80 FBDEEAL BBPOBTEB. �ley, it would not have changed His opinion. He fonnd one sale, and evidence which inclined him to think that there might have been two or three more, and yet sustained the patent. I find myself differing from my eminent predecessor upon the effect of the evidence, and have had great doubts as to my duty. I have determined that, as new evidence bas been produced •which I must act upon, my action must be in accordance with my own views of its effect. I cannot under- stand the new evidence without opening the whole record, and I must act upon what I find in the old and new evidence together. �In my opinion the evidence tends to show a sale of the invention. True, some sales were conditional; that is to say, the stoves were to be returned if they were not satisfactory to the buyers; but this does not, without further explanation, prove that they were experimental. It may show that the purchaser had doubts about the article, but does not prove any on the part of the seller. Sales in the usual course of business, whether absolute or conditional, if they are sales of the patented thing, work a forfeiture. A single sale has this effect, as well as a hundred sales. It is very unlikely that a buyer would take what he understood to be an experimental thing ; but if he did, the evidence should be unequivocal that a test of the invention was one of the purposes of the seller. This article could be tested by the inventer as well in his own house as in any other place ; and when he sold it in its com- pleted form, though with warranty or on condition, he sold it. �Upon the facts, the stove appears to bave contained the invention, within the doctrine of Am. Tlide, etc., Go. v Am. Tool Co. 1 Holmes, 503, 513, in which Judge Shepley charged that the thing sold need not be perfect in the mechanical sense, but only in that it embodied the completed invention in a form which would be operative. Indeed, this stove appears to have been perfect in both senses, for it has been in suecessful use for about twenty-four years. �Judge Shepley found that the sales in evidence before him were made while the inventer was still conducting experi- ����
