The bill of complaint in this case charged the defendants, who are here the appellants, wilh infringement of United States letters patent No. 465,911, granted on December 29, 1891, to Thomas William Ballard for improvements in gas stoves, the title to which patent had become vested in the complainant, the Dixon-Woods Company, here the appellee. At the beginning of the specification of the patent the general nature of the invention is stated thus:

“This invention relates to that class of gas stoves known as ‘cosy stoves,’ employed for heating rooms or apartments, and the main object of the inven*350tion is the cpnstruction of a cosy stove which will give a more even diffusion of flame for "heating the asbestus fiber and fire clay over the whole heating surface of the stove, and will also economize the consumption of gas.”

Then, after referring to tbe accompanying sheet of drawings showing a cosy store constructed in accordance with the invention (Fig. 1 showing the front elevation, and Fig. 2 a transverse section), the specification proceeds as follows:

“A is the framing of the stove, of any suitable design, and preferably provided with an outlet,, At, to which a pipe may be attached for, drawing off the products of combustion. B is a hollow slab supported in said framing, and made of lire clay, or other suitable refractory material. The front of said slab may present a lumpy or undulating surface, as shown in Fig. 2, in imitation of pieces of coal or other fuel, or such front surface may be smooth. Through the said front portion perforations, C, are made, communicating- with the interior, Bi of the slab. Certain of said perforations are filled or partly filled with asbestus fiber, D, so disposed that lengths of the same will be distributed over the front surface of the slab. The remaining holes form passages or burners for the emission of the heating- gas. In Fig. 1 every alternate perforation, C, is shown as filled with asbestus fiber. But I do not confine myself to such arrangement. Thus, as a modification, one row of perforations may be left open for burners, and the next row filled with fiber, and so on. The bottom of the slab, B, is provided with an atmospheric or Bunsen burner, E, of any suitable design, connected with a gas pipe, F, so that when the gas is turned on it will pass into and fill the interior space, B*, with a mixture of gas and air, and the said mixture will issue through the perforations left open in the front of the slab, and when ignited at the said perforations the flames will play uniformly over the whole front surface, among the asbes-tus fiber, and so diffuse a uniform heat thereto. The heat thus diffused raises the temperature of the interior space, Bi, and so rarefies and makes more intense the heating properties of the gas as it escapes from the burners or perforations, thereby economizing the consumption.”

Tbe patent bas a single claim, of which the following is a copy:

“In a gas stove, the combination of the frame, A, with the hollow slab, B, provided with perforations, C, the burner, E, for supplying a mixture of gas and air to the interior of said slab, the asbestus fiber, D, and the outlet flue, Ai, all substantially as and for the purpose set forth.”

*351We cannot carefully read (lie specif!cation without seeing that this patent is for a mere structural improvement upon a pre-existing type of gas stoves. The impression thus produced is deepened when we look into Ballard’s provisional specification, filed on March 22, 1887, in his application for a British patent for this same invention, which he obtained in 1887. In that English specification Ballard referred t.o the old style of cosy stoves in the words following:

“The so-willed ‘cosy stoves,’ as at present constructed, are provided at their base or up tlin sides with a gas pipe from which the flame, when ignited, burns upward amongst ilie ashes his, thereby burning more intensely at the bottom or base of the stove iban at any oilier part, whereas, according to my construction of-a si ove, T diffuse the heat uniformly over the whole surface.”

It is inferable from the inventor’s above-quoted statements, as, indeed. it otherwise appears, that the supposed novelty of Ballard’s construction consisted in the perforations in the face of the upright hollow slab, and (he distribution of fibrous tufts of ashes tus among the perforations. Now, this record is full of proofs touching the prior art, and when they are examined it becomes perfectly clear that, if there is any patentable novelty at all in Ballard’s improvement, it lies within the narrowest compass. Let us here consider the British patent, No. 14,333, issued in the year 1852 to Boggett <& Pettit, relating to (he warming of buildings by means of gas. Those patentees, in their specification, state:

“The principal features in our manner of accomplishing this object consist of various stoves or apparatuses for using asbestos in combination with gas, either carbure!ted or otherwise, the employment in connection therewith of tubes or hollow chambers, with or wiihout the addition of heat-absorbing substances, to produce, warm currents of air, and in constructing the said stoves or apparatuses, and 1lie flues belonging to them, in such a manner as to effect the perfect removal of the heated vapors after having uliltaed them to the utmost extent before their final escape into the chimney,” etc.

*352These features are illustrated by numerous attached drawings, accompanied by explanatory observations in the body of the specification. For example, Fig. 1 shows a vertical cast-iron burner for a gas fireplace cast with an interior hollow space from top to bottom. The top is tightly closed, and the bottom is provided with a boss connected with a gas supply pipe. In the front of the hollow burner are perforations serving both as outlets for the gas, and as holders of small tubes having asbestus secured to their outer ends. Fig. 2 is a front view of several of these burners, arranged perpendicularly side by side, with a space of about a quarter of an inch between them, and inclosed in a frame to form the front of a gas fire. “The frame,” the specification states, “is carefully closed, except in front, and the air, having no other access to the flues, rushes with great velocity between the bars and in the direction of the arrows, Fig. lc, as soon as the gas is lighted, carrying with it all the products of combustion, and rendering the asbestus vividly incandescent.” But Figs. 17, 18b, and 18c are still more in point; the specification explaining them thus:

“Fig. 17 represents an elevation of an apparatus for burning air and gas. * * ⅜ The air and gas mixes at A, passing by a tube in the back, as shown in section Fig. 18, to supply the three bars, B, B, B, and issues through perforations in them, as shown in Fig. 18a, in which are also shown the tufts of asbestus. 1<⅛. 18b is a vertical section, and 18c. an elevation of a similar method to the preceding of burning air and gas. The air enters at A, as before described, and, mixing with the gas at c, passes into a square iron box, perforated all over with fine holes, through which it burns; some tufts of asbestus, such as shown in Fig. 13, being suspended in front of it.”

The words “perforated all over with fine holes” are shown by the drawing to mean all over the front face of the “square iron box.”

We need not go outside of this British patent to find all the general features of Ballard’s construction. That earlier patent shows, in gas stoves for heating buildings, a frame, an outlet flue, an upright hollow burner, with perforations through its front face communicating with the interior chamber, provision for supplying to this inner chamber a mixture of air and gas which will issue through the perforations, and burn, when ignited, among tufts of asbestus secured to or suspended in front of the burner. The only thing here lacking is Ballard’s special arrangement,, whereby the tufts of asbestus fiber are attached to the face of the hollow slab by direct insertion in certain of the perforations. Without, then, particularly discussing the other proofs leading to the same conclusion, we here content ourselves with saying that the prior art imperatively requires that the patent in suit be confined and restricted to the precise construction thereby illustrated and described; assuming that the patent can be sustained at all. Such limitation, we think, is also imposed upon the claim of the patent by the proceedings- which took place in the patent office. Ballard originally made the following claim, namely:

“In that class of gas stoves known as the ‘cosy,’ the employment of a hollow slab of fire clay or other suitable refractory material, provided on its front surface with burners and asbestus fiber, so arranged that the mixture of gas and air supplied to the interior of the slab will, when ignited at the burners, uniformly heat the said fiber, and economize the consumption of gas, substantially as herein set forth.”

*353This claim was rejected by the office upon references to several prior patents. The applicant them canceled the above claim, substituting therefor the one now found in the patent. In a letter dated May 21, 1891, addressed by-the applicant: to the commissioner of patents, canceling the claim first made and substituting the other, Ballard wrote thus: “Applicant now claims his exact construction only, and, as none of the references have this, it is hoped that we will get a speedy allowance.” The grant of the patent having been thus procured, certainly the complainant below was in no- position to insist ujíon a con-síiucüon of the allowed claim which would make it practically the; equivalent of the rejected claim. Morgan Envelope Co. v. Albany Perforated Wrapping Paper Co., 152 U. S. 425, 14 Sup. Ct. 627, 38 L. Ed. 500. Moreover, the terms of (he claim are most specific, and it must be read with reference to the particular description contained in the specification and illustrated by the patent drawings. Fay v. Cordesman, 109 U. S. 408, 421, 3 Sup. Ct. 236, 27 L. Ed. 979; Knapp v. Morss, 150 U. S. 221, 228, 14 Sup. Ct. 81, 37 L. Ed. 1059. Thus read, no infringement of the claim appears. The construction shown by lía Hard is not followed by the defendants. According to Ballard’s specification, as we have already seen, the asbestus is inserted directly into certain of the perforations on the front face of the slab, and is thus held in place. These perforations are used as sockets for the asbestus. No other means for securing the asbestus on the slab are suggested. The perforations are widely apart. The asbestus fiber is arranged in isolated tufts, and large surface spaces are entirely bare. The gas stoves manufactured by the defendants are materially different from the stove of the patent in suit. In the defendants’ construction the face of the slab is provided with a series of horizontal rows of numerous minute, closely arranged, burner perforations; and between those rows, and in immediate proximity, continuous strips or banks of asbestus fiber are cemented upon'the fiat, unbroken surface of the slab, in such maimer as to cover the whole face of the slab with a coat of asbestus. The asbestus is permanently and effectually secured to the face of the slab by silicate of soda cement, which, it seems, was unknown in the art at the date of the patent in suit. The evidence is quite convincing that the two constructions are substantially different both in form and in results. Waiving-'the question whether the patent possesses any patentable novelty in view of the prior art, we are clearly of the opinion that it must be confined to Ballard’s exact means for securing the asbestus to the slab, and that the defendants do not infringe the patent. The decree of the circuit court is reversed, and the case is remanded to that court, with directions to enter a decree dismissing the bill, with costs.