of the strong electromotive force, the oxygen, and SO,, they rapidly disintegrate. Platinum seems to be the only resource. But platinum is expensive, and unless roughened by an electro-deposit of more platinum on its surface, offers great resistance, by reason of the retention of oxygen upon its smooth surface.

As it was desirable to deposit three pounds of copper per hour, it is necessary to use three pounds of zinc in each of three cells, or nine pounds in all, for each three pounds of copper produced. This was an expense of $1.12 per three pounds of copper, besides sulphuric acid, and labor, and waste, amounting to nearly as much more; rolled zinc suitable for batteries costing 12 cents per pound. This makes rather expensive copper; say 60 cents per pound.

The expense by dynamo-electric machine was figured as follows: Force, or energy, of 9 lb. zinc, and equivalent of H,SO,, less force of equal amount of copper, is 9,105,469 foot pounds per hour, or about

4.6 horse power.

This is the amount of available force necessary under the conditions. A very few, if any, dynamo-electric machines utilize more than 50 per cent. of the force in foot pounds applied to them ; double that number of foot pounds of force must therefore be applied, or 18,210,938 foot pounds per hour, equal to 9.2 horse power. This, with coal, attendance, etc., from an ordinary steam-engine, would cost 42 cents per hour for 3 lbs. copper, or 14 cents per pound; coal costing $8 per ton in the locality.

I did not deem it advisable to place two or more depositing cells in series, since not only the resistance increased with each addition, but also the counter-electromotive force, so that would necessitate a change in the construction of the machine, so as to increase its electromotive force.

While canvassing the merits and demerits of iron as a soluble anode for the purpose, I tried a plan for the use of iron in reducing the copper, which proved very successful. After a short consideration the question arose, Why use a current of electricity when iron. alone is sufficient to reduce copper from the solution? If I apply the current with an iron anode, copper will still be reduced upon it by local action, and I will have the same fine powdery deposit, the same formation of insoluble basic salts of iron mixing with the copper deposit, and the expense for producing the electric current. As these objectionable results seemed to arise from the direct contact and association of the iron, copper, and copper solution, as well as the iron solution already present and synthetically formed, I decided

to try to separate them, and did so by placing iron in a less than saturated solution of sulphate of iron (free from copper), contained in an ordinary porous cell, such as is used in various galvanic batteries. This porous cell and contents I placed in a large vessel containing some of the copper liquor and a sheet of metallic copper. I connected the iron and copper, external to the solutions, by means of a clamp, and the work commenced. In 36 hours the liquor was completely freed from copper, which was deposited upon the copper sheet as a beautiful velvet-like coat, pure, reguline, and coherent.

No formation of basic salt of iron; no copper powder; none of the defects of the ordinary precipitation of copper by means of iron. The expenditure of iron was but the equivalent for the copper deposited, namely, 56 of iron for 63.5 of copper. All the attendance requisite was for the occasional removal of some of the nearly saturated solution of iron from the porous cell, filling the space made with water.

There was then procured ten of the largest porous cells obtainable, ready made, and set up in series, that is, the iron of one connected with the copper of the next vessel, and so on through all, forming a ring or closed circuit. The result was the same, all the copper deposited in 36 hours. Eighteen large porous cells have been made, measuring 12 inches in diameter and 32 inches long, and large-sized oil barrels will be used for the vessels to contain the copper liquor. A modification of this arrangement calculated for the continuous treatment of cupriferous solutions places the vessels so that the solution may run from one to another through as many as may be needed to complete the deposition of copper. A low percentage of copper increases the speed of exhaustion. Scrap iron may be placed loosely in the porous vessel, and may be added from time to time to take the place of that which has been dissolved. It is necessary to remove portions of the solution of iron as it approaches saturation, in case it be desirable to save that material, and fill again with water, or part can be displaced by water, allowing it to overflow into the outer vessel. Speed of operation, as regards quantity, may be gained by increase of size and number of vessels.

In this way any concern, whether producing a gallon of copper solution, or thousands of gallons daily, may produce fine, merchantable copper by inexpensive apparatus at, say one cent per pound, more or less, as scrap iron may be worth more or less than $20 per ton.

NOTES ON FIRE-BRICK STOVES FOR BLAST FURNACES. BY JOHN M. HARTMAN, PHILADELPHIA.

(Read at the Wilkes-Barre Meeting, May, 1877.)

Two systems are used for heating air in blast-furnace operations: I. The double surface system, in which a cast-iron pipe is heated on the outer surface, and, at the same time, heats the blast from its inner surface. This is simple in operation and gives a continuous effect, but is limited by 1100° F. as a maximum temperature of the blast.

II. The single surface system, by which large surfaces of fire-brick are heated, and air passed over the heated surface, absorbing the heat and carrying it on to the furnace. This system is more complex than the double surface, as it involves the reversing of the air and gas every hour and a half.

The single surface system has two advantages:

1. The blast can be heated to a temperature of 1800° F.

2. The stoves are indestructible.

From recent experience it has been found that 1300° F. to 1400° F. is the best average temperature for economy of coal for safe working. This is equivalent to a saving of 1 to 2 cwt. of coal per ton of iron over the extreme limit of cast-iron stoves.

Independently of this, it is a strong point in favor of this system that the blast can be raised to a temperature of 1800° F. within an hour when the hearth is getting cold. All furnacemen know the value of a hot hearth for quality and quantity of iron. Cooling of a hearth occurs from leaky tuyeres, scaffolds, or heavy burden.

When there is not sufficient coal at the tuyeres to seize on the oxygen of the entering air and convert it at once to carbonic oxide, there will not be enough heat to liquefy the cinder. Black cinder and poor iron are the results. The remedy is additional heat from the blast. If the stoves will give 1800° in place of 1100°, it is obvious that the furnaces will get around sooner, and without waiting for a change of burden at the tunnel head to bring extra coal to the hearth. The heat absorption caused by a leaky tuyere will chill the hearth and drive the zone of fusion higher up in the furnace. This loss must be supplied, and the calorics lost to the hearth must be regained before good iron can be made.

Take another case: A furnace carefully burdened on No. 1 iron, during a spell of damp weather, goes on to No. 2 or No. 3. Heat is lost to the furnace through absorption by the moisture, and less burden must be carried in order to get back on No. 1. When the

weather clears up, this burden is too light, and, unless changed promptly, silicized iron is produced.

The difference between a dry and wet day in heat absorption is equivalent to two tons of coal per day, when using 10,000 cubic feet of air per minute. With stoves of good capacity an additional amount of air can be poured into a furnace to maintain its temperature and avoid change of burden or grade of iron. In the case of a furnace working a light burden and doing a carbon duty of 2.3 or 2.4, the difference above given would not be so appreciable, as there would be a large surplus of heat above actual requirements; but, when running on a carbon duty of 2.7 to 3, all these small differences must be closely watched, or the running of a furnace on light burden in these latter days will not pay.

For some years past we have been collecting results of brick stoves, and declining to give up the iron stoves until we could find good results from actual workings of the brick stoves both at home and abroad.

The Cedar Point Iron Company have demonstrated that they can save fuel by the brick stoves, and we find the failure at other places is due to the stoves being too small. The superintendent of the Cedar Point works, with a foresight not always found, put up large stoves, and to this is due his success so far as hot blast is concerned. They use four stoves, 22 by 30 feet, having a total heating surface of 35,200 square feet. The average temperature is 1375° F., with a maximum of 1750° F. They have four square feet heating surface to each cubic foot of air passing per minute, and get a carbon duty of 3.13 on a basis of No. 3 iron. They change a stove on the furnace every two hours. The gas escapes from the stoves at a temperature of 200° F.

At Rising Fawn, Georgia, with three stoves, 18 by 30 feet, and having 17,400 square feet of heating surface, they average but 1000° F. with 1200° F. for a maximum temperature. They have 2 square feet surface for each cubic foot of air passing per minute, and get a carbon duty of 2.35. The escaping gas goes off at a temperature of 650° F., which is a loss of 450° F. in the gas, and a loss of 375° F. in the blast. They change stoves every hour. This shows that economy is only to be obtained by using plenty of surface to absorb heat.

In stoves where brick walls are used to absorb heat, the thickness prevents the heat from the interior of the walls becoming available. It has been found that, when using 9-inch walls, and changing every

two hours, the exterior of the walls would become hot within three hours or so, if the heat was reduced to a minimum and the stoves were shut up. This was repeated twice in succession. This shows the necessity of thinner walls and increased heating surface, as the storage of the heat in the interior of the wall is not available in the time required to lower the temperature to the minimum. The slow conducting power of the fire-brick is the cause of it. The valves of the stove require the attention of a careful man.

Where a bell and hopper is used the escaping gas goes off at so low a temperature that there is no danger of harming the gas valve. When the heating surface is small, and the escaping gas goes off at a high temperature, the chimney valve must be cooled with water. The hot-blast valve is cooled either by water or cold blast. There are objections to the use of water, as it often causes explosions in case of leakage. It is advisable to use as few valves as possible, to prevent leakage and handling. The cleaning of the fire-brick stoves is no more difficult than that of cast-iron stoves. Scraping the walls and blowing off the dust by blast are the methods employed.

After a careful comparison of fire-brick stoves, we have taken up the Siemens-Cowper-Cochrane stoves as being the most simple and inexpensive in construction. These are the original Cowper stoves, modified by Dr. Siemens and Mr. Cochrane. These patents cover the use of all fire-brick stoves. The stoves consist of a wrought-iron shell, lined with, say, 18-inch brick; inside of the shell is a vertical, circular flame flue, say 4 feet in diameter by 14 inches thick. The flame flue is eccentric to and built against one side of the 18-inch lining. Around this flame flue are built the vertical regenerating cells. They are composed of split-brick, 1 inches thick on the edge, which leave a vertical opening of 33 by 33 inches from top to bottom. This cellular arrangement gives a large surface of contact, while the thinness of the walls admits of the heat being thoroughly abstracted, so that there is no waste storage, or heat stored that is not available in the 13 hours during which a stove is on the furnace. We propose three stoves, two on gas, one on the furnace, and to use 5 square feet heating surface to each cubic foot of air passing through the stove per minute. This will allow the escaping gas to go off at 150° F,, which is important for economy, as the gases are less rich in carbonic oxide and less in volume as the burden and hot blast are increased at the furnace. The pressure of the blast at anthracite furnaces is double that where coke is used, and hence twice the gas will be required to generate steam for anthracite furnaces as compared with

VOL. VI.-30