Although at trial there was contest on the question of infringement, even though validity were established, and the findings on that head are assigned for error, defendant’s argument in this court has been confined to validity, and that alone shall we consider, merely noting that we regard the matter of infringement as quite clear and properly treated below.

The economic advantage sought by the patent is to use no more gas in a storage water heater than enough to get the water to a predetermined temperature, then reduce the heating flame to a pilot light, and when the temperature falls to the lower predetermined figure turn up the flame again. The general method of doing this is by thermostatic regulation, which in some form is older than the patent. The new result said to be attained is the immediate reduc*861tion, and immediate restoration of the heating flame, and the element in the claim constituting the means of accomplishing this result is the “lost motion connection between the thermostat and valve” controlling the fuel gas supply. The apparatus to which Ruud applies these words needs no other description than the subjoined patent drawing.

It is contended that- the phrase “lost motion” is confusing, and without exact meaning. We think not; the-words are defined in lexicons of general acceptance as “any difference of motion between the parts of a machine that communicate motion from one to another” a meaning harmonious with the finding below that a lost motion connection is shown between gas valve and thermostat, because the thermostatic action is not transmitted to- the valve for a considerable period; i. e., until the spring (18) is passed over its fulcrum or center under tension, whereupon it instantly snaps the hinged arm (IT) to one end or the other of the latter’s traverse (80 and 81) and by so doing as instantly opens or closes the gas port. It is obvious that the disclosed device uses no power other than or extraneous to' that, of the thermostat. Mechanically we agree with appellants that this part of the apparatus does not differ in principle from the disclosure of patent 542,708 to Johnson for a “mechanical movement,” and indeed in some form the “snap action” resulting from “upsetting” a spring under tension is well known.

[1] Objection to validity may be stated under several heads: (1) The prior art tells so much concerning what Ruud claims that inven-don is as matter of fact absent. (2) A water heating apparatus installed in the house of one Barton five years before Ruud filed application is a complete anticipation.. (3) The course of the application through the Patent Office proves that the claim in suit was intended to cover just what was done in the Barton house, and the claim is therefore void.

1. By pointing out that storage water heaters1 with thermostatic regulation of heat, and devices for translating into quick or “snap” action slow initiating movements,2 are old, and then combining them on paper, appellants have ingeniously designed a machine which looks as if it would operate as does Ruud’s.

Attempted demonstrations of this kind are not uncommon and they always raise the question: If the patentee, whose achievement looks so simple, had not done the thing first, would the alleged in-*862fringer be able even to do it on paper? Invention is neither disclosed nor disproved by such methods; it is a question of fact, to be settled by trying to find out how much thought and skill was required to “do the trick” at the time it was done. We find the direct application of the thermostat as a slow initiating force to a snap action instantaneously and directly operating on the fuel valve an extremely happy thought, simply solving for many purposes a very real problem. The solution of an actual difficulty, simply, cheaply, practically, and along sound mechanical lines, is not a definition of invention, but is a description of many excellent ones.

2. The Barton installation seems to' us perfectly to justify Ruud’s patent rather than show its invalidity. Companies affiliated with the same Johnson whose mechanical movement patent has been noted put in the heater, which utilized for gas regulation a compressed air system which the Johnson concern quite widely used for heat regulation. In this instance a thermostat in the water storage tank gave an impluse to the compressed air (so to speak), which in turn actuated the gas valve, and we infer that such original impulse! consisted in some use of Johnson’s mechanical movement; i. e., it was a “snap action.”

But there was no direct action by thermostat on valve, nor was there any “lost motion” at all, as Ruud applies it. The Barton thermostat switched on a wholly extraneous power, from a different source and requiring its own system of generation, i. e., compressed air, and that turned the gas on or off, up o'r down. If the compressed air for any, reason was out of order, the mechanical connection (if there was one) between thermostat and gas port might be intact; but it was useless. There was no- anticipation by that heating apparatus.

[2] 3. Ruud’s disclosure was fully contained in his original specification. As the claims first submitted showed, he conceived himself entitled to claim broadly any means, if actuated by the thermostat, of fully opening and completely closing the controlling gas valve. He also suggested, at the end of the specification as it now stands, that his mechanism, instead of acting directly on the gas valve, might set in motion water under pressure, or electromotive force, wherewith to move the gas valve, with some advantageous results. All the claims submitted having been rejected on certain prior patents, and properly we think, and the applicant being called on to furnish drawings of his alternatives, he withdrew' the suggestions aforesaid and substituted (among others)- the claim in suit, which restricts invention to a combination containing the lost motion element.

It is now said that the claim must be read as asserting itself to cover the withdrawn suggestions of variant Structures, which suggestions would clearly have covered the Barton installation; wherefore the claim is void, and Engel v. Sinclair, 34 App. D. C. 212, is relied on. No doctrine is there announced, other than that an amended claim cannot be given a construction which makes it the equivalent of a rejected one. Such is not the case here. The alternative suggestions in the original specification, which would have covered the *863Barton use, were consonant with the claims first propounded; they could not be covered by the claim in suit. It may be that the matter ■was withdrawn because Ruud did not care enough for it to furnish drawings; but the, fact remains that it was withdrawn when the rejection of claims that were appropriate thereto was acquiesced in. We see nothing in this file wrapper inconsistent with a proper trimming down of the specification to suit a much restricted claim, which these defendants infringe.

Decree affirmed, with costs.