Mr. Fitzgerald, the chemist of the Acheson Company, pointed out to me a curious glassy cavity in one of the half-dismantled furnaces. "Here the heat was only a fraction of that in the core," he said. But still the fire-brick--and they were the most refractory produced in this country--had been melted down like butter. The floors under the furnace were all made of fire-brick, and yet the brick had run together until they were one solid mass of glassy stone. "We once tried putting a fire-brick in the centre of the core," said Mr. Fitzgerald, "just to test the heat. Later, when we came to open the furnace, we couldn't find a vestige of it. The fire had totally consumed it, actually driving it all off in vapour." Indeed, so hot is the core that there is really no accurate means of measuring its temperature, although science has been enabled by various curious devices to form a fairly correct estimate. The furnace has a provoking way of burning up all of the thermometers and heat-measuring devices which are applied to it. A number of years ago a clever German, named Segar, invented a series of little cones composed of various infusible earths like clay and feldspar. He so fashioned them that one in the series would melt at 1,620 degrees Fahrenheit, another at 1,800 degrees, and so on up. If the cones are placed in a pottery kiln, the potter can tell just what degree of temperature he has reached by the melting of the cones one after another. But in Mr. Acheson's electrical furnaces all the cones would burn up and disappear in two minutes. The method employed for coming at the heat of the electrical furnace, in some measure, is this: a thin filament of platinum is heated red hot--1,800 degrees Fahrenheit--by a certain current of electricity. A delicate thermometer is set three feet away, and the reading is taken. Then, by a stronger current, the filament is made white hot--3,400 degrees Fahrenheit--and the thermometer moved away until it reads the same as it read before. Two points in a distance-scale are thus obtained as a basis of calculation. The thermometer is then tried by an electrical furnace. To be kept at the same marking it must be placed much farther away than in either of the other instances. A simple computation of the comparative distances with relation to the two well-ascertained temperatures gives approximately, at least, the temperature of the electrical furnace. Some other methods are also employed. None is regarded as perfectly exact; but they are near enough to have yielded some very interesting and valuable statistics regarding the power of various temperatures. For instance, it has been found that aluminium becomes a limpid liquid at from 4,050 to 4,320 degrees Fahrenheit, and that lime melts at from 4,940 to 5,400 degrees, and magnesia at 4,680 degrees.