There are essentially two arrangements for particle-beam experiments in elementary-particle physics. Before the early 1970s, fixed-target experiments were the only means of probing the structures of subatomic particles. In a fixed-target experiment a beam of accelerated particles is directed at a stationary target. An example of such an experimental arrangement is the Bevatron at Berkeley, built to produce antiprotons by striking a proton at rest with a high-energy proton: p + p → p + p + p + ̄p. According to the mass–energy equation (see mass) some of the kinetic energy of the accelerated proton becomes the mass energy of the proton-antiproton pair. Since this newly formed mass has only the mass equivalence of two proton rest masses, it is remarkable that in practice this fixed target experiment requires an incident kinetic energy equivalent to six times the proton rest mass to create a single proton–antiproton pair. This example illustrates the inefficiency of scattering using a stationary target; conservation of momentum (see conservation law) means that the products of a fixed target experiment must possess some residual kinetic energy which therefore cannot be involved in forming the mass-energy of collision products. This problem does not arise in colliding-beam experiments introduced in the early 1970s. In these, the beams of particles are caused to collide with their antiparticle counterparts moving in the opposite direction at the same speed. The total momentum of the colliding particles is therefore equal to zero, and their total energy is available to form the mass energy of the collision products. Colliding-beam experiments are known to have lower collision rates than their fixed target counterparts. However, the likelihood of collision is increased by using finely focused beams and by allowing particles to accumulate in storage rings before they are made to collide. New machines in high-energy physics are mostly based on a colliding-beam arrangement.