the knives upon the rapidly revolving cylinder. Now, it may be that the other machines called "disintegrators" accomplish the same result, but they accomplish it by the use of two cylinders, each armed with cutting, crushing, or disintegrating projections, which intermesh, and which effect the disintegration in the same general method as crushing is effected by smooth rollers. The speeds with which the two rollers are operated differ but little in the Moore, Rossi, and Watson machines, and, while they may have effected disintegration, they certainly did not do it in the same way and upon the same principle as that seen and employed in the Potts devices. Conceding that disintegration of the clay, sufficiently complete to introduce it at once into the pugging mill or the brick machine without having recourse to the soak pit, was accomplished before the Potts invention, we are nevertheless of opinion that the operation of the Potts device is so different from that of the prior devices, and is so much more efficient than they are shown to be, that it is still entitled to the reward of a limited monopoly. It is difficult at first to distinguish between the pulverizing and crushing operation and that of disintegrating. In a wide sense, "disintegration" necessarily takes place in the operation of crushing and pulverizing. The supreme court uses the term, however, in the sense of tearing apart, piece by piece, or shredding. In no clay machine but the Potts do we find this kind of disintegration. It is true that the Moore, Rossi, and Watson machines are described as shredding the clay, but, on cross-examination, it was made quite apparent that the machines as operated would not tear the clay piece by piece, for the differential speed of the two rollers was not great enough for this. Moreover, the product was afterwards subjected to the "soak pit," an indication that the clay was not in the desired condition, for it was to make the soak pit unnecessary that the Potts machine was invented. We do not think that the disintegration of apples or of cocoanut kernels is so analogous to the disintegration of wet clay that the ingenuity shown in the adoption of the device for disintegrating clay can be minimized by reference to these other arts. We are justified in this conclusion by the fact that the supreme court did not find that the Butterworth pat ent, for disintegrating apples, which was quite as near in operation and principle to the Potts device as either the Mills or the Archenbroon mill, was a reason for depriving Potts of the right to a patent. The Gregg patent, it is quite clear, was impracticable, and, though it professes to comminute the clay, it is shown by the evidence to have probably been a complete failure, and never to have gone into use at all. The Dodson patent was an instance of the intermeshing of pro jections on the surface of an abutment with corresponding projec tions on the surface of a cylinder. It is not shown to have produced disintegration of the clay, and is one of those wrecks and failures of inventive genius that are constantly found lining the path of the successful inventor, who takes the last step which wins.

We think Judge Sage erred in his conclusion that the supreme court's decision rested wholly on the pioneer or primary character of the Potts machine in accomplishing an entirely new result, and, even if that were the ground of the decision, it does not at all follow

that the supreme court would not have reached the same conclusion, because of the difference in the means or method employed, and greater efficiency thereby secured, in the Potts machine over any shown in the preceding art. If, as contended by counsel for defendants, the Newell machine will disintegrate clay better with its intermeshing surfaces on the face of two rollers than the Potts machine, the defendants have the right to use the Newell machine.

In reaching our conclusions of fact in this case, we have not been unmindful of the abuses which the strict rules enforced in the allowing of the rehearings were adopted to prevent. Where an elaborate opinion of the court of last resort upon the evidence is published, and the weaknesses of the losing side are clearly brought out, and the defeated party is thereafter given an opportunity to strengthen the defects of his case by evidence as to transactions long past and machinery long since cast into the scrap heap, there is great danger that the exigencies of the case may lead witnesses to round out evidence beyond that which exact truth would permit. Such evidence must be taken with great caution, and weighed in the light of this danger.

Objection is made by the defendants that the sixth claim of the complainant's patent is simply for a cylinder with longitudinal grooves and scraping bars, adjustably secured in a groove, and that the other elements shown in the patent cannot be read into it in order to make it a combination patent. We understand the supreme court in its decision to have treated this as a combination claim, or, at least, to have held that it was for the element in the clay-disintegrating machine to be used in combination with the opposing and other elements necessary to secure disintegration of the clay by the methods specified in the patent. The supreme court intimates in its opinion that the substitution for the vibrating plate of a smooth roller did not involve invention, and therefore the second patent was not valid as an improvement over the first. It supports, however, the validity of the first patent, and the sixth claim thereof, as for the element in a disintegrating machine of the cylinder, constructed according to the claim, in combination with a vibrating plate or the equivalent thereof, which shall feed the mass of clay to the armed cylinder, and hold it to be clipped off or disintegrated, piece by piece, by the rapid revolution of the cylinder. It is impossible, therefore, to determine the validity of the claim in the light of the prior art, without considering the other parts of the old machines. with which the alleged anticipatory cylinder co-operates in them. We have therefore felt justified, as the supreme court did, in looking at the combination in each of the old patents, to determine whether the revolving cylinder, with its peculiar functions, in the Potts patent, had been anticipated. We hold, therefore, that the first Potts patent is valid and is infringed, but that the second Potts patent is invalid, because it shows no patentable improvement over the first. Other questions have been made by appellant, but, in view of our conclusion, it is not necessary for us to consider them. The decree of the circuit court is affirmed in so far as it dismisses the bill on the second Potts patent, and is reversed in so far as it dismisses the bill

on the first Potts patent, with directions to enter a decree finding the validity of the first Potts patent, and ordering a reference for damages. The costs of appeal are taxed against the appellees.

MAGIC LIGHT CO. v. ECONOMY GAS-LAMP CO.

(Circuit Court of Appeals, Seventh Circuit. October 3, 1899.)

No. 584.

1 PATENTS-INFRINGEMENT-DESIGN FOR GAS-GENERATING Lamp.

The Williams design patent, No. 30,147, for a fixture for generating and burning gas from liquid hydrocarbons, held not infringed.

2. Same—Gas-Generating Lamps.

The essential feature of the device shown in the Williams patent, No. 606,435, for an improvement in gas-generating gas fixtures, is a generating tube with the bore tapering towards its discharge end, to permit the pressure from the gas generated to be exerted more freely backward in the tube, and to secure an even flow from the discharge end; and, in view of the limitations placed on the claims of the patent by the amendments made to meet objections of the patent office, it cannot be held to cover a device having a generating tube of uniform bore, and in which uniformity of flow is secured by packing the tube with asbestos wicking. 3. SAME-MECHANICAL EQUIVALENTS.

A patent covers only known equivalents, and where it is clear from the proceedings in the patent office that when a patent was granted it was not known that a different device was the mechanical equivalent of one described therein, and its equivalency was expressly denied by the patentee, the mere fact that such device is subsequently shown to accomplish the same result as that of the patent does not render its use for that purpose an infringement.

4. SAME-CONSTRUCTION OF CLAIMS-LIMITATION BY AMENDMENT.

The effect of the amendment of claims to meet the requirements of the patent office as a limitation of the scope of the patent does not rest upon the doctrine of equitable estoppel, but the estoppel in such case is in the nature of one by contract, and its scope in a particular case is a matter of interpretation and construction of the terms used, according to their fair meaning, and with reference to the intention of the parties as disclosed by the proceedings.

5. SAME METHOD OF CONSTRUCTION.

Where it is plain that an amendment of his claims by an applicant was for the purpose of making a particular construction a distinctive feature of his invention, and he explicitly denied the utility of a different construction to accomplish the result intended, his patent cannot be construed to cover such method of construction.

Appeal from the Circuit Court of the United States for the Northern Division of the Northern District of Illinois.

James H. Peirce, for appellant.

Chas. L. Dobson and Ephraim Banning, for appellee.

Before BROWN, Circuit Justice, and WOODS and JENKINS, Circuit Judges.

WOODS, Circuit Judge. This appeal is from an interlocutory decree of injunction against infringement of claims 1, 2, 3, 4, 5, 11, 12, and 14 of letters patent No. 606,435, issued on June 28, 1898, to John F. Williams for improvements in gas-generating gas fixtures,

and of a design covered by letters patent No. 30,147 granted to the same patentee on February 7, 1899. Of the accompanying drawings, Fig. 1 of the first patent sufficiently represents the design of the second patent. Fig. 2 is an enlarged representation of the corresponding parts of Fig. 1. The third figure represents the design and construction of the lamp made by the appellant. The design covered by the second patent is described in the specification as follows:

"The fixture consists of two curved arms, A and B, connected at their upper ends to a fitting or casting, C, from which extends a straight section, D. A reservoir, E, is mounted upon said section, D, and a hook, F, extends up from the reservoir. To the lower end of the arm, B, there is fixed a fitting or casting, G, and from the horizontal branch of the fitting extends an arm, H, which stops just short of a projection, I, extending out from a coupling or casting, J, secured to the lower end of arm, A. Extending down from the coupling, G, is an arm, K, its lower end being bent at right angles, forming a horizontal arm, L. A similar arm, M, extends down from coupling, J, and is formed with a horizontal arm, N, the ends of the arms, L and N, being connected to a casting or fitting, O, which forms the burner-support."

In respect to this it is enough, besides referring to the drawings, to say that the lamp of the appellant resembles no more closely the design of the patent than it resembles other lamps and gas fixtures which are shown to have been in earlier use. In other words, infringement has not been proved.

It is necessary to quote only the first and fourteenth claims of the first patent, since the others mentioned are like the first in respect to the point of controversy. They read as follows:

"(1) In a device for generating and burning gas from liquid hydrocarbons, the combination of a suitable supply-pipe; a gas-generating pipe connected thereto and provided with a gradually-decreasing internal bore towards its discharge end; a discharge-outlet for said generating-pipe; a burner; a mixingchamber intermediate said burner and outlet; and means for admitting air into said mixing-chamber." "(14) A fixture for generating and burning gas from liquid hydrocarbons, comprising in combination a tube or pipe, 9, having a reservoir connected thereto; supply-pipe, 4; a gas-generator, 12, connected to pipe, 4, and provided with a suitable valve; a mixing and conducting pipe, 1, into which the gas is discharged; an air-inlet for said mixing-pipe, and means for controlling it; a pipe, 3, connecting pipe, 9, and pipe, 1; a pipe, 22, communicating with the mixing-pipe; a chimney mounted on said pipe, 22, below the generator, 12; and a mantle of incandescing material mounted in said chimney."

Pipe 12 of the drawings is the "gas-generating pipe" of the claim. The tapering passage therein, according to the specification, is preferably formed from a tube having a straight passage by placing therein a rod, filed or dressed to form the passage, the rod being held in place by frictional contact. That form of construction is explicitly covered by the thirteenth claim of the patent. The operation of the apparatus, as affected by the narrowing passage, is thus described:

"The liquid hydrocarbon passing from the vessel, 10, through the tubes, 9 and 4, to the vaporizing-tube, 12, is there vaporized by the heat from the flame concentrated through the chimney arranged in close proximity under said tube. The tube, 9, being larger than the tube, 4, and the tube. 4, larger than the vaporizing-tube, and the passage in the vaporizing-tube decreasing in size from inlet to outlet, it is apparent that by reason of the constant decrease in size of the passage from the inlet of the liquid into the pipe, 9, to the exit of the vapor from the vaporizing-tube, any back pressure in the tubes generated during the process of vaporization meets the least possible frictional resistance, and the exit of the vapor will be steady, and cannot puff and blow and cause uneven pressure. In other words, by the arrangement and construction of the passages as shown, but leaving them otherwise clear of any packing or filling material, the particles or molecules of the liquid as they expand or are converted into gas may exert their pressure in a rearward direction, and not tend, as in the usual burners, to force the gas out through the tip under uneven impulses."

This feature of the invention is emphasized by the proceedings in the patent office. The claims first presented did not embody it, and were rejected upon reference to prior patents. To avoid the refer ences the feature was introduced, first without specifying in the first and third claims in which direction the passage should narrow, but finally, in response to the specific demand of the office, defining the direction by inserting the words "towards its discharge end." The advantages of the tapering bore were explained at length by the attorneys of the applicant in the letter of January 14, 1898, in which the following statement was made, distinguishing the tapering bore from other means for retarding the flow of the liquid: