delivered the opinion of the Court:
This is an appeal from a decision of the Commissioner of Patents in an interference case, having the following issue:
. “1. In a furnace, the combination of a porous hearth supporting the material to be treated, and means for passing the air or gases through the porous hearth to the material.
“2. In a reverberatory furnace, the combination of porous material for supporting the material to be treated, an air chamber or passage underneath said porous material, and means for introducing air or other gases to said chamber or passage.”
Charlie E. Mark’s application, filed November 8, 1897, claimed an improvement in reverberatory furnaces for malting and purifying metal through a porous hearth. John A. Greenawalt’s application, filed December 27, 1901, claimed an improvement in reverberatory furnaces for roasting ores, with a like introduction of air through a porous hearth. An earlier interference than this was declared between the two applications with a narrow issue relating particularly to a furnace for roast*255ing ores as described by Greenawalt. Questions were raised as to the operativeness of Mark’s furnace for the purpose described by him, and as to his right to make the narrow claims of the issue. When the case reached the Commissioner, he dissolved the interference on the ground that Mark’s furnace was inoperative as a melting furnace for metals, and remanded the application for a further consideration of the question of operativeness. Affidavits of experts were filed by Mark, which were satisfactory to the Primary Examiner in respect of operativeness; but he expressed the opinion that Mark had no right to make the narrow claims of the issue.
The difficulty of the situation was this: The essential feature of the invention claimed by each was the arrangement of a reverberatory furnace with a porous hearth and the means of forcing air or gases containing oxygen through the interstices of the porous hearth to assist in the roasting or melting process to be carried on in the melting chamber of the furnace. Greenawalt’s expressed object was to roast ore, particularly that containing precious metals, preparatory to the extraction of such metals by a further process; the air being introduced during the melting process to oxidize certain combustible elements of the ores. Mark’s expressed object was to melt metal, particularly pig iron, as commonly done in reverberatory furnaces; and the introduction of air through the porous hearth was to aid in the melting and purification of the metal treated.
The invention lay in the forcing of the air through a porous hearth, which is common to both applicants, and not in the different, though related, uses intended. Consequently, a patent could not properly be issued to each for the same furnace, with a limitation to a different practical use.
Mark met the situation presented by the Examiner’s ruling by inserting the broader claims of the present issue, which Greenawalt also adopted, and then the present interference was declared.
It was found by all of the tribunals of the office, and is beyond’ doubt, that Mark was the first to conceive the invention. The question then was, as stated by the Examiner of Interferences : *256“Mark being the first to conceive and the first to file his application, the decision in this case depends on whether that application discloses an operative device, and Mark is entitled to it as n constructive reduction to practice.” Greenawalt was charged with the burden of showing that Mark’s furnace was inoperative, and the case turns upon the evidence offered by him showing practical tests of the same in Denver on February 3, 5, and 8, of 1906, respectively. The tribunals of the office, in turn, held that this evidence established the fact of inoperativeness, and was not overcome by the evidence on behalf of Mark, who has never built or operated a furnace after his design, because, as claimed, of ■the want of means and opportunity so to do.
The evidence offered by Greenawalt, taken a few days after the tests, tended to show that he built at his works in Denver a furnace with a porous hearth, and provided it with the means of supplying air thereto, as described by Mark, and furnished with appliances for measuring the amount of air supplied. Drawings ■of the furnace were exhibited.
The tests were made by Greenawalt and his brother, with the active assistance of two experts, who were provided with a modern pyrometer for testing the temperature of the furnace ■from time to time. The first test made was to ascertain if the air introduced through the porous hearth would combine with the combustible gases coming from the fire box and increase the temperature of the melting chamber of the furnace with no metal therein. The second was to determine the effect of passing the air current through the hearth with pigs of metal lying thereon. The third was to determine the effect of passing air through the hearth while covered with molten metal. The testimony then tended to show: (1) That the temperature was not increased as the result of the first test; (2) that the air so applied was of no assistance in melting the iron; and (3) that, in attempting to introduce the air through the molten metal, the pressure at once rose and disrupted the hearth. Mark objected to the competency of this evidence on the ground that the experiment was made without notice to him, or opportunity for him to be present and observe the same.
*257He also attacked its weight on the ground that the experiment was privately carried on by the interested party with experts on his behalf, and that the furnace had been dismantled or broken up immediately afterwards.
He also introduced several expert witnesses who analyzed the evidence describing the experiments, and undertook to sho.w that the furnace was not erected in accordance with the description of Mark, was not a properly constructed reverberatory furnace, and that the temperature readings were not properly taken, and particularly that in the third test the air was not properly introduced.
There can be no doubt that the fairest way to make these tests Avould have been to notify Mark and give him the opportunity to Avitness the same; or, at least, to leave the furnace intact, and give him and his assistants the opportunity to make other tests in their presence and that of their experts, together with the presence of other disinterested experts if such could be reasonably procured. By pursuing either course all possible doubt might have been easily removed. We are not prepared to say, however, that the testimony was incompetent for the reasons given. But, at the same time, the one-sided nature of the tests, and the subsequent dismantling or partial dismantling of the furnace, ought to have great importance in respect of the weight to be given to the evidence relating to the tests.
A party deeply interested in showing that his rival’s device will not operate or accomplish the object claimed for it, who secretly tests the same with the aid of an expert who had previously testified to the opinion that it would not operate, must necessarily rest under the burden of showing, with great certainty, that they conformed strictly with all the requirements of the description, and in good faith and with great care took all the steps necessary and important to obtain fair and accurate results.
After careful examination of the voluminous testimony relating to the tests, we are of the opinion that the tribunals of the office did not give sufficient weight to these considerations.
It is readily seen that Mark’s application is rather indefinite *258in its description, but it is to be remembered that, while one is required to point out and distinctly claim the part, improvement, or combination which he claims as his invention or discovery, his specifications are not addressed to the public generally but to those skilled in the art. In the language of Mr. Justice Bradley: “He may begin at the point where his invention begins, and describe what he has made that is new, and what it replaces of the old. That which is common and well known is as if it were written out in the patent and delineated in the drawings.” Webster Loan Co. v. Higgins, 105 U. S. 580, 586; 26 L. ed. 1177, 1179; Carnegie Steel Co. v. Cambria Iron Co. 185 U. S. 403, 437, 46 L. ed. 968, 986, 22 Sup. Ct. Rep. 698. Mark distinctly pointed out the invention which' he claimed, which was arranging in a reverberatory furnace a hearth permeable to air but impermeable to the metal melted thereon, and providing means for the introduction of air thereunder and forcing the same through the hearth and into the metal thereon. “In this way,” he said, “the air can burn out the combustible portion of the metal and so materially aid the melting action, without removing from such action any of the advantages incident to its performance in a reverberatory furnace.”
It was unnecessary to describe the reverberatory furnace which had been long in common use. This furnace has a fire box and adjacent thereto what is called the melting chamber, separated at the bottom by a bridge wall over which the heat and fuel gases pass from the fire box into the melting chamber. The hearth was ordinarily constructed of silica sand on which the metal to be melted was placed. It was customary to heat the furnace before introducing the metal, so that the sand would be baked into a thin crust on the surface, and hold the metal without breaking up. Naturally the flame and heat will flow from the fire box to and along the arch or top of the melting chamber towards the chimney flue, unless deflected and made to turn down upon the hearth. The reverberatory furnace used in making malleable-iron castings gets its name from the fact that it is so constructed as to cause the heat to reverberate down upon the hearth and as near the fire-bridge wall as possible. The object *259in firing seems to be to gasify a large part of tbe fuel in the fire box and turn as much as possible of the unconsumed gases and the hottest part of the flames into the melting chamber. Part .of that construction, commonly adopted at the time of Mark’s application, consisted of a projection from the arch or roof of the furnace, directly, or almost directly, above the fire-bridge wall for the purpose of assisting in deflecting the heat current and combustible gases downward towards the hearth. It was also a general, if not universal, practice in furnaces of good size, to force quantities of cold air into the melting chamber through what are known as “tuyeres” inserted in the top or roof directly, or nearly so, over the fire-bridge wall. This air current is so directed as to aid in the deflection of the heat downward. It also effects the combustion of the fuel gases which pass out of the fire box, and thereby increases the heat in the melting chamber. No invariable rule can be laid down as to the quantity of this air which, if in excess of the amount necessary to effect the combustion of all of the fuel gases, would have a detrimental effect. Therefore, the pressure and consequent volume of air must be regulated by the judgment of the operator of the furnace, who must take into consideration all of the conditions attending the operation from beginning to end.
The same conditions apply in the process of Mark for introducing air through the hearth, the effect of which, according to his idea, is, first, to increase the heat at the hearth by operation upon the unconsumed gases deflected thereon, and, second, to purify the molten iron by burning out its impurities.
This brings us to the consideration of the evidence relating to the tests which are claimed as demonstrating the inoperativeness of Mark’s invention.
It would prolong this opinion to an intolerable length were we to undertake to review, in detail, the evidence offered in support of the sufficiency of the tests to demonstrate the inoperativeness of Mark’s apparatus and process, or the evidence of Mark’s experts, in whose opinions the tests were not properly and satisfactorily made. We shall therefore confine ourselves to a statement of the chief reasons why, in our opinion, the evi*260deuce of the test falls short of the certainty of proof heretofore mentioned as necessary under all the circumstances.
1. The reverberatory furnace in most general use in making malleable-iron castings has a melting chamber from 18 to 20 feet in length. The fire-bridge walls in such furnaces apparently vary in width from 27 inches to more, and the size of the fire box is in reasonable proportion to the size of the melting chamber. The shape of the top of the melting chamber varies according to the ideas of different manufacturers, but the object is to obtain a construction that will cause the deflection and reverberation of the heat. In many constructions this is accomplished in part, as before stated, by arranging a projection from the roof of the furnace over the fire-bridge wall that will tend to give an abrupt downward deflection of the flames and fuel gases. This projection is indicated in Mark’s drawing. In all large furnaces, at least, it is customary to provide tuyeres that have been before described.
The demonstration furnace, built shortly before the test at Greenawalt’s works in Denver and apparently for the purpose of this test, was quite a small one. The fire box may have been unduly large, or the fire-bridge wall of 21 inches may have been too wide; but we need not consume time with the criticism of their size, etc. The melting chamber was 3 feet, 9 inches long, and was curved from the outer edge of the fire box to the chimney flue. There was no projection from the curved surface over the fire-bridge wall or anywhere else. There were no tuyeres, and the curved construction was alone relied upon for the deflection of the heat. The experts on behalf of Greenawalt testified that this construction was more efficient than that indicated in Mark’s drawing, and that in so small a furnace tuyeres were not necessary. We will pass by the probable effect of tuyeres in such a furnace. On the other hand, Mark’s leading expert testified that the furnace curve tended to deflect the heat towards the end of the furnace near the chimney flue, and not towards the hearth at and near the fire-bridge wall, where most important. It seems quite clear that the heat should have been deflected upon the hearth at the *261nearest point to the fire-bridge wall to insure as far as possible uniformity of heat along the entire surface of the hearth. Without undertaking to determine that the one construction would be as effective as the other in the small furnace, we are not satisfied that such is the case. Whatever the opinion of G-reenawalt, we think it was incumbent upon him to use a furnace conforming strictly to Mark’s description, or else to show with certainty that the deflection in the test furnace would be fully as effective as the other.
2. In view of what has been said in regard to the construction of the furnace, the first of the tests may be briefly disposed of. If, as contended by Mark’s expert, the construction of the test furnace was such as to cause the greatest heat and the current of unconsumed gases to follow the top line of the arch, and to be deflected to the hearth at or near the flue end, then it would seem natural that the introduction of cold air through the hearth, that would not come into immediate contact with the fuel gases, would lower the temperature, especially in the neighborhood of the fire-bridge wall. As before said, we are not satisfied by the evidence that the curved construction would not have the effect attributed to it by Mark’s expert witness.
3. The same considerations apply in the second test where the air was introduced through the hearth while the charge of 1,100 pounds of pig iron was being heated and melted. The amount of air introduced into the melting chamber through the tuyeres, as we have before remarked, is a matter of discretion to be determined by the operator under the conditions presented from time to time; and the same must be true as regards the introduction of the air through the hearth of the furnace. Considering the conditions of heat at the time of this second test, it seems highly probable that an excessive amount of air was forced in through the hearth. The experts who made the tests immediately prepared a statement of their actions and observations, which they compared with their notes taken throughout the different processes. Mark’s experts have vigorously attacked the accuracy of the heat records disclosed in the statement. We deem it. unimportant, however, to consider them, save in one respect.
*262It is agreed by all the experts that average pig iron will ordinarily melt at a temperature of 2000, F., though it is customary, to secure certainty and quicker results, to- get a heat averaging, say, 2500, F. It appears from the statement that the empty furnace was heated for four hours without introducing air. At 2: 48 p. m. the pyrometer readings were registered as follows: Hearth, 2723, F.; arch, 2723, F.; bridge, 2732, F. Within twenty minutes a charge of 1,118 pounds of pig iron was introduced. With the heat recorded, the melting of this small amount of iron ought to have commenced and been completed with great rapidity; apparently within a half hour or very little more. At 3: 25 p. m., twenty-four minutes after the charging was completed, 2:30 cubic feet of air per square foot of hearth area was forced in under a water-pressure indication of 17 inches. This was about all that the hearth would resist, for within twenty minutes a blowhole action was observed. Now, if an excessive amount of air was introduced, it would have a detrimental action upon the heat, as heretofore observed. And it seems that it would have been reasonable to introduce the air in small quantities at first, and, after observing its effect, to gradually increase it. Be that as it may, at 3: 50 p. m (forty-five minutes after the iron was charged an'd the air turned on), the following reading was taken: Hearth, 2075, F.; arch, 2775, F.; bridge, 3200, F. At 4: 45 p. m., the reading was: Hearth, 2003, F.; arch, 2775, F.; bridge, 3047, F. Now, it is further stated that some of the iron in the upper layer began to melt and run down upon the lower layer, where .it was chilled to solidification. Pieces of this iron broken from the pigs with which it had come in contact were exhibited.
Without intending to impugn the veracity of the persons conducting the test, we are constrained to say that we are not satisfied with the accuracy of the pyrometer readings, an error in which might have been due to the construction of the pyrometer, or to the apparent difficulties attending the accurate ascertainment of the position, of the indicator, which the evidence tends to show. Aside from the character of the readings first given, we cannot escape the conclusion that there was some mistake, be*263cause of the statement as to the melting. As the heat over the upper layer was far above the melting point of all iron, the melting of that layer should have been very rapid. As the molten metal ran down upon the pigs, it ought to have increased their heat and assisted their melting. And, if the heat at the hearth was in fact 2075, F., and then as low, even, as 2003, F., it was still above the melting heat. Coming down with excessive heat in it, we are unable to understand why this metal should have solidified at a temperature when still about the melting point.
Grant that the metal may have been of a character that required more than 2000, F., to melt it, it seems reasonable that, when melted, it would have remained in that state at the lesser temperature. Moreover, the iron at the bottom was not cold iron, but must have been near the melting point at least; hence, its chilling effect ought to have been more than counteracted by the excessive heat of the molten metal flowing \ipon it.
3. We come now to the third and last test in which the air was forced through the hearth when it was covered with several inches of molten metal.
It is evident that the demonstrators forced the air with extraordinary pressure so that it disrupted the hearth and blew parts of its contents through and to the top of the molten bath. On this occasion the furnace was first heated for four hours without introducing air through the hearth. At 1:25 p. m., the pyrometer read: Hearth, 2542, F.; arch, 2732, F.; bridge, 2804, F. Without considering whether the heat at the bridge was out of due proportion, owing either to imperfect firing, or the construction of the furnace, it is to be observed that there is a difference between the heat in this and the preceding two at a similar period of the operation, in that there was an excess of 190 F. at the arch over that registered at the hearth, there having been no introduction of cold air either at the top or through the hearth of the melting chamber. The charging with 1,118 pounds of pig iron was finished at 1: 30 p. m. It began to melt at 2: 45 p. m., and the melted iron dropped on the hearth and remained molten. The statement then reads as follows: “At 5: 30 p. m., when the iron was molten and flooded the hearth to a depth of about 3 *264inches, air was introduced through the hearth and molten iron. The air was turned on very gradually. When the air meter registered 1.7 cubic feet of air per minute per square foot of hearth area, the first indication was seen on the molten iron. It manifested itself by coming through the iron in three or four blowholes. There was no other manifestation of air in the furnace. The pressure gauge at the time registered 60 inches of water. Within about a minute after the blow-hole action began, there was an upheaval of the hearth in the vicinity of the blowholes which consisted of sand, chilled iron, and slag.” It goes on to state that, within two minutes after the air was turned on, the hearth was completely destroyed; no molten metal remained on the surface, and none could be drawn out through the tap hole.
It appears from Dr. Chauvenet’s testimony that the long time that was taken to melt the whole was due to “the fact that at one corner of the furnace, close to the bridge wall, there was some iron which resisted the melt longer than at any other point; the heat was continued for some time after the great bulk of the metal was molten, in order that the turning on of the air should not be tried until the whole of the pig iron was in a fluid condition.”
We are not satisfied from all the evidence that the air was properly turned on in this test. Notwithstanding the statement, from which we have quoted, that the air was turned on “very gradually,” Dr. Chauvenet, in cross-examination, said that the air turned on registered 1.7 cubic feet per square foot of hearth, and “at this time the pressure rose at once to 5 feet in the tube gauge, — that is, the difference in the height of the water column in the two arms of the tube was 5 feet.” Professor Sadtler, the other expert demonstrator, said in his direct examination: “The air was turned on at the rate of 14 cubic feet per minute, and the water gauge immediately jumped from zero to 60 inches pressure. I ran outside to observe the influence of the introduction of this amount of air, and first noticed nothing on the surface of the molten bath of metal, but some air was blowing out around the edge of the hearth. Then several small blowholes *265made their appearance near the center of the bath of metal, the metal around such blowholes being dark in color and apparently chilled by the air. In a very short period thereafter, probably a minute, a mass of iron rose to the surface of the metal accompanied by a very liberal spouting of the sand of the hearth which remained on the surface of the metal, which gradually disappeared or sunk in level, leaving the sand or chilled fragments on top.” (It is to be remarked that the statement of 14 cubic inches of air had reference to the whole hearth area, and accords with the statement of Dr. Chauvenet that it amounted to 1.7 cubic feet per square foot of hearth area.)
Under all the evidence, we cannot but feel uncertain as to the probable result had the air been in fact turned on “very gradually.” We are not prepared to assume that, had there been a. gradual turning on of the air, stopping short of a pressure that would disrupt the hearth, there would still have been no beneficial result.
In explaining the sudden and immediate rise of water in the gauge, indicating extraordinary pressure, it is attributed to the condition of the crust of the hearth, which, it is contended, became practically impermeable to air, and resisted the pressure from below until it broke as related. It is upon this theory that the Commissioner, in the original interference between the parties, decided that Mark’s furnace was inoperative. He said that it was necessary for the materials of the hearth to he well fritted or fused together to support molten metal. He said: “In producing such a bottom in practice, the material, generally a refractory sand, is spread in a thin layer and subjected to a high temperature until -the grains reach a state of semifusion and weld into a compact mass. Then a fresh layer of sand is applied and the process continued until the bottom is of the required thickness. * * * It is thought the bottom would have to be so hard and dense as to preclude the passage of an. appreciable quantity of air therethrough.”
The evidence in this case shows that, while some hearths may be constructed as above set forth, it is sufficient to make the hearth of one layer of sand, and apparently that is the way in *266which the hearth in the demonstration furnace was made. Such sand, while permeable to air, is impermeable to molten metal to any appreciable extent. It is usual, however, to give the furnace a preliminary heating for the purpose of baking and hardening the top of the sand bed; but this is done to prevent the pigs, when thrown in, from sinking in the sand and otherwise destroying the level of the hearth. When there is no preliminary heating, boards are laid on the hearth for the purpose -of receiving the pigs. The sand is neither fused and glazed by the preliminary heating, nor by contact with the molten metal, for it resists a temperature that is ample to heat the metal. While this crust is less permeable to air than loose sand, it remains porous and penetrable. It appears that, even after a number of successive meltings, this crust, which increases in thickness each time, remains porous and will absorb water. Such of it as burns and mingles with other impurities becomes slag, which, instead of lying on the hearth and increasing the diffi•culty of forcing air up through it, rises up through the molten metal to the surface of the same. Hence, it is not seen why air might not have been forced through it without disruption under a carefully regulated pressure. With the air forced through without disruption of the hearth, it does not seem improbable that it might have assisted in the complete melting and then in purifying the metal by causing the burning of some of the impurities. At any rate, the evidence does not satisfy us that such would not have been the result. In the ore-roasting furnace of Greenawalt, it does accomplish this burning and purifying resnlt, though in that process there is no melting of the ore.
The experts on behalf of Greenawalt seem to have confounded Mark’s process, to some extent, with that of the Bessemer direct process of malting steel, in which a current of air under tremendous pressure is driven into and through the molten metal in the converter. In that process the converter is provided with openings or tuyeres through which the metal would escape unless prevented by the great air pressure applied thereto. Mark’s process, on the other hand, does not require an air pressure that will support the metal and drive the air through it, but is limit*267<ed to forcing the air into the metal. “In this way,” he says, “the air can burn out the combustible portion of the metal and so materially aid the melting action, without removing from such action any of the advantages incident to its performance in a reverberatory furnace.” Forcing the air into the metal lying on the hearth may be understood as forcing it into contact with the metal, and in part penetrating it through the burning operation that begins with the contact. It certainly does not mean forcing the air through the mass of molten material, as was done in the test under consideration.
We are not to be understood as finding that Mark’s apparatus .and process will operate successfully and commercially. That •question is not before us. His application was allowed as' disclosing operativeness, and Greenawalt was charged with the burden of showing that it was not really operative for any practical or useful purpose. The effect of the evidence of his tests upon which he relied was the single subject of consideration by the tribunals of the office, and that is the question brought up to us by this appeal.
We repeat that this evidence is not of sufficient weight and certainty, in our opinion, to overcome the burden imposed; and for that reason we are constrained to reverse the decision appealed from.
The decision will be reversed, and the clerk is directed to certify this decision to the Commissioner of Patents as required by the statute. Reversed.