tering upon the inquiry, he soon found himself involved in preliminary difficulties with the spark itself. His phosphorescent investigations remain yet to be carried out, but the results obtained relative to the electric flash are of extreme interest. The full account of the research is given in a series of papers published in Silliman's Journal, and, if the reader finds the following statement insufficient in its details, he will know where to go for further explanations.
Since the time of Franklin, the lightning-flash has been regarded as a gigantic electric spark produced in the atmosphere; the inquiry, therefore, involved the nature of the meteorological discharge, as well as of the spark artificially produced. Various attempts to determine the duration of lightning have been made, with varying results. Faraday observed it, without any instruments for measuring the time, which seemed to last for a second, but he was doubtful if part of the effect was not due to the lingering phosphorescence of the cloud. Decharme observed the lightning-flashes from a distant storm, which also appeared to last for from a half to an entire second. Prof. Dove employed a revolving disk with colored sectors, and satisfied himself that single flashes of lightning often consisted of a number of instantaneous discharges. It is well known that, when a rapidly-moving train of cars is illuminated at night by lightning, it seems to stand still, that is, the duration of the flash is so brief that no motion of the train is perceptible while it lasts. The wheels are sharply defined as if perfectly motionless, but if they had a blurred aspect we should know that the illumination lasted sufficiently long to render the motion perceptible. Prof. Rood extemporized a simple contrivance for observing lightning, which acted upon this principle. It consisted of a white card-board disk, five inches in diameter, with a steel shawl-pin for an axis, on which it was made to revolve by striking the edge. He traced black figures near the circumference of the disk, and when it was in rapid motion these figures were sometimes seen as sharply as though they had been stationary, although they were often blurred as though the disk had moved through a few degrees during the act of discharge. He then cut narrow, radial apertures into the circumference of the disk, and observed the lightning through these openings. Here, again, the apertures were sometimes seen quite unchanged, but they were more frequently elongated into well-defined streaks some degrees in length. He afterward measured the average rate of rotation imparted to the disk in this way, and arrived at the conclusion that the lightning-flashes on the occasion referred to had a duration of about one five-hundredth of a second. Dissatisfied with the roughness of these observations, Prof. Rood arranged a small train of toothed wheels driven by a spring, which rotated a circular pasteboard disk with four open sectors. This instrument gave more regular and precise results; and, while it was shown that the flash sometimes lasts for a whole second, the suggestion of Dove was clearly verified that each flash "consisted of a consider-
