The Luminescence of Air , Glass and Quartz under a - Particle Irradiation

The luminosity of a zlOPo source has been detected using a photomultiplier technique. It is also found that photons are produced in air, glass and quartz when these materials are irradiated by a-particles. It is shown that the ' scintillations ' observed by Richards and Cole probably originated in these materials, and not in the various thin films to which they were ascribed. T is well known that zlOPo and other cc-particle sources are feebly self-luminous. The spectrum of this luminosity was first examined with a quartz spectro-I graph by Sir William and Lady Huggins (1903, 1905 and 1906) and found to coincide with that of nitrogen. They concluded that the effect was due to a-particle excitation of nitrogen occluded in the source, and in the air irradiated by the source. This has been confirmed by later observers, the most recent being Ortner and Salim (1952). They photographed the visible spectrum of the light emitted from a zlOPo source, and from the air near the source. They identified the various lines and bands in the two spectra with those from nitrogen, and also oxygen. This luminosity and similar effects have been investigated using an EM1 type 5060 photomultiplier tube, which has a 1 cm diameter semi-transparent photo-cathode, deposited on the inside of a 1.5 mm thick flat glass window. A 2.4 mc 'loPo source, deposited on the end of a 0.9 mm platinum wire, was located at the end of a brass collimating tube, 15 mm long and 1.5 mm internal diameter, giving a narrow emergent beam of about 4 x los a-particles per minute. The source was mounted in air directly above the photo-cathode, on a micrometer screw, so that the distance from the tube could be varied. The output from the photomultiplier was fed through a cathode follower, linear amplifier and discriminator to a fast scaling unit. Throughout the experiments all surfaces were kept clean and free from materials not specifically mentioned. With the source in position, a large number of small amplitude pulses, greatly exceeding the noise count, was observed. The integral pulse-height distribution curves gave no indication of a plateau, and comparative measurements were therefore made of the counting rate at constant gain and a fixed discriminator bias. The number of counts observed at this bias was approximately 0.18 of the total number of counts, extrapolated to zero bias. Figure 1 shows the observations on the …