HOUGH, Circuit Judge

(after stating the facts as above). [1] The Schram Case, supra, holds the claims in suit valid for a mechanical device, and denies that what Brooke patented is in truth a method. We agree with this, and plaintiffs must therefore prove that the alleged infringing machine, not only produces the same results as does the device of Brooke, but that its operation, when in use, is substantially the same. Davis v. Perry, 120 Fed. 945, 57 C. C. A. 231.

It is not suggested that defendant has copied the various mechanical components of the device disclosed in plaintiff’s specification, and the record is commendably free of expert evidence relating to any machine. Hardinge, etc., Co. v. Abbe, etc., Co., 195 Fed. 936, 940, 115 C. C. A. 624. The reason is that in this case plaintiffs are only interested in showing that Brooke conceived and disclosed in his patent (in the words of a brief)—

“tbe entirely original and. novel idea [of operating], upon molten glass flowing, continuously from tbe furnace and dropping in a stream — body or column— tbroiigb tbe air, by severing or cutting tbat flowing stream, at a point below tbe outlet and distributing tbe cut-off portions into molds.”

Brooke admittedly disclosed only one means of embodying or utilizing this conception, and defendant’s means are very different in mechanical arrangement; but it is urged that-Brooke was entitled (as he states in his specification)—

“to broadly cover all means for cutting or separating a stream of flowing molten material into unformed molten masses and discharging tbe cut-off portions.”

Exactly what defendant does is a matter as to which much evidence has been given; but whatever difficulty exists in answering that question does not depend upon difficulty in discovering what the defendant would like to do or is trying to do, but from the fact that what is actually happening with molten glass at the edge of a spout, at a temperature of about 2,000 Fahrenheit is not easy to see with the human eye, and (judging from exhibits) quite impossible to perpetuate by photography.

What the defendant tries to do — by the calculated agitation of a paddle in a mass of molten glass resting in a container whose crest *429or dam is uniformly higher than the glass level — is to propel over said crest and into a spout (which is itself slightly above said level) precalculated portions of molten glass, to the end that such portions or “gobs” of glass shall separately and individually be distributed to the molds awaiting them.

If defendant’s machine could do this with accuracy, and push over the crest, into the spout, and so on to the mold, separate definite weights or volumes of melted glass, with not one of them connected with or touching another, the machine would arrive at perfection in its class. But it cannot do this; when several agitations of the paddle start the melted glass to surging, each forward paddle motion shoves glass over the crest, and each backward movement of the paddle retracts the glass body. Thus in actual operation there may be 30 surges a minute produced by the paddle, and 30 “gobs” per minute will drop from the spout; but they are connected by a thinner band, string, or line of glass, because the viscosity of the substance will not permit it to .drop from any spout like a shot or marble. The visible result of the allegedly infringing apparatus is to produce from the end of the container spout, not a rope of molten glass, but (again to quote from argument) a “string of sausages” of the same material. Each sausage is a “gob,” and destined for one mold; and the size — i. e., length and weight — of each can be and is measurably predetermined by the amplitude of the paddle motion.

Brooke by knives severs at predetermined intervals his continuously flowing stream of molten glass, and so produces “gobs”; there is’ absolutely no preformation until the knife cuts. In defendant’s machine, knives cut the “sausage strings”; but “gobs” are preformed by the calculated movement of the paddle before the knife cuts. It is true that what defendant’s machine discharges by surges suffers solution of continuity only by and through the knife, and the same is true of Brooke’s stream; but by all the evidence the two streams look no more alike than a stream of sausages looks like a stream of sausage meat. The contest of fact in this case may be said to' rage only over the size or thickness of the string that connects the sausages. It seems plain that this contest is immaterial, if there is really no more than a string connection, between masses visibly formed, and formed for use, before severance; and we agree with the court below that such is the case.

The essence of Brooke’s concept or idea is that (as disclosed) he always has a “stream of flowing molten material” existing and moving by gravity alone; the contention here is that defendant’s “string of sausages” is such “flowing stream.” The fact that the stream said to infringe relies on gravity only after the material has been by the paddle shoved or lifted, paddleful by paddleful (so to speak), over the container’s crest, makes no difference, because (as is argued), once over, gravity takes hold, and the stream is that of Brooke “substantially as described.”

The argument for infringement may be thus summarized: (1) Brooke’s is a pioneer patent, and (2) therefore entitled to a “favorable construction commensurate with the advance which he made.” ’ (3) *430Defendant’s machine is made to work on a continuous supply of molten glass, which is cut with a knife, and (4) this is arriving at the same result by “analogous means” — which is enough to constitute infringement of a pioneer patent.

The argument avoids the question suggested by Brooke’s specification, viz.: Is a patentee ever entitled “to cover broadly all means” for doing a desirable thing, when he discloses but one means and suggests no substitute? Perhaps, in the light of this evidence, the question is whether, if Brooke was the first to use a severed stream of glass for filling molds, anybody else can so use such a stream, irrespective of means.

But we take the argument as made, and of course accept the description of a pioneer invention furnished by Brown, J., in Westinghouse v. Boyden, etc., Co., 170 U. S. 537, 18 Sup. Ct. 707, 42 L. Ed. 1136. It is as near a definition as we are likely to get. Auto Piano Co. v. Amphion Co., 186 Fed. 163, 108 C. C. A. 291.

Of such an invention we have said that, when it “inaugurates a new industry,” courts should be “zealous so to construe the claims as-to give validity to what is a meritorious invention.” Auto Vacuum, etc., Co. v. Sexton Co., 239 Fed. 900, 153 C. C. A. 26. It is also true-that, under restrictions not necessary to dwell upon, a patentee is entitled to the benefit of properties or functions inherent in his invention whether fully comprehended by him at the date of disclosure or not. Electric, etc., Co. v. Gould, etc., Co., 158 Fed. 610, 85 C. C. A. 432; Van Epps v. United, etc., Co., 143 Fed. 869, 75 C. C. A. 77.

Whether Mr. Brooke’s invention is of such a primary nature as to-merit the application of the word “pioneer” we shall assume, but not decide; for whatever the proper adjective applicable to this patent, the-legal rule of construction is the same. Outlook, etc., Co. v. General, etc., Co., 239 Fed. 878, 153 C. C. A. 5, et seq. It is always necessary, even after granting the widest range of equivalents, to find as a matter of fact that what the defendant has done is the invention of the plaintiff “substantially as described.” The range of decision, the limits imposed by law on the triers of the facts, are indicated by the word “substantially” ; an infringer may easily substantially imitate a big thing— i. e., a deeply rooted and wide-spreading inventive thought; whereas, without “Chinese copying,” imitation of a little thing is oftentimes difficult.

[2] But, whether the invention is large or small, primary or trivial, it remains true that, when a claim is clear and distinct, the patentee cannot go beyond the words thereof for the purpose of establishing infringement; the specification may be referred to for the purpose of limiting, but not of expanding, a claim, and the range of equivalents is measured by what is described and claimed. Westinghouse, etc., Co. v. New York, etc., Co., 119 Fed. 874, 56 C. C. A. 404; Universal, etc., Co. v. Sonn, 154 Fed. 665, 83 C. C. A. 422; Loraine, etc., Co. v. General, etc., Co., 202 Fed. 215, 120 C. C. A. 615; Fowler, etc., Co. v. McCrum, etc., Co., 215 Fed. 905, 132 C. C. A. 143.

Applying these rules, we are of opinion that defendant’s apparatus, does not present a flowing stream of molten material, whether we con*431sider that phrase standing alone, or ask whether it is the stream conceived by Brooke, or suggested by his patent. If the phrase be interpreted verbally, the word “stream” by universal definition conveys the idea of uniform and unbroken succession in movement; and “flowing,” which indicates movement, as if in a current or stream, only emphasizes the drought. The idea of continuity, uniformity, and indeed of steadiness, is inherent in the phrase.

'Much criticism of the lower court has been made, in that it declared the words to mean a steady discharge; yet by definition a “steady motion” means in respect of a fluid that the velocity at each point remains “constant in magnitude and direction.” We think the word was well applied.

But if the meaning of the phrase be referred, not directly from the claim to lexicographers, but to the disclosure as illuminated by the evidence, it is clear that what Mr. Brooke desired to get away from, and did most ingeniously avoid, was the formation of “gobs” before their severance from the general molten-mass. He did that by chopping up whatever fluid came by gravity out of an orifice in the container. Out of such orifice he could only get, and only wished to get, a run of material as from a spigot. Such a stream he had in mind, and he had no other; nor would any other suit what he wanted to do. The relation between defendant’s and plaintiff’s supplies of material is that both are continuous and both are of glass; and that is not enough.

The decree below is affirmed, with costs.