We pass the question whether, when suit is brought upon two patents in one bill, the earlier may be taken as an anticipation of the later. For the sake of argument we shall assume that this is permissible. Nevertheless we think that the issue of invalidity is not so clear in the case at bar as to deprive the plaintiff of any opportunity to try its case.
The structure of the two condensers is concededly in some respects different. The terminals of Van Deventer are the tubular rivet and the lug in contact with the outer easing. Those of Dubilier are the two tubular rivets themselves. The question is whether these acknowledged differences in structure would upon any evidence make Van Deventer’s disclosure a patentable improvement over Dubilier’s. Verbally his claims in suit, 5 and 6, cannot be made to read on the earlier disclosure. We must therefore say, if we are to affirm the decree, that the changes are such as must inevitably be within the ingenuity of the ordinary artisan. However tempting it may be so to conclude by mere inspection of the two condensers, it seems to us too hazardous.
Van Deventer’s condenser may, if one choose, be regarded as no more than a more compact form of Dubilier’s. He left out one of the rivets and used the casing as a terminal of opposite polarity. Nevertheless this required an internal rearrangement of the condenser itself, and, while his notion and its execution may seem obvious, we cannot say that it was inevitably within the powers of the ordinary journeyman. Suppose the proof should be that Dubilier’s condenser was at once superseded by Van Deventer’s, because the art especially needed a condenser of as small size as possible. Must we say that the man who eliminated one rivet and used the easing in its place made no invention? In all *725inventions the safest test is the condition of the art before and after the putative invention, appears. At least that is an immeasurably safer test, when available, than any á priori conclusions as to what is or is not an obvious step. To the last we should not resort, except in cases of absolute necessity.
We do not mean even remotely to indicate that the claims in suit are valid. That cannot be told until the proof is in, but it is only in the plainest eases that the invalidity of a patent can he ascertained merely on its face. International Mausoleum Co. v. Sievert, 213 F. 225 (C. C. A. 6); Bonnie-B. Co. v. Giguet (D. C.) 269 F. 272, affirmed 269 F. 1021 (C. C. A. 2); Bayley v. Blumberg, 254 F. 696 (C. C. A. 2); Scott v. Aristo (D. C.) 266 F. 382. While it is true that we have here, not only such common knowledge as we may take judicial notice of, but a part of the prior art, the case seems to us to fall within the principle of the cases just cited.
Decree reversed, and cause remanded, with instructions that the defendant shall answer the bill.