The study of the motions and physical characteristics of celestial bodies could be undertaken on a case-by-case basis. Cosmology, on the other hand, by definition had to mold from the whole body of observations a coherent theory of the structure of the universe. From the time of Aristotle until the Copernican revolution this structure had been conceived firmly as Earth-centered, in spite of Aristarchus' heliocentric views. Between the inner Earth and the outer sphere of fixed stars each planet was considered embedded in a crystalline sphere that was in uniform circular motion. Within this neat compact structure the whole history of humankind and the universe unfolded. The work of Kepler and Newton on planetary motion and of Galileo and others on the physical nature of the planets resulted in the gradual acceptance during the 17th century of the Copernican heliocentric hypothesis as a physically valid system. This acceptance had profound implications for cosmology. Not only did Kepler's theory of ellipses make obsolete the theory of crystalline spheres of Aristotle, the absence of any detectable stellar parallax - apparent change in the direction of a star - even when measured from opposite sides of the Earth's orbit, demonstrated the necessity of acknowledging the enormous size of the universe. The subsequent shift from the closed, tightly structured world to an infinite homogenous universe was one of the landmarks in the history of astronomy. Giordano Bruno and Thomas Digges were among the new concept's earliest exponents, and René Descartes and Newton incorporated it as a standard part of the new view of the universe.

The Galaxy. The downfall of the concept of the sphere of fixed stars opened the way for investigations into the distribution of the stars. The phenomenon of the Milky Way, shown by Galileo to consist of myriads of stars, hinted that some previously unsuspected structure might exist among the stars. In 1750 the English theologian and astronomer Thomas Wright sought to explain the brilliantly luminous band of the Milky Way as a collection of stars that extended further in the direction of the band than in other directions. In 1780, William Herschel initiated an observational program of star counts, or gauges, of selected regions of the sky. By assuming that the brightness of a star is a measure of its distance from the Earth, this program resulted in a picture of a flattened disk-shaped system with the Sun near the center. In spite of the fact that his assumptions were incorrect, Herschel's research marked the beginnings of an understanding of the structure of the stellar system; his research also earned him the title of founder of stellar astronomy.

Because of the lack of a direct method for determining stellar distances, progress in cosmology lagged for more than a century after Herschel's star gauges. Only with improved instrumentation did Friedrich Bessel, Wilhelm Struve, and Thomas Henderson (1798-1844) in 1838 succeed in measuring the first stellar distances. An annual parallactic shift of .31 in in the position of the star 61 Cygni implied a distance equivalent to 590,000 times that of the Earth from the Sun. Difficulties innate to the method, however, resulted in fewer than 300 known stellar distances by the end of the century - far too few to solve the problem of determining the structure of the stellar system.

Along with continuous attempts at parallax measurements, a major task of 19th-century astronomers was the compilation of astronomical catalogs and atlases containing the precise magnitudes, positions, and motions of stars. The work of Friedrich Argelander, David Gill, and J. C. Kapteyn is especially notable in this regard, with the latter two astronomers utilizing new photographic techniques. Such surveys yielded the two-dimensional distribution of stars over the celestial sphere. Heroic efforts were made, especially by Hugo von Seeliger, to extrapolate from this data to an understanding of three-dimensional structure.

Real progress was finally made only through an analysis of the extremely small motions of stars, known as proper motion. Building on the astronomical catalogs of James Bradley, G. F. Arthur von Auwers (1838-1915), and Lewis Boss (1846-1912), and on his own observations, Kapteyn, exploiting the new field of statistical astronomy, applied statistical methods of distance determination to find an ellipsoidal shape for the system of stars. In 1904 he found that the stars streamed in two directions. Only in 1927 did J. H. Oort, working on the basis of Bertil Lindblad's studies, determine that Kapteyn's observational data could be accounted for if the galaxy were assumed to be rotating. In a cosmological shift comparable to the Copernican revolution four centuries before, the Earth's solar system was found to lie not at the center of the Galaxy but, rather, many thousands of light-years from the Galaxy's center.

Other Galaxies. The question of whether or not the Galaxy constituted the entirety of the universe came to a head in the 1920s with the debate between H. D. Curtis and Harlow Shapley. Curtis argued that nebulae are island universes similar to but separate from our own galaxy; Shapley included the nebulae in our galaxy. The controversy was settled when E. P. Hubble detected Cepheid stars in the Andromeda nebula and used them in a new method of distance determination, demonstrating that Andromeda and many other nebulae are far outside the Milky Way. Thus the universe was found to consist of a large number of galaxies, spread like islands through infinite space.

Such was the progress of astronomy from the time of the Babylonian observations of planetary motions to within a few degrees' accuracy, to the Greek determinations of positions within a few minutes of arc, to the 19th-century measurements of parallax and proper motions in fractions of a second of arc. The concern of astronomers evolved from the determination of apparent motions to the observation of planetary surfaces and ultimately to the measurement of the motions of the stars and galaxies.