The realization of the great heating value of the top gases of the blast-furnace was very slow in taking useful shape. For centuries, furnaces ran with their tops wide open, discharging the gas into the air, with which it united in a great column of flame, visible for miles by night, even in the small furnaces of that early day.

Even after the invention of the steam engine, the possibilities of blowing furnaces by steam power were not practically realized for several decades and furnaces continued to be located primarily with reference to a suitable water power for blowing. In the same way after the invention of the hot-blast in the third decade of the nineteenth century the stoves were not at first heated by gas from the furnace, but were independently fired with coal.

Later a small chamber filled with iron pipes was built over the top of the furnace and the flame from the open top passed up through this, heating the blast which passed through the pipes. Still later flues were built into the brickwork near the top but just below the stock line and connected with a powerful chimney whose base was approximately level with the top of the furnace; this chimney drew some of the gas from the furnace, naturally much mixed with air drawn in from above, and the mixture burnt on its passage from the furnace top through the boilers and stoves to the chimney.

The final step was taken by closing the top of the furnace with a bell and forcing the gas by the pressure from below out through the flues into a pipe leading down to the ground level, where the stoves, boilers and stacks were located as previously described. This pipe early received and still retains the name of the "downcomer".

These top gases in Lake ore practice contain about 50 per cent. of the total thermal value of the fuel charged. In districts where leaner ores are used and more fuel is consequently required per ton of iron the gas is correspondingly richer and contains a proportionately larger percentage of the thermal value of the fuel. In Lake ore practice an average analysis is about as follows: carbon monoxide 25 1/2 per cent., hydrogen 3 per cent., carbon dioxide 12 1/2 per cent., nitrogen 59 per cent. The thermal value comes entirely from the first two components.

The hydrogen comes partly from the moisture in the blast which is dissociated in the hearth, and partly from the moisture of the charge which is decomposed in the upper levels of the furnace. This gas under standard conditions of 32° F. and 30 inches of mercury contains from 94 to 102 B.t.u. per cubic foot and has a specific heat of about 0.02 per cubic foot, so that it contains about 2 B.t.u. of sensible heat for each 100° F. above that temperature.

This gas is then somewhat more than half as high in heating value per cubic foot as ordinary producer gas. This, with the enormous volume produced (some 140,000 cu. ft. per ton of iron in good practice and more where the practice is poorer), give it enormous commercial value, so great in fact as to equal a considerable percentage of that of pig-iron itself.

This gas having been in intimate contact with the charge in its course up through the furnace, and having carried off all the moisture of the charge not dissociated, necessarily contains much moisture and picks up much fine dust from all the different components of the charge. It also carries a fume consisting of solids volatilized in the hearth of the furnace which sublime into the solid state, in the form of an impalpable powder, as the gas cools in its progress toward the furnace top.

It is this fume that gives the gas its characteristic appearance, so that an experienced furnaceman can tell more quickly by looking at the gas discharging from the boiler and stove stacks than by any other way if his furnace is working "hot" or "cold," since a relatively few degrees drop in hearth temperature make a great difference in the amount and appearance of the fume thrown off.

This fume consists mainly of silica, alumina, lime, potash from the ash of the coke, and iron, and in case the charge contains much manganese, this element is present in the fume to a great extent, changing the white color of the fume of a good-working furnace on ordinary iron to a light yellow when running on "spiegel" and to a heavy yellow or brown when running on ferromanganese.

In the early days of slow driving the velocity of the gas current through the furnace was small and the quantity of fine ore in the charge was almost negligible in comparison with the quantity of fine ore used to-day, which runs from 50 to 100 per cent. Mesabi in the majority of American practice.

Under those conditions a single, comparatively small outlet at one side of the furnace top was sufficient and the quantity of dust carried by the gas was extremely small, therefore a single pipe of moderate size running straight or almost straight down and branching directly off to the stoves and boilers was all that was necessary to handle the gas. This constituted but a small, almost an insignificant fraction of the plant as a whole. Conditions at the present are vastly different from this in three respects:

First, the velocity of the gas through the furnace is several times greater than it was in those early days. The carrying power of the fluid current increases at an enormously rapid rate with an increase in velocity. I have seen the statement that this increase was as the seventh power, which would mean that doubling the velocity would increase the quantity of material carried 128 times. Whether this be true or not it is certain that the rate of increase in carrying power is extremely high.

Second, the ore used instead of being practically all hand-prepared lump consists very largely of fines, much of which is capable of passing through a 100-mesh sieve.

Third, we have come to realize the enormous cost of attempting to utilize gas in its dirty condition for stoves and boilers, even leaving out of consideration gas engines which require for their successful operation gas cleaner than the air we breathe.

These three causes have had an enormous effect on the gas-handling system of the furnace, which has in recent years become one of the largest and most conspicuous features of the whole blast-furnace plant.