DENISON, Circuit Judge.

This case comes in succession to the decision of this court between the same parties reported in 268 F. 201. The patent there involved and sustained was issued to Luther, and its claim was: "The process of making arsenate of lead, which consists in reacting upon lead oxide held in suspension in water with arsenic acid in the presence of a catalytic agent; the lead oxide and arsenic acid being used in the proportion of their combining weights."

Pursuant to our opinion, the court below referred the case to the master. Defendant claimed that, as a standard of comparison, it could depend on the Schaeffer and the Hedenburg-Pratt patents, which it controlled and which it thought did not infringe. It also announced that it had changed its process, thereby getting outside the scope of the opinion of this court and of the accounting. Thereupon plaintiff filed a new bill — the present one — to reach this second process. Upon the hearing counsel presented to the court below two questions: (a) Did the Schaeffer and Hedenburg process infringe? (b) Did the defendant's second process infringe?

The court answered both these questions in the affirmative. On motion for rehearing, defendant presented yet a third process, said to have been used just before the hearing of the present suit and thereafter until the decision thereof, and made the claim that, as this third process did not infringe, comparative analysis showed that the second process did not. Plaintiff disputed this premise and this conclusion. The court denied the rehearing. The appeal presents these three questions.

It was the theory of the Luther patent that the action of the arsenic acid upon the litharge (lead oxide) held in suspension in water, which action tended to convert the oxide into arsenate, was too slow or inefficient to be commercially practicable, and that by adding to the compound a "catalytic agent" (a stable accelerator) the reaction became rapid and the liquid released by the precipitation of the arsenate attacked new litharge particles, thus setting up a "cyclic action." Luther in his specification proposed nitric acid as a suitable catalyst, and somewhere between 1 and 3 per cent. of the litharge as a suitable proportion. His claim does not expressly require that the catalyst should be an element physically added to the litharge and water, but we think that it is the plain and clear implication. He knew that the litharge and water unmodified and treated with the arsenic acid would in time produce some arsenate; he knew (because it was common knowledge) that under heat the reaction would be accelerated and more arsenate would be produced; but these results were not commercial. His discovery was that by adding a substantial amount of some selected other agent, he could speed up the process so as to have a factory instead of a laboratory.

In a general way Schaeffer and Hedenburg may be considered together. They claim to have discovered or observed that by substituting fumed litharge for the ground litharge, which Luther and defendant's first method had used, they would so increase the surface area of the litharge particles exposed to the liquid as to increase greatly the rapidity of the action, and that by also using heat they could get a commercial process. It is said that the area exposed was thus increased 50 or 100 times; the increase is conceded to be fourfold. Whether the reaction in the presence of this increased area and this heat alone would be speedy enough to make the product so rapidly that it could meet market competition, we need not decide; but that the process would be operative on a substantial scale is fairly clear. No catalyst being added, it seems not to infringe the Luther patent; but plaintiff's theory is that when the fumed litharge is suspended in water and exposed to continued heat, there develops in the liquid what defendant calls an electrolyte, but which plaintiff says is a catalytic agent, and in the presence of which the accelerated reaction takes place. It would seem that sometimes, but not always, an electrolyte is a catalyst, and that sometimes, but not always, a catalyst is an electrolyte; but we think it unnecessary to follow out these definitions; they are not material.

The total amount of this developed agent is very small, about one-tenth of 1 per cent.; but it is not necessarily to be disregarded for that reason. We have no doubt that the fuming of litharge and the heating of water were expedients of common knowledge in the pre-Luther art, and indeed were variant forms of the base upon which he rested his improvement; and we conclude that the Luther patent cannot extend to the mere use of a catalytic agent (if there is one) inevitably developed as incidental to a familiar manipulation of the old materials and as contrasted with the deliberate addition of a new element, in full catalytic action as soon as added. We find therefore that the practice of the Schaeffer and Hedenburg patents, according to their respective theories as set out in their specifications, is not an infringement of Luther. Of course, this conclusion does not necessarily indicate that these two patents show available standards of comparison in determining profits or damages; that question involves additional elements.

This conclusion also is confirmed, though not required, by the conduct of the former cases. Evidently it did not then occur to plaintiff that defendant's first process infringed even without the added catalyst; no such theory of the patent was suggested.

Infringement by the second process depends upon what is claimed to be an accidental or inadvertent use of an added catalyst. The arsenic acid which defendant was using was found to contain arsenious acid and sulphuric acid and or other impurities of catalytic capacity amounting to a total of .88 of 1 per cent. of the litharge volume. That this is added along with the arsenic acid, instead of being added to the water before this acid is put in, is plainly not important. Defendant did not buy this arsenic acid in the market as and for a pure article. It was the manufacturer. The process of manufacture employed (or the one which the testimony indicates was thought to be employed because it is the common one) was to take arsenious acid already prepared, the formula of which is As O₃ and treat it with nitric acid to produce oxidization — resulting in arsenic acid, the formula of which is As₂O₅. To insure the necessary oxidization it was necessary to use an excess of nitric acid, leaving probably a residue which, at the end of the process, must be eliminated. There were two recognized methods for this elimination. One was to employ a deliberate process for that purpose, involving time and expense. The other method was to use an excess of arsenious acid whereby there would be, in the ultimate arsenic acid, some of the original arsenious acid not combined and transformed.

The sulphuric acid was an impurity likely to remain in some degree in the nitric acid, resulting from the process of its manufacture, and in that event it would be carried along into the arsenic acid.

It was not necessary that any of these impurities should be found, in more than a trace, in the arsenic acid; the fact that they were so found might be the result of intention or might be the result of inadvertence and a desire to save the expense of careful elimination.

The record is satisfactorily convincing that in the defendant's second process this .88 per cent. of catalytic agents performed a substantial function in accelerating the reaction. This was sufficient to make out infringement of the Luther process; and it becomes unimportant to determine whether the additional .15 per cent. of the so-called electrolytes were or were not of importance. The specification indicates a minimum of 1 per cent. for the added catalytic agent; but this is only Luther's idea at that time of the preferred form, and the claim contains no such limitation. It only provides that the reaction should be "in the presence of a catalytic agent," and this requirement is met if the agent is present in sufficient quantity to have a substantial effect. If Luther made the useful discovery which the former opinion attributed to him, and if the additional refinements of treatment which experience developed enabled him to get along with a smaller amount of the catalytic agent than he had supposed necessary, nothing except an express limitation in his claim should deprive him of the benefit of these improvements in the process.

We think it unnecessary to decide whether the presence of these impurities should be attributed to inadvertence, as defendant claims, or to what was at least a willing carelessness, as plaintiff claims. Defendant was under obligation, created by the patent and emphasized by the decree of this court, to see to it that no added catalytic agent crept in. It was bound to know that these agents were in the neighborhood and must be kept out.

It is said that in the third process defendant used an arsenic acid chemically pure (as near as may be) and employed no catalytic agent whatever, unless by the use of the so-called electrolytes which developed from the fumed litharge and water under heat. If this is true, the third process did not infringe.

The decree below found expressly that the Schaeffer and Hedenburg process infringed, and this involved a finding that defendant's third process was an infringement. In these respects the decree was erroneous; otherwise we think it was right. The decree as entered will be reversed and the case remanded for the entry of a new decree pursuant hereto.