51 



ON A TABLE OF STANDARD WAVE-LENGTHS OF THE 



SPECTRAL LINES 1 







[Memoirs of the American Academy of Arts and Sciences, XII, 101-186, 1896] 

 PRESENTED MAY 10, 1898 



Investigations on Light and Heat, made and published wholly or in part with appro- 

 priation from the Rumford Fund 



Some years since, having made a machine for ruling gratings and dis- 

 covered the concave grating, which placed in my hands an excellent 

 process for photographing spectra, I applied myself to photograph the 

 solar spectrum. The property of the concave grating, mounted in the 

 method which I use, of producing a normal spectrum gave me the 

 means of adding a scale of wave-lengths, and so producing a. photo- 

 graphic map of the solar spectrum on a very large scale and of great 

 accuracy. I soon after constructed a very much better ruling engine, 

 which is kept at a uniform temperature in the vault of the new physical 

 laboratory of the Johns Hopkins University, with which I have made 

 very much better gratings. I therefore went over the whole process 

 once more, extending the map to include B, and making new negatives 

 of the whole spectrum very much better than the old. This set of ten 

 photographic plates is now familiar to most spectroscopists. 



In order to place the scale on the negatives, it was necessary to know 

 the wave-lengths of certain standard lines. Of course my first thought 

 was of Angstrom, whose measurements were the wonder of his time. 

 On trying to place my scale according to his figures, I found it impos- 

 sible to make them and my photographs agree ; and I finally was forced 

 to the conclusion that a new series of standards was needed before I 

 could go further. Here again the concave grating came to my rescue. 

 All the spectra are in focus at once, and relative measures can thus be 

 made at once by micrometric measures of the overlapping spectra. 

 Again, the spectrum is normal, and so a micrometer of very long range 

 could be used. To obtain the primary standards by means of overlap- 

 ping spectra, I have used gratings with from 3000 up to 20,000 lines to 



1 An abstract of this paper has recently appeared in 'Astronomy and Astro-Physics,' 

 and in the 'London Philosophical Magazine.' 



