Cosmologists attempt to understand the origin and structure of the universe as a whole. They begin their search with an assumption about the nature of the universe – namely, that in looking out from our vantage point in the cosmos, we see essentially the same kind of universe that an observer stationed in any other part of it, no matter how remote, would see. As far as our telescopes can reach, we see galaxies and clusters of galaxies distributed in more or less the same way in every direction. This assumption that the universe is uniform on a large scale is called "the cosmological principle." One thing that is certain is that the universe is expanding. In every direction we look, distant galaxies are moving away from each other. Until the 1960s, the expansion of the universe was the primary fact of cosmological significance that cosmological theories had to accommodate. There were two general classes of cosmological theories that fit with the expanding universe: the evolutionary (Big Bang) theory and the steady-state theory. The essential idea of the evolutionary cosmology is that there was a beginning – a moment of creation at which the universe came into existence in a hot, violent explosion – the Big Bang. In the beginning, the universe was very hot, very dense, and very tiny. As the explosion evolved, the temperature dropped, the distribution of matter and energy thinned, and the universe expanded. From the current observed rate of expansion, we conclude that the creation event occurred between ten and twenty billion years ago. The steady-state theory is based on an idea called the "perfect cosmological principle." It is "perfect" in that it maintains that the universe is uniform not only in space but in time. Thus it is the hypothesis that the large-scale universe has always been the way it is now and will be this way forever in the future. This view is consistent with philosophical approaches that reject the notion of an absolute beginning of the universe as unacceptable. The steady-state universe would have no beginning and no end. In an expanding universe, the galaxies move away from each other, spreading matter more thinly over space. On the other hand, the perfect cosmological principle requires that the density of matter in the universe remain constant over time. To make the steady-state theory compatible with the expanding universe, its proponents introduced the notion of continuous creation. As the universe expands and the galaxies move farther apart, new matter – in the form of hydrogen – is introduced into the universe. The rate at which the hypothesized new matter is created is far too small for this creation to be detected with available instruments, but continuous creation provides just enough matter to form new stars and galaxies that fill in the space left by the old ones. Thus in the steady-state universe there is evolution of stars and galaxies, but the general character and the overall density of the universe remains unchanged over time. In this special sense, the steady-state universe itself does not evolve. Both of these views – steady-state and Bing Bang – allow for cosmic expansion. However, the discovery in the 1960s of a comparatively small star-like objects called quasars tipped the scales in favor of the Big Bang cosmology. Astronomers determined that almost all quasars are very distant. Given how bright quasars appear even at such great distances, astronomers concluded that quasars typically have an output of light that is 1,000 times greater than that of a whole spiral galaxy composed of billions of stars. Quasars are such distant objects that the light now reaching us from quasars left them billions of years ago. This means that when we observe quasars today we are seeing that state of the universe billions of years ago. Thus the fact that almost all quasars are very far away implies that earlier in the history of the universe quasars were developing more frequently than they are now. This evolution is consistent with the Big Bang theory. But it violates the perfect cosmological principle, and so it is inconsistent with the steady-state view.