274 ANNUAL REPOET SMITHSONIAN INSTITUTION, 19 31 



lie lines are still relatively sharp. Since hydrogen shows the effect 

 of an electric field far more than any other element, it is believed 

 that such fields are present in these hot stars. Plate 1 (6') and (Z>) 

 shows the spectra of Sirius with a Cygni again for comparison. It 

 is easy to see that each member of the series of hydrogen lines in 

 Sirius is considerably wider than in a Cygni. Spectra on a larger 

 scale show little difference in the metallic lines. Electric fields in 

 a stellar atmosphere are probably caused chiefly by the negatively 

 charged electrons and the positively charged ions passing close to 

 the atoms and disturbing them. The number of charged particles 

 close enough to disturb an atom naturally increases with the pressure, 

 and so it begins to look as if the widths of spectral lines, particularly 

 of the hydrogen lines, may in the future serve as another indicator 

 of the pressure. 



In everything thus far said we have considered the atmosphere 

 of the star as a whole, since we can not see its disk; and the atoms 

 have shown themselves only by absorbing out light from the con- 

 tinuous spectrum of the white light which passes by them from the 

 inside of the star. In one case only can we study the atmosphere of 

 a star without interference from the light inside and that is our 

 own star, the sun, when the moon passes across it at the time of a 

 total eclipse. For several seconds the main body of the sun is then 

 entirely covered, while the atmosphere at its edge still shows. A 

 spectrogram made at such a time shows a spectrum of the atmos- 

 phere entirely by itself, and the atoms of the vaporized metals above 

 the sun give their own pure spectrum of bright lines. Plate I {E, 

 F, G) shows such a spectrum taken at the eclipse in northern Cali- 

 fornia on April 28, 1930. The light of the disappearing crescent of 

 the solar atmosphere is split up into more than a thousand separate 

 crescents of different colors. Each of these crescents shows by its 

 intensity the amount of the element to which it belongs in the solar 

 atmosphere. The bright bands of continuous spectrum were caused 

 by minute bits of the sun itself, which still showed through deep 

 valleys on the edge of the moon when the photograph was made. 

 The intervening mountains projected beyond the solar disk, and 

 there we obtained a spectrum of the upper atmosphere without any 

 of the lower strata. 



Another way to take a spectrum of the solar atmosphere at 

 an eclipse is to use a spectrograph with a slit and to move the plate 

 by means of a screw at an even rate of speed while the moon is 

 covering the sun's atmosphere. One of the best plates of this kind 

 was taken by Doctor Campbell, of the Lick Observatory, at the 

 eclipse of 1905. Tlie moon moves in its orbit about half a mile in 

 every second, which means that it covers up about 200 miles of solar 



