50 

 A NEW TABLE OF STAND AKD WAVE-LENGTHS 



[Johns Hopkins University Circulars, No. 106, p. 110, 1893; Philosophical Magazine [5], 

 XXXVI, 49-75, 1893; Astronomy and Astro- Physics, XII, 321-347, 1893] 



PREFATORY NOTE 



During the last ten years I have made many observations of wave- 

 lengths, and have published a preliminary and a final table of the wave- 

 lengths of several hundred lines in the solar spectrum. 



For the purpose of a new table I have worked over all my old observa- 

 tions, besides many thousand new ones, principally made on photo- 

 graphs, and have added measurements of metallic lines so as to make 

 the number of standards nearly one thousand. 



Nearly all the new measurements have been made on a new measur- 

 ing machine whose screw was specially made by my process 1 to cor- 

 respond with the plates and to measure wave-lengths direct with only 

 a small correction. 



The new measures were made by Mr. L. E. Jewell, who has now be- 

 come so expert as to have the probable error of one setting about T ir?nF 

 division of Angstrom, or 1 part in 5,000,000 of the wave-length. Many 

 of these observations, however, being made with different measuring 

 instruments, and before such experience had been obtained, have a 

 greater probable error. This is especially true of those measurements 

 made with eye observations on the spectrum direct. The reductions of 

 the reading were made by myself. 



Many gratings of 6 in. diameter and 21^ feet radius were used; and 

 the observations were extended over about ten years. 



The standard wave-length was obtained as follows: Dr. Bell's value 

 of D 1 was first slightly corrected and became 5896-20. C. S. Peirce's 

 valfce of the same line was corrected as the result of some measurements 

 made on his grating and became 5896-20. The values of the wave-length 

 then become 



See Encyc. Brit., art. Screw. 

 35 



