The District Court held claims 2 and 3 of patent No. 1,-371,572 valid and infringed. The claims are:

“2. A process of coloring leather, which consists in suspending pulverulent, insoluble pigment in a fluid and agitating said fluid and leather with respect to each other, so that the pigment is uniformly deposited in the interstices of the leather. '

“3. A process of coloring leather, which consists in' suspending pulverulent, insoluble pigment in water and agitating the leather therein in the presence of heat, so that the pigment is uniformly deposited in the interstices of the leather.” ’

This is a process patent for coloring leather splits with pigments. Witness Peterson had, to some extent, used pigments, prineipally chrome yellow, 20 years before, and when the price of anilin dyes became high during the World War he again took up the use of the pigments, in 1915. His description of his method of use is very like that of the patent. His testimony is not cont-rádieted. The witness Poll learned the method of using pigments from Peterson, and pracfaced the method from the early part of 1916. Martin, another witness, who showed Pfingsten how to use it, commenced the use of yellow pigment coloring in February, 1916.

Early in 1917, Pfingsten, the inventor and appellee’s assignor, without any previous knowledge of coloring by the use of pigments,' having seen some splits colored by the George Martin Leather Company, went to them for the purpose of buying 20 pounds of the coloring matter used by them. After making some inquiries, he was told to return the next day, and they would show him how it was used. That was done, and Pfingsten bought 25 pounds of the coloring matter, made some experiments with it in the plant where he was employed, and, on July 2, 1917, filed his application for patent.

The District Court sustained the claims, because it found an irreconcilable conflict between the testimony of Martin and that of Pfingsten, as clearly stated in its opinion:

“One of them testifies that in 1916 at his tannery he disclosed this identical process to the patentee. The patentee, on the other hand, admits his presence there, witnessing a demonstration, examining the product, but tells this further story: That in the practice of the process there was no water used, except such as left the hides in their moist condition, in which they would be left after draining off the so-called fat liquor; that the dry pigment was inserted at that time, and the drum revolved and the hide agitated, with the idea that thereby the dry pigment could be introduced into the hides, and that as the result of it the hides came out, and it was discovered that the coloring matter was upon the surface of the hides; that thereafter he made experiments of his own, and found that in practicing what he saw at the Martin tannery the result was a spottiness of hides, likewise .the deposit or presentation of the pigment upon the surface of the hide, which was discoverable upon staking the hide and using a certain machine for stretching it; that the result of those operations disclosed the scraping off of the' coloring matter from the surface of the hide; and from that he proceeded to the step of introducing a quantity of water, with the result that he found the hides were clear, and that is what his patent, his idea, is.”

Where the court has seen and heard the witnesses, and has given preference to the testimony of one or another, it is unusual to disturb the conclusion. However, in this case, the court found that the scales were evenly balanced and were only turned in Pfingsten’s- favor because “the place which the process has taken in the art strongly fortiñes the presumption of the inventive quality of his act.” If the process was disclosed by Martin, it would seem that the testimony of public approval should have gone to him.

We cannot agree that there is no preponderanee in the evidence. We must dismiss the idea of the necessity of a large amount of water because the claims cover any quantity that, under agitation, will keep the pigment in solution; how much that may be Í3 not disclosed by the record,

Considering the evidence, we find:

(a) Pfingsten had his patent at stake; Martin, so far as appears, had no interest,

(b) The quality of the Martin dyeing had brought Pfingsten to Martin to buy some of the coloring. He saw Martin’s demonstration produce a uniform, unspotted coloring, That he thought the coloring and Martin process were both right is conclusively shown by his testimony that he bought the coloring matter, went home, and tried to get Martin’s results by Martin’s process,

(c) The record shows that spots, in dyeing, come from two causes: (1) An improper preparation of the leather before the dyes are applied; (2) a failure to use sufficient water. Pfingsten’s patent is based on the fact that he did use sufficient water. He only sueceeded, he says, in producing a spotless job by the use of more and more water. If that is the only way to produce a spotless job, the *577evidence is strong that Martin used enough water, because his splits wore spotless.

(d) When asked as to how much water Martin used, and as to the quality of the work produced, Pfingsten made such indefinite and evasive answers as convince us that he either had no very definite recollection about what he saw at the demonstration or was lacking in candor. Either would very much weaken Ms testimony. As to the quantity of water, several answers were made, and then, “as far as I recollect, there was hardly any.” Finally admitting the good quality of the work, Pfingsten conclusively shows that Martin had sufficient, when judged by Pfingsten’s notions of the quantity of water necessary.

Pfingsten said: “Of course, after I saw in how small a mill he (Martin) colored, I could really see that he would be able to get a perfect colored split by using a small amount of leather.” Then he was asked: “As a matter of fact, then, the amount of water used was small, beeatuse he was just coloring a small amount of leather?” He did not answer the question, but said, “He did not use hardly any water,” and then digressed to say why he could not have used much. How much “very little” or “hardly any” water, in a tank 6x8 feet and holding several hundred gallons, would be, we cannot tell. One looking at a barrel with 25 gallons of water in it would probably say it had a great deal in it. Looking at a tank of several hundred gallons’ capacity, with only 56 gallons in it, the same person would probably say it had very little in it.

Martin’s spotless job, the evasiveness and lack of definiteness in Pfingsten’s answers, the fact that Martin used some water, and that Pfingsten saw how Martin, by coloring a small amount of leather, “got a perfect colored split,” are facts that seem to us to support the truth of Martin’s testimony.

The suggestion is made that, where too little water is used, there is an excess of coloring matter, that drops when the splits are staked. We will later discuss that feature. There is no doubt that it is true, but no one found any such thing on the Martin splits. It is urged that, where there is insufficient water, the coloring matter will be pasty. That is self-evident, but Pfingsten did not discover that Martin’s coloring was pasty.

We do not find Pfingsten’s experiments in coloring either enlightening or persuasive. He nowhere told of the steps taken by him in the experiments that he says failed. He said notMng more than: “I went to my f actory, * * * and made an experiment along the lines of Mr. Martin’s instructions. * * * I had probably 30 or 40 per cent, of spotty leather.” It would seem that that result is evidence that he did not follow Martin, and there is further evidence that he did not, because he had a surplus deposit of coloring matter that came off when the splits were staked. Martin had nothing of that kind.

In his first experiment, Pfingsten says he used “a very small amount of water, or practically dry.” There is Ms evidence that Martin used “very little” water, but there is no evidence that, when he made the demonstration to Pfingsten, Martin used leather that was “practically dry.” He made several more experiments by adding more water, and when he finally arrived at about the amount that Martin says he used he got the “perfect colored split.” To say the least, that is not evidence that he was not then following Martin. Neither Birkholz nor Lakatos saw the Martin demonstration; so their testimony does not support Pfingsten as to the amount of water used. That Pfingsten got both good and bad Results needs no support.

There is in favor of the Martin story the character of the powder used and of the splits to be colored. The splits were large, the size of half a cowhide, there were many of them, several hundred pounds placed together in a drum, and they were wet. The powder was dry when put in, and, if there was not enough water and agitation to hold it in suspension, it would simply become pasty from contact with the wet hides, and, unless carried between the hides by the wash of water in the turning and tumbling of the hides in the slowly rotating drum, excessive amounts would stay where it fell upon the damp hides, and inevitably not reach and col- or all parts of the hides. Pfingsten claims that he found out that fact by a few experiments. Martin had colored hides with the same coloring matter for a year, and it is not possible that he had not learned and did not disclose to Pfingsten the very obvious necessity of using water sufficient to carry the coloring matter to every part of the hides.

It is urged that, by the patent method, the coloring matter penetrated into the interstices of the leather in a way different from and better than by the Martin method. There is no such evidence in the record. We are, of opinion that Martin taught Pfingsten all the essentials of the patent, and, in addition thereto, that putting in water enough to hold the powder in suspension under agita*578tion was the obvious thing to do, and did not constitute invention.

The decree is reversed,' with directions to dismiss the bill.