The Ionization of Potassium Vapor by Light

Photo-ionization of potassium vapor by ultra-violet light to λ 1850 {}.---(1) Jet method. To insure that the light was not appreciably absorbed before it reached the vapor under observation, and to reduce thermionic and photoelectric emission to a minimum, a tube was used in which a jet of potassium vapor was directed from a heating chamber into the vacuum observation chamber (pressure about 10^-4 mm.) and was there condensed on the walls, while a carefully diaphragmed beam from a quartz mercury arc was directed through the jet into a conical light trap. The presence of positive ions in the illuminated vapor was determined bv means of the usual accelerating and retarding electric fields. (2) Variation with wave-length. By the use of a series of screens, the long wave-length limit for ionization was found to be between 2,800 and 3,100 {}., in agreement with the limit, 2,856 {}., calculated from the ionization potential by means of the quantum relation. No positive ions were obtained when the radiation was limited to wave-lengths greater than 3,100 {}. or when the jet was intercepted by a sliding shutter. As the wave-length was decreased below 2,800 {}., the ionization rapidly increased in amount; being apparently a continuous function of the wave-length. (3) Variation with pressure of the vapor in the jet was found to be linear. (4) Proportion of exposed atoms ionized was computed to be roughly 1 in 10¹⁰ for an exposure time of 10^-5 seconds or 1 in 10⁵ per second for continuous exposure.Photo-electric emission from potassium metal.---The electrodes, being coated with potassium, gave a photo-electric emission as a result of scattered light, which had to be corrected for in the measurements for the vapor jet. The long wave-length limit occurred quite consistently at 4,500 {}.; but there were indications of increased sensitivity below 3,000 {}. These results agree with the view that the limit is the same as for the vapor, provided the emissions observed for wave-lengths 3,000 to 4,500 {}. are due to impurities. A rough calculation indicates that the proportion of exposed atoms ionized was about the same as in the case of the vapor, for the same light intensity.Series of ultra-violet absorption screens, transmitting to various short wavelength limits from 4,000 to 2,300 {}., including various solutions and glasses, are described and a spectrogram showing the transmission of each is reproduced.