(after stating the facts as above). The patent can succeed over Conover only by virtue of the position of the screen and of the wheels. To place the screen between the nozzle and the suction device was old in other forms of the art. It existed in Kenney, in C. J. Harvey, and J. J. Harvey. The only color for invention in this particular arises from the fact that these latter are not hand-driven machines, deriving their power from traction wheels, upon which their weight rests. As a new question, and without its subsequent success, we should have no question that the position of the screen on the near side of the bellows was not invention, in the face of the disclosures in this case. Indeed, we should have great doubt whether, without those disclosures, it would be invention to screen out the dust before it reaches the bellows, rather than to let it pass through and clog the valves. We find it difficult to look at this feature, except as one which the inevitable course of experience in the art would have soon suggested, if it has not been done at the outset, and as one which, once suggested, required no mechanical invention to adapt to a hand-driven carpet sweeper.
[1] The plaintiff must rely, therefore, and, indeed, it does largely rely, upon the success which its sweeper has attained, and we are forced, therefore, to a consideration of the propriety of that test here. It is true that the books are full of cases in which courts have regarded the success of the plaintiff’s patent as an important test of invention, and we are in no sense disposed to question its value in proper cases. Yet it is a hazardous rule, and one which is quite likely to result in confusing genuine invention with imagination in advertising and energy and business skill in promotion. Where the art presents a case of earlier efforts, unsuccessful because of the absence of what the patentee contributed, and followed by a wide success after that contribution was added, it is reasonable to infer that the art needed that feature, and that it was not so easy to invent as might seem to us, who necessarily have no proficiency in the art.
It by no means follows that every successful exploitation of a patent complies with these conditions; there are many other reasons for success, which need not be. detailed. In the case at bar three companies commenced the manufacture of these cleaners, all substantially similar in kind, and each competed with the other two. The Domestic Company commenced selling them as early as 1910, and its sales rapidly increased. The Wright Company began its sales early in 1911, and likewise quickly became successful. The National Company began at some time not definitely stated, before which and in 1908 Baender had • made some 2,000 machines — at least, such is the assertion. The mutual competition of these companies compelled a consolidation of their interests, which occurred in January, 1916. Now, we may assume for argument’s sake that each of these companies manufactured from the outset under each of the three patents, and that the credit of the invention is to be attributed equally to all; yet it is obviously quite impossible to say that the success was due to any one of the *199elements contributed by all. It is, indeed, impossible, upon this record, or upon all three records, for ns to say that the success was due to all three together, assuming that would serve if we did. We do not question that success awaited any successful hand-driven vacuum cleaner; but we must be able to attribute the success to the invention, and the dominant idea rested, as it seems to us, rather in the general conception of such an appliance, set forth certainly in Conover, which, if indeed it was as it stood, likely to become clogged in the valves, only required changes sure to be realized as the art progressed, and already disclosed in three varieties of existing cleaners of other species.
The “idea” of such a cleaner seems to have become fruitful at about this time in many instances: Buell, Dudley, Baender, Sturgeon, Apple-gate, and Quist, all between 1907 and July, 1910. Conover it is true, was earlier, his British patent dating from 1904; and if there was any basic disclosure it was this, upon these records. Yet, disregarding Conover, the art presents a picture suggestive rather of a commercial opportunity discovered than any necessary invention. Each of these patents would, so far as we can see, have answered as well as any other, with such modifications as would be sure to suggest themselves in the progress of the art. In such a case, we ought to scrutinize, not without some jealousy, the claims of any single contributor to a monopoly .of his own details.
The conclusion of the examiners in chief finally granting the patent scarcely justifies the assumption that the position of the screen weighed very much in their minds. They appear to have depended quite as much upon the distribution of the load between the wheel and the suction nozzle, an advantage which turned out to be of no moment, or perhaps upon the general compactness and advantage of the arrangement of the parts, which is not claimed, and hardly could be. Furthermore, the appearance in the art, even if it was in fact later than Baender’s invention, of the device of Ander’s British patent, suggests that there was no divination necessary to the location of the screen where Baender placed it.
We agree, therefore, with the District Court in holding the patent void for lack of invention, and the decree will be affirmed, with costs.