PULLMAN PALACE CAR CO. v. WAGNER PALACE CAR CO. et al

Circuit Court, N. D. Illinois.

April 17, 1889.

v.38F, no.5-27

1. PATENTS FOR INVENTIONS—PATENTABILITY—NOVELTY—UTILITY—CAR-BUFFERS.

The invention described in letters patent No. 878,098, November 15, 1887, to H. H. Sessions, is a frame-shaped plate applied vertically and transversely to the end of a railway car. The frame is of about the height of the car, and is so fixed that it can have no lateral motion except with the lateral motion of the body of the car, and is projected a short distance beyond the end of the car by backing springs. It is claimed that when the cars are coupled the plates act as spring buffers in frictional contact under constantly opposing spring pressure between the superstructures of the cars, and that part of the force of the bottom springs is transmitted along the plates to their tops, and that the plates diminish the shock of a collision, and resist the forces tending to create oscillation. Prior devices lacked such frictional contact under constantly opposing spring pressure, and such transmission of force. The invention was promptly adopted by the principal railroad companies, and there was evidence that trains on which it is used suffer less from collision, and that the cars sway less, than in the case of trains not using it. Held, that the device possessed patentable novelty and utility.

2. SAME—SUFFICIENCY OF SPECIFICATION.

As a skilled mechanic could construct the three forms of buffers described from the specification and drawings, the specification in that respect is sufficient, though but one form is shown in the drawings.

3. SAME—STRENGTH OF PARTS.

The patentee was not obliged to limit himself to top springs of any particular strength, but it was sufficient that he described springs powerful enough to diminish shocks and furnish frictional resistance to counteract wholly or partly the forces producing oscillation.

4. SAME—INFRINGEMENT—MECHANICAL CHANGE.

The substitution of an elliptic spring for a coiled spring, the patentee not limiting himself to the latter, is merely a mechanical change, and is insufficient to avoid a charge of infringement.

In Equity.

Suit by the Pullman Palace Car Company against the Wagner Palace Car Company, the Lake Shore & Michigan Southern Railway Company, and others, to restrain the infringement of a patent.

Offield & Towle, B. F. Thurston, and Runnels & Burry, for complainant.

George Payson and Coburn & Thacher, for defendants.

Before GRESHAM and BLODGETT, JJ.

GRESHAM, J. This suit was brought for an injunction and damages for infringement of letters patent No. 373,098, issued to Henry Howard Sessions, on November 15, 1887, for a “new and useful improvement in the construction of railroad cars,” and assigned by him, before issue, to the complainant. The improvement is thus described in the specifications:

“The invention hereinafter particularly described is embodied in the application to the individual cars, which, when coupled, will compose a train, of a frame-shaped plate arranged in a vertical plane parallel With a vertical transverse frame passing through the car body, and projecting, by means of backing springs, for a short distance beyond the end of the car. The height of said frame-plate, for the best results, should be substantially that of the height Of the car to which it is attached, and the same should be so shaped as to allow free communication between the ends of adjacent cars for the passage of persons through such frame plates. The purpose of the improvement is twofold: First, to diminish the racking effect upon a car body, due to its momentum when it is suddenly brought from a state of motion to a state of rest from any cause, as well as the same injurious consequences when a car is suddenly started from a state of rest; arid, secondly, to diminish the tendency to a swaying or oscillating movement, which is developed whenever a train is running at high speed upon an ordinary railroad track. The end to be accomplished is to cause the frame plates to act as spring buffers whenever cars are being coupled, or whenever a train is suddenly checked or started, and also to act as frictional resistance plates to oppose or counteract the influences which tend to induce a swaying or oscillating movement in the several cars of a train. In place of the arrangement of springs shown to exert pressure upon the frame plate, it is obvious that any other can be substituted which will meet the requirements of necessity or convenience, according to the judgment or choice of the constructor. * * * To employ this improvement it is not necessary that the ordinary spring buffers in use should be dispensed with. IP my judgment it would be well to retain such appliances to diminish the effect of shocks. By my improvement the body of the car is stayed against the racking effect of such shocks by the yielding frame-plate buffer, which is applied not merely in the line of horizontal planes of the platforms, but also in the lines of vertical planes extending substantially to the top of the superstructure, whereby the duration of the life of the car is greatly promoted. 418* * * The front ends of the upper set of backing springs take their bearings at the two upper corners, respectively, of the frame-plate, or, as shown in the drawings, against shoulders on the bars, c, c′, which bars are jointed to the frame plate at v′. The rear ends of the springs abut against the ends of the keepers, k, k′, and through the eyes of these keepers the bars, c, c′ can slide. These keepers are shown in the drawings as bolted to the sides of the vestibule extension of the car body, and the coiled springs, t, t′ are, for convenience, wound around the rods or bars, c c′. In case there should be no vestibule extension of the car body, the keepers may be attached in any convenient way to the main body of the car, so as to furnish resisting abutments for the pressure springs, and guides for the rods connecting with the frame-plate. The spring pressure to act against the lower portion of the frame-plates is obtained, as exhibited in the drawings, from the coiled spring, m, which takes a bearing at one end against the solid frame-work of the car, and at the other end against a cross-head beneath the, entrance platform car, which cross-head, by means of the rigid links, s, is connected with the threshold of the frame-plate, a, the said links or bars s, s′, being knuckle-jointed to the threshold-plate, o.”

The two claims, both of which it is alleged are infringed, read:

“(1) The combination with the end of a railway car of a frame-plate or equivalent series of buffers backed by springs, arranged with its face in a vertical plane, and normally projecting beyond the end of the car, whereby, upon the coupling of two cars, a spring buffer will be interposed between the superstructures of such adjacent cars above their platforms, and also frictional surfaces under opposing spring pressures to prevent the racking of the car frames upon sudden stoppages and to oppose the tendency of the cars to sway laterally when in motion, substantially as hereinbefore set forth. (2) The combination of a spring buffer or friction plate with the ends of the adjacent cars of a train, said buffers being, located on the ends of the superstructures of the cars, respectively, and substantially at the tops of the same, and so arranged that when, the two cars are coupled the faces of the buffers will bear against each other in contact under pressure, substantially as and for the purposes specified.”

It is claimed that by attaching the Sessions organization rigidly to the car bodies, so that the frame-plates have no capacity to move sidewise independently of the movement of the cars, the frame-plates act as spring buffers in frictional Contact under constantly opposing spring pressure between the superstructures of the cars, and that, thus combined with the ends of the cars, and acting in co-operation with the platform springs, the buffer-plates are far more effective in dissipating the force, of shocks in collisions, and in resisting the forces which tend to create oscillation, than if they were in close proximity, or even in frictional contact, but not under such pressure. It is not claimed that, the improvement possesses the merit of entirely overcoming the tendency to oscillation in cars in motion. All that is claimed for it is the power to dissipate the force of shocks, and to resist and overcome the swaying tendency to an appreciable and useful degree, as compared with anything previously known in the art. The unevenness and Curvatures in tracks cause cars in motion to vibrate, and it is claimed at, the. Sessions device possesses special merit, as compared with previous devices of the same character, in its ability to counteract the first impulses to these oscillatory movements. “It is common experience,” says the specification, “that when a train of drawing-room or 419sleeping-pars is traveling at high speed, there is induced in each car a: tendency to sway or oscillate laterally. The force which induces this tendency may be relatively a slight matter; but its continued repetition results in an aggregation of impulses which accelerate the oscillations and cause unpleasant effects upon the passengers especially when the road-bed has reverse curves, even of great radius. Especially is this experienced in; trains of sleeping-cars which are provided with upper sleeping berths, constituting, when occupied, a weight elevated high above the center of gravity the effect of my improvement is to provide a resistance to this; tendency, to oscillation by checking the same at the outset before the impulses which produce it have accumulated. The surfaces of the spring-backed frame-plates in contact are capable of resisting all ordinary impulses to oscillation induced by the movement of the train. Moreover, as the cars of a train do not generally sway in unison, but oscillate according to the effect of particular accidents or influences, the effect of combining the cars of a train by the aid of frictional surfaces in contact under considerable pressure, such as I have shown, is to dissipate all the lateral movements of each car throughout all the other cars so connected, and thus give steadiness-to the whole train.” The value of the Sessions device as a means of making collisions less injurious, and of diminishing the oscillation of cars in motion, and thereby promoting the safety and convenience of the traveling public, is abundantly established by the evidence in the, record. Its prompt adoption by the leading railroads of the country, especially the Baltimore & Ohio, with its numerous sharp, curves, attests its merits, and the large number of intelligent witnesses, many of them disinterested, who have testified from experience and observation of its steadying effect, leave no room for doubt upon this branch of the case. Persons, traveling on cars equipped with the improvement, are less affected by swaying movements than when traveling on cars not so equipped. Persons troubled with nausea experience less discomfort on trains operated with the improvement attached than when, traveling* on cars not so operated. Persons occupying upper berths of cars operated with the improvement attached experience less discomfort than previously, and trains operated with the improvement have sustained less injury in collision than trains colliding without it. These results are attributable to the Sessions device. On the evidence they can be accounted for in no other way.

Before discussing the patents, machines, and publications, which are relied on as anticipations of the Sessions improvement, we will briefly advert to the marked difference between English and American railway cars. American cars of the Pullman and Wagner class are from 50 to 80 feet long, 10 feet or more highland rest at either end upon six and eight wheel trucks. The cars and trucks are connected by means of swiveling joints, to allow the cars to keep the track, and turn curves. The floors or platforms rest upon the trucks; and the cars, which are heavy, are necessarily strongly braced and trussed to stand the severe strain to which they are subjected. English cars are about one-third the length and two-thirds the height of American cars. They are mounted on two axles in 420rigid down-hangings, secured to the sills of the platforms, so as to admit of a vertical spring movement. They have no swiveling attachment or movement, for the easy turning of curves, and the car bodies are necessarily short,—so short as to be within the unit of each curve they travel, otherwise they would leave the track. Owing to their comparative great length and height, our American cars, in motion, have a stronger tendency to sway than English cars. If it were possible to operate American cars with platform buffers three or four feet from the center of oscillation, as in England, there would be less necessity for the Sessions device, or indeed, for elevated buffers of any kind. As to the platform buffers described in the English and American patents found in the defendants' record, it is sufficient to say that in construction and mode of operation they are unlike the Sessions frame-plates in frictional contact under constantly opposing spring pressure interposed between the superstructures of the car bodies, and so attached to them that part of the force of the bottom or platform springs is transmitted along the frame-plates to their tops. While the Miller coupler and buffer, with its single central pair of coupling faces, and the Janney coupler and buffers, with their faces 26 inches apart, were great improvements on previous means of coupling American cars, they lacked the essential elements or features which characterize the Sessions organization. The English patent granted to Symons in 1847 describes buffers attached to the ends of cars at or near the top as distinguished from platform buffers, or buffers lower down on the car ends, so that in case of shocks or collisions the top buffers may come in contact and press upon each other, and thereby prevent the force of the lower buffers from causing the cars to rise and fall upon or override each other. The idea of these top buffers being in contact under spring pressure to prevent or diminish oscillation, is nowhere suggested in the specifications. It is only in case of collisions and sudden concussions that they are forced into contact. In order to steady the car bodies, and prevent them from swaying under ordinary conditions, the faces of these elevated buffers should be in frictional contact, which is not the case. It is true, the patent speaks of a means of diminishing oscillation, but it is in connection with an improvement for the purpose of traction or propulsion. In explaining the advantage of connecting the carriages for traction at two points instead of one, as was then usual, the specifications say:

“It is obvious that when carriages in a train moving at high velocity are attached to each other only at one point in the center, as at present is the practice, they must have a separate tendency (especially the last carriages) to oscillate or rock from rail to rail if the slightest obstacle be presented on the rail to either of the wheels on the side of the flange. It is also obvious that, if the carriages are attached at two points, as above provided for, the tendency to oscillate would be counteracted, if not obviated entirely.”

The top buffers described in this patent are shown in sections of spheres, for which reason, as well as others already mentioned, they are not capable of diminishing oscillation. The English patent granted to Dyer in 1864 shows elevated buffers, intended “to alter or change the 421line of concussion, in order to lessen the tendency of the carriages to overturn, or rise on end, in the event of collision.” These buffers, like the Symons top buffers, are not in contact under ordinary conditions, and, in view of the office which they are expected to perform, it is not necessary that they should be. It is only in case of collisions, or violent, sudden shocks, that they are brought into play. The means designed to prevent oscillation are thus described:

“I propose to place the wheels at the sides of the carriages, to bring the center of gravity within or between the wheels, and thus prevent oscillation, and give increased steadiness and security to the carriage. I propose to make the wheels revolve on their axles, instead of being fixtures thereto, so as to prevent the danger attendant on traveling round curves, and also to lessen the injurious results of oscillation upon the nervous system of passengers. I also propose to increase the diameter of the wheels to about five feet, or about equal to the base or breadth of the carriage, or less, as may be required, to resist the overhanging weight and tendency to overturn, and to give greater steadiness to the motion of the carriage generally.”

If the elevated buffers were intended to be in contact as a means of preventing oscillation, and if they are in fact so shown in the patent and drawings, why did the patentee describe other and different means for that purpose without mentioning the elevated buffers in the same connection? But, treating them as in contact, they are not under such spring pressure as to produce friction which will resist the tendency of cars to sway to any appreciable extent, as the Sessions buffers do. The English Garvey patent of 1852 is “for more effectually dissipating the shock of collision in railway trains, reducing the surface exposed to atmospheric resistance, and diminishing oscillation.” The means employed by the patentee for accomplishing the last-named purpose are-described in the specifications to be shields or frames attached to and covering the end of the cars, which frames are supported on the ends of four iron rods, attached rigidly to the car body, at or near its four corners. The shields are attached to the ends of the rods by means of volute springs. They are light, and covered with felt, and are forced into close contact when the cars are screwed together into a train by means of draw-links. The shields are capable, however, of moving vertically or “laterally, independently of the rods” which support them, and they would therefore be worthless as a means of resisting the tendency of American cars to oscillate. Whatever merit this device has in diminishing lateral movements of cars consists in the pressure of the volute springs. In speaking of the action of these shields, the specifications say:

“As these surfaces are covered with thick and strong felt, they cohere sufficiently to prevent them from sliding over each other, whilst the shields, possessing a universal mobility, will remain in close contact with each other, whatever may be the position of the carriages, whether they are turning a curve, or passing over a rise, or running on a line.”

Although this device may somewhat diminish oscillation, not by the friction of the shields in contact, but by the force of the volute springs, it is different in principle from the means employed by Sessions 422for accomplishing the same purpose; namely, frictional resistance afforded by two vertical frame-plates in contact under constant opposing spring: pressure capable of moving on each other, and yet resisting the tendency to move, and so attached to the car bodies that the force of the platform buffers is effectively transmitted along the plates, even to their tops; The Waller provisional specification of 1871 is relied on as a publication. The means employed, “by which, in case of a collision taking place, or a tire being broken; the injury to passengers is greatly diminished” are as follows: The entire ends of the cars are covered with a sheet of rubber one inch thick, which, in turn, is covered with sheet iron, three-eighths of an inch thick, upon which fourteen, or more, rubber sockets, each one foot, square and one inch thick, are placed to receive an equal number of spiral flat springs, two feet long when not compressed India-rubber sockets cover the outer ends of these springs, which sockets are covered with sheet-iron plates, tied together by flat strips of metal. Over all this is placed another sheet of India-rubber, half an inch thick. “The carriages,” say the specifications, “constructed in this way, when coupled, touch one another on a surface of not less than 30: square feet, the springs being then reduced to one foot nine inches in length, and the whole train becomes a solid and flexible mass, so that it is impossible for one buffer not to act upon another. It will be remarked that I never have two ordinary surfaces touching each other, which is an important feature.” There is a wide difference in organization and mode of operation between this mass of compressed rubber and springs attached to the ends of railway cars, and the Sessions device, as already explained. The English patent of 1845 to Fuller describes elastic cushions, or leather or rubber sacks filled with wool, or other flexible material, and interposed between the ends of cars, so that in case of collision, the cushions may diminish the effect of the concussion. The cushions are made on strong wooden, or metal, frames attached vertically to the ends of the cars, and backed by strong springs on sliding rods attached to the cars at the top and bottom by long iron sockets. These cushions, which act as buffers, are intended to be used with or without the ordinary buffers. If these cushion buffers are shown in contact, under ordinary conditions, they do not operate as frictional surfaces, like the Sessions frame-plates. The principle upon which the two organizations act is not the same. The English Bessemer patent of 1847 is, in part, for a hood-like organization to close the open space between the ends of cars in a train; and thereby avoid the resistance of atmospheric pressure when the cars are running at a high rate of speed. The hood is described as similar to the hood of the ordinary road carriage. The frame-plate or bow is pivoted at the bottom to the buffers, moves with them, and is forced out at the top by a spring. The device is somewhat difficult to understand, but it is so constructed and provided with hinges, springs, pins, rods, cranks, and levers as to be capable of being expanded and folded back. In one form it is described as attached to a wooden structure built out from the car equal to half the space between the cars in a train. Frictional resistance, as a means 423of opposing the tendency of the car bodies to sway, was not contemplated by Bessemer, and his device, which never went into practical use, so far as the evidence shows, is not capable of preventing lateral movements of the cars. His aim was to inclose the space between the ends of the cars in a train with a hood, so constructed that it would adjust itself to the movement of the cars and keep the space inclosed, and thereby prevent the pressure of the atmosphere from impeding the progress of trains. In mode of operation it is unlike the Sessions improvement. The patent granted by the United States in 1852 to Smith was for ad? justable flexible hoods so constructed and attached to the ends of cars as “to afford ready and convenient means of passage from one car to the other without danger to passengers.” Instead of using a single frame, or carriage bow, as Bessemer does, Smith uses several, which he unites by cloth, or other thin material. His hoods, which were composed of flexible strong material, are supported and held in position by light ribs or frame pieces. His forward, or front bows, come together to prevent the entrance of cinders and dust, and they are covered with packing. It is plain that the faces of these bows or plates, thus brought into contact, are not intended to counteract or diminish the lateral movement of cars. Frictional resistance to the tendency of cars to sway is not contemplated. In speaking of the action of the hoods and their ends the specifications say:

“They can freely slide against each other and accommodate themselves to the vibrations or rockings of the car, and that in a perfectly independent manner, and need no other fastenings. One end of the hood is permanently secured to the end of a car, while the other end is free to move in all directions.”

It is undoubtedly true that in some of its elements the Smith hood is like the Sessions organization. Smith's patent describes bows or frames in contact between the superstructures of cars under spring pressure, but they are free to move in all directions, and do not therefore resist the tendency of the car bodies to sway. The English patent granted to Rock Chidley in 1865, like the Smith and other patents of the same class, is for a means of inclosing the space between the ends of cars in a train. A single hood is attached to the end of a car, and its construction is such that it is capable of being extended or expanded so as to reach the end of the adjacent car. The frame-plate is attached rigidly to the car and supported upon heavy iron rods, and the hood is extended by spiral springs which force the frame plates into contact with the adjacent car. “Above the platform,” says the specification, “I provide a sliding frame over which is stretched a hood or covering, of water proof material. This frame is kept expanded by springs, but when pressure is applied to it, by the ends of adjoining carriages approaching each other, it will partially collapse, like a carriage hood, and still afford protection from the weather.” Chidley's hood, as it is described in the patent, is so, constructed and attached to cars that “when a number are brought together they will form a saloon, the whole length of the train, and thus afford easy communication throughout the whole length of the train and, afford 424the passengers protection from the weather.” His organization is provided with springs strong enough to expand the hood and cause its face, or frame, to bear against the end of the adjacent car. The idea of diminishing oscillation in this way never occurred to Chidley, and it is “plain that his device is useless for that purpose. He had in mind and described a flexible hood, composed of cloth, leather, or rubber, supported by a frame capable of collapsing in order that the device should yield freely to the movement of trains. It is urged that this patent shows a spring strong enough to keep the frame-plate in contact with the end of the adjacent car, and if these springs are not as strong as the Sessions springs, the difference is one of degree only. It is not material, in the operation of this device, that the face-plate be kept in frictional contact with the car end; and, if it is shown in contact, it is by no means clear that oscillation can be thereby diminished. The patent does not show two frame-plates in frictional contact under opposing spring pressure, and the difference between the Chidley and the Sessions device is not in degree, but in kind. Previous to 1873, the Chicago, Burlington & Quincy Railroad Company attached to the ends of its cars, or some of them, canopy covers to prevent cinders from the locomotives falling upon the car platforms. None of these attachments have been used since the last-named date, when they appear to have been discarded as useless. The defendants insist that these hoods operated as spring buffers, and fully anticipated the Sessions patent. Four inch iron rods, inclosed by springs, were attached rigidly at the tops of the car bodies. Two of the rods supported half of the canopy frame, or board, which was one inch or more thick, faced with sheet or thin iron, and hinged in the middle. The frame-work was forced out by the springs, and when the cars were coupled the springs were compressed and the faces of the boards, or ends of the hood-frames, were brought into contact, This rude device was not capable of diminishing oscillation.

It is also insisted that the Sessions patent is void for want of a definite or accurate specification; that it claims three different forms of buffers, and shows but one in the drawings, and that the strength of the top springs is nowhere stated. Treating the drawings as part of the specifications, the first claim covers two forms of buffers, the frame-plates used by both the plaintiff and the defendants, and their equivalent series of buffers, and the second claim covers buffers under similar pressure, located at the top or ends of the care. A skilled mechanic, with the drawings and specifications before him, would have no trouble in understanding how to make the buffers covered by the two claims, and Sessions was not obliged to limit himself to top springs of any particular strength. He described springs powerful enough to accomplish a result, namely, springs of sufficient strength to diminish the force of shocks, and furnish frictional resistance to counteract in whole or in part the forces which produce vibration or oscillation of the car-bodies.

It is further insisted by the defendants that they do not infringe, because they use an elliptic spring to force out the upper ends of the buffer-plates, instead of the coiled springs described in the Sessions patent. 425Sessions did not limit his invention to the use of coiled springs, and the use of the elliptic spring was contemplated by the patent. In all other respects, the defendants' device corresponds exactly with the patented device, and the substitution of an elliptic spring for a coiled one was purely a mechanical change. Again, the defendants' counsel insist that friction does not depend at all upon the extent of surface, but solely upon the force with which the surfaces are compressed, and that, therefore, the frictional resistance of the Sessions frame-plates in contact is precisely what it would be if the surfaces were larger or smaller under the same degree of pressure. While that may be true, and doubtless is, when the surfaces in contact are perfectly smooth, it is equally true that in proportion as the surfaces are roughened, the frictional resistance increases. It may be admitted that, if the buffer spring-plates were not forced together at the top as shown in the patent, the platform springs would exert little, if any, influence at the upper end of the frame-plates. With the buffer-plates rigidly attached to the top ends of the cars, and held in a substantially vertical line, without capacity to move sidewise independently of the car bodies, the pressure of the platform springs against the foot of the plates must necessarily be transmitted, to a greater or less degree, along the plates to their very top. If it were not so; the Sessions device would, indeed, be a worthless incumbrance, and the defendants would have abandoned it as promptly as they adopted it without right. In this respect the Sessions organization differs from all prior buffing devices. The record fails to show elevated spring buffers, or frame-plates, co-operating with platform buffer-springs, prior to the Sessions invention, It required more than mere mechanical skill to see that the pressure of the platform buffer-springs could be made effective in vertical lines between the superstructures of the cars, as well as in the longitudinal lines of the platforms. Sessions discovered a means whereby our long, high, American cars might be made to run as steadily as the low, short, English cars, and the fact that for years his now simple device occurred to none of our many car builders is a circumstance strongly favoring the claim that his invention possesses novelty. If his device, or anything operating like it, and capable bf producing the same useful results, was known in the prior art, it is remarkable that its practical utility was not sooner recognized and understood. All prior buffing structures lacked what was necessary to give them the effective force that the Sessions elevated spring buffer-plates are capable of exerting. We have already seen that the value of the Sessions improvement as a means of diminishing the force of shocks and counteracting the tendency of cars to sway when in motion was promptly recognized by the principal railroads of the country, and, while utility is not conclusive proof of invention, it is strongly suggestive of it. Owing to the differences already alluded to between American and English cars, there was greater necessity for additional means of steadying cars of the former class than of the latter, and yet no one suggested the elevated spring buffer-plates. The defendants are at liberty to use the vestibule structure without the Sessions invention, as well as all the various prior buffing devices, whether 426described in patents or not, and yet they persist in asserting their right to use the Sessions buffer-plates while denouncing them as worthless. If they are sincere in thus characterizing this improvement, why do they squander money in attaching it; to their cars? Practical railroad men do not adopt and use devices that are of no value. The frame-plates used by the defendants are covered by both claims of the patent, and an injunction will issue, as prayed for in the bill, and the case will go to a master, to take testimony and report the profits and damages.

This volume of American Law was transcribed for use on the Internet
through a contribution from Google.