The Evolution of the Stars

F. Hoyle and R. A. Lyttleton

Editor’s Note

Astronomers Fred Hoyle and Raymond Lyttleton review how recent advances in nuclear theory were fundamentally changing astronomers’ understanding of stellar evolution. It was now possible, as George Gamow had recently argued, to probe the nuclear chemistry likely to be important at the densities and temperatures prevailing in stars. But Hoyle and Lyttleton think Gamow went rather too far in suggesting most problems of stellar evolution were now solved. In Gamow’s view, for example, all red giant stars must be of very recent origin, which seemed most unlikely. Hoyle and Lyttleton were correct: physicists then understood only the rudiments of nuclear chemistry, and knew nothing of nuclear fusion, a key process driving all stellar activity and evolution.　　中文

PROF. G. Gamow has recently discussed in Nature1 the consequences of recent developments in nuclear theory on the problem of stellar evolution. In the light of the exact knowledge that is now available of a large number of nuclear reactions, it is possible to decide which processes rise to importance at the densities and temperatures prevailing in the stars, and on this basis trustworthy estimates have been given for the rate of liberation of subatomic energy. These results, which are the outcome of laboratory investigations, furnish the mathematical theory of internal constitution of the stars with a new equation that enables the luminosity of a star to be calculated by direct methods. The information so obtained has been utilized to attempt to resolve the many paradoxes and discrepancies encountered in discussing the general problem of stellar evolution. All this recent work has been authoritatively summarized by Gamow with great clarity and understanding in the article referred to above. It is therefore with some surprise that we find that Prof. Gamow concludes his article with the impression that these new developments practically solve the problems of stellar evolution. This seems to us to be so far from being the case that some further discussion of the claims of nuclear theory as the main factor in stellar evolution would be desirable.　　中文

In the first place, it should be noticed that the application of nuclear theory in its present state inevitably leads to a result at least as embarrassing as any of the questions that it might possibly resolve. For the theory maintains that no synthesis of atomic nuclei from hydrogen is possible within the stars except for the very light elements. This would imply that the stars can no longer be regarded as the building place of the heavy elements, which must have formed before they became part of the star—if indeed they ever were formed. Now although such a conclusion does not itself constitute a logical contradiction, it seems to us to present such overwhelming difficulty that it is much more reasonable to conclude that the basis of nuclear theory is in need of revision rather than that the heavy elements were not formed by synthesis. On the other hand, many investigators seem to have accepted the former result as satisfactory, and in particular Gamow has proceeded to make it the basis of a theory of the red giant stars.　　中文

Secondly, Prof. Gamow apparently regards the conclusion that various classes of stars should be of totally different ages as a natural one. Thus the fact that the present theory leads to a life-time for certain massive stars of order 10－3 the life-time of the Sun is not regarded as a difficulty at all; indeed it is merely supposed that this is the case and the theory remains unquestioned. In point of fact, it is an essential part of the theory as proposed by Gamow that all red giant stars are considered as of very recent formation, since the presence of lithium, etc., is required to enable them to radiate with their supposed low internal temperatures.　　中文

Now if all the stars could be regarded as single autonomous bodies, it would be difficult to dispute the validity of these views by direct means. But it so happens that the frequent occurrence of binary systems enables a simple test of the theory to be made, for in the case of binary stars both components must be of comparable age. It is immediately clear that the well-known difficulty concerning the relative emission per unit mass of the components of doublestars must remain in any theory that appeals only to the internal properties of the stars, although by making very artificial assumptions in Gamow’s theory some of these discrepancies might be avoided. For example, it would require as a general result that the less massive components of binary systems form with a hydrogen content differing systematically from that of their companions. In certain cases this would lead to even more dubious initial conditions: thus, whilst Sirius must have formed with high hydrogen content, it would have to be assumed (according to the more generally accepted theory of degenerate matter) that the companion formed almost solely from heavy elements for it to have been practically exhausted during the whole of its existence. Such a solution of this difficulty could scarcely be regarded as satisfactory even if there were no other objection to the theory described by Gamow.　　中文

It should be particularly noticed that the foregoing suggestion assumes that the components have not evolved by fission, for this latter process (even if it were dynamically satisfactory) would clearly lead to two stars of closely similar compositions. We have been able to show, however, that to produce close binary systems, periods of order 5×1010 years are necessary2. Thus the existence of spectroscopic binary systems in which one component is a red giant or any highly luminous star presents an immediate contradiction of the theory given by Gamow. Moreover, there seems to be a general tendency for the mass of binary systems to increase with decreasing separation.　　中文

While on the subject of fission, it is perhaps only proper to point out here that although the mathematical investigations of the development of rotating fluid masses clearly demonstrate that binary stars cannot be generated by fission3, many astronomers do not yet seem to have realized the physical significance of the mathematical results. As a consequence of this there are still many who “believe” in the fission theory. But as will be seen, even an appeal to fission could not save Gamow’s theory of the red giant stars, in addition to which it would raise afresh the difficulty of the relative emissions of the components in binary stars. Even if some process of break-up of a single star led to a binary system, it is evident that the components must have similar chemical compositions, while in close binary systems consisting of a red giant star and a class B star, Gamow’s theory would require the red star to be the less massive component on account of the mass luminosity relation. But observation shows that in such pairs the giant star tends to be the more massive component. This is the case, for example, in the three stars υ Sagittarius, V V Cephei and ζ Aurigae. Thus it seems that no matter from what angle we approach the questions raised by the observed properties of binary systems, the paradoxes already recognized by astronomers must remain in one form or another.　　中文

However, quite apart from the foregoing objections to the constructive portion of Gamow’s article, it is very noticeable that no reference is made to the wide class of dynamical features that is associated with the stars. This, of course, is the direct result of attending only to the internal physical properties of the stars; but the dynamical features we have in mind are altogether too marked to remain unaccounted for in a satisfactory theory. Thus such questions as the formation of individual stars, and of binary and multiple systems, together with the general increase of mass with decreasing separation, and the observed approximation to equipartition of energy among the stars seem to present the real key to any theory of stellar evolution. An internal theory can give no explanation for the correlation between peculiar velocity and spectral class or the observed tendency for massive stars to lie in the galactic plane, features that must be related to the previous history of the stars.　　中文

It has been customary during recent years for investigators on stellar evolution to devote attention to internal constitution with little or no regard for the dynamical features. It appears that Prof. Gamow has followed essentially in this tradition and therefore confined his article to the modifications effected by the introduction of modern nuclear theory. Thus, in dealing with the properties of variable stars, no attempt is made to account for the three distinct periodicity groups comprised by stars of periods of order half a day, four days and 300 days. These variables also show a marked preference as regards spectral class, the first being largely of classes B and A, the second of F and G and the third of class M. Moreover, the two short-period groups exhibit a most remarkable property in that none of them, out of more than two hundred available examples, possesses a close companion, whereas about one star in five of normal stars of similar spectral classes does possess a close companion. On the other hand, long-period variables appear to possess a normal complement of companions. The first and third types are stars of moderate luminosity and show no pronounced galactic concentration, whereas the variables of intermediate period, the Cepheids, are strongly concentrated to the galactic plane and are among the most luminous known stars. Thus it is clear that very remarkable dynamical properties are intimately connected with even the different types of variability, and therefore that purely internal considerations are most unlikely to prove capable of elucidating the nature of the connexion.　　中文

From these and many other dynamical qualities associated with various types of stars, it appears to us that Prof. Gamow has over-estimated the importance of nuclear theory in the problem of stellar evolution. Indeed, in our opinion nuclear physics has very little to add to the results already conjectured by astrophysicists, and can merely serve to confirm these conjectures, a typical instance being the mass-luminosity relation itself. Finally, we wish to point out that although the present article consists largely of criticism, we have discussed elsewhere a number of the questions raised2, and it has been found that purely dynamical considerations may be sufficient to provide a natural explanation of many of the difficulties mentioned in this article.　　中文

(144, 1019-1020; 1939)

F. Hoyle and R. A. Lyttleton: St. John’s College, Cambridge.

References:

1. Nature, 144, 575, 620 (Sept. 30 and Oct. 7, 1939).

2. Proc. Camb. Phil. Soc., (4), 35 (1939).

3. Mon. Not. Roy. Astr. Soc., 98, 646 (1938).