150 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1923 



These do not fit into the scheme at all, and they present such an 

 extraordinary combination of high temperature, small luminosity, 

 and considerable mass that it is very difficult to form any consistent 

 idea of the physical conditions which exist on their surfaces. There 

 are indeed more "worlds for theory to conquer — and some of them 

 look as if it would take hard fighting. 



But there are other ways in which our knowledge of the properties 

 of matter may be applied to the stars. A simple calculation shows 

 that the gravitational pressure at the center of the sun must be some- 

 thing like a hundred million tons per square inch. The pressures 

 in other dwarf stars are of the same order of magnitude. Those in 

 giant stars are smaller, but are usually measurable in thousands of 

 tons per square inch, even when the density can not be many times 

 greater than that of air. To withstand such a pressure, at this 

 density, the gas must have a temperature of many millions of 

 degrees. 



What can we say of the properties which matter would exhibit 

 at these temperatures? Twenty years ago the only answer w^ould 

 have been, " Very little " ; but now, with our knowledge of atomic 

 structure, we can say a good deal. The extreme violence of the 

 collisions between the atoms would knock off all the electrons of 

 the outer shells, and keep them off. The lighter atoms — perhaps as 

 far as sodium or even beyond — would lose all their electrons, and 

 be reduced to bare nuclei. The heavier ones would retain their 

 innermost one or two rings or shells of electrons, but lose the outer 

 ones, which contain a considerable majority of the whole number 

 of electrons originally present. We can be certain, however, that 

 the nuclei themselves would emerge quite unscathed from these col- 

 lisions, and that if an isolated nucleus, or the battered fragment of 

 a heavier atom, had a brief interval of relative quiet it would begin 

 to pick up electrons again from those which passed by slowly enough, 

 and to reconstitute the atomic structure. Could we remove a por- 

 tion of the matter in this strange state and let it cool, the familiar 

 atoms would thereupon rebuild themselves, bit by bit, and at the 

 end they would be the same as ever. 



The principal differences at the high temperature, from our 

 present standpoint, are, first, there would be a vast multitude of 

 free electrons flying about, as well as the far heavier atomic nuclei, 

 so that the average " molecular weight " — in determining which 

 every free-moving particle in the gas counts as much as any other — 

 would be much diminished* Secondly, the gas at this temperature 

 would emit a tremendous flood of radiation, most of it of such short 

 wave length that it would resemble X rays rather than ordinary 

 light. This radiation would not go very far before it was scattered 



