236 



CARNEGIE INSTITUTION OF WASHINGTON. 



of 3 km. per second. The accompanying observations, made by Mr. St. 

 John at Mount Wilson with the grating spectrograph, show no change 

 in the wave-lengths of the atmospheric lines from sunrise to sunset. 



Wave-lengths of atmospheric lines of oxygen. 



Since it has long been customary to employ the absorption lines of 

 our atmosphere as fixed standards of wave-length in measurements 

 upon solar and stellar spectra, it seemed desirable that the question 

 should be independently examined by Mr. Babcock in Pasadena, 

 using interference methods. Observations were accordingly carried 

 out during March and April, using for the most part the a group of 

 air-lines having its head at X6276. Instrumental conditions were 

 favorable to the detection of very minute displacements of the spectral 

 lines, but no evidence was found of diurnal variation of wave-length. 

 This is in agreement with Mr. St. John's work with the grating, and it 

 is hard to account for the discordance between these results and those 

 of M. Perot, whose data indicated daily variations amounting to about 

 0.030 A for lines of the B group. 



SOLAR WAVE-LENGTHS WITH THE INTERFEROMETER. 



Preliminary work on the application of the interferometer to certain 

 classes of solar observations was begun by Mr. Babcock during the 

 summer, using the Snow telescope in conjunction with the new vertical 

 spectrograph recently completed. The aim has been to ascertain the 

 best working conditions and the reliability of the results obtained 

 when wave-lengths are observed at the center of the sun. The pho- 

 tographs are also being used for further study of the wave-lengths of 

 atmospheric absorption lines. Up to the present, the auxiliary disper- 

 sion has been produced by a 7-inch grating having an extremely bright 

 first-order spectrum, used at the 18-foot focus of the auto-collimating 

 spectrograph. Plates of excellent quality have been obtained, on 

 which the definition is satisfactory over 100 a. The few measurements 

 made thus far indicate that, for good solar lines of medium intensity, 

 the relative wave-lengths can be depended on to about 0.001 a, i.e., 

 the accuracy is about the same as in the grating method. Further 

 tests are to be made with the aid of higher auxiliary dispersion 



