spark was absolutely instantaneous, its image thrown upon the ground glass would be exactly the same, whether the mirror was motionless or was revolving at the highest speed. But, if the spark had an appreciable duration, its image would be prolonged or drawn out into a streak, the length of which must depend upon the time of discharge. The rate of the mirror's rotation being known, also the distance, m i, and the length of the streak, it was easy to calculate the total duration of the spark.
Prof. Rood now had the subtile agent he was pursuing pretty effectually in his grasp, and the results that came out were very striking. The ordinary spark was found to be a highly-complex effect; to consist of diverse and successive elements, and, in fact, to have its periods and orderly history just like the geology of the globe. But, while the "vast durations" of Lyell and Dana are vague and inferential, these infinitesimal periods could be demonstrated with the greatest exactness. The previous discordant results were reconciled, Feddersen being justified in assigning a longer period for the total duration of the spark, and Wheatstone's time holding true of its elements.
With a Leyden jar of about a quart capacity (114.4 square inches of coating), and all the connections as short as possible, so as to offer the least amount of resistance to the electric flow, with brass balls as electrodes, with a striking distance of about the twenty-fifth of an inch, and the velocity of the mirror up to 223 per second, the image of the spark thrown upon the ground glass and viewed by the naked eye was drawn out into a streak one and a half or two inches long, the length, however, varying with the speed of the mirror. The aspects of the image are represented in Fig. 5. The first part was pure white, which shaded into a brownish-yellow tint, passing on into a pretty distinct green. When a polished plate of glass was substituted for the ground glass, and a small magnifier was used to observe the image, a series of bright points, on each side of the streak, became visible, in the positions indicated by the dots in Fig. 5. With high velocities, this succession of points was beautifully developed, and it consists of a series of separate discharges following the first. It was thus found that the