Annu. Rev. Astron. Astrophys. 1999. 37: 445-486 Copyright © 1999 by Annual Reviews. All rights reserved

3. STELLAR EVOLUTION

3.1. Stars Move Off the Main Sequence Rather Than Up or Down

With the discovery of the main sequence turn-off of the globular clusters and the heretofore unseen connection of the main sequence to the subgiant and the giant branches, the path of stellar evolution had suddenly become clear. Stars move off the main sequence rather than either down or up it as they age. Although trivial now, it was a revelation in 1952.

By 1940 it was well known that the main sequence could be explained as the locus of stars with different fractions of hydrogen throughout a chemically homogeneous stellar interior. Hence, a fully convective star, mixing the products of hydrogen burning uniformly throughout its interior, will progressively increase the mean molecular weight of its gas owing to the progressive increase in the helium content, and will move up the main sequence (Gamow 1940, Figure 36; Oke & Schwarzschild 1952, Figure 3; Bok 1946, Stromgren 1952). Likewise, stars on the bright part of the main sequence will move down the sequence if they lose significant mass as they age.

Even as late as 1946, these two processes were believed to be the keys to an understanding of the HR diagram and the time-dependent evolution of stars (see Section 8 of Bok 1946). Stromgren (1952) also discussed the process. Soviet astronomers continued to discuss the process even into the 1960s (Ambartsumian 1952, Massevich 1954, 1959, Fessenkov 1952, Fessenkov & Idlis 1959). Cowling (1958) provides an excellent review.

However, with the discovery of the globular cluster main sequence, its turn off near M_V = +4, and the identification of the turn-off with the Schonberg-Chandrasekhar 10 percent mass limit for a hydrogen exhausted isothermal core (Schonberg & Chandrasekhar 1942, Sandage & Schwarzschild 1952), it had become clear that a mapping off the MS was the principal feature of the evolution. This mapping also gave a satisfactory explanation of Trumpler's (1925) system of classification for the variety of HR diagrams of open clusters, and his related discovery of the turn-up of the main sequences of open clusters near their MS termination points (Oke 1955, Sandage 1958a, b, 1988).