certain high degree of vacuum, formed behind the electrode, and prevented the use of very high potential, but by having the wire extend far into the tube and providing wrappers, the sparking was much less likely to occur. By proper adjustment as before intimated, Tesla could produce just about enough to compensate for the electrical increase of the vacuum. Another difficulty that presented itself in connection with high E. M. F. was the undue formation of streamers heretofore referred to, apparently issuing from the glass, and so often disabling it. He therefore immersed the discharge tube in oil as pointed out in detail hereinafter. The walls of the tube served to throw forward to the thin glass many of those rays that otherwise would have been scattered laterally. Upon comparing a long thick tube of this kind with a spherical one, the sensitive plate was acted upon by the rays in 1/4 the time with the tube. A modification consisted in surrounding a lower portion of the tube, with an outside terminal _e_, indicated in dotted lines in Fig. 1. In this way the discharge tube had two terminals. The greatest advantage probably in using a long tube, was that the longer it was, within the proper limits, the greater the potential which could be applied with advantages. As to the aluminum electrode, he noticed that it was superior, in comparison with one made of platinum which gave inferior results, and caused the bulb to become disabled in an inconveniently short period of time. 146. PERCENTAGE OF REFLECTED X-RAYS. He performed some preliminary experiments, testing roughly as to whether any appreciable amount of radiation could be reflected or not from any given surface. Within 45 minutes he was enabled to obtain clear and sharp sciagraphs of metal objects, and the same could have been obtained only by the reflected rays, because he screened the direct rays by means of very thick copper. By a rough calculation he found that the percentage of the total amount of rays reflected was somewhere in the neighborhood of 2 per cent. Prof. Rood, of Columbia University, N.Y., (_Sci._, Mar. 27, ’96.) by means of an experiment with platinum foil, § 80, concluded that the per centage was about .005, the incident angle being 45 degrees. He regarded this figure as the mere first approximation. Judging from Roentgen, § 85, Tesla, Rood and others, therefore, it seems to be established that the percentage of X-rays reflected is very small.