This suit charges infringement of letters patent 661,640, granted November 13, 1900, to Alfred Heald for an improvement in woven fabrics. It is not for a machine or for a process of manufacture, but for the manufactured article itself, irrespective of the means by which it may be produced. Of course, as the fabric belongs to the art of weaving, it will as a practical matter be produced by using a loom with some of its adjuncts and appliances; but it may be .as well to call attention to the fact that the invention has nothing to do with the loom or with any of its parts, but is concerned solely with a specific article, which is built up by combining threads in a particular fashion. The specification thus describes the fabric:
“My invention consists of a woven fabric intended especially for carpets, rugs, and the like, and of such character that, while it can be readily produced upon a loom such as is ordinarily employed for weaving ingrain carpet, it will form an acceptable substitute for an ordinary brussels carpet, as it bears a close resemblance to the latter in appearance and diversity of coloring, while it has the additional .advantage of being, like an ingrain carpet, double faced and reversible; the fabric being, moreover, of close texture and having its faces or plies well tied together so as to be free from the objectionable pockets of an ingrain-carpet fabric. The fabric is also considerably cheaper than a brussels carpet fabric of corresponding quality.”
Four kinds of thread are used by the patentee, heavy warp-threads, heavy weft-threads (called also figuring, or patterning, warp-threads and weft-threads), fine or binding warp-threads and fine or binding *892weft-threads. The heavy warp-threads and the heavy weft-threads, either or both, are used to form the figure or pattern, when a figure is intended to be displayed, and the binding threads are used to fasten the fabric together, so as to give it firmness and stability. The method of construction is described in general terms as follows:
“My improved fabric in its preferred form consists of sets of patterning warp-threads alternating with pairs of fine binding warp-threads, and pairs of heavy weft-threads alternating with pairs of fine weft-threads. Each set of patterning warp-threads may comprise two, .three, or more, and those threads may constitute either in whole or in part the surface-threads of the fabric whereby the pattern is produced. Each set of these warp-threads forms on each face of the fabric successive loops very similar to the looped pile-threads of an ordinary brussels carpet; the difference being that, instead of being tied into a backing fabric and otherwise unsupported, these loops are filled and supported by the heavy weft-threads of the fabric. The heavy weft-threads may serve simply as pile-loop filling or supporting threads, or may, in addition to performing this duty, in certain parts of the fabric appear upon either or both faces of the fabric in other parts, so as to diversify the coloring of the pattern. The fine weft-threads hind down and indent the pile-forming warp-threads between the successive pairs of heavy weft-threads, so as-to-form the desired pile-loop effect, and the fine warp-threads bind together the fine weft-threads of each pair and also the heavy weft-threads of each pair, so as to give the fabric the desired firm texture.”
It will he observed from this description that the invention is not rigidly confined to one variety of fabric. Laying aside the plain or solid colored variety, it is clear that at least two other varieties are contemplated: One, where the pattern is produced by differently colored heavy warp-threads; and, another, where the pattern is produced by the combined use of such threads and of the heavy weft-threads. The heavy weft-threads may be brought to either surface of the fabric and may thus diversify the coloring of the pattern, or they may be buried out of sight, and in that event they will simply stuff out, or fill, the loops that are made by the binding-threads as they tie down the heavy warp-threads at frequent intervals. In the present suit the controversy is wholly about the variety first referred to — where the pattern is produced by differently colored heavy warp-threads — for it is only the first claim of the patent that is in issue, and it is clear from the language of the claim that in the fabric there described the heavy weft-threads do not come to the surface at all:
“1. A woven fabric in which sets of pile-forming warp-threacls alternating with binding warp-threads are interwoven with heavy weft-threads alternating with binding weft-threads; said pile-forming warp-threads producing pile-loops, on both faces of the fabric, the heavy weft-threads distending said pile-loops, the binding weft-threads tying and indenting the pile-forming warp-threads between the heavy weft-threads, and each binding warp-thread passing around a heavy weft-thread, thence to and around a binding weft-thread, and thence back and around a heavy weft-thread, substantially as specified.”
The complainants concede that the prior art has several examples of double-faced fabrics, in which all the threads used by the patentee are employed to produce pile-like loops on either face; but it is declared in the brief of counsel that in none of them—
“were pairs of binding warp-threads alternated with sets of pile-forming warp-threads, and interwoven with pairs of binding weft-threads alternated *893with heavy weft-threads in a manner to pass from the respective binding weft-threads engaged thereby reversely under and over adjacent heavy weft-threads within the pile-loops in.a manner to exert in opposite directions a tying and binding force upon the binding weft-threads, and upon and about the heavy weft-threads, as comprehended by claim 1 of the patent. Or, in other words, the essential structural characteristic so succinctly described on page 1, lines 59 to 66 of the specification, is not found in the prior fabrics.”
This essential characteristic (which has already been quoted) may be repeated:
‘The fine weft-threads bind down and indent the pile-l'orming warp-threads between the successive pairs of heavy weft-threads, so as to form the desired pile-loop effect, and the fino warp-threads bind together the fine weft-threads of each pair, and also the heavy weft-threads of each pair so as to give the fabric the desired firm texture.”
I have read attentively the record and the briefs of counsel, but I do not think it necessary to discuss the arguments in detail. The defendants insist that the claim has been anticipated, especially by the British patent 344 to Alfred Newton, dated February 2, 1872; the patent No. 388,682 to John Humphries, dated August 28, 1888; the patent No. 638,410 to Thomas Dornan, dated December 5, 1899; the patent No. 312,220 to Priestley & Kunkler, dated February 10, 1885; the patent No. 335,567 to George Crompton, dated February 9, 1886 (reissued July 6, 1886) ; and the patent No. 541,645 to William Weaver, dated June 25, 1895. Without following the expert testimony that attacks or supports the defense of anticipation, I think it is clear that, in view of the prior art disclosed by these references, the claim under consideration can only be saved if it be strictly construed— confined in effect to the specific fabric thus described. The complainants do not make and sell the article that is shown in the drawings ; but in my opinion they have not departed from the claim if it is read in the light of the specification. The essential characteristic to which their counsel refers, namely, the course of the binding warp-thread as described in the last few lines of the claim, is undoubtedly present in the fabric that they do make and sell — the so-called Kashmir rug or carpet — and the other variations from the drawings that appear in the fabric are, I think, warranted by the specification and by the other language of the claim. But, while the patent may thus be saved, the conclusion is unavoidable that (if this view is correct) the defendants do not infringe, for in their fabric the course of the binding warp-thread is plainly different. It is correctly traced in the brief of their counsel (page 41):
“Each binding warp-thread, after passing around a heavy weft-thread — or, more properly, two unseparated heavy weft-threads — passes between a pair of binding weft-threads and thence around two ««separated heavy well-threads, and thence upward and over and around a binding weft-thread, and thence downward and backward around another binding weft-thread, and then resumes the course as first described. The two binding warp-threads alternately and separately perforin their binding functions. The result is an altogether different tying of the threads. The threads are interwoven differently.”
Indeed, the defendants’ fabric in this respect is nearly identical with the fabric of the Weaver patent, to which reference has been made; *894the principal difference being in a part of the course taken by the binding warp-threads.
It should be said, also, that the commercial success that has attended both the complainants and defendants is, in my opinion, due in large part to a change in the course of trade. Probably this change has been as influential as any improvement in the respective fabrics that is due to the method of weaving. The earlier fabrics made under the patents to Priestley & Kunkler and to Crompton were also double faced, and imitated brussels carpets with a fair degree of success; but it was then necessary that they should be made of wool. The trade demanded that wool should be used, and would not have accepted cotton, even if the art of dyeing had at that time been adequate to the producing of colors in cotton that would be as durable and bright as colors in wool. The consequence was that the Priestley & Kunkler and the Crompton carpets were not able to compete with the cheaper tapestry carpets, and were compelled to surrender the market; but the conditions had changed when the complainants arid the defendants began to offer their goods to the public. The art of dyeing had improved, so that better and more variegated designs could be produced in cotton as well as in wool, and the trade and the public had now become willing to accept carpets made of cotton. As a result, fabrics that closely resembled the productions of the prior art were received with favor, but by no means for the sole, or even for the principal, reason that the method of manufacture had been materially improved.
Another ground for construing the claim in question strictly is to be found in the proceedings before the commissioner of patents. When the patent was first presented for allowance, claim 1 described the course of the binding warp-threads as follows:
“Each binding warp-thread passing alternately over a heavy weft-thread and under a binding weft-thread, or vice versa, substantially as specified.”
The claim in this form was rejected on the patents to Crompton and Newton. If it had been allowed'to stand, the language is broad enough to include not only the Crompton and Newton fabrics, but also the fabric made by the defendants. Being rejected, however, the applicant amended the claim so as to present the language that now appears, and stated that the claim had thereby been “rendered more specific than before in (its) definition of the course of the binding warp-threads.” No doubt such a result was thus produced, for the claim now is unquestionably more specific than it was in its original form; but the patentee, having limited his claim in order to succeed, is forbidden to insist upon a construction that would give him what he expressly surrendered. Morgan Envelope Co. v. Albany, etc., Co., 152 U. S. 425, 14 Sup. Ct. 627, 38 L. Ed. 500; Computing Scale Co. v. Automatic Scale Co., 204 U. S. 609, 27 Sup. Ct. 307, 51 L. Ed. 645. Since therefore the original claim was broad enough to cover the defendants’ fabric, while the claim as amended does not cover it, except by applying the doctrine of equivalents, it seems to me that the patent must be construed according to its limited sense, and must be confined to the particular article that is described.
*895Accordingly, for the purposes of this case, I assume that claim 1 of the patent is valid, when restricted to the fabric there pointed out, and [ hold that the claim, thus restricted, is not infringed by the fabric that is made by the defendants.
A decree may be entered dismissing the bill, with costs.