The Origin of Large Cosmic-Ray Bursts

Abstract

Integral size-frequency distribution curves for cosmic-ray bursts of more than 100 particles were obtained by using an ionization chamber shielded by 1.25, 12, and 35 cm of iron. These distribution curves can all be represented by an inverse power law with exponent = 2.0\ifmmode\pm\else\textpm\fi{}0.2. The transition curve plotted from these data shows a pronounced maximum. Five G-M counter coincidence sets were arranged to register on the ionization chamber trace so that coincidences between bursts and air showers could be measured. For a thickness of 1.25 cm of iron, about 85 percent of the bursts were coincident with air showers, while for 12 cm of iron, this fraction was about 20 percent and for 35 cm of iron, only 5 percent of the bursts were observed to be coincident with extensive showers. A discussion is given to account for the origin of the bursts which were not coincident with air showers. For the analysis of the transition curve three different types of bursts are considered: bursts due to extensive atmospheric showers, bursts produced by narrow air showers or single high energy electrons, and bursts produced by mesotrons (knock-on and bremsstrahlung processes). Data for bursts of more than 500 particles under 12 cm of lead are compared with the corresponding data obtained under 35 cm of iron in order to determine the dependence of burst frequency on atomic number; the results are in substantial agreement with production of these bursts by bremsstrahlung of the mesotron. Integral size-frequency distribution curves are also plotted for data obtained during a five-year period at Huancayo (3350-meters elevation) and at Cheltenham (72 meters) for bursts under 12 cm of lead. The Cheltenham data are compared with the theoretical calculations of Christy and Kusaka for burst production by mesotrons of spin 0, \textonehalf{}, and 1 and it is concluded that either the spin 0 or the spin \textonehalf{} curve fits the experimental data but the evidence definitely excludes spin 1. The altitude dependence for burst under 12 cm of lead shows that the ratio of the burst frequencies at Huancayo and Cheltenham is constant for bursts up to 2400 particles, but increases sharply for still larger bursts. This increase is discussed on the basis that bursts under thick shields at higher altitudes may be caused either by extensive showers or by possible spin 1 mesotrons having so short a mean life that most of them fail to reach sea level.