British 2,395 of August 21, 1867, is for like improvements, wherein oxide ores are reduced or deoxidized in vertical hoppers or retorts above the smelting chamber, into which they discharge between sides and center at the "surface of the metallic bath, or nearly so, in order to cause a more gradual and uniform descent of the ore into the furnace with as little exposure as possible of the spongy metal reduced in the hoppers to the oxidizing influence of the flame."

British 1,892 of 1868 is for like improvements, wherein for the vertical retorts of the 1867 patent is substituted a horizontal cylinder or muffle, which discharges through a vertical hopper "into the metallic bath of the melting furnace" at its center. This patent refers to that of 1867, both refer to that of 1866, and all three are of reference in that of April, 1871, next following. And the materiality of the 1867 patent and the 1868 patent consists in that the reference aforesaid in the patent of 1871 has been obvious

ly misinterpreted as an admission by Siemens of failure of the process of wall protection by ore of his 1866 patent, and to this misinterpretation is mainly due the issuance of Carson's second patent.

United States patent 113,584 of April, 1871, is identical with British 594 of 1870. For like "improvements in processes and apparatus" in ore to steel manufacture, it refers to the patents of 1866, 1867, and 1868, as follows:

"In my British patent of 1866, No. 2,413, I described a process and apparatus for the production of cast steel directly from the ore, with or without the addition of pig iron, and of ferromanganese or spiegeleisen.

"This process consisted, essentially, in effecting the reduction or deoxidation of the ore by the reaction upon it of carbonaceous matter under the influence of intense furnace heat, and in effecting the liquefaction of the reduced ore within the same furnace by the further addition of carbonaceous mat

ter. It was found, however, in practice, that the deoxidation of the ore and the adjust ment of the chemical condition of the metal

lic bath were attended with great difficulty under those circumstances, to remove which I subsequently separated the process of deoxidizing the ore from that of effecting its liquefaction and chemical adjustment by the introduction of vertical or other retorts, or of rotating muffles, as described in my British patents of the 21st of August, 1867, No. 2,395, and of the 10th of June, 1868, No. 1,892, respectively.

"Notwithstanding these improvements, it was difficult to realize all the conditions necessary to insure a satisfactory result. The reduction of iron ore in close retorts or muf

repairing the sides from time to time"; that in the furnace preferred.

"In this furnace the crushed ore AA, mixed with the requisite percentage of flux and carbonaceous matter, as before described, is fed continuously onto the bed B through the hopper C, into which it is charged from trucks W, running on rails, as shown at Fig. 1. The ore is caused to pass gradually from side D of the furnace, situated opposite the the hopper in a thick layer down the inclined gas and air ports, which are all placed on one and the same side of the furnace, in a similar manner to arrangements already patented by me both in England and in the United States."

This furnace in longitudinal section follows:

fles is essentially a slow and expensive process, and the pulverulent iron produced thereby, upon being introduced into the melting furnace, floats upon the metallic bath for a considerable length of time without being being incorporated with it. Being exposed, in the meantime, to the oxidizing and sulphurizing action of the flame, the metallic oxide thus produced corrodes the banks of the metal bath, and, being a nonconductor of heat, causes the fluid metal below to set."

In description of a furnace, the specifications state that "refractory iron ore, such as hematite, may be used with advantage in

In respect to the misinterpretation aforesaid, it would seem that the Patent Office had not discovered the April, 1871, patent, for the only Siemens patent to which it referred is that of 1866. Apparently, however, Carson knew of others, for in his final argument before the Examiners he cited the April, 1871, patent and contended that in the foregoing quotations Siemens admitted that the incidental process of wall protection by ore of his 1866 patent had failed, and that, declared Carson, because the ore would not descend the inclined walls to the hearth, nor "prevent the oxide of iron from dissolv

The Patent Office accepted and acquiesced in this error, and because of it issued the second patent. Apparently the Patent Office thus perpetuated Carson's error, without even perusal of the April, 1871, patent; for it is clear, too clear for argument, that wherein that patent states the "oxide thus produced corrodes the banks of the metal

bath," it refers to the furnaces of the 1867 and 1868 patents, center-charged and without ore wall protection, and not at all to the furnace of the 1866 patent, side-charged and with full ore wall protection.

Moreover, the reference in the April, 1871, patent to "repairing the sides from time to time," yields to the character of the

the drawing above, the ore charged at the end, down which wall the ore by gravity continuously descends "in a thick layer" fully protecting it, the sides not thus protected. And note well that in this patent 113,584, which is alleged to contain Siemens' admission of failure of his ore wall protecthe ends are fully thus protected. See Fig. tion process, he prefers a furnace wherein 2 above. Of course, sides and ends are equiv

alents.

British 3,077 of November, 1871, is for like improvements, and refers to British 594, and so to United States 113,584. Therein Siemens describes, as the "best means which I know of carrying" the invention "into effect," the operation of the furnace in the drawing below:

That is, "the ore to be fused, mixed with about 5 per cent. of carbonaceous matter, is supplied to the hoppers and sinks down through the openings into the bed of the furnace, forming a sloping heap at either side. The ore acted on by the flame is fused, and the liquid ore collects in the basin formed in the middle of the furnace between the two side heaps of ore," and that the metal might be drawn off through the trap door in the hearth, "or the fused ore may be tapped from time to time, so as to issue by a tap hole in the front of the furnace, whence it runs by a movable gutter O to the converting furnace P."

French 93,906 of 1872, for like improvements, and wherein are drawings of furnaces like those aforesaid of patents 113,584 and 3,077, the latter with and without a trap in the hearth. This patent is notable for that, whereas, in all other patents Siemens relied on common intelligence and knowledge of physical laws to select an angle for walls which would operate his plainly described ore wall protection process, in this patent he defines the angle as "about 50 centimeters (a 62 degree angle)."

All the foregoing issued to Charles Siemens, who died in 1883, and to his brother Frederick issued the last in evidence, viz.: British 14,143 of 1885, for "continuous reduction of iron ore." The specifications state that next to the furnace end, where heat is supplied

"next to this end of the furnace I form a deep bed or basin, from which I carry a slope up to the extreme end of the furnace where there is an opening through the roof. Through this opening I feed pulverized ore mixed with reducing materials and fluxes. This as it gradually descends the slope is subjected to the radiant heat of the flame which has the effect of reducing the ore, and the metal produced flows down to the basin where it is collected with the slag above it. I provide a tapping hole from which the slag is allowed continuously to flow, and a lower tapping hole by which from time to time I draw off a portion of the molten metal, leaving always a large mass of the metal in the basin. I also provide a third tapping hole at the bottom of the basin for the purpose of running off the whole contents when desired.

"The furnace may be made in duplicate form, that is to say, with a slope on each side and the basin bed between them, the flues being arranged as described above so that the flame takes a direction at right angles to the lines of descent along the slopes." And the claims are that:

"I make no general claim to a furnace in which ore descends a slope to feed a basin bed where in admixture with carbonaceous substances it is subject to the action of a gas flame but what I claim is:

"The construction of a regenerative gas furnace for continuous reduction of iron ore, having at one of its ends or sides the air and gas flues which communicate with the regenerators, having a deep basin bed provided with tap holes, and having beyond or at each side of the bed a slope fed with ore mixture from the crown of the furnace, the flues being so arranged and the furnace so proportioned that the flame acts by radiant heat alone, substantially as herein described."

The drawings are of furnaces like those aforesaid of patents 113,584 and 3,077, the latter without a trap in the hearth. Of other patents defendant presents Taylor's United States 843,776 of application in 1902 for an electric smelting furnace. The specifications state:

"As will readily be understood, the furnace is designed to obviate to the greatest possible degree loss of heat by radiation from the furnace walls and the corrosive action of the products upon the fixed elements of the furnace. The walls are protected at all points by the incoming charge and streams of conductive material."