PROGKESS IN ASTROPHYSICS ABBOT. 177 



of these lines in the hiboratory. AVhen about the year 1895 Rowland 

 completed his investigation of the spectrum of the sun and of the 

 chemical elements, it was thought that the last word had been said 

 upon this, and that no greater accuracy of positions of the spectrum 

 lines was necessaiy, or indeed possible, than he had obtained. But 

 in recent years it has been found necessaiy to go over the whole 

 ground again, and to determine the positions of the lines of the 

 chemical elements and the lines in the spectrum of the sun with a 

 still greater accuracy than that of Rowland. This work has been 

 taken up under the auspices of the International Solar Union, and is 

 now approaching a satisfactory completion. 



In the year 1893 a remarkable piece of work was carried out by 

 Prof. Michelson (now of the University of Chicago) in the measure- 

 ment of the wave length of light in terms of the standard meter of 

 the International Bureau of Weights and Measures at Paris. Several 

 of the spectrum lines were investigated, and among them the red 

 line of cadmium, whose wave length as determined by Michelson is 

 6438.4722 Angstrom units.^ In pursuance of the investigations re- 

 cently recommended by the International Solar Union, Messrs. Fabry 

 and Perot remeasured the wave length of the cadmium line and found 

 the value 6438.4696, which, it will be seen, differs by less than 3 parts 

 in 6,000,000 from that obtained by Michelson. On this value of 

 Fabry and Perot will rest the system of Avave lengths adopted by the 

 International Solar Union. 



It had been determined at the meeting of the Union on Mount 

 Wilson in 1910 that only wave lengths which are independently de- 

 termined with satisfactor}'' agreement by three observers with the 

 most approved apparatus should be accepted as secondaiy wave- 

 length standards. In pursuance of this action of the Solar Union, 

 Messrs. Fabry and Buisson in France, Pf und at Baltimore, Eversheim 

 and Burns in Germany, have been determining with the highest 

 possible accuracy the wave lengths of certain lines in the spectra of 

 iron and nickel, selected at nearly equal intervals of wave length. 

 About 85 such lines have now been measured with satisfactory agree- 

 ment in three or more independent investigations, and have been 

 adopted by the International Solar Union as secondary standards 

 of wave length. These lines cover the spectrum from a wave length 

 3370.789 Angstrom units, which is far beyond the visible limit of the 

 spectrum, in the violet, to wave length 6750.163 Angstrom units, 

 which is near the limit of the visible red. It is expected that further 

 investigations will carry the lists of secondary standards as far as 

 wave length 2,000 in the ultra-violet, and perhaps as far as wave 

 length 10,000 in the infrared. The astonishing accuracy of the re- 

 sults obtained may be inferred when it is said that the three inde- 



1 The Angstrom unit is one ten-billionth of a meter. 

 44863°— SM 1913 12 



