This is an appeal from an order for preliminary injunction against infringement ■ o.f the Tufford reissued patent No. 14,049, dated January 11, 1916, for rubber heel. This is the fourth branch of the litigation on this patent which has reached this court. In Fetzer v. I. T. S. Co., 260 Fed. 939, 171 C. C. A. 581, and in U. S. Rubber Co. v. I. T. S. Co., 260 Fed. 947, 171 C. C. A. 589, we construed various claims of the patent and held that they were valid and infringed. In Elyria Co. v. I. T. S. Co. (C. C. A.) 263 Fed. 979, we decided that the form of heel there involved did not infringe. Reference to these opinions will make it unnecessary to repeat here the general situation.
[1] We have thought that the presence in most of the claims of the phrase “concavo-convex on every line of cross section” imported into them a requirement that the upper surface of the heel be “saucer-shaped” rather than “scoop-shapéd,” and the present case presents, for its first problem, a question of the application of these definitions. It is quite clear to us that the lift of the present defendant, the “T. P.” heel, must, unlike that of the United States Rubber Company, be classified with that of the older patent to Nerger, rather than with Tufford. *251Every test suggested in our former opinions and every theoretical principle of operation requires this conclusion. Because there has been misunderstanding of the principle upon which we intended to distinguish, and for a clearer presentation of the matters to be considered, we here insert drawings indicating the longitudinal cross section and the transverse arch at the breast, each coincident with the upper surface, of the four varying forms of lifts so far involved in the litigation.1
a — a Line of under surface of leather heel.
b~b Line of central attachment point.
c — c Middle of longitudinal center.
d — d Center of rear upper edge.
e — e Upper breast corners of transverse arch.
f — f Center of breast edge.
In order that these lines should truly represent their,relative situation, they are shown in each case as they will be when the upper edge of the lift is horizontal; i. e., when the lift is placed up against the horizontal bottom of the heel (indicated by the dotted line a a) ready for attachment, hut before applying any distorting pressure. As we pointed out in the Fetzer Case, 260 Fed. 940, 171 C. C. A. 581, some such common basis for positioning of the various lifts must be adopted, or there can be no intelligent comparison of the different forms. If the lifts are merely rested loosely upon the flat surface of a counter or desk, and are allowed to position themselves, it is obvious that whether the upper surface will hold water like a saucer, or allow the water to run out like a scoop, will be dependent, not at all upon the form of that surface, but rather upon how the weight of rubber may happen to be distributed, so as to fix the center of gravity, or upon such irregularities as there may be on the under surface. On this rather natural, but deceptive, test, the T. P. will rock rearwardly into such a position that the top seems to be saucer-shaped; but when the test is corrected, to bring it to the common standard, the above drawings make it clear that on the longitudinal cross section, the lowest spot is at the breast, and is not centrally disposed.
Not only is this theoretically true, but inspection and careful measurement of the numerous exhibits now presented show that every one of the I. T. S. and the U. S. Co.’s lifts has a low spot centrally disposed, *252while in every one of the T. P. lifts the distance from the top plane to the longitudinal center of the upper surface constantly increases as the breast is approached, and is greatest at that point. Applying the. practical test of whether liquid is retained in this cavity when the top edges are made horizontal, we find that there is no such retention; it all runs out, except that minute quantity which will not flow over a relatively dry surface. The affidavits which state the contrary result must be based upon an abnormal positioning of the lift, unless water has lost its inherent tendency to run down hill.
We do not overlook that Nerger’s upper central longitudinal line, from center to breast, is horizontal, while the T. P. is a curve; but this is immaterial, as it is a downward curve — not upward — and so is even further away from Tufford than Nerger is. If Nerger is to be distinguished, because horizontal, the T. P. is yet more to be distinguished, because more than horizontal.' There is no escape from thp conclusion that the T. P. lifts do not infringe those claims of the patent which call expressly for the concavo-convex form on every line of cross section.
It 'is said that claims 7 and 9 do not contain this limitation, and that therefore a broader question is presented. Claims 5, 6, and 8 have, for one element, “a body portion of concavo-convex form on every line of cross section.” Claim 7 (and the same is true of claim 10) describes this element as “a body portion, the attaching face of which is concave and the tread face of which is convex on every line of cross section.” It is now urged that in claim 7 the descriptive restriction, “on every line of cross section,” applies only to the immediately preceding “convex” and not at all to the more remote “concave.” This is to say that these claims contemplate a structure which, upon its lower or tread surface, should be convex on every possible cross section, but on the upper surface need not be of corresponding fohn, but might be of any shape which would generally respond to the word “concave.” Grammatically, this construction is possible, but we think it unnatural. No such variant is suggested in the specification or the file wrapper proceedings, and, as indicated hereafter, we should regard that construction as a departure from the principle of Tufford’s invention, as that principle is imported into claim 7 by the. later requirement in that claim that there must be uniform pressure upon the edge all around when the central portion is forced up into position. We think the form, of this phrase used in claims 5, 6, and 8, and the form used in 7 and 10, as above quoted, are intended to express precisely the same thought.'
[2] Claim 9, as it is found in the patent, is quoted in the margin.2 So far as concerns the particular phrase so far treated, this claim re■quires only that the attaching face should be concave and the tread con*253vex. If these words could have the broad quality or meaning to which they would often be entitled, the T. P. lift would in this respect respond, in the common sense of the words, and perhaps in the technical sense. It would, not be impossible to speak of the scooped-out upper surface of the T. P. lift as “concave.” It is said that the presence of the steel reinforcing plate in Nerger prevented the suction effect which this claim calls for, and hence that this claim is rightly to be distinguished from Nerger in that particular, and would be infringed, by a.lift, even though it were in Nerger’s exact form, without the reinforcing plate. For the purpose of this opinion we may assume, without intending thereby to decide, that the mere omission of this plate from Nerger would be invention, if the lift then would accomplish a useful new result. That assumption does not end the question; the actual limitations of claim 9, expressed or necessarily implied from its language, remain, and their true interpretation makes important a resort to the file wrapper history. Although there is some confusion, caused by the redrafting of a body of claims, rather than by preserving their individual identity, this one is fairly traceable to claim 10 in the application as filed, which was:
“A heel lift, formed, of yieldable material and provided with a normally convex tread face and a normally concave attaching face, defining a suction area, whereby, when the lift is positioned on the flat surface of a heel and pressure is applied to the convex face of the lift, the (margin) entire attaching face 3 of said lift will be retained in engagement with the (margin) face of the heel by the action of the suction area.”
This claim, with others, was rejected. Thereupon applicant amended by adding, as No. 11, a claim which, in still other language, led to the functional statement:
“The attaching face forming an extended suction area to hold the lift in intimate contact with a shoe heel when flattening pressure is applied to the convex tread face of the lift.”
These claims were again rejected on the same reference, and on reference to Ferguson. See opinion in Fetzer Case, 260 Fed. 943, 944, 171 C. C. A. 581. Thereupon applicant added to his specifications a statement of the suction effect, whereby the lift would be held against the heel while the workman was nailing it on, and amended claims 10 and 11 by substituting “entire attaching face” for “margin,” and claim 11 by substituting “face” for “member.” These claims were all rejected on reference to Nerger. Applicant canceled them all, and substituted the following under the numeral 8, being the immediate antecedent of issued 9 (as filed, it contained the words in parentheses and omitted those in italics):
(8) Í). “A heel lift of resilient material comprising a body portion, the attaching face of which is concave and the tread face of which is convex (and normally held in such form by its inherent resiliency only), the concave face of the lift being (free from projections), unbrolcen, and lying entirety below a plane passing through the rear upper edge and breast corners of the lift whereby, when the convex tread face is depressed to flatten said lift, a suction *254will be created between tbe lift and tbe beel to bold tbe attaching face of tbe lift throughout its entire extent in contact with the exposed face of tbe beel.”
This again was rejected, with the statement by the examiner that Nerger’s rubber heel lift had an upper surface concave on any line of cross section and tread surface concentric thereto, that Nerger fully responded to all of these claims, except as applicant omitted the reinforcing plate, and that there was no invention in the mere omission of the plate. The examiner also said that the temporary adhesion and suction, relied upon to draw the lift into contact with the heél before nailing down, were not disclosed in the specification, and in any event would not occur, because the nail holes would prevent a vacuum. The statement of nondisclosure was true, as the amendment to the specification which had been introduced, describing this preliminary suction effect, had been, by another amendment, erased. Applicant then amended the specification by inserting the matter found in lines 83-99 of page 2 of the specification as issued, and redrafted and revised the claim, so that it took the amended form shown above. After some further discussion, the claim in this form was allowed.
Mindful of the danger of giving too great weight to the mere arguments of the attorneys in Patent Office proceedings, we pass by the repeated statements and arguments which show the clear understanding of the attorneys that the “saucer shape” was of the essence of Tuf-ford’s invention, and coming to the insertion or retention in claim 9 of language which must be interpreted, we find that Tufford was presenting a combination which was expressly distinguished from Ner-ger by the phrase “normally held in such form by its inherent resiliency only,” and was perhaps distinguished from Nerger by the further requirement that the upper face was of such concave shape arid so smooth that the flattening of the concavity would give a suction which would hold the entire extent of the surface of the lift in contact with the exposed surface of the heel. Being rejected upon a repeated reference to Nerger, he amended by striking out one clause and inserting another. He erased “normally held in such form by its inherent resiliency only.” He thereby ceased to rely upon merely the absence of Nerger’s plate as distinguishing. He inserted the requirement that the concave face should be below the plane passing through the rear upper edges and the breast corners of the lift. We have held, in the U. S. Rubber Co. Case, 260 Fed. 948, 171 C. C. A. 589, in effect, that this insertion of itself would not distinguish from Nerger; hence we-can now only say that its addition at this time serves to emphasize that the suction effect was thus presented and relied upon as the substantial feature distinguishing from, and showing invention over, Nerger. In effect, the Patent Office said:
“To omit Nerger’s plate is not invention, and, unless you bave some better distinction than that, we will not grant you a patent.”
Tufford replied:
“The important distinction is in tbe shape and bind of concavity of tbe upper surface. X bave a concavity which, by its flattening, produces an effective suction over tbe entire surface; Nerger has not that bind of a concavity; *255and I am, in claim 9, distinguishing by calling for an unbroken concave face lying entirely below a certain plane, and of such shape that [whereby] this extent of suction, thus effective, will be created.”
We can see no reason why the settled principle of estoppel does not apply to this situation. Unless Tufford had put into his claim, originally or by amendment, language which distinguished from Nerger as to the shape of the upper concavity, he would not have received his patent, and he must be held to the restriction which he has thus accepted.
This construction of claim 9 does leave as much distinction between that claim and some others as there otherwise might be; but, even then, the case is the common one where the applicant has used in different claims different forms of expression for slightly varying ideas, and it is quite conceivable that there should be forms of concavities not geometrically perfect which would, nevertheless, bring the substantial result called for by claim 9; indeed, we so held in the U. S. Rubber Co. Case.
Nor can we think this construction of claim 9 overnice. Rather is it merely confining Tufford to his actual invention. We quite agree with Judge Brown in his conclusion (I. T. S. Co. v. United Race, etc., Co. [Dist. of R. I., June 11, 1920 ] 266 Fed. 375), that Tufford’s meritorious advance was in the effect he secured, and we think this was an effect of which suction is not the operative element, but is rather the symptom. Tufford aimed at a lift of such material, and with a surface so shaped that the flattening of the center would produce an automatic intensive sealing at all.the edges, including the breast. Merely omitting the plate from Nerger would not get this result, because Nerger’s shape was not right. Nerger’s breast edge, without the plate, would not seal, except as nails might be driven in far enough and near enough to the edge, so that they would produce the sealing pressure, even if the lift were flat. The omitting of Nerger’s plate by itself will not get the desired result; the reshaping of Nerger’s upper surface is also necessary ; both must co-operate; and hence the construction which we give to claim 9 is essential to its validity.
Referring to this suction effect as an indication or symptom of the presence of the essence of the invention, we are satisfied that the T. P. lifts do not have it in substantial extent, either theoretically or practically. Theoretically they cannot. The depth of the transverse arch at the breast is such that, when this part of the face is pressed up flat, there is a strong tendency for it to spring down away again and resume the transverse arch form. To resist and overcome this there must be a greater force applied on the longitudinal arch, caused by the pressing up of its center. This greater force cannot exist when the transverse arch in its normal shape, to which it tries to return, is deepest at the breast (unless the effect of this condition is in some way neutralized). Referring back to the drawings, it seems quite plain that, if the center is driven up and nailed to the heel (but without compressing the rubber between the nail head and the leather heel), the effect will be: In Ner-ger, it will tend to cause the line c f to lie up flat against the heel, and the “retractive” counterforce from the transverse arch at the breast will *256tend to make an opening at this point. In the T. P. heel, line c f will become an arch, fastened up at one of its ends only, and the other end will not be held up with any more force, but with less, than if the arch were a horizontal line. In both the Tufford and the United States Rubber Company heels, the natural tendency of both ends of the longitudinal arch to return to shape and to take a higher place than the center'will produce a maximum amount of pressure at the breast.
Practical experiments with the samples submitted confirm these theories. It must not be forgotten that perfectly plane surfaces of rubber will adhere by suction to a hard smooth surface, like glass, if sufficient pressure is applied to squeeze out any air that might be between them. That is not the kind of suction to which Tufford refers. He refers to the kind which will be more persistent, and which will arise undér imperfect conditions, because the sealing pressure is hardest at the edge where the air would naturally come in. We do not find that any, one of the T.- P. samples submitted will adhere by suction to a smooth surface, even momentarily, as the result of merely flattening, and without a general squeezing pressure. Observation shows that, as the concavity is flattened by pressure at the center, when the center comes into contact with the heel, the breast arch still remains open, or tends to, and only additional pressure and compression at the center, further distorting the shape, firmly close the breast. With any release of pressure, the breast springs open first, or tends to.
We are satisfied that, whether it should be said to be due to the depth of the transverse arch at the breast or to the fact that in its normal form the longitudinal surface does not rise from the center to the breast (which we think two forms of expression for the same fact), the performance of the T. P. lift, in respect to this general theory of a sealing pressure strongest at the breast edge, does not follow Tufford, but rather Nerger. Merely by the central attachment Tufford gets the strongest sealing pressure at the breast, while with the same center attachment the T. P. gets no substantial pressure at this point, and it can get a further and effective sealing only by further fastening or nailing near to the breast edge. Samples have been presented, comprising lifts fastened to a board, and showing an irregular edge in the board sealed by the breast edge of the Tufford and T. P. lifts in apparently the same manner; but this effect is produced in the Tufford lift (in some measure) by its kind of concavity, and in the T. P. by driving the fastening nails nearer the edge, and continuing the driving until the rubber is compressed by the head of the nail and thereby squeezed into the irregularities. We get measurably the same effect with the O’Sullivan flat heel; hence this test is not persuasive.
We conclude that claim 9 is not infringed, and we think the preliminary injunction against the patent infringement should not have been issued. The order, therefore, will be reversed, and the case remanded for further proceedings in accordance with this opinion.
The drawing of the line in the Nerger patent is criticized by Tufford as not definite enough to rely upon. We therefore adopt the form insisted upon by Tufford as the true one, as indicated by the Nerger heels manufactured.
The inserted drawings are inaccurate representations of the exhibits, in that, in the latter, the depth of the point o below the line a a varies, while in the drawings this is arbitrarily made equal; but the relative shapes of the upper concavities are correctly indicated.
Claim 9: A heel lift of resilient material comprising a body portion, the attaching face of which is concave and the tread face of which is convex, the concave face of the lift being unbroken and lying entirely below a plane passing through the rear upper edge and the breast corners of the lift, ‘whereby, when the convex tread face is depressed to flatten said lift a suction will be created between the lift and the heel to hold the attaching face of the lift throughout its entire extent' in contact with the exposed face of the heel.
The words in parenthesis are the original form; those in italics were substituted by an early amendment.