The HEAO X-ray experiment has shown that remnants of exploded stars, supernova remnants, are enriched in the medium-heavy elements such as oxygen, silicon, and sulfur. These observations provide strong confirming evidence for the evolution of stars, according to which medium-heavy elements are synthesized in massive stars and thrown into space by the explosion of these stars.
Another, more elusive prediction of stellar models is that many of the heavy elements, for example, arsenic, strontium, radium, and uranium, are built up very rapidly from lighter elements during the actual explosion. Add to this the long-standing belief that cosmic rays are also produced as a result of a supernova explosion, and you have a testable prediction, namely, that cosmic rays should show an enrichment of the rapidly processed, or reprocess, heavy elements.
Several cosmic ray experiments before the HEAO program seemed to support this prediction; thus, when the results from the HEAO cosmic ray experiments came in, many astronomers were in for a surprise. The Heavy Nuclei Experiment and the Cosmic Ray Isotope Experiment showed convincingly that there is no significant enrichment of reprocess elements in the cosmic rays. The relative amounts, or abundances, of each element are quite similar to the relative abundances of the elements in the solar system. In addition, the HEAO C Gamma Ray Spectroscopy Experiment set limits on the abundance of certain nuclei that should have been produced in supernova explosions over the lifetime of the galaxy. These limits are significantly below theoretical predictions.
What are the implications of these findings? They mean that either the reprocess buildup of heavy elements does not occur in supernova explosions, or that cosmic rays do not originate in the remnants of supernovas. Recent theoretical and observational work suggests that both these statements are true.
Recent theoretical work has shown that the reprocess is probably not very efficient in supernova explosions after all, that the most likely place for it to occur is in the core of a star about the size of the Sun. Such stars are very common and will never become supernovas because their masses are not large enough. However, when the core of a solar-type star exhausts its hydrogen fuel, it collapses until it ignites thermonuclear reactions based on the fusion of helium nuclei. This ignition, which is called the helium flash, has the character of a thermonuclear explosion and apparently occurs  several times until the core of the star settles down. The helium flashes are not sufficiently violent to blow the star apart, but they can apparently produce large enough fluxes of neutrons to make the reprocess work. Thus, the standard dogma that all the heavy elements are produced in supernova explosions may have to be revised.
What about the idea that cosmic rays are produced as a result of supernova explosions? As we shall see, the standard view may be in trouble.
First of all, remember that cosmic rays are high energy particles. Their speed is very close to the speed of light, the ultimate speed that any particle can attain. How cosmic rays reached such high speeds is the central problem in understanding the origin of cosmic rays. There are many proposed solutions to this problem, but no accepted ones. All the proposals have a common thread, namely, that the force that accelerates the cosmic rays to high energies is the electrical force.
Charged particles moving through a high electrical voltage drop can achieve high energies if the voltage drop is large enough. The voltage drops required to explain the cosmic rays observed by the HEAO 3 experiment range upwards from millions of volts to many billions of volts.
The cosmic rays must have moved through a very powerful electrical field at one time, or else traversed smaller electrical fields many times. The first situation could be realized in the vicinity of the rapidly rotating, highly magnetized collapsed stellar cores, known as pulsars or neutron stars, that are sometimes left behind as remnants of supernova explosions. Alternatively, a charged particle could be trapped in the violent shock wave produced by a supernova explosion. As it bounces back and forth in the shock wave, the charged particle would traverse an electrical field many times and be accelerated up to very high energies. Radio, optical, and X-ray observations of the compact pulsars and extended supernova shells have shown that electrons can in fact be accelerated to very high energies in both situations. The protons and other heavy particles that make up most of the cosmic rays do not radiate as efficiently as electrons, so they cannot be observed directly. Nevertheless, it was assumed, with good reason, that supernova explosions were responsible for the cosmic rays.
Here again, experiment had another surprise in store for the theoreticians.
The relative abundances of the elements in the cosmic rays observed by the HEAO 3 experiment are similar to the relative abundances of the elements in solar system material, but they are not identical. The differences may contain important clues as to how cosmic rays originate.
It is observed that the elements that are more difficult to ionize are less abundant in the cosmic rays than in solar system material. For example, it takes about twice as much energy to tear an electron away from, or ionize, an oxygen atom as it does to ionize an iron atom, and the amount of oxygen relative to iron in cosmic rays is about half the relative amount found in the solar system.
 This general trend, which had been suggested on the basis of earlier experiments, was confirmed by HEAO 3 for a wide range of elements, from oxygen through zirconium. This behavior is difficult to understand if cosmic rays are produced by violent events such as supernova explosions. The atoms should be completely stripped of all their electrons under such conditions; the amount of energy required to tear off the outermost electron should be irrelevant.
On the other hand, if cosmic rays originate in a comparatively low energy environment having a temperature of about 10000 degrees, the....
 ....ionization energy may have an important selection effect. Under such conditions, some elements would be completely ionized and some would not. For example, relatively more iron atoms than oxygen atoms would have lost an electron. When an atom loses a negatively charged electron, the electrical neutrality of the atom is destroyed. It becomes a positively charged ion. This has important consequences for the acceleration of the particles. If an atom, which is electrically neutral, is placed in an electric field, it is not accelerated by the electric field. On the other hand, an ion, being charged, will feel an electrical force and will be accelerated.
This means that if a gas composed of a mixture of neutral atoms and ions is subjected to an electric field, only the ions will be accelerated. In particular, if the gas has a temperature of about 10000 degrees, then relatively more iron will be accelerated than oxygen, since relatively more iron atoms will be ionized. This is what is observed in the cosmic rays.
The implication is that cosmic rays do not originate in the high energy environs of exploding stars, but in comparatively more peaceful conditions. These conditions may be found on the edges of hot bubbles carved in interstellar space by hot young stars, around red dwarf stars known for their frequent, intense flares (flare stars), or in the wake of a shock wave which has been plowing through the interstellar medium for hundreds of thousands of years and lost much of its intensity.
Another idea, which may fit in with a supernova origin and the observed relative abundances, traces the origin of cosmic rays to dust grains in the interstellar medium. The observed abundances of these grains are consistent with the relative abundances of cosmic rays. Epstein proposes that the grains themselves are swept up and accelerated by the shock front of a supernova explosion. While interacting with the ambient gas in the shock wave, the grains' electrons are blown off. The ionized grains are accelerated to high speeds and break up through collisions with other grains or with high energy particles into a number of heavy ions. These ions then reenter the shock front and are accelerated to speeds near the speed of light.
These new ways of thinking about the origin of cosmic rays have been stimulated by the excellence of the high energy astronomy experiments. Once again, as it always must, the data, the cold hard facts, have tempered the theories and opened up new pathways for exploration. Bold ideas are needed, to be sure, but only by gathering data to check these ideas can we hope to arrive eventually at some reasonably accurate picture of the universe in which we live.