50O 



NATURE 



[June 30, 192 i 



parable with that of the earth. Our sun and our 

 nearest stellar neighbour, o Centauri, are marked 

 as typical dwarfs of type G, and Sirius is a repre- 

 sentative A-type star. 



From the known luminosity and surface tem- 

 perature of any star it is easy to calculate its 

 surface and so its density. Giants of types G 

 and K are found to have densities of the order of 

 0004 and 00005 respectively, agreeing with the 

 known densities of binary stars of these types. 

 Sirius, with a luminosity of forty-eight times, and 

 a surface temperature about one and a half times, 

 those of our sun, must have a surface nine times 

 as great. Its mass is 34 times the solar mass, 

 so that its density must be about 02. In general 

 it is found that all giant stars must be gaseous, of 



Luminosity 



(Son ' I) 



'4pOO 



100 



01 



'c epoo'c. 3,000'c. 



Fig. 2. — Luminosity-temperature diagram. 



density so low that the ordinary gas-laws will be 

 approximately obeyed. Dwarf stars may be gase- 

 ous or liquid or solid, but, if gaseous, they 

 are so dense that the gas-laws will be no- 

 where near the truth. It is now easy to see 

 why, in the giant stars, increase of temperature 

 and density go together; this is merely a conse- 

 quence of Lane's law. But the dwarfs may be 



thought of as approximating rather to masses of 

 fixed dimensions, and for these the luminosity falls 

 off as the temperature decreases. 



Our sun radiates hght at a rate of about 2 ergs 

 per second per gram of its mass. Gravitational 

 contraction, as Lord Kelvin showed, could provide 

 energy at this rate for only about 20,000,000 

 years, and radio-active and chemical energy could 

 only slightly lengthen this period. For a giant 

 star, radiating at 1000 times the rate of the sun, 

 the maximum period would be only a few thou- 

 sand years. This period is far too short, and it 

 is now generally accepted that, so far from gravi- 

 tation and known sources of energy providing 

 the whole of a star's radiation, they can provide 

 only an insignificant fraction. Energy of adequate 

 amount can originate only from sub-atomic 

 sources, as, for instance, from internal rearrange- 

 ments in the positive nuclei of the atoms Or from 

 the transformation of a small fraction of the star's 

 mass into energy. It is a matter of simple calcu- 

 lation to show that all other stores of energy in 

 a star can constitute only an insignificant reservoir 

 of energy which, unless continually replenished 

 from sub-atomic sources, would be exhausted in, 

 astronomically, a moment. Thus the rates of 

 radiation and of generation of sub-atomic energy 

 must be practically equal, and the luminosity of 

 a star will be determined by the latter rate at any 

 instant. 



We may now think of the evolution of the 

 stars as represented by the march of a vast army 

 through our diagram (Fig. 2), the individuals 

 keeping, for the most part, within the marked 

 belt. Each individual takes his marching orders 

 from the supply of sub-atomic energy, and so 

 long as we remain in ignorance of the exact 

 source and nature of this we cannot be certain 

 whether the motion of the army is up or down, 

 or even that it is all in the same direction. But 

 if we are right in conjecturing that the stars were 

 born out of a nebula of very low density, the 

 order of march will be from low density to high ; 

 our army will be marching downwards in the 

 diagram. Its tail, except for a few stragglers, is 

 about at absolute magnitude —4, its head is lost 

 in darkness. In the next lecture we must study 

 the incidents which may occur during the march 

 of this army of stars. 



{To he continued.) 



Obituary. 



Dr. a. M. Kellas. 

 T)Y the death of Dr. A. M. Kellas we have 

 -L' lost one of the best authorities on the 

 effect of high altitudes on the human system. No 

 one else had so great a practical knowledge, or 

 worked scientifically at the subject with more 

 persistence than he. 



Born in Aberdeen, he was educated there, and 

 afterwards went to Edinburgh, London, and 

 Heidelberg. For some time he was assistant to 

 NO. 2696, VOL. 107] 



Sir William Ramsay, and afterwards lecturer on 

 chemistry at Middlesex Hospital. 



As a teacher he was most successful, taking 

 endless trouble in helping backward students. In 

 pure chemistry he did little research, his chief 

 contribution being a long and careful investigation 

 on "The Determination of the Molecular Com- 

 plexity of Liquid Sulphur," published in 1918. But 

 during the last ten years he gave up most of his 

 spare time to study the physiological and physical 



