Specialty Mach. Co. v. Ashcroft Mfg. Co.

213 F. 35 | 2d Cir. | 1914

LACOMBE, Circuit Judge

(after stating the facts as above).

[1] Judge Holt evidently thought that Hazeltine’s advance on the prior art was unimportant; that all he did was to substitute one kind of well-known gearing for another; that he merely used meshing cogs, instead of frictional contact, to drive the reciprocating driven shaft. We db not concur with him, and think he was misled by two models which were introduced by the expert for the defendant as being constructed m conformity to patents in the prior art. If the Taylor patent or the Millingar patent disclosed devices such as the models represent, invention as to the subject-matter of these three claims would no doubt be negatived; but the models do not represent the disclosure of either of these prior patents.

The Taylor patent (255,208) shows a machine in which the rotation, on its longitudinal axis, of the shaft which carries the part to be ground is effected by means of a conical head on the shaft engaging with beveled flanges on driving wheels arranged face to face; this pair of driving wheels is called a pulley in the patent. Parts of the flanges of the wheels are made thicker than the rest of the flange. In consequence the cone is first engaged .by one thickened flange, frictionally, and revolved in one direction, and is next engaged by the other thickened flange and revolved in the opposite direction. The model which the witness produced has not only this capacity for thus reciprocating cone-motion, but also by reason of the circumstance that the thickened part on one flange extends circumferentially further than the like part on the other flange, the position of the part to be ground is positively advanced by the action of the pulley at each revolution of the same. In other words the model has precisely the mode of operation of complainant’s machine, except that the engagement with the shaft head is frictional instead of cogged. For this inequality in the length of flange, however, there is absolutely no authority in the Taylor patent. It says merely:

“The pulley is made with two like flanges M and N, one on either side, the parts of each flange being of unequal thickness, and said flanges being so arranged in respect to each other that’ the thick part of each flange is opposite the thin part of the other flange. The thick part of either flange extends about halfway around the flange, and may be a separate piece attached to the flange. The inside of each flange is beveled and made to fit the cone D,”

There is nothing in this to indicate any difference in length between the two thickened portions. Figs. 3 and 4 of the drawings are “views of the inside of the flanges,” the thickened portions are shown extending each about halfway around its flange, and are of precisely the same length. Referring to the operation of his machine Taylor says:

“The thick part of one flange rotates the cone in a’ given direction during about one-half of the revolution of the pulley, and the thick part of the other flange rotates the cone in the opposite direction during the remainder of the revolution.”

Further down he says:

“The thick portion of each beveled flange being equal to about one-half its diameter, more or less, but not exactly one-half, will cause the cone to revolve in such a way that as (its) motion is reversed the grinding in of the plug will not commence and leave off at the same point, thus obviating excessive grinding at any point.”

The meaning of this last quotation is so obvious from examination of the specifications and drawings that it would seem, even an “expert” might perceive it. Since neither thickened portion extends quite halfway around its flange, there is a brief moment when one thickened portion has let go and the other thickened portion has not yet caught on; the cone will then be free, with its last acquired motion of rotation to run on till the other thickened portion arrests and reverses it. While it thus overruns it necessarily at each revolution of the pulley shifts the position of the grinding plug on- its seat, so that grinding “will not commence and leave off at the same point.”

We have given so much space to this discussion in order to call attention to the fact that, although models are often helpful when descriptions and drawings are intricate, their helpfulness depends on their accuracy; when they are inaccurate, they only increase the labor involved in a study of the case. So, too, the so-called illustrative models, -in which the designer has taken fragments from several prior patents and put them together as he thinks he might have done if called upon to produce a machine before the structure of the patent was disclosed, are often confusing.

The same witness testified as to the Millingar prior patent (177,873), and produced a model Exhibit B, which he said was intended to illustrate, both the Millingar and Taylor patents. This model uses toothed gears, as Millingar does, instead of the frictional engagement of Taylor. The model shows the opposite toothed segments of. unequal length, so that, as in the patent in suit, there will be a positive shifting (through the action of the teeth) of the valve on its seat. He testified that the Millingar patent showed engaging segments of unequal length. The specifications of the Millingar patent disclose no such inequality ; the witness assumed he had found support for his statement in one of the Millingar drawings. On cross-examination he was asked to measure the segments shown on this drawing, and admitted they were of exactly equal length. The draughtsman, however, made a mistake in drawing the teeth, putting 10 on one segment and 11 on the other. If constructed to conform to the drawing, the teeth would be of unequal thickness and the machine would not operate at all, a fact the witness was forced to admit. Certainly Millingar does not show periodical positive change of seating position.

As to another prior patent (Chadwick 430,142) the testimony of the same witness is directed, not so much to the patent, as to a model which he had constructed to illustrate it. With the experience derived from the study of the other models, we have confined our investigation to the patent itself. It shows an intricate and highly complicated machine, but apparently shifts the valve on its seat by frictional means applied at intervals by independent hand-operation. The passages of the specifications are too long to quote; they may be found on page 2, lines 22 to 57, and page 3, lines 50 to 100. It does not disclose Hazeltine’s device; it is suggestive that it was 17 years after Chadwick before the positive automatic, unitary structure of the patent in suit appeared.

In Porteous (427,904) there is positive mechanism for reciprocating the valve key or plug through equal arcs, and the valve seat is intermittently knocked loose longitudinally from the plug for a moment, and connected frictionally to the driving mechanism, so that for a fraction of a revolution the seat and valve travel together in the same direction and then co-operate again “in a fresh relative position.” Mercer (322,840), a washing machine, is in a nonanalogous art. The other prior patents, which have been examined, need not be referred to; we find none of them which negatives invention.

[2] As to re-issue. The original claims will be found in the opinion below. It seems to us that the change effected by re-issue was not a broadening of the claims, but a narrowing of them, and that, thereforé, the re-issue more than two years after the granting of the original patent was proper under Eames v. Andrews, 122 U. S. 40, 7 Sup. Ct. 1073, 30 L. Ed. 1064. We think that Miller v. Brass Co., 104 U. S. 350, 26 L. Ed. 783, and. the other cases cited do not apply. As originally drawn the claims would cover devices of the prior art; as re-issued they were narrowed down to cover the unitary, positively acting, novel combination of Hazeltine.

There has been much discussion between the applicant and the Patent Office, but we find no action taken by the patentee which can be held to be an abandonment of claims for the device which his original application fully disclosed, or which restricted him to any particular number of mutilated segments or any specific interarrangement of them to effect a particular cycle of movement. Between the successive engagements of gear segments with beveled pinion there is a movement of advance of the valve on the seat; whether such movement occurs once in a revolution or more frequently is unimportant, the claims cover both contingencies, and we find nothing in the file wrapper to restrict them to any specific number of advances. It is understood that defendant does not dispute the charge of infringement, if. the claims be not restricted to a particular cycle. But if this be an error, we have, from examination of his device, reached the conclusion that it does infringe.

The decree is reversed.