Ard. 458, 1878.

and not such an equivalent in another situation.' Springs and weights are generally equivalents; but when the environment is such that a spring will operate successfully while a weight will not so operate, then they are not equivalents. While in most cases a mere handle is not an equivalent of a lever, it is such an equivalent where its connections are such that it performs the same function in substantially the same way. But infringement is averted. by using a hand, instead of a lever of a claimed combination, to work a rod.' However, in one case, the Supreme Court went to the length of deciding a confined column of water in a cylinder, worked by a pump and working a piston, to be an equivalent of a combination of a vibrating arm, toggle joint, and other mechanical devices, when used to transmit vibratory power.

6

In view of the foregoing adjudications, and of the nature of the question, it is at present impossible to say precisely wherein the required sameness of way of performing a function really consists; and in view of the vast variety of present and future mechanical devices and modes of operation, it will probably be always impossible to formulate a criterion which will be applicable to every case. The safest reliance is upon the analogies of precedents, and upon the instructed sagacity of those who are called upon to advise and to decide in new cases.

§ 354. Whether a device, in order to be an equivalent of another, must have been known at the time of the invention or of the patent, is the final inquiry relevant to the characteristics of equivalents. It is a point of large importThe affirmative view seems to have originated in the

ance.

mind of Justice CLIFFORD, and to have originated there in the Fall of 1865. In November of that year he instructed a jury that a thing to be an equivalent of an article in a patent, must have been known at the date of the invention, and of the patent.' The rule of the Supreme Court, at that time, seems to have been that a thing might be an equivalent of a thing in a patent, if known to be such at the time of the infringement. Indeed, Justice CLIFFORD himself, in 1861, had delivered a carefully prepared instruction to a jury, in which he said: "You will regard the substantial equivalent of a thing as being the same as the thing itself, so that, if two machines do the same work, in substantially the same way, and accomplish substantially the same result, they are the same; and so if parts of the two machines do the same work, in substantially the same way, and accomplish substantially the same result, those parts are the same, although they may differ in name, form, or shape. Engaged, as he then was, in defining an equivalent, he wholly omitted the element of age. Indeed, as late as October, 1865, he found a defendant's device to be an equivalent of a thing in a patent, because it performed the same function, in substantially the same way, and produced substantially the same result. Here is no allusion to any element of age in the characteristics of an equivalent. That element was introduced into the definition later in the Fall. But no statute relevant to the subject was enacted in the interval, nor indeed at any other time, either before or since Justice CLIFFORD changed his views. Nor did Justice CLIFFORD ever announce the reasons which carried him to his altered definition. He, however, repeated the later proposition in several subsequent cases.

993

In Seymour v. Osborne,' he said, that in order to be an equivalent of a thing in a patent, a device must have been

Union Sugar Refinery v. Matthiessen, 2 Fisher, 629, 1865.

* O'Reilly v. Morse, 15 Howard, 123, 1853.

Cahoon v. Ring, 1 Clif. 620,

1861.

1865.

Roberts v. Harnden, 2 Clif. 506,

5 Seymour . Osborne, 11 Wallace, 556, 1870.

known at the date of the patent. In support of that statement he cited two cases, neither of which appears to relate to the precise point.' But after formulating the doctrine he was content to ignore it, and he reasoned out the question of infringement in the case with sole reference to the character of the defendant's machines, and without any regard to the point of time when any parts of those machines first became known. Indeed he expressly said that questions of infringement are best determined by a comparison of the machines. Now, it is evident that a comparison of two machines has no tendency to show whether a particular device in one of them, was or was not known when the other was patented.

In Gould v. Rees' Justice CLIFFORD delivered the opinion of the Supreme Court, reversing a judgment in an action at law, because the judge had charged the jury that the omission of one of the elements of a combination, and the substitution of another mechanical device to perform the same function, will not avoid infringement. Now that instruction was clearly wrong, because it did not provide, as it ought, that the substitute must perform the same function in substantially the same way, in order to be an equivalent for the thing for which it is substituted.' The Supreme Court, therefore, decided that the instruction was erroneous, but Justice CLIFFORD, in delivering the opinion, said, in effect, that it ought also to have been limited by the proviso that the thing substituted should have been known at the date of the patent. The case did not call for any decision on that point, because the undoubted error in the charge was enough to require a reversal of the judgment. Nor is the dictum supported by any reasoning, or by any citation of authority other than the previous dictum of the same justice in the case of Seymour v. Osborne. It is true that

Prouty v. Ruggles, 16 Peters, 341, 1842; Johnson v. Root, 2 Clif. 123, 1862.

Gould v. Rees, 15 Wallace, 193,

1872.

3 Burr v. Duryee, 1 Wallace, 573, 1863; Sickles v. Borden, 3 Blatch. 535, 1856.