the same mode or principle of operation. The defendant's switch is made under a patent granted for an invention by Schmidt and Werner in 1903. Figure 1 is selected for explaining it.

At first sight it resembles the second Werner patented switch. But on analysis of its construction it is found to be an essentially different organization. There are two track sections, 16 and 17, having each a rail on its margin, but here the similitude to any new feature of Werner's ends. The rail sections of the defendant are not separate from and movable on each other, as in the Werner patent, but are cast in one piece rigidly on a common pivot around which both the switch plates revolve. Neither of the plates moves horizontally, as does the lower one in Werner's. A chamber is made below the lower plate, into which it drops on being brought backward and downward off the lugs which support it at either end. These lugs are beveled off on their inner faces so as to form an incline from top to bottom, down which the face or edge of the lower section slides. The operation by which the shifting of one rail section for the other is effected is peculiar. Slitted bearings are made in the casing at either end. These slitted bearings extend upward and backward, and in them are journaled the ends of the pivot, on which the rail sections are arranged at about a right angle. When the lower section is pulled backward and upward, the pivot rises in its bearings, the rail section is drawn off the lugs on which it normally rests, its face or edge slides down on the inclined inner face of the lugs into the chamber. At the same time the other rail section is coming forward and down until it rests on the same lugs. Thus the switching to one of the diverging rails is effected. Simply reversing the process, the upper section is thrown over and back to its former place, and the lower section comes up to its place and by the effect giv

en to the slotted bearings of the pivot, as the pivot descends therein the rail section is thrust forward into place as soon as it reaches the top of the lugs, and then rests thereon.

The mode of operation and the adaptation of the means to secure it in the Schmidt and Werner patent are so far different from that of the switches shown in the two Werner patents that it cannot be held to infringe them. The difference is in the vital part of the mechanism for shifting the rails which is the purpose of a switch. In the second Werner patent, the devices in which are most like those of the defendant's patent, no means for making the shift being contained in the second and fifth claims, if we should bring in by the reference at the end of the claims the devices pointed out in the specifications for that purpose, this would complete the claims. And this would be the basis. for comparison with the defendant's. The spur on the under side of the upper rail section and the opening in the lower rail section constitute the individual elements for making a shift of the rails. In other claims they are expressly included as elements of the combinations. The defendant's switch has no such elements. Moreover, the Werner patent has the ordinary simple bearings on which the pivot of the upper section rail revolves. The lower swings on a pivot at the base of one of the arms of the bracket. The defendants lodge both sections on the same principal pivot and provide slotted and curved bearings for the pivot which effect a peculiar mode of operating the rail sections not found in the Werner patent. And, further, the lugs with inclined inner faces which also affect the mode of operation in the defendant's device are not found in the complainant's. All of the claims of the complainant involved in the controversy are for combinations. The elements of those of the respective parties are not only unlike in some of their characteristics, but those of the complainant's are some of those omitted by defendant and some of those of the defendant which contribute to the operation of the switch are not found at all in the complainant's. It would be flying in the face of well-settled principles in the patent law to hold that there is infringement in such circumstances as these. Prouty v. Ruggles, 16 Pet. 335, 341, 10 L. Ed. 985; McClain v. Ortmayer, 141 U. S. 419, 425, 12 Sup. Ct. 76, 35 L. Ed. 800; Black Diamond Coal Co. v. Excelsior Coal Co., 156 U. S. 611, 617, 15 Sup. Ct. 482, 39 L. Ed. 553; Cimiotti Unhairing Co. v. American Fur Refining Co., 198 U. S. 399, 25 Sup. Ct. 697, 49 L. Ed. 1100.

The case seems to us an excellent illustration of the rule that when each of the two inventors improve upon the former art, each in his own distinct and separate way, they shall each be credited with his own improvement. It does not necessarily follow that, because the separate elements unaffected by their special adaptations might be found in both patents, the combinations are therefore the same. The observation of Mr. Justice Brown in Westinghouse v. Boyden Power Brake Co., 170 U. S. 537, 18 Sup. Ct. 707, 42 L. Ed. 1136, at page 568, is very pertinent. Of course, it would be likely that some of the necessary features of such an apparatus would coexist, but, as we have said, the elements of the combination do not operate in the same way, although they attain the same object. The second and fifth claims of the second Werner patent are very artfully framed, and seem to aim

at including more than the specifications warrant. It is only by restricting them by the reference to the specifications that they can be supported and as so construed the defendant's switch does not infringe either of them.

The decree of the Circuit Court will be affirmed, with costs.

GENERAL ELECTRIC CO. v. ALLIS-CHALMERS CO. et al.

(Circuit Court, D. New Jersey. May 26, 1909.)

1. PATENTS (§ 328*)-INFRINGEMENT-ATTACHMENT FOR MOTOR Controller. The Potter patent No. 671,232, for an attachment for notched quadrants, the purpose of which is to secure automatically a notch-to-notch movement of the handle of the contact-cylinder on an electric car, construed, and held not infringed.

[Ed. Note. For other cases, see Patents, Dec. Dig. § 328.*]

2. PATENTS (§ 312*)-INFRINGEMENT EVIDENCE.

The fact that a device is within the language of a claim of a patent does not necessarily prove infringement.

[Ed. Note. For other cases, see Patents, Dec. Dig. § 312.*]

In Equity. On final hearing.

L. F. H. Betts and Ramsay Hoguet, for complainant.

Thomas F. Sheridan and Clifton V. Edwards, for defendants.

CROSS, District Judge. There are two defendants in this suit, one only of which, the Allis-Chalmers Company, has been served. The bill of complaint alleges infringement of letters patent No. 671,232. for an "attachment for notched quadrants," issued to Wm. B. Potter, assignor to the General Electric Company, April 2, 1901.

The specifications define the purpose of the invention as follows: "This invention relates to controllers for electric motors; and it is especially intended for the large controllers used in electric railway equipments. In certain types of such controllers the handle by which the contact-cylinder is rotated is provided with a spring-latch which engages with notches in a quadrant secured upon the top of the controller-casing. The latch normally engages one of the notches, and can only be disengaged to release the handle by pressing down a thumb-piece in the handle. In operating this device it is necessary to keep the thumb-piece depressed until the latch has cleared the edge of the notch, the result being that the latch is sometimes retained so long as to skip the next notch.

"The object of my invention is to positively insure a notch-to-notch movement and simplify the operation necessary to obtain this result. I accomplish this by providing a spring-actuated auxiliary support movable independently of the quadrant, so that when the latch is raised said support instantly operates to hold it in this position above the teeth of the quadrant, irrespective of the pressure on the thumb-piece, so that the handle is positively released and is free to be turned if the motorman gives only a momentary pressure on the thumb-piece."

A controller is used on street cars to regulate the speed of the motors, and consequently of the cars. As this patent has nothing to do directly with the internal mechanism of the controller, it seems unnecessary to give a detailed description of its method of operation

For other cases see same topic & § NUMBER in Dec. & Am. Digs. 1907 to date, & Rep'r Indexes

farther than to say that it is essential that the motorman turn the controller drum to such a point that the proper switch connections will be made. For, as the counsel for the complainant in substance says, if it should be turned too far the contact fingers and segments within the drum might be separated far enough to break their actual contact, but not far enough to prevent the formation of an arc between the contact fingers and segments which would result in seriously burning them while at the same time the current might not be efficiently carried to the motors. Hence, in order that the motorman may stop the controller handle at the proper places to make the desired switch connections, the correct positions are marked on a dial plate under the controller handle, by the use of which dial the motorman may gradually move the controller notch by notch to the indicated positions. The object of the inventor of the patent in suit was to provide a handle for the controller "that would operate automatically to insure a notchto-notch movement." It was intended to provide a gradual movement of the controller handle, and prevent the operator from overrunning. the notches and turning the power on too abruptly and suddenly or at improper places. The first claim of the patent, and the only one in issue, is as follows:

(1) "In a controller, the combination with the handle of the contactcylinder, of a latch pivoted thereon, a notched rack coacting with said latch, and means for automatically supporting said latch when raised out of a notch, until the handle is turned to bring the latch over the next notch."

It will be seen that the claim embodies four elements in combination: The handle of the contact-cylinder, a latch pivoted thereon, a notched rack coacting with said latch, and means for automatically supporting said latch when raised out of a notch until the handle is turned to bring the latch over the next notch. The Priest patent No. 556,862, of the prior art, shows substantially the arrangement of the patent in suit, save for what it called the supplementary rack or auxiliary quadrant. Indeed, the complainant's expert practically admits this when he says:

"I do not believe that there is any material or substantial difference between the device of the Priest patent and the device of the patent in suit, aside from auxiliary quadrant and the means for operating it."

Whatever of invention exists, therefore, in the patent in suit, must be found in the "means" referred to in the fourth element in combination with the other elements. The main contest at the argument was upon the question of infringement. The validity of the complainant's patent was not seriously controverted. With the exception of the Priest patent, the patents cited as anticipations all belong to a more or less remote art and are incapable of adjustment to a controller, and no one of them, not even that of Priest, contains the last element of claim 1, and hence does not show the combination of that claim. But the same thing is true of the defendant's device, which likewise does not contain that element, or any equivalent thereof. The claim calls for means for automatically supporting the latch when lifted out of the notch; means which, upon pressure being exerted upon a button on the controller handle by the operator, operate instantly and independ

ently of his will. The "means" of the patent constitute devices entirely separate from the rack, while working in co-operation therewith. The rack is itself the second element of the claim, and cannot therefore be the means of the fourth element, which is an independent and additional element, and shown to be such by its separate description. The method of the patent for automatically supporting the latch is by the use of a supplementary rack or auxiliary quadrant. This, as already indicated, is not only an essential element of the claim, but is the only novel one. The patentee, while suggesting modifications and variations of the means therein referred to, nowhere attempts to get away from the auxiliary quadrant or something equivalent thereto pivoted or attached to the main quadrant. It is true that in the specifications he speaks of other and separate quadrants for series notches each having its own spring; also of a separate spring tooth or dog pivoted on the main quadrant adjacent to each notch; and again at another place says that, while the drawings show the invention applied to a curved rack or quadrant, it can within its scope also be applied to a straight rack, if desirable; but no matter what form or variation is suggested, it is always something outside and independent of the quadrant; that is, something which, as the specifications say, can be "applied" to it. For instance, at one place it is said:

"While the drawings show the invention applied to a curved rack or quadrant, yet it is manifestly within the scope of my invention to apply it to a straight rack, should it be desired so to do."

This language conveys a decided intimation that the patentee's entire invention exists outside of the rack or quadrant, to which, moreover, it is capable of application. The defendant's device does not have any supplemental rack or quadrant to support the latch. It has but one rack, and that is like Priest's in the prior art. The latch when lifted from the rack is supported manually and not automatically; of course it has to contact with the rack at some time and place, otherwise it could not take the next notch of the rack, but it is not supported by it in the way that the latch in the patent in suit is supported by the auxiliary rack or quadrant. The claim in issue provides means for supporting the latch when raised out of a notch, which means operate automatically and immediately once the latch is raised, and that, too, without any intervening movement of the controller handle. The specifications say that, "when the latch is lifted by a momentary pressure on the thumb-piece, the instant the latch clears the top of the auxiliary quadrant, the spring, I, throws this quadrant endwise so that its tooth, h, comes under the latch and prevents its falling." This is done, it will be noticed, automatically and not manually. The following question addressed to the defendant's expert and his answer thereto show the method of operation of the defendant's latch:

"Q. Please state exactly during what part of the operation of the controller handle the latch is out of contact with the rack in defendant's structure? A. As the handle moves to the rear, the lug at the outer end presses on the inner end of the latch and finally disengages it from the tooth. At this instant the latch is pulled forward and put into a position where it cannot enter the notch it has just left. The latch does not touch the rack at this time, but is at some distance from it. Being now in a position where