This gas containing, as has been said, roughly 50 per cent. of the total energy of the fuel charged, supplies the heat necessary for all the operations around the plant outside the furnace itself. Its first and foremost use is for the purpose of heating the air, with which we have no present concern. This, in good practice, consumes about 25 per cent. of the total gas. In first-class modern practice another 12 to 16 per cent. is required for blowing power, and 5 to 10 per cent. for miscellaneous purposes around the furnace, hoisting, pumping water, etc., and loss, leaving about 50 per cent. of the gas, or nearly 25 per cent. of the total energy of the fuel charged, available for purposes entirely independent of the furnace.

The steam plant at the blast-furnace has been until very recent years a poor and inefficient one, measured by the standards prevailing at similar periods in other lines of industry. The reason for this is not far to seek. When the closed-top furnace was introduced, and it was found that the gas could be forced by this means to any desired point within reason, it soon developed that the furnace possessed a source of power ample for its needs, even with the cheapest and simplest type of steam plant.

Steel mills at that early day were not generally built immediately in conjunction with furnace plants as they are to-day, and the latter, therefore, did not generally have any local market for the suplus power, and until the general introduction of electricity for lighting and power transmission there was no disposition for power except such as could be developed immediately where it was to be consumed. Hence there was no use for surplus power which might be developed by the use of economical machinery, and as this machinery was higher in first cost, and more expensive to operate than the simplest and plainest type, the latter was in almost universal use until within the last fifteen or twenty years. Slide-valve or piston-valve engines with late cut-off and little expansion, using low boiler pressure, and running non-condensing were the rule.

The boilers likewise were of the cheapest, simplest and generally most inefficient types. Plain cylinder boilers with their enormous settings and proportionately excessive air leakage were the prevailing style until within the last twenty or thirty years. These boilers were literally plain cylinders, or sometimes a pair set side by side, from 3 to 4 feet in diameter and 40 to 60 feet long, entirely without flues. Sometimes there were two pairs, one above the other, the lower one joined to the one vertically above it by necks on each outside or "high" sheet, the whole suspended from two or three suspension points and exposed to the flame up to a height just below the water level in the upper cylinder, the gases making only one pass straight through to the chimney connection.

On account of their enormous water contents in comparison with their heating surface, these boilers possessed a great quantity of reserve energy, and steam pressure would not drop injuriously during a few minutes' suspension of the gas supply. They were also very easy to clean because containing no tubes. Men could be put inside them and chip off the scale with chisels or brick axes, which, of course, is impossible in other types of boilers except as to a small fraction of their heating surface.

Other than this these old types of boilers had little to recommend them, as the great cost of the setting and the high price per square foot of their heating surface made their cost per horsepower installed about as high as that of better types.

The great increase in the tonnage of iron produced in a furnace of a given size, which might almost be said to have culminated with the Duquesne revolution, and the increase in the size of the furnaces themselves, both acted in the same direction, that of requiring a higher blast-pressure to force the desired quantity of blast into the furnace. This in turn required more power, and therefore the practice which, had been adequate if not ample for the old conditions became inadequate for the new, because its consumption of steam was so great that the combustion of all the furnace gas available was frequently unable to supply the quantity required. When this occurred it was necessary to burn coal under the boilers, and such coal represented a dead loss.

During the same period the integration of iron and steel plants developed extensively. Furnaces were built at large steel plants, or steel plants at furnaces in order that direct metal from the furnace might be used at the steel plant with a great saving in expense for both, the furnace sending the iron in ladles to the steel plant and so avoiding the cost of casting and handling it, while the steel plant avoids the greater cost of remelting it.

The well-equipped furnace plant having generally, if not always, a considerable excess of steam at its command, advantage was taken of this fact by running a steam pipe from the furnace boilers to the steam mains at the steel mill, the latter using enormous quantities of power, but having no source of supply of its own except by burning coal or other commercial fuel.

The great value of the power so supplied became more and more appreciated and the power plant at the blast-furnace was gradually improved so as to leave a larger margin of gas available for the steel mill, and this process has now gone on until some of the finest and most economical power plants in the world are those employed in blowing blast-furnaces, while the subject of power development for' this purpose has enlisted the closest attention in Europe and more recently in this country as well.

It is a rather interesting fact that this gradual progress from the utter indifference which came from having a prodigal supply of power and no use for the surplus, to the present condition in which the expenditure of every thermal unit is watched with a jealous eye, lagged a little behind the development of the steam boiler from the relatively crude types, which I have described, and which were in practically universal use at one time in all other industries as well as at the blast-furnace, into the modern water-tube types. In other industries after the single-flue and two-flue boilers, came a gradual increase in the number and decrease in the size of the flues, the multitubular boiler, which was for so many years the standard in good steam practice. Then came the later evolution, forced by rising steam pressures, from the horizontal tabular boiler to the water-tube boiler, which became well established about the time the demand developed for more economical boilers for blast-furnaces.

In this way blast-furnaces skipped that period of development during which the horizontal tubular reigned supreme in most industries, and passed immediately from the cruder types to the modern water-tube boiler.