MOUNT WILSON SOLAR OBSERVATORY. 183 



Standards of Wave-length. 



As a practical test of the degree of precision obtainable by the plane-grat- 

 ing spectrograph, the absolute wave-lengths of a score or more of iron lines 

 belonging to the same pressure groups as the international standards of the 

 second order were determined with similar instruments of 30 feet focal 

 length on Mount Wilson and in Pasadena by Mr. St. John, assisted by Miss 

 Ware. The mean of the differences, Pasadena-Mount Wilson, taken with- 

 out regard to sign, is 0.0010 Angstrom ; the mean probable error for the Pasa- 

 dena series is ± 0.0007 A., and for the Mount Wilson series, ± 0.0006 A. 

 These results are of interest in view of the possible use of the plane grating 

 in absolute measurements and in a study of the comparative precision obtain- 

 able by plane and concave gratings in such measurements. 



In the case of lines not belonging to the same group as the standards, when 

 classified according to displacements under pressure, the measurements on 

 the Mount Wilson and Pasadena plates showed consistent differences in 

 wave-length that seemed only referable to the difference in atmospheric pres- 

 sure between the two stations. The elevation of the Mount Wilson station 

 is 5,886 feet (1,794 m.) ; the elevation of the Pasadena laboratory is about 

 800 feet (244 m.), so that the difference in altitude corresponds to slightly 

 less than one-fifth of an atmosphere. When lines of the same widths on the 

 two series of plates were compared it was found, ( 1 ) that in the case of 20 

 lines the wave-lengths determined from the Mount Wilson plates were on 

 the average 0.009 Angstrom shorter than those from the Pasadena plates ; 

 these lines belong to group d of Gale and Adams, which consists of lines 

 showing enormous displacements to the red under pressure, and, therefore, 

 displacements to the violet under the decreased pressure on Mount Wilson ; 

 (2) that in the case of 15 lines the wave-lengths from the Mount Wilson 

 plates were on the average 0.014 A. longer than those from the Pasadena 

 plates. Upon the plates taken by Gale and Adams under a pressure of 9 at- 

 mospheres, the maxima of these lines are greatly displaced to the violet, the 

 lines, however, being too diffuse for measurement. There is, therefore, suf- 

 ficient reason for considering them a distinct group, the members of which 

 are displaced to the violet under pressure, and hence to the red under the 

 decreased pressure on Mount Wilson. 



The investigation offered the means of examining the relative accuracy of 

 the international standards of the second order, which had been questioned 

 at the Mount Wilson meeting of the International Union for Cooperation in 

 Solar Research. In the regions A 5371 to A 5658 and A 5975 to A 6494 there 

 are 21 standards of the second order, and of these 16 were reported as re- 

 quiring adjustments of 0.001 to 0.011 A. Such results, if confirmed, would 

 tend to destroy confidence in the precision of the interferometer method of 

 obtaining standards of wave-length. With standards 50 A. apart on the 

 present photographs an error in one of them of 0.002 A. is clearly shown. 



