ant, is shown in the following drawings and description thereof, taken from the specifications and claims found in the patent:

"Referring to the drawings, R represents a radiator of any heating system, to which leads a steam-supply pipe 1 and from which leads the delivery pipe, 2. Interposed at the discharge end of the radiator, between the latter and the pipe, 2, is my attachment or valve. The latter comprises an outer valve-casing, 3, to the upper open end of which is secured a screw-cap, 4, from the center of the inner face of whose top depends a guide-tube, 5, designated to receive the central tubular stem, 6, of a cup or float, 7, adapted to be normally seated about the edge of the opening, 8, formed at the bottom of the casing. The bottom of the hollow stem is provided with an opening, 9, for the free passage of the currents, as seen by arrows in Fig. 2.

"Under ordinary circumstances the water of condensation incidental to the operation of the system will mechanically lodge and in a short time fill the cup up to the lower edge of the tube, 5, the water in the casing corresponding to the same level. The condensation-water by thus filling the cup to the lower edge of the tube, 5, acts as a seal against a too-violent draining of the radiator under the action of the vacuum-generator (not shown) to which the pipe, 2, leads, It being understood that the said condensation water is maintained at this level

in the cup, any excess being carried off mechanically by the currents induced as a result of the vacuum or rarefaction referred to. The water thus car ried off passes from the radiator into the cup, thence upward through the space between the stem, 6, and guide-tube, 5, and thence down through the stem, 6, and opening, 9, into the pipe, 2. Should the condensation take place more rapidly than the currents could carry off the same mechanically, the water which would thus accumulate around the cup would raise the latter of its seat, permitting the water to discharge through the opening, 8, into the pipe, 2.

"It is to be understood that the present device may be applied to any vacuumheating system similar to that set forth in the aforesaid patent.

"A review of the foregoing discloses the fact that the openings, 8 and 9, both communicate with the common outlet for the casing, that the opening 9 is located in the path of the opening 8 and registers with it, and that the open ing 9 is always open to the pipe 2 and to the radiator, R. "Having described my invention, what I claim is:

"1. A radiator attachment comprising a casing having a suitable inlet-open ing, an inner tube depending from the top of the casing, an outlet-opening a the base of the casing located below the tube, an exit-pipe in communication with said outlet-opening, a cup or float surrounding the tube and adapted to seat over the opening at the base of the casing, a tubular stem open at both ends carried by the float and located within the tube and spaced therefrom thereby affording communication by the passage thus formed, between th radiator and the exit-pipe, substantially as set forth.

"2. A radiator attachment comprising a casing having a suitable inlet-open ing, a screw-cap covering the top of the casing, an inner tube depending frou the center of the cap, an outlet opening at the base of the casing located belov the tube, an exit-pipe in communication with said outlet-opening, a cup or floa surrounding the tube and adapted to seat over the opening at the base of the casing, a tubular stem open at both ends carried by the float and located withi the tube, and spaced therefrom, thereby affording communication by the pas sage thus formed, between the radiator and the exit-pipe, substantially as se forth."

It is not contended here that the valve described in the Wade pat ent is an infringement upon the complainant's device. It is insisted however, that the defendant has made additions in the Wade valv which entirely change its character, and that these changes convert th cup in defendant's valve into a pressure-motor. The valve complaine of is shown in the following drawing:

It will be seen by comparing it with the drawing in the Wade patent that the changes consist in the additions of the depending wall, 10, on the cap and the extension of the rim of the cup, 12, within that wall and form the annular space, 13.

The conditions which must exist on the inlet side of the valve in complainant's patent to effect the necessary differential pressure by which the pressure-motor is moved to raise the valve piece are "the pressure or the presence of steam or water." Either of these conditions will result in a working differential. The condition, however, which is of chief importance and which is responsible for the differential under which the motor responds, is the accumulation of water of condensation on the inlet side of the valve. When such a condition obtains, the vent of the motor by such condition is controlled through the instrumentality of the float, m, shown on the drawing, or its equivalent.

This float consists of one of the devices controlled by the conditions existing in the inlet side of the valve, or, in the language of the patentee:

"The operation of the valve-actuating-motor will be operated by the differecce in pressure in the returns and in the inlet side or radiator, while that difference in pressure will be controlled by the devices controlled by the conditions existing in the inlet side."

It will thus be seen that the device which is controlled by the conditions on the inlet side of the valve is the float, m, indicated on the drawing, and the means controlled by the conditions referred to in claims 2 and 7 of the patent being sued for refer to the float or any other means which, acting under the influence of the conditions on the inlet side, will secure the effective sealing of the vent to enable the pressure-motor to lift the valve piece. In other words, when the vent, 1, is closed by the valve on the float so that all communication between opposite sides of the motor are cut off, it allows an increase of the steam pressure to accumulate below the motor in excess of the pressure above, and this excess or differential serves to move the motor and raises the valve piece, marked "d" on the drawing. It will be noticed that this motor, or diaphragm, is "actuated by steam pressure," and that the only function performed by the water is to control the float, or equivalent device, by which, in turn, "the difference in pressure will be controlled." In the defendant's patent the patentee's conception of his invention undoubtedly was that his cup was a float; and in this we think he was right. However that may be, we do not see upon what theory it can be said that this patent infringes complainant's patent, for as was said by the Supreme Court in Stow v. Chicago, 104 U. S. 547, 26 L. Ed. 816:

"A patentee who is the first to make an invention is entitled to his claim for all the uses and advantages which belong to it."

And it was further said in that case that it was immaterial whether he perceived and stated such advantages in his patent. But, in the view we have taken of this case, it is not necessary to consider that question. It is contended by the complainant that the valve described in the Wade patent was purely a float valve and operated hydrostatically by flotation alone, and as such was inoperative in a vacuum system be

cause of the suction in the return which prevented it from floating; that the defendant then made the changes complained of, and thereby changed the valve from a float to a pressure-motor. And, in support of this contention, some of the witnesses for complainant testified that, when steam is admitted to the radiator through the inlet pipe, the air passes out of the radiator and into the valve. As the air passes out, the steam is allowed to enter. The air passes into the body of the valve through the inlet, and then between the inside of the body of the valve and the outside of the cup up through the small space between the upper part of the cup and the lower part of the cap, then downward into the cup and upward between the outside of the tube 6 shown on the drawing and the inside of the tube 5. The air then passes down through the inside of the tubular stem, 6, and out through the orifice, 9, into the return pipe. As the air passes out, the steam continues to enter, and the air passes out and the steam enters, because the pressure in the return pipe is less than the pressure of the steam in the supply pipe. There is less pressure in the return pipe than there is in the supply pipe. After a certain amount of steam has entered the radiator, the steam commences to condense, and the water, due to condensation, falls to the bottom of the radiator and flows towards the valve on the outlet end of the radiator; as the water accumulates in the body of the valve it accumulates around the cup, until finally the water rises and closes the small space between the upper part of the cup and lower part of the cap, and thereby cuts off communication between the interior of the valve-body above the cup and the space between the valvebody outside of the cup and communicating with the inlet side of the valve-body. When this sealing action takes place, the suction from the returns, acting through the small aperture in the valve piece, reduces the pressure on the upper or inner side of the cup, or float, relatively to the pressure which exists outside of the cup and in the body portion of the valve casing on the inlet side. The action of this is to produce an effective differential in pressure to act upon the cup, there being a less pressure on the inside than upon the outside, with the result that the cup, or float, is forced to rise and carry with it the valve piece to permit a free discharge of the water of condensation from the body into the valve orifice, 8, shown on the drawing, into the returns. This operation, it is contended, results in the discharge of the water of condensation, and breaks the water seal between the upper part of the cup and the lower part of the cap, and permits the cup to descend to its seat. We cannot accept this theory; for the word "sealed," as here used, can mean nothing more than that the water flows into the space around the cup until it reaches the space between the upper part of the cup and the lower part of the cap, and it can hardly be said that water seals an opening against the passage of water itself, or as Prof. Woodward, in his testimony, puts it:

"The very entrance of the water into that space shows that the water can enter the space, and therefore the opening is not sealed; so I regard the use of the word 'sealing,' at that point, with reference to the filling of that space. whether the space be large or small, as inappropriate and improper.”

When the water of condensation rises above the top of the inlet pipe, the inlet pipe may properly be said to be sealed, because the steam

in the radiator cannot then enter the casing of the valve. In other words, the water closes the opening against the passage of steam. In the complainant's patent the sealing takes place, not only when the inlet pipe is wholly under water, but the passageway through the diaphragm is sealed by a valve. The difference is we think manifest. The sealing which takes place in complainant's valve and causes it to operate as a motor is when the vent is closed by the valve on the float, and this can never take place in the defendant's valve either in the form shown in the drawing of the valve complained of, or in any form of defendant's valve offered in evidence. In all of defendant's valves only one kind of sealing takes place, and that is when the water of condensation rises above the top of the inlet pipe. It follows, therefore, as it seems to us, that this water performs but one function and one object—that is, to float the cup-and that it cannot be said that the cup in defendant's valve is in any sense the equivalent of the complainant's valve. In complainant's valve the vent can never be omitted, for without it the device would become inoperative, the vent being a part of the pressure-motor, so that whatever be the means controlled by the conditions these means must be either the float or some equivalent of the float. Fay v. Cordesman, 109 U. S. 408, 3 Sup. Ct. 236, 27 L. Ed. 979. As stated by the patentee:

"The preferred means for controlling the operation of the pressure-motor consists of a float controlled by the water of condensation on the inlet side and ontrolling a vent in the pressure-motor, by which the differential pressure and the operation of the motor are regulated."

It is undoubtedly true that if any considerable vacuum existed in the discharge pipe immediately below the opening, 8, shown on the drawing, that the cup in defendant's device would not rise, but there is testimony in the case to the effect that there was almost continuously a steam jet escaping through the orifice, 9, as indicated on the drawg, and that the pressure from this steam jet at the point of impact, reacting from the walls of the discharge pipe, increased the pressure in such a way that the cup would lift with slight assistance from the water surrounding the cup in the body of the valve. Prof. Woodward testifies from actual experiment that the cup did operate between the steam pressures in the radiator on one side and the vacuum pipe on the ther, and that the cup operated irrespective of the diameter of the cup and irrespective of the fact that it did or did not reach the flange, 10, connected with the cap, shown in the drawing. If this theory be correct, then the relations between the steam pressures in the radiator, the pressure within and above the cup and the pressure in the outlet cipe are not important, since the cup operates, not by reason of them, but in spite of them, and whatever these pressures may be, it would not take away the cup's capacity to float. Neither is it important for s to determine how much the pressure within and above the cup may be less than the pressure of the steam in the radiator, because to the extent that it is less it becomes supplemented by the hydrostatic head around the cup, so that this head, together with the pressure within and above the cup, would always balance the steam pressure in the radiator; otherwise, the water would run over the edge of the cup and ink it. These forces being equalized, nothing further remains except