In Equity. Final hearing on pleadings and proofs of a suit for infringement of United States letters patent No. 574,894, January 12, 1897 (applied for June 2, 1894), to F. P. Ferry, assignor of one-half to Theodore Clark & Co., for hat-packing ring.

J. Edgar Bull and F. M. Smith, Jr., for complainants.
Chamberlain & Newman, for defendant.

LACOMBE, Circuit Judge. The first impression formed upon reading this patent is that there could be no patentable invention in so simple a modification of earlier forms. But the evidence is most persuasive to the conclusion that, trivial though it seems, the improvement is one which has commercially proved exceptionally successful; and if it were, as defendant contends, a natural development from earlier forms, it is difficult to understand why, in view of the demands of the trade, no one produced it during the interval succeeding the earlier patent of January 6, 1891, unless it were that more than the mere technical skill of the handicraftsman in the art was required for its conception.

The patent relates to what are known as hat-packing rings or stays. The manufacturers of hats ship these articles in tall boxes, each containing several hats. To keep the hats in the box separate, so they will not rub against or mar one another, hat-packing rings are employed. Hat-packing rings of various forms have been employed for years. Any plain strip of pasteboard of suitable width, curved to conform to the contour of the hats, might be employed for the purpose. It is obvious, however, that the sharp or rough edge of a piece of pasteboard would chafe the hats where it was in contact with them. Various expedients had been adopted, prior to this patent, to overcome this difficulty. Strips of paper had been pasted over the raw edges of the cardboard, or they had been bound with flannel or other soft material, or the edges had been broken over, so that they stood at an angle with the body of the strip, forming a flange or broader strip for the hat to

rest on.

The patentee, Ferry, in 1888 applied for a patent, issued January 6, 1891, covering a ring of a general cylindrical shape to contain the hat-crown; the edges of said ring being curled outwardly, so as to present a perfectly smooth, unbroken surface for contact with crown and brim of the hat. Fig. 3 sufficiently indicates the curved edges of the ring.

In this earlier patent, 444,343, the specification states:

"I am aware that packing-rings have been made with their edges boundwith felt or other soft material, so as not to chafe the hats, but my invention contemplates no such construction. I am also aware that a packingsupport for a single hat has been made wherein the top edge has been flanged or curled outwardly in order to afford a nonchafing support for the hat-brim, but this has never, to my knowledge, been done except in the instance of a separable ring; and, moreover, such construction could not be of advantage in packing a nest of hats, since it is essential that both the top and bottom edges of the ring should be so formed that the brim of the hat should not be chafed either on the upper or under surface. I therefore wish to be understood as distinctly disclaiming any flaring or curling of the edges of a hatpacking ring, save the outward curling of both the top and bottom edges of a closed pasteboard ring."

While affording theoretically a sufficiently broad and easy seat for the hat, this earlier ring was, comparatively speaking, expensive and difficult to manufacture. A considerable proportion of the product was unsalable because defective, and the beads, D, were liable to uncurl, kink, or break and chafe the hat. Moreover, this liability to uncurl or kink made it impracticable to dispose of them, save as completed rings, fastened together at the ends. They could not be nested and shipped in bulk. Freight to any distance, by reason of the bulk of parcel, was prohibitive.

The improvement of the patent in suit is sufficiently indicated in Fig. 2, and the following excerpts from the specification:

"In carrying out my invention, I take a piece of pasteboard, 1, of any suitable width and length, and curl over the opposite edges as shown at 2, 3, to form hollow beads, 4, 5, a surplus from such edges being left to form petticoats, 6, 7, which latter lie closely against the body of the strip. The object of the petticoat portions, 6, 7, is to give stability to the beads, 4, 5, and to prevent them from crawling back, and also give them greater body and flexibility. The hollow beads at one end of the strip are readily inserted within the beads at the other end, since the latter beads yield readily to permit of this, and exert a grasp firm enough to prevent any accidental slipping of an adjustment. The hollow bead affords great advantages, in that it is not stiff or unyielding, but is flexible and resilient [in

*

earlier devices the edge had been turned over and glued down, sometimes with an extra thickness of material inserted-manifestly an unyielding and nonresilient finish], while its shape is preserved by the petticoats, 6, 7, which, as above set forth, snugly lie against the body of the strip."

To reduce the field of discussion presented on the record, touching infringement, the third claim, only, need be quoted. It reads as follows:

“(3) As a new article of manufacture, a pasteboard hat-packing ring, having closed beads at its edges, and, with the extreme edges of the stock beyond the beads disposed flat against the body of the strip, substantially as described."

Infringement of this claim seems plain, upon a mere inspection of the exhibits. The only question of difference between the parties is as to the "hollowness" of the beads. The beads of the patent, as shown in the drawing, are "hollow," in the full sense of the term. Hollowness is achieved by turning over a mandrel, or otherwise, sufficient material to form a tubular structure. In defendant's rings sufficient material is thus turned over, but it is also collapsed upon itself from the top downward, so that it no longer looks like a "hollow bead," but it is not pasted together or otherwise held rigidly, so that it cannot perform the useful function of end into end insertion, which the evidence shows to be the great commercial advantage of complainant's patent. The superabundance of material, giving a greater bearing surface than would the mere turning over of the material on a knife edge, affords a similar support for the hat, and it also secures the necessary play for insertion of the other end of the ring-a play which would not be secured without such superabundance of material. Of course, if it were pasted down or rigidly secured, it would be merely the device of earlier patents; but since it performs all the functions of the device of the patent, and in the same way, the difference between the two is verbal only. A web hose is a hollow structure for the transmission of water, although, when water is not flowing through it, it may lie perfectly flat, and be folded upon itself, with no suggestion of a tube about it.

The evidence establishes with a conclusiveness rarely found in patent suits that the advance from Ferry's patent of 1891 to the one in suit has produced a marked saving in the cost of manufacture and in the amount of waste, and has vastly enlarged the output field of the manufacturer. The earlier rings had to be completed as rings before shipment; that is, the ends had to be fastened together, or the edges would uncurl. Then, since freight is regulated to some extent by the size of the package, the manufacturer could supply only his immediate neighborhood. The device of the patent may be "nested" and shipped to remote places, each ring to be there fastened by insertion. when put to use.

A number of patents have been introduced in evidence. It is sufficient to say that none of them anticipate. Some of them would seem to indicate that the improvement of the patent-now that we know of it was to be expected of the skilled workman; but when it appears that an enormous increase in sales awaited the man who devised it, and that from the issuance of the first Ferry patent, of 1891, to the

application for the patent in suit, June 2, 1894, no one but the patentee came forward with such a device, the court is constrained to conclude that it exhibits patentable invention.

The evidence introduced by defendant is mainly directed to establishing the proposition that the complainants made and sold petticoated rings more than two years prior to the date of application. The testimony on this branch of the case is conflicting. Application was filed June 2, 1894. The burden is upon defendant to show public use and sale prior to June 2, 1892. There is not a scintilla of evidence that petticoated rings were made by hand. On September 30, 1891, a manufacturer named Wilkins supplied complainants with a two-part machine adapted for curling over the bead edgings and fastening the rings as shown in the 1891 patent. This was the first machine for making packing rings used by complainants. On June 24, 1892, the same manufacturer delivered a new machine to complainants, which was. adapted to curl over a bead with a petticoat-not so long a petticoat as shown in the patent, or as is produced by later machines, but still sufficiently long to come within the patent. Both these machines exist to-day. They have been put in evidence and operated. The dates of their delivery are fixed beyond peradventure or dispute by book entries, dated bills, and letters. The complainants concede that they made rings with petticoats on the second machine and sold them shortly after its receipt. The question, then, is narrowed down to the single issue, were petticoated rings made commercially on the first machine? Those made on 'the second machine are out of the case. The weight of evidence seems to call for a negative answer to this question. Indeed, it would seem that such rings could not be made. upon the first machine, except experimentally by altering the adjustment of parts in the bead-forming portion of the machine beyond the capacity of the mechanism, wedging the guides out of alignment, while the subsequent passing of such experimental rings through the ringforming portion of the machine would practically ruin the greater portion of them. The defense of prior use is not established.

The evidence does not sustain the contention that Ferry was not an original inventor, by reason of a certain ring alleged to have been shown to him by the witness Hallock in February, 1890. That ring had its edges simply turned over flat down upon the stock. There was no suggestion in it of a bead, hollow, or collapsed. It was, as the witness says, "practically all petticoat."

There is no force in the proposition that complainant abandoned invention by stamping the petticoated rings made prior to the issuing of the patent in suit with the patent date of the earlier Ferry patent, on which they were an improvement.

Complainant may take the usual decree.

COMPTOGRAPH CO. v. MECHANICAL ACCOUNTANT CO.

(Circuit Court, D. Rhode Island. April 4, 1904.)

No. 2,644.

1. PATENTS-INFRINGEMENT-COMPUTING MACHINES.

A preliminary injunction against infringement of the Felt patent, No. 465,255, for improvements in recording computing machines, denied on the ground that the proofs left in doubt both the, validity of claims and the question of infringement.

In Equity. Suit for infringement of letters patent No. 465,255, for a computing machine, granted to D. E. Felt December 15, 1891. On motion for preliminary injunction.

Munday, Evarts & Adcock and Henry Love Clark, for complainant. Wilmarth H. Thurston and Warren R. Perce, for defendant.

BROWN, District Judge. This suit is for infringement of letters patent No. 465,255, to D. E. Felt, dated December 15, 1891, for improvements in recording computing machines. The subject-matter of claims 7 and 8 is "subtraction cut-offs." Claims 9, 10, II, and 12 relate to means for returning all numeral wheels to zero.

In performing subtraction on an adding machine, the numeral wheels are given the same movements as in addition. Subtraction is performed by the addition of complementary numbers. Since the numeral wheels rotate as in addition, the carrying-mechanism, whereby a complete rotation of a wheel of lower order moves forward one step a wheel of higher order, will continue to operate, and to produce an error in the result. This error will occur at the left of the wheels operated. The patentee says:

"Another object of my invention is to prevent the carrying of tens from any column to the next higher whenever a subtraction is made, by means of adding a complementary number."

Referring to the machine of a prior patent to Felt, the patentee says: "In my former machine, the 1 carried to the next higher numeral-wheel to the left of the highest numeral-wheel of the subtrahend was removed by adding a 9 to said numeral-wheel, and to all other numeral-wheels at the left of it. To save the carrying of the 1 above mentioned, and afterward adding the 9's as above described, I provide the following device."

The substance of the improvement was to provide for each carryingmechanism an individual cut-off, so that the operator may at will prevent the occurrence of the carrying. It is necessary to put but one carrier out of operation. When the operator makes a subtraction, a lever at the left of the highest column or series of keys in which a key is struck is pushed backward. This raises a carrying-pawl, and thus prevents the carrying from the highest series. The defendant. shows nothing in the prior art which will accomplish all that is accomplished by individual subtraction cut-offs.

It is quite clear that devices whereby all the carriers are simultaneously thrown out of operation are not full anticipations, since to throw out all the carriers would prevent the addition of complementary numbers. Felt, so far as appears, was the first to provide a comput