August. 26, 1920] 



NATURE 



^11 



these investigations, as in Fowler's experiments 

 on helium and oxygen, and Merton's further work 

 on carbon, will yield results of high value in the 

 interpretation of the spectra of stars at the highest 

 stages of temperature, and possibly also of the 

 nebulae. 



Another of the chief subjects which attracted 

 Lockyer during a great part of his life was the 

 classification of stellar spectra, and the order of 

 celestial evolution which might be inferred. He 

 was at first mainlv dependent upon stellar observa- 

 tions made by others, but he soon saw the neces- 

 sity for first-hand data, and, following Pickering's 

 remarkable success with the objective prism, he 

 adopted this form of instrument in most of his 

 work at Kensington, and afterwards at Sid- 

 mouth. He early adopted the suggestion made 

 by Tait that in nebulas and comets the luminosity 

 may be due to solids heated by impact, as well 

 as to heated gas generated by the impacts, and 

 about 1887 developed it into his "meteoritic 

 hypothesis." The fundamental idea is that all self- 

 luminous celestial bodies are composed either of 

 swarms of meteorites, or of masses of meteoritic 

 vapour produced by heat, the heat being developed 

 by condensation due to gravity, and the vapour 

 being finally condensed into a solid globe. The 

 classification of stellar spectra which he based 

 upon this theory has undergone modifications in 

 detail, chiefly in the direction of subdivision and 

 more complete definition -of the criteria for the 

 various stellar groups ; but the essential idea has 

 remained unchanged throughout. In common with 

 other astronomers, Lockyer adopted the view that 

 the spectroscopic differences between the various 

 classes of stars are mainly due to differences of 

 temperature, but, unlike most of them, he in- 

 sisted that in place of a single line of evolu- 

 tion from hot (white) to cool (red) stars the pro- 

 gression must be from cool to hot stars and back 

 again to cool stars. That is, in accordance with 

 the theory of condensing swarms of meteorites or 

 masses of gas, the classification made a distinc- 

 tion between stars of increasing temperature and 

 those which are on the down-grade towards the 

 extinction of luminosity. 



Some of the earlier evidence for the separation 

 of the stars on the two branches of the " tempera; 

 ture curve " which Lockyer pictured may be of 

 doubtful validity, but the valuable photographic 

 data accumulated later, in combination with 

 laboratory researches, placed his classification on 

 a much firmer basis. It was found, for example, 

 that when stars at any given stage of tempera- 

 ture were brought together by reference to the 

 relative intensities of enhanced and arc lines, they 

 were definitely divisible into two groups, showing 

 that the spectra were dependent in part upon 

 physical conditions other than those imposed by 

 temperature alone. This difference was attributed 

 to differences in the state of condensation, one 

 group being less condensed than the other, and 

 therefore to be considered as being in an earlier 

 stage of evolution, notwithstanding equality of 

 temperature. The Harvard classification, w'hich 

 XO. 2652, VOL. 105] 



has been adopted by most astronomers, is along 

 one line of temperature only, and accordingly dis- 

 regards this difference. It is clearly of great 

 importance, however, that the difference should 

 be tal^en into account in questions relating to 

 stellar distribution and other matters connected 

 with the structure of the sidereal universe, and it 

 was a source of profound regret to Lockyer that 

 greater attention was not given to it. In the fcase 

 of the helium stars, however, Lockyer's classifica- 

 tion has received substantial corroboration from 

 a discussion by Herassimovitch of their radial 

 velocities and absolute magnitudes, in which the 

 catalogues of Lockyer were utilised. Among the 

 results it was shown that the stars which Lockyer 

 had located on the ascending branch of the tem- 

 perature curve were brighter than those on the 

 descending branch, and, assuming the average 

 masses to be equal, it would follow that the 

 former were of greater volume and lower density 

 than the latter, in accordance with Lockyer's 

 hypothesis. 



The theory of stellar evolution put forward a 

 few years ago by Prof. H. N. Russell resembles 

 that of Lockyer in its main outlines, though based 

 mainly on deductions as to the densities and abso- 

 lute magnitudes of the stars. The criteria are 

 thus somewhat different in the two cases, but 

 there can be little doubt that in one form or other 

 the recognition of an ascending, as well as of a 

 descending, line of stellar temperatures will take 

 an important place in the astronomy cf the 

 future. 



The observation of total eclipses of the sun also, 

 occupied much of Lockyer's attention. He per- 

 sonally took part in nine eclipse expeditions, and 

 was responsible for several others in which the 

 observations were undertaken by his assistants. 

 On several occasions, when H.M. ships were de- 

 tailed to assist the expeditions, his exceptional 

 organising ability enabled him effectively to utilise 

 the services of officers and men so as to cover the 

 widest possible range of observations. The out- 

 standing feature of his work in this connection, 

 however, was the introduction and use, first of a 

 visual spectroscope without slit or collimator, and 

 afterwards, when photographic methods could be 

 adopted, of the prismatic camera. With instru- 

 ments of this type he was able clearly to differ- 

 entiate between the coronal and chromospheric 

 radiations, and, besides detecting several new 

 coronal lines, he obtained splendid records of the 

 "flash " spectrum. He was thus able to determine 

 the various heights to which the different vapours 

 extended, and he identified a multitude of the 

 bright lines with enhanced lines which he had so 

 diligently investigated in the laboratory. 



Lockyer would have been the last to claim that 

 his work was wholly free from errors, but it was 

 almost invariably of a stimulating character, and 

 has played a leading part in the development of 

 the science of astrophysics practically from its very 

 beginning. Much of his work will have an endur- 

 ing place in the history of the science to which he 

 devoted his great gifts. 



