The “pour” is preceded by a shower of sparks, consisting of little particles of molten steel which are projected fully a hundred feet in the direction of the open mouth of the converter. In the line of this my apparatus was stationed in an open window, at a point where its view could be directed down into the converter on one side, and up at the sun on the other. This apparatus consisted of a long photometer-box with a _porte-lumière_ at one end. The mirror of this reflected the sun’s rays through the box and then on to the pouring metal, tracing their way to it by a beam visible in the dusty air (Fig. 57). In the path of this beam was placed the measuring apparatus, both for heat and light. As the best point of observation was in the line of the blast, a shower of sparks was driven over the instrument and observer at every “pour;” and the rain of wet soot from chimneys without, the bombardment from within, and the moving masses of red-hot iron around, made the experiment an altogether peculiar one. The apparatus was arranged in such a way that the effect (except for the absorption of its beams on the way) was independent of the size or distance of the sun, and depended on the absolute radiation there, and was equivalent, in fact, to taking a sample piece of the sun’s face _of equal size_ with the fluid metal, bringing them face to face, and seeing which was the hotter and brighter. The comparison, however, was unfair to the sun, because its rays were in reality partly absorbed by the atmosphere on the way, while those of the furnace were not. Under these circumstances the heat from any single square foot of the sun’s surface was found to be _at least_ eighty-seven times that from a square foot of the melted metal, while the light from the sun was proved to be, foot for foot, over five thousand times that from the molten steel, though the latter, separately considered, seemed to be itself, as I have said, of quite sunlike brilliancy. [Illustration: FIG. 57.--PHOTOMETER-BOX.] We must not conclude from this that the _temperature_ of the sun was five thousand times that of the steel, but we may be certain that it was at any rate a great deal the higher of the two. It is probable, from all experiments made up to this date, that the solar effective temperature is not less than 3,000 nor more than 30,000 degrees of the centigrade thermometer. Sir William Siemens, whose opinion on any question as to heat is entitled to great respect, thought the lower value nearer the truth, but this is doubtful. [Illustration: FIG. 58.--MOUCHOT’S SOLAR ENGINE. (FROM A FRENCH PRINT.)] * * * * *