121. VARIATION OF VACUUM BY DISCHARGE AND BY REST.—Prof. Pupin was among the first to test the efficiency of external electrodes for generating X-rays. Independently of the quality of the glass and of the kind of pump and of the presence or absence of phosphoric anhydride, the following peculiarities were noticed, which Edison attributed to a kind of atomic electrolysis. § 47. 80 per cent. of the lamps exhibited the phenomena as follows: First, such a high vacuum was obtained by the pump that the line spectrum disappeared and pure fluorescence and generation of X-rays at a maximum occurred. The lamp was then sealed off. After three or four hours of rest, the vacuum deteriorated, so that striae and other characteristics of low vacuum were obtained when connected up in circuit, but upon continuing the current, the high vacuum gradually came back, the line spectrum vanished, and suddenly X-rays were generated. Again the bulb was left at rest for 24 hours, after which X-rays could not be generated until the discharge had been continued for 4-1/2 hours. [Illustration: DISCHARGE TUBE, FIG. 2. § 120. ] 122. EXTERNAL ELECTRODES DISCHARGE THROUGH HIGHER VACUUM THAN INTERNAL.—A vacuum that was so high that no discharge took place with internal electrodes was made luminous by the use of electrodes on the outside of the glass bulb. Then he made the vacuum so high that even with a 12-inch spark from Leyden jars, no discharge took place with external electrodes, and the tube was dark, this part of the experiment indicating another limit at which an extremely high vacuum is not a conductor and appearing to overthrow, as Edison intimated, Edlund’s theory that a vacuum is a perfect conductor. § 25. 123. DEPOSIT ON GLASS FROM ALUMINUM ELECTRODE.—It has always been common to employ aluminum for electrodes in vacuum tubes, on the ground that no deposit took place, and therefore no blackening, nor whitening of the glass wall. § 40. Edison observed also that no blackening was visible, but stated that his glass blower, Mr. Dally, upon breaking the bulb and submitting the interior surface of the glass to an oxydizing process, the oxide of aluminum was so thick as to be opaque to light. With magnesium, also, a mirror was produced, of a lavender color, by transmitted light. In the case of aluminum, he was able to obtain a visible spot at the phosphorescent portion, but only after a great many hours of use. See cut from a photograph of a discharge tube used for several months by Prof. Dayton C. Miller, and having a heavy aluminum deposit opposite the aluminum cathode. With the increase of the deposit, the power of the X-rays diminished, but, he thought, not on account of the absorption, but because, “through lack of elasticity at the surface.” [Illustration: DISCHARGE TUBE, § 123. ]