480 Prof. H. A. Rowland's Table of 



the coincidence with the lower wave-lengths, a method which 

 gives them nearly eqnal weight with those of the visible 

 spectrmn. 



The full set of observations will be published hereafter; but 

 the present series of standards can be relied on for relative 

 wave-lengths to "02 division of Angstrom in most cases, though 

 it is possible some of them may be out more than this amount, 

 especially in the extreme red. 



As to the absolute wave-length, no further change will be 

 necessary, provided spectroscopists can agree to use that of 

 my table, as has been done by many of them. 



By the method of coincidences with the concave grating 

 the wave-lengths have .been interwoven with each other 

 throughout the whole table, so that no single figure could be 

 changed without affecting many others in entirely different 

 portions of the spectrum. The principal difference from the 

 preliminary table is in the reduction to the new absolute wave- 

 length, by which the wave-lengths are about 1 in 80,000 

 larger than the preliminary table. I hope this difference will 

 not be felt by those who have used the old table, because 

 measurements to less than -^^ division of Angstrom are rare, 

 the position of the lines of many metals being unknown 



to '^ ^'^''^le^divi.ai'^^-.^- Angstrom. As the new map of the 

 ol ty 'tt ^/J ir<ji. -oT o-T^. --oiv'iTl ^'- -':^,, ^j.' j-t ' *• + Vvl T 



spectrum has been made according to this"r?e"vV i^-.^.L©,/ see no 

 further reason for changing the table in the future. 



No attempt has been made to reduce the figures to a vacuum, 

 as the index of refraction of air is imperfectly known ; but 

 this should be done where numerical relations of time-period 



are desired. . , ,i • l a a 



In the column giving the weight, the primary standards 

 are marked S and the other numbers give the number ot 

 separate determinations of the wave-length, and thus, to some 

 extent, the weight. 



Many of these standards are double lines, and some of them 

 have faint components near them, which makes the accuracy 

 of setting less. Tliis is especially the case when this com- 

 ponent is an atmospheric line whose intensity changes with 

 the altitude of the sun. The principal doubles are marked 

 with d; but the examination has not been completed yet, 

 especially at the red end of the spectrum. 



