A is a tube screwed on to the supporting base G. Bis a round piece covering and fitting into the tube A. Through the centre of B goes a screw-threaded spindle C, which bears on its lower end, between a fork-like cut, a pulley D. F is the pin on which the pulley turns and extends through opposite cuts in the tube, to prevent C and D from turning when B is turned to lower or raise the pulley D.

Over the pulley the suspending spring e lies, and its two ends are fastened between the two bars KK. The two plates HH are screwed against the under part of G, and just far enough apart to allow the ends of the spring to extend laterally through the slit they form.

If now the pendulum hangs on the L its weight will bear equally on either side of the spring by means of the pulley, and adjust its tension automatically.

How perfectly that suspension adjusts the tension can be seen if we bring the bars KK into the position shown by the dotted lines. Even this abnormal position will not interfere with the pendulum in performing its oscillations in a true plane. While with other suspensions it would be impossible to avoid the wabbling motions if K and H would vary the least from being parallel.

THE PUDDLING PROCESS, PAST AND PRESENT.

By PERCIVAL ROBERTS, JR.

Read before the American Institute of Mining Engineers.

It may seem necessary to offer an apology for presenting for consideration a process which is conspicuous by its absence in the literature of the Institute, and which may be thought by some to belong to the past in metallurgy, and to have been already superseded. But the large capital invested in puddling calls for a careful consideration of the question whether the time has certainly arrived when the puddling furnace must be replaced by the converter and open-hearth furnace. May there not still be a place for puddled iron alongside of molten iron and steel, and is not the improvement of the puddling process itself worthy the attention of engineers equally with the Bessemer and open-hearth processes?

The changes involved in the conversion of pig iron into wrought iron are well understood and need only be briefly alluded to. The patent of Henry Cort bears the date of 1784. Since that time the improvements in the process have mainly consisted in the replacement of sand by iron bottoms by Samuel Baldwyn Rogers in 1818, and the still more recent substitution of iron oxide for the refractory materials used for the sides and bridge of the hearth, which distinguishes the wet or boiling process from the dry or puddling process. Chemically, the process consists in the removal of the metalloids from the pig iron, a result effected mainly by the iron oxide. Silicon is first oxidized, then the phosphorus, and finally the carbon. The silicic and phosphoric acids produced pass into the cinder and the carbonic oxide burns as it escapes from the bath of metal.

It ie interesting in this connection to note the effect of temperature on the removal of the phosphorus from the iron. As is well known, no phosphorus is eliminated under the oxidizing influences prevailing in the Bessemer converter, while from 70 to 80 per cent. is removed in puddling. But we find, if in working cold short irons the temperature of the furnace is much increased towards the end of the process, that a considerable amount of the phosphoric acid is deoxidized and phosphorus again combines with the iron. This reverse process WHOLE NO. VOL. CX.-(THIRD SERIES, Vol. lxxx.)

4

is aided by a siliceous cinder arising either from the use of a very siliceous pig iron, or of an over-siliceous ore for fix. The fact which has been known for some time that only a basic cinder can retain phosphoric acid has given rise to the "basic lining" which now attracts so much attention in the Bessemer process. For the conditions affecting the removal of phosphorus from pig iron I would refer to the careful and complete experiments of I. Lowthian Bell, in England. Notwithstanding the recent progress in the metallurge of iron the puddling process is essentially what it was three-quarters of a century ago-laborious, crude and unsatisfactory. The attempts at improvement in the process may be classified under two heads: 1. economy of labor; and 2. economy of fuel.

Increase of yield and improvement of quality are so intimately connected with both of these two classes that it is not easy to consider one apart from the other.

1. ECONOMY OF LABOR.

For the successful accomplishment of the operation of puddling it is necessary to bring the molten metal into contact with the solid oxides by agitation effected either by human or mechanical agency upon a stationary hearth, or by giving motion to the whole body of the furnace. One of the first attempts for lessening the labor of the puddler is recorded in a drawing at Dowlais which has been traced back to the year 1834. It is a reverberatory furnace with a revolving hearth, driven with a vertical shaft by bevel gearing. Whether this machine was ever used I do not know, but it is of interest as showing that most of subsequent improvements are not new in principle. Coming to more recent times, we have the Richardson process of blowing air into the molten bath through a tubular rabble. The advantages claimed for this method are that it hastens the boil, reduces the labors, and produces a tough metal of uniform and high quality. After the iron has come to a boil the rabble is withdrawn and the working continued in the ordinary manner. I believe this process has never been used in this country and but sparingly in England.

Morgan's puddling machine consists of a reverberatory furnace of the usual form, which has an opening in the roof through which a vertical shaft is lowered with a horizontal arm. The shaft is set in motion by suitable machinery, and the arm revolves in the furnace, doing away with the labor of the puddler and helper until the heat is

ready for balling, when the shaft and arm are withdrawn, the opening in the roof closed, and the balling proceeded with in the usual manner. The wear and tear connected with this method must be enormous, and the results, I should think, not very satisfactory.

Griffith's and Whitham's devices are similar in idea, but different in mechanical details. Their object is to give an oscillating movement to a rabble of the ordinary shape by means of machinery, the puddler or helper merely guiding the rabble. The balling is accomplished in all cases by hand labor. None of the above-mentioned improvements do away with the skilled workman, but merely lessen the laborious work of the early stages of the heat, which requires brute force rather than experience.

In a work by Kohn upon the Manufacture of Iron and Steel will be found more detailed statements concerning these processes. That any of them has proved satisfaetory I question. One of the imperfections common to them all is the difficulty of keeping the raw iron from gathering in the crevices of the fix and settling on the bottom and in the corners of the furnace into which the rabble does not enter, leaving the furnace at the conclusion of the heat in a very dirty condition.. We all know the importance of a thorough working of iron in the jambs of a furnace, as it is there that the metal begins to gather when coming to nature, requiring careful working for good results. Another serious objection which may be advanced against these processes is, that they require the same skilled workmen to operate them as are needed for the old style of hand puddling. No increase in the number of heats is obtained, for the men, instead of encouraging experiments, look upon them with great distrust as inimical to their best interest, and when a workman and his tools do not agree good results cannot be expected.

About 1867 a change in the direction of improvement took place, and it was reserved for an American, Samuel Danks, to have the boldness to propose an entire revolution in the puddling process. The Danks furnace was the first rotary furnace to be put into successful operation, although its success was not assured until many improvements and alterations were made upon the original designs.

In England this same idea was elaborated, and several machines were brought out differing in details. The one of most novel construction was the Godfrey-Howson furnace, which had but one -opening into which the heat enters and the products of combustion

escape, a blowpipe on a large scale being substituted for the ordinary fireplace. Later, in this country, we find a rotary furnace designed by the Edgemoor Iron Company, of Wilmington, Del., worthy of mention from the fact that this company is at present equipping their workswith these furnaces, which would seem to indicate great confidence upon the part of the proprietors in the success of the rotary process.

2. ECONOMY OF FUEL.

In the utilization of coal for puddling two methods are employed. The one in almost universal use, where coal is directly burned on the grate of the furnace, is irrational and wasteful. The other method,. consisting in the conversion of the coal into combustible gases, which are burned on the hearth of the furnace, though more economical and rational, is but seldom used.

The attempts which have been made to improve the old system may be divided into two classes: First, those having for their object the prevention of smoke by feeding the coal below the surface of the fire, which is always kept bright. The mechanical devices for accomplishing this object are found in the Frisbie & Sweet furnaces. The system has not come into general use. An objection in the case of coal forming clinkers is, that the clinkers are forced to the top of the fire.. Second, those having for their object the utilization of the volatile. matters of the coal by a partial coking of the coal before it reaches the fire. This is effected by the employment of a separate magazinein connection with the fireplace. The gases from the coal are caused to pass over the fire and are there burned. Of this variety of furnace may be mentioned the Wilson furnace, and of more recent date the Price furnace, which has given very good results.

When we consider, however, the cost of introducing these improved. furnaces, and the trouble and annoyance of teaching workmen to use them, it is evident that we might just as well go a step further and introduce the gas system in its entirety. The great advantages to be gained in the use of gas in puddling are well known. We may distinguish here two systems, the continuous-acting furnace, of which Swindell's furnace is an example, and the well-known Siemens regen-erative furnace.

Of the use of water-gas in the place of the ordinary generator-gasit is too soon to speak, but reference may be made in passing to the astonishing results said to have been obtained at Washington by the